Previous day | Jump to hour: 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 | Next day

Seconds: Show Hide | Joins: Show Hide | View raw
Font: Serif Sans-Serif Monospace | Size: Small Medium Large

Click in the nick column to highlight everything a person has said.
The Logo icon identifies that the person is a core developer (has commit access).

Notice: Only Gecko based browsers prior to FF4 support the multipart/mixed "server push" method used by this log reader to auto-update. Since you do not appear to use such a browser, this page will simply show the current log, and not automatically update.

#rockbox log for 2007-11-11

00:01:11DerPapstcd /mnt
00:01:16DerPapstmkdir sansa
00:01:22linuxstb_Then type "mount /dev/sda /mnt"
00:01:28DerPapstmount /dev/sda /mnt/sansa
00:01:37 Quit moos (Read error: 110 (Connection timed out))
00:01:44DerPapstor what linuxstb_ said
00:02:45amiconnprzemhb: If you bother to compare - this is the h1x0 driver taken as is, with just the port fiddling macros (and DELAY) changed
00:02:53TPM_Debianafter moutn, it will show device data
00:03:09TPM_Debiani wil try copy os-files
00:03:12linuxstb_TPM_Debian: After copying the files, type "umount /mnt" before disconnecting the usb cable.
00:03:22TPM_DebianTHX SO MUCH!!!!!!!!!!
00:03:52DerPapstbetter eject /dev/sda no?
00:04:01Bagderumount is fine
00:05:40przemhbamiconn: I'll take a look...
00:06:40linuxstb_DerPapst: Do you know if the IPL devs are aware of the big speed improvements recently made to the Rockbox driver for the 5g's LCD?
00:06:54DerPapstyes.. they are
00:06:55TPM_Debianlinuxstb, must i rename the downloadet os-file "SKU_E-PPS022.mi4" to "PPS022.mi4" befor i can upload it?
00:07:15BagderPP5022.mi4, yes
00:07:19 Quit newbyx86 ()
00:07:29DerPapsti told it and GodEater did so too.
00:07:33Bagderthat's a five, not an S
00:08:20amiconnDerPapst: I wonder if they care.... like about the power consumption and clock setup fixes on PP5002...
00:08:30TPM_Debianhmm . my sys does not have a "unmount" command
00:08:57amiconnTPM_Debian: Where did you read the 'n'? It's umount
00:09:32Bagdernot unmount
00:09:34TPM_Debiana .. k
00:10:13DerPapstamiconn: not too much i'd say.. unfortunately
00:11:15TPM_Debianhmmm .... it caaaannnnt be ... the display will not going on ... :[[[[
00:11:29linuxstb_Bagder: Shouldn't TPM_Debian also copy a bootloader file?
00:11:37Bagderindeed he should
00:12:09Bagderan mi4 only will hardly help
00:12:20TPM_Debianrecovery mode will also show black display .... strange strange strange
00:12:52TPM_Debiani cant explain me that
00:14:27TPM_Debiancan i reset the device ... like a "LOW-LEVEL-FORMAT" ? ... perhaps the flash is hard corrupt ..
00:14:43amiconnprzemhb: Hmm, the patch looks still hackish to me. GPIO B2 isn't tuner power, it's the audio mux. The method should work though, but the placement should be changed before commit
00:14:54TPM_Debianor i has the device killed
00:15:11linuxstb_TPM_Debian: You need to put it in recovery mode again (with e200tool), and then copy the bootloader file to /mnt
00:15:15przemhbamiconn: I do agree
00:16:03amiconnHmm, and we need a radio context button map
00:16:28 Quit Llorean (Read error: 104 (Connection reset by peer))
00:16:29amiconnI think the FF/REW buttons are ideal candidates for tuning
00:16:56 Join newbyx86 [0] (
00:17:40 Join saratoga [0] (i=9803c6dd@gateway/web/cgi-irc/
00:17:57TPM_Debianohh man .... now e200tool cant found the device .... ..... ....
00:19:09markunTPM_Debian: maybe you should give it a rest and try again tomorrow
00:19:14TPM_Debianoh . ok .. i forgot the hold-key
00:21:32TPM_Debiancan i copy "BL_SD_boardSupportSD.rom" and "PP5022.mi4" on the device, or first the bootloader and then (after reboot) the other file ?
00:23:18Bagderyes you can
00:23:55 Quit spiorf (Remote closed the connection)
00:25:02TPM_Debianafter copy files turn power off or try it via e200tool ?
00:25:16Bagderumount, then pull the usb cable
00:25:27TPM_Debianand then?
00:25:48Bagderthen I believe it reboots and it should boot fine
00:25:53Bagderinto the OF
00:25:59TPM_Debianno ... display is on, and nothing will do
00:26:13TPM_Debianlast message is "LUN0 unlocked"
00:26:59 Quit midkay ("Leaving")
00:28:29 Join spiorf [0] (
00:28:55TPM_Debiani think if i turn the power off then the device will not update the files ... but .. what i can do ?!
00:29:23Bagderswitch it off and on again
00:30:10TPM_Debianok .... result: JOG-LEDs on , Display black ... hmmmm .......
00:30:55TPM_Debianso i think slowly the device it dead
00:31:34BagderI don't
00:32:25 Join BigBambi [0] (n=Alex@rockbox/staff/BigBambi)
00:32:57 Quit spiorf (Remote closed the connection)
00:34:08TPM_Debianwell ... i got no more idea's what i can do ....
00:34:42 Quit davina (Remote closed the connection)
00:36:50 Join iamben_ [0] (
00:36:55Buschelinspired by saratoga's tests i played around with the DEV_EN1/EN3 register on ipod video
00:37:24Buschelwith EN1 = 0x424192e and EN3 = 0x00000037 the unit works fine for >8 hours now
00:37:27theclawis there any documentation for the rockbox source? (i'd like to know where the "entry point" is, that is, which function the bootloader calls)
00:37:52Buschelall other bits must be set to have full functionality
00:38:13Buscheldoesn't save a single mA :/
00:38:34linuxstb_theclaw: The entry point is in crt0.S (in firmware/). This is the low-level startup-code, and it then calls main() in apps/main.c
00:41:21TPM_DebianYEAH ... it works now! ... Bagder after disconnect usb cable its need to reconnect to it ... then the device will start the writing data folloging by reboot and run the os
00:41:34TPM_Debianoh man ... THANK Y SO VERY VERY MUCH!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
00:41:48TPM_Debianand all other people!!!!!!!!!!!
00:42:19amiconnprzemhb: I have a usable keymap now
00:43:50mrkikoDoes someone know any open-source software for phone accessibility?
00:45:03Buscheland: DEV_EN1 = 0x00400008 (both bits) must be set for USB-detect
00:45:08krazykitmrkiko, that has nothing to do with rockbox. please stay on topic.
00:45:17 Quit theclaw ("n8")
00:46:09amiconnprzemhb: Patch uploaded
00:46:38saratogaBuschel: should probably put that in the wiki then
00:46:46amiconnNow the remaining question is why there is no sound........
00:46:57 Quit Robin0800 (" HydraIRC -> <- The alternative IRC client")
00:47:28Nico_Pamiconn: would #if (LCD_DEPTH > 1) be better than #ifdef HAVE_LCD_BITMAP for album art, or do you think a special define or another check are needed?
00:47:30przemhbamiconn: thanky you for the patch
00:48:25amiconnNico_P: Of course, but we should probably have a separate macro
00:49:07Nico_Pamiconn: maybe two actually. HAVE_ALBUMART and HAVE_REMOTE_ALBUMART
00:49:19*amiconn only read the first half of the question, but thinks that his answer might still be understandable
00:50:20Nico_Por maybe not the second... and should I define the first in the config*.h files?
00:52:07saratogaamiconn: did you realize that the DEV_EN3 register is already defined as CACHE_PRIORITY?
00:52:27 Quit iamben (Connection timed out)
00:52:44amiconnThat's why I added the comment that it needs verification
00:53:01amiconnBut from a routine in the G5 rom it looks like it's a DEV_EN
00:53:18amiconnI posted it on pastebin somewhere - should be in the logs
00:54:47amiconnBagder: What happened to the Recorder build in the last round? :>
00:55:08 Nick fxb__ is now known as fxb (
00:56:08Buschelsaratoga: updated wiki with the "must haves"
00:56:37 Join iamben [0] (
00:57:13 Quit iamben_ (Read error: 104 (Connection reset by peer))
00:57:54 Join Mouser_X [0] (i=cf9bb003@gateway/web/cgi-irc/
00:58:09 Quit Mouser_X (Client Quit)
00:58:19 Quit ender` (" "Religion is the opiate of the masses." -- Karl Marx "Winners don't do drugs." -- The FBI")
00:58:45saratogaBuschel: interesting that the video needed bit 5 set in DEV_EN3
00:58:49saratogathe sansa did not
00:59:29Buschelit shows the same effect as bit 2
01:00:46Nico_Pamiconn: should I enable AA on targets with 2bpp screens?
01:01:17 Join Mouser_X [0] (i=cf9bb003@gateway/web/cgi-irc/
01:01:19 Part toffe82
01:01:33saratogaBuschel: where did you set the bits?
01:01:53saratogai noticed that doing it during startup caused some crashes that could be avoided by setting the same bits later
01:02:06DerPapstNico_P: if it uses dithering maybe.. else i would say better not :)
01:03:00Buschelsaratoga: not at the best place i guess... i set them in the ide_power_enable()-routine. this is first called after ~10s.
01:03:32saratogathat seems late enough
01:05:35***Saving seen data "./dancer.seen"
01:07:14 Join xxl3w [0] (
01:07:23xxl3wdoes rockbox work with windows?
01:07:58Mouser_XYou can run the simulators.
01:08:31 Quit stewball`ghost (Read error: 113 (No route to host))
01:08:39Mouser_X(But that's not necessarily going to give the best impression of what Rockbox is.)
01:11:35krazykitxxl3w, you can install rockbox while using windows, and you can put music on your device while using windows, so yes.
01:13:08 Part przemhb
01:14:04 Join TPM_Debian_ [0] (
01:20:44 Quit n17ikh|Lappy ()
01:22:13 Nick fxb is now known as fxb__ (
01:24:41 Join n17ikh|Lappy [0] (
01:29:23 Join davina [0] (
01:30:38 Quit TPM_Debian (Read error: 110 (Connection timed out))
01:32:49 Quit xxl3w ()
01:40:14TPM_Debian_gn8 @all .. and thx :]
01:41:20 Quit TPM_Debian_ ("")
01:45:49 Join iamben_ [0] (
01:46:48 Quit bertrik ("bye")
01:47:19 Quit BigBambi ("Please insert girder")
01:49:48 Join JdGordon [0] (
01:51:22 Join nomel [0] (i=42ed06d2@gateway/web/cgi-irc/
01:53:41nomelhow do i find who worked on the lv24020lp tuner driver?
01:53:46 Join criznach [0] (
01:53:51nomelis it Ivan Zupan, or jethead71+
01:55:08n1snomel: doesn't the svn history say?
01:57:15nomelyes :) just found that command.
01:58:32 Quit iamben (Read error: 110 (Connection timed out))
02:02:45 Join BigBambi [0] (n=Alex@rockbox/staff/BigBambi)
02:05:06 Quit iamben_ (Read error: 104 (Connection reset by peer))
02:07:18 Quit davina (Remote closed the connection)
02:08:43 Join iamben [0] (
02:09:27 Join johny1234 [0] (
02:11:49 Quit RaRe (Read error: 113 (No route to host))
02:11:57 Quit BigBambi ("Please insert girder")
02:12:39 Join RaRe [0] (
02:13:09Nico_Pwow AA causes a big binsize increase
02:13:23JdGordonwouldnt you expect that?
02:13:25n1sno surprise really
02:13:40Nico_P4k on gigabeat
02:13:41*n1s guesses on 10k
02:13:58Nico_Phehe then I might not get bashed too hard :p
02:14:03JdGordonNico_P: are you copying the pic from the ram into a specific buffer for the wps? or displaying it straight from the MoB buffer?
02:14:15Nico_PJdGordon: straight from the main buffer
02:14:45JdGordoncuesheet and lyrics is next then ? :)
02:14:47Nico_PI think it's just about ready to be committed
02:14:54Nico_PJdGordon: yeah, cue is next
02:15:09Nico_Plyrics I don't think I'll do
02:15:19n1swith scaling, jpeg and from-tag support I guess 10k is more like it though
02:15:28Nico_Pn1s: indeed
02:15:41JdGordonis scaling going to be added?
02:16:19Nico_PJdGordon: it's only planned... there are some issues I don't know how I will solve yet
02:16:56Nico_Pand the patch on the tracker has a bad scaling algorithm and will be completely outdated
02:16:58JdGordonoh, dont the good AA themes need scroling margins patch to work properly?
02:17:20Nico_PJdGordon: on most targets with a landscape display, yes, unfortunately
02:17:34Nico_Psansa and gigabeat are good though :)
02:17:47JdGordonthe other 23 though :p
02:18:08johny1234a question
02:18:14JdGordonan answer!
02:18:33Nico_PJdGordon: let's hope it will motivate amiconn to implement viewports
02:18:34*n1s was going to type that! :P
02:18:37johny1234first, my english is not wll
02:18:57johny1234is the cowon x5 a good player
02:18:58Nico_PI think unofficial builds won't disappear until he does
02:19:38n1sNico_P: there will always be some hackish patch tat users want to have
02:19:40mrkikojohny1234: read guideliness; don't use words such as "wll": not all of us can understand them.
02:19:56JdGordonor we should think about adding that patch in way that it can be removed easily if/when viewports happens
02:20:31n1sjohny1234: it depends on what you want to do with it etc...
02:21:06JdGordonNico_P: its not likely to motivate amiconn to do it.. but more likely to get someone else working on it
02:21:46n1sJdGordon: did you get anywhere with the customizable statusbar?
02:21:55johny1234n1s hear musik
02:21:56JdGordonnot yet
02:22:18Nico_PJdGordon: someone lile you or me? :p
02:22:20n1sjohny1234: I am certain it can play music :)
02:22:55JdGordonNico_P: you me _jhMikeS_, anyone... its nto a small task
02:22:58n1sand do it well I was more asking about other stuff as recording/radio/video
02:23:39Nico_PJdGordon: indeed it's not... I'd probably be afraid of starting it :)
02:23:59 Join BigBambi [0] (n=alex@rockbox/staff/BigBambi)
02:24:09n1sJdGordon: have you decided which route to take?
02:24:42JdGordonIts going to be similar to the list icons.. but im not sure the best way to integrate it there to save on duped code
02:24:48JdGordonand to not waste ram
02:25:30n1sJdGordon: will the size be customizable?
02:25:46 Join J3TC- [0] (
02:25:49 Join johny1235 [0] (
02:25:51n1sah, that's really all I want, great :)
02:25:55JdGordonno point doing it if it wasnt :p
02:26:11n1sfor teh shineh!
02:26:33johny1235but with which quali
02:26:38 Join daurnimator [0] (n=daurnima@unaffiliated/daurnimator)
02:28:04 Quit Buschel ()
02:28:27*Nico_P is off to bed... will probably commit tomorrow
02:29:45Nico_Pbye :)
02:29:48 Quit Nico_P (Remote closed the connection)
02:30:02n1sjohny1235: I have never listened to one but I think the audio quality is good
02:34:55 Quit johny1235 (Read error: 104 (Connection reset by peer))
02:36:33 Join iamben_ [0] (
02:41:10 Quit johny1234 (Read error: 101 (Network is unreachable))
02:44:13 Quit linuxstb_ ("ChatZilla [Firefox]")
02:48:51 Quit iamben (Read error: 110 (Connection timed out))
02:49:41 Quit DerPapst ("So Long And Thanks For All The Fish!")
02:51:05 Quit mrkiko ("leaving")
02:54:22rasherJdGordon: the text editor is ... tricky, if the file contains blank lines (you cant see which one you've selected)
02:54:51JdGordonyeah, any ideas how to fix that though?
02:55:35rasherI don't know the code, so no.
02:56:16rasherI was under te impression that it was specifically hidden though? Can't you just not hide it?
02:56:56JdGordonwhat was specifically hidden?
02:57:32rasherThe selector bar
02:57:58 Quit Thundercloud (Read error: 104 (Connection reset by peer))
02:58:27JdGordonno, the problem is with the text drawing code.. it doesnt invert if there is no text being displayed
03:01:56GX9900WZHow come when I skip to a few tracks too fast, I hear a squeeky sound through the headphone, those sound you hear when the harddrive is seeking data.
03:02:14 Quit n1s ()
03:05:14JdGordonstupid bloody touchscreen!
03:05:38***Saving seen data "./dancer.seen"
03:06:32JdGordonanyone good at maths here?
03:06:52krazykiti'm good at arithmetic.
03:07:25 Join jhulst [0] (n=jhulst@unaffiliated/jhulst)
03:11:27 Join TMM [0] (
03:11:48 Join kingwen [0] (n=kingwen1@
03:30:48kkurbjunJdgordon, what are you trying to do?
03:31:06JdGordontrying to calibrate it for landscape
03:31:22JdGordonI thought the math would stay working once the values were put in
03:31:30JdGordonbut its rotated 90 deg which is a bit wierd
03:32:07kkurbjunoh nice, yeah, the landscape mode is a bit strange, where are the calculations done?
03:32:48JdGordon is my patch for that file
03:32:56JdGordonwith the new values
03:33:28 Quit nomel ("CGI:IRC (EOF)")
03:34:16JdGordonbtw, SCREEN_ROTATE probably inst the best define... SCREEN_PORTAIT or _LANDSCAPE would be better
03:35:34kkurbjunSCREEN_PORTRAIT would be an easy enoguh chang
03:36:16kkurbjunI was just thinking that too actually because it's a matter of perspective
03:37:48 Join Dan[eq] [0] (i=dced10a9@gateway/web/cgi-irc/
03:41:25 Join barrywardell [0] (
03:43:46*JdGordon back in 10
03:45:09 Join iamben [0] (
03:48:19 Quit PaulJam (".")
03:53:34JdGordonkkurbjun: see anything funny in the maths which i missed?
03:54:09kkurbjunyeah, I'm tring a compile out now
03:54:29kkurbjunbut I want to see if I can figure out a way to make it so there doens't have to be an ifdef for it
03:54:31JdGordonok cool
03:54:48JdGordonan ifdef for which part?
03:54:55kkurbjunthe calculation
03:55:15JdGordonwhats wrong with an ifdef? we have no plans for making it runtime witchab anyway
03:55:42kkurbjunyeah, I suppose, I just like to avoid them if possible
03:56:26JdGordonx1 and x2 need to be switched?
03:56:41kkurbjunx2 and val
03:57:06kkurbjunand x1 needs to be x2 on the x calc
03:57:21JdGordonah ok
03:57:27 Quit iamben_ (Read error: 110 (Connection timed out))
03:57:54kkurbjun:), looks like I'm wrong too though
04:00:27 Quit Dan[eq] ("CGI:IRC (EOF)")
04:00:37 Join male [0] (
04:01:07maleHas anyone done any work on a browse-by-album-art plugin?
04:02:04JdGordonnever heard of such a plugin.. so no
04:02:08maleI'm not talking about a full on 3D cover-flow like thing. Just fullscreen album art browsing an selection plays the album.
04:02:09JdGordonkkurbjun: :) oh well
04:03:20maleI might try adapting the bmp viewer then.
04:04:06JdGordonkkurbjun: do you have a wirless mouse/kb?
04:05:16 Quit iamben (Read error: 104 (Connection reset by peer))
04:05:20kkurbjunJdGordon: no, why?
04:05:43JdGordonkkurbjun: my mrobe seems to make my keyboard act very wierdly when its on!
04:06:07kkurbjunhuh, that's odd
04:06:59JdGordonanyway, swapping x1 and x2 everywhere doesnt work :p
04:07:03*JdGordon gets some lunch
04:10:57 Join iamben [0] (
04:13:03 Join tedrock [0] (
04:17:01 Join BHSPitLappy [0] (n=Me@unaffiliated/bhspitmonkey)
04:27:15 Part kingwen
04:31:37kkurbjunJdGordon: Just committed the fix for it, I have to go
04:34:49 Quit barrywardell ()
04:38:27 Quit iamben (Read error: 104 (Connection reset by peer))
04:43:52 Join iamben [0] (
04:48:14 Join psycho_maniac [0] (
04:59:42 Join TTThomas [0] (
04:59:45 Join aliask [0] (
04:59:47 Join stevenm [0] (n=stevenm@
05:05:05 Join alex[slx] [0] (n=Alex|slx@unaffiliated/alex-slx)
05:05:42***Saving seen data "./dancer.seen"
05:06:30 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
05:06:48 Join miepchen^schlaf [0] (
05:15:34 Quit bb (Nick collision from services.)
05:15:41 Join bb_ [0] (n=bb@unaffiliated/bb)
05:15:46 Quit criznach ("ChatZilla [Firefox]")
05:18:40 Part stevenm ("Connection reset by beer")
05:24:34 Quit jhulst ("Konversation terminated!")
05:28:59 Join Calcipher [0] (
05:32:21 Join ball [0] (n=ball@
05:32:58 Quit TTThomas ("Lost terminal")
05:33:21ballIs Rockbox only available for pocket MP3 players? Is there a stand-alone version or an X client?
05:33:41aliaskball: There's a simulator which runs under SDL
05:34:09JdGordonwhats a good define for CONFIG_ for SCREEN_PORTAIT or SCREEN_LANDSCAPE?
05:34:41 Join TTThomas [0] (
05:34:48JdGordonsomething shorter :p
05:34:50ballaliask: thanks
05:35:24CalcipherJdGordon IS THAT WHAT I THINK IT IS?
05:35:33aliaskCalcipher: Probably not :)
05:35:38Calcipherlandscape mode? really really?
05:35:43JdGordonits not
05:35:52*Calcipher dies
05:35:53JdGordonthe mrobe will have a compile time oiption though
05:36:23 Join kingwen [0] (n=kingwen1@, you know what I was gonna ask, I'm still advocating for a landscape mode on the Sansa e200s
05:37:26psycho_maniacmay i ask why? just curious
05:37:27JdGordonall it needs is someone to fix the lcd driver to do it.. and hen the bitmaps...
05:37:37JdGordonits not that hard really cor a compile time option...
05:39:36ballDoes Rockbox work with an iPod mini, with Firewire?
05:39:49Calcipherpsycho_maniac I was asking for that feature, because as a user who needs large fonts, I see using the longer length of the LCD screen as better use of space for scrolling text, so you can read more things on the screen when you have larger fonts, it makes sense, and doesn't cause any issues on long lists since they would scroll down normally anyway
05:39:58aliaskball: Yes.
05:41:02psycho_maniacwould be a cool feature to show off i think.
05:41:10ballWould I need iTunes to install Rockbox?
05:42:12 Nick HellDragon is now known as ifehf (i=jd@unaffiliated/helldragon)
05:42:17 Nick ifehf is now known as HellDragon (
05:42:42psycho_maniacno. read the install instructions. ball
05:43:06aliaskball: Here's a handy link too!
05:43:16TTThomasso helpful!
05:43:32aliaskI try
05:46:56 Nick HellDragon is now known as rrrrrr (i=jd@unaffiliated/helldragon)
05:46:58 Nick rrrrrr is now known as fwefe (
05:47:27 Nick fwefe is now known as HellDragon (
05:50:49JdGordonCalcipher: although... there is a bit more to rotating on the sansa because the buttons need to be "remapped"...
05:50:56JdGordondo you know c?
05:58:31psycho_maniacis this going to get implemented? just curious because i know gigabeat owners might want this feature.
05:59:11*Mouser_X is fine with his 240x320 screen.
05:59:54JdGordonI doubt both orientations would be offered as dailys.. but I dont see why a patch to add the option wouldnt be accepted
06:00:00psycho_maniacwell if its landscaped on the gigabeat then its the same resolution as the ipod video correct?
06:00:10 Quit J3TC- (Read error: 110 (Connection timed out))
06:00:20Mouser_Xpsycho_maniac: Yes.
06:05:43Calciphersweet, I'm sorry JdGordon I was doing some accounting for my pops, and no I do not know C
06:06:03Calciphersorry, I do want to learn a language though
06:07:07 Join Llorean [0] (
06:07:43psycho_maniacwhat is the button you press to show the bootloader info on start up?
06:08:08JdGordonwhich target?
06:09:00JdGordonany button prob?
06:09:24psycho_maniacis it? i just tried right and that worked. what would this be under the manual?
06:09:26JdGordonright button
06:09:54JdGordonits not really useful for everyday users.. so no need for it in the manuak
06:10:43ballI'd be happy if I could just find a Pascal compiler that worked.
06:10:46psycho_maniacwould it be in the wiki?
06:10:53Calcipheranyone around who has generated voices in cygwin?
06:11:07CalcipherI ran into some problems doing that
06:11:14JdGordonpsycho_maniac: could be... its defiantly only the right button thoygh
06:12:11JdGordondamn, the mrobe lcd is soo much slower in landscape mode
06:12:27JdGordonno wonder the OF's screen is so boring!
06:13:04LloreanJdGordon: re "landscape mode and compile time options", it'd be much more useful as a runtime option. There's already a compile time option patch for the gigabeat and as far as the discussion I've heard, it's not particularly wanted since really the official build will just be one way or another.
06:14:03JdGordonyeah, well having it runtime is probably not going to happen, at least a compile time option helps those that want to use it
06:14:15LloreanWhy not just pick one?
06:14:26JdGordonwhy is one better than another?
06:14:31 Quit homielowe (Connection reset by peer)
06:15:05LloreanWell, since I doubt we'll have the buildservers producing two builds for every target in the future, one kinda needs to be picked if it's not a runtime option...
06:15:20JdGordonI wouldnt complain if there were 2 gigabeat builds... after all.. its not extra work adding wps or bmps for the rotated build
06:15:34JdGordonnot for all targets... just selected ones
06:15:44LloreanThat's somewhat unfair to all the other targets...
06:15:51JdGordonbecause not all fell right rotated
06:16:06LloreanWhy do you get to decide which do and which don't?
06:16:16JdGordonbecause im special :p
06:16:21JdGordonbut yeah, i guess your right
06:16:29psycho_maniacwould be cool if it was an option. landscape: yes/no
06:16:33JdGordonbut having the option in the code doesnt add anything bad to it...
06:16:44CalcipherDoes my argument sound reasonable? as to why I would like this implemented
06:16:50LloreanIt just seems like a waste of time to work on a feature that's already been called "not wanted" for another target.
06:17:06JdGordonso why not accept it?
06:17:15JdGordonisnt the gb one for the sim?
06:17:21saratogaif someone wants to #define in orientation settings, power to them
06:17:28LloreanCalcipher: Please, we're not discussing whether Portrait/Landscape option is bad. We're discussion whether he should do it the _wrong_ way and waste time on something that'll later need to be implemented the right way if it's going to be included in official binaries.
06:17:49*JdGordon is farily confident runtime rotating will never happen
06:17:55JdGordondefinatly not before viewports anywa
06:17:57LloreanJdGordon: The GB one is for target.
06:18:10LloreanPeople use it with a custom bootloader that allows booting from either .rockbox or .rockbox_1
06:18:21LloreanAnd I _really_ don't like features that encourage people to use custom / outdated bootloaders
06:18:40LloreanBut I'm not the one who said compile-time is a bad idea
06:18:54LloreanJust relaying what I observed in a previous discussion on the matter that I actually wasn't involved in.
06:18:55JdGordonwhos against the option?
06:19:20LloreanDon't remember who all was.
06:19:29LloreanI just remember it mostly being a "If it's going to be done, it really should just be done the right way."
06:19:46CalcipherI'm not accusing anywone of saying its bad, I was just curious if you think my reasoning is seinsible
06:19:47psycho_maniacshouldnt everything be done the right way first?
06:20:01JdGordonnow thats an odd idea :p
06:20:19LloreanCalcipher: Why does the reasoning matter if pretty much everyone agrees a runtime option is worth having if someone could do it.
06:20:28Lloreanpsycho_maniac: Well, that's why Album Art was kept out so long... wrong way.
06:20:38 Join homielowe_ [0] (
06:20:43 Nick homielowe_ is now known as homielowe (
06:20:49JdGordonAA is going in tomorow probably !
06:20:57LloreanI know!
06:20:59saratogai can't wait
06:21:04Calcipheroh ok, I didn't mean the method, thats all up to you all, as the devs
06:21:05saratogai'm so sick of hearing about it
06:21:20saratogai've never been so excited about a feature I will never use
06:21:27psycho_maniacgreat now i can download every theme for my player on the wiki ]
06:21:33JdGordonsaratoga: haha ditto
06:22:09LloreanJdGordon: Honestly, if you're planning on implementing compile time rotation for *all* targets, *I* won't complain too much.
06:22:20psycho_maniacwhat is album arts fs#?
06:22:27saratoganow if we could just figure out the remaining PP registers, and get Zagor to commit USB support, hell will be about frozen over
06:22:37Mouser_Xpsycho_maniac: 3035?
06:22:44JdGordonviewports before hell can freeze!
06:22:56saratogawhat do viewports even do
06:23:01Calcipherwhat is viewports?
06:23:04Mouser_Xpsycho_maniac: (It was mentioned in the logs fairly recently.)
06:23:16JdGordonLlorean: I'm not planning on adding it.. but I dont see why we cant commit the patches as they come in...
06:23:35ballThanks for you help chaps.
06:23:40 Part ball ("bye!")
06:23:56JdGordonviewport will in theroy split up the display into sub "windows" each being able to scroll differently and stuff...
06:24:12JdGordonand more than that
06:24:12LloreanJdGordon: Because since an official "rotated" build probably won't be available anyway, if they have to compile a custom build to use it either way, they might as well patch and keep the code from being more complicated?
06:25:14JdGordonnup.. I dont tinhk thats a good argument... the code wont change in that section anyway.. so commitnig the patch makes the custom builders lives easier.. and doesnt effect anyone else
06:25:46LloreanJdGordon: So why do we make custom builders' lives easier and our support guys' lives harder?
06:25:58LloreanWhy exactly do the people who aren't willing to help contribute directly the ones we should be helping out?
06:27:47psycho_maniacAA is 3045
06:28:35Mouser_Xpsycho_maniac: I was pretty close...
06:31:04psycho_maniacwhat is stopping "scrolling margins" from being comitted?
06:31:29Lloreanpsycho_maniac: It does the same thing that viewports will do.
06:31:31JdGordonits another "not the right way of doing it" patch
06:31:37LloreanOr rather, what it does is a small portion of what viewports will do
06:33:27psycho_maniacwith vireports will you really be able to have the menu in the wps like the example on the wiki page?
06:34:16LloreanDepends on how they're implemented, and more specifically, whether someone codes up a patch to do that afterward
06:35:48Mouser_XWith viewports, you could have a spectrum anylizer in the WPS. Well, at least as I recall, you could have plugin that runs in a viewport on the WPS, and that plugin could be an anylizer... I don't know how capable a DAP would have to be to actually do that though.
06:36:23JdGordonits all ideas still....
06:36:27LloreanYes, a TSR plugin drawing to a viewport is one aspect brought up today.
06:36:33JdGordonnothing is coded yet so we dont know what will be possiblw
06:36:57LloreanJdGordon: True, but amiconn suggested a plugin drawing to a viewport, so if he does end up being the one to implement it, it'll probably end up possible.
06:39:00*JdGordon thiks vp will be such a massive change noone will actually be bothered starting it :p
06:41:04 Part kingwen
06:41:41psycho_maniaci dont really like the menu in wps idea really.
06:42:56Calcipherhaha, sounds crazy, I sure would never care for the analyzer, or any visualizations probably, all my player does is hang out in my pocket, and play me some music
06:43:16 Part alex[slx]
06:44:01Mouser_XI'd like visulizations, but only if you have the option to disable them. They'd be disabled 95% or more of the time for me, but occasionaly I'd like to have them up.
06:44:03psycho_maniaci liked the wps that had a "fake analyzer" on it.
06:44:19Mouser_XWhich was that?
06:45:35psycho_maniaci couldnt tell you the name. its for the ipod video and the artist lables his themes that always starts with "phk" or something like that.
06:46:12Mouser_XAh. I did actually check that WPS gallery once. I might look through it again.
06:46:16 Join linuxstb_ [0] (n=linuxstb@rockbox/developer/linuxstb)
06:46:52psycho_maniacits not on rb-t i dont think .
06:47:44Mouser_XI've only looked at the iPod video WPS gallery in the RB wiki.
06:49:23maleHas anyone ever done anything like have a different WPS for when the backlight is off?
06:49:47 Join hannesd_ [0] (
06:49:58maleI find that a color-intesive WPS on the Nano is unreadable with the backlight off.
06:50:08psycho_maniacit seems to me the wiki is slower then rb-t
06:50:59maleIs there a conditional the backlight status?
06:51:05males/the/for the/
06:51:07psycho_maniacmale: isnt that why you turn the backlight on
06:51:32saratogaa lot of players work pretty well with the backlight off
06:51:35malepsycho_maniac: Seems a waste to turn the backlight on at track change just for a glanec.
06:51:41maleHow about a strobe backlight?
06:51:49maleFlashes once a second or so.
06:51:58maleThat might be sufficient.
06:52:06*JdGordon can hardly see the e200 display when the backlight is on and there is any sun around :'(
06:52:12saratogatheres no way that would be comitted
06:52:16maleWell, the nano is very dark with the BL off.
06:52:33Mouser_XAs is the Gigabeat.
06:52:38psycho_maniaci heard the gigabeats screens go off when the backlight is off. correct?
06:52:39saratogaJdGordon: I think theres a brightness control on the sansa, might be possible to make it brighter
06:52:46LloreanJdGordon: Pretty much the Nano and iPod 5G are the only ones with really visible color displays with BL off
06:52:58Mouser_Xpsycho_maniac: It used to, but they took that out.
06:52:58LloreanUnfortunately, bright sunlight is very similar to "backlight off"
06:53:18saratogaregarding the gigabeat, the LCD and controller should probably be powered down when the light is off to save power
06:53:29Mouser_XApparenlty, there wasn't any savings having it turn off, and it slowed it down a *lot.*
06:53:40maleI think a patch for this strobe idea should be pretty easy to implement.
06:53:49JdGordonit doesnt help that my glasses tint when in the sun so it makes it even harder than usual :p
06:53:56Mouser_Xsaratoga: It used to be. toffe test it thought, and said that it wasn't worth turning off.
06:54:10psycho_maniacactually there was a patch that would turn the lcd to the beat of the music.
06:54:26saratogaMouser_X: I don't understand the things the gigabeat people say
06:54:44saratogaapparently the LCD, and cpu frequency scaling doesn't save power
06:54:50saratogawhich makes no sense to me
06:54:54JdGordonijuts :p
06:55:08Mouser_XToffe tested power usage with the LCD controller on, and off, and the differences in power usage were minimul to none.
06:55:09Lloreansaratoga: CPU scaling doesn't as far as we can tell.
06:55:20Mouser_XAt least, that was what I was told.
06:55:20saratogaerr, I mean apparently the frequency scaling doesn't, and now the LCD
06:55:27LloreanLCDs (not backlights) really do drain little enough power in most cases.
06:55:40LloreanThe LCD wasn't "None" just "trivial amounts"
06:55:51saratogawas the controller disabled too?
06:55:58saratogai'd imagine between the two there'd be some savings
06:56:14saratogaof course compared to that arm9 core maybe it was too small to care about
06:56:18Mouser_XBacklight was turned off, and the LCD controller was turned off. as I understand it.
06:56:21LloreanAs for the CPU, if I recall someone theorized that if we just left it alone it handled scaling itself or something.
06:56:37saratogathat'd be easy enough to test
06:56:49LloreanI have no idea if they were just talking out their ass though.
06:57:07saratogasee if power usage scales linearly with CPU load or not
06:57:08LloreanBut I do know runtime tests show either equivalent or slightly worse runtime when we scaled from 100 to 300
06:57:35LloreanOf course, most instances had a boost ratio of nil, or nearly so, I'd imagine, even then.
06:57:51 Quit linuxstb (Read error: 113 (No route to host))
06:58:04saratogathough i'm not sure how a CPU could know to throttle itself better then rockbox could
07:00:05psycho_maniachas there ever been talk about selecting more then one thing in rockbox? it would work good for deleting.
07:01:58LloreanIt's been suggested many times.
07:02:36JdGordonand nocked back every time...
07:02:40JdGordonwell almost every time...
07:02:50JdGordonnot really easy to do with the current code
07:03:03JdGordonwould be ncie to have though
07:03:14LloreanI thought the idea of a selection widget was smiled upon, just not really done yet.
07:03:26Mouser_XI'd like it for playlist control, when moving multiple items to a different location.
07:05:11 Quit hannesd (Read error: 110 (Connection timed out))
07:05:12 Nick hannesd_ is now known as hannesd (
07:05:29 Quit saratoga ("CGI:IRC")
07:05:43***Saving seen data "./dancer.seen"
07:06:38JdGordondoes it really make sense to allow up to 24x24 pixel icons for the statusbar?
07:06:55JdGordonit makes the code slightly simpler.. but wastes potentially more ram
07:08:14LloreanIt'd depend on the target, I think.
07:08:37LloreanPeople replace the status bar with a rather large bar in the WPS quite often, so I'd imagine them wanting to everywhere if they could.
07:08:39JdGordonyeah, but im looking for a way to do it the same for all tagets
07:09:00JdGordonmaybe we think about dropping the max size on the smaller lcds
07:09:10LloreanWe'll probably want to
07:09:22LloreanEspecially for Archoses and maybe even X5/M5s
07:09:26LloreanAnd the iFP if that ever happens. :-p
07:10:14JdGordonhow much crap would I be in if i removed the number display in the statusbar?
07:10:31LloreanYou mean numeric battery and volume?
07:10:43LloreanA lot, I think
07:10:50JdGordonthats what I was afraid of :p
07:10:57LloreanMost of us prefer knowing an objective volume level
07:11:19JdGordonI'm tihnking of using the sysfont if the ui font is too small on targets without an rtc
07:11:33psycho_maniaci like the numbers more then the battery because i can see the numbers more clearly.
07:11:38JdGordonthe ones with an rtc would load a seperate iconset with the numbers bmp's for the clock
07:11:49JdGordonsane? or stupid?
07:11:56LloreanWhat about volume and battery?
07:12:08JdGordoni'm trying to keep ram usage down...
07:12:13Mouser_XI'd prefer a percentige count, rather than numbers, for battery display stuff.
07:12:30LloreanMouser_X: Battery is a percentage...
07:12:32psycho_maniacwhat does it mean when a wps loads good in the simulator but not so good on the actual player.
07:12:55Mouser_XLlorean: Ah. Goes to show how often I look.
07:12:58LloreanJdGordon: Why does the RTC need custom numbers, and more specifically, why can't those custom numbers be reused for anything numerical in the status bar?
07:13:39JdGordonthe user font may not fit, the sysfont may be too small.. and/or the user may want a differnt font for htme than the other 2
07:13:54JdGordonLinus agreed it was good to allow a seperate set for the numbers
07:14:22JdGordonthey could be reused.. but it adds alot of waste to non rtc builds which would only use them for the batt/vol display
07:15:06LloreanI think that using them for battery and volume is less wasteful than RTC...
07:15:14LloreanThat's two displays rather than one.
07:15:54LloreanIt's only wasteful for users who don't use either as a numeric display *and* don't have an RTC.
07:16:14LloreanAnd it'd look really, really odd for the RTC to have one set of numerals and the Volume and Battery use a different set within the same status bar
07:16:34JdGordonthey would use the same if the rtc was there..
07:16:38JdGordonok, ill always add them
07:16:56JdGordonas a seperate set though so they dont have to be the same dimensions as the icons
07:17:14LloreanProbably a good idea
07:18:03 Quit Mouser_X ("CGI:IRC (EOF)")
07:19:40 Join midkay [0] (n=midkay@rockbox/developer/midkay)
07:19:48psycho_maniacthen people may use the status bar in wps'es. i know i only like it in the menu as of now.
07:20:47 Join Mouser_X [0] (i=cf9bb003@gateway/web/cgi-irc/
07:21:02 Quit Mouser_X (Client Quit)
07:21:43 Join Mouser_X [0] (i=cf9bb003@gateway/web/cgi-irc/
07:23:45Calcipherwhat is RTC?
07:23:58JdGordonreal time clock
07:24:50JdGordon <- /me's latest peice of nasty code :p
07:27:07Calcipherwhat does that do for rockbox?
07:28:22JdGordonits the clock
07:28:34Calcipheroh, so what was changed for it?
07:29:35Calcipheroh, but hasn't that been in there awhile?
07:30:04JdGordonthat paste was for soemthing different
07:30:12psycho_maniacdoes violate copyright ? it seems to resemble rockamp
07:30:46psycho_maniacwoah its already been removed i guess.
07:31:23psycho_maniacok nm looked at wrong one. sorry :S
07:33:43psycho_maniacit looks exactly like rockamp. just a different color.
07:40:05 Quit BHSPitLappy (Read error: 110 (Connection timed out))
07:43:29psycho_maniacwhen AA is comitted. do you need a special sized picture for it to work and format?
07:43:37JdGordoni belive so
07:43:49JdGordonbmp and I dont know the size
07:44:36Mouser_XWhat I'm reading from the logs is that it will look for the size requested by the WPS, and if it doesn't find it (cover.100x100.bmp), then it looks for cover.bmp.
07:45:02Mouser_X(Replace 100x100 with whatever numbers the WPS requests)
07:45:35Mouser_XScaling will be implemented later.
07:46:36psycho_maniaci dont understand this, " No embedded album art support" what does embedded mean in that sentence?
07:46:46Mouser_XIn the ID3 tags.
07:47:02Mouser_X(Of the MP3, or whatever supports images in the music file.)
07:47:30psycho_maniaci never actually download album art. only when wmp downloads them for me without knowing.
07:47:53Mouser_XI've never downloaded it intentionally, though occasionally it comes with the albums...
07:48:41psycho_maniaci once had "show system/hidden" files un ticked and i noticed i have LOTS of album art.
07:49:22aliaskThat'll be windows media player
07:49:42aliaskCreates 3 or 4 files per album.
07:50:19*Mouser_X uses Winamp, and thus doesn't have this problem.
07:50:35psycho_maniacyup actualy these were just for songs. dont know why i used wmp. i usually get bored and try out new programs.
07:50:39CalcipherI still haven't been able to generate sapi voices in cygwin, and have no Idea why
07:51:21Calcipherno one else seems to know whats going on with the error message I get
07:52:27Calcipherand I can try the vmware method, but I don't really know how to use other sythesizers like sapi5
07:53:06amiconnYou cannot use sapi in the vmware image of course
07:53:17amiconnsapi is a windows feature
07:53:21Calcipherin the debian setup, or how to access my Win XP partitions from the vmware environment
07:53:31 Quit homielowe (Success)
07:53:40Calcipherwhere I have a possible alternative tts engine for linux, waiting to be tested!
07:54:35CalcipherI have Loquendo TTS for the tom tom GO 501 GPS system, and that is for a linux system
07:55:00CalcipherI just don't know how to access those files from within VMware
07:55:34 Join midgey_ [0] (
07:55:43 Quit midgey (Read error: 104 (Connection reset by peer))
07:57:00Calciphernobody here seems to be familiar with the cygwin voice generating method, or atleast no one has said anything in a few days
07:57:29 Join sup [0] (
07:58:20amiconnI have no idea what could be wrong on your end
07:58:43amiconnI did quite some things to speed up the sapi voice generation method, and it works like a charm here
07:59:15 Join bertrik [0] (
07:59:17Calcipherwould you like to see the erorr I'm recieving
07:59:27amiconnThere should be a specific error message before the ***make: error N messages
07:59:41amiconnAlso, make sure you have all required programs installed
07:59:59CalcipherI do use non default tts packages, like scansoft, neospeech and loquendo
08:00:34amiconnWell, as long as those are sapi voices (preferably sapi5), there should be no problem
08:00:40amiconnI'm using at&t voices
08:01:10Calcipherwell I was recieving a arm-elf-gcc path error message, but Mouser_X told me of a fix for that particular issue
08:01:36psycho_maniacbooting the OF first, would that be under the bootloader or could that be a patch?
08:02:21Calcipheryeah it was complaining that a path was set incorrectly, but I don't get any other errors after I specify the correct path at startup
08:02:29psycho_maniacIm thinking its in the bootloader, but just want to make sure.
08:02:57Calcipherlet me get the message for you amiconn, one minute
08:03:07amiconnWell, you need a working build environment for rockbox in order to be able to build voices
08:03:25amiconnThat includes a working gcc for your target(s)
08:03:32amiconn...and the path set correctly
08:03:36 Join Crash91 [0] (i=d98bd903@gateway/web/cgi-irc/
08:03:46Lloreanpsycho_maniac: One could make a patched version of the bootloader to do it, yes
08:03:55amiconnAnd you need the encoder (lame) in the path
08:03:57LloreanA quick and dirty hack can be done with a single-character change to the source code.
08:04:39 Quit Crash91 (Client Quit)
08:04:55psycho_maniacLlorean: To load the OF first?
08:08:09 Join homielowe_ [0] (
08:08:14 Nick homielowe_ is now known as homielowe (
08:09:56Calciphercrap, now I can't even go through the build configuration process
08:10:03Calcipherso somethings different
08:10:37CalcipherI must have done the export PATH=$PATH:arm/opt/gcc wrong
08:10:53Calcipherdon't even remember if thats correct
08:11:42Calcipherlame: command not found, I know its there because it went through the other day without errors
08:11:51 Part Llorean
08:11:54psycho_maniacdo the rules here also apply in #rockbox-community?
08:12:17amiconnThe path can't be correct this way, there's a leading slash missing
08:12:22Mouser_XCalcipher: If what you're doing is the command I told you yesterday, then it's "export PATH=$PATH:/opt/arm/bin"
08:12:26psycho_maniacie: enter is not the space button.
08:12:45CalcipherI thought that made more sense
08:12:53Calcipheropt for optional
08:12:55amiconnAnd lame isn't part of cygwin. Get the win32 version and install it somewhere in your path
08:13:29amiconnAnd why do you need to input it every time?
08:13:51*Mouser_X doesn't know how to set the path permanently.
08:14:00amiconnYou should put such additions to the path to your .bashrc
08:14:18Mouser_XI didn't bother finding that file.
08:14:31amiconnIt's in your cygwin home dir
08:14:39Mouser_XI memorized the comand, and did that at startup instead...
08:14:40Calcipherwell folks, I jacked up my path completely
08:14:52Calcipherwhat is the main path normally?
08:15:03Mouser_XCalcipher: Then exit, and restart.
08:15:06amiconnDepends onyour system
08:15:26psycho_maniacyou know how we have a "gradient" selector bar. how about gradient backround color?
08:15:36Calcipherah, ok because of the home folder name?
08:16:16amiconnpsycho_maniac: Make a background image containing that gradient. No extra code, very flexible.
08:16:53psycho_maniacgood point.
08:16:57Calcipheramiconn, how do I add that path permenantly?
08:17:33Calcipheras you mentioned to the .bashrc file?
08:21:55JdGordonwhere do the charcell statusbar icons come from?
08:24:03JdGordonah.. found em
08:24:29JdGordonand dinner time!
08:26:35 Join pengo [0] (
08:27:10 Join Rob2222 [0] (
08:27:19psycho_maniacMouser_X: you have a gigabeat correct?
08:27:53Calcipheramiconn, how do I install lame to be used with cygwin?
08:28:24Mouser_Xpsycho_maniac: Yes.
08:28:32pengoany chance of a new stable release of rockbox coming out someday?
08:28:40Mouser_X(And soon an e250!)
08:29:14psycho_maniacMouser_X: check this out. seems pretty neat and could get comitted sometime. probably not soon though.
08:29:34Mouser_XWhy would I want the OF?
08:29:57psycho_maniacno no that.
08:30:31Mouser_XAh, sorry. I just looked at the link, I didn't click it yet.
08:31:12 Join Agrajag__ [0] (
08:31:21amiconnCalcipher: Just install it somewhere in your OS path. I do usually put it in C:\Windows
08:31:34amiconnA bit dirty, but then it's available from everywhere
08:31:44amiconnkkurbjun: around?
08:32:22Agrajag__Guys, I have some MP3's that are actually just audible books converted to MP3. Everything sees the tags including my iPod Photo (default) but in Rockbox, the tag only appears during playback. In the Track listing under Artist it comes up "Unknown". Any ideas?
08:32:33Calcipherreally ok, thanks amiconn
08:33:59 Part Agrajag__
08:35:53CalcipherI'm looking on sourceforge right now, is that where to get the lame encoder
08:36:22psycho_maniacAgrajag__: maybe you only have one version of the tag?
08:37:17Calcipherwow, I have no Idea what to do with the lame package I just downloaded
08:37:47Calcipherit has a acm, linux, mac, debian, and more folders inside
08:39:05Mouser_Xpsycho_maniac: Interesting patch. I'd probably use it, if it was commited.
08:40:05psycho_maniaci dont know if the guy who said he'd want it commited could actually comit it though. just thought of that.
08:40:42amiconnCalcipher: You just need the lame.exe, nothing else
08:41:17psycho_maniaclinuxstb_: are you around? i was wondering if you could take a look at . maybe it needs to be synced with current svn?
08:41:27CalcipherI'm looking for that now
08:41:31Calcipherwhere did you find that?
08:41:39Mouser_Xpsycho_maniac: "Karl Kurbjun (kkurbjun)" He's commited lots and lots of stuff. He was the one that talked about committing it the most it looked like.
08:41:58 Join Crash91 [0] (i=d98bd903@gateway/web/cgi-irc/
08:42:26pengoi'm just asking about a stable version because i love rockbox and all, but i'm a bit tired of carrying around a safety pin to reset my iriver
08:43:09CalcipherI see a lame encoder on free-codecs
08:43:10psycho_maniacoh alright. im not filmiliar developers and who they are.
08:44:01 Quit Rob222241 (Read error: 110 (Connection timed out))
08:44:01Mouser_Xkkurbjun is in the channel now. He did a lot of work on the Gigabeat port.
08:44:07amiconnCalcipher: "LAME 3.97 release"
08:44:17Mouser_XAnd is working on the m:ROBE now.
08:44:19amiconnThis site is linked from ...
08:46:03advcomp2019pengo, what is the issue
08:46:28Calcipheroh nice, so I will insert it into my windows folder in its own folder? it wouldn't be in the system32 folder would it?
08:46:44 Quit male (Read error: 104 (Connection reset by peer))
08:47:35amiconnNo, you just take the lame.exe and put it in c:\windows
08:47:52amiconnThe rest is just documentation you can keep somewhere if you want
08:48:03 Quit Crash91 ("CGI:IRC")
08:48:15Calcipheroh ok, very nice, thanks
08:49:13CalcipherI included the dll as well
08:49:41amiconnBasically you could put it anywhere you like, but then you would have to edit your system's path in order to access it from anywhere. Putting it in c:\windows is easier because that's already in the system-wide path
08:50:03pengoadvcomp2019: random crashes
08:50:38amiconnYou don't need the dll for using lame on the command line
08:51:23pengoadvcomp2019: not sure if i could be more specific.. sometimes when pressing too many buttons in a row, sometimes at the end of a track
08:51:46advcomp2019pengo, do you have the latest version of rockbox on there
08:51:59Calcipherstill says lame: command not found
08:52:26advcomp2019pengo, plus what player do you have
08:52:29 Quit bertrik ("bye")
08:52:31pengoadvcomp2019: updated about a week ago.. for the first time in a few months.. i was having crashes before, and still having (different) crashes now.. only very occassional
08:52:47pengoadvcomp2019: iHP-120
08:53:56pengo(= iriver H120)
08:54:01 Join desowin [0] (
08:54:11Mouser_XCalcipher: You're using cygwin, right? Try putting "lame.exe" into your cygwin directory? Or, the "home" directory withing your cygwin directory?
08:54:37Calcipherok, I was thinking of doing that too, I'll try that
08:54:56amiconnI'd rather put it somewhere where it's available system-wide
08:55:08amiconnc:\windows should really work
08:55:37Mouser_XWell, I'm just wondering if, for some reason, cygwin can't access it/doesn't look in the windows directory.
08:55:46amiconnIf not, you did something wrong, or your windows is installed somewhere else
08:55:57Mouser_XI'm not saying it does, or does not, but that might be it.
08:55:59amiconnMouser_X: The system wide path is part of the cygwin path
08:56:08advcomp2019pengo, since i do not have the iriver h120 i cant help you there.. all i can say is try the current build and see if it fixes the issues.. sorry if i am not much help
08:56:17Calcipherhmm, well even though I saw that lame : Command not found. message I still went ahead with the make voice command
08:56:20amiconnecho $PATH in cygwin will tell you
08:56:32Mouser_Xamiconn: Ah.
08:56:35Calcipherand its processing and no errors yet
08:56:39amiconnOne component is /cygdrive/c/WINDOWS for sure
08:57:25Calcipherso maybe I was simply getting a no lame encoder message?
08:57:39Calcipherit was error 141
08:57:58*Calcipher gasps
08:58:10CalcipherI recieved no errors?!
08:58:29*Calcipher looks at the manual for the rest of the instructions
09:01:10 Quit midgey_ ()
09:04:47Calcipherwill try his voice file now!
09:05:21Calcipheroops, so happy I forgot the /me
09:05:46***Saving seen data "./dancer.seen"
09:06:37 Join Llorean [0] (
09:07:18 Join [1]psycho_maniac [0] (
09:07:42 Quit psycho_maniac (Nick collision from services.)
09:07:43 Nick [1]psycho_maniac is now known as psycho_maniac (
09:09:07CalcipherI can't wait till zagor finishes up doing his USB magic and I don't have to ever see that Sansa OF
09:09:42 Join midgey [0] (
09:10:41Calcipherjust to make sure, all I need in the lang folder is english.voice, which is what was generated
09:11:23amiconnLooks like it would be useful to put some error handling in the vbscript's EXEC method...
09:12:05 Quit n17ikh|Lappy ()
09:12:08 Part Mouser_X
09:12:48psycho_maniacCalcipher: you are correct.
09:13:42Calcipherhm, well something must have been done wrong, I don't hear a word, and the voice file was replaced with the new one, and I checked that voice menus is enabled
09:15:39CalcipherIt seemed like it was actually doing the processing, since it took some time and then gave no errors, so I don't know what happened
09:19:11CalcipherI think I'll try again after updating the svn build to current build, since the voice file was not for the build I had on the player
09:21:33psycho_maniacthat would really suck if you have to have your own build just for a different voice
09:22:35Calcipherno I just mean that the current build, the build i used when generating the voice, and the build on the player were all different
09:22:55Calcipherand ... hmm svn update said skip
09:23:03Calcipherdidn't update anything
09:23:20Calcipherdoes it do that automatically sometime?
09:23:33 Quit Llorean ("Leaving.")
09:23:51psycho_maniacthats in cygwin?
09:24:26 Quit midgey ()
09:25:48Calcipherhow do I updated the build I'm working with in cygwin, thats what I thought I was going to do when I did the SVN update command
09:26:04Calcipherto current build
09:26:41psycho_maniacsvn up i think
09:27:43psycho_maniac"svn up" updates the source and then you run your "make" commands.
09:27:49Calcipherstill says skipped
09:28:27psycho_maniacmaybe that means your already up to date?
09:28:43CalcipherI did svn status and got a warning message
09:29:11Calciphersays '.' is not a working copy
09:29:42 Join Llorean [0] (
09:30:14Calcipherso I'm thinking my svn settings aren't right
09:31:33Calcipheroh ok I'm an idiot
09:31:35psycho_maniacbut how does it work one second and then not work the next?
09:31:42CalcipherI was in the wrong directory
09:32:01psycho_maniaci was gonna ask that but figured you already checked hehe. i shoulda
09:32:05CalcipherI'm not used to shells
09:33:23pengocrashed again. :/
09:33:31 Quit homielowe (Read error: 104 (Connection reset by peer))
09:37:54psycho_maniacpengo: have you updated to the current build?
09:38:28pengopsycho_maniac: about a week ago
09:38:40pengoi did it cause i thoguht it might be more stable
09:38:55psycho_maniacthen that is not the current build. the current build is updated every time something is changed in the code.
09:38:55pengoi came here to ask if there's any plans for a stable release?
09:39:07pengopsycho_maniac: is it going to be worth my while?
09:39:15pengohow about that
09:39:27psycho_maniacit wouldnt hurt.
09:39:43pengoit might :/
09:40:33psycho_maniacusually in order to talk about a problem you have to have the newest current build.
09:40:51pengothe problem is that there's no stable release.
09:42:03pengoa stable release is where you don't add any features, and just concentrate on fixing bugsx
09:42:40pengoso that your users don't get sick of using buggy software all the time
09:42:54aliaskpengo: There have been a number of commits to rockbox in relation to these crashes recently, it'd be worth upgrading
09:43:02pengoaliask: cheers
09:43:33 Quit Rob2222 (Read error: 104 (Connection reset by peer))
09:43:52 Join Calcifer [0] (
09:43:59 Quit Calcipher (Read error: 104 (Connection reset by peer))
09:44:07 Join Rob2222 [0] (
09:44:18 Quit amiconn (Nick collision from services.)
09:44:26 Join amiconn [0] (n=jens@rockbox/developer/amiconn)
09:45:41 Join Feisty [0] (
09:46:39 Quit Feisty (Client Quit)
09:48:08 Join homielowe_ [0] (
09:48:13 Nick homielowe_ is now known as homielowe (
09:48:22 Join Thundercloud [0] (
09:49:22 Quit psycho_maniac (" HydraIRC -> <- \o/")
09:52:38 Join J-23 [0] (
09:53:20 Join TPM_Debian [0] (
09:53:30Calciferamiconn I see you made some changes to the sapi voice script, what does that add?
09:53:41J-23How big's Sansa c200 screen?
09:53:50JdGordondoes anyone use a custom iconset for the viewers icons on a remote lcd target?
09:53:50TPM_Debianhi @all
09:54:04amiconnIt just adds error reporting for a case like yours, e.g. if lame is missing
09:57:38CalciferI had noticed the changes since I saw the update in svn, while trying to get everything up to date
09:57:59Calciferthe generated voice file didn't seem to work
09:58:28 Join linuxstb [0] (n=chatzill@rockbox/developer/linuxstb)
09:58:28Calciferand when I go into configure I still see lame: Command not found.
09:58:46Calciferbut I don't recieve any errors when I run make voice
09:59:26linuxstbaliask: Around?
10:06:38*amiconn points to the device comparison chart
10:07:46amiconnJdGordon: In fact I have almost forgotten about the custom icon sets. I don't use them on any of my targets
10:08:11amiconnBut not I remember that there's something left to do for charcell...
10:08:16CalciferI just tried the entire process with the current build, although I still don't know if its recognizing the lame.exe in my windows folder
10:08:22JdGordonim redoing the way they are stored to hopefully make it much nicer to work with multiple cds
10:08:38Calciferbut I get a generated voice file with no errors
10:08:53Calciferand I don't hear the voice in rb
10:09:22amiconnwhat does 'which lame' tell you?
10:09:40*JdGordon wonders if this should be commited before the statusbar icons are done or not
10:09:47Calcifercommand not found
10:10:02amiconnSHould be /cygdrive/c/WINDOWS/lame if you installed it where I suggested
10:10:52amiconnJdGordon: I don't want cds in my mp3 player :P
10:11:20J-23hmm, 132x80 movies?!?
10:11:23Calciferwell the lame.exe is inside C:/windows
10:11:34amiconnecho $PATH
10:11:53amiconnThere should be a component /cygdrive/c/WINDOWS
10:12:21 Nick bb_ is now known as bb (n=bb@unaffiliated/bb)
10:12:49Calciferlooks like my windows paths are for /system32
10:12:53Calcifernot just windows
10:13:19 Join n17ikh|Lappy [0] (
10:13:27amiconnThere are several, of course
10:14:32Calciferactually yeah your right c:/windows is one of them
10:14:41Calciferso it should be detecting it
10:15:05TPM_Debianhello linuxstb .. i got a question about sansa connection to debian ... i found the device in recovery and manu-mode, but not in normal-boot mode (OF) ... what can i do to fix this problem?
10:15:16Calciferthrew a copy into system32
10:15:22Calciferand now which lame sees it
10:17:03Calciferis there a command for seeing which voices are available?
10:17:24CalciferI see something about a list voice vbs on the wiki
10:20:08 Join linuxstb__ [0] (
10:20:49aliasklinuxstb__: I am now
10:21:33TPM_Debianhello linuxstb__ .. i got a question about sansa connection to debian ... i found the device in recovery and manu-mode, but not in normal-boot mode (OF) ... what can i do to fix this problem?
10:24:43 Join ddalton [0] (
10:24:52linuxstb__aliask: Did you read my questions in the logs about merge0.cpp ? I'm trying to find the current version.
10:25:18linuxstb__The wiki has some amusing text saying "see below", followed a few paragraphs later with "see above"...
10:26:07amiconnstkov wiki ;)
10:26:24amiconnkkurbjun: ping
10:26:29aliaskLatest one is here in here:
10:26:31linuxstb__TPM_Debian: In normal mode (when the original firmware is running), it probably appears as /dev/sda1
10:26:57TPM_Debianlinuxstb__, hmm .. i cant found it in fdisk -l
10:27:00aliaskArgh, sorry, no it's not...
10:27:30linuxstb__aliask: That's the one I was looking at, but it seems buggy - it stops reading the input file (the bootloader in my case) when it finds 0 in the input file.
10:27:53aliaskYes, if you just remove that behaviour it works fine, ptw put that in there for some reason
10:27:56linuxstb__And in the case of the bootloader, this happens at about 64 bytes.
10:28:14aliaskI think it's something like "if(!*ptr) break;"
10:28:20linuxstb__Ah, so it's not the version you're using...
10:28:29aliaskI thought I had uploaded my one
10:28:32Calciferwell now lame is detected, and I leave both sapi and lame on defaults
10:28:43Calciferwhich I hope is ok
10:28:55Calciferyet the voice file does nothing in rb
10:29:05JdGordonamiconn: do you have a problem with pulling the charcell statusbar code out of gui/statusbar.c and putting it into player/statusbar.c?
10:29:32JdGordonby the time im done there will be almost no shaed logic
10:29:45linuxstb__aliask: Do you have any ideas about how we can make the bootloader dual-boot? I don't understand the gigabeat S boot process yet...
10:30:30amiconnSomeone should go ahead and rename player/ and recorder/ ...
10:30:37aliaskI'm not 100% clear on it either, but I can try to explain it as I best understand it
10:30:47 Quit linuxstb_ (Read error: 104 (Connection reset by peer))
10:30:48Calciferamiconn, is there anything I have to specify when configuring the voice makefile?
10:30:57 Join linuxstb_ [0] (n=linuxstb@rockbox/developer/linuxstb)
10:31:13aliasklinuxstb_: Bad connection?
10:31:17ddaltonCalcifer: whats the problem? I have built voice files before
10:31:28linuxstb__aliask: I would like to try and make the bootloader both dual-boot, and be able to load a binary from the FAT32 partition - e.g. a file disguised as a .mp3 transferred via MTP.
10:31:39 Quit linuxstb (Read error: 110 (Connection timed out))
10:31:40CalciferI don't know what the problem is
10:31:42 Nick linuxstb__ is now known as linuxstb (
10:31:49Calciferthey don't sound in the player
10:31:51 Quit linuxstb_ (Remote closed the connection)
10:31:55ddaltonCalcifer: First run configure.
10:31:58amiconnJdGordon: player/ and recorder/ actually mean charcell/ and bitmap/ most of the time. A few files are different though
10:32:01ddaltonhave you done that?
10:32:06CalciferI don't recieve any errors in the generation process
10:32:18ddaltonfrom a build dir which should be in trunk/build_dir
10:32:26ddaltonor what ever you call build_dir
10:32:28JdGordonamiconn: yeah, player is charcell... but i dont like just renaming recorder to bitmap...
10:32:49ddaltonCalcifer: are you on linux?
10:32:55ddaltonoh actually don't worry...
10:33:00Calcifer•ddalton• I have named my dir "build" and I'm in cygwin
10:33:05linuxstbJdGordon: I don't see a problem with renaming it that way - we can later move files that don't belong there elsewhere.
10:33:14ddaltonCalcifer: Ok select your actual target. Make sure it is the right one!
10:33:29ddaltonCalcifer /trunk/build?
10:33:55ddaltonrockbox is the trunk right?
10:34:09aliasklinuxstb: The OF bootloader loads each chunk of the nk.bin file to the designated memory locations (specified in the nk.bin file), and then jumps to the insertion point in the code (also specified in nk.bin). By adding another "chunk" in the nk.bin at the same address as the insertion point, it overwrites the OF in RAM, and when it jumps to the insertion point, it jumps into our code.
10:34:19amiconnJdGordon: Btw, does the m:robe play music yet?
10:34:24JdGordonnot yet
10:34:35JdGordonwaiting on eigima to get the dsp going
10:34:44Calcifer•ddalton• I run tools/configure and select my target, advanced, Voice, language, tts engine, and encoder
10:34:50Calciferno errors recieved
10:34:51linuxstbaliask: So could we load it to a different address in RAM, so as to not overwrite the OF?
10:35:03ddaltonOk once selecting the voice hit enter twice...
10:35:10ddaltonthen type make voice.
10:35:20Calciferyeah I do thta
10:35:23ddaltonthen tell me how big the voice is in mb/kb/bytes etc
10:35:34Calciferand don't recieve errors
10:35:48CalciferI saw my last attempt was 1.0mb
10:36:10linuxstbaliask: I noticed in the button driver that the hold switch is simply attached to a GPIO, so the code in crt0.S could check that GPIO (as the very first thing it does), and then branch to OF if hold is off. Booting with hold on would run our code.
10:36:34aliasklinuxstb: I suppose so, as long as we change the insertion point in nk.bin as well.
10:36:52aliaskYes, good idea
10:37:04ddaltonCalcifer: Try installing the fonts. BTW what target are you building for?
10:37:09Calcifer•ddalton• 1058kb
10:37:19linuxstbOr another option would be to append our bootloader to the existing OF chunk, and change the first instruction to branch to our code.
10:37:50amiconnJdGordon: Hmm, that's something I don't understand. Why work on some fancy gui stuff if the basics aren't working?
10:38:27Calciferddalton, what do you mean installing the fonts?
10:38:34ddaltonCalcifer: You select e200 r? Your not getting confused with another sansa player?
10:38:34JdGordonbecause im not a hardware hacker... and this has been bugging me on my sansa for ageas...
10:39:29 Join StOaN [0] (
10:39:45CalciferI don't think so
10:40:03aliasklinuxstb: I like the first idea better personally
10:40:16*amiconn was never bugged by the status bar
10:40:25*JdGordon not surprised
10:40:31Calciferddalton, no I see here it lists as e200r before generating the voice file
10:40:32amiconnIt works and shows what I need
10:40:36Calciferso no mistake there
10:40:45amiconnI even use it in the wps, instead of reinventing the wheel
10:41:19 Quit daurnimator ("Cyas")
10:41:26amiconnCalcifer: Try e200 (not e200r)
10:42:07amiconnAfaik we run the 200 build on the r, and the e200r option in configure is a leftover
10:42:19amiconnMaybe it's needed for building the bootloader, not sure
10:42:30TPM_DebianLlorean, if i list the fdisk tabel, some wrong devices will show from sda1 to sda4 (in recovery mode) ... sda0 is the 16MB device ... but in normal mode will no other sda device show ....
10:43:00advcomp2019Calcifer, what are you trying to do again
10:43:46linuxstbaliask: OK, so getting dual-boot working seems relatively easy. Any ideas about the second problem - locating a binary transferred in MTP mode? n1s did a test and was able to transfer a binary file renamed to .mp3 using a Linux mtp tool.
10:44:07GX9900WZHi, is it possible to load two seperate font together?
10:44:15 Join ompaul [0] (n=ompaul@gnewsense/friend/ompaul)
10:44:15Calcifergenerate sapi5 voice files for my Sansa e200r
10:44:35GX9900WZI meant two different type of fonts together.
10:44:54aliasklinuxstb: I have experimented with using a custom merge0 (btw the updated one is now on the wiki) which added another chunk for the rockbox.gigabeat file. The OF bootloader even loads it into memory for us.
10:45:18aliaskBut I think I got held back by memory configuration issues, it wouldn't boot properly.
10:45:18Calciferwhen generating voices, it reads "TTS engine (options) : sapi ()
10:46:17aliasklinuxstb: However that method is a little clunky if you want to update the build
10:46:54StOaNis there anything i can do when my ipod is freezes after choosing "recording" in the main menu?
10:47:52Calciferoh wow, staying up trying to stubbornly get this to work probably ruined my entire day tomorrow
10:49:37*Calcifer celebrates quietly since its 4:49am here
10:49:57Calciferamiconn your idea of using the e200 option did it!
10:50:04linuxstbaliask: I'm just thinking of an easy way to run test code for the moment - so we flash a bootloader once (think of it as a development bootloader), then we can start work on the real bootloader, which we transfer with MTP as a fake .mp3 file. I also had the idea of transferring a .tar file as fake mp3, and making our bootloader extract the files (.tar is a very simple format to extract).
10:50:18Calciferthank you so much! I can sleep easy this morning
10:50:57CalciferI guess someone should note that the e200r option doesn't work with voice generation
10:51:16aliasklinuxstb: A tar of .rockbox?
10:52:46Calcifernice, much nicer sounding than the speex or espeak voice
10:53:00 Join tywickra [0] (i=tywickra@nat/redhat-in/x-6920e22946ec65fe)
10:53:02aliaskrockbox boots without it though, couldn't we just omit that temporarily (hopefully usb will be coming soon, so once we have the firmware image we can use usb msc and transfer .rockbox normally)
10:53:27tywickraaliask, how about the battery life with MP3s ??
10:53:33tywickraon rock box ?
10:54:09tywickraGuys.. someone could help me with the battery life of the RockBox, on iPod classic ??
10:54:09Calciferwhoa, its not reading things correctly at all though
10:54:29aliasktywickra: Rockbox doesn't support ipod classic
10:55:01tywickraaliask, ah.. what am i suppose to do then :( ?
10:55:50Calciferwow, I have no idea how the voice is reading things wrong, for example for every menu item past now playing it reads the selections wrong
10:56:07aliasktywickra: You have no choice but to use the original firmware.
10:56:25Calciferand under settings its also reading selections wrong or not reading certain selections at all
10:56:26tywickraaliask, oops.. i'm stuck it seems..
10:56:52ddaltonCalcifer: I haven't been following but did you get the fonts?
10:57:01ddaltonand whats wrong now? It doesn't say the right menus?
10:57:17linuxstbaliask: Sure, but adding support for untarring a file wouldn't take very long at all (I already have some .tar handling code I wrote as part of a patch for loading .wps files from a .tar), and it could be useful whilst we wait for USB.
10:57:38Calciferddalton, right, its reading the wrong things
10:57:40StOaNis there a way to restart my "ipod video(ma146fd)" when it is frozen?
10:57:54aliasklinuxstb: Oh ok. I was concerned that it'd be sidetracking us from the main point, but if it's really easy, we may as well.
10:57:54Calciferddalton, no I did not get the fonts, how do I do that?
10:58:01ddaltonCalcifer: Ok go and install the fonts. and go to extras.
10:58:28ddaltonunzip them to /.rockbox/fonts and make sure the fnt files are located at: /.rockbox/fonts*
10:59:29Calciferddalton, so I don't have to change anything in my build folder?
10:59:45ddaltonCalcifer: If Im understanding you right no
11:00:19Calciferddalton, my build is updated to 155769
11:00:50amiconnIf you build yourself, just use 'make fullzip' instead of 'make zip'
11:01:05amiconnThe .zip generated by this contains all fonts
11:01:20amiconnYOu don't need to do this everytime, as the fonts don't change often
11:01:30 Join stewball [0] (n=WTFOMGBB@
11:01:30Calciferddalton, ok installed the fonts
11:01:40linuxstbaliask: Also, I would like to commit a merge0.cpp program to SVN (similar to the other mkboot.c utilities for other targets). So I started rewriting it in plain C, following Rockbox coding guidelines. Now I know that the merge0.cpp I was looking at is correct (minus that break;), I'll finish it and commit.
11:02:09 Quit linuxstb ("ChatZilla [Firefox]")
11:02:36 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
11:02:44 Quit StOaN ("leaving")
11:02:52aliasklinuxstb: Excellent. That file always kinda irritated me.
11:03:22Calciferamiconn, what are some of the lame options I can specify? and what are the defaults set to anyway
11:03:44ddaltonCalcifer: lame opts won't fix your problem.
11:03:55ddaltonwhats the path to the fnt files?
11:04:25Calciferddalton, Oh I know, I just thought it would be useful to know those
11:05:20Calciferddalton, the path is the same as always, when I extracted I was asked to overwrite the old fonts so I know its in the right directory
11:05:36ddaltonCalcifer: not exactly...?
11:05:42ddaltoncan you check for me?
11:05:48***Saving seen data "./dancer.seen"
11:06:16Calciferddalton, you mean where they are located in my .rockbox folder on my player?
11:06:33 Join Robin0800 [0] (
11:06:46ddaltonCalcifer: not exactly
11:06:59ddaltonCalcifer: Whats in /.rockbox/fonts?
11:07:03ddaltonwhat do you see there?
11:07:45Calciferddalton, 77 .fnt files totalling 9.22mb
11:08:34ddaltonCalcifer: And your problem still occurs?
11:09:07Calciferwell lets try it again, let me generate a new voice file and try it, I'll let you know
11:09:21ddaltonCalcifer: why. Whats changed?
11:09:26ddaltonjust the fonts?
11:09:37ddaltonupdate your build then to. But that probably won't do much
11:09:53Calciferddalton, no nothing, I just wanted to try a different voice
11:10:19ddaltonCalcifer: the voice-synth isn't the problem. Sounds like english.voice is been generated ok.
11:10:37Calciferddalton, yeah it is
11:10:55ddaltonyou don't have any patches installed? Your using the same version of english.lang as your build is?
11:11:18Calciferddalton, I'm generating a new one just to hear a different voice, I know reading the wrong thing is a due to some other problem
11:11:50Calciferddalton, I don't even know how to install patches yet, so no no patches
11:11:51ddaltonCalcifer: one thing at a time! :-)
11:12:05ddaltonCalcifer: ok
11:12:15ddaltonCalcifer: are you using english?
11:12:43Calciferddalton, also you should note that when I tried using the e200r target, the voice never read a thing
11:13:34Calciferddalton, so as amiconn suggested I set the target to e200 spec and then the voice did work, even though it read most items incorrectly
11:13:50Calciferddalton, yes I am using english
11:15:08 Join pixelma [0] (n=Marianne@rockbox/staff/pixelma)
11:16:24Calciferddalton, strange, I set a new default voice in windows tts options, and the generation process still used the same voice, so I'm going to have to figure out how to specify which I want to use
11:16:29 Join davina [0] (
11:17:09Calciferddalton, but aside from learning that, nothing came of the changes, the voice reads "files,database,now playing" correctly
11:17:56Calciferddalton "resume playback" also, but after that it reads "playlist" instead of "rocording"
11:18:18Calciferddalton, and from there on its all wrong
11:18:18ddaltonCalcifer: Strange. Latest build and lang file?
11:18:49ddaltondo I need to be root for
11:18:58Calciferddalton, wouldn't the lang file be updated through "svn update"
11:19:17ddaltonCalcifer: it would. Then did you make your own build?
11:19:38Calciferddalton, I don't know how
11:20:13Calciferddalton, what I did was create the voice file for the same current build I had installed at that time
11:20:33ddaltonCalcifer: cd rockbox mkdir build_dir cd build_dir ../tools/configure number n make
11:20:33 Join n1s [0] (
11:20:41ddaltonreplace number with your target's number
11:22:50 Join daurnimator [0] (n=daurnima@unaffiliated/daurnimator)
11:24:25Calciferddalton, ok it said created makefile
11:25:22Calciferddalton, when asked for the platform i selected 50 which is the e200, and type i selected n for normal mode
11:25:44 Join webguest03 [0] (i=dcfd4b9c@gateway/web/cgi-irc/
11:27:38ddaltonCalcifer: run make
11:27:54Calciferddalton, make all?
11:28:59Calciferddalton, I ran make help, and saw the options, but I ran make alone since you didn't specify any option
11:29:00 Quit J-23 ()
11:29:13ddaltonCalcifer: make will do
11:29:24ddaltonyou should see it compile or the files
11:30:14Calciferddalton, sounds like this process is not far from compiling with patches installed right?
11:30:37 Join J-23 [0] (
11:31:02Calciferddalton, so maybe after sorting this out that will me my next project, to include those nice accesibility patches you mentioned to me
11:31:47ddaltonCalcifer: it will be a big one. Best to email me and get me to send you combined.diff and that way no dependencies
11:32:04ddaltonjust one patch to apply. But I haven't re-synced for ages.
11:32:11ddaltonIm on 15273
11:32:28Calciferddalton, wow
11:33:04 Quit daurnimator (Remote closed the connection)
11:33:22 Quit amiconn (" HydraIRC -> <- s0 d4Mn l33t |t'z 5c4rY!")
11:33:26 Join daurnimator [0] (n=daurnima@unaffiliated/daurnimator)
11:33:48Calcifermy boss, whos blind and now using rockbox with my help, might be better off with your custom build
11:34:12Calciferhe also uses a e200r
11:38:03TPM_Debianwhat does need i to use to write my owen plugins for rockbox?
11:39:23CalciferI see a ton of code went in to adding support for different media formats
11:48:14 Join mf0102 [0] (n=michi@
11:49:33 Join GnosticE [0] (
11:53:09 Join ender` [0] (
11:55:21 Quit ddalton ("I was using BOFHNet IRC version 1.2 by fmillion - get your copy today from !")
11:55:58 Quit aliask ("ChatZilla [Firefox]")
11:58:10 Part Robin0800
11:58:29 Join Robin0800 [0] (
11:59:42 Join spiorf [0] (
12:02:08 Part pixelma
12:09:20 Join Nico_P [0] (n=nicolas@rockbox/developer/NicoP)
12:20:21 Join dj-datavirus627 [0] (
12:22:00GX9900WZhi, i installed rockbox
12:22:15GX9900WZbut my player is not restarting itself when i disconnect from usb in OF
12:22:59J-23my too.
12:23:10dj-datavirus627haj bot GX9900WZ
12:24:26GX9900WZ... i think it's suppose to restart itself, right?
12:24:41linuxstbWhat kind of player is it?
12:24:59GX9900WZsansa e250
12:25:30linuxstbThen I've no idea. Did it use to restart before you installed Rockbox?
12:25:40n1sCalcifer: sounds like you have an old build with a new voicefile, update your build to the same revision that you used to build the voicefile
12:26:41Nico_Plinuxstb: I just read the logs... I see you're catching up nicely on the S :)
12:28:10dj-datavirus627ok cool box :)
12:29:01 Part dj-datavirus627
12:29:24GX9900WZI am just afraid that I bricked the player. It doesn't wake up when I connect it to usb anymore. It doesn't restart after disconnecting from usb.
12:34:19n1scrap, found the windows laptop but no powerbrick :(
12:37:23 Join barrywardell [0] (
12:39:12 Quit spiorf (Read error: 110 (Connection timed out))
12:41:40 Join amiconn [0] (n=jens@rockbox/developer/amiconn)
12:43:29amiconnkkurbjun: around?
12:44:37Robin0800Nico_P, managed to get your album art patch installed this morning all working, players been on for 2 hours with no problems
12:44:42 Join pixelma [0] (n=Marianne@rockbox/staff/pixelma)
12:45:10 Join moos [0] (
12:49:46 Nick parafin|away is now known as parafin (
12:50:00 Join bertrik_ [0] (
12:50:25 Quit davina (Remote closed the connection)
12:52:12 Join davina [0] (
12:53:12 Join Seed [0] (
12:53:24Robin0800Nico_P,is there anything else to do with this patch before it can be commited?
12:53:33 Join lee-qid [0] (
12:56:11Nico_PRobin0800: not much, no. I think I'll be committing it today
12:56:12 Join mirak [0] (
12:56:38Nico_Pamiconn: do you think I should enable AA on targets with 2bpp displays? I'm thinking no
12:56:39linuxstbDoes anyone know ptw419's real name?
12:56:56amiconnNico_P: Why not? Just tell the bmp loader to dither
12:57:03 Join spiorf [0] (n=spiorf@
12:57:19Nico_Pamiconn: ah, right. thanks :)
12:57:41Nico_Plinuxstb: James Espinoza I think
12:59:17Nico_Pyes, that's it. he's james83 in SVN
12:59:32amiconnNico_P: In fact enabling dithering makes sense on all targets
13:00:10pixelmalinuxstb: you can even find that in the Rockbox wiki ;)
13:00:26linuxstbHe has SVN access? Seems he's not in CREDITS though...
13:00:54linuxstbpixelma: Where? I couldn't see him in IrcNicks...
13:01:18Nico_Plinuxstb: and I added him to IrcNicks
13:01:35 Join Frazz [0] (
13:01:38JdGordonanyonw feel like being creative?
13:01:52Nico_Plinuxstb: he should probably be added to CREDITS too
13:02:00*JdGordon needs some icons made up :p
13:02:53Nico_Pamiconn: how is dithering enabled?
13:04:03amiconnJust or FORMAT_DITHER into the format parameter
13:04:10Nico_Pon bitmap load?
13:04:23Nico_Phmm obviously
13:05:03Nico_Phmm does loading the AA bitmap with FORMAT_TRANSPARENT make any sense?
13:05:23JdGordonprobably not
13:05:50***Saving seen data "./dancer.seen"
13:06:14rasherHm, why not? If someone sets up a replacement albumart
13:06:36rasherSay, for a "mixed" dir for which he has no albumart
13:06:49rasherDon't know how expensive it would be though?
13:07:50Nico_Prasher: if there is no albumart, nothing will be displayed
13:07:59Nico_Por maybe I don't get your point...
13:08:47rasherNico_P: Well, say I have a directory of all sorts of crap. I could put a cover.bmp with some sort of symbol there. Which might have transparency. Don't know if it's terribly likely though.
13:08:50amiconnSpecifying FORMAT_TRANSPARENT when loading currently does nothing
13:09:09amiconnThis is meant to build a transparency mask on low depth targets
13:09:19amiconnBut that's not implemented yet
13:09:50Nico_Pok... on display I don't call the transparency version of bitmap_part
13:12:04 Quit Seedy (Read error: 110 (Connection timed out))
13:18:35Nico_Pwell I think I'm ready to commit... anyone want to see the patch first?
13:20:50linuxstbNico_P: Did you say it causes a 4K bin-size increase?
13:21:14Nico_Plinuxstb: yes, on gigabeat... I'll check on coldfire
13:21:31moosNico_P: we will see the deltas...
13:22:32*Nico_P does a few test compiles
13:22:40rasherIsn't it usually less on coldfire? Assuming the same code
13:24:00Nico_Pseems to be ~3k on coldfire
13:26:12 Join linuxstb_ [0] (n=linuxstb@rockbox/developer/linuxstb)
13:26:18 Quit linuxstb (Nick collision from services.)
13:26:22 Nick linuxstb_ is now known as linuxstb (n=linuxstb@rockbox/developer/linuxstb)
13:27:44*linuxstb goes to look for a Gigabeat album-art theme that requires no patches
13:28:26n1srasher: coldfire also cheats by using -Os flag when compiling ;)
13:29:19moosNico_P:will you commit a AA wps too?
13:29:34Nico_Pmoos: not yet
13:30:20Nico_Poh no please don't tell me the build server is stalled
13:31:02 Quit webguest03 ("CGI:IRC (EOF)")
13:31:25Nico_PBagder: ping
13:31:41 Nick fxb__ is now known as fxb (
13:31:47n1shaha, that build system is a real bastard :P
13:31:55rasherThat's not a terribly good time to break
13:32:05Nico_PI should have checked :/
13:33:39 Join linuxstb_ [0] (n=linuxstb@rockbox/developer/linuxstb)
13:33:41 Quit linuxstb (Nick collision from services.)
13:33:43 Nick linuxstb_ is now known as linuxstb (n=linuxstb@rockbox/developer/linuxstb)
13:36:34moosdon't you notice the buils system hang sundays those weeks?
13:37:09moosand no sweedish the sunday in the day :(
13:38:59pixelmaBagder implemented a build system reset after a while (I think it's 20 minutes or so)
13:39:51moospixelma: heh, good news
13:40:13linuxstbSpeaking of which - it seems to have triggered...
13:40:25moosyup cool
13:40:38*linuxstb gives Bagder a cookie
13:40:59linuxstbNico_P: What WPS have you been using to test?
13:41:08moospixelma: this reset system is recent, no?
13:41:14Nico_Plinuxstb: iCatcher with AA tags hacked in
13:41:43linuxstbIs it usable?
13:42:03Nico_Plinuxstb: for everyday use, not really
13:42:10pixelmaI think 1 or 2 weeks ago
13:42:35moospixelma: that's nice
13:43:41linuxstbNico_P: Can you see any patch-requiring tags in this WPS?
13:43:48 Quit Robin0800 (" Want to be different? Try HydraIRC -> <-")
13:44:25Nico_Plinuxstb: no
13:45:35 Join Arathis [0] (
13:45:45*Nico_P closes FS #3045
13:47:27moosNico_P: congrates, you have surely made lots of AA users happy (I'm not), and maybe reduce the number of use o unsuported builds
13:47:51 Join Domonoky [0] (
13:47:58moosnot too much red ;)
13:49:26JdGordonnice big delat :p
13:50:09moosindeed, but not too much anyway
13:50:28moosnice red builds too for mrobe and ifp
13:50:43Nico_Pit's a very stupid mistake
13:50:59*amiconn wonders what aa has to do with the firmware layer
13:51:28JdGordonits playback.c ....
13:51:30JdGordonnot firmware
13:52:17linuxstbI'm guessing it's because hwcodec has metadata handling in firmware
13:52:30 Part J-23
13:52:49Nico_Pamiconn: what do you mean?
13:53:10Nico_Pthe audio.h thing?
13:53:30rasherIf this doesn't deserve a MajorChanges entry, few things do (in terms of how many people care)
13:53:40amiconnThe change in firmware/export/audio.h
13:54:16JdGordonwe now need rbutil to manage the aa bmps :p
13:54:19Nico_Pamiconn: I agree it's a bit out of place, but I wanted to put it next to audio_current_track
13:58:43rasherI almost dare not ask, but would it make sense to look in a static location for albumart, in addition to the "current-dir" relative ones? (./rockbox/albumart/ perhaps)
13:59:04n1srasher: why
13:59:40linuxstbitunes users...
13:59:49rashern1s: to avoid cluttering the music dirs mostly, and in admittedly rare occassions, avoid duplicates
13:59:58rasherAlso itunes users I guess
14:00:17n1srasher: you mean like images matching dirnames or?...
14:00:39n1salbum tags I suppose
14:01:31rasherWell currently "albumname.bmp" is looked for in the same dir as the file (and parent dir, perhaps) - I'm suggesting looking in /.rockbox/albumart/ as well
14:04:50DomonokyJdGordon: depending on where to get the aa from, it woudnt be too hard for rbutil to manage the bmps (resize, convert)
14:05:43 Quit GnosticE ()
14:05:57amiconnHow is albumname.bmp supposed to work?
14:06:17amiconnAlbum tags can easily contain characters which are illegal on the filesystem
14:06:38rasherThen it won't work, I guess. Don't think any sanitizing is done
14:06:43Nico_Pamiconn: then file_exists will fail
14:06:56amiconnI would have expected it to look for some static filename, like cover.bmp
14:07:00Nico_PIMHO that's the user's problem
14:07:06Nico_Pamiconn: it does too
14:07:07rasheramiconn: it does
14:07:08Domonokyis it posssible to have bmps with different sizes for the same albums ? (for different themes..) ?
14:07:15Nico_PDomonoky: yes
14:07:26*Nico_P is updating the wiki page
14:07:30 Quit advcomp2019 (Read error: 104 (Connection reset by peer))
14:07:46 Join advcomp2019 [0] (n=advcomp2@unaffiliated/advcomp2019)
14:10:32pixelmaNico_P: ifp sim is still red
14:10:35amiconnNico_P: Still red...
14:10:43amiconnBtw, iFP is a mono target
14:10:57Nico_Phaha that one is to far right, I didn't see it :p
14:11:34pixelmalame excuse... I saw it on the small display I'm currently working on :P
14:12:16 Join karashata [0] (
14:13:01Nico_PI don't understand why that part gets compiled on the iFP... it's mono :p
14:13:37TPM_Debiancan i show movies whis rockbox? and when yes .. which type of movies can i show?
14:14:08Nico_Phmm I guess I'm just stupid
14:14:21n1sTPM_Debian: read the PluginMpegplayer wiki page
14:16:02amiconnNico_P: Hmm, why does the m:robe 500 see 400 bytes red delta even though aa isn't enabled in its config-*.h ?
14:17:02Domonokywhere is album art normally stored on PC ? (embedded in the file, or somewhre on the hardrive ? )
14:17:18*Domonoky thinks about rbutil albumart support :-)
14:17:32JdGordonDomonoky: grab it off the net :p
14:17:34Nico_Pamiconn: because of what I fixed in my latest commit... but I just saw I didn't completely fix it :(
14:18:26DomonokyJdGordon: do you have an url for me ? :-)
14:18:48JdGordonamarok seems to try getting it of amazon
14:19:10amiconnA target where aa isn't enabled shouldn't see a delta, should it?
14:19:26amiconnI mean swcodec targets of course
14:19:48Nico_Pamiconn: no. I simply made a mistake for the #ifdef in buffering.c
14:19:50*Domonoky would prever a free album art service... using amazon, doesnt sound to legal.. :-)
14:20:45linuxstbDomonoky: Aren't there already apps to do that?
14:20:53rasherDomonoky: amazon offers an api for getting it - I duobt it's terribly illegal ...
14:21:02SoapWhat "free" service are you going to find whose business model is redistributing the artwork off of other people's albums?
14:21:39TPM_Debiannls .. i cant foudn the mpegplayer in plugins/app and the video files in "file browser" it does not show
14:22:03linuxstbTPM_Debian: Are your videos .mpg files?
14:22:17Domonokyfor rbutil, the best would be to only convert /resize/.. them... but i dont know where rbutil should find those images.. :-)
14:22:42TPM_Debianmp .. does't show all files in the browser?
14:22:51Domonokybut if amazon has an api for it, we may try it.. (if i find enough time ) :-)
14:23:45JdGordon... if you can grab it from amaroks databse that would be sweet also :)
14:24:21 Join J3TC- [0] (
14:24:59TPM_Debianlinuxstb, the files named as .mov
14:25:18Nico_Pdamn... red again
14:26:30J3TC-whoa...album art is implemented? :3
14:27:50Nico_Pyes :)
14:28:06amiconnyes :(
14:28:11BigBambiyes :)
14:28:20bertrik_(confused smiley)
14:28:31linuxstbTPM_Debian: That is the wrong type of movie. You need to convert them to .mpg files - the PluginMpegplayer wiki page tells you how.
14:28:56Nico_Psorry about all the red... :/
14:29:02TPM_Debianok .. but why will not show the files in the browser?
14:29:11linuxstbIs there a reason we don't expose the lcd_setmargins() function to .wps files?
14:29:39linuxstbTPM_Debian: You probably have the "show files" option set to "supported" - and .mov files are not supported.
14:29:45DomonokyTPM_Debian: if you want to see all files, you have to change the file view in rockbox to all
14:29:55TPM_Debianin the config?
14:30:18DomonokyTPM_Debian: somwhere in the mnü, read the manual :-)
14:30:34TPM_Debianok .. thx
14:33:37J3TC-To enable for a target, #define HAVE_ALBUMART in its config file.
14:33:42J3TC-I don't quite get this
14:34:28J3TC-Ah...I see the files updated :3
14:35:07Nico_PJ3TC-: its something for developers, not users
14:35:21 Join Buschel [0] (
14:35:57J3TC-Oh ok but I can still compile the latest code and have the album art support?
14:36:47Nico_Pwhoo! I did it!
14:36:51J3TC-Now let's see how many of maxwen's patches break :3
14:36:53 Join tb [0] (i=865fb382@gateway/web/cgi-irc/
14:37:14J3TC-Last compile I have is r15305 :3
14:37:21tbHi! Can I get write access to the wiki, please? Name is TillBayer
14:39:23amiconnmarkun: Hmm, the gigabeatf/x backlight fading is not configurable? Or am I missing something?
14:40:00BigBambitb: done
14:40:28 Nick bertrik_ is now known as bertrik (
14:40:40moosamiconn: it isn't, tha's right
14:40:40tbBigBambi: thanks!
14:40:47 Quit tb (Client Quit)
14:40:53amiconnmoos: Hmm, so it's always enabled?
14:41:08*amiconn would be annoyed by that fade-in then
14:41:23moosamiconn: plus the LCD power off option is under debug menu and then isn't a config saved...
14:41:32moosyup always enebled :(
14:42:01amiconnMost devs seem to not care about getting the low level working properly
14:42:08 Join PaulJam [0] (
14:42:23moosamiconn: and you don't have gigabeat :(
14:42:27mooshehe ;P
14:42:27amiconnSame goes for finishing features
14:43:53 Join MethoS- [0] (
14:48:00markunamiconn: it's not configurable. Didn't we talk about this already?
14:48:51amiconnI asked about the fading times, but I expected at least enable/disable
14:48:54moosmarkun: hi, do you know why the LCD power off option, isn't an option either?
14:49:06markunamiconn: ah, I didn't know it couldn't be disabled
14:49:42amiconnI greped for HAVE_BACKLIGHT_SET_FADING - doesn't occur on the app layer at all
14:49:47amiconnHence, no setting for it
14:50:12amiconnI am asking because I need to restructure backlight handling a bit
14:50:23amiconnThat change will touch all targets
14:50:24markunmoos: it made the remote control unresponsive and didn't save any battery, after moving is back to the debug menu nobody worked on it anymore
14:50:50moosmarkun: oh okay, missed this episod, thx for the info
14:51:49amiconnBasically, _backlight_on() and _backlight_off will be moved to target tree, and some other definitions shuffled around
14:51:58amiconn...and made more generic
14:52:29markunamiconn: thanks, because last time I looked it was a complete mess
14:52:36markun#ifdef hell
14:52:53amiconnA lot of ifdefs have to remain, unfortunately
14:52:54moosmarkun: pitty that's the F/X port seems to be "aborted"...
14:53:09amiconnBut this way I won't have to introduce yet another bunch
14:53:35 Join ompaulafk [0] (n=ompaul@gnewsense/friend/ompaul)
14:54:21markunmoos: what are the biggest problems you have with the port?
14:55:27 Quit ompaul (Nick collision from services.)
14:55:39moosmarkun: not really problem, just seems sort of "unfinished", I mean LCD things, fast restart, rom box... kkurbjun seems to enjoy with m:robe now
14:56:16 Quit JdGordon (Remote closed the connection)
14:56:23*moos still prefers his iaudio ;)
14:56:40markunwhat's the advantage? recording?
14:57:05*Nico_P misses RoLo
14:57:13moosradio, recording,...
14:57:40moosyeah Nico_P no RoLo yet (yet another thing that kkurbjun wanted to do)
14:58:39Nico_Plinuxstb: I have a patch to give AA to iCatcher on the gigabeat
14:58:50 Quit Frazz ("Leaving")
15:03:36moos_jhMikeS_: is your new batterie working well on your iaudio? (I'm thinking to buy one)
15:04:00 Part Llorean
15:04:04moosNico_P: is your patch somewhere for test
15:04:21Nico_Pmoos: not yet
15:04:32*moos is curious to see it on gigabeat
15:04:36moosok np
15:04:39Nico_PI'll probably upload it to the tracker
15:04:50moosok thanks
15:05:53***Saving seen data "./dancer.seen"
15:07:20_jhMikeS_moos: sure is...and I'm silently watching the conversation about dedicated follow-through on implementing features :)
15:07:27 Nick _jhMikeS_ is now known as jhMikeS (n=jethead7@rockbox/developer/jhMikeS)
15:07:51mooshehe, hiden ;)
15:08:21moosis all working well for your iaudio now? If so CONGRATULATIONS
15:09:45jhMikeSyeah. batt replacement is a bit of a pain with it being a soldering job and the location of the wires (the HD bracket in the way).
15:10:12moosjhMikeS: I'm still afraid to wait *months* if I send my iaudio back for the joystick...
15:10:44jhMikeSblah...yeah there's that little nasty thing
15:10:56moosaie then, I hoped that the batterie replacement could be less tedious
15:11:02jhMikeSas long as they don't break something else, I can't imagine it taking months
15:11:21 Quit pengo ()
15:11:53moosyeah, I did count the eventuallity to send them back again until they really fix things, they have really bad reputation
15:12:12 Join robin0800 [0] (
15:12:56 Quit robin0800 (Client Quit)
15:13:23Nico_Pmoos, linuxstb:
15:13:42moosNico_P: merci, I'll try it out...
15:15:55 Join robin0800 [0] (
15:17:58rasherNico_P: Getting some weird behavior.. When skipping to next a few times, quickly, it will play the first couple of second, then skip, and do that a bunch of times until finally ending with Undefined instruction at 00124274
15:18:11 Join Robin_0800 [0] (
15:18:34rasherWith a wps that displays albumart
15:18:46Nico_Prasher: could you test it with the rev before AA?
15:18:56 Quit Robin_0800 (Client Quit)
15:18:58Nico_Pand maybe with a WPS that doesn't have AA
15:19:06rasherYeah, I'll make sure it reliably does it first
15:19:31 Quit robin0800 (Client Quit)
15:19:36Nico_PI don't seem to get it here
15:19:47 Join robin0800 [0] (
15:20:08rasherand try with headphones, to see what's being played
15:20:49jhMikeSamiconn: is the only thing missing from the H10 tuner work the audio path setup?
15:20:58amiconnI don't think so
15:21:11amiconnIf it were, radio recording should work, but it doesn't
15:21:17jhMikeSyou said you could read and write it
15:21:19rasherNico_P: okay, seems pretty consistent.. I'll try without an aa wps first
15:21:31jhMikeSamiconn: did I miss and update? you got playback?
15:21:41amiconnYeah, I mean it's audio path setup, but not just the monitoring you mentioned
15:22:10amiconnIf it were the monitoring, fm recording should work even if normal fm playback doesn't
15:22:22barrywardelljhMikeS: I added the monitoring, but it still doesn't work
15:22:35jhMikeSso, sound at all
15:22:39amiconnBtw, I get a strange display when I try to record on H10 with source set to fm or line-in
15:22:44 Join jean [0] (
15:22:53amiconnIn both cases I get a constant reading on the peakmeter
15:22:57 Quit jean (Client Quit)
15:23:02rasherHrm, seems like automatic database update blocks any file operations until it's done (or something like that - I went into "show themes" and it didn't actually show them until the database update was done)
15:23:07amiconnThere should be no signal at all afaiu...
15:23:12jhMikeShmmm...that didn't happen before. was the line-in setup altered?
15:23:22 Quit ompaulafk (Client Quit)
15:23:30Nico_Prasher: thanks
15:23:34amiconnI can't actually test line-in as I don't have the required cable
15:23:58 Join jean [0] (
15:24:00rasherNico_P: it's not happening without AA, it seems
15:24:04 Quit jean (Read error: 104 (Connection reset by peer))
15:24:14 Join ackbahr [0] (
15:24:28barrywardellamiconn: I think line-in recording on h10 doesn't work, but can't test myself either
15:25:23rasherNico_P: eh.. now it's not happening with AA either. I guess I can't reproduce reliably. Dammit.
15:25:55karashataI could test line-in recording on the H10 if necessary, I have the cradle for it
15:26:16jhMikeSis the line-in mux known at all?
15:26:54barrywardellwe know the line-in/tuner switch
15:27:19 Join darkapostrophe [0] (
15:29:28 Join kingwen [0] (n=kingwen1@
15:32:11 Join MethoS-- [0] (
15:32:49 Quit ackbahr (Remote closed the connection)
15:32:55karashatanope, line in recording doesn't work
15:33:20jhMikeSmaybe the GPIOB2 is reversed?
15:35:02karashataon the other hand, the line-out does work
15:36:58 Quit kingwen (Read error: 104 (Connection reset by peer))
15:37:15desowinNico_P: yay, albumart commited
15:38:30*jhMikeS likes the idea and will use it but, oh boy :)
15:39:32Nico_PjhMikeS: what?
15:40:01jhMikeSjust wondering if all the MoB bugs were fixed first
15:40:24preglowjhMikeS: can we gut out much code when we eliminate codec swapping?
15:40:32Nico_PjhMikeS: seems quite stable now... but maybe AA will make new bugs show up :)
15:41:09 Quit Dark_Apostrophe (Read error: 110 (Connection timed out))
15:41:10 Nick darkapostrophe is now known as Dark_Apostrophe (
15:41:14jhMikeSpreglow: I think so...maybe even a whole thread.
15:41:47jhMikeSI think turn taking will need to be preserved when playing
15:42:39preglowwill be sweet
15:43:20jhMikeSmeh...maybe not the's really the swapping that complicates things...threading is the easy part of anything usually.
15:47:43jhMikeSit's finding that stooping one-liner bug that usually halts progress on stuff...which just happened. :p
15:47:56jhMikeSstoopid even
15:48:04TPM_Debianif i will run .mpg file, i got a error message "missing packet start code prefix: 00 at 1214" .. what does this mean?
15:48:32jhMikeSit means mpegplayer can't resync for crap, but my redo stops that sort of thing
15:49:09jhMikeSin SVN any packet error will make it barf
15:49:51jhMikeSreorganize it nicely. make it more robust against errors. have it setup to add whatever functionality.
15:51:45 Join darkapostrophe [0] (
15:51:52jhMikeSAbout half the test files I have won't play with the SVN version but play fine with the changed code.
15:53:42 Quit MethoS- (Read error: 101 (Network is unreachable))
15:53:53 Part pixelma
15:59:26kkurbjunamiconn, here now
16:02:58J3TC-Hrmm...the bmpresize patch is already broken :3
16:03:20J3TC-rc = read_bmp_file(gwps->state->id3->albumart_path, &temp_bitmap, data->img_buf_free, FORMAT_ANY|FORMAT_TRANSPARENT);
16:03:23J3TC-That doesn't exist anymore
16:03:40 Quit amiconn (Nick collision from services.)
16:03:41J3TC-In apps/gui/gwps-common.c
16:03:49 Join amiconn [0] (n=jens@rockbox/developer/amiconn)
16:04:20jhMikeSI thought it was committed?
16:05:32kkurbjunmoos, the gigabeat hasn't really stopped, I know at least I'm interested in continuing work on it, I just don't have time to work on two ports at once at the moment, plus I still haven't moved my debugging setup out of storage for the gigabeat. Once I have that out of storage I'll be able to work on the flash writing again. The fast reboot feature right now is a bit messy and leaves the player in a strange state. It really needs
16:05:41Nico_PjhMikeS: not the bmp resize patch
16:05:51linuxstbkkurbjun: Were you also working on rolo for the F/X ?
16:06:16jhMikeSNico_P: ah. and no WPS's were modded to use it?
16:06:29Nico_PjhMikeS: yes they were but they won't be broken
16:06:52J3TC-Really? Hrmmm
16:06:54kkurbjunlinuxstb: yes, but the iram that's documented in the manual doesn't seem to be accessable once rockbox is fully running
16:07:18kkurbjunI can access it if I freeze the player before the full startup takes place
16:07:21jhMikeSNico_P: my SVN up didn't reveal any wps stuff
16:07:34linuxstbkkurbjun: So does Rockbox do any memory remapping?
16:07:40Nico_PjhMikeS: that's weird because I changed it
16:07:45jhMikeSkkurbjun: I managed to write and read from 0x40000000
16:07:58Nico_PjhMikeS: ah I'm sorry I didn't understand your question
16:08:01kkurbjunbut once rockbox is fully booted, I can't write to it which I don't understand.. Yes it does, but that area should be mapped onto itself..
16:08:08kkurbjun:), maybe it does work then
16:08:09Nico_PjhMikeS: I didn't change any WPS in SVN
16:08:18mooskkurbjun: no problemo, that wasn't a complaint ;), but glad to hear you will resume your works in the future...
16:08:22 Quit Dark_Apostrophe (Success)
16:08:22 Nick darkapostrophe is now known as Dark_Apostrophe (
16:08:29jhMikeSNico_P: oh, ok...where's the modded one?
16:08:31kkurbjunjhmikes when did you try that?
16:08:47jhMikeSkkurbjun: on a whim not long ago
16:08:51Nico_PjhMikeS: I have a patch for iCatcher on gigabeat:
16:09:11jhMikeSNico_P: perfect...since that's what I use :)
16:09:19kkurbjunhmm, was it after I moved the mmu stuff so that it's common on the m:robe and gigabeat?
16:09:38J3TC-I shouldn't be looking at rockbox stuff
16:09:47J3TC-I have a paper to write about professional ethics
16:09:56J3TC-I should plagiarize =d
16:10:18linuxstbNico_P: I've just tried that iCatcher patch - did you try it with smaller AA? It seems a bit too big to me...
16:10:26kkurbjunI'll give it another try then, see if the linker scripts I was working on will accept it
16:10:49jhMikeSI tried the 0x4000000-4k and nothing though which is is depends on mode. I did it with raw addresses though.
16:10:50Nico_Plinuxstb: it'll work fine with smaller AA
16:11:25kkurbjundid you have the mmu disabled?
16:11:44jhMikeSnope, just as booted
16:12:26kkurbjunso you were wriitng to 0x4000000 to 0x4000000+4k?
16:12:34 Join Crash91 [0] (n=evil91@
16:12:55linuxstbNico_P: I also meant that the text is now too small - I would prefer bigger text and smaller AA, but maybe that's just me...
16:12:56jhMikeSjust 0x40000000....put the framebuffer there which it didn't like very much for too long but the image showed up
16:13:12Nico_Plinuxstb: ah... I find nimbus19 too big
16:13:39jhMikeSof course not a perfect image but it showed that data was written and read
16:14:43kkurbjunyep, that's what I would expect to see :)
16:15:17kkurbjunjhMikeS: I think the gigabeat could d some really fast updates if we just swapped the fram address and the screen pointer
16:15:24jhMikeSNico_P: so what happens if I use smaller AA? Freeze? :)
16:15:32Crash91Nico_P: Was AA comitted?
16:15:37Nico_PCrash91: yes
16:15:38kkurbjunI think the key is to wait for a sync signal before it's done
16:15:42jhMikeSkkurbjun: how would that work exactly?
16:15:44Crash91Nico_P: w00t
16:15:45Nico_PjhMikeS: what? no, it will be displayed
16:16:15linuxstbNico_P: It seems the album-art is drawn in front of the text - is that intended? Doesn't that mean that it's drawn with every WPS refresh, or does a WPS refresh always redraw the entire screen anyway?
16:16:58jhMikeSwaiting for a sync signal 1) Limits the number of updates so batching is needed. 2) Batching can starve an interrupt if that is used. Was not easy to solve at all.
16:17:20kkurbjunwait for the controller to be in an inactive area and then move the frame pointer to the lcd_framebuffer, at the same time time - how do you mean batching?
16:18:00jhMikeSa lot of things do lots of incremental rect bubbles for instance.
16:18:05Nico_Plinuxstb: hmm actually it might be drawn too often, yes
16:19:01jhMikeSand now that lcd_framebuffer is the current frame, when you move back to the driver framebuffer, what happened to the updates in lcd_framebuffer? they'll not be in sync anymore.
16:19:13kkurbjunjhMikeS: couldn't we just add a while loop in the lcd_update - I wonder how much it would end up affecting it
16:19:35Nico_Plinuxstb: I'll investigate further once I've finished with my playback fix
16:19:37jhMikeSkkurbjun: _Bad_ and using the interrupt is more efficient if you do that
16:20:29*barrywardell has a working fm tuner on his H10 :)
16:20:31jhMikeSit locks it to exacly 60 updates per second
16:20:42barrywardelljhMikeS: you were right, the mux bit was backwards!
16:20:51kkurbjunbad as in it would definately be slower - it would depend on how often we hit it at the top of the fram vs the bottom
16:21:01jhMikeSbarrywardell: wow...that was a lucky guess on my part :)
16:21:02kkurbjunbut that's all luck I suppose
16:21:51jhMikeSyes, totally random. what I'd rather do is provide a timing service to apps that might want sync.
16:23:09jhMikeSin that way, they can perform all the updates they want in the framebuffer, then tell the driver to do a swap or copy when frame sync interrupt happens.
16:23:10kkurbjunso when the service runs all the apps do the updates they need to?
16:23:16 Part Crash91
16:25:38jhMikeS1) update lcd_framebuffer 2) wait for the interrupt which performs the task of making the new data visible in some way or another.
16:26:09kkurbjunthat would be an interesting way to do it also - definately much more involved, how much of an impact do you think it would have? I guess the other side of it is how fast do we need updates to run in the core, mpegplayer already writes directly to the FRAME doesn't it?
16:26:15kkurbjundoom definately does
16:27:01kkurbjunit could end up reducing quite a few cycles on the frame updates though which gives more time to sleep
16:27:05jhMikeSit's the usual way it's done to have apps that expect to operate in a synced manner
16:28:45jhMikeSthe framebuffer avaiable to the controller can't be cached though. the current one isn't (it would appear much different if it were)
16:29:07kkurbjunoh true
16:29:36kkurbjunand the mmu can't do very fine sections right now
16:32:07 Quit BHSPitMonkey (Read error: 110 (Connection timed out))
16:32:28 Quit stewball ()
16:32:43jhMikeScan we do uncached aliasing like on PP?
16:33:09kkurbjununcached aliasing?
16:33:17jhMikeSbtw, yielding or sleeping isn't accurate enough for the timing so the interrupt must be used
16:33:51kkurbjunthere's other ways the data cache with the mmu could be setup too, we could do write through
16:34:03kkurbjunI don't knwo what the effects of that are on performance
16:34:35jhMikeSPP is setup so that all of RAM is mirrored starting at 0x10000000 but is not cached at addr+0x10000000
16:35:16kkurbjunoh, yeah, that would be doable - it might already be setup that way actually
16:35:31kkurbjun0x0 +32M is cached
16:35:32 Join lazka [0] (
16:35:43kkurbjunbut 0x3000000 is not I believe
16:35:56kkurbjunwith the mmu that is
16:36:00jhMikeSso reading/writing addr+0x30000000 is uncached?
16:36:09kkurbjunit should be
16:36:14jhMikeSall you need is to ensure line alignment of such buffers
16:37:21jhMikeSso I guess aligned and padded to 32-byte boundaries ensures that?
16:38:40kkurbjunline alignment of the FRAME addresses?
16:40:34kkurbjunwell we're writing shorts in so it could be 16 bit aligned right? I don't know the requirements of the FRAME address for the controller though
16:40:48jhMikeSimportant so a buffer boundary isn't in the middle of a cache line...on PP required so COP/CPU don't clobber each other's private data and also when using uncached addresses so the processor doesn't clobber it's own. On meg-fx, the 2nd applies.
16:41:43kkurbjunahh, I didn't realize that - I don't understand how the caches work on a deep level
16:42:23kkurbjundo you have a document that explains the cache lines and how they work?
16:42:25jhMikeSIf the buffer boundary is in the middle of a line, the cached line won't fully agree with the uncached address.
16:42:52jhMikeSit's just caching priciples in general. googling should be plenty informative.
16:43:23 Join johny1234 [0] (
16:45:12johny1234whats going on
16:45:26johny1234who Knows Heroes?
16:45:52 Join darkless [0] (
16:46:24 Part johny1234
16:46:29 Quit robin0800 (" HydraIRC -> <- The alternative IRC client")
16:49:04*jhMikeS sees AA which looks good but the graphics and title info are pushed down :\
16:50:29jhMikeSwoops, forgot the new BG :p
17:01:18preglowwhat profile aac can ipods decode? only lc?
17:02:09preglowbarrywardell: why not add him to credits if he did hard work?
17:02:18barrywardellhe's already there
17:02:52preglowour faad is set up to decode all kinds of weird aac files
17:02:53amiconnSomething is seriously borked on 1st/2nd gen
17:03:08preglowi don't think i've even seen anything but lc and lc+sbr
17:03:10amiconnIt crashes as soon as the hdd is powered of the first time
17:04:15amiconn...independent of where I am when that spindown happens. The way it crashes differs though. Sometimes it just freezes, sometimes I get data aborts etc, with addresses from all over th eplace
17:04:46Nico_Pbarrywardell: congrats on the commit :)
17:05:53kkurbjunamiconn: you were looking for me?
17:05:55***Saving seen data "./dancer.seen"
17:06:02amiconnkkurbjun: yes
17:06:16barrywardell:) przembh really deserves the credit though. If he hadn't figured out the i2c over gpio, then I would never have got anywhere with it
17:06:16amiconnIt would be nice if you could do a quick check on your mini 1st gen
17:06:28jhMikeSNico_P: ok, the progress bar is a couple pixels too low for the background so how I move things up a few px's? (I know crap about WPS format atm.)
17:06:32amiconnI am interested in the value of LCD1_CONTROL
17:06:46*Arathis is going to update his rockbox to test the fm tuner :3
17:06:53karashatabarrywardell: what are the chances that'll end up commited soon?
17:06:55Nico_PjhMikeS: in the pb tag
17:06:56kkurbjunI have hdd powerdown disabled on mine - I use CF
17:07:03Nico_PjhMikeS: I think the y value is the last one
17:07:06kkurbjunI can check that register though
17:07:07 Join Genre9mp3 [0] (
17:07:16jhMikeSthat seems to be the width
17:07:18barrywardellkarashata: what end up committed soon?
17:07:27karashataworking FM radio
17:07:32barrywardellit's already done
17:07:42amiconnbarrywardell: Why did you commit the ugly hack in power-h10.c instead of putting this selection where it belongs?
17:07:54karashataoh? where is it hiding then?
17:07:58amiconnThis gpio port is definitely not tuner power...
17:08:28amiconnOh, sorry
17:08:32Nico_PjhMikeS: :)
17:08:33*amiconn confused something
17:09:28 Join pixelma [0] (n=Marianne@rockbox/staff/pixelma)
17:09:33amiconnbarrywardell: I noticed that several keymappings on H10 need improvement....
17:09:36rasherI know I've asked before, but why doesn't e200 reset on usb-insertion?
17:09:55barrywardellkarashata: the build hasn't finished yet. maybe the build server has frozen again
17:10:14karashataoh, it only *just* got commited then, okay...
17:10:31karashatawas gonna say, the build I have doesn't have it yet
17:10:43barrywardellamiconn: yeah, I moved the gpio port out of power-h10.c before my commit
17:10:53barrywardellamiconn: which keymappings in particular?
17:11:18amiconnRecording screen, mpegplayer are some particularly bad examples
17:11:58jhMikeSNico_P: yeah, yeah...don't be smartass :). Thought you'd just have a quick fixeroo since you modified it.
17:12:17barrywardellah, ok. I never use them, which explains why I haven't noticed. the recording is bad indeed
17:12:34barrywardellkarashata: try downloading now ;)
17:12:38Nico_PjhMikeS: it's fine for me...
17:12:46karashatajust did, gonna try it out in a sec
17:14:49jhMikeSNico_P: yeah, but I use the 19-point font :)
17:15:37karashataouch, got a data abort trying to scan in reverse
17:15:45karashataat 40087B08
17:17:35karashatahmm, I'm gonna have to play with this...
17:18:28karashataI'll be back in a bit, gonna go get a decent signal and get some presets saved
17:18:44Genre9mp3Nico_P: congrats for the album art commit! :)
17:18:51Nico_Pthanks :)
17:18:57 Join Arathis2 [0] (
17:19:28amiconnkarashata: There's a large selection of presets in the wiki
17:20:28Genre9mp3Nico_P: It's pity thought that it's not so useful for all the landscape targets (in other words: it would be nice if we had scrolling margins in SVN, too)
17:20:47Genre9mp3^at least until the viewports implementation
17:21:26moosbarrywardell: you must have to update
17:21:42Nico_PGenre9mp3: yeah, but something might be possible with left margins
17:21:46amiconnbarrywardell: The fm screen should also have a remote button map: I didn't add one because I have no remote, and hence no idea how the buttons are arranged and what would make sense
17:21:50*jhMikeS is more than happy to add some content to the barren gigabeat screen landscape
17:22:10*barrywardell doesn't have a remote either
17:22:17markunNico_P: planning on updating the rockboxed theme for the gigabeat as well?
17:22:26Genre9mp3Nico_P: left margins?
17:22:27Nico_PjhMikeS: please post your 19px version of iCatcher with AA to the tracker ;)
17:22:52Nico_Pmarkun: not really because I don't use it, but maybe I'll get round to it
17:22:56 Quit SirFunk (Remote closed the connection)
17:23:30barrywardellI wonder if line-in recording is working now too?
17:23:52jhMikeSNico_P: sure, when I'm done fiddling around...but I also have it for small AA
17:24:12*Genre9mp3 would be interested in making AA versions of iCatcher for all colour targets
17:25:41jhMikeSok, it lines up now on the original background but need to get the format graphic back where it was
17:25:47markunNico_P: I do use it so maybe I should do it myself
17:26:16 Quit idnar (Nick collision from services.)
17:26:18 Join idnar_ [0] (i=mithrand@unaffiliated/idnar)
17:26:36barrywardellmoos: updated
17:27:35*moos just notices that iriver h10s aren't very expensive on ebay...
17:28:30jhMikeSah...much better...nice decent text size and a reasonable AA size
17:28:31 Quit amiconn (Nick collision from services.)
17:28:38karashatahmm, FM radio works nice
17:28:38 Join amiconn [0] (n=jens@rockbox/developer/amiconn)
17:29:02*pixelma would like to say that about c200 radio too :\
17:29:03*amiconn got a bluescreen
17:29:06Nico_Pbarrywardell: I think this also deserves a MajorChanges entry :)
17:30:06Genre9mp3Nico_P: I would like to rework this: for a possible inclusion in SVN
17:31:23 Join Mouser_X [0] (i=cf9bb003@gateway/web/cgi-irc/
17:32:33Nico_PGenre9mp3: linuxstb is working on a patch to support left scrolling margins, so your idea would be possible. also it would be nice to use the AA conditional to center the texts when there is no AA
17:33:41 Join Tavnos [0] (
17:33:41Genre9mp3Nico_P: then I'll have to wait for that I guess
17:33:47TavnosHi there
17:34:05TavnosHi Nico_P
17:34:22linuxstbGenre9mp3: I'm just investigating to see how intrusive it would be - the feature is already in the LCD driver, so it seems trivial to allow WPSs to use it, especially as it's more or less required for album art...
17:34:40jhMikeSNico_P: guess it's not quite proportioned how it should be but works for now =>">
17:35:02TavnosJust a little question about CVS changes... Does rockbox support now album arts natively Nico_P ?
17:35:11Nico_PTavnos: yes
17:35:21krazykitjhMikeS, that's pretty awesome.
17:35:35Nico_PjhMikeS: why is the first line of text so low though?
17:35:45linuxstbjhMikeS: How big is the AA bmp?
17:35:46Nico_Pyou could make the AA bigger
17:36:04TavnosNico_P :) Great News, I think I should update !
17:36:16 Quit Arathis (Read error: 110 (Connection timed out))
17:37:25amiconnlinuxstb: Why not start working on the right thing instead...
17:37:46linuxstbBecause this is a 30 minute job...
17:38:02kkurbjunjhMikeS: were you writing to 0x4000000 or 0x40000000
17:38:11Nico_Pamiconn: IMO left marhins can still be useful even with viewports
17:38:33amiconnlcd_set_margins() will be replaced by viewports
17:39:01Genre9mp3why not just go with scrolling margins _until_ the viewports implementation, anyway?
17:39:53amiconnBecause it will get in the way
17:39:55jhMikeSNico_P: just 75x75
17:40:24Genre9mp3jhMikeS: too small for Gigabeat screen IMO
17:40:28jhMikeSthat's what was on hand and I kind of liked it like that
17:40:37jhMikeSjust an icon size
17:41:10jhMikeSkkurbjun: whatever address the datasheet mentioned
17:41:18Genre9mp3amiconn: it would just get obsolete with viewports
17:41:18kkurbjunjhMikeS: could you look at this linker script and see if you see any problems, my gigabeat won't boot with it.
17:41:46amiconnGenre9mp3: It would become obsolete, but not "just"
17:41:55lostlogiccongrats, nico.
17:41:56amiconnIt will require extra work to remove it
17:42:09Nico_Plostlogic: :) I'm trying to track down another bug now
17:42:19Genre9mp3amiconn: I see
17:42:27lostlogicNico_P: who you hunting now?
17:42:28jhMikeSkkurbjun: because of all the other IRAM stuff? I think a special section is needed...BOOTSRAM_ATTR or something
17:42:28Nico_Pamiconn: I don't see why if it just uses the left margins that already exists
17:42:29markunamiconn: do you feel like working on viewports in the near future?
17:42:50amiconnI doesn't have to be me
17:43:00kkurbjunwhy another section?
17:43:05Nico_Pamiconn: then you should write down your concept :)
17:43:10amiconnThere's so much low level stuff that's half-baked
17:43:17amiconnI already explained my idea
17:43:34jhMikeSkkurbjun: what happens for all the various Ixxx_ATTR stuff throughout rockbox?
17:43:35Nico_Plostlogic: some weird thing that happens when a track fails to load, the next is skipped
17:43:58 Part Mouser_X
17:44:01lostlogicsounds fun
17:44:04Genre9mp3"I doesn't have to be me" << a golden quote I guess?
17:44:19Nico_Pamiconn: yes but finding that info would be a pain... maybe you don't have to do it but a spec should be written down
17:44:34 Quit mf0102 ("Verlassend")
17:44:46Nico_Plostlogic: seen my latest commit in playback.c?
17:44:52markunamiconn: did you write your explanation down somewhere?
17:45:19markunNico_P: oops, I guess your question answers mine :)
17:45:34lostlogicNico_P: in comment only, reading diff now
17:46:37 Join FOAD_ [0] (
17:48:05lostlogicNico_P: I think you should be able to use audio_finalise_track_change to push the runtime data onto the buffer?
17:48:20*jhMikeS did a folder skip on gigabeat and things failed to unmute once
17:48:31Nico_Plostlogic: indeed. I hadn't tought of it :p
17:48:41 Join darkapostrophe [0] (
17:48:57lostlogicNico_P: me neither until I saw the name of that function :-\
17:48:59kkurbjunjhMikeS: they are ignored if they are using teh C defines
17:49:23kkurbjunmemcpy, memmove, memset, etc are all in there
17:49:28kkurbjunand the yuv routines
17:49:45preglowam i the only one thinking that audio should stop immediately when you press prev/next?
17:49:45kkurbjunbecause they all use the direct attribute
17:50:08kkurbjunbut i dont see why having those few in there would be a problem as long as it fits
17:50:12 Join criznach [0] (
17:50:16 Join midgey [0] (
17:53:16 Join toens [0] (
17:53:19jhMikeSkkurbjun: you changed the crt0.S?
17:53:28 Part toens
17:53:33jhMikeSor that should work already?
17:54:11n1spreglow: i don't think so but a few people don't :-)
17:54:21kkurbjunthat should work already, it's using the same code for copying the iram as other targets to my knowledge
17:54:51n1sbtw I threw together a crude sinewave generating plugin but I need to improve my fixedpoint skills before it is usable
17:55:04 Join BHSPitLappy [0] (n=Me@unaffiliated/bhspitmonkey)
17:56:32jhMikeSkkurbjun: and 0x40000000??
17:56:33 Join Omega_ [0] (n=chatzill@
17:56:44 Quit Buschel ()
17:56:48 Nick Arathis2 is now known as Arathis (
17:57:15kkurbjunthat's what's in the ds
17:57:17TPM_Debiani think i found a bug in Rockbox ... if i show a video and my desplay-config use a invertet color the video will negative show
17:58:02Omega_Is there anyway to dual boot ipodlinux and rockbox?
17:58:58krazykitOmega_, the rockbox loader does this already. i would think the information will be in the manual or the wiki.
17:59:19 Quit Thundercloud (Remote closed the connection)
17:59:21kkurbjunis anyone around who was working on crt0.s for the gigabeat s?
17:59:35Omega_I know it dual boots the apple frimware and rockbox, but can it do ipodlinux aswel?
17:59:47krazykitOmega_, as i said, yes.
17:59:53jhMikeSkkurbjun: 0x40000000-0x1000?
18:00:19Omega_I will do some further research, thanks for your help.
18:00:29jhMikeSoh, fudge that
18:00:49jhMikeSwhat is OM[1:0] set to anyway?
18:01:00kkurbjunI dont' know, I think those are pins
18:02:49kkurbjunI can write to that address before rockbox starts though
18:02:50jhMikeSprint out BWSCON
18:03:20jhMikeSbits [2:1]
18:03:35preglown1s: i've already made such a plugin, i think
18:03:43 Quit FOAD (Read error: 110 (Connection timed out))
18:03:44 Nick FOAD_ is now known as FOAD (
18:04:01Arathisbarrywardell: fm tuner works really great on my h10/20gb :D
18:04:51barrywardellArathis: great!
18:05:00kkurbjunit should be 10
18:05:00 Quit Dark_Apostrophe (Connection timed out)
18:05:01 Nick darkapostrophe is now known as Dark_Apostrophe (
18:05:14kkurbjunI explicedly write that in the bootloader
18:05:50jhMikeSit says it's read only though and set by OM[1:0]
18:06:34jhMikeSthose particular bits
18:06:45kkurbjunoh yeah, haha, one second, let me recheck
18:07:02kkurbjunI may have pulled that from a register read
18:07:21kkurbjunthe bits then would be 01, but I need to recompile to verify that
18:08:02Arathisbarrywardell: just recording doesn't yet, but I think you know that already ;)
18:08:18 Quit daurnimator (Connection timed out)
18:08:33barrywardellhmmm. karashata's data abort is weird. 40087B08 isn't a valid address, is it?
18:08:59barrywardellArathis: it must be the same problem affecting line-in recording
18:09:33karashatabarrywardell: haven't been able to reproduce it, I think it was a fluke
18:09:52jhMikeSthe highest IRAM address should be < 0x4000c000 right?
18:10:31kkurbjunit's only 4k right
18:10:40jhMikeSwas talking about PP :) sorry for confusion
18:10:56barrywardelljhMikeS: exactly. that's what confuses me
18:11:33jhMikeSunless there's a nice cache line clobber happening somehow but I can't imagine what
18:12:32 Join ompaulafk [0] (n=ompaul@gnewsense/friend/ompaul)
18:14:15Omega_Would you have to replace the rockbox loader with a other one to be able to dual boot with ipodlinux?
18:14:49jhMikeSbarrywardell: probably would be prudent to 32-bytes align the interrupt stacks at this point
18:15:15jhMikeSso .balign 32 in crt0-pp.S
18:15:36jhMikeSit's 16 now but I'm honestly not sure it's enough
18:16:35barrywardellis there much negative effect from 32-byte aligning?
18:17:15jhMikeSand the COP shouldn't be using them right now either. of all the things...the cache stuff between CPU and COP can cause weirdness
18:17:16 Join goffa [0] (n=goffa@
18:17:50jhMikeSup to 31 bytes thrown out but the stacks themselves are multiples of 32-bytes so only the start needs aligning
18:19:30jhMikeSI also can't say for sure if the COP could be picking up on something that belongs to the CPU while it's sitting in shouldn't...idle stacks are also non-cached.
18:19:33kkurbjunjhMikeS: BWSCON: 01055102
18:21:09amiconnkkurbjun: Did you check LCD1_CONTROL on your mini?
18:21:21goffaso.. anyone attempting to get the ipod classic going? 160gb would be nice
18:21:40jhMikeSdata bus with = 16bit then
18:21:57kkurbjunamiconn: not yet, I don't have a cable around
18:22:09barrywardelljhMikeS: 31 bytes isn't much if it's gonna improve stability
18:22:30kkurbjunthat would fit with the flash width
18:23:46jhMikeSany chance the cache could be trying to cache SRAM? (entertaining crazy notions I know :)
18:24:39amiconnI don't think th ecurrent pp instability is a dc problem
18:24:51amiconnPP5002 is *incredibly unstable atm
18:25:00jhMikeSthis came from a PP5002?
18:25:10 Join linuxstb_ [0] (n=linuxstb@rockbox/developer/linuxstb)
18:25:48*jhMikeS goes "wtf?"
18:26:12 Quit linuxstb (Nick collision from services.)
18:26:14 Nick linuxstb_ is now known as linuxstb (n=linuxstb@rockbox/developer/linuxstb)
18:26:54Nico_Pis there a reason why I can't enter aac.c in gdb?
18:27:18Nico_PI get error messages about _temp_codec0.dll when playing an AAc file
18:27:30*jhMikeS also wonders where all the new instability came from when a short time ago he couldn't crash the H10 at all
18:27:43amiconnIt would be really nice if someone with a PP5002 target could test a current build
18:27:54amiconnOn 2nd Gen, rockbox is completely unusable atm
18:28:08Nico_Pcould it be my fault?
18:28:11jhMikeS3g is enough or specifically that?
18:28:57amiconnData aborts, undefined instructions, stkov... the whole thing, with seemingly random addresses
18:29:42linuxstbThere have been a few similar reports here in IRC recently of people with PP5002 ipods
18:29:47amiconnHappens everytime after boot, either (with dircache disabled) immediately or (with dircache enabled) at the end of the dircache scan
18:30:16jhMikeSabout when did it start?
18:30:37amiconnNot sure... I noticed it today, but not in my first tests
18:30:56amiconnI did 'svn up' somewhere in between
18:31:16amiconnIt started before the AA commit
18:31:26amiconn..and most probably not before last weekend
18:33:05 Join kfazz [0] (
18:35:48kfazzhas anyone been able to get zagor's patch to apply correctly? i cant, even on the revision in the patch
18:40:37*preglow wonders why faad has to have so many hacks to make it compile at all
18:42:41 Join Lear [0] (i=chatzill@rockbox/developer/lear)
18:43:30jhMikeS3g seems fine with the 10th daily
18:43:38barrywardellamiconn: have you had a chance to look at the pp lcd registers?
18:43:55LearNico_P: Aware of the sim warning "bad file name lacks slash"? From album art...
18:44:18Nico_PLear: no
18:44:53Nico_PLear: I don't think I ever saw it
18:45:11*jhMikeS doesn't understand why rocklife is a game instead of a demo
18:45:12 Join linuxstb_ [0] (n=linuxstb@rockbox/developer/linuxstb)
18:45:42Nico_PLear, preglow: would you happen to know why I get error messages like "Error while mapping shared library sections: "archos/_temp_codec0.dll": not in executable format: File format not recognized."?
18:45:54 Quit linuxstb (Read error: 104 (Connection reset by peer))
18:45:55LearNico_P: Seems to come if you play a file in the root...
18:46:07Nico_PLear: ah, very possible
18:46:11LearNico_P: Which OS?
18:46:17Nico_PLear: linux
18:46:32Nico_Pthe aac codec seems to work but I can't enter it with gdb
18:46:33LearI would've guessed Windows from the filename. :)
18:47:23LearNico_P: If it is AAC, probably because the sim codec is larger than a buffer in codecs.c.
18:47:44LearI get problems on Windows too.
18:47:49 Nick thegeek_ is now known as thegeek (
18:47:56 Quit Omega_ ("ChatZilla [Firefox]")
18:48:13Nico_PLear: ah, I'll try increasing the buffer, thanks
18:49:05Nico_Phmm I doubled the size but still have the same prob
18:49:31Nico_Poh no I'm stupid
18:49:38LearYou need to change the CODEC_SIZE define too, at least within codecs.c.
18:52:01Nico_Pah, nice :)
18:56:32 Join Frazz [0] (
18:56:47jhMikeSbarrywardell: audiohw_set_monitor shouldn't turn the DAC off of mess with the sidetone settings
18:57:13barrywardelljhMikeS: no? the wm8731 datasheet suggested it should...
18:57:26jhMikeSyou won't hear voice with the dac off
18:57:40 Quit ompaulafk (Client Quit)
18:58:00barrywardellpage 31
18:58:06 Join ompaul [0] (n=ompaul@gnewsense/friend/ompaul)
18:58:22barrywardellah, ok. I'll see if it's still ok when I leave them on
18:58:43jhMikeSit should be. sidetone will only be turned on by recording anyway
18:58:52 Join mf0102 [0] (n=michi@
18:59:37 Nick linuxstb_ is now known as linuxstb (n=linuxstb@rockbox/developer/linuxstb)
18:59:59jhMikeSI think it should also be audiohw_enable_recording(recording); in the radio case
19:00:14 Quit J3TC- (".•«UPP»•.")
19:02:20barrywardellI don't think that's the case. We need to always select line-in
19:03:00*barrywardell thinks the argument to audiohw_enable_recording() is deceptive
19:03:09 Join Domonoky_ [0] (n=Domonoky@
19:03:29jhMikeSah, yeah. ;p
19:04:05Nico_Pkkurbjun: is there a chance of having RoLo soon in light of what you committed and your earlier discussion?
19:04:05jhMikeSthat's why I hate bool arguments
19:05:50kkurbjunNico_P: I just seperated them because crt0 was getting to be a mess to work with all the gigabeat s #ifdefs. I still don't think that the sram is working as I would expect it to, and I'm not sure what's causing it though.
19:05:58***Saving seen data "./dancer.seen"
19:06:15jhMikeSbut...audiohw_disable_recording() is missing for when the recording flag is false
19:06:41 Join J3TC- [0] (
19:07:14kkurbjunyou could also do it without sram, and that would probably be easier now with the lds separated also
19:08:10Nico_Pkkurbjun: me? I don't think so :p
19:08:29barrywardelljhMikeS: but that would disable line-in
19:08:44billeniumi am restarting my computer
19:08:59Nico_Plostlogic: ping
19:09:21lostlogicNico_P: quack, what'd I not do?
19:09:33jhMikeSyeah, it should do audiohw_disable_recording, then audiohw_set_monitor() immediately after
19:09:58 Quit Genre9mp3 ()
19:10:04 Quit billenium (Remote closed the connection)
19:10:14 Join przemhb [0] (
19:10:19Nico_Plostlogic: nothing, don't worry :) from my investigation I think the block at playback.c:1992 should be rempved
19:10:26Nico_Por removed too
19:10:33przemhbHi all
19:10:48Nico_Plostlogic: I can't see any negative side effects but maybe you know more than me about that part
19:10:57barrywardellah, right. and audiohw_set_monitor() should do the power config too?
19:11:14przemhbgreat job with the H10 tuner guys!
19:11:42*jhMikeS gets an error about LANG_FMR
19:11:45TPM_Debianwhy does my file-speech does't work? ... need i a plugin to work it?
19:11:45przemhbI'm building the code now
19:12:34jhMikeSbarrywardell: the mute stuff and such? yes...totally independent.
19:12:39pixelmajhMikeS: after a make clean/reconfigure
19:13:22jhMikeSI did a reconfig...will make clean and try once more
19:13:40pixelmashould have been a questions... I remember the same errors when c200 radio support got in...
19:14:24lostlogicNico_P: That may not be the right call to make any more, but if you don't do something there and new_track is set then you are potentially repeating the same track
19:14:29jhMikeSit's ok now...the reconfig wasn't enough
19:14:42 Quit lazka (Remote closed the connection)
19:14:53lostlogicNico_P: at least that's what it was meant for −− but in the MoB world that might be different
19:15:18lostlogicNico_P: something to do with rapid skipping I think
19:15:24Nico_Plostlogic: on codec failure it results on skipping twice. once in the codec, and once there
19:16:02lostlogicNico_P: anytime a skip has 'happened' new_track should be reset to prevent that.
19:16:25Nico_Pwell in the current code this bit gets executed even if new_track == 0
19:16:45lostlogicahhh, I see
19:16:54lostlogicmaybe just needs a conditional modification
19:17:12Nico_Pyeah I think so... removing it might prevent regular skips... I'm checking
19:17:33lostlogicit would most likely only affect certain cases of rapid skipping
19:17:41jhMikeSbarrywardell: I'll muck about and check every possible thing...should only need write_or/and for that
19:17:54Nico_Pyeah regular works too with it disabled
19:17:56przemhbI have a question: could Leszek Psuty name be placed in the credits? All the hardware investigetion concerning H10's tuner was done by me and Leszek Psuty on his player.
19:18:09 Join random_desu_is_s [0] (
19:18:16barrywardelljhMikeS: i'm working on it now. I'll have a patch for you to look at shortly...
19:18:32lostlogicbut if the codec exits with an error, I"m still not clear on how the track it was playing gets skipped over so that we don't end up in an infinite loop of playing the same broken track.
19:18:32 Quit tywickra (Remote closed the connection)
19:18:42 Part random_desu_is_s
19:18:45barrywardellprzemhb: sure. I'll do it in a few minutes
19:19:00Nico_Plostlogic: the codec always requests a new track
19:19:03przemhbbarrywardell: thanks
19:19:17lostlogicNico_P: how can we be sure of that if it exits with an error...
19:19:36Nico_Plostlogic: the only case where we'd end up in a loop of trying to play a broken track is if the track is alone on a playlist with repeat enabled
19:19:47 Join iamben_ [0] (
19:19:47Nico_Pbut it is already the case in the current code
19:20:07lostlogicNico_P: how is the codec requesting a new track if it's exiting with an error
19:20:09lostlogicthat doesn't make sense
19:20:20lostlogicif it exits with error we have no idea what state we've been left in and should move to the next track
19:20:31Nico_Plostlogic: it sets an error var, then asks for a new track to see if it must stay active
19:20:40Nico_Plook at mpa.c or aac.c
19:20:42 Quit Domonoky (Read error: 110 (Connection timed out))
19:21:00jhMikeSbarrywardell: ok. I just threw in the changes myself and made a build. :p
19:21:16barrywardellcool. i'll leave it to you in that case ;)
19:21:38barrywardellcan you add Leszek Psuty to the CREDITS when you're committing it?
19:22:10amiconnjhMikeS: Does your 3rd gen crash with current rockbox?
19:22:24jhMikeSamiconn: I tried a 10th daily and had no trouble
19:22:38amiconnHmm, and today?
19:22:59lostlogicNico_P: what about the early return? That would not trigger a track skip and would keep tryign to run a broken codec
19:23:00jhMikeSI couldn't download the 11th...I'd have to build one
19:23:24lostlogicNico_P: perhaps we need another return code from codecs, but that block must stay to prevent looping forever in some form.
19:23:25Nico_Plostlogic: true
19:24:19Nico_Plostlogic: but how did this use to work?
19:26:06lostlogicNico_P: uncertain −− I think the thing to do is something like "if (ci->request_next_track()) goto next track; else return STREAM_ERROR;" where STREAM_ERROR is handled differently than CODEC_ERROR
19:26:43amiconnjhMikeS: A daily from the 10th works for me too
19:26:52lostlogichmm, or something... I dunno...
19:27:05amiconntrying 11th now
19:27:15lostlogicNico_P: it's possible we'd skip two tracks on error before too :-P \
19:27:41Nico_Plostlogic: a solution that doesn't involve changing all the codecs would please me more
19:28:00Nico_Pbut that one still does sounds good
19:28:20Nico_Pamiconn: have you seen linuxstb left scroll margin patch?
19:28:23lostlogicnothign is coming to mind in that realm unfortunately ;)
19:28:33amiconn11th works too... weird
19:30:02amiconnSO it looks like AA is the culprit...
19:30:57 Join tywickra [0] (i=tywickra@nat/redhat-in/x-a4d466a836e2d15e)
19:31:06tywickraGuys.. need an urgent help
19:31:22 Quit iamben (Read error: 110 (Connection timed out))
19:31:24tywickrahow can i format my iPod on FC 8 ?
19:31:36lostlogictywickra: mkdosfs /dev/ipoddevice
19:31:48tywickrathat's it ?
19:31:52linuxstbtywickra: You _don't_ want to format it...
19:31:56lostlogic... yes...
19:32:01tywickrai want to format it
19:32:08lostlogiclinuxstb: hey, I just answer the questions as asked :-D
19:32:18linuxstblostlogic: ;)
19:32:54linuxstbtywickra: What's the problem you want to fix?
19:33:02tywickra[root@localhost ~]# mkdosfs /dev/sdb1
19:33:03tywickramkdosfs 2.11 (12 Mar 2005)
19:33:03tywickra[root@localhost ~]# mount /dev/sdb1 /mnt/
19:33:03DBUGEnqueued KICK tywickra
19:33:03tywickramount: /dev/sdb1: can't read superblock
19:33:05lostlogictywickra: note: there are two partitions on the ipod, the firmware and the data... you almost assuredly don't want to format the firmware partition and only maybe want to format the data one
19:33:31linuxstbtywickra: That's the firmware partition - you didn't want to format it...
19:33:33lostlogictywickra: see now you've gone and formatted the firmware partition
19:33:45Nico_Pamiconn: the crashes are due to AA??
19:34:02lostlogictywickra: don't worry though, it's not a brick... permanently
19:34:13linuxstbtywickra: You may want to follow these instructions -
19:34:33tywickraah i get the ipod menu and all
19:34:51tywickraok .. sure .. will follow the instructions..
19:35:07linuxstbtywickra: Wait, so your ipod is still working?
19:35:12tywickrabut how can i overcome this problem , which i'm getting when i try to mount the ipod ?
19:35:25linuxstbYou should mount the second partition - mount /dev/sdb2 /mnt
19:35:26tywickrayeah ipod is still working .. but can't mount on FC 8
19:35:48amiconnNow a current build works too ?!???
19:36:34tywickralinuxstb, still i'm getting the error : mount: /dev/sdb1: can't read superblock
19:36:49lostlogictywickra: ... sdb1 != sdb2
19:37:30Calciferhey, yesterday I had to go to bed while ddalton was helping me compile rb
19:37:50tywickralostlogic, [root@localhost ~]# mount /dev/sdb2 /mnt
19:37:50tywickramount: special device /dev/sdb2 does not exist
19:38:11Calciferthe last thing I did was "make" and it went through a long process and created some files and folders
19:38:21Calciferbut how do I use them?
19:38:36lostlogictywickra: what model ipod is this? are you sure the ipod is on sdb? have you somehow messed with the partition table on the ipod?
19:38:45karashataCalcifer: are you compiling for a device or a sim?
19:39:01linuxstbtywickra: Can you type "fdisk -l > my_disks.txt" and then upload the my_disks.txt file to
19:39:16tywickralinuxstb, sure
19:39:24karashatathen you'll want to do "make zip" and extract the contents of the zip file it makes
19:39:37lostlogiclinuxstb: nice play, I'll hafta store it in my playbook.
19:39:42karashataer, extract the contents to your device
19:39:48Calciferwill it take the same amount of time as "make"
19:40:04krazykitit should be far faster
19:41:16Calciferhave any of you used the cygwin voice generation method using sapi?
19:41:32Nico_Plostlogic: about the runtime data gathering, shouldn't we call the "unbuffer" callback on track change? because one could play a track many times between it being buffered and it being unbuffered
19:41:49 Join Thundercloud [0] (
19:41:52CalciferI wanted to figure out how to specify what voice was being used, under sapi options, but I don't know the switches
19:41:57tywickralinuxstb, yes.. i've uploaded it.. name of the file is tywickra_my_disks.txt
19:42:29linuxstbDid it give you a link?
19:42:30tywickralostlogic, it's a 4th Gen ipod, 80gb color one
19:42:43lostlogicNico_P: I don't know −− I hate/don't care about/think are stupid runtime data gathering/scrobbler ;)
19:42:46CtcpIgnored 2 channel CTCP requests in 8 seconds at the last flood
19:42:46*jhMikeS left some wigged-out weirdness in audiohw_enable_recording :\
19:43:11lostlogic80gb color... they made such a thing?
19:43:38linuxstbIt's not an "ipod classic" ?
19:43:43Nico_Plostlogic: hehe... I have roughly the same opinion
19:43:45tywickralostlogic, lol.. no idea .. it's a 4th gen ipod
19:43:59tywickrawith 80gb
19:44:35linuxstbtywickra: Assuming it's a 4th gen ipod, then that's not the correct partition table - I'm surprised it's still working normally....
19:45:00tywickralinuxstb, what am i suppose to do now ?
19:45:09lostlogiconly 5.5th generation and classic come in 80gig models. If it's a 5.5th generation then that's the wrong partition table, if it's a classic, then rockbox doesn't support it.
19:45:27CalciferI need help using the "make voice" command, I would like to know the sapi options
19:45:43linuxstbtywickra: If you type "fdisk -l /dev/sdb", does it display the message about "Note: sector size is 4096 (not 512)" ?
19:46:04tywickralostlogic, linuxstb this is what it says .. MB 147ZP/A iPod classic 80GB
19:46:18BigBambiSo it is a classic then
19:46:21lostlogictywickra: rockbox does not support that model, we have no help for you here.
19:46:25 Join Rondom [0] (
19:46:25BigBambiWhich rockbox doesn't support
19:46:26 Join StarMan [0] (i=4cb431de@gateway/web/cgi-irc/
19:46:40linuxstbAnd I've no idea what the correct partition table should look like for the classic...
19:46:47 Join PaulJam_ [0] (
19:47:17tywickralinuxstb, yes.. it does display that message
19:47:26StarManMy iRiver H10 20gb freezes when scanning for channels using the new FM Radio feature
19:47:28tywickralinuxstb, [root@localhost ~]# fdisk -l /dev/sdb
19:47:28tywickraNote: sector size is 4096 (not 512)
19:47:31StarManjust thought I'd let someone know
19:48:36tywickralostlogic, oh .. but one of my friends using RockBox on the same
19:48:43lostlogictywickra: no they aren't.
19:48:47BigBambitywickra: No he doesn't
19:48:55tywickralinuxstb, what am i suppose to do now ?
19:48:55lostlogicBigBambi: highfive!
19:49:09tywickralostlogic, oh
19:49:15 Quit StarMan (Client Quit)
19:49:20 Quit chris__ (Remote closed the connection)
19:49:39 Quit PaulJam (
19:49:39 Quit Seed (
19:49:39 Quit jhMikeS (
19:49:39 Quit animeloe (
19:49:39 Quit sd__ (
19:49:53 Nick PaulJam_ is now known as PaulJam (
19:52:25tywickralinuxstb, so to restore things back to normal.. i have to plug it into a windoez or mac box.. isn't it ?
19:52:26BigBambitywickra: Sory, that link I gave you was garbled, try:
19:52:56NJoinSeed [0] (
19:52:56NJoinanimeloe [0] (n=animeloe@unaffiliated/animeloe)
19:52:56NJoinjhMikeS [0] (n=jethead7@rockbox/developer/jhMikeS)
19:52:56NJoinsd__ [0] (n=sd@
19:52:57 Quit sd__ (Read error: 104 (Connection reset by peer))
19:52:57 Join sd [0] (n=sd@
19:53:11tywickraBigBambi, it's confusing .. on the Invoice it says iPod classic 80GB *3rd* generation
19:53:23BigBambiThe importnat word here is classic
19:53:28 Quit Seed (Nick collision from services.)
19:53:34 Join Seedy [0] (
19:53:39lostlogictywickra: <−− use it, love it.
19:53:57BigBambiI don't know what the third gen there means, but the classic is the 6th gen of 'big' ipod
19:54:40 Join stripwax [0] (
19:55:12stripwaxNico_P - great job on putting albumart on buffer
19:55:20Nico_Pthanks :)
19:57:19stripwaxNico_P - seeing as you used the %C and %Cl tags am I right in saying WPS written for the albumart patch would now work on official rockbox build?
19:57:21 Quit maddler (Connection timed out)
19:57:31Nico_Pstripwax: yes
19:57:41stripwaxNico_P - well you are just the dude
19:58:09Nico_P:) It would have been silly of me to break compatibility when the tags are perfectly fine
19:59:07tywickralostlogic, well... it's a 5th Generation iPod
19:59:11stripwaxNico_P - is there an announcement in the WPS or Unofficial builds? would be worth stickying
19:59:24tywickralostlogic, thanks for the cool page ..
19:59:29Nico_Pstripwax: no idea
20:00:29 Quit merbanan ("Leaving")
20:01:54tywickralostlogic, BigBambi actually i bought this ipod focused on RockBox
20:02:08tywickralostlogic, can i still give a try ?
20:02:17Soaptywickra: why do you say it is a fifth generation iPod?
20:02:33 Quit amiconn (Nick collision from services.)
20:02:38SoapWhat is the front case material? Plastic or matte-finished metal?
20:02:42 Join amiconn [0] (n=jens@rockbox/developer/amiconn)
20:02:50lostlogictywickra: no, rockbox does not work on the classic.
20:02:59stripwaxNico_P - ok, I just added a quick post to the WPS forum. Great stuff
20:03:35Soaptywickra: Classic (as BigBambi mentioned ten minutes ago) = 6th generation.
20:03:44tywickralostlogic, are you guys working on classics ?? to make it enable ?
20:04:26tywickraSoap, ah.. means it's a 6th Gen
20:04:37stripwaxtywickra - nobody is working on rockbox for the IPod Classic
20:04:43lostlogictywickra: not exactly. the firmware is encrypted and nobody has a good way to break the encryption to enable hacks
20:05:02Soap(or a bad way) ;)
20:05:24tywickralostlogic, ah.. means i have hit the brick
20:05:35lostlogictywickra: brick? no, just hafta use the ipod firmware on it ;)
20:05:56BigBambilostlogic: Same thing :)
20:06:23Calcifercan anyone tell me if these instructions work under cygwin, using the "make voice" command, I want to specify what voice will be used using sapi5
20:06:24tywickralostlogic, which means .. that we can't be sure of a RockBox release for this version of iPod .. isn't it ?
20:06:26lostlogicdepends... fo rme yes, I have 80% of my music in formats that ipod OF doesn't play... but most people have mp3s
20:06:49lostlogictywickra: correct, there is no guarantee that rockbox or any other alternative firmware will come out on the classic.
20:07:02lostlogictywickra: if you want it to happen, I suggest learning cryptography and breaking the firmware encryption.
20:07:04BigBambitywickra: If somone breaks the encryption then works out the new, undocumented hardware in it, sure it is possible
20:07:11CalciferI tried something along those lines after using the listvoices.vbs to see what the voices were called
20:07:36BigBambitywickra: But it isn't being worked on by anyone we know of, so don't hold your breath.
20:08:14 Join billytwowilly [0] (
20:08:43lostlogicor maybe apple will stop being retards and just release specs. hmm, maybe not.
20:09:22tywickraBigBambi, ok .. i got the point.. so the bottom line is i've got to stick to this ugly ipod for some time it seems
20:09:41BigBambiqute some time if not always I would have thought
20:09:48 Quit billytwowilly (Remote closed the connection)
20:10:17tywickraBigBambi, are you a contributor for RockBox ? :)
20:10:26CalciferI get a "sed: -expression #55, Char 21: Unknown option to 's'
20:11:01Calciferso my syntax is wrong, but where do I find out the right way to input the options?
20:11:02BigBambitywickra: Only in support
20:11:26tywickraBigBambi, ok . ... is someone working on iPod classic ? :)
20:11:28BigBambiI'm not much of a programmer (I dabble a little, but not to Rockbox standards)
20:11:45BigBambitywickra: I quote myself "<BigBambi> tywickra: But it isn't being worked on by anyone we know of, so don't hold your breath."
20:12:01stripwaxCalcifer - what exactly is your command line
20:12:03tywickraBigBambi, oops.. yeah .. you did say that
20:12:28stripwaxi.e. when you say you're doing 'something' along these lines, what exactly are you doing?
20:13:48Calciferits using /tools/configure , make for e200 using advanced, voice, english, sapi, then for sapi options I want to specify the voice to be used
20:14:13stripwaxwhen do you get the sed error?
20:14:18tywickraBigBambi, let's hope that some will come out with the good soon .. to rescue us
20:14:58CalciferI used listvoices.vbs to tell me the voice I want to use is Land= 409;9, Voice= VW Kate, Gender= Female
20:15:16BigBambitywickra: Well, personally I don't have any ipod so I'm not too bothered, but it is always nice to support more targets
20:15:39Calciferwell after I try inputing some options for sapi and trying to have it create a make file
20:15:48tywickraBigBambi, cool.. sounds encouraging
20:15:52Calciferso pretty much at the very end of the configure process
20:16:09 Quit moos (Read error: 104 (Connection reset by peer))
20:16:19stripwaxCalcifer - so at the point when configure asks you "Enter sapi options", what did you type in ?
20:16:22 Join moos [0] (
20:16:36Calciferbtw, since I'm new to this, what do you call the configure process, is it configuring what will be compiled
20:17:12CalciferI typed 409;9, VW Kate, Female
20:17:39 Quit Thundercloud (Read error: 104 (Connection reset by peer))
20:17:56BigBambiCalcifer: Voice generation or online dating? :)
20:18:21stripwaxCalcifer - what happens if you just type instead Land= 409;9, Voice= VW Kate, Gender= Female
20:18:43Calciferlets look and see
20:18:51Nico_Pkkurbjun: "This lays the groundwork to get rolo running." whoo \o/
20:18:54stripwaxshould work I think
20:19:48lostlogicBigBambi: that's what I thought when I saw that line :-\
20:20:17tywickralostlogic, linuxstb
20:20:27tywickrait seems there's a workaround for this ..
20:20:32BigBambitywickra: Irrelevent to a rockbox port
20:20:41tywickraoh :(
20:20:49lostlogictywickra: that's just the database crypto, not the firmware
20:20:58BigBambiThat is just for generating the iTunes database, nothinh to do with the firmware encryption
20:21:00 Join billytwowilly [0] (
20:21:08Calciferalmost exact same message except now its char 27
20:21:24tywickralostlogic, oops..
20:21:56CalciferI get a "sed: -expression #55, Char 27: Unknown option to 's'
20:22:43Calciferthat section should really be clarified
20:24:22Calciferwhen I leave it to defaults it uses the same voice, which is not the first voice listed, or the default sapi voice I have set
20:29:02Calciferuh oh, something looks really off
20:30:11stripwaxCalcifer- looks like you cannot enter any comma in the sapi options so this must not be the correct way to specify them
20:31:05stripwaxWhen you get that sed error it is because of punctuation in the tts options. This results in a malformed makefile - you won't be able to "make voice"
20:31:18Calciferwhen I ran make voice this time, at the point where it starts to generate the clips, it lists the perameters, and usually for target it just lists e200, instead this time e200 is followed by :backlight_brightness:flashstorage:hold_button: ect
20:31:26stripwaxI've never tried voicefiles so not sure what the correct args should be
20:32:20 Quit karashata ("Leaving.")
20:32:55Calciferwell I've successfully created one that atleast read a few items correctly, and when I had, I only saw Target: e200
20:33:06Calcifernot that really long mess of stuff listed
20:33:19Calciferso something new is wrong
20:33:36 Join karashata [0] (
20:34:29 Part pixelma
20:38:24stripwaxCalcifer - So it looks like 1) you should not specify LANG on the options, 2) you should specify the voice using /voice:VW Kate . Does that work for you?
20:38:58Calciferlets see
20:40:06stripwaxI don't have that voice file so I can't test that myself unfortunately, but I'll see if I can get a regular voice build working in a similar way
20:40:49Calciferhey, got no strange message
20:41:10 Quit HellDragon (Connection reset by peer)
20:41:34Calciferso now I went ahead and ran make voice
20:42:14 Join Weiss [0] (
20:42:46Calcifergot a no matching voice for english vw message
20:43:28stripwaxBah, try instead with /voice:"VW Kate"
20:43:34 Join Arathis2 [0] (
20:43:48stripwaxit might be that the sapi_voice tool doesn't like options that have spaces in them
20:43:58 Quit Arathis (Nick collision from services.)
20:44:05 Nick Arathis2 is now known as Arathis (
20:45:00Calcifersounds like it since it only said vw, not vw kate, also when I ran listvoices.vbs, it came up with lang as 409;9 instead of simply english, which is the language
20:46:27Calciferwe'll see in a second if it worked
20:47:00Calcifersame error
20:47:17CalciferI saw a voice listed as dave
20:47:41CalciferI can try that since it doesn't include spaces, and trust me I don't want "Dave" as my voice
20:48:33 Join Daniel_S [0] (i=57b0d4b4@gateway/web/cgi-irc/
20:49:25stripwaxCalcifer - yes, mine reported the language as "409;9" too, but don't forget - you specify the language when asked by ../tools/configure
20:49:51Calciferso could that be the problem?
20:50:08Calciferor does it know that value as english anyway
20:50:28Calciferaww, son of a bitch lol, Dave is working
20:51:20stripwaxCalcifer - the language isn't a problem - looks like the space in the voice name is.
20:51:26Calciferso yeah, I guess you were right stripwax, must be it doesn't like names with spaces
20:51:46Calciferwhich means either someone can edit that and fix the issue, or I can somehow rename my voices?
20:52:16 Quit iamben_ (Read error: 104 (Connection reset by peer))
20:54:07Calciferone second
20:54:17Calciferreconnecting to original nick
20:54:22 Quit Calcifer ("—I-n-v-i-s-i-o-n— 2.0 Build 3515 with A Pack Fix By")
20:54:32 Join robin0800 [0] (
20:54:37 Join Calcipher [0] (
20:55:08 Join thegeek_ [0] (
20:56:31Calcipherlets test out the Dave voice
20:58:41 Join barrywardell_ [0] (
21:00:11Calcipherdamn, thats great, now to get a different voice, for some reason now that I used the compiled install, the voice is reading the correct information
21:01:30Calcipherdo you know of anyway to change the way a voice is listen in sapi?
21:02:38 Quit amiconn (Nick collision from services.)
21:02:47 Join amiconn [0] (n=jens@rockbox/developer/amiconn)
21:06:00***Saving seen data "./dancer.seen"
21:06:01stripwaxCalcipher - I reckon it would be in the registry somewhere - search for VW Kate using regedit?
21:06:06 Quit Daniel_S ("CGI:IRC")
21:06:55stripwaxCalcipher - ok try this. for your sapi options, enter /voice:""VW Kate"" (yes, two quotes either side)
21:07:04kkurbjunmarkun: have you run any tests with llvm to see what the code/speed results are vs gcc?
21:07:15 Quit barrywardell (Read error: 110 (Connection timed out)) strips out the first set, and we still need spaces around the args when we call make_voice.vbs
21:10:48 Quit TPM_Debian ("")
21:12:04 Quit thegeek (Read error: 110 (Connection timed out))
21:12:30 Quit amiconn (" HydraIRC -> <- Nine out of ten l33t h4x0rz prefer it")
21:12:48Calciphertwo quotes didn't fix the problem, but incase its that voice, I'll try a different voice with spaces
21:13:41Calcipherand I'll also include the two ""
21:15:15 Quit spiorf (Remote closed the connection)
21:15:39 Join metzelmetzger [0] (
21:16:29Calcipherits definitely the spaces problem, I tried a voice callled ScanSoft Jill_Full_22KHz
21:16:57Calcipherand it gave me an error saying it couldn't find english for "ScanSoft"
21:17:13 Join einhirn [0] (
21:17:42stripwaxCalcipher - I'm still investigating
21:17:55stripwaxtrying to see when the spaces disappear..
21:18:46preglowlinuxstb: why was a lot of arrays in functions made global in faad?
21:19:27*midgey kicks his gigabeat
21:20:05 Join Thundercloud [0] (
21:20:10preglowlinuxstb: i get the big ones, but not the small ones
21:20:27Nico_Pmidgey: mine doesn't work either
21:20:37 Join Agrajag__ [0] (
21:20:59midgeyNico_P: hang at the rockbox logo with the disk spinning?
21:20:59Nico_Pkkurbjun: r15583 seems b0rked
21:21:04n1spreglow: nice plugin, I kicked it a little to get it compiling
21:21:19n1swhy haven't you committed it?
21:21:33Nico_Pmidgey: no, codec failures
21:21:38linuxstbpreglow: How small are the small ones?
21:21:47midgeyhmm, mine wont even boot
21:22:15Nico_PI also get "incompatible model" on plugins even though I built them
21:22:22Nico_P... I think
21:22:43*Nico_P rebuilds everything
21:23:09jhMikeSr15584 :\
21:23:23Nico_PjhMikeS: I don't think it's my fault
21:23:25Agrajag__Hi guys. I have a Rockbox problem with my iPod Photo. Everything is running great but I have one audio book saved as 20 MP3 files. The default Apple firmware sees the tags in these and displays them. In Rockbox under the track view it just shows "Unknown" but once playing it shows the title correctly. This makes listening impossible as I have to guess which file is which. Any ideas?
21:23:43linuxstbTag them correctly?
21:23:50Agrajag__They appear to be tagged just fine.
21:23:51jhMikeSI'll love that one with a COP codec :)
21:24:00Agrajag__Media Monkey and Apple see them fine.
21:24:08preglowlinuxstb: one example is three arrays of 64 ints, for example
21:24:26linuxstbI would guess I just got carried away...
21:24:30preglown1s: yeah, it's quite old
21:24:48preglown1s: haven't commited because it's not finished
21:24:56Nico_Pyep, r15583 is completely broken
21:25:02n1spreglow: I got it working fine, only thing is that it crackles and pops a bit in the sim but not on target
21:25:34linuxstbAgrajag__: Apple's firmware will use the itunes database, not the tags in the files. I've no idea what media monkey does.
21:25:54linuxstbAgrajag__: But the fact that Rockbox doesn't see the tags would indicate that that's the problem...
21:25:55crashddoes anyone want to buy an ipod 5g, the only thing wrong with it is a slightly loose headphone jack
21:25:59Agrajag__linuxstb: I don't have iTunes even installed.
21:26:25Agrajag__linuxstb: Media Player sees the tags just fine as well.
21:26:47stripwaxlinuxstb - the problem described above is that the tags only appear when the file is playing, but not when viewed from the file list
21:27:01Agrajag__stripwax: Exactly.
21:27:15jhMikeSNico_P: Why does it matter if the codec requests data first?
21:27:22Agrajag__I also have 1600 other MP3's on the iPod and they all appear just fine.
21:27:45linuxstbstripwax: What do you mean by "file list" ?
21:28:13Nico_PjhMikeS: because of the yielding in the AA loading... the codec would start requesting for data on handle -1
21:28:33Agrajag__linuxstb: When I choose "Database" and then "Track" is says "Untagged" 20 times in a row.
21:28:43stripwaxlinuxstb - well, my interpretation of what Agrajag said would be that he's using the database and viewing the track names
21:29:01stripwaxwhich he just confirmed :)
21:29:02SoapAgrajag__: which WPS are you using?
21:29:07linuxstbOK, so you're saying that the database isn't showing the tags, but the tags appear in the WPS?
21:29:09Agrajag__The first 20 files happen to be the book.
21:29:18Agrajag__Soap: New to RB. What's a WPS?
21:29:23Nico_PjhMikeS: maybe the proper solution would have been to make the codec wait for bufseek to return other than an error
21:29:27Soapthe WhilePlayingScreen.
21:29:36n1spreglow: if you want I can post a diff somewhere, I adapted it to the new menu api, new iram macros and changed long->int32_t in some places to make it work on my sim
21:29:38linuxstbAgrajag__: Can you make one of these files available for us to look at?
21:29:43midgeyNico_P: 15580 boots fine
21:29:52Agrajag__That's correct. This happens with all the themes I tried.
21:30:03jhMikeSyeah, but this can't depend on not yielding or we're down that same path as before and a COP codec will just flat out start running.
21:30:04Agrajag__linxstb: Yes I can. Where can I send it?
21:30:14Nico_Pmidgey: the problem is kkurbjun's changes
21:30:14SoapAgrajag__: The reason I ask is that some of the WPSs have a conditional wherein they show the file name if the tags can not be found. I guess I'm asking is it possible that you are seeing the _filename_ when the track starts playing, and not the tag?
21:30:43Nico_PjhMikeS: then I'll also add what I just said
21:30:50*midgey wades through r15581 to see if anything looks suspicious
21:30:56Agrajag__Soap: The filenames, according to APPLE are filenames like AVGF.MP3 and I'm not seeing that.
21:31:15Agrajag__Soap: However, that might just be some sort of temporary iPod naming thing.
21:31:19linuxstbAgrajag__: Can you upload it to a webserver somewhere?
21:31:30stripwaxmaybe your database just needs updating?
21:31:30Calciphervery nice, I am able to use some of my Sapi5 voices for voice menus
21:31:31preglowlinuxstb: faad seems kind of undermaintained...
21:31:36preglowunused variables everywhere
21:31:38Soapno, Agrajag__ - that's the file names of an MP3 put onto an iPod with iTunes.
21:31:43Agrajag__linux: If you grab it right away I'll put it on my personal one in a minute or two. Hold on.
21:31:57Agrajag__Soap: It's very possible that's what it's doing.
21:31:58preglown1s: would be cool
21:32:04Calciphernow how do I do that for directories and files?
21:32:17Nico_Pmidgey: I can give you a patch that reverts kkurbjun's changes if you want
21:32:22Agrajag__Soap: I mainly use iCatcher
21:32:24 Join Eismann [0] (n=Gunnar@
21:32:36midgeyNico_P: i'm quite capable, but thanks
21:32:45stripwaxCalcipher - I have a fix. in the last line of the generated Makefile in that directory, change the end of the last line from -S="$TTS_OPTS"; done \ to -S='$TTS_OPTS'; done \ i.e. single quote
21:33:29stripwaxand further up the file, change the line that says export TTS_OPTS to read export TTS_OPTS=/voice:"VW Kate"
21:35:04 Quit kfazz ("Leaving")
21:35:23 Join JavaMan22 [0] (
21:35:35 Nick Eismann is now known as MortimeR (n=Gunnar@
21:35:42JavaMan22does the current build from the website have album art?
21:35:49stripwaxCalcipher - in fact just change line 2061 in ../tools/configure to replace the -S="$TTS_OPTS" with -S='$TTS_OPTS' then rerun ../tools/configure and specify /voice:"VW Kate" as your options (just the one set of quotes now)
21:35:53JavaMan22thank you!!!
21:35:56jhMikeSNico_P: really, the only way to control the codecs properly is to have them listen to the message queue, wait on it and take commands and never actually set things directly in ci
21:36:04 Join HellDragon [0] (i=jd@unaffiliated/helldragon)
21:36:25JavaMan22im surprised there arent new wps for the album art
21:36:32JavaMan22actually i saw one
21:36:33Agrajag__linuxstb: Are you ready for the URL to grab that example?
21:36:41 Part MortimeR
21:36:53stripwaxJavaMan22- there's loads that will work, possibly with none or only minor modifications
21:36:55Nico_PjhMikeS: but that would require changing all the codecs, wouldn't it?
21:37:05linuxstbAgrajag__: Downloading now - it's going to take about 2 minutes
21:37:09stripwaxfeel free to make your own or modify an existing one
21:37:23Agrajag__linuxstb: It would have gone faster but someone also jumped in with you.
21:37:25JavaMan22i do
21:37:31JavaMan22i edit wps on my ipod
21:37:32jhMikeSNico_P: sure it would to a minor degree. they'll have to be. whatever needs changing needs changing no matter how extensive.
21:37:44 Quit HellDragon (Client Quit)
21:37:49Calcipherok I'll try that stripwax
21:37:51 Join HellDragon [0] (i=jd@unaffiliated/helldragon)
21:38:05JavaMan22linuxstb are you the main dev for mpeg player
21:38:08stripwaxJavaMan22 - since albumart was only just added to the build, it's not too surprising there are no 'official' wps with albumart yet
21:38:18JavaMan22o tru
21:38:37Agrajag__stripwax: thanks. That was one of my noob questions.
21:38:41linuxstbJavaMan22: I did the initial work, but others have carried on with it, and I haven't touched it for months.
21:39:04 Quit HellDragon (Client Quit)
21:39:15 Join HellDragon [0] (i=jd@unaffiliated/helldragon)
21:39:29linuxstbAgrajag__: OK, got it now.
21:39:35 Quit criznach ("ChatZilla [Firefox]")
21:39:55JavaMan22does rockbox use a library to draw stuff like sdl?
21:40:10JavaMan22or does it have its own drawing methods
21:40:12Agrajag__linuxstb: Any chance for a quick look just to see if you see what I see? I wondered if maybe I didn't rebuild the database right after putting them on there, etc. Only Day 2 with RB.
21:40:18jhMikeSI suppose when they first load, they'd just stay blocked until they get a command. Much of the actual processing can be in the core.
21:40:32linuxstbAgrajag__: The file doesn't have a "track number" tag.
21:41:20Agrajag__linuxstb: That's correct. Is it really critical to appear?
21:41:40linuxstbIf you want the audiobook to play in the correct order, yes...
21:41:48stripwaxJavaMan22 - there is a basic line/pixel/blit api in the core and some further apis in hte plugin library (filled triangles etc)
21:41:55Agrajag__linuxstb: the default firmware plays these all sequentially.
21:42:14Agrajag__lnuxstb: it's quit seemless to the point I don't even realize it changed files!
21:42:17CalcipherI don't have a proper editor to find that line
21:42:35Calcipherand I tried notepad and searching for the -S="$TTS_OPTS"
21:42:40 Quit Frazz ("Leaving")
21:42:43Calcipherand there were no results
21:42:58JavaMan22btw for mpeg player after i start playing a video, quit, start it again but with start at time, rb crashes
21:43:23stripwaxCalcipher - um, Notepad probably will screw up line endings. Do you really not have anything else you can use?
21:43:25JavaMan22it crashes when i press select
21:43:34linuxstbAgrajag__: That's probably because the software transferring the files created a track number entry in the itunes database, which is where the Apple firmware reads the tag information from - the Apple firmware doesn't look at the tags in the files, which is what Rockbox uses.
21:43:50jhMikeSJavaMan22: "Crashes" meaning?
21:43:52JavaMan22and another thing is the bar no longer has the right minutes
21:43:54stripwaxCalcipher - sry, line is actually -S="\$TTS_OPTS" in configure
21:44:06JavaMan22the first time it would work 0.0 - 10.0
21:44:10stripwaxgah. -S="\$(TTS_OPTS)"
21:44:15JavaMan22then the second time its 0.0 - 58437553
21:44:15Agrajag__linuxstb: Good answer. I'm manually adding Tracks now right on the iPod. When done do I need to rebuild Database?
21:44:21*Nico_P ponders whether he should revert kkurbjun's changes in SVN
21:44:33Agrajag__linuxstb: If so, what option specifically?
21:44:35JavaMan22it crashes as in it keeps saying Loading...
21:44:38Calcipheroh ok I think I saw that one
21:44:54linuxstbAgrajag__: You should simply use a tag editor application to add the proper track number tags to the files.
21:44:58jhMikeSJavaMan22: ah...yeah it's just hanging trying to find the time
21:45:04stripwaxCalcipher - cool, that's the one. Hopefully one of the devs can commit that change
21:45:07 Quit HellDragon ("xchat upgrade")
21:45:10JavaMan22yeh but i waited a pretty long time lol
21:45:10jhMikeSNico_P: what happened? gigabeat doesn't work?
21:45:16Nico_PjhMikeS: yes
21:45:25linuxstbAgrajag__: I'm not sure how to recreate the database - I don't use it... The manual should tell you.
21:45:41Agrajag__linuxstb: Okay, I don't have those files locally any longer so I was trying to skip the step of copying them all off the iPod. hehehe
21:46:07jhMikeSNico_P: then I concur. I already svn up'ed but haven't made a build so didn't notice yet.
21:46:46markunAgrajag__: you can edit them on the ipod
21:47:22 Join HellDragon [0] (i=jd@unaffiliated/helldragon)
21:47:30 Join maddler [0] (
21:47:33JavaMan22once album art, and mpeg player is better i would never need the original firmware :D
21:47:37jhMikeSone other problem with it is it won't take lcd buffer alignment fudging into account
21:49:22jhMikeSJavaMan22: it's better here :) But not quite commit ready...this video software is really complex to implement thoroughly.
21:49:41stripwaxCalcipher / devs with svn write access :
21:49:43 Quit Rondom ("Ex-Chat")
21:50:24markunJavaMan22: what about battery life?
21:50:31Soapand accessories
21:50:33JavaMan22i forgot about that lol
21:50:42JavaMan22ok also when brightness is here
21:50:54*jhMikeS just spent two days looking for a stoopid bug that was simply a statement in the wrong place :\
21:51:01Calciphernice, stripwax, its generating for "vw kate"
21:51:16stripwaxCalcipher - great!
21:51:37Calcipherwithout the quotes it won't work for voices with spaces, but thats fine with me!
21:51:49stripwaxthat's right. that would actually be quite a bit harder to fix
21:52:00stripwaxbut at least quoting options now works
21:52:27Calciphervery nice, great job stripwax, thanks a ton
21:53:49kkurbjunNico_P: feel free to revert it all, or I can do it if you can tell me how, I was just trying to setup the start of rolo support if someone else wanted to debug it without having to touch or crt0.s
21:54:23Nico_Pkkurbjun: done. it's a pity though, as I'm very eagerly awaiting rolo on gigabeat
21:55:06barrywardell_jhMikeS: thanks for fixing that up. works great. I wonder if line in recording is working on the h10 now...
21:55:08 Join Guile`` [0] (n=Guile@
21:55:09 Part JavaMan22
21:55:56jhMikeSbarrywardell_: np. I think it will. the meters look the same as if you turn radio down all the way.
21:56:02Calciphervery freakin nice!, voice sounds great clear and understandable
21:56:51kkurbjunyeah, I'm kinda tired of working on the gigabeat, there are too many problems with the way things are setup and it's very unrewarding when half the time is spent debugging oddities in different players. half the time I've made changes to the core and it breaks someone else's player the failures are not reproducible on my own player.
21:57:28CalcipherI have a question about compiling, for next time I want to update with svn and compile a current build, the process I should go through is simply ../tools/configure select target and normal mode, then run make zip
21:57:31Agrajag__linuxstb: Thanks for the help. I added tracks right on the iPod files, rebuilt the database and now they appear and work fine. Will keep an eye on that in the future.
21:57:38Calcipheris that right? or should I run "make" alone requires that change too btw
21:57:50kkurbjunI can fix that in a second
21:57:54markunkkurbjun: working on flashing the gigabeat will be fun again
21:58:02linuxstbWhy does IRAM need to be faked for the Gigabeat F? I thought the F had a small amount of IRAM (4KB?)
21:58:24stripwaxCalcipher - that's right. If you've built before you don't need to rerun configure, just "svn update", "make zip"
21:58:29kkurbjunyeah, but I'm afraid that it will cause strange problems on those players that had troubles with other changes I made
21:58:33preglown1s: updated signal_gen.c, thanks
21:58:34 Part Agrajag__
21:58:38krazykitCalcipher, you have to run make before running make zip again
21:58:43kkurbjunlinuxstb: the iram does not seem to work as expected
21:58:49kkurbjunat least i can't get it to work
21:58:58kkurbjun(after rockbo has started)
21:59:01Calcipherah ok, and that process is the long one
21:59:08markunkkurbjun: although in the manual they say that you should be able to use it as normal sram
21:59:19kkurbjunyeah, I've read that
21:59:36markuncan access it in the bootloader?
22:00:07kkurbjunI was just testing in my jtag setup, I would freeze after a reset
22:00:15kkurbjunand I was able to access it then
22:00:29 Quit MethoS-- (Remote closed the connection)
22:00:30Calcipheroh nice, thanks stripwax
22:00:33kkurbjunbeing read and write to it
22:00:43kkurbjunI don't know if the mmu is causing some problems with it
22:00:50markuncould be
22:01:04kkurbjunbut I can't write to it with the jtag setup even using physical addreses
22:01:37stripwaxyep - svn update, make, make zip
22:02:23kkurbjunjhMikeS: has been able to write to it though
22:03:01Calcipherand then I just re do the voice steps, cool
22:03:14Calciphernow atleast I have an idea of what to do
22:03:31CalcipherI'm not used to using shells and compiling
22:04:18stripwaxCalcipher - so I *think* you can just svn update, make, make voice, make zip
22:04:40Calcipheryeah that sounds about right
22:04:47Calciphernext I'll try patching rb
22:05:05Calcipherhehe, actually I should see about makind dir and file clips
22:07:04stripwaxDevs : is it possible to configure a single build for Normal and Voice ? I can see Simulator+Voice works but can't seem to specify Normal+Voice as a combination in configure
22:09:26linuxstbIsn't that what happens when you just select "Voice" in the advanced options?
22:09:51 Nick idnar_ is now known as idnar (i=mithrand@unaffiliated/idnar)
22:09:58stripwaxah, is it?
22:10:08 Quit karashata ("Leaving.")
22:10:27linuxstbI've just tried selecting a voice build, and typing "make" seems to be building a normal build.
22:10:47 Join karashata [0] (
22:11:27stripwaxlinuxstb - great. Calcipher - there's your answer
22:12:33Calcipherreally hmm
22:14:30Calciphercan anyone direct me to where on the site I can find info on voice files and directories clips
22:17:05Calcipherbeautiful, thank you karashata
22:17:25Arathisbarrywardell_: why is the fm tuner commit been removed from svn?
22:18:35barrywardell_Arathis: it wasn't. Just a problem with the original code was fixed. I think this was the last thing stopping line-in recording from working too.
22:19:29Arathisbarrywardell_: but it's not on the list anymore
22:20:05karashataArathis: it still shows up on mine
22:20:16BigBambiArathis: Which list?
22:20:47 Quit mf0102 ("Verlassend")
22:20:57Arathisboth the list of last commits on the mainpage as
22:21:08BigBambiIt is between the two
22:21:23barrywardell_I can see it on this page:
22:21:32LearHm, when unpacking a 16-bit color value to 24 bits, why should the low bits be repeated? Repeating the high bits seem more correct.
22:21:33BigBambiOnly the last few are shown on the front page, and there have been enough commits to reomve it from that
22:21:39karashatathe main page only shows the 5 most recent commits and the commits in the last four weeks is a little slow to update
22:21:43barrywardell_the since-4weeks page sometimes lags behind the main page
22:21:49BigBambiAnd the since-4wekks hasn't yet updated
22:22:09Arathis shows it. fine than :)
22:22:48BigBambiThat is the same as the front page, just more than 5
22:23:07BigBambirelax, you would have to see another commit actually removing something, it can't just vanish
22:24:03Arathisthat's what confused me ;)
22:24:46Arathisbarrywardell_: so fm recording works now too?
22:25:00barrywardell_and probably line-in too
22:25:02LearWhen doing so, the RGB value reported by the color picker actually match the bitmap. Or, put otherwise, when drawing using the reported color, a background is no longer dithered...
22:25:17Arathisbarrywardell_: if I just had a cradle or a modified cable ;)
22:25:23barrywardell_me too
22:26:02 Quit barrywardell_ ()
22:26:08*karashata can test in a sec
22:26:19Arathisnever tried to make such a cable though it should be worth it
22:27:05 Part przemhb ("User is away.")
22:27:48karashataline-in still doesn't work
22:31:19*Arathis needs to get rid of the midi-patchset download in his update script ... takes too long ...
22:33:08 Join SirFunk [0] (
22:33:16 Join barrywardell [0] (
22:33:49Arathisbarrywardell: ever tried to make a line-in/out cable?
22:34:12Arathisand regarding to karashata line-in recording still doesn't work
22:34:38karashatayeah, tested it quick since I actually have the cradle for it
22:34:48karashatano signal
22:34:58 Join pixelma [0] (i=pixelma@rockbox/staff/pixelma)
22:35:01*Calcipher is happy, he has rockbox reading in a sexy lady's voice, including files and dirs, now the only thing missing is actuall track info, I think stephen doyon made some patch for that
22:36:32Calcipherah, its good when that voice says Iron Maiden
22:36:39Calciphersounds right for some reason
22:36:49barrywardellArathis: yeah, I thought about it, but never got around to it
22:38:45ArathisI think those jacks needed for the H10 are quite hard to get in europe or at least not very cheap :/
22:40:11Dark_ApostropheCalcipher: Can one get Text To Speech?
22:40:29Dark_ApostropheCalcipher: I'd love to download craploads of articles from news sites and listen to them as if they were podcasts
22:42:59 Quit mirak (Remote closed the connection)
22:42:59CalcipherDark_Apostrophe, I know that you can have a tts program generate audio files of documents, thats pretty much what your looking for
22:43:05 Join Daniel_S [0] (i=57b0cd0d@gateway/web/cgi-irc/
22:43:29Dark_ApostropheCalcipher: Problem is, Festival generates craploads of .wav files
22:43:36Dark_ApostropheI just need a single mp3/ogg file per article
22:45:41Calcipheroh, well I am on Windows, and there are programs that do just that for windows
22:46:09Calciphertry looking for a text to speech program that can use festival
22:46:40barrywardellArathis: I saw the cables on ebay for $1
22:46:45Calcipheryou should be able to get a hold of one that can let you specify encoding format
22:46:48stripwaxDark_Apostrophe - you wamt to then convert the wav to ogg?
22:47:04stripwaxLear - hm, converting 16bit->24bit should just zero the low bits, not repeat them
22:47:08 Quit barrywardell ()
22:47:11 Quit Daniel_S (Client Quit)
22:47:12Dark_ApostropheThat is not the problem, rockbox can play wav
22:47:21Dark_ApostropheThe problem is, it outputs a file for every sentence
22:47:21Arathisbarrywardell: do you know how they are named and perhaps what else is needed to build the cable?
22:47:30Dark_ApostropheAnd in a 5 page article, there are quite a few sentences
22:47:44Calcipherwow, can you use an other engine:?
22:47:46stripwaxDark_Apostrophe - Oh! I'm sure there's a way to tell it to not do that..
22:47:56Calcipherthere must be
22:48:00Dark_ApostropheCalcipher: I don't know of any other
22:48:08Dark_Apostrophestripwax: Not with the KDE GUI I use
22:48:38crashdwhat's the sansa support atm?
22:49:44stripwaxDark_Apostrophe - if you just use the Festival command line directly, I'm sure it would work.
22:49:52 Join jhulst [0] (n=jhulst@unaffiliated/jhulst)
22:50:03Dark_Apostrophestripwax: I don't know how to tell it to read articles in my RSS feed reader :p
22:50:04stripwaxOr concatenate the output yourself into a single file
22:50:18 Join obaid [0] (i=5253546a@gateway/web/cgi-irc/
22:50:31stripwaxI'm sure you could pipe the output from your rss feed reader into festival
22:50:36stripwaxanyway, this is getting off topic
22:50:41obaidhey guys, how can i put my .m3u playlist on rockbox?
22:51:02stripwaxobaid - copy it
22:51:21obaidno matter which folder?
22:51:31stripwaxwhichever one you want to put it into
22:51:53karashataif the playlist was created in the same folder as the tracks it lists, any folder should work
22:52:20obaidallright, great :D
22:52:35stripwaxif the .m3u was created with absolute paths it will work in any folder
22:53:26 Join amiconn [0] (n=jens@rockbox/developer/amiconn)
22:54:11amiconnstripwax: For voices with spaces around them, a better (imho) method is to use single quotes around the entire argument
22:54:23amiconnI.e. '/voice:Microsoft Mary'
22:54:44 Nick parafin is now known as parafin|away (
22:55:00stripwaxamiconn - would that actually work?
22:55:10amiconnIt *does* actually work
22:55:37stripwaxIt sounds worth adding to that wiki page
22:59:28stripwaxor at least documented *somewhere*
23:00:36Learstripwax: nope, zero-padding is not correct.
23:01:19stripwaxamiconn - would you agree that putting single quotes around the entire argument is non-obvious and most people would be far more used to trying to put quotes around the value?
23:01:43stripwaxLear hm, why?
23:01:44amiconnTook me a while to find this solution
23:01:59stripwaxWith that simple patch, both ought to work
23:02:09Learstripwax: Would never show "FF" then.
23:02:19 Quit obaid ("CGI:IRC (Ping timeout)")
23:02:25amiconnbash fidldes with double quotes afaik, but not with single quotes
23:02:35stripwaxLear mm, right.
23:02:40jhMikeShmmm...DEFAULT_BUFFERING_FILECHUNK set to 8k on e200 doesn't last long...everything locks up
23:03:14amiconnI was able to build voice files before, but only by editing the generated Makefile
23:04:10stripwaxamiconn - yep but we're talking about the arg to cscript which would now look like 'voice:/Microsoft Sam' . Is quoting the entire arg work for tools in general?
23:05:10amiconnYour slash is in a wrong position ;)
23:05:26stripwaxLear - right so you want to scale the lower bits up according to the fraction. sounds like copying the top bits would do the right thing.
23:05:49amiconnThe sapi voice generation is a bit odd because several different parsers have a chance to clobber the argument
23:05:56stripwaxamiconn - oops, right :)
23:06:05***Saving seen data "./dancer.seen"
23:06:11amiconnFor the vbscript, I used named arguments as the easiest solutiion
23:06:35Calcipherwell I'm going off to the gym to enjoy my Sansa using rb, now with nicer voices!
23:06:43amiconnBut the tokenizer breaks at spaces first - unless the whole thing is quoted
23:06:47stripwaxI didn't know that you could quote the entire arg to cscript in that way (so didn't try it..)
23:07:03CalcipherI will be back later and will try adding patches to my build, thanks so much guys
23:07:08stripwaxbut the patch just quotes the whole thing in single quotes to fix that
23:07:28amiconnWhat patch?
23:07:38amiconnYou must not quote the entire argument string
23:07:48 Join Mouser_X [0] (i=cf9bb003@gateway/web/cgi-irc/
23:07:52amiconnSo the quotes must be put by the user in configure
23:08:02stripwaxamiconn - ?
23:08:06stripwaxamiconn - the patch I listed above
23:08:10 Join safetydan [0] (n=safetyda@rockbox/developer/safetydan)
23:08:11 Quit Domonoky_ ("Trillian (")
23:08:16amiconnOtherwise things like '/voice:Microsoft Mary' /speed:1 are impossible
23:08:32stripwaxamiconn - no: /voice:"Microsoft Mary" /speed:1
23:08:48stripwaxor would that not work?
23:08:56jhMikeShmmm...if I make the e200 ata driver yield every 1ms instead of 2ms the buffering/PCM proportion is identical to the old...with no specialized buffer yield
23:09:10amiconnI'm not sure, iirc that didn't work
23:09:25 Quit Lear ("ChatZilla [Firefox]")
23:09:31amiconnWhen I tried, I mean
23:09:48stripwaxamiconn - right, and with the patch, now it would. i think
23:10:26stripwaxamiconn -
23:11:24amiconnIirc the quotes must be put around the entire argument
23:11:53amiconnAnd it looks to me that this patch just exchanges non-working double quotes for non-working single quotes -> nothing gained
23:12:30amiconnOnce you know that you have to use single quotes and put them around the entire argument, it just works, without a patch
23:12:47stripwaxright but that's imho an unusual way to treat options
23:13:09stripwaxtyping /voice:"Microsoft Sam" is more usual imo
23:13:35stripwaxLet me just check if /voice:"Microsoft Sam" /speed:1 works with the patch now. Not sure what you mean by 'non-working'
23:13:36amiconnDid that work?
23:13:43stripwaxDid what work?
23:13:50amiconnYour quote placement
23:14:22stripwaxFor which argument, /voice:"Microsoft Sam"? yes. that's why I put the patch on the tracker
23:14:31stripwaxThe second example, above, I haven't tried yet
23:14:44stripwaxI am doing it right now, when I'm done typing into irc
23:15:01amiconnMaybe single quotes would also work around the voice name only, I never tried
23:15:17amiconnPutting the quotes aroud the whole thing just felt more natural
23:15:43stripwaxis it common? (that's why I asked earlier...)
23:16:08stripwaxi.e. would average user just try that outright rather than /voice:"blah blah"
23:16:24amiconnI don't know if it's common, but there is no space between the argument name and value, hence placing a quote there seems odd to me
23:16:29stripwaxusing /voice:"Microsoft Sam" /speed:1 as opts works with the patch too
23:17:04stripwaxI don't think I've ever seen putting the opt name *inside* a quote before
23:17:07amiconnIf it were /voice[space]"Blah blah" it would look okay, but not with the colon
23:17:35amiconnAnd /voice:[space]Name doesn't work either, there must be no space inbetween
23:18:05amiconnAnyway, quote placement seems to be a non-issue, and your patch changes broken double quotes for broken single quotes
23:18:35stripwaxagain, broken why? it now works, and before it didn't
23:18:52amiconnYes, but with the patch I'm quite sure that single quotes won't work
23:19:00amiconn...which did work before
23:19:47amiconn(iirc perl handles single and double quotes interchangeably; in a string with single quotes it leaves double quotes intact and vice versa)
23:20:13stripwaxI understand what you mean now
23:21:46 Join Llorean [0] (
23:21:48stripwaxIn that case, the comment in is misleading : "-S=<TTS engine options> .. Enclose in double quotes if the options include spaces"
23:22:24amiconnThat comment isn't mine...
23:22:30stripwaxI didn't say it was...
23:22:47*amiconn wonders if perl can be told to leave both single and double quotes alone...
23:23:05stripwaxWhatever you think the 'best' solution is, it should be documented somewhere. Calcipher and I wasted a bit of time there
23:27:22 Join nerve [0] (
23:28:11nerveanyone here familiar with USB plugs? I'm currently looking at a USB A, with five pins and I can't find a schematic that describes the pin's behavior
23:28:40stripwaxnerve - here.
23:28:53nervethere on;y four pins there..
23:29:05stripwax1 2 3 x 4
23:29:15stripwaxor are you saying it's not a mini?
23:29:21nerveyea, its not a mini..
23:29:40stripwaxit's not USB then
23:30:04nervereally? cause it comes with my mp3 and plugs into my USB jack :)
23:30:10nervemp3 player*
23:30:35stripwaxnerve does the other end plug into your pc? does the other end also have 5 pins?
23:32:11nervethe end that plugs into the pc has 5 pins
23:32:20nervebut those 5 pins connect to a 4 pin slot
23:32:29stripwaxyeah. i wonder how that works.
23:32:33nervei'm trying to find the vcc and gnd so i can charge with out the device thinking that its connected
23:32:47 Quit Arathis ("Bye, bye")
23:33:12stripwaxeither the extra pin is a dummy or two pins on the cable physically touch one conductor in the socket
23:35:32preglowit seems we can use faad with fewer hacks now
23:36:33Lloreanpreglow: You mentioned you'd only seen LC and LC+SBR if I recall?
23:37:02LloreanI seem to remember someone with a parametric stereo file at one point, I think.
23:37:39 Quit desowin ("use linux")
23:43:29 Quit karashata ("Leaving.")
23:43:38 Quit HellDragon (Client Quit)
23:45:37 Join thegeek [0] (
23:45:37 Quit Xerion (Read error: 104 (Connection reset by peer))
23:45:44 Join Xerion_ [0] (
23:45:49 Quit metzelmetzger ()
23:46:31 Quit XavierGr (Read error: 110 (Connection timed out))
23:51:06 Quit bertrik ("bye")
23:54:00 Join HellDragon [0] (i=jd@unaffiliated/helldragon)
23:55:00pixelmahmm... my just freshly compiled c200 sim crashes as I try to play some music (just mp3s that played fine in Rockbox so far)
23:55:47pixelmarev 15589
23:56:23IsolinearI've noticed some instability in the more recent builds as well...
23:57:19stripwaxAnyone know how rbutilqt finds Crashbox theme for ipod 5g but it's not on this page
23:57:58 Join billenium [0] (
23:58:18pixelmaI think it looks for builds that require no patches in too (not a 100% sure though)
23:58:21PaulJamdoesn't rbutil take the themes from ?
23:58:22preglowLlorean: oh, yeah, seem parametric stereo too
23:58:35pixelmaerr *themes
23:58:42stripwaxHm. rbutil no longer shows Crashbox theme. I guess it was removed from ?
23:58:47IsolinearI updated from r15184 to r15526 and noticed sluggish scrolling through my giant playlist, as well as that weird bug where the song would crash and you would just hear that "static" type sound.
23:58:58stripwaxOh, no, there it is, right at the bottom.

Previous day | Next day