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).

#rockbox log for 2004-11-18

00:00:02 Quit [IDC]Dragon ()
00:05:52***Saving seen data "./dancer.seen"
00:06:12 Join lImbus [0] (
00:06:48amiconnhi again!
00:07:00amiconnCould you please look up the AN5 value?
00:07:11lImbusof course, just a moment
00:07:30lImbusunfortunately, it's too late to scan now. postponed to tomorrow morning
00:08:00amiconnNo hurry
00:08:04quelsaruktime to go, cu another day
00:08:08quelsarukgood night!
00:08:13 Quit quelsaruk ("Using KVIrc 3.0.1 'System Virtue'")
00:08:28lImbusAN5 is mostly 001, but sometimes 000
00:09:07amiconnAh, interesting. AN1 around 0 as well?
00:09:38lImbusAN1: 3FF
00:09:42lImbusis that -1 ?
00:09:52 Join midk_ [0] (
00:10:27amiconnlImbus: No, it's 1023 (these are the ADC input values, which are unsigned)
00:11:24amiconnAh, I was looking at the wrong place. AN1 is 3FF here as well. However, AN5 is not around 0 with the newer hardware, as I suspected
00:16:22lImbusdoes the ondio really need the anti skip buffer setting ?
00:17:27amiconnMost probably not. We just didn't remove it, perhaps it's a good idea to do that
00:19:19lImbusmay confuse noobs, then they fiddle around, then they complain about something, and it takes ages to find out one of the symtomps of his problems is this.
00:20:51 Join AciD [0] (
00:26:57 Join AimVector [0] (AV@
00:27:33 Part scott666_
00:27:43 Quit midk (Read error: 110 (Connection timed out))
00:27:48AimVectori have recently flashed my jukebox, when ever i try to start it the HD spins and makes a clunking sound and has the error "ATA Error: -42"
00:27:51AimVectorwhat does that mean?
00:32:04LinusNsounds like the drive doesn't get enough power to spin up
00:32:28AimVectorhappens while it is plugged in too
00:32:40AimVectorand i have tried charging the batterys with an external charger for a lil while too
00:32:58LinusNthen i think your battery connectors might be broken
00:33:18AimVectorill check
00:33:27AimVectorany visible signs to look out for?
00:33:51LinusNopen it up and flex the small pcb:s with the springs on them
00:34:12LinusNthen you see if they are properly connected to the chassis
00:35:02AimVectorcould it be the batterys?
00:35:54AimVectori need one of those T screwdrivers :|
00:36:14LinusNtry with a set of fresh batteries
00:36:30AimVectoronly fresh batterys i have are AA
00:36:42LinusNa bad battery may very well show a good voltage, but fail to deliver enough current
00:36:46AimVectornot the Ni-MH ones that are for the jukebox
00:36:59 Join bagawk [0] (Lee@bagawk.user)
00:37:17AimVectorwouldnt the power supply bypass that though no?
00:37:36bagawkHi amiconn and LinusN
00:37:39LinusNthe power supply can't deliver enough current alone
00:37:48LinusNit needs the batteries for that
00:37:53LinusNbagawk: hola
00:37:59AimVectorany tips on how to remove the screws without one of those T screwdrivers?
00:38:08bagawk?? To Run only from power supply?
00:38:29bagawkAimVector: that is just a head, no tips needs
00:38:37bagawkYou will need size t10
00:39:45AimVectorgonna try and find a t10 screwdriver
00:39:48AimVectorshould have one somewhere
00:40:08amiconnhi bagawk
00:40:22bagawkLinusN: I need more mods for my jukebox :)
00:40:57lImbushi bagawk/dstar5/lee :-)
00:41:19bagawklImbus: Helloo
00:44:39bagawkLinusN: has anyone ever made a IR remote control for the jukebox?
00:44:57LinusNi've heard of a project like that
00:45:15AimVectorcant find a t10 screwdriver, ill get one at my parents work 2morow
00:45:31bagawkAimVector: what are you planning to do?
00:46:04AimVectormy jukebox has a power problem, LinusN says it may be the battery connectors have came off the pcb
00:46:17AimVectorcommon problem i have read
00:46:35bagawkahh I guess you have a studio?
00:46:46AimVectorjukebox recorder
00:47:11bagawkLinusN: - Object not found!
00:47:48LinusNbagawk: oops
00:48:01AimVectorbagawk what have you done to ur jukebox?
00:49:01bagawkAimVector: CPU board transplant, RTC wake up and white backlight
00:49:17AimVectorcpu board transplant??
00:49:24AimVectorwhat did that envolve and why?
00:50:05bagawkI replaced the cpu board so i could flash (it the the bottom board in the recorder)
00:50:40AimVectori thought ya could flash all v1 models
00:51:04bagawkMost of them
00:51:15bagawksome have a older flash chip that you cannot
00:51:35AimVectorhave you got a pic of ur jukebox?
00:51:42AimVectornever seen a white backlight before heh
00:51:56bagawkI do have a ugly picture taken with a webcam
00:52:31amiconnAimVector: Yes I've done it as well :)
00:53:03AimVectorthats a nice backlight
00:53:50amiconnLinusN: Are there (probably old) recorders with mask bit 0 == 0?
00:55:18lImbusbagawk: 404
00:55:30bagawkhold on...
00:56:10lImbusany plans on getting light to the ondios ?
00:56:22lImbuswould be leet :D
00:57:19amiconnIn fact it's already prepared on the PCB, but not populated. Looks like some pads for an EL driver chips and associated parts
00:57:41AimVectori was tempted to buy an ipod today
00:57:53 Quit midk_ ("just STOP it arspy")
00:58:57lImbusamiconn: I thought on seeing that.
00:59:21bagawkthat might work
00:59:26bagawkfreewebs deleted me hehe
00:59:33 Quit edx ()
00:59:39bagawki did not put a link to there main page :P
01:00:38bagawkwrong name :P
01:00:52amiconnlImbus: We just have to find out which chip fits in there. Most likely (according to [IDC]Dragon) Sipex SP4403
01:01:20AimVectornice :)
01:02:08bagawkA freidn took some nicer pictures, but i dont have them right now
01:02:16 Quit AciD (Read error: 104 (Connection reset by peer))
01:02:30AimVectorits cool
01:02:51lImbusamiconn, what does that chip cost, how hard is it to build it in, and what happens if its the wrong one ?
01:03:52amiconnI didn't find a source for that chip. Anyway I'd like to check whether it fits (should be possible to tell from the pinout and some beeping)
01:04:15bagawktime to go
01:04:18bagawksee you later
01:04:39 Quit bagawk ("umount /dev/brain")
01:05:02amiconnThen there is the next question - where to get a suitable EL foil. There is a little web shop in Sweden that sells an EL backlight circuit along with an EL foil that looks like it could fit
01:06:30lImbusmounting leds from the side of the glass like jukebox-style is no option ?
01:07:48amiconnThat would require a light diffusor - I think there is not enough room for it.
01:08:23lImbusoh, yeah, of course
01:17:17amiconnLooks like it's time for some beeping...
01:17:54lImbustüüüüt üüüüü tüüü üt
01:26:49LinusNi have contact with the iriver hard disk!
01:27:17LinusNIDENTIFY says "TOSHIBA MK2004GAL"
01:28:32LinusNnow i think i have earned some sleep :-)
01:29:10LinusNnite all
01:29:15lImbusdon't forget to switch the topic, else you will dream of your registers and stuff
01:29:15 Part LinusN
01:38:43lImbusme too, off for bed...
01:39:22 Quit lImbus (" HydraIRC -> <- Nine out of ten l33t h4x0rz prefer it")
01:40:25 Quit mecraw_ ("Trillian (")
01:44:28 Quit AimVector (Read error: 104 (Connection reset by peer))
02:05:54***Saving seen data "./dancer.seen"
02:13:39 Quit matsl (Remote closed the connection)
02:14:44 Join scott666_ [0] (
02:31:39 Join amiconn_ [0] (
02:36:19 Quit scott666_ ("i'll be back...eventually...")
02:47:10 Join webguest37 [0] (~18221819@
02:47:53 Quit webguest37 (Client Quit)
02:49:28 Quit amiconn (Read error: 110 (Connection timed out))
02:49:28 Nick amiconn_ is now known as amiconn (
03:11:27 Quit Stryke` ("Friends don't let friends listen to Anti-Flag")
03:54:47 Part amiconn
04:05:56***Saving seen data "./dancer.seen"
04:21:52 Join rakaus [0] (
04:22:39rakaushey all, just had a quick question about flashing my ROM on my FM recorder
04:24:05rakauswhat happened to the FM-recorder.bin files that used to be included in the daily build .zip files?
04:26:31 Quit rakaus ()
04:38:55 Quit unibrain (Read error: 113 (No route to host))
05:38:55 Quit lImbus_ (Remote closed the connection)
06:06:00***Saving seen data "./dancer.seen"
06:37:57 Join midk [0] (
07:10:20 Join AciD [0] (
07:36:18 Join oxygen77 [0] (
07:45:40 Join iwok [0] (aixela20@
07:54:11 Part iwok
08:06:02***Saving seen data "./dancer.seen"
08:22:53 Quit AciD (Read error: 104 (Connection reset by peer))
08:36:58 Join einhirn [0] (
08:39:53 Join AciD [0] (
08:44:58 Join amiconn [0] (
08:59:32 Join Zagor [242] (
09:30:52 Join LinusN [0] (
09:34:36 Join [IDC]Dragon [0] (
09:34:59[IDC]Dragon'morning chaps!
09:35:26LinusNcheck the irc log for iriver news
09:35:50[IDC]DragonI did, my deep congratulations!
09:37:09[IDC]DragonI eventually should get an iriver, from a clearance sale or such :-/
09:38:24gromit`_hello !
09:39:12[IDC]Dragonamiconn: your board is already in the mail again
09:39:36[IDC]Dragonso you should have it back at the weekend
09:42:31[IDC]DragonLinusN: so you can attach the file system soon? That should be a leap forward.
09:42:49LinusNmy ata driver is a mess right now
09:43:09LinusNbut it will work very soon, i hope to be able to spend some time this evening
09:43:22[IDC]Dragonare buttons and LCD in shape for Rockbox?
09:44:23LinusN[IDC]Dragon: what was the url for the backlight modules again?
09:44:52LinusN[IDC]Dragon: buttons are not yet in shape, but that's just lazyness from my side
09:46:17[IDC]Dragonthe product is
09:47:00LinusNhow many?
09:47:30[IDC]Dragonone, unless Jens wants to join in right away
09:47:46[IDC]Dragonhe didn't sound too much like it
09:48:41[IDC]Dragonwell, maybe two, to be on the safe side
09:49:06LinusN[IDC]Dragon: i tried to /msg you
09:49:11[IDC]Dragonsomebody will want one, too
09:54:13Bagdernice work on the ATA code LinusN
09:59:08 Join lImbus [0] (
09:59:21lImbusgood mornin' everybody
10:06:04***Saving seen data "./dancer.seen"
10:08:50Zagormajor snowfall today
10:10:21LinusNand i'm supposed to go out on a field test today...
10:10:22Zagorno no, snow
10:10:44Zagorcould be worse
10:10:46lImbusshowfall ?
10:10:47Zagorcould be raining
10:10:50BagderLinusN: bring the showel
10:12:24ZagorlImbus: you know, white stuff from the sky? :)
10:13:03lImbushey, I'm living next to the highest point of belgium, we already had this year :)
10:13:30lImbustbh, that hghest point is not even 700m above the see, and we had snow once :-/
10:13:32Bagderhighest point? is that above sea level? B-P
10:25:51BagderZagor: I see your new box runs rc5 very fine :-)
10:26:36Zagoryeah, it's cranking out quite a few keys
10:40:43lImbusI'm about to finish the scans of my ondio (older pcb version). a few issues:
10:41:38lImbus- my one has hf-cases around a few components (the surface between mcc-slot and radio pcb, underneath the lcd) and the radio-pcb can't be lift off without desoldering
10:42:08lImbusif needed, I can do that (not now), but I'm not going to tamper around with that unless it's really needed.
10:44:54lImbusI did 1200dpi, rotated and cropped.
10:45:31lImbusI'm driving to my workplace now to deposit them on a slow but biiiig webspace. bmp and jpg is ok I assume ?
10:47:47*oxygen77 is away: chui pas là
10:49:25lImbusme silly. should be more careful with new toys...
10:51:10LinusNi did a similar thing on my Player a few years ago
10:51:22LinusNonly the screw perforated a power line :-)
10:51:32lImbuseven more ouch
10:52:25lImbusbut: it's fixeable. my one keeps a bullet hole unless I start with some filler...
10:52:37*lImbus is taking photographs of the bullethole
10:53:45*LinusN had a bullet hole in a wall in his previous apartment
10:54:09LinusNrough neighborhood
10:54:31lImbusa real bullet hole ? from a gun ? in sweden ?
10:54:56LinusNfrom a hunting rifle
10:55:27LinusNthe bullet stopped in the mains fuse box :-)
10:55:33lImbuswhat if you would have been in the line of that at the wrong moment ?
10:56:17Zagorgangsta linus
10:56:47LinusNin fact, it wasn't a fight
10:57:08lImbusI suppose it was an *accident*
10:57:21LinusNmy neighbour two flats away played with the rifle and didn't know it was loaded
10:57:37LinusNthe bullet passed one whole apartment and stopped in mine
10:57:47LinusNnobody was hurt
10:58:00lImbuswow. too much people die during playing with guns.
10:58:17LinusNlet's just say that he learned a lesson :-)
10:58:44lImbusdid he sell that weapon ? if not, he did not learn...
10:59:47LinusNi don't know
11:00:10lImbusBTW: does everybody agree with my toughts and uttering about scans of ondio ? so I can put the scanner back in the shelve
11:01:07LinusNyou mean to not remove the shield?
11:01:47 Quit einhirn (
11:02:15NJoineinhirn [0] (
11:02:35lImbusnooo. im only removing the shield if really needed. the current cvs-version is already running on my ondio. so I suppose it's NOT needed immediately
11:05:03LinusNthen what am i supposed to agree with?
11:05:27Zagorlunch (early)
11:05:43lImbusI just want to know if 1200 dpi is ok, and if I forgot to do anything if I scanned front and back
11:05:51lImbusenjoy your meal
11:06:10LinusN1200 dpi is fine, as long as you don't destroy it with a hard JPEG compression
11:06:28lImbusenjoy your meal
11:16:23lImbusoff for work ...
11:16:26 Quit lImbus (" HydraIRC -> <- s0 d4Mn l33t |t'z 5c4rY!")
11:51:57 Quit AciD (Read error: 104 (Connection reset by peer))
12:06:05***Saving seen data "./dancer.seen"
12:10:03 Join AciD [0] (
12:26:20*oxygen77 is back (gone 01:38:33)
12:51:20 Quit Bager ("CGI:IRC (Ping timeout)")
12:51:25 Quit [IDC]Dragon ("CGI:IRC (Ping timeout)")
13:51:32 Join ripnetuk [0] (
13:51:46ripnetuk"TOSHIBA MK2004GAL" - Nice work Linus :)
13:53:16ripnetukhas any thought been given to how non-bdm owners are going to boot code on the iRiver? i assume you are still using the wiggler to run it
13:53:25 Quit oxygen77 ("Cho")
14:01:12 Join oxygen77 [0] (
14:02:23 Part oxygen77 ("Tcho")
14:06:08***Saving seen data "./dancer.seen"
14:18:05 Quit AciD (Read error: 104 (Connection reset by peer))
14:28:31 Quit ripnetuk ("CGI:IRC")
14:28:51 Join quelsaruk [0] (~kvirc@
14:36:38 Join AciD [0] (
14:50:39 Join lImbus_ [0] (
14:51:28 Quit elinenbe (" HydraIRC -> <- The dawn of a new age")
14:51:40 Nick lImbus_ is now known as lImbus (
14:53:10 Join Bagder_ [0] (
14:53:27 Quit Bagder (Read error: 104 (Connection reset by peer))
14:53:34 Nick Bagder_ is now known as Bagder (
14:53:40lImbusamiconn: how / where should I drop the images (scans) ?
14:54:00 Join elinenbe [0] (~elinenbe_@
15:05:00 Join scott666_ [0] (
15:11:53 Quit AciD (Read error: 104 (Connection reset by peer))
15:12:51 Join AciD [0] (
15:13:58 Quit scott666_ ("i'll be back...eventually...")
15:15:21 Part LinusN
15:46:36 Join amiconn_ [0] (
15:46:37 Join [IDC]Dragon [0] (~d90a3255@
15:47:14[IDC]DragonI was disconnected for hours without noticing
15:47:30amiconn_hi [IDC]Dragon
15:47:39[IDC]DragonAldi Nord is bringing an mp3 HD player:
15:47:54Zagor[IDC]Dragon: can you try whatever didn't work with twiki before?
15:48:26[IDC]Dragonlooks again different than all I've seen
15:48:51[IDC]Dragonfull description:
15:49:46[IDC]DragonZagor: I'll try, hang on
15:51:26lImbusamiconn: are you ok ?
15:51:41lImbusamiconn_ that is
15:53:21[IDC]DragonZagor: works, hooray!
15:54:00[IDC]Dragonwhat was it?
15:55:02amiconn_lImbus: I am here. My rdp connection doesn't like me today (again)
15:57:15Zagorformatted for the logs (and thus google): "Connection reset by peer: ap_content_length_filter: apr_bucket_read() failed" with twiki was remedied by changing apache2 multi processing module from mpm_worker to mpm_prefork
15:58:02[IDC]Dragonbut the dilithium crystals and the warp core is ok?
15:58:26lImbusamiconn_: so where should I get them ? and which version ? I can tell you the url where it is on my slow but huge webspace, I can upload it to wiki, I can post it on the list.
15:58:59lImbusmaybe you want to have the bmp-files (about 24 mb) but only the jpgs (700kb) should go towiki.
15:59:20lImbusand finally, I can make a browsing people friendly version.
16:03:03amiconn_I think the version that goes to the wiki should be approximately the same size as the ones already there (for inline display). Then you could add the full size jpegs and link them
16:06:07lImbusok, I'll try that
16:06:09***Saving seen data "./dancer.seen"
16:08:43lImbusmhmm. amiconn_, where should it go in wiki ? there are already pictures...
16:14:48amiconn_Just attach the pictures to the ArchosOndio topic. I think it is best to have the inline versions side by side with the pics of the newer hardware
16:28:25 Join methangas [0] (
16:32:54 Part Zagor
16:33:49 Nick quelsaruk is now known as Quel|away (~kvirc@
16:43:42 Join mecraw_ [0] (~lmarlow@
17:01:54 Quit lImbus (Read error: 104 (Connection reset by peer))
17:31:36 Join Stryke` [0] (~Chairman8@resnet-241-86.resnet.UMBC.EDU)
17:35:08amiconn_[IDC]Dragon: r u there?
17:37:35[IDC]Dragonnow I am
17:37:47[IDC]Dragonjust looking at that alien Ondio
17:39:29amiconn_Yes, I did already have a quick look as well.
17:40:19amiconn_Last night I tried to identify the necessary EL driver chip for the empty footprint. No success yet. Your candidate (Sipex SP4403) does not match...
17:41:02amiconn_The closest match I could find was the Sipex SP4425, which has all pin at the necessary places, but mirrored...
17:42:42[IDC]Dragonmaybe they mount it upside down :-|
17:43:51[IDC]Dragonwhat is the stuff under lImbus LCD?
17:44:57amiconn_I don't know. Looks like simple white plastic to get a bit more contrast (as the lcd is semi-transparent)
17:45:02 Join edx [0] (
17:45:30amiconn_[IDC]Dragon: What I figured about the pads so far (please correct me if you think I got these wrong):
17:46:34amiconn_Left row, top down: (1) Enable input, connected to the CPU, (2) Vcc, (3), (4) EL foil
17:47:25amiconn_Right row, top down: (1) unknown, (2) ground, (3) coil (4) unknown
17:48:06[IDC]Dragon(1) can have a component towards Vcc
17:48:20[IDC]Dragonperhaps a pullup
17:48:42amiconn_Either a resistor or a diode (some chips employ external diodes)
17:48:53[IDC]Dragon(4) can have a component towards the coils, perhaps an external diode
17:49:05amiconn_Ah, yes
17:49:44amiconn_oil is to be placed between the 2 large pads below the chip.
17:49:56[IDC]Dragondid you notice the wireinng for lImbus is different?
17:50:32[IDC]Dragonthe EL foil seems to go to the bottom pin of each row
17:50:47amiconn_Ah, I did not look for that. That looks much more like SP4403 (EL between 4 and 5)
17:52:49amiconn_There are pads for small components that go between each end of the foil and the chip. What could that be?
17:52:50[IDC]Dragoncoil is pin 3
17:53:08[IDC]Dragondunno, maybe current limiting resistors?
17:53:40 Join ^ShAtZMaN^ [0] (
17:53:50[IDC]Dragon(as a safety feature)
17:56:07[IDC]Dragonlooks like Archos tried to shield the digital part, probably to quieten the tuner
17:56:31[IDC]Dragonunder the LCD is more than just white
17:57:07[IDC]Dragonlooks like something is connected to the upper edge
17:57:20[IDC]Dragonmaybe a shield
17:58:48amiconn_There is much more shielding in that unit than in yours. The samsung tuner board looks a bit strange...
17:59:11[IDC]Dragonwith the air coils?
17:59:52[IDC]Dragonthe philips tuner is more sophisticated, doesn't need such external parts
18:00:13[IDC]Dragonbut could also use some shielding...
18:00:14amiconn_Yes, both the air coils and the non-smd ceramic capacitors
18:00:29[IDC]Dragonthose are filters
18:00:59amiconn_Ah, yes
18:02:17[IDC]Dragonwhy is the first pic duplicated as the last?
18:02:29amiconn_I dunno
18:02:37^ShAtZMaN^hello everyone =]
18:02:56amiconn_There is still the question which EL driver chip goes into the empty space.
18:03:46amiconn_I found a comprehensive data sheet collection on the web (which does not require login). I'll look through the EL driver data sheets (there are >100 of them), perhaps I'll find the matching one
18:03:48^ShAtZMaN^i wanted to see if i can help... so i was told to set up the enviroment on my comp... i did that but i got the archos sim while i wanted the iriver one... even though i chose iriver on the configure... help?
18:03:57[IDC]DragonI don't know. Perhaps SP4403 for that old one, a different for ours
18:04:27^ShAtZMaN^sorry for my english.=\
18:05:35[IDC]Dragonamiconn_: where is that collection?
18:06:07^ShAtZMaN^i did that...
18:06:14***Saving seen data "./dancer.seen"
18:06:46^ShAtZMaN^the doc's only address the archos simulator can you please address me to the iriver doc's?
18:07:48amiconn_[IDC]Dragon: Sorry, I can't look it up atm (have the link on my home box, which I can't reach atm).
18:07:54[IDC]Dragon^ShAtZMaN^: run the configure script, it should give you a choice of iriver
18:08:56^ShAtZMaN^[IDC]Dragon; i ran it... and chose iriver. but still after the "make" when i run "./rockboxui" i got the archos one. the archos picture and all.
18:09:35[IDC]Dragon^ShAtZMaN^: dunno, I never did (no iriver people here atm)
18:09:50^ShAtZMaN^[IDC]Dragon; Maybe i need different CVS's ?
18:10:13[IDC]Dragondo you use linux or cygwin?
18:10:25^ShAtZMaN^i have a running linux system... will it easy're there ?
18:10:34^ShAtZMaN^im usind cygwin at the moment
18:10:54[IDC]Dragonperhaps, the sim was adapted by a linux guy
18:10:57^ShAtZMaN^i have MDK10... if it matters...
18:11:08^ShAtZMaN^i'll check that... 10x
18:11:16^ShAtZMaN^bye for now. =]
18:11:26[IDC]Dragonso the Windows version may not support iriver
18:11:52 Quit ^ShAtZMaN^ ()
18:12:58[IDC]Dragonamiconn_: I did a comprehensive button table
18:18:04[IDC]Dragonleaving this place...
18:18:08 Quit [IDC]Dragon ("CGI:IRC")
18:24:55 Join scott666_ [0] (
18:24:56 Quit Quel|away (Read error: 54 (Connection reset by peer))
18:28:26 Quit amiconn_ ("quite quiet")
18:29:00 Join oxygen77 [0] (
18:53:33 Join Tang [0] (~chatzilla@
18:57:27 Quit einhirn (Read error: 104 (Connection reset by peer))
19:07:46 Quit AciD (Read error: 104 (Connection reset by peer))
19:08:17*oxygen77 is away: Chui pas là
19:09:49 Join AciD [0] (
19:11:53 Quit AciD (Read error: 104 (Connection reset by peer))
19:12:25 Join AciD [0] (
19:54:42 Part amiconn
19:58:37 Join amiconn [0] (
19:59:53 Quit AciD (Read error: 104 (Connection reset by peer))
19:59:59 Quit Tang ("Chatzilla 0.9.66 [Mozilla rv:1.7.5/20041108]")
20:00:55 Join AciD [0] (
20:06:14 Join matsl [0] (
20:06:15***Saving seen data "./dancer.seen"
20:39:56 Quit AciD (Read error: 104 (Connection reset by peer))
20:58:16 Join AciD [0] (
21:00:00 Join einhirn [0] (
21:00:23 Join Digital007 [0] (
21:00:51Digital007hi Dwihno
21:01:05Digital007whats happening?
21:01:27Digital007There was some aspects of the Archos i miss in the iRiver
21:02:16Digital007like the on-the-fly playlists
21:02:33dwihnowell, there might be rockbox for the iriver too :)
21:03:13Digital007MIGHT? should be a resounding WILL!
21:34:43 Quit AciD (Remote closed the connection)
21:36:54 Join AciD [0] (
21:37:59 Join [IDC]Dragon [0] (
21:38:12amiconnhi again
21:38:23[IDC]Dragonhi there
21:38:48[IDC]Dragonwhat's happening?
21:39:11amiconnI try to debug talk.c :-/
21:39:29[IDC]Dragonthe missing start?
21:39:45amiconnOne problem (and cause for swallowed clip beginnings) already solved, see my latest commit
21:39:54*[IDC]Dragon looks
21:40:11amiconnNow there is another one, introduced with the silence scheduling
21:41:22amiconnshutup() now works correctly for all regularly scheduled clips, so no more swallowed beginnings. However, shutup() doesn't work for the added silence, since this is not scheduled into the queue...
21:41:54[IDC]Dragonoh, yes
21:42:40amiconnSo if new clips arrive while the silence is playing, the silence gets stopped mid-frame, and the next clip start is swallowed
21:43:05amiconnI wonder why you don't simply schedule the silence into the queue
21:45:38[IDC]DragonI didn't want to schedule within the interrupt
21:45:53[IDC]Dragonbut perhaps I should split that code
21:46:06amiconnSomething like this: if (is_voice_clip(last_clip) && last_clip != silence) schedule(silence);
21:46:41[IDC]Dragonand schedule needs to be a small, efficient worker function
21:46:56[IDC]Dragonnot the full-blown thing
21:47:04amiconnWhy not simply call queue_clip()
21:47:22[IDC]Dragonexcuse me for some time
22:06:17***Saving seen data "./dancer.seen"
22:18:10 Quit AciD (Read error: 104 (Connection reset by peer))
22:19:38 Join AciD [0] (
22:39:03 Quit methangas (" HydraIRC -> <- :P")
22:45:09[IDC]Dragonback again
22:45:23amiconnI think I have a working solution...
22:45:40amiconn...a bit ugly (using goto) but simple
22:46:02[IDC]Dragonlike what?">
22:51:59[IDC]DragonI see
22:53:09[IDC]DragonI'd prefer to extract the "heart" of queue_clip() and call it from there
22:53:28[IDC]Dragonbut I'm not sure if it works out
22:54:06[IDC]Dragonbecause when the queue is found empty, it's too late for normal scheduling
22:54:17[IDC]Dragonsame problem you ran into
22:54:35[IDC]Dragonjust avoiding a few lines of duplicate code
22:55:09amiconnI don't know if it's worth to have a separate function for those few lines. The queue add and queue read could even be shortened to 3 lines of code, if the queue length matches a 2^n-1 value
22:55:46[IDC]Dragonit is a power of two
22:55:57[IDC]Dragonwhy -1 ?
22:56:11amiconnAh, not -1, I already thought of the mask
22:56:39amiconnqueue_write = (++queue_write) & queue_mask;
22:57:15[IDC]Dragonyes, I wonder why I haven't done it like that
22:57:27[IDC]DragonsinceI mandate 2 ^n
22:57:34amiconnA similar approach is also possible for other lengths than 2^n, however it requires a modulo division
22:58:02amiconn(much more expensive)
22:58:37[IDC]Dragonthen the if() limiter is better
22:58:52amiconnQUEUE_MASK is already there
23:00:30[IDC]DragonI'm still puzzled about that frame detection bug
23:00:50[IDC]DragonI swear I have carefully checked its operation
23:01:22[IDC]DragonI made a debug list, tracking from where to where was played
23:01:40[IDC]Dragonhad a debug entry to dump all that into a file
23:01:52[IDC]Dragonand verified the file offline
23:02:14amiconnIt does not bug in from where to where it played, but with storing the previous frame header, causing the header search in shutup() go wrong
23:02:43[IDC]Dragonwhich determined the end point
23:03:02[IDC]Dragonwhich was part of my debugging, too
23:03:23amiconnFrom checking the disassembly, it is clearly visible. *start[1] first resolves the subscript, then dereferences. Adding the parentheses reverses that operation
23:03:46[IDC]Dragonyou can use p_lastclip[n]
23:03:48amiconnI caught that bug with the catch mem accesses debug feature
23:04:03[IDC]Dragonnice feature ;-)
23:04:24amiconnYup. I wonder why noone else than me seems to use it though
23:04:50[IDC]Dragonbecause it's not on by default?
23:05:29amiconnThat was impossible until a few days ago. The playback engine caused loads of hits
23:05:40[IDC]DragonI know">
23:07:58[IDC]Dragonis it really done 3 times?
23:08:08amiconnyes, once per line
23:08:13[IDC]Dragonor does the compile re-use *start ?
23:08:46[IDC]DragonI was expecting better
23:08:47amiconnNo, it doesn't. r4 is start
23:09:10[IDC]Dragonanother reason to use p_lastclip[n]
23:09:18amiconnThat's why I suggested -O2 some time ago. Iirc this solves a number of those stupid redos
23:09:19[IDC]Dragonor memcpy()
23:09:42[IDC]Dragonah yes, we still don't have that
23:09:53[IDC]Dragonit there a sane reason?
23:10:16amiconnI don't know. Perhaps I should try it
23:10:40[IDC]Dragonhow about -Os ?
23:11:08[IDC]Dragonmin size is what we really want
23:12:36amiconnI think we want both a small size and a decent speed. I wouldn't use -Os for normal building, the only reason for -Os would be that it could make rombox fit on the FM devices
23:13:08amiconnBtw, did you notice my disasm snippets were from rombox?
23:13:39[IDC]Dragonnow that you tell...
23:14:50amiconnThe somewhat ugly goto could be avoided if the code block within if (QUEUE_LEVEL) {...} gets duplicated instead.
23:18:14amiconnusing p_lastclip[n] instead of (*start)[n] doesn't generate much better code
23:25:59amiconn-O2 caused an IllInstr exception :(
23:26:42[IDC]Dragonoh, I thought we already compiled it like that
23:26:50[IDC]Dragonsomeone did...
23:26:58[IDC]Dragon-Os, iirc
23:29:59amiconnWith -Os, jpeg.o doesn't link... and it crashes as well
23:31:04[IDC]Dragonsuboptimal, indeed
23:37:44amiconnI tried changing the queue pointer arithmetics as suggested. Now I get 3 warnings from gcc :(
23:37:57amiconn"talk.c:203: warning: operation on `queue_read' may be undefined"
23:38:56amiconnAh, easy to avoid :)
23:44:12amiconn[IDC]Dragon:"> (again). What do you think?
23:48:21[IDC]Dragonlooks good, yes
23:49:36[IDC]DragonI leave it in your capable hands tonight
23:49:59amiconnNite [IDC]Dragon
23:49:59[IDC]Dragon(and thanks for debugging!)
23:50:08 Quit [IDC]Dragon ()

Previous day | Next day