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 2008-10-05

00:01:11ameok, some one can help me, i've install rockbox and now, my Ipod Video 30GB can't connect to my computer
00:03:08 Quit freqmod_qu (Read error: 104 (Connection reset by peer))
00:03:53bluebrothercan't connect in what sense?
00:04:37amehum it's not detected by my computer, or any other computer
00:04:56bluebrotherwell, is Rockbox running? Does it reboot into disc mode? What exactly is the ipod doing?
00:06:07amei'm french, so excuse my english :/ Rockbox is running, perfectly, but it can't be reboot into disc mode
00:06:48bluebrotherok, so when connecting to the PC does anything happen? Btw, I'm non-native too (not french though)
00:07:11smacnaythe sandisk seem to be well supported (dual boot) and affordable + they can use the sd cards.
00:08:56 Join freqmod_qu [0] (i=quassel@2001:700:300:1430:213:d3ff:fee9:5ed0)
00:09:04amethe ipod screen show the batterie's level and it show that the batterie is charging... the ipod does not detect the computer
00:11:17bluebrotherand your sure the cable works? It seems the Ipod is detecting your PC as charger
00:12:03ameyes my PC is detecting as charger, i try an other cable, i'll come back
00:12:36bluebrotheralso, have you tried resetting your settings? Not sure if you can disable the reboot permanently
00:13:02bluebrotherand if you hold menu while plugging the ipod it won't reboot into usb mode but only charge −− that's a feature
00:13:32 Join funman [0] (n=fun@
00:14:39 Join webguest1 [0] (n=6391efa2@gateway/web/cgi-irc/
00:14:54webguest1ne one have a snisk c200
00:15:32mcuelenaerewebguest1: Please use real English words..
00:15:44webguest1ne one have a sandisk c200
00:15:56webguest1sry im using a new keyboard
00:16:24*bluebrother recommends proof-reading before hitting enter
00:16:29mcuelenaerejust ask your question, then someone perhaps might be able to help you
00:16:47amebluebrother : no, my second cable (a new cable) do the same effect as the first
00:17:20bluebrotherwell, you can always put it into disc mode manually. What build have you installed?
00:17:21 Quit bimbel ("Woah!")
00:17:30webguest1i want to put rockbox on my microsd card will rockbox still work
00:17:48bluebrotherwebguest1: no. Rockbox has to be on the internal flash
00:18:09alleleI had a similar thought webguest1, I was thinking just putting plugins on flash though
00:18:16 Join dabujo [0] (
00:18:31allelesince plugins can be run from the filebrowser
00:18:59webguest1ok , but will the playlist show up if i just extract the folders
00:19:14bluebrotherwhat playlist?
00:19:33*bluebrother looks after the crystal ball
00:21:13amebluebrother : in the menu : info rockbox : Version r18551-080919
00:21:15funmanbluebrother: 42
00:21:27 Join bitmouse [0] (
00:21:32amebut i'm not sure that is the build version:/
00:21:45bluebrotherame: that's heavily outdated. Please update your build
00:21:53bitmouseis there anyway to get rockbox to load apple's IpodOS as the default OS?
00:21:56bluebrotherthat is the build version ;-)
00:22:07bluebrotherbitmouse: no. There is no point in this anyway
00:22:30bluebrotheras appleos won't shutdown the player but just sleep it. Thus to go back to Rockbox you need to hard reset it.
00:22:43amebluebrother : well, ok but how can i update it without a computer acces ?
00:23:04bitmouseright, but doing so would make my girlfriend more tempted to try rockbox
00:23:32bluebrotherame: put the ipod into emergency disc mode ...
00:23:47amehow can i do that ?
00:24:16Grundybeari think the rockbox playlist corrupted my sansa's internal flash :(
00:24:39bitmouse: (
00:24:58bluebrotherame: the IpodFAQ describes it ...
00:25:04BagderGrundybear: a _playlist_ corrupted the flash?
00:25:15bluebrotherGrundybear: what makes you thinking so?
00:25:16allelegrundy, you didn't defrag your player, did you???
00:25:20bitmousehm, is there anyway to turn of voiced menus from within rockbox itself?
00:25:24ameok thank you bluebrother i'll look at the FAQ ;)
00:26:01Grundybearallele how do you defrag?
00:26:16alleleno don't defrag grundybear
00:26:26allelebut similar problems have arisen from doing so
00:26:28linuxstbamiconn: The license. The one I found looked to be equivalent, and was GPL'd.
00:26:50allelewhat suggests that the flash would be corrupted though?
00:26:56Grundybearanyway i created a new playlist, and it went through adding all the songs, but after it was finished my sansa locked up
00:27:14Grundybearso i let it sit for a while thinking it might be doing something
00:27:37Grundybearbut it sat forever and still locked, so i did the 15 second reboot thing, and ever since then windows won't read the drive anymore
00:27:55amebluebrother : thank you very much, my Ipod is now detecting ;)
00:28:18amegood bye and thanks again ;)
00:28:36Grundybearand rockbox won't play any songs without locking up
00:28:45 Quit ame ("leaving")
00:30:06funmanGrundybear: try scandisk on the sansa drive
00:30:48 Join saratoga [0] (n=41becb3b@gateway/web/cgi-irc/
00:31:27Grundybearwon't run
00:31:38Grundybearwindows shows it as having 0 bytes out of 0 bytes
00:32:18Grundybearchkdsk won't run either
00:32:50Grundybear"Cannot open volume for direct access."
00:33:07 Quit funman (Read error: 113 (No route to host))
00:33:54 Quit webguest1 ("CGI:IRC")
00:34:25 Quit bluebrother ("leaving")
00:34:48linuxstbGrundybear: Which Sansa do you have?
00:35:56druidudoes anybody have FWDN?
00:36:28 Join Twisty [0] (
00:36:39linuxstbGrundybear: Then I would put it in recovery mode, and reformat by creating a file called sansa.fmt in the recovery drive.
00:36:41 Join taurelilomea [0] (
00:37:00*gevaerts was just going to suggest that, but linuxstb got there first
00:37:20Grundybeardoes sansa.fmt have to say anything?
00:37:25Grundybearor just the filename
00:37:58linuxstbJust the filename
00:37:59gevaertsJust an empty file, called sansa.fmt
00:38:16gevaertsBe careful you don't end up with sansa.fmt.txt or something like that
00:39:06 Join funman [0] (n=fun@
00:39:44taurelilomeaHey #rockbox, I have a question about m3u handling... does rockbox pay any attention to the song length in the playlist itself?
00:39:45Grundybearwhat is this version.txt ?
00:40:07gevaertsGrundybear: you can ignore that. It contains the version of the OF
00:40:16Grundybearbut it's empty
00:40:23 Quit allele ("Java user signed off")
00:40:24 Quit Tetracomm (Read error: 110 (Connection timed out))
00:40:39gevaertsI'd still ignore it for now
00:40:48 Join Tetracomm [0] (n=nicholas@
00:40:51gevaertstaurelilomea: I don't think so
00:41:42funmantaurelilomea: m3u doesn't specify song lengths, only filenames
00:41:42taurelilomeaAlright, cool. I'm making a script to convert xspf to m3u, but the numbers in my xspf files are all wrong. Thanks!
00:42:17taurelilomeafunman, I've never had to use m3u before I got rockbox, but says it does. :-/
00:42:35rasherfunman: extm3u does
00:42:39funmanthat is extended m3u, I'm not sure if there is a common format
00:42:46taurelilomeaIn any case, I'll just use a placeholder. thanks, gevaerts, funman, I'll tell you if it works.
00:42:55rasherfunman: Winamp writes them, that pretty much settles it
00:43:05Grundybearit worked
00:43:18Grundybearlooks like i'll have to reinstall rockbox and everything though, but hey it's working now
00:43:21Grundybearthanks for the help
00:43:27funmantaurelilomea: you can simply use filenames, this will make a valid m3u playlist, without extension
00:44:13linuxstbdruidu: Whilst searching for fwdn.exe, I found this forum thread - it contains a SDCFG value for the YP-P2 -
00:44:29Grundybeari wonder what caused it though
00:44:38Grundybearit was hanging on something like 'writing m3u file'
00:44:43taurelilomeafunman: so just /album/blah.flac, newline, /album/blah.flac, newline? that's it?
00:44:46Grundybearand it sat there for a good hour
00:45:10funmantaurelilomea: exact, quite simple ;)
00:45:21linuxstbdruidu: Forget that, you've already told me that doesn't work.... (and it's in tcctool)
00:46:16taurelilomeathat makes my job quite easier, funman. one last question−−how does rockbox like them formatted, e.g. back or front slashes, whole or partial directories?
00:46:35gevaertsGrundybear: you had some nasty file system corruption. That could have been there from before the m3u writing, and only made worse by it. It's not easy to know where that comes from
00:47:17taurelilomeaactually, lol, I can find out myself, just have to make a playlist in rockbox...
00:50:06linuxstbdruidu: It seems to be linked from this forum thread -
00:51:52 Quit Twst (Read error: 110 (Connection timed out))
00:53:25 Join bitmouse1 [0] (n=taylor@
00:53:32 Quit bitmouse (Read error: 104 (Connection reset by peer))
00:53:53***Saving seen data "./dancer.seen"
00:54:39 Part bitmouse1
00:55:53 Quit petur ("Zzzzz")
00:57:32 Join mr_broken_gigabe [0] (i=180a8a6f@gateway/web/ajax/
00:58:46 Part mr_broken_gigabe
01:00:47 Join Akira_Scythe [0] (n=18630b99@gateway/web/cgi-irc/
01:01:06Akira_ScytheOh hold on BRB
01:02:44 Quit ender` (" Information travels more surely to those with a lesser need to know.")
01:05:10 Nick fxb is now known as fxb__ (
01:08:01Akira_ScytheI need some technical help with my sansa e250
01:08:36blkhawkAkira_Scythe: Sandisk tech support might help you?
01:09:01Akira_Scythethis is about Rockbox on my e350
01:09:19blkhawkahh - who would have guessed
01:09:26Akira_ScytheSO can you help me, it's about the original firmware
01:09:29blkhawkwhile i have an e280
01:09:37blkhawki might not be the right guy to help
01:09:49blkhawkbut tell me about it anyway
01:09:49Akira_Scytheit don't matter
01:09:57Akira_Scythethere the e200's series
01:10:02Akira_Scythewell kk here i go
01:10:06blkhawkis it a v1 or a v2?
01:10:34*gevaerts looks at 20 lines of backlog without any actual question
01:10:44Akira_ScytheWell when I connect my sansa250 to my laptop it said it could not load or find the original firmware
01:11:02linuxstbAkira_Scythe: What says that?
01:11:20linuxstbAh, the Rockbox bootloader will say that.... Has it always said that?
01:11:22Akira_ScytheMy thingy when I connect it, I guess it's the bootloader
01:11:31linuxstbYour thingy?
01:11:33 Join mf0102 [0] (
01:11:44Akira_ScytheAnd then i started without connectin it to my pc
01:11:54Akira_Scythewhile pressing the left button
01:12:05Akira_ScytheIt said the same thing could not load original firmware
01:12:19linuxstbHas it always said that?
01:12:23Akira_Scytheit said could not find system/0F.bimn
01:12:44Akira_ScytheWell....this is the first time i tried rockbox
01:12:58linuxstbIs it an e250R or a normal e250?
01:13:05Akira_Scythenormal e250
01:13:19linuxstbAnd how did you install Rockbox? WIth rockbox utility?
01:13:42 Join Skail [0] (
01:13:46Akira_ScytheBut i instal the bootlaoder automaticly
01:13:48linuxstbSo exactly what did you do?
01:14:09Akira_ScytheI just Moved the .rockbox to he folder
01:14:19linuxstbI'm guessing you copied a Rockbox bootloader mi4 to the device and used the OF to upgrade to it?
01:14:52Akira_ScytheNo no I manualy instaalled the rockbox and automaticly installed the bootloader
01:14:52saratogaAkira_Scythe: you did more then that . . .
01:15:33*Akira_Scythe Akira wonders why they lookin at him so wierd
01:15:41linuxstbAkira_Scythe: Please try and be specific with your answers. What does "automaticly installed the bootloader" mean? What did you do _exactly_ ?
01:15:57Akira_ScytheDo u know the automated installer?
01:16:16Akira_ScytheNot the utility but the automated installer that has complete and small installation
01:16:30saratogai don't either
01:16:31Akira_Scytheand it has tabs for games, fonts, and themes
01:16:36saratogabe specific, use names
01:16:40saratogaand words about things
01:16:41linuxstbWhere did you download it from?
01:16:42Akira_ScytheLeme get link
01:17:14Skailhey guys, what's the key for power/menu on the simulator?
01:17:35linuxstbSkail: Try running the simulator with the "−−background" option. That shows you the buttons.
01:17:56 Quit faemir (Remote closed the connection)
01:18:06Akira_Scythe it even has a pick of it
01:18:25Skaillinuxstb: ahh... thanks.
01:18:30saratogaso you used rockbox utility
01:18:46Akira_ScytheYEs to ainstall the bootloader
01:19:20linuxstbAkira_Scythe: Does your sansa display any other messages before saying it can't find System/OF.bin ?
01:19:36Akira_Scythelet me reboot it and check it out again
01:20:07Akira_Scythecan not load from parrtion
01:20:11Akira_Scythebad check sum
01:20:37Akira_Scythetrying system/of.mi4 could not load
01:20:50Akira_Scythetrying system/0f.bin could not load
01:21:07linuxstbThat first error is the important one - "can not load from partition". It means that your firmware partition is corrupted somehow.
01:21:17Akira_ScytheOH FUCK
01:21:23linuxstbYou should enter recovery mode and put an original firmware mi4 file back.
01:21:28Akira_ScytheI hope i still can use my sansa and take of rockbox
01:21:34linuxstbNo need to panic - Sansas have 101 ways to recover
01:21:37Akira_ScytheHow do i do recovery mode
01:21:42Akira_ScytheTHANK YOU LOARD JESUS
01:22:14linuxstbI've no idea, I don't own a sansa, but ask google.
01:22:14Akira_Scythehow do i do this
01:22:42Akira_Scythegoogle =p um that's gona take a while and idk what to even search
01:23:01linuxstbHow about "sansa recovery mode" ?
01:23:02saratogause real words
01:23:42 Join faemir [0] (
01:23:44 Join dandin1 [0] (
01:25:55*amiconn spotted a possible optimisation in the libdemac entropy decoder, but unfortunately it applies only to 24 bit streams
01:26:03amiconnSaves a division though
01:26:25 Quit GodEater_ (Client Quit)
01:27:05amiconnlinuxstb: Are there 24 bit test files?
01:27:08taurelilomeagevaerts, funman, with the exception of a few unicode problems that aren't your problem, it worked perfectly. I'd lurk, but I've gotta go. Thanks again!
01:27:13 Quit taurelilomea ("Leaving")
01:28:41Akira_Scythenow it says Enter USB2.0 MSD MODE
01:28:47linuxstbamiconn: No, I don't think so. In the past, I've downloaded things from the live music archive - (enter "flac24" as a search term)
01:29:00amiconnah thanks
01:29:28Skailthanks to rockbox, the sansa's are a pretty good player for the price
01:29:33amiconnlinuxstb: Btw, the possible optimisation is in entropy.c line 181
01:29:47Skailmine lacks a reset button, though. which can be annoying.
01:30:02amiconnbase_lo = range_decode_culfreq(1 << lo_bits); ==> base_lo = range_decode_culshift(lo_bits);
01:31:25amiconnrange_decode_culfreq() does a division while range_decode_culshift() does a shift, and x / (1 << bits) == x >> bits
01:34:23linuxstbamiconn: Yes, nice spot.
01:34:56 Quit faemir (Remote closed the connection)
01:35:27 Quit {phoenix} (Remote closed the connection)
01:37:43 Quit Skail ("Concoction recent by Pier.")
01:45:11 Join ameyer [0] (
01:46:24 Join mc2739_ [0] (
01:50:42druiduI GOT IT!
01:50:49druidufound the sdcfg for X20!
01:51:23 Quit bertrik ("Leaving")
01:52:41 Quit Grundybear ("There are many but none in CANADA")
02:00:54linuxstbdruidu: Congratulations. What is the full line to add to tcctool.c
02:02:24 Quit mc2739 (Read error: 110 (Connection timed out))
02:02:46druiduI have confirmation!
02:03:00druiduI modified the in the X20.hex file into
02:03:23druiduthen booted the firmware, held the battery with my hand, unplugged the usb cable and voilla, system info shows new url
02:03:41druidu {"x20", "iRiver X20", 0xb051, 0x20000000, 0x02e92010 },
02:04:53druiduthey didn't follow the datasheet, 0x22e92010 doesn't work, although that's the value that should be for 128 MB as they say
02:05:23advcomp2019druidu, nice.. have you checked the insignia DV firmware because the insignia dv series is the same hardware as the iriver x20
02:05:44advcomp2019minus the mic tho
02:09:50 Join AndyIL [0] (i=AndyI@
02:10:10linuxstbdruidu: A new USB PID as well? I assume you had to change tcc.inf (if you're on Windows)?
02:20:36druiduyes, I changed tcc.inf
02:20:40funmanlooks photoshopped
02:20:46 Quit AndyI (Read error: 110 (Connection timed out))
02:20:52druiduthe PID is also in the wiki page
02:21:28druidufunman: you try holding the battery with one hand and making a picture with a webcam with the other :P
02:21:59*linuxstb wonders how changed to (more characters), and so also suspects photoshop ;)
02:22:36funmanlet's ask the support hotline
02:22:38druidunope, there are exactly two nulls after in the firmware
02:22:45druiduprobably padding
02:23:13funmandon't exclude software bugs
02:23:28 Quit nuonguy ("Leaving")
02:23:31saratogais the X20 the same CPU as the Iaudio7?
02:23:41linuxstbNo, it's the tcc8200
02:23:49linuxstbiaudio7 has tcc77x, and D2 has tcc7800
02:23:59linuxstbThe TelechipsInfo wiki page explains all...
02:24:07druiduARM946ES, 200 MHz, 0.13 μm
02:24:51linuxstbBut all three are very similar - code for one will either work as-is, or is easy to adapt to another.
02:24:55druiduis there any way to "escape" from usb msc mode without unplugging the usb cable?
02:25:12saratogaso the 7800 is actually faster then the 8200?
02:25:18 Join newbie [0] (n=18ec7c6f@gateway/web/cgi-irc/
02:25:22druiduif I unplug it I loose power
02:25:23advcomp2019druidu, here is what i am talking about with the insignia dv and iriver x20:
02:25:49 Join nuonguy [0] (
02:25:57funmanMaybe manufacturers reduce CPU frequency when they can achieve desired features with peripheral chips ?
02:26:10newbieIt seems that the sims at"> are missing the rockboxui.exe file ??
02:26:16druiduadvcomp2019: thanks, I probably read all the X20 hacking threads there are anywhere on the net
02:26:46druiduso the checksum is only checked when installing flash via normal way?
02:27:27rashernewbie: interesting, I'll look into it
02:27:30funmanBagder: do you welcome patches to ?
02:27:53Akira_Scytheang i still haevn trouble
02:28:07Akira_Scythebut i think i almost there
02:28:08linuxstbBagder: He will be sleeping now, but my guess is that he would.
02:28:17linuxstb(I mean funman:...)
02:28:34druidulinuxstb: do you want a patch for tcctool to allow setting pid and sdcfg in the cmdline?
02:28:41Akira_ScytheAfter This I swear i will never download rockbox again
02:28:49linuxstbdruidu: Did you see my PMs?
02:28:52newbieBagder: I've tried about 4 different sims - same issue with all
02:28:59druiduno, sorry
02:30:08 Quit kugel ("ChatZilla 0.9.83 [Firefox 3.0.3/2008092510]")
02:30:54 Quit Bjoern-Erik (Remote closed the connection)
02:31:16 Quit newbie ("CGI:IRC")
02:32:31 Quit bughunter2 ("bye")
02:33:05 Quit |AhIoRoS| (Remote closed the connection)
02:33:43 Nick JdGordon|zzz is now known as JdGordon (n=jonno@rockbox/developer/JdGordon)
02:35:12 Quit saratoga ("CGI:IRC")
02:35:56 Join perrikwp [0] (i=4aa794a0@gateway/web/ajax/
02:37:24linuxstbdruidu: So are you interested in porting Rockbox to the x20?
02:37:55druiduwell, I'm thinking of it... for the moment I want to get some basic code running
02:37:57 Quit dabujo (Read error: 104 (Connection reset by peer))
02:39:01druiduto be honest, I was thinking of porting Android on it, but the first part is common for any port −− get lcd to work, get flash to work etc
02:39:29Akira_ScytheGuys were on the sansa can i check my sansa version
02:39:38 Join MethoS- [0] (
02:40:18Akira_ScytheGuys were on the sansa can i check my sansa version
02:40:20linuxstbdruidu: I thought Android was a mobile phone platform? Is it more than that, or are you just making it more?
02:40:49druiduspeaking of which: LCD is IZT2155714 and other stuff written on the ribbon: Varitronix FPC-12155_1-1_62 and a date: 2006.06.23
02:40:53 Quit druidu ("CGI:IRC")
02:41:00Akira_Scythecan anyone help me
02:41:02 Join druidu [0] (n=d5a4e705@gateway/web/cgi-irc/
02:41:32linuxstbAkira_Scythe: Patience...
02:41:42druiduI know, sounds silly −− porting Android, a mobile phone platform, to a device which has no phone support, no wifi, no bluetooth no nothing
02:41:46Akira_Scythe=p well everyones going blah bla blah
02:42:01Akira_ScytheI still need help after my sans is on recovery mode what do i do next
02:42:06druidubut I would learn a lot and would be really cool if I succeded
02:42:09Akira_Scythesansa e200*
02:43:33 Join Grundybear [0] (
02:43:40druidubasicly, Android is a platform and a really cool one, you can make it whaterver you want :)
02:43:55linuxstbAkira_Scythe: Patience... No-one here is being paid to give support, and it's the middle of the night in Europe, where a lot of us are. Eventually someone will come who can help you.
02:43:58Grundybearso, what media organizer do you guys recommend for keeping your media players synced and organized and stuff
02:44:17Grundybearsongbird looks promising, but it's pretty buggy currently
02:44:28Grundybearwinamp sucks at managing tags and things
02:45:06linuxstbGrundybear: Rockbox is independent of any PC software, so that's more a topic for #rockbox-community
02:45:14scorchedruidu: porting android wouldnt make much would have to change the entire input scheme, not to mention many other things...rockbox actually makes sense in my opinion..
02:45:46*ameyer wonders if Akira_Scythe read
02:46:13*Akira_Scythe wonders if ameyer is his savior
02:46:45druiduanyway, I found rockbox after I decided to port it to X20 and was researching, so who knows where it will take me? :P
02:47:39*Akira_Scythe gives ameyer an internet hug
02:47:42druidubut anyway, rockbox port is much easier to make than android port, so I'll probably start with it
02:47:42scorcheAkira_Scythe: unnecessary...
02:47:42linuxstbdruidu: You should do whatever you will find interesting - assuming you release your code under the GPL (or something compatible), someone else can always steal it and use it for a Rockbox port
02:48:11linuxstbBut Android does sound like an odd choice.
02:49:24ameyerlinuxstb: for gpl code, is "steal" really the right word</wording_police>
02:49:45druidubasicly, if I could boot android on it I could have everything else except whatever requires network / phone support −− single player games, music player, video player, etc... of course, input would be tricky but as I said, it would be cool
02:50:15druidujust like if it would be cool to put linux in your toaster or frigde −− if that could be possible
02:50:28linuxstbameyer: Technically, no, but I like to think of the English language as flexible...
02:51:00Akira_Scytheameyer were are u it worked
02:51:04druiduso, about the LCD... I couldn't find a datasheet, does rockbox already have something that could be useful?
02:51:27linuxstbI don't recognise the name.
02:51:42Akira_Scythethxs for the hell rockbox team =D
02:51:55druiduTFT QVGA 320x240, 2.2inch
02:52:26Akira_Scythebye scorche ameyer linux and druid
02:52:28 Quit HBK ()
02:52:31 Quit Akira_Scythe ("CGI:IRC")
02:52:36linuxstbdruidu: But the actual lcd controller is often hidden somewhere, and the usual way to find it is to reverse-engineer the OF to find the init sequence, and hope someone recognises it.
02:53:08druiduany patterns on that?
02:53:52funmanit looks like the standard capabilities of the screen, nothing specific to your model
02:53:54druidutill now the 0xF0000000 helps that I can find places where the tcc8200 is configured
02:53:55***Saving seen data "./dancer.seen"
02:54:09druiduin the disass of the firmware that is
02:54:50druidulcd controller is embedded in the tcc8200 chip, the specs say it can drive lcds
02:55:05linuxstbThen look for writes to those registers in the OF.
02:57:42funmanAm I mistaken when I think an interrupt can occur when the CPU is already handling a previously generated interrupt ?
02:57:44 Quit culture (Read error: 110 (Connection timed out))
02:59:05druiduthere is that stuff about non-maskable interrupts
02:59:11linuxstbfunman: I don't think so. Isn't that what interrupt priorities mean? i.e. interrupts of higher priority can interrupt the handler of a lower priority interrupt?
02:59:51 Quit ompaul (Client Quit)
03:00:03funmanall this is very complex
03:00:40funmannon-maskable interrupts seem to be the highest priority
03:01:33druiduso basicly, if you do a divide by zero in an interrupt handler, you would get another one right?
03:01:54funmanthat's how I understand it
03:02:20 Quit mf0102 ("Ex-Chat")
03:02:50funmanI'll read Linux code to get better understanding
03:04:00 Join |AhIoRoS| [0] (n=ahioros@
03:07:25 Join saratoga [0] (n=9803c6dd@gateway/web/cgi-irc/
03:09:21 Quit herrwaldo (Remote closed the connection)
03:16:35 Join phixxor [0] (
03:17:22 Part phixxor ("Leaving")
03:20:13 Quit ajonat ()
03:23:05 Quit saratoga ("CGI:IRC (EOF)")
03:25:53 Quit mirak (Remote closed the connection)
03:34:32druiduhow do I get ns-dvXg firmware?
03:34:53 Join mmadia [0] (
03:34:55druiduthey save an exe file here, I just need the binary file
03:36:33 Quit Schmogel (Read error: 104 (Connection reset by peer))
03:36:47 Join webguest05 [0] (n=cfff245c@gateway/web/cgi-irc/
03:38:22advcomp2019druidu, hold on i think i can get one from Tobey which is a the insignia moderator at anythingbutipod
03:41:32 Quit mcuelenaere ("Zzzzzzzzz")
03:44:02advcomp2019druidu, here are two file to 1.200 if that is what you need:
03:51:48 Quit DerDome (Read error: 110 (Connection timed out))
03:52:30 Join DerDome [0] (
03:56:37druiduthanks, but I managed to extract the binary image from the exe
03:56:50druiduit was in the resources section of the exe file
03:57:48druiduI managed to flash NS-DVxG-V1.200.rom and it works but no sound and screen is rotated 180 degrees
03:57:57druiduplus hold button is reversed
03:59:29 Quit Winkie ("brb")
03:59:51 Join bmbl [0] (n=Miranda@unaffiliated/bmbl)
04:01:31 Quit Thundercloud (Remote closed the connection)
04:02:41 Quit nuonguy ("Leaving")
04:06:53druiduv1.105.4 doesn't even boot
04:08:40 Join ejl [0] (
04:08:40 Quit EspeonEefi (Read error: 104 (Connection reset by peer))
04:08:43 Quit bmbl ("Woah!")
04:12:08druidu1.9 beta works too, still no sound, and looks like a iRiver UI clone
04:14:58 Join Rondom [0] (
04:15:37 Quit mc2739_ ()
04:15:48XavierGrFor anyone getting dizzy by forward scrolling lines FS #9455
04:16:40 Join Darksair [0] (n=user@
04:23:32 Quit webguest05 ("CGI:IRC")
04:30:26 Quit toffe82 (Read error: 104 (Connection reset by peer))
04:30:31 Join toffe82_ [0] (
04:30:41 Nick toffe82_ is now known as toffe82 (
04:30:43 Quit mmadia ("Vision[0.9.7-Z-101305]: i've been blurred!")
04:38:48 Quit jhulst (Read error: 110 (Connection timed out))
04:43:38 Join blkhawk- [0] (
04:47:52 Quit MethoS- (Remote closed the connection)
04:49:55 Quit Rondom (Read error: 110 (Connection timed out))
04:51:08 Join jhulst [0] (n=jhulst@unaffiliated/jhulst)
04:53:59***Saving seen data "./dancer.seen"
04:55:34 Join phinze [0] (
05:01:02 Quit blkhawk (Read error: 110 (Connection timed out))
05:01:36 Nick blkhawk- is now known as blkhawk (
05:01:51 Quit dandin1 ()
05:08:11 Join miepchen^schlaf_ [0] (
05:14:09 Quit coatman (Remote closed the connection)
05:15:44 Quit miepchen^schlaf (Connection timed out)
05:18:30 Join EspeonEefi [0] (i=espeonee@CERULEANCITY.MIT.EDU)
05:18:38 Join coatman [0] (
05:25:55JdGordonLlorean: hmm... just listening now and i tihnk its been made bleedingly obvious that we should have a "getting started" page somewhere on the frontpage
05:26:15 Quit agaffney (Read error: 104 (Connection reset by peer))
05:26:32 Join agaffney [0] (n=agaffney@gentoo/developer/
05:26:40 Quit funman ("leaving")
05:33:25 Quit |AhIoRoS| ("Abandonando, see you")
05:39:16 Join massiveH [0] (
05:40:57 Join massiveH_ [0] (
05:42:01 Quit fdinel ("Miranda IM! Smaller, Faster, Easier.")
05:42:07 Quit massiveH (Nick collision from services.)
05:42:09 Nick massiveH_ is now known as massiveH (
05:47:10 Join BHSPitLappy [0] (n=BHSPitLa@unaffiliated/bhspitmonkey)
06:02:22 Quit phinze ("leaving")
06:07:22JdGordonamiconn: is it 4 or less function parameters which can speed things up? or is it less than 4?
06:07:53JdGordon... or more to the point, should i bother trying to get the param list down from 5 or is there no point
06:11:45 Join nuonguy [0] (
06:13:39 Quit XavierGr ()
06:16:53 Quit Seed ("cu, Andre")
06:29:52 Join webguest19 [0] (n=45e82ce4@gateway/web/cgi-irc/
06:29:56webguest19may someone help me
06:30:11webguest19i dont know how to sync music onto my c240 rockbox
06:31:53wpyhBagder: I have submitted the playlist shuffle patch:
06:32:04wpyh(since you said you'd agree to it)
06:33:25 Quit webguest19 (Client Quit)
06:40:12 Quit Tetracomm (Read error: 110 (Connection timed out))
06:40:48 Join Tetracomm [0] (n=nicholas@
06:54:00***Saving seen data "./dancer.seen"
06:58:45 Quit JdGordon (Read error: 110 (Connection timed out))
07:00:15 Quit scorche (Read error: 104 (Connection reset by peer))
07:05:45 Join solrize_ [0] (n=phr@unaffiliated/solrize)
07:06:00solrize_if anyone cares there is a good deal on refurbed sansa e280's tonight
07:07:48massiveHand there goes his battery
07:13:21 Join HBK [0] (
07:16:21 Join scorche [0] (i=Blah@rockbox/administrator/scorche)
07:21:16 Quit Zarggg ()
07:23:33 Join alkos333 [0] (
07:23:48alkos333How do I remove iPod_Control after installing rockbox?
07:24:10alkos333I've tried mounting with root and rw options, but when I run rm -rf on it, it says mounted as read-only
07:25:09scorcheis the entire device read-only?...i have always been able to just delete it
07:25:30alkos333So have I
07:25:51alkos333Yes, apparently the entire device
07:26:45alkos333I just tried deleting the Playlists folder created by rockbox
07:27:56 Part solrize_ ("Leaving")
07:28:35 Quit druidu ("CGI:IRC (EOF)")
07:28:55 Part toffe82
07:30:03alkos333scorche: So what do I do?
07:32:16 Quit alkos333 (Remote closed the connection)
07:37:40 Quit Darksair (Client Quit)
07:41:19 Quit massiveH ("Leaving")
07:43:17 Join JdGordon [0] (
07:47:37 Quit setkeh (Read error: 110 (Connection timed out))
07:49:45 Part wpyh
07:51:13 Join alkos333 [0] (
07:51:25alkos333scorche: Sorry, had some problems here..
07:57:07 Join JdGordon_ [0] (
07:57:49 Quit coatman (Read error: 113 (No route to host))
07:58:54 Join kushal_12_27_200 [0] (n=kushal@
07:59:05Unhelpfulalkos333: could be permissions, could be filesystem corruption. linux forces fat filesystems to be remounted read-only if it detects corruption. see if there's anything logged via dmesg about fs corruption. if so, try fsck with the filesystem unmounted, and then mount it again.
08:02:28alkos333I remounted with -o rw
08:02:30alkos333It worked this time..
08:02:34alkos333Don't know why it didn't work last time
08:05:12 Quit JdGordon (Read error: 110 (Connection timed out))
08:05:21 Join J-23 [0] (
08:06:10 Join Darksair [0] (n=user@
08:07:14 Quit nplus (Read error: 110 (Connection timed out))
08:09:44 Nick JdGordon_ is now known as JdGordon (n=jonno@rockbox/developer/JdGordon)
08:20:36 Join ajonat [0] (n=ajonat@
08:34:08 Join devslashnull [0] (n=delvslas@
08:34:51devslashnullive got a 5th gen ipod video and theres a lag at playback. does it make a difference as far as playing speed if i add music thru itunes or drag and drop it onto my ipod
08:37:31devslashnullis anyone here
08:39:13TetracommI doubt it.
08:39:44 Join Rob2222 [0] (
08:43:36devslashnullive got a weird problem
08:44:07devslashnullon one of my albums the last letter is cut off for the artist name,album name and song title
08:46:11 Quit BHSPitLappy (Remote closed the connection)
08:49:31 Quit kkurbjun ("Lost terminal")
08:51:46 Join smacnay_ [0] (
08:54:05***Saving seen data "./dancer.seen"
08:58:09 Quit Rob2223 (Read error: 110 (Connection timed out))
09:02:09 Quit smacnay (Read error: 110 (Connection timed out))
09:07:38 Quit Tetracomm ("Get free, legal, and high quality downloads (including MP3 and OGG) from:")
09:11:49 Quit jfc (Read error: 104 (Connection reset by peer))
09:12:16 Join jfc [0] (
09:13:22 Join massiveH [0] (
09:20:14J-23Was Sansa bootloader bug (bad binary file type) fixed?
09:26:30amiconnJdGordon: It is 4 or less parameters which will help both ARM and SH1 (function will use regparms then)
09:33:48 Quit [omni] (Remote closed the connection)
09:40:49LloreanJ-23: Is this a bug that's in the tracker or something?
09:42:24 Quit jhulst (Read error: 113 (No route to host))
09:42:33J-23Llorean: no.
09:42:38 Quit devslashnull ("This computer has gone to sleep")
09:42:42J-23hmm, installing older bootloader helped
09:42:48J-23and stable Rockbox
09:43:14JdGordonamiconn: ok, ta
09:43:18LloreanJ-23: So, what exactly were you asking?
09:43:37LloreanMost devs won't know the bug exists if nobody's ever reported it, and if it were fixed you'd see a note about it in the changelog.
09:44:02JdGordonLlorean: hey, dunno if you saw my message before... that podcast I think made it obvious we need some sort of "getting started" page...
09:44:24*J-23 compiles Rockbox from SVN to check if this happens for newest revision too.
09:44:34J-23maybe it was already fixed.
09:45:13LloreanJdGordon: I think so too. Cut the "Releases" and "Current Builds" link and replace it with "Install"
09:45:30LloreanJ-23: We only recommend people use the bootloader the RBUtil installs.
09:45:53LloreanJ-23: we do also recommend people actually report bugs with newer code if they're experiencing them by actually documenting how to reproduce them in the tracker.
09:46:17JdGordonLlorean: yeah something like that, and with a walkthrough of using rbutil and hope people are smart enough to be able to follow it if the target being walkthroughed isnt the one they are using
09:46:56LloreanJdGordon: From "Install" it can have a brief "Getting Started" paragraph, then a link to RBUtil, and a link to the builds page after a description of why most users shouldn't want to go ther.
09:47:21JdGordonI think we need to start limiting how many links people need to follow
09:47:26J-23works properly with stable bootloader
09:48:29LloreanJdGordon: Well, overall that would be the same (or less) links than the current "RButil" install path.
09:48:37LloreanAnd it'd be one less link in the menu on the left.
09:49:24JdGordonI like the idea of 1 link from the main page to a full install walkthrough
09:55:23 Join globi [0] (n=Drk@2001:470:1f12:d3:0:0:0:2)
09:57:34 Join stoffel_ [0] (
10:03:35 Join fragilematter [0] (n=fragilem@
10:19:13 Join bmbl [0] (n=Miranda@unaffiliated/bmbl)
10:19:13 Quit nuonguy ("This computer has gone to sleep")
10:23:45 Join pixelma [0] (n=pixelma@rockbox/staff/pixelma)
10:24:57 Join ompaul [0] (n=ompaul@gnewsense/friend/ompaul)
10:25:11 Quit jeffdameth (Read error: 110 (Connection timed out))
10:25:26 Join jeffdameth [0] (
10:29:54 Join Skail [0] (
10:32:17 Quit miepchen^schlaf_ ()
10:40:33 Quit Grundybear ("There are many but none in CANADA")
10:42:49amiconnlinuxstb: I think I know why libdemac (especially entropy decoding) is still significantly slower on PP5002 even with the relevant functions in libdemac/ put into IRAM, and why it is so slow in general: The entropy decoder does several divisions per sample (for 3.98 streams, it's 3 divisions without overflow, and 5 divisions with overflow)
10:43:25amiconnAnd since ARM has no div instruction gcc uses a library function - which isn't put into IRAM on PP5002...
10:44:24linuxstbAh... Can that be fixed easily? e.g. are library functions in their own sections?
10:45:50amiconnNo, just .text
10:47:08amiconnI wonder whether the range decoder could be implemented in a way that it uses fewer divisions.
10:48:16amiconnThe generic ffmpeg range decoder uses none, but it decodes a bit at a time instead of multiple bits
10:50:18 Quit massiveH ("Leaving")
10:53:04 Join massiveH [0] (
10:53:43alkos333Ok, so I enabled the log, how do I use it?
10:54:06***Saving seen data "./dancer.seen"
10:57:05alkos333massiveH: Thanks
10:57:34alkos333Another question is regarding the artwork. All of my artwork is saved as /Artist/Album/cover.jpg
10:57:47massiveHpersonally, I suggest qtscrobbler
10:58:05alkos333I'm using the newClix red theme and for some reason instead of the artwork it shows Cover.bmp 404
10:58:18alkos333I tried renameing cover.jpg to cover.bmp, but that didn't work..
10:58:34massiveHyou can't rename a jpg file to .bmp and have it work
10:58:42massiveHyou need to convert the jpeg file to a bitmap file
10:58:57linuxstbalkos333: See for some suggested image converters
10:59:11alkos333massiveH: I don't use QT
10:59:26alkos333massiveH: But they all have to be in BMP, right?
10:59:57massiveHalkos333, I don't use coverart, so I'm not sure, but I do know they have to be a bmp of certian dimensions
11:00:20alkos333100x100 I believe
11:00:28massiveHand QT is a program that logs to
11:01:03linuxstbalkos333: The dimensions of the album are are specific to the theme - see that wiki page I linked you to...
11:01:09JdGordoneach wps uses a different size bmp... you can check the one your current wps is expecting in system > rockbox info
11:01:46 Join nplus [0] (n=nplus@141.25.Globcom.Net)
11:01:50alkos333JdGordon: linuxstb: Thank you guys!
11:03:06JdGordonis buffering:949-951 a possible deadlock? mutex_unlock() is called after rm_handle() which im pretty sure locks on entry?
11:03:12alkos333JdGordon: Hmm.. doesn't show anything artwork related in System > Rockbox Info
11:03:22JdGordonand if so, shouldnt mutex_unlock happen straight after load_bitmap()?
11:03:30JdGordonalkos333: you need a newer build then
11:04:11alkos333JdGordon: I got r18607 - 3.0
11:04:38JdGordonthen yeah, you need a current build
11:04:45JdGordonit was added shortly after 3.0
11:05:14alkos333I see
11:07:10alkos333JdGordon: So if I were to unzip the latest build, would that overwrite the theme/configuration files, or just the firmware itself?
11:07:36JdGordonjust the firmware
11:07:49JdGordonyour config/themes are safe
11:07:51alkos333JdGordon: Sweet :)
11:08:04massiveHas long as you don't delete anything you are safe**
11:08:51alkos333So where is this jpeg decoder plugin, so I could use jpeg artwork?
11:11:30linuxstbJdGordon: I was wondering about this commit -;r2=18688;pathrev=18688 Shouldn't last_screen be defined as a "signed char" ?
11:12:21JdGordondont they default to signed?
11:12:31 Quit Skail ("Concoction recent by Pier.")
11:12:50linuxstbThe C standard says chars can be either - it's up to the compiler.
11:13:05linuxstbOn ARM and (I think) MIPs, it's unsigned, on others, it's signed.
11:13:31JdGordonhmm, ok
11:14:20linuxstbAnd I seem to be missing something in that commit - isn't it just saying "if (x==-1) { x = -1}" ?
11:16:23JdGordonumm... it was supposed to fix crashes when the nvram was corrupt if start in screen was set to previous
11:16:31JdGordonI may have stuffed it though...
11:18:30linuxstbShouldn't you just check if last_screen contains a valid value, and set it to GO_TO_ROOT if it doesn't?
11:18:40alkos333So where's the JPEG decoder plugin?
11:19:21alkos333for the artwork that is
11:19:37pixelmaalkos333: it's a viewer (that's explained in the manual as well). You can use it to watch your cover jpeg but it is not used to show cover art in the WPS
11:19:38scorchelook in flyspray...
11:20:07alkos333pixelma: Ah, got ia
11:20:13linuxstbscorche: There's a patch for that?
11:20:32scorchelinuxstb: i misunderstood what he was asking for..
11:20:34pixelmamaybe a misunderstanding here
11:20:58pixelmascorche: I'm not sure anymore either
11:21:29alkos333Odd, I enabled the log, rebooted the ipod, then listened to 3 minutes of a song, then mounted the ipod and did ls -a , but .scrobbler.log is not there
11:21:48JdGordonwas the song more than half played?
11:22:14alkos333it was 4 minutes long
11:22:19alkos333I'll try again.
11:22:27JdGordonlinuxstb: hmm, yeah i tihnk your right... ill revert the whole thing and hope the crashing doesnt come back.,.. it shold work without it
11:22:31scorchewasnt there some bug with the first file?
11:23:57pixelmaI thought that was with "gather runtime data" but maybe they suffer both from it
11:28:12alkos333Is there any advantage to using the Database?
11:28:52alkos333Right now, I just do browse and the Database is not used at all.
11:31:09 Join JdGordon_ [0] (
11:31:47 Join stripwax [0] (
11:33:40 Quit globi ("Lost terminal")
11:34:18 Quit massiveH ("Leaving")
11:36:07alkos333I don't know understand why the scrobbler file isn't being created
11:36:13alkos333I played several songs from start to end..
11:37:27 Quit amiconn (" l8er")
11:38:05linuxstbalkos333: After you enabled it, how did you reboot your ipod?
11:40:28alkos333Menu + Select
11:40:48linuxstbThen check if it's still enabled - that's like pulling the power plug
11:41:05linuxstbYou should always shutdown rockbox cleanly (hold PLAY for a few seconds)
11:41:23alkos333it says no
11:41:25 Join faemir [0] (
11:41:40alkos333Oh.. well I see now
11:41:48alkos333I thought that's what they meant by reboot
11:42:04linuxstbNo, "menu+selectr" is known as a "hard reset".
11:42:20alkos333Ok, I see now.
11:42:38alkos333Ok, it says Yes now
11:42:55alkos333linuxstb: Have you ever used ?
11:46:38 Quit JdGordon (Read error: 113 (No route to host))
11:47:08 Join bertrik [0] (
11:52:02 Quit stripwax (Read error: 110 (Connection timed out))
11:55:24 Quit ajonat ()
11:59:36 Nick JdGordon_ is now known as JdGordon (n=jonno@rockbox/developer/JdGordon)
12:01:49alkos333So what's the purpose of a database?
12:02:55 Join culture [0] (
12:03:58 Join lasser [0] (
12:06:25 Join stripwax [0] (
12:12:06 Join J [0] (
12:12:34 Nick J is now known as Guest25178 (
12:24:16 Join hannesd [0] (
12:29:10 Part fragilematter
12:33:26 Join Bagder_ [0] (
12:37:18 Join herrwaldo [0] (
12:39:16 Quit JdGordon (Remote closed the connection)
12:41:10 Join JdGordon [0] (
12:49:33 Quit fyrestorm ("irc protip: press alt-f4 to open the super secret mirc console window!")
12:49:46 Join einhirn [0] (
12:53:55 Join fyrestorm [0] (
12:54:10***Saving seen data "./dancer.seen"
12:55:57 Quit Bagder (Read error: 110 (Connection timed out))
12:59:51 Nick Bagder_ is now known as Bagder (
13:06:34Bagderalkos333: you decide that yourself, use it or not
13:07:51 Join Tsukasa-Ujiie [0] (
13:08:11Tsukasa-Ujiiehey, is it possible to build rockbox using mingw?
13:08:23Tsukasa-Ujiiedidnt think so
13:09:57 Quit Tsukasa-Ujiie (Client Quit)
13:12:06linuxstbBagder: It's impossible? ;)
13:12:30Bagderyeah, the question should've been that ;-)
13:14:19*Bagder showed his human side and tried to answer to what I think he meant rather than the exact question he asked
13:15:38 Join Thundercloud [0] (
13:27:23 Quit Thundercloud (Remote closed the connection)
13:36:36Bagder"This article or section needs sources or references that appear in reliable, third-party publications."
13:36:48Bagderwikipedia comment on the Rockbox article
13:37:17BagderI don't understand for what statements they would "need" that
13:38:18scorchewhich comment?
13:38:27Bagderthe big one at the top
13:40:01LloreanI swear that wasn't there just a short while ago.
13:40:25scorcheoh...i was looking in discussion...well...time to take it there, i suppose
13:40:36Bagderyeah, nothing in discussion mentions it
13:41:57scorchehaha...though the last discussion topic is entertaining
13:42:11LloreanDiscussion is pretty entertaining on that page period.
13:44:14LloreanYou could just dump the contents of "Articles About Rockbox" on the discuss page and say "add references as you will"
13:47:26 Join ender` [0] (
13:49:26 Quit einhirn (Read error: 104 (Connection reset by peer))
14:02:42 Join GodEater_ [0] (n=ge@rockbox/staff/GodEater)
14:03:11 Join kugel [0] (n=chatzill@unaffiliated/kugel)
14:11:55 Join mf0102 [0] (
14:16:54kugelJdGordon: ping
14:17:49 Join {phoenix} [0] (
14:19:26 Quit {phoenix} (Remote closed the connection)
14:21:34 Join bluebrother [0] (n=dom@rockbox/staff/bluebrother)
14:29:27 Join bughunter2 [0] (n=Jelle@
14:29:49 Nick Platyna is now known as Padme (
14:31:26JdGordonkugel: pong
14:31:55kugelJdGordon: Just uploading v8 with minor changes and a some comments
14:32:18JdGordongonna be around for a bit?
14:32:50JdGordonhmm.. whats the fs#? its not in my email yet
14:34:19JdGordonarg.. misread uploading as uploaded :p
14:34:32*JdGordon rebooting into linux so we can commit this
14:34:41 Quit JdGordon (Read error: 104 (Connection reset by peer))
14:40:28 Join JdGordon [0] (
14:41:53kugelJdGordon: I've done uploading
14:41:59JdGordonlooking now...
14:42:46JdGordonjust on point 2... if you dont have the parent vp there it will be a better delta :p
14:43:18kugelWell, it still breaks the whole point
14:43:32kugelAlso, a parent is needed for fix_viewports
14:43:34JdGordonna, depends how the eventual custom vp patch happens
14:44:02kugeland for the icons (they are drawn into the parent directly)
14:45:23kugeland the icons are only drawn once (i.e. not in pitchscreen_draw, but at initializing)
14:47:16kugelWait I think I can find a way to "remove" the parent (still use it but use one of the pitchscreen vps instead)
14:48:01JdGordonna, its better how it is than with hacks
14:48:55 Quit stripwax (Read error: 104 (Connection reset by peer))
14:49:47JdGordonkugel: actually... i think pitch_mode should go back to being global
14:50:16JdGordonill fix that and commit
14:50:22kugelOk, pitch could global too then
14:50:41 Nick fxb__ is now known as fxb (
14:51:39JdGordonwas it global before?
14:52:54kugelI don't think so, but it's basically in the same situation (being remembered after quitting the pitchscreen, being a parameter of pitchscreen_draw)
14:53:55 Join miepchen^schlaf [0] (
14:54:11***Saving seen data "./dancer.seen"
14:55:22kugelAlso, globals aren't that bad imho :)
14:55:30JdGordondoes the hwcodec targets have the pitchscreen?
14:55:50bertrikI dislike global symbols
14:55:54kugelif they have HAVE_PITCHSCREEN #defined ;)
14:55:55JdGordondepends how and why, but yeah i was more concerend with having it global and being passed around... one or the other
14:56:47kugelbertrik: we're not talking about a symbol
14:57:48JdGordonkugel: the screen actually works as expected still yeah?
14:58:31kugelUhm, most likely ;) I could do a last test in the h300 sim
14:58:59kugelthe functionality doesn't change with my patch
14:59:09JdGordoni know... but just in case :p
14:59:50JdGordonrebuilding a full h300 sim so i can double check
15:02:03kugelI think my pitchscreen has some nice optimizations for small screens and remotes, that alone is worth committing :P
15:02:22JdGordonfinally :p
15:02:37Lloreankugel: "optimizations" how?
15:02:38JdGordonjust over 2 months... i thought it was longer than that :p
15:02:40kugelwooh great! cool
15:03:05bluebrotherglobals can be dangerous ...
15:03:10kugelLlorean: graphic'ish optimazations
15:03:14*pixelma gets scared
15:03:18Lloreankugel: That's a complete non-answer.
15:03:51kugelThe screen will omit some text which isn't necessary for the functionality if it doesn't fit, instead of overlapping
15:03:56bluebrother"break the GUI once more"?
15:04:44Lloreankugel: What text?
15:05:00 Quit miepchen^schlaf ()
15:05:07kugeldepends on the screen. The only text which is always there is the pitch value
15:05:22pixelmayes, hwcodec has the pitchscreen and I wonder if it was tested at least in a (bitmap) Archos sim
15:05:41*bluebrother wonders since when enums are named like defines
15:05:41pixelmaexcept the Player
15:05:56LloreanIt seems to me like it'd make more sense to default back to sysfont if everything won't fit.
15:06:22kugelLlorean: Well I had this idea too, but got said from several people "No"
15:06:49Lloreankugel: Are these in comments on the tracker? What FS task is it?
15:07:10kugelthe main reason was afaik that sysfont doesn't support all languages and pitchscreen is quite localized
15:07:18JdGordonthe whole point of this is to remove any sysfont fsallbacking...
15:07:23kugelall languages = all char sets
15:07:54kugelLlorean: no it was here in irc
15:08:07LloreanJdGordon: It seems to me, thought, "making the screen indecipherable" isn't much better.
15:08:23pixelmaI thought items should scroll if they don't fit, much like the quick screen?
15:08:38LloreanJust not showing them seems a pretty non-ideal solution.
15:09:12kugelLlorean: That's plain wrong. The icons will stay and the pitch value will stay. That's just enough to use the pitchscreen
15:09:43Lloreankugel: "Plain wrong"? Why would there be text on the screen at all if it had no point?
15:09:48JdGordonpixelma: there is more text on this screen than the quickscreen so scrolling would look terrible
15:10:11*bluebrother wonders if anyone agrees that "enum PITCHSCREEN_VALUES" is prone for confusion −− PITCHSCREEN_VALUES is not a define ...
15:10:41Lloreanbluebrother: Do we have a guideline for constant naming?
15:10:47kugelthe "optimizations" mostly happen with on the remotes with a rather big font
15:11:34bluebrotherLlorean: not sure. defines are all-caps, and from a quick grep in apps/ it doesn't look like enum FOO is used at all but enum foo
15:11:36linuxstbLlorean: Yes, in docs/CONTRIBUTING...
15:11:45 Quit bughunter2 ("bye")
15:12:06linuxstbbluebrother: Yes, I was thinking the same. enum names should be lower case, or in those cases, probably just anonymous, and no variables are declared of that type.
15:12:09LloreanThen if we've got a guideline on it, there's no question it's wrong, right?
15:12:29bluebrotherah, just found it. So I can scream ... enum FOO is forbidden!
15:12:54*bluebrother screams
15:12:58*linuxstb wonders what ICON_BORDER is for, and why it's 12
15:13:03pixelmathe Archos' screen is even smaller than the lcd Iriver remote, but I need to take a look later
15:13:34*bluebrother wonders why those defines aren't defines anymore and why it's better to have them in an enum
15:14:03kugellinuxstb: there're icons on the pitchscreen. and ICON_BORDER is there width/height + some pixels to look better
15:14:28bluebrotherthen it isn't a border size
15:14:31linuxstbkugel: The icons are always the same size?
15:14:37Lloreankugel: Wouldn't a name like PITCHSCREEN_ICON_SIZE or something make more sense?
15:14:58kugelwell, the icon+a small border.
15:15:04kugelfeel free to rename...
15:15:33bluebrotherwhy aren't they named properly in the first place?
15:15:58bluebrotherdealing with confusing names is a real pita. I get to do that at work much too often ...
15:16:10kugellinuxstb: the icons are all 7x8
15:16:50kugelSorry, I didn't know the name is that confusing for you
15:16:52pixelmalinuxstb: I think it's the same in the quickscreen (same icon size)
15:17:01 Quit kushal_12_27_200 ("Leaving")
15:17:29JdGordonthose icons will possibly change one day...
15:17:35bluebrotherany name that does not read what it actually is is confusing. At least if you aren't used to the misuse (which is almost always the case)
15:17:52*bluebrother still screams by the coding style violations
15:17:56linuxstbkugel: It's always best to be explicit as possible. So if your "12" comes from the width of an icon, plus some padding, it's far clearer to define it in those terms.
15:18:44kugelYes, you're probably right.
15:18:55kugelNot that anyone of you guys could've come the past few month and complain about it then.
15:18:56 Join MethoS- [0] (
15:19:05 Quit Dieterbe (Read error: 110 (Connection timed out))
15:19:11Lloreankugel: Frankly, that was JdGordon's job before committing the patch.
15:19:31JdGordonbut then youd all have nothing to complain about....
15:19:38LloreanJdGordon: And we'd like it that way...
15:19:44kugeland he didn't gave me a hint on that define-name
15:20:05*bluebrother thinks this is something one shouldn't need to explain
15:20:13Lloreankugel: But the all-caps names for defines is a pretty common thing, and I guess it's in the file linuxstb mentioned as well.
15:20:14*pixelma is not surprised ;)
15:20:17bluebrotherand as if the others(tm) didn't have enough stuff to do ...
15:20:41kugelLlorean: my defines are all caps, that was not the point
15:21:01Lloreankugel: No, the point was non-defines being named with all caps...
15:21:43kugeloh, that's already changed. I though we're discussing about the name of the ICON define
15:22:29LloreanThe ICON define is at least a case of "well, some people might disagree as to whether it's a logical name" (I can't see how), but standard coding guidelines are things that definitely should've been checked (both by the author, you, and the committter)
15:23:14LloreanYou might want to double check other patches you've got sitting in the tracker, at least.
15:24:28kugelI've read CONTRIBUTING, but I must have missed the part with enums allways lower case. Anyway, I'm sorry, and it's already changed. The name of the enum was of no use anyway
15:31:44 Quit Llorean (Read error: 104 (Connection reset by peer))
15:32:09 Join Llorean [0] (
15:33:17linuxstbkugel: Also, wouldn't it be simpler to pass pitch_viewports[screen] directly to the pitchscreen_fix_viewports() function, instead of two parameters - screen and pitch_viewports[][] ?
15:33:29 Join Onj [0] (
15:34:49linuxstbkugel: Also, looking in gui_syncpitchscreen_run, there's a comment saying that the parent viewport should be a parameter to that function. Why isn't it?
15:35:09kugelbecause JdGordon didn't like it
15:35:44JdGordondidnt i suggest it gets put there first? then changed my mind?
15:35:48*JdGordon doesnt really remember
15:35:57 Join peerlessdeepak [0] (n=peerless@
15:37:03JdGordonlinuxstb: on the firt point yeah your right, and one the 2nd.. we need to decide how we want to eventually do custom viewports... untill then it doesnt matter if its a param or not
15:37:20 Quit J-23 (Read error: 104 (Connection reset by peer))
15:37:27kugellinuxstb: looks like you're right on the fix_viewport thing. I remember having some problems with passing the array, and this was the first way that worked
15:37:57kugelAlso, in an earlier version I needed screen
15:38:02 Quit reacocard (Read error: 110 (Connection timed out))
15:38:47 Join reacocard [0] (n=reacocar@WL-415.CINE.HMC.Edu)
15:38:56 Join Dieterbe [0] (
15:40:12 Join HBK- [0] (
15:40:20kugellinuxstb: ah yes, I did screens[screen].getstringsize() in an earlier version, so I needed screen. Then I replaced this is font_get()->height, and didn't change the parameters
15:42:37 Join funman [0] (n=fun@
15:46:13kugelfunman: trying to get lcd to work?
15:47:00funmanyes, but we may miss some initialisation
15:49:28kugelfunman: haven't atomicpunk done some lcd work?
15:50:09 Quit HBK (Read error: 104 (Connection reset by peer))
15:50:47 Join J-23 [0] (
15:52:53funmankugel: yes, mainly copy of the OF initialisation
15:53:09funmanbut I'm not sure about the clock speed of the DBOP module in AS3525
15:53:21 Join fdinel [0] (
15:53:48kugelyou mean reengineering :) Copying sounds like violating someones's rights ;)
15:54:28funmanmy English is limited but you know what I mean
15:56:11 Quit pixelma (" HydraIRC -> <- Organize your IRC")
15:56:52peerlessdeepakhey guys i am new to rockbox
15:56:58funman < here is an opensource driver for the SSD1303 lcd controller
15:57:16funmanhello fdinel
15:57:17bluebrotherpeerlessdeepak: happens ...
15:57:21peerlessdeepakjust checked out the source code
15:57:42fdinelhey funman :)
15:57:59peerlessdeepakis there something which I can start working on ..or should i passively wait to figure that out for few more days
15:58:00fdineleveryone :P
15:58:33 Quit Onj ("I can get up and I can get out, so I did.")
15:58:36funmanfdinel: I notice in your code you don't set a clock divisor for the DBOP module, isn't that a problem if the clock is too fast ?
15:58:43bluebrotherpeerlessdeepak: well, what are you looking after?
15:58:49scorchepeerlessdeepak: work on something you would like to improve or do...selfish reasons are the best motivator for learning ;)
15:59:01bluebrotherif you want start developing, we have a nice for developers section in the wiki
15:59:03 Join MethoS-- [0] (
15:59:19peerlessdeepakI have an iriver h10 5GB and want to play video in it
15:59:45peerlessdeepakthanks bluebrother
16:00:01bluebrotherwe have mpegplayer
16:00:39fdinelfunman: I may have missed the divisor code, but hmm it would be strange since I thought it was there :P
16:00:57peerlessdeepakIs it possible to port vlc in to it?
16:02:34bluebrothernah. The cpu isn't powerful enough
16:02:36scorchepeerlessdeepak: it wouldnt be very productive and a lot of work
16:03:15peerlessdeepakoh ok
16:03:22kugelsounds interesting though :)
16:03:33peerlessdeepakI am trying to compile and run the ui simulator
16:04:16peerlessdeepakdo i need to have a svn version to do that
16:04:36 Join shotofadds [0] (n=rob@rockbox/developer/shotofadds)
16:04:50kugelpeerlessdeepak: you could at least port the codecs to rockbox (most wouldn't perform good enough on most targets though)
16:05:03scorcheyou dont, but it really ought to read the "for developers" bit in the wiki
16:05:11kugelSVN is recommended, but you can also download the source package
16:05:39funmanfdinel: I notice also when you check status register you miss a shift (1<<0xA)
16:05:41peerlessdeepakI have downloaded it directly
16:05:41bluebrotherfor serious development svn is a must −− unless you want to make yourself additional work :)
16:06:04peerlessdeepakthanks guys ..i will come back after doing the basics
16:06:19 Quit EspeonEefi ("さよなら")
16:06:40funmanor 1<<9 rather (10th bit)
16:06:41kugelfunman: is your rockbox integration done so far? I'm wondering if it can be committed
16:06:57funmankugel: no, it's very hackish
16:07:20funmanfdinel: no, 1<<A (bit 10 = 11th bit) ..
16:08:19fdinelbut I'm checking directly using &, not bitshifting, no?
16:08:30funman& 0xA = bits 3 and 1
16:08:49fdinelright :D bad conversion on my side ;)
16:09:01fdinelshould be 0x400
16:09:08funmanstill no luck
16:09:24funmanfdinel: I prefer to use (1<<x) to understand which bit exactly is used ;)
16:09:36fdinelI wrote the code quite fast, there still may be some errors, especially on the sleep side
16:10:10fdinelwe could use a macro for this, would be clean and easier to understand (less error-prone)
16:10:34kugelfunman: can you until it's less hackish push into the git repo?
16:10:43 Join fragilematter [0] (n=fragilem@
16:10:53fdinelhey fragilematter
16:10:58funmanalso the controller is for a screen with 132 columns and our screen only hs 128, maybe we should loop over the screen (I already try to draw somewhere else than the borders)
16:11:10funmankugel: well when someone wants to code on it, why not
16:11:45kugelthat's why we have the git
16:11:54fdinelfunman: yes we should but I didn't see anything (yet) limiting the display size in the OF, maybe they just "soft-limit" dimensions
16:12:08kugelbtw, can someone give me "push access" on the git repo?
16:12:10funmanand also it's useless while linuxstb hasn't committed a version of mkamsboot which supports custom code
16:12:22funmankugel: you must ask daniel_at
16:12:37 Quit stoffel_ ("leaving")
16:12:55fdinelfunman: and there's that area stuff I don't know much about, maybe it is simply the same with different back-colors, but there may also be something else to consider...
16:13:50kugelpossibly, but unlikely imho
16:15:14 Quit MethoS- (Read error: 113 (No route to host))
16:15:58funmanfdinel: I'm not sure what's the GDDRAM (mentioned in commands B0/BF), this command is used after 'inverting' the screen columns
16:16:30funmanI think I'll read the code of 'ssd1303oled' project
16:17:08funmanIt helped me understanding that 'D/C' was Data/Command# pin, not some electrical power :P
16:17:25fdinelhaha yeah ;)
16:17:41fdinelwell at least have a look at the SSD datasheet
16:18:00funmanyes I read the commands list, but it still is a bit obscure
16:18:20funmanmaybe we would need to reset the device before using it
16:22:54 Quit RandalSchwartz (Remote closed the connection)
16:23:33funmanI think first step is understand exactly which clock frequency the controller needs
16:24:37funmanwe use a dividor in the AS3525 (set for cgu_dbop) AND in the lcd controller
16:26:01fdinelyes well we can have a look at the OF, everything I coded was from the same function (which I named InitLCD) but there may be more elsewhere...
16:26:04linuxstbkugel: A few other comments on your patch.... Why use "n" for a loop variable (in the fix viewports function) - "i" is more conventional. So is "int", not "short".... In pitchscreen_draw, reusing the max_lines variable for other purposes is just nasty...
16:27:24 Join Thundercloud [0] (
16:28:19fdinelfunman: about the GDDRAM, read para. 7.7 in SSD datasheet
16:28:41fdinelin brief, it is a bitmap to be displayed
16:28:49fdinellike 132x64 bits or something
16:28:51bertrikint is probably also more efficient than short (smaller and faster) for the processor
16:29:17linuxstbbertrik: Exactly.
16:29:48kugelI didn't know that int is more effective
16:30:09fdinelkugel bertrik and linuxstb: exactly and GCC even add code to "remove" unused upper bits so using an int skip those instructions too
16:30:56linuxstbkugel: That's the whole point of the "int" type - it's the integer type the cpu is most comfortable with, so you should generally always use it.
16:32:48kugelOk, will do. I just thought a short always takes 2bytes and an int can take 4, so short wouldn't do anything but save ram
16:32:49funmanfdinel: the dbop clock divider is set in 0x151C , the value stored in RAM (at 0x2022C) is 64000000 (64MHz)
16:33:49funmanfdinel: and this function is called at 0x1E70
16:34:09funmanbut I lost track of the argument given to it :/
16:34:20fdinelfunman: ok I'll have a look, I must go, we talk again later :)
16:34:52kugellinuxstb: the i convention isn't mentioned in CONTRIBUTING, I didn't know that we're supposed to use i
16:35:36linuxstbkugel: No, that's not a rule as such, more like something the whole world does out of habit... ;)
16:36:01kugelI'm quite used to n though
16:37:06 Quit scorche (Read error: 104 (Connection reset by peer))
16:37:13kugelso a short is basically a no-go in rockbox?
16:37:14linuxstbYou wouldn't use i and n as follows? for (i=0;i<n;i++) { } i.e. n representing the number of items?
16:37:37linuxstbkugel: No, they have they're place, it's just that in that specific case, int is the better choice.
16:37:58 Join scorche [0] (i=Blah@rockbox/administrator/scorche)
16:38:26funmanisn't short used when you need 16 bits data type ?
16:38:43kugellinuxstb: well if I have both n and i I'd do it the way you wrote
16:39:13bluebrotherdon't you consider it confusing if i is the running value in one case and not in a different one?
16:39:16bertrikI think shorts are used sometimes in rockbox as members in structure, especially if the structure is used in a big array. I don't think using n instead of i is a big deal
16:39:18 Join {phoenix} [0] (
16:40:33kugelbluebrother: Not really. If you only know i then it's probably confusingm but it's definitely not confusing for me
16:41:10bluebrotherkugel: because you're used to strange conventions in contrast to what the whole world does? ;-)
16:41:44kugelNo, because I don't blindly follow conventions, but actually look what variables are used for what
16:42:01linuxstbkugel: What about max_lines?
16:42:35bluebrotherwho says that following conventions means doing it "blindly", and not looking what variables are used for?
16:42:40kugelif you consider this as nasty feel free to change. It's fine for me, and I commented it so it's not confusing for other people
16:43:02bluebrotherbelieve me, if you have to deal with code where stuff is named differently than what it actually does you'll start to love such conventions
16:44:19 Join XavierGr [0] (n=xavier@rockbox/staff/XavierGr)
16:44:29kugelI just don't feel like using more variables than I need to, also to safe space and such
16:44:40linuxstbkugel: So you think the code is clear? Lines like "if (max_lines < pitch_viewports[PITCH_MID].width)" ?
16:44:59linuxstbEspecially when max_lines is a parameter to that function...
16:45:05kugelof course it's not clear if you didn't read the comment. That's why I wrote it
16:45:20bluebrotherif variables are allocated on the stack they don't take up more space −− the compiler can optimize them away out of their scope
16:45:41 Join DaaT [0] (n=dani@
16:45:51linuxstbkugel: The comment doesn't help. Saying "max_lines" no longer means the maximum number of displayable lines does't make the actual code clear...
16:46:58linuxstbUsing meaningful variable names is pretty fundamental to writing readable and maintainable code.
16:47:26bluebrothercode that needs a multiline comment to explain is kinda broken by design
16:49:35 Join robin0800 [0] (
16:49:37DaaThi everyone. I'm on the Database page over at the RB site, but I'm not seeing it listed. Does RB support the album art tag so that it displays it in the WPS?
16:49:44kugelThen change it. I totally agree that it's not clean coding style and confusing without reading the comments. I did that to safe a varialble. And I didn't know that another variable doesn't take space at all, which was the reason I did that
16:50:07scorcheDaaT: see the AlbumArt wiki page
16:50:16 Quit lasser ("ChatZilla 0.9.83 [Firefox]")
16:50:38DaaTscorche, thanks
16:51:07 Quit Thundercloud (Remote closed the connection)
16:54:15***Saving seen data "./dancer.seen"
16:55:38 Join {-phoenix-} [0] (
16:56:31 Join PaulJam [0] (
16:58:24 Join jhulst [0] (n=jhulst@unaffiliated/jhulst)
17:02:44 Quit pixelma2 (Read error: 110 (Connection timed out))
17:03:21 Join pixelma2 [0] (n=marianne@rockbox/staff/pixelma)
17:06:07linuxstbkugel: The problem is that the more I look at that code, the more problems I find... So IMO the patch shouldn't have been committed as it was (and is).
17:06:09 Join toffe82 [0] (
17:06:10 Join miepchen^schlaf [0] (
17:06:25linuxstbAnother example is all those snprintf() calls - why not just use the str() value directly?
17:06:58kugelthat was before me
17:07:31 Quit ejl ("Ex-Chat")
17:08:34kugellinuxstb: tell me the problems. I'm willing to learn, they that I can do better next time
17:08:51 Quit {phoenix} (Read error: 110 (Connection timed out))
17:09:00kugelBut I didn't touch those snprintfs, I only deleted one
17:13:16linuxstbThat's true, sorry for blaming that on you ;)
17:14:04kugelNo problem. I'm used to get blamed here
17:15:10kugellinuxstb: I would've reviewed my patch once more, but I was just compiling a sim to test once more, when JdGordon suddenly committed
17:15:49kugelAlso, JdGordon said it's fine, so I didn't know that it's that bad
17:17:06kugelBut of course, that's not an excuse for the bad coding style on my part
17:20:45 Join {phoenix} [0] (
17:21:00kugellinuxstb: Any more problems?
17:23:54 Nick JdGordon is now known as JdGordon|zzz (n=jonno@rockbox/developer/JdGordon)
17:24:29 Join stripwax [0] (
17:26:17kugelseems not
17:26:36 Join thegeek_ [0] (
17:28:03 Join Schmogel [0] (
17:29:57 Join wpyh [0] (n=william@
17:30:24 Quit thegeek (Read error: 113 (No route to host))
17:32:38bertrikit looks like the sansa clip charges with a current of 100 mA
17:32:46linuxstbkugel: I'm just trying to find where to start... I think my main problem is that I would do the whole thing completely differently. So for example, I wouldn't have an array of viewports, I would have one struct per screen (i.e. main and remote) which stores information about how the pitch screen is formatted. So three viewport structs - top, mid and bottom, plus some booleans to say whether the various elements are displayed or not.
17:33:03linuxstbI would find that clearer, as the logic to decide how to lay out the pitch screen will be separated from the actual display.
17:34:01kugelIt's 3 viewports per screen
17:34:52linuxstbI know.
17:35:05linuxstbWell, 4 if you include the parent viewport.
17:35:06kugelthey get formatted in fix_viewports. The decision if stuff is drawn happens upon drawing.
17:35:28linuxstbI know... I'm saying that's not how I would have done it.
17:36:13kugelhmm I think the code is quite logical and straight forward, but it's certainly possible with structs too
17:36:42 Quit {-phoenix-} (Read error: 110 (Connection timed out))
17:37:06kugelI mostly oriented on the viewport'ified quickscreen, which happens to do it quite in the same way
17:46:28 Join crash91 [0] (n=7aa23d61@gateway/web/cgi-irc/
17:58:04 Join amiconn [50] (n=jens@rockbox/developer/amiconn)
18:00:49 Quit pixelma2 (Remote closed the connection)
18:01:00 Join Lear [0] (i=chatzill@rockbox/developer/lear)
18:01:09 Join phinze [0] (
18:03:41 Join pixelma2 [0] (n=marianne@rockbox/staff/pixelma)
18:04:09 Quit XavierGr ()
18:05:55 Quit crash91 ("CGI:IRC (EOF)")
18:09:03linuxstbkugel: I was thinking of something along these lines -
18:10:15linuxstb(it's just a quick attempt, so show you my intention)
18:12:38 Join mmadia [0] (
18:14:20kugellinuxstb: Ok. Now I'm wondering why haven't done it yourself in the first place. Could've saved me much work.
18:14:46 Join clemare [0] (n=clemare@
18:15:23kugelwell, it doesn't really differ that much from my code. Just the structs and bools put around
18:15:32clemareI just find out rockbox firmware.... but I don't see the iPod I have.
18:15:52 Join coatman [0] (
18:16:10clemareMy iPod is the fat iPod Nano (latest I think)... somebody know if it is supported?
18:16:46linuxstbkugel: No, it's not that different, just a reorganisation of where things are done
18:16:59linuxstbclemare: No, only the very first generation of Nanos can be used with Rockbox.
18:17:20linuxstbApple completely changed the hardware after that (compared with all previous ipods)
18:17:53clemareOk, thanks.... It was a gift anyway :-)
18:18:52kugellinuxstb: well, the decision moved to the init. That means that it can't change due to redrawing. With my code, it was possible that the "+2%" and "-2%" popped up when going from 100.0% to 99.9%
18:19:06kugelit is, rather
18:21:23linuxstbkugel: Yes, I thought about that, but I think that's a bad thing.
18:22:02kugelHm, I considered that as good.
18:22:41kugeland that's why I didn't let it decide on what to draw in the fix_viewports function
18:25:25 Quit alkos333 (Read error: 104 (Connection reset by peer))
18:25:48 Join alkos333 [0] (
18:27:31kugellinuxstb: I find it a bit sad, that 2 month nobody is interested in this patch (and the patch begins to rot), but post-commit everyone joins a discussion with loads of complaints here and there, and someone who says "I'd have done it differently"
18:30:40linuxstbkugel: Speaking for myself, it's simply that I've had almost no time for Rockbox over the last few months, but the commit of your patch co-incided with me having some free Rockbox time, so I had time to look at the commit closely and think about the problem.
18:39:17 Join Zarggg [0] (
18:44:50clemarethanks, bye
18:45:01 Quit coatman (Read error: 110 (Connection timed out))
18:45:02 Part clemare ("Ex-Chat")
18:47:34 Quit DaaT ("Ex-Chat")
18:53:18 Quit nplus (Read error: 104 (Connection reset by peer))
18:54:19***Saving seen data "./dancer.seen"
18:55:14 Join Seed [0] (
19:01:46 Join dabujo [0] (
19:07:12 Part fragilematter
19:09:12 Quit {phoenix} (Remote closed the connection)
19:09:22 Join hannesd_ [0] (
19:14:14 Join AlexStacey [0] (
19:15:22 Join lasser [0] (
19:16:11AlexStaceyhi, i've got a Sandisk e260 v1 player and was about to install Rockbox but wondered about the USB issues - it say's that Windows will recognise the device but no more
19:16:18AlexStaceydoes anyone know about that?
19:17:04AlexStaceyi need to be able to charge the device from usb and to transfer music but it sounds s though i won't be able to with Rockbox
19:17:53AlexStaceyIs it possible to switch between the two firmwares? (officil and rockbox)
19:18:42 Join CrazyFoam__ [0] (
19:18:56bertrikAlexStacey, it will reboot into the OF automatically if USB is plugged and reboots back to rockbox when unplugged after charging/transferring data
19:19:06vitjastrange thing: bidi_l2v doesn't work with -O but works with -Os, then if insert dummy call it works again... compiller bug?
19:19:11bertrik(OF = original firmware)
19:19:28AlexStaceybertrik: wow - that sounds perfect
19:20:06CrazyFoam__is there another way to reboot rockbox (ipod-video) besides holding menu+select?
19:20:14CrazyFoam__was playing with rockboy and it's frozen
19:21:10bertrikvitja, in my experience it's more likely a bug in our source code rather than the compiler itself
19:21:54vitjabertrik: ok.. after changes in bidi.c nand_identify fails as it seems memory is corrupted
19:22:15funmanwhy does nand_identify needs code from bidi.c ?
19:22:19vitjabtw if I insert strlen() for example in bidi_l2v() it works ok
19:22:51vitjaI was trying different revisions and found that this happend after optimizations in bidi
19:23:37CrazyFoam__perhaps only option is to wait 15 hours for battery to drain? :(
19:24:15vitjais there a way to check stack for corruption?
19:24:21 Join webguest64 [0] (n=c77ebdec@gateway/web/cgi-irc/
19:24:38 Quit hannesd (Read error: 110 (Connection timed out))
19:24:39 Nick hannesd_ is now known as hannesd (
19:24:51 Join Thundercloud [0] (
19:26:34*bertrik wonders how easy/hard it would be to run a static code analyser (e.g. splint) over the rockbox code
19:27:21 Quit jhulst (Read error: 60 (Operation timed out))
19:29:05 Quit robin0800 (Read error: 110 (Connection timed out))
19:30:10 Join DismalArcadia [0] (
19:32:51 Join {phoenix} [0] (
19:34:29 Quit vitja ("Ex-Chat")
19:34:46webguest64I am not a programmer but i want to add plugins to rockbox that are in the Patches section of the website. How do I do it?
19:36:02linuxstbwebguest64: Easiest way is to find "unofficial builds" of rockbox that already include those plugins, and use those builds. But if you want to apply the patches yourself, you need to set up a development environment and compile Rockbox yourself - there are plenty of guides to that on the Rockbox wiki
19:38:23kugellinuxstb: I wonder nobody changed the nasty reuse of max_lines. If you want I can do a patch doing this and also your struct stuff (and possibly the snprintf)
19:40:13 Join bughunter2 [0] (n=Jelle@
19:40:53 Join XavierGr [0] (n=xavier@rockbox/staff/XavierGr)
19:44:28 Join Horschti [0] (
19:45:20 Quit Horscht (Nick collision from services.)
19:45:20 Quit webguest64 ("CGI:IRC (EOF)")
19:45:26 Join ajonat [0] (n=ajonat@
19:46:44 Join nplus [0] (
19:47:15 Join petur [50] (n=petur@rockbox/developer/petur)
19:47:24 Join nuonguy [0] (
19:48:16 Quit fyrestorm (Read error: 104 (Connection reset by peer))
19:50:25 Quit mmadia ("ChatZilla 0.9.83 [Firefox 3.0.3/2008092417]")
19:51:14 Join fyrestorm [0] (
19:54:15 Quit CrazyFoam__ ("Ex-Chat")
19:55:47 Quit Darksair ("Do you hear that? This is the sound of inevitability. This is the sound of your death, Mr. Anderson.")
19:56:25AlexStaceyrockbox seems very nice. props to the developers =]
19:56:41 Quit fyrestorm (Read error: 104 (Connection reset by peer))
19:58:49gevaertsThanks :)
19:59:47 Join fyrestorm [0] (
20:01:10 Quit nplus (Remote closed the connection)
20:02:19 Join Nibbler [0] (
20:02:46 Join jhulst [0] (n=jhulst@unaffiliated/jhulst)
20:03:21 Join nplus [0] (
20:11:57 Quit kugel (Read error: 110 (Connection timed out))
20:15:12 Join kugel [0] (n=chatzill@unaffiliated/kugel)
20:20:10 Quit funman ("leaving")
20:20:32 Quit nplus (Remote closed the connection)
20:21:20 Join nplus [0] (
20:23:31 Quit {phoenix} (Remote closed the connection)
20:26:30 Join {phoenix} [0] (
20:30:39 Quit Thundercloud (Remote closed the connection)
20:31:05 Join coatman [0] (
20:36:03 Quit jhulst (Remote closed the connection)
20:38:50 Join funman [0] (n=fun@
20:39:03funmanI'm just back because my LCD screen just lighted up :P
20:45:11 Join webguest81mitche [0] (n=5657aa5c@gateway/web/cgi-irc/
20:46:03webguest81mitchehello I have bought the sansa e280. but I have a problem I have the v2 version and I want install rockbox has somenoe an solve for me?
20:47:29 Part J-23
20:47:37linuxstbwebguest81mitche: Rockbox doesn't work on the v2 versions...
20:47:38bertrikno, rockbox doesn't work on the v2 yet and there's no schedule on when it will
20:49:40 Part AlexStacey
20:49:49 Quit bughunter2 (Read error: 104 (Connection reset by peer))
20:52:50 Join Dhraakellian [0] (
20:53:25 Join EspeonEefi [0] (i=espeonee@STRATTON-SEVEN-TWENTY-ONE.MIT.EDU)
20:54:21***Saving seen data "./dancer.seen"
20:54:57 Quit webguest81mitche ("CGI:IRC (EOF)")
20:55:25agaffneywtf was that about?
20:56:00Dieterbehe likes you
20:58:06 Nick HBK- is now known as HBK (
20:58:22 Quit peerlessdeepak ("Leaving")
21:01:01 Join tvelocity [0] (
21:01:38 Join denes [0] (
21:02:37denesgevaerts: markun: ok, I have sent in the meizu m3 lcd driver, it's fs #9459
21:05:59*gevaerts goes to try it
21:06:48 Join myndzi [0] (
21:07:26myndziso i had to fix the theme i'm using again and decided to try and pload it .. i followed all the instructions but i can't for the life of me locate the login link for the wiki
21:08:00myndziwhat do i do from here? i already have a username from sometime before
21:08:43gevaertsmyndzi: the wiki will pop up a login dialog when you click on any link that needs a login (like "edit")
21:09:19myndziok, though that's kinda like closing your document to save it :P
21:11:43 Quit Schmogel (Read error: 104 (Connection reset by peer))
21:11:58 Join fragilematter [0] (i=5c526163@gateway/web/ajax/
21:13:15denesgevaerts: ok, I am interested in whether it works for you
21:14:02gevaertsIs dfu on the m3 the same as on the m6sl? (hold M+play while plugging in)
21:14:38denesgevaerts: yes, and then release play when plugged in. release M when it says "Clearing status ...".
21:14:52gevaertsOK. code uploads, and I get a white screen. Now to draw something on it...
21:16:08Dhraakellianhow would Rockbox's MPEG-2 playback compare with the Sansa e200 MJPEG playback in terms of battery life?
21:17:09*gevaerts thinks that Dhraakellian asks a very interesting question, and he can't wait for his test results ;)
21:17:51bluebrotherthe sansa OF has mpeg playback? I thought they do only some special avi
21:17:58 Join bughunter2 [0] (n=Jelle@
21:18:08Dhraakellianbluebrother: MJPEG in a .mov container, iirc
21:18:21linuxstbbluebrother: Motion JPEG...
21:18:29funmanDhraakellian: is that standard mjpeg in standard mov, or some format similar ?
21:18:36funmanaka AMV
21:19:07gevaertsdenes: I just see a white screen, no text.
21:19:12Dhraakellianfunman: I would think so, but I could very easily be moving out of my depth if I try to answer further
21:19:16 Join domonoky [0] (n=Domonoky@rockbox/developer/domonoky)
21:19:17denesgevaerts: did you apply the whole patch - in which case there should be something written on the top - or just the lcd driver part?
21:19:57denesgevaerts: is this old or new m3? patch only works with new m3
21:20:06gevaertsEverything. Let me check for other local changes that could cause this
21:20:31gevaertsdenes: this is an m3 that was purchased from last week. The back says it's an SP
21:21:01 Join phinze_ [0] (
21:21:01denesgevaerts: mine says sp too, but it's definietly the "v2" version
21:21:33gevaertsDoes it say that somewhere on the case?
21:21:53denesgevaerts: afaik no
21:24:15denesgevaerts: ok, I checked with vanilla svn+patch and works for me...
21:24:25 Part fragilematter
21:24:44denesgevaerts: does your player work with 2003.XX firmware? the new m3 only works with 2004.X or later
21:24:58gevaertsIf I comment out the clock setting, the touchpad works, so I definitely have running code
21:25:03denesgevaerts: check the version of the firmware in the menu
21:25:28denesgevaerts: then it's the other LCD :(
21:25:35 Join Thundercloud [0] (
21:26:20gevaertsYou're looking at this the wrong way. Now we can at least decently support both!
21:26:58denesgevaerts: well :)
21:27:25denesgevaerts: so can you check the firmware version?
21:27:50gevaertsCurrently 2.004.3. I'm downloading some other versions right now
21:28:30denesgevaerts: maybe they use whatever lcd they have at hand
21:29:32denesgevaerts: if you get picture with 2003.X firmware, then it's likely the old type lcd
21:29:57gevaertsstill downloading. Their site seems a bit slowish
21:31:22 Quit phinze (Read error: 110 (Connection timed out))
21:32:07gevaertsdenes: what copyright year does your m3 show on the back? Mine says 2007
21:32:21denesgevaerts: mine too
21:32:40gevaertsThat doesn't help identify them then :(
21:33:09denesgevaerts: there is no 2004.3 fw for download on the page, which is strange
21:33:12deneskind of
21:33:32denesgevaerts: mine came with 2004.6
21:33:44gevaertsMine is a dane elec one, maybe the OEMs get different versions
21:33:59 Join jhulst [0] (n=jhulst@unaffiliated/jhulst)
21:34:10denesgevaerts: yes, possible
21:38:17*petur tries to wrap his brains around sample_to_mono()
21:40:04 Quit myndzi (Remote closed the connection)
21:41:00 Quit linuxstb (Remote closed the connection)
21:41:31gevaertsdenes: 2003.5 works, so I guess this is the old one
21:41:47denesgevaerts: yes, seems like that
21:43:41 Join przemhb [0] (
21:44:20 Join myndzi [0] (
21:45:04myndziok, think i did it! also one other question: what's the state of the MOD support on rockbox? it doesn't really say much under SoundCodecs but i'm wondering if it has been improved past the basic patch that was in there some time ago
21:46:18przemhbAm I right believing that I will be able to format a new HDD connected to iriver H10 after replacing an old one?
21:46:40gevaertsmarkun: ping
21:47:49gevaertsdenes: I could probably verify this myself, but did the stack size exceed available RAM?
21:48:28*gevaerts is a bit surprised that the OUTPUT_FORMAT thing didn't cause heaps of trouble
21:50:02denesgevaerts: I don't think the stack exceeded available RAM, but .bss definietly did
21:50:20 Quit nplus (Remote closed the connection)
21:50:32gevaertsThat could explain some weird issues I saw I guess
21:50:41 Quit amiconn (" l8er")
21:51:24denesgevaerts: on the m6sl? what issues?
21:51:47gevaertsdenes: if the code is a bit too big, it stops working
21:52:04denesgevaerts: try maskikng the interrups. that's what I did
21:52:15denesgevaerts: INTMSK=0;
21:52:59denesgevaerts: the code appeared to work, but broke when "Clearing status appeared", and masking the interrupts fixed it
21:53:27denesgevaerts: I am not sure whether the iram wraps around (my guess) or what happens when it gets overwritten
21:53:47 Join tim__b [0] (
21:54:01gevaertsI guess we need both. Disabling interrupts certainly doesn't hurt as long as we don't actually need any
21:54:30 Join ztx_- [0] (
21:54:30denesgevaerts: what do you mean "both"?
21:54:45ztx_-hey guys
21:54:53gevaertsBoth disabling interrupts and not overflowing RAM
21:55:01ztx_-any idea where i can get a sansa e280 v1 to use with rockbox?
21:56:03funmanztx_-: probably on ebay
21:56:28denesgevaerts: imho overflowing ram is not a problem. if the code is loaded at the start of iram (like on m6sl), then it's not easy to overflow the ram. it's possibly to overwrite important stuff though, and maskin all interrupts takes care of that issue.
21:57:06gevaertsYes, but that only solves part of the problem. You never know what you overwrite
21:57:19denesgevaerts: on m3 (where stuff is loaded at IRAMSTART+128k) it probably isn't a problem, as ram apparently wraps around
21:57:32gevaertsztx_-: if you have contacts in the US, you can try Unfortunately they don't deliver to most of the world
21:57:35denesgevaerts: but why do you care, code using that data won't ever be executed
21:58:02denesgevaerts: after our code got loaded
21:58:02ztx_-gevaerts, i dont have contacts in the US :/
21:58:24gevaertsToo much speculation about what exactly happens. Keeping it small enough just is safer
21:58:54denesgevaerts: the bigger problem is that the bootrom hands execution to us in user mode, so irq and such stuff cannot really be set up.
21:58:55gevaertsztx_-: It's very unfortunate, I know. We just have to live with it I guess
21:59:14denesgevaerts: it is done in crt0.s
22:00:24denesgevaerts: but yes, if you can keep the code small, then that's even better
22:00:52myndziis there a link to walk a guy through compiling rockbox / adding patches?
22:02:46gevaertsdenes: anyway, I just committed your patch, so the work will get a bit easier at least for you
22:03:03gevaertsmyndzi: have you seen ?
22:03:52myndzithat looks like about what i wanted.. is it just me that has a very hard time finding his way around the wiki?
22:04:38gevaertsThe wiki does have a search feature
22:04:54myndziwhich doesn't seem to produce useful results whenever i try it :(
22:05:16myndziit took me montsh to remember "SoundCodecs" when i wanted to dig up that page heh!
22:05:28myndzioccasuionally that is.. god, i hate telnet
22:06:27myndzii think i'll asve this for later when i have speedier internet. t hanks.
22:07:16 Join kronflux [0] (
22:07:40 Join _any1 [0] (
22:07:47 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
22:09:06funmanlinuxstb: when you use BX to branch to the unpack function, shouldn't the function start address be aligned on 32 bits ?
22:09:49linuxstbfunman: No, it's thumb code, so just needs to be aligned to 16-bits. The BX parameter has 1 added to it (i.e. bit 0 set) to indicate a switch to thumb mode.
22:10:45funmanlinuxstb: I remember as or ld being insulting when my functions didn't start on 32 bits aligned address
22:11:03funmanand in the Clip OF I see a lot of dummy 16 bits 0, before function starts
22:11:24 Quit Lear ("ChatZilla 0.9.83 [Firefox 3.0.3/2008092417]")
22:12:05linuxstbas/ld were complaining about thumb code not being 32-bit aligned?
22:12:27funmanno, function starts
22:12:39funmanie, what you would use in bx instructions
22:12:46linuxstbThat's what I mean - the start of thumb functions?
22:13:58 Quit kronflux ("Leaving")
22:14:11funmanIf I remember correctly, yes
22:14:58linuxstbI can't think of a reason why a bx to thumb code on a 16-bit (but not 32-bit) boundary would cause problems.
22:15:27domonokylinuxstb: i just saw, you moved mkamsboot from utils into the rbutil dir, and we wanted to move everything in the rbutil dir to the utils dir in the future :-)
22:15:50linuxstbdomonoky: bluebrother already mentioned that - see the logs from soon after my commit...
22:16:04*domonoky goes log checking...
22:16:46funmanlinuxstb: I may have been confused earlier, because I can't reproduce this problem
22:18:28 Join mirak [0] (
22:19:39 Quit myndzi (Remote closed the connection)
22:19:45domonokyah, thx... :-)
22:20:14domonokythats also a good idea, but if utils is for devs, the wpseditor should stay there a bit longer :-)
22:20:47 Join tessarakt [0] (
22:20:54 Join massiveH [0] (
22:23:59linuxstbdomonoky: Thinking into the future, mkamsboot is very similar to the iriver h1x0/h3x0 mkboot - it will take an original firmware upgrade file and a Rockbox bootloader, and produce a patched firmware upgrade file which the user will need to copy to their device and flash using the OF.
22:24:11 Quit jeffdameth (Read error: 110 (Connection timed out))
22:24:40linuxstbWhat would rbutil need from the mkamsboot code to do that?
22:25:21bluebrotherhmm. Shouldn't the fwpatcher / (un)scramble move to its separate folder below rbutil too in that case?
22:25:50linuxstbscramble is used as part of the build system
22:25:59 Join jeffdameth [0] (
22:26:00domonokyisnt fwpatcher an outdated gui tool which should be deleted ?
22:26:04bluebrotherit is?
22:26:16bluebrotherI thought it's only used when patching the hex file
22:26:42linuxstbbluebrother: No, it's used for lots of things. It generates the file as part of most builds
22:27:08bluebrotherah, ok. So the actual iriver scrambling is only a small part of it.
22:27:14domonokylinuxstb: rbutil would need the code to "patch" the original firmware from mkasmboot, preferably with some checking (md5sum or alike)
22:27:58linuxstbdomonoky: I know, but with what kind of API? A single function that takes filenames, or open file descriptors, or pointers to buffers...
22:28:06bluebrotherQt can do md5sums itself (as in the new bootloader stuff)
22:28:33 Join avis [0] (n=ident@pdpc/supporter/student/avis)
22:28:40bluebrotherlinuxstb: I don't think that matters too much −− iriver patching currently takes filenames
22:29:53markungevaerts: pong
22:29:57bluebrotherbut Qt can return usual FILE* from files it opened, so there shouldn't be an issue if we get those.
22:30:07domonokyi think we could cope with every type of API, but it would be good, to make a seperate main.c file for mkasmboot so we can easly link t it
22:30:22gevaertsmarkun: I checked by now, but it seems that the s5l8700 was a bit buggy
22:30:32domonokyalso all error should go into a buffer, and not directly to stdout
22:30:37markungevaerts: what happened?
22:31:11markundenes: thanks for the LCD code
22:31:15gevaertsOUTPUT_FORMAT was wrong. It was set to elf32-littlearm. denes spotted that one
22:31:32gevaertsI'm actually a bit puzzled that it produced working code at all
22:31:45bluebrothererrors should get returned as return value if possible, like I adapted the scrambling code a while ago
22:32:08bluebrotheralso, IMO it would be a good thing to make the API at least similar to what's used for the hex patching
22:32:19linuxstbbluebrother: Yes, that's what I'm asking - what's that API?
22:32:40linuxstbCan you point me to the hex patching functions in rbutil?
22:33:21*linuxstb spots it's all Qt-ized cpp and backs away....
22:33:53markunhm, I can't find fs#9459
22:33:54bluebrothercurrently we include tools/mkboot.h and tools/iriver.h
22:34:10bluebrotherQt-ized cpp? Not in the tools folder ;-)
22:34:21gevaertsmarkun: too late. I committed it
22:34:35bluebrotheroh, and put the extern "C" stuff in the header that exports the API ;-)
22:35:08bluebrotherso it's basically some functions we call while passing them filenames
22:35:10linuxstbbluebrother: I was looking at bootloaderinstallhex.cpp, but realised that wasn't it...
22:35:36markungevaerts: I should have checked that :)
22:35:59funman < clip lcd
22:36:25bluebrotherno, that's the actual implementation of the calling logic in Qt. Isn't really different from the "other" main function
22:36:43domonokyso for irivers we use iriver_decode(..),mkboot(..),iriver_encode(..) :-)
22:36:48bluebrotheror at least, shouldn't be different. Except that it isn't plain C anymore
22:37:40linuxstbWell, I can simply provide "int mkamsboot(const char* infile, const char* bootfile, const char* outfile);" that does everything, returning an error code.
22:38:26 Quit ompaul (Read error: 60 (Operation timed out))
22:38:27funmanif you want to do *very* ugly code, bootfile argument could even go away
22:38:39bluebrotherif there's nothing else to handle that would be fine.
22:38:50linuxstbfunman: ?
22:39:26bluebrothernah, we want to download the most current bootloader from the download server ;-)
22:39:43linuxstbbluebrother: There's no encryption, just various checksums that need updating. That function will do that.
22:39:53denesgevaerts: thanks for comitting
22:39:54funmanhm I must be confused
22:40:18domonokylinuxstb: sounds good, should be easy to use in rbutil.
22:40:23bluebrotherlinuxstb: I guess this is no lenghty process?
22:40:27linuxstbCan rbutil detect the v2 sansas yet?
22:40:48domonokylinuxstb: it should detect them as incompatible devices, yes
22:40:58*bluebrother too slow
22:41:01domonoky(via usb ids)
22:41:10 Join ompaul [0] (n=ompaul@gnewsense/friend/ompaul)
22:41:14bluebrotherdang, to slow again :o
22:42:01linuxstbHow does rbutil handle bootloader patching on the iriver? Does it write the patched file directly to the device without prompting the user for a filename?
22:42:06 Quit ryanakca (Connection reset by peer)
22:42:12bluebrotherdomonoky: had a long weekend? You seem so ... vital
22:42:27 Join ryanakca [0] (n=ryan@ubuntu/member/ryanakca)
22:42:32bluebrotherlinuxstb: yes, as there is no choice when flashing the file
22:43:23linuxstbOK, so when rbutil calls mkamsboot it will already know the kind of device the user is installing onto?
22:43:33domonokybluebrother: yes :-)
22:44:10domonokylinuxstb: correct, at this time rbutil already know all it needs about the device (mountpoints, etc)
22:45:02 Quit pixelma2 ("-")
22:45:15bluebrotherwell, the output file is still a temporary file first which gets copied later (after md5summing)
22:45:30 Join pixelma [50] (i=pixelma@rockbox/staff/pixelma)
22:47:22linuxstbHmm, md5sums could be trickier for the V2s - the firmwares are still being actively developed by Sandisk, and we are nowhere near a stable bootloader...
22:47:34 Quit phinze_ (Read error: 60 (Operation timed out))
22:48:06domonokylinuxstb: can a broken "patched" firmware cause any harm ?
22:48:29bluebrotherwe could provide the md5sums on the download server
22:48:43domonokyso we should make md5sums for known good combinations, and md5sum them..
22:48:49 Join BigBambi [0] (n=Alex@rockbox/staff/BigBambi)
22:48:56bluebrotherbrick as in iriver brick? Or as in unbrick-without-special-hardware?
22:49:27bluebrotheri.e. like unbricking my router by shortening some pins
22:49:43linuxstbSome devices (e.g. e200v2) can enter a recovery mode by shorting some pins, but not all
22:49:59bluebrotherhmm. Sounds not-so-good
22:50:06linuxstbI think funman still has a Clip with a broken firmware installed...
22:50:12domonokymd5sums on the server would be a good idea.. and with possible bricks we can not go without md5summing..
22:50:37bluebrotherbut is there a point in offering support in rbutil unless we have a stable bootloader?
22:50:48linuxstbbluebrother: I was about to say the same thing
22:50:56linuxstbSo we should probably forget this for now...
22:51:29bluebrotheryep. We could add support without safety-checks and #ifdef that out so only people compiling themselves can use it
22:52:50linuxstbI don't think it's worth you spending time on it - developers can just use mkamsboot... I just wanted to try and write mkamsboot in such a way that it would be easy for you to use when the time comes.
22:54:24***Saving seen data "./dancer.seen"
22:55:06 Join amiconn [50] (n=jens@rockbox/developer/amiconn)
22:57:49 Quit alkos333 (Read error: 104 (Connection reset by peer))
23:00:26 Quit Nibbler (Read error: 110 (Connection timed out))
23:00:44 Join Nibbler [0] (
23:01:14CtcpPing from gevaerts!n=fg@rockbox/developer/gevaerts
23:02:03 Part przemhb
23:02:23CtcpPing from gevaerts!n=fg@rockbox/developer/gevaerts
23:05:38CtcpPing from gevaerts!n=fg@rockbox/developer/gevaerts
23:08:47 Join Nico_P [50] (n=nicolas@rockbox/developer/NicoP)
23:09:23 Quit _any1 ("Ping timeout")
23:10:06 Quit tvelocity (Read error: 104 (Connection reset by peer))
23:11:43 Join fml [0] (n=4fd3f165@gateway/web/cgi-irc/
23:12:18fmlbluebrother, domonoky: hello. Have you seen FS #9426?
23:12:20markungevaerts: wouldn't it be better to use an enum for the LCD type in the M3 driver?
23:12:42gevaertsmarkun: maybe, once we know a bit more about it
23:13:22markunah, we don't know the controller?
23:14:45gevaertsNot as far as I know. I still need to run my controller-id flashing code on the m3 though
23:16:49markungevaerts: but you can display stuff on your M3 now?
23:17:06gevaertsNo. Mine seems to be the older version
23:17:41gevaertswpyh should be able to however. If I understand his meizume comment correctly he has the newer one (same as denes)
23:17:51*amiconn experienced a seemingly new bug on the beast
23:18:51amiconnVoice doesn't work after boot before playing music. As soon as music is started, voice starts working, and then stays like that even if music is stopped again
23:19:15 Join ompaul_ [0] (n=ompaul@gnewsense/friend/ompaul)
23:19:15amiconnQuite annoying :\
23:22:57 Join alkos333 [0] (
23:25:04 Quit domonoky (Read error: 104 (Connection reset by peer))
23:25:41 Quit fml ("CGI:IRC (EOF)")
23:27:17 Join tvelocity [0] (
23:28:34 Quit ompaul (No route to host)
23:32:41 Quit Dieterbe (Read error: 110 (Connection timed out))
23:34:38 Quit bughunter2 ("bye")
23:35:00 Quit hannesd (Read error: 148 (No route to host))
23:35:37 Join cas [0] (
23:35:49cashello Rockbox people :-)
23:35:55peturin the lang file, can you specify that a string is only used when two defines are present? I have a string that is only used for targets with swcodec _and_ recording...
23:35:56 Nick cas is now known as casainho (
23:36:32*petur silently looks at Bagder
23:36:37Bagderpetur: only if you create a "feature" that depends on those two variables
23:36:54casainhoI would like to know where I can add "" option on makefile of RB bootloader...
23:37:17Bagderits not already there?
23:37:59casainhowell, there is a map file but is with a lot of information, I was looking for a simpler map file like this one:
23:38:03BagderI get a map when I build a test bootloader here
23:38:27peturBagder: any hints on how/where you create such a feature?
23:38:46Bagderapps/features.txt defines all the lang features
23:39:04peturBagder: thanks!
23:39:41Bagdercasainho: so you want -Map but with _less_ details?
23:39:46BagderI don't understand
23:39:57casainhobut I am getting this one from RB bootloader:
23:40:19*petur discovers an existing feature called 'recording_swcodec'
23:40:39casainhoI would like to compare both files... however one have much more information... it there any options to generate both files in the same way?
23:41:10casainhoI must say that one file is generated with GCC-4.1.1 and the RB the GCC-4.0.3...
23:41:12Bagdercasainho: since both were done with ld I'm sure that's possible, if nothing else with a little scripting
23:42:10casainhoBagder: hmmm... I just did add "" to generate the map file on that other code... how can I do the same on RB bootloader?
23:42:24Bagderit already does that
23:42:55linuxstbcasainho: The two maps look the same to me, apart from the fact that they're from compiling different code...
23:43:15 Quit shotofadds (Read error: 110 (Connection timed out))
23:43:32casainholinuxstb: the same? but they have the same code on different addresses...
23:43:52Bagdercasainho: that's controlled by the lds file
23:43:55linuxstbNo, I mean the format looks the same.
23:44:12linuxstbThe _content_ is obviously different, but they are from two different projects...
23:45:58casainhobut both lds files are almost the equal
23:46:11linuxstbYes, but the C code you're compiling is completely different.
23:46:24casainholinuxstb: no, is almost the same code...
23:46:49*Bagder says something about posting patches to get better help...
23:46:53casainhobut the same code as RB bootloader is not working, so, I was trying to compare the memory map
23:47:10linuxstbcasainho: So this mapfile is from a Rockbox bootloader, linked against the Rockbox code in firmware/ ?
23:47:20pixelmapetur: did you see the already existing recording_swcodec feature? ;)
23:47:38*pixelma should have read on
23:48:17*amiconn thinks the MM3 LCD driver looks like the port uses the wrong lcd data format
23:48:50casainholinuxstb: here is one main():
23:48:53linuxstbcasainho: The "simple" map file looks like it is just three object files - cstartup.o, lowlevel.o and main.o. The Rockbox one has lots of object files - i.e. all the Rockbox code.
23:48:58 Quit {phoenix} (Remote closed the connection)
23:49:09peturpixelma: ah, you read it already ;)
23:49:40casainholinuxstb: look here the RB bootloader:
23:50:35casainholinuxstb: yes, only that three object files, because I am in the begining of build the bootloader, I don't need no more for now
23:51:09amiconnSlow unstuffing of 16 bit lcd data into 24 bit...
23:51:22linuxstbcasainho: All I'm saying is that that is the reason your "simple" map file is simple. I can't help debug your code.
23:51:32amiconnWhy not use a 32 bit framebuffer if the lcd is capable of full 24 bit colour?
23:51:36 Nick ompaul_ is now known as ompaul (n=ompaul@gnewsense/friend/ompaul)
23:52:00 Quit MethoS-- (Remote closed the connection)
23:52:14casainholinuxstb: ah, okok.
23:52:35 Quit tvelocity (Remote closed the connection)
23:52:45casainholinuxstb: can you find some problem on where interrupt code are being putted on memory?
23:53:23casainholinuxstb: because my problems are with interrupts, they don't work... main() is executed however interrupts give problems...
23:53:41denesamiconn: is there code for that? like ldc-32bit.c ?
23:53:57amiconnNot yet, but it shouldn't be difficult
23:54:02 Quit nuonguy ("Leaving")
23:54:25 Quit bmbl ("Woah!")
23:54:26denesamiconn: the rationale from my part was to have something working soon
23:54:51amiconnOf course, if the controller doesn't support real 24 bit internal, but only the interface is now in 24 bit mode, we should switch the interface into a more suitable mode instead
23:54:57linuxstbcasainho: Sorry, I can't help now. But you don't need to enable interrupts during the early stages of development - you can do other things first, and then work on enabling interrupts later, e.g. when you have a lcd driver working to help you debug. Many Rockbox bootloaders never enable interrupts.
23:54:58funmancasainho: when googling for interrupts I read that you need the volatile keyword when you access registers, if gcc is optimizing the code
23:55:14amiconnKeeping 24 bit precision if the controller can't handle it would be a waste
23:55:23denesamiconn: yes. we don't know what kind of lcd chipd there is.
23:55:46linuxstbdenes: What's the resolution of the LCD?
23:55:51denesamiconn: this is how it's done in the OF
23:55:59amiconnThat's detail which needs research
23:56:02deneslinuxstb: 176*132
23:56:06Bagdercasainho: I would suggest using objdump on the files to compare the interrupt vector setups etc
23:56:18casainholinuxstb: thanks for that idea of never enable interrupts :-)
23:56:23casainhofunman: tnahks.
23:56:31linuxstbdenes: And how much RAM does the device have?
23:56:42casainhoBagder: like doing disassembly?
23:56:47deneslinuxstb: 256KB iram, 16 (imho) MB ram
23:57:33linuxstbdenes: OK, so a 32-bit framebuffer shouldn't be too much of a burden.
23:57:33casainhoBagder: but If I do, I will just compare code... I can't compared the addresses, right?
23:57:50deneslinuxstb: no, if somebody writes the necessary code
23:57:57Bagdercasainho: but if your code actually runs (partly) then why is there an address problem?
23:58:09linuxstbdenes: As amiconn said, it should be relatively trivial to adapt lcd-16bit.c to lcd-32bit.c
23:58:18linuxstb(just change shorts to ints...)
23:58:39amiconnSome other files would also need to be adjusted/ extended, e.g. lcd.h
23:58:43ztx_-linuxstb, where do i find the lcd files?
23:58:57amiconn(fb_data, pixelformats...)
23:59:00casainhoBagder: I believe that interrupts are not ok, like If there is an Periodic Interrupt, system jumps to some error place hand system may hang...

Previous day | Next day