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

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

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

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

#rockbox log for 2007-08-14

00:00:01Bagdermicrosd cards are everywhere
00:00:29IA4004Bagder I dont mean micro sd cards,I mean the memory module of the sansa
00:00:52advcomp2019i do not know if you can
00:01:03BagderIA4004: what memory module?
00:01:15Bagderthe actual nand flash chips?
00:01:32 Join Rob2222 [0] (
00:01:33IA4004for an e250 non rhapsody
00:01:49IA4004yes the nand chips
00:02:05Bagderit would surprise me greatly if you can buy them
00:02:21Bagderask sandisk, I doubt anyone else knows
00:02:32advcomp2019you can try to get them to send −−>
00:02:45DerPapstaren'T they soildered on the mobo?
00:03:00Bagderno, they're on a daughterboard
00:03:19Bagderso it's technically feasable
00:03:27DerPapstah ok
00:03:30 Quit Tempis (Read error: 110 (Connection timed out))
00:03:38DerPapstmaybe get a dead one from ebay?
00:03:40 Quit jgarvey ("Leaving")
00:03:53DerPapstwith brolen buttons, lcd and whatnot
00:03:59rasherGreat, now the patches generated by my translating web-interface even have the correct filepaths. Should be good to go, then. lists "Sansa e280 8GB Flash" repair for 120USD
00:04:56DerPapstthat's expensive
00:05:23advcomp2019but that is because you have to send your player to them and they fix it
00:05:33DerPapsta dead one from ebay is cheaper then
00:05:46DerPapstand you can resell the unneeded parts again ;)
00:06:10Bagderadvcomp2019: yes but the 120 doesn't include shipping
00:06:28advcomp2019i know
00:06:28Bagderwhich is at least another 15
00:07:01Bagderseems like a cool place nonetheless
00:07:07advcomp2019i thought i should throw that site out to people
00:07:15DerPapsti would either buy a new one then (new warranty) or get one of ebay
00:08:20 Join roolku [0] (
00:09:01 Quit JdGordon ("Konversation terminated!")
00:12:13 Quit robin0800 (" HydraIRC -> <- *I* use it, so it must be good!")
00:14:34IA4004anyone has a better bet?
00:15:41roolkupetur: any chance on having a 4th option with the old behaviour? It is really awkward now to quickly check the file extension (it used to be just the quick screen now it is several button presses and menu navigations away)
00:16:07*roolku is talking about the "show extensions" commit
00:16:08IA4004so if I cant get my sansa back to the game,I'm going to put it for parts
00:16:20peturI didn't know people used it that way...
00:16:28peturI'll see what I can do
00:16:34roolkuthank you
00:16:49IA4004everything of the sansa is great except the memory modules(nand)
00:17:14 Quit nick89 ("Gotta Go")
00:18:47 Quit Rob222241 (Read error: 110 (Connection timed out))
00:19:06 Quit roolku ()
00:21:37IA4004I'm going to search for spare parts for the sansa,if I find great things,I'm going to paste links
00:23:44 Quit IA4004 ("ChatZilla [Firefox]")
00:31:28 Join vmx [0] (
00:32:37vmxhi. can i see the current changelog somewhere? i only found the changelog for daily builds, but i'd like to see the new changes
00:32:40 Join IA4004 [0] (n=chatzill@
00:33:25Lloreanvmx: The front page of the site...
00:33:45vmxah, sorry :)
00:33:53vmxthanks llorean
00:34:10peturvmx: check the link ^^
00:34:55rasherHrm, can't Rockbox somehow detect if it's running the 64mb version on a 32mb ipod?
00:35:09vmxi just wanted to check out what fixes a bug i tried to fix quite a few hours
00:35:28bluebrotherwasn't that related to the codecs running from the end of the ram?
00:36:15rasherYes, but I mean if the user installs the wrong version, would it be possible to flash a big "OH NO" screen, to at least tell the user he's messed up?
00:36:47rasherI'm not talking about adapting to 32/64mb, just telling the user if he has the wrong version.
00:37:01rasher(64mb build on the 32mb ipod)
00:37:38rasherAnyway, why's it labelled "iPod Video 64MB", rather than 80GB? I doubt people would know how much RAM their player has.
00:38:45Lloreanrasher: Because it also covers the 60gb
00:38:53LloreanMeanwhile, the 32mb build will work on either player
00:38:59rasherwell, then 60GB/80GB
00:39:01LloreanAnd, for some users, this is actually preferable.
00:39:07 Join krazykit [0] (
00:39:28 Join DB42 [0] (n=wy@
00:39:30rasherThe point being, people should not by accident run the 64mb build. There's a pretty good chance they'll do that as it is now.
00:39:30*petur wonders how to describe the old file extensions behaviour
00:39:45DB42is it possible to access the ipod nano's serial port from an rockbox application ?
00:40:20Lloreanrasher: The 32mb one is simply called 'iPod Video', people who're confused should be expected to install that since it doesn't mention RAM at all, right?
00:40:49DerPapstpetur: just keep 'em secret :P
00:40:57rasherLlorean: But people might install the 64MB version thinking it's better, not realizing it won't run on their ipod video.
00:40:58DB42 <−− why no ipod nano ifno ?
00:41:31DerPapstrasher: that already happend a few times
00:41:52rasherDerPapst: Exactly.
00:42:02Lloreanrasher: Well, 60/80gb might work, but then they don't know what the real difference between builds is, nor can they have any chance of picking the right one for thir playback behaviour.
00:42:14LloreanIt might be better if they were described clearly in the manual
00:42:22DB42anybody can help ?
00:42:38LloreanDB42: Can you fill in the missing ones?
00:43:07DB42ahh, i have no information on that, i was just wondering
00:43:23DB42i'm looking for a way to access the serial port of the ipod using rockbox
00:43:27DB42and was wondering if it's possible @ all
00:43:39DB42(via a plugin code)
00:43:47LloreanNot yet, until someone writes a driver
00:44:06DB42i see
00:44:35rasherLlorean: But really, I'm going to bet that less than 1% of all users know how much RAM their player has and even less would think of running a build utilizing less RAM. And those few users would probaby know to pick the 30GB build anyway, even if it was labelled as such. We're giving them a choice they can't make an informed decision about.
00:45:34 Join Echelon [0] (n=SweetOne@
00:45:34IA4004I'm an user who dont know how much ram has the player I use
00:45:53 Quit tihoc4n ("Leaving.")
00:46:11DerPapsti think a splach screen that informs the user that he has picked the wrong one would be ok (meaning using a 64MB one on a 32MB one. not vise versa). that wouldn't be to costy
00:46:23 Quit DB42 ()
00:46:37IA4004but thats the lazy way
00:47:11rasherIt's better than just letting the user crash and burn
00:47:42DerPapsthe will come here and start complining
00:47:46IA4004I think is correct to put a "blue screen of the death"
00:48:05Lloreanrasher: As I said, it should be documented in the manual
00:48:17LloreanAnd the splash screen might be nice.
00:48:25rasherLlorean: we both know that won't help much, but yes, it should.
00:48:41Lloreanrasher: I think it'll actually help if you put it in the right place, and in red.
00:48:56Lloreanrasher: Another thing is: iPodPatcher could install slightly different bootloaders
00:49:04LloreanThe one for 30gb iPods could simply refuse to load the 64mb version
00:49:06IA4004wich color for the splash screen?,I guess that a red color screen of the deat will be fine
00:49:08pixelmawell, he said he read the manual over and over
00:49:19rasherLlorean: That would make sense
00:49:48LloreanRight now they're the same bootloader, but since there's no user-choice in the bootloader install, it'd be fine to add a simply check in the 32-mb one, assuming the builds can be compared easily
00:50:34DerPapstcodewise the splash would be easier
00:50:50IA4004Lorean why you just make them compatible between each other?
00:51:07pixelmacouldn't that be achieved with the help of the rockbox-info.txt? - if ipodpatcher or rbutil know what kind of Ipod is connected
00:51:10IA4004why not you just make them compatible between each other?
00:51:13 Quit Febs (Read error: 104 (Connection reset by peer))
00:51:25DerPapstand i'm not sure if ipodpatcher can 100% distinguish between 32/64MB
00:51:26rasherIA4004: because this is a hard problem to solve.
00:51:53DerPapsti've seen an iPod Video with a 60GB HDD and 32MB ram
00:52:21IA4004why dont just use virtual memory?,to make them compatible
00:52:22DerPapstbecause its mobo was replaced by some company
00:52:35DerPapstno mmu perhaps?
00:52:40pixelmacouldn't it also find out what size the RAM is?
00:52:41LloreanIA4004: Because virtual memory would entirely defeat the purpose of the compressed audio buffer
00:53:19Lloreanpixelma: I seem to recall there was some discussion on how to do that, but nobody came up with a good way so the separate builds were done instead.
00:53:22DerPapstpixelma: no. it can only detect the size of the HDD and guess the correct size
00:53:48LloreanThat's another problem
00:54:00DerPapstpixelma: where you talking about rockbox or ipodpatcher?
00:54:14pixelmaipodpatcher (or rbutil)
00:54:18*DerPapst was talking about ipodpatcher
00:55:13linuxstbDerPapst: Maybe the scsi inquiry method could distinguish, but ipodpatcher doesn't do that (yet).
00:55:31DerPapstlinuxstb: i doubt that too
00:55:31pixelmahow does the appleOS know?
00:55:54DerPapstpixelma: it distinguishes on runtime
00:56:46pixelmaah ok
00:57:13DerPapstthe same apple os runs on 32MB ram and 64MB ram ipods-
00:57:27DerPapstthey have seperate builds for G5 and G5.5
00:58:41DerPapstwasn't amiconn talking on devcon about a way to unify 32MB and 64MB builds?
00:59:07linuxstbYes, a solution has been talked about even before then - it's just that no-one has implemented it.
00:59:25linuxstb(reshuffling how Rockbox uses memory)
00:59:44DerPapstputting the music buffer at the end or ram or something similar
00:59:55IA4004using virtual memory
00:59:58pixelmayes it's been discussed and the knowledgable people agreed on that it's a good solution but it'll need some more work
01:00:08DerPapstvirtual memory = BAD
01:00:17linuxstbDerPapst: Exactly, so it can expand into the extra 32MB if it exists. (or extra 6MB on Archos, extra 16MB on iriver h1x0...)
01:01:10IA4004one way to have space is to shrink things,or remove plugins
01:01:28DerPapsta good solution... but difficult to realize i guess...
01:02:00DerPapsti bet it's not done with "simply" changing the linker scripts ;)
01:02:00IA4004just put a light version of each DAP
01:02:08LloreanIA4004: Removing plugins doesn't add any space unless you simply remove the ability to play plugins.
01:02:14pixelmaand so it was decided to provide a seperate build for the 64MB ram version as a fast way to provide one at all, faster than waiting for the rclean implementation
01:02:48*DerPapst is however thankfull to the 64MB version
01:04:46IA4004wish I could help on this,I dont know how to program
01:05:49DerPapstout of couriosity.. what has to be done to change the ram structure?
01:06:17linuxstbMove the plugins and codecs to the start of RAM (but after the interrupt vectors). Currently they're at the end of RAM.
01:06:58linuxstbCurrently it's something like [rockbox.bin,audio buffer,codec,plugin] and we want it to be [codec,plugin,rockbox.bin,audio buffer]
01:07:14 Quit ender` (" I have the heart of a small boy. I keep it in a jar on my desk.")
01:07:44pixelmaDerPapst: petur/devcon2007/images/dscf0014.jpg"> ;)
01:08:02DerPapstthe linkerscripts are... umm "responsible" for the ram layout?
01:08:32linuxstbThe limitations are the the codecs and plugins need to be a fixed address in RAM, and rockbox.bin varies in size between builds.
01:09:23linuxstbDerPapst: They're part of it. The low-level startup code (crt0.S) may need adapting for each target, as well as the bootloaders.
01:09:27DerPapstbut isn't that an issue with the current design too? (<−− talking about rockbox.bin)
01:10:37linuxstbAnother solution of course is relocatable plugins and codecs...
01:11:37IA4004 I suggest a way to put those things(codec,audio buffer,plugin,rockbox.bin)
01:12:18DerPapstand what is the benefit?
01:12:24*DerPapst can't see any
01:12:37LloreanIA4004: Your last sentence didn't make any sense...
01:12:42tumuwould it be possible to uncompress things on the fly to save ram?
01:12:54linuxstbtumu: What things?
01:12:54LloreanDerPapst: Relocatable means they don't have to be put in any specific place in ram, right now they have to be at a certain address
01:13:05tumuplugins, codecs
01:13:22DerPapsto.O Llorean has a mindreader too
01:15:14IA4004you are like a holy man in a robe who reads minds
01:15:17DerPapstLlorean: my "benefit" comment was for IA4004...
01:15:23 Quit Entasis (Read error: 104 (Connection reset by peer))
01:15:33LloreanDerPapst: Oh, I thought you were asking of the benefit of relocatable codecs.
01:15:53DerPapstSo it looked to me like you were reading my mind and guessing my next question ;)
01:17:14DerPapstrelocatable codecs and plugins sound nice...
01:17:25IA4004what means relocatable?
01:17:36DerPapstlook ~ 15 lines above
01:18:13IA4004ok I've took a look
01:18:23DerPapstbut i guess it's even harder to realize than the new ram layout
01:18:28linuxstbtumu: How would that save RAM? You need to uncompress them to run them...
01:19:00IA4004thats the trouble
01:19:00LloreanDerPapst: I seem to recall other benefits too, but don't remember 'm
01:19:32linuxstbDerPapst: One problem with rearranging RAM is that the current linker scripts are a bit of a mess and could do with cleaning up before changing how they work.
01:19:49linuxstb(mess == #ifdef hell)
01:20:20IA4004that is another problem,when you do arrangements in anything,you may be encounter with troubles
01:20:29DerPapstcan those scripts be split up to go in the target tree?
01:20:53linuxstbDerPapst: That's been suggested.
01:21:25pixelmaI think that's what amiconn suggested... when looking at the left side of the pixture, I linked, you'll see the "vectors" at the top. IIRC correctly that's what he added but I must admit that I didn't follow this discussion too closely as I don't know much about those details...
01:21:44 Quit bluebrother ("leaving")
01:21:54*pixelma gives up correcting typos
01:22:23DerPapsthahaa.. my typoclasses are a huge success :D
01:22:53*petur sits at a back bench, sleeping
01:23:46*DerPapst wonders how petur can type while asleep
01:23:55peturwatch me
01:24:03 Quit petur ("Zzzzzz")
01:24:21DerPapstgood night petur (for the logs)
01:25:11 Join webguest03 [0] (i=425c883c@gateway/web/cgi-irc/
01:25:19 Part Llorean
01:26:28DerPapstpixelma: afaiu there are the interrupt verctors. they are already at the beginning of the ram and stay there (the only thing that doesn't change with the new ram layout)
01:26:33webguest03hello. i have a question about rockbox i 'm trying to play mpeg videos on my ipod video but there seems to be problem with the audio can someone please help me?
01:27:22webguest03i think that either the audio or the video is to fast and cant keep up.
01:27:41krazykitholy... sansa e250 for fifty bucks?
01:27:56krazykitwebguest03, have you been following the instructions for making video on the wiki?
01:28:14*DerPapst doesn't know anythin about mpegplayer. but suggests to use the OF for video playback because it's a lot faster
01:28:28webguest03not really because i dont really understnd what to do.
01:28:35DerPapststop the missig "g"
01:28:44DerPapstthe first gets a virtual cookie
01:28:56webguest03i download the elephants dream movie and i dont have a problem with that video
01:29:00DerPapst*missing even
01:29:23krazykitwebguest03, then you're making the videos wrong. the instructions are pretty clear, really. what part don't you understand?
01:30:04webguest03all the movies are converted the right way its jusyt that sound is a litte faster than the movie.
01:31:03DerPapstgood night at all :)
01:31:14 Quit DerPapst ("So Long And Thanks For All The Fish!")
01:31:25webguest03how am i supposed to make the videos play the right way? Please help me? Please?
01:32:30scorchewebguest03: have you read the wikipage Plugin Mpegplayer?
01:32:57webguest03yes but i don'treally understnd what i;m supposed to do?
01:33:08 Quit Soap (Read error: 104 (Connection reset by peer))
01:33:25 Part pixelma
01:33:25 Join Soap [0] (n=Soap@rockbox/staff/soap)
01:35:52 Join hannesd_ [0] (
01:36:06webguest03i have another two the Brickania plugin or game how many levels are there?
01:36:17IA4004I'll be back,I'm going to get off this piece of junk for a while
01:36:22 Quit IA4004 ("ChatZilla [Firefox]")
01:36:27scorchelay it and find out...
01:36:36scorcheand what do you not understand about that page?
01:37:10webguest03um...everything...sorry im not good at computers
01:37:25scorchewell, i cant help you with "everything"
01:37:28webguest03yoou guy's are the smart ones
01:37:35webguest03hold on
01:37:36scorcheyou need to be specific about a specific peoblem
01:37:48 Quit thegeek (Read error: 110 (Connection timed out))
01:37:49webguest03ok sorry i'll try
01:38:40webguest03i don't understand the stuff about the frameskipping.
01:38:52scorchewhat dont you understand about it?
01:38:57webguest03will that make my ipod play videos better?
01:39:48webguest03the part about something about the audio i think it says 44khrz....something like that
01:39:52SoapFrameskipping will help, but using ffmpeg (WinFF, as discussed in the wiki is an easy front-end) to encode your videos to a lower framerate works better.
01:40:13Soapyes, the sample rate needs to be 44.1, just like a CD.
01:40:37webguest03i already tried the frameskipping ad nothing happens sometimes it makes things worse
01:40:50SoapiPod video I assume?
01:41:17webguest03how can i make the audo 44.1?
01:41:43Soapis it already? If not many audio editors, and even Foobar2000 can resample it.
01:42:12 Quit bdgraue (Remote closed the connection)
01:42:50webguest03do you guy's know about a certain movie converter i can use to convert movies to Mpeg and play them on my ipod without any problems?
01:42:51 Quit hannesd (Connection timed out)
01:42:51 Nick hannesd_ is now known as hannesd (
01:43:52SoapWinFF (the front end for ffmpeg (discussed in the wiki)) is about foolproof if you follow the instructions we posted there. (assuming you are running windows)
01:43:57markunwebguest03: what kind of ipod?
01:44:17webguest03ipod video 5th genreation 30gb
01:44:33markunI think you're better off sticking to the apple OS for playing movies then
01:44:34webguest03generation) sorry
01:45:21webguest03 the thing is i like rockbox a little better and i want to ditch itunes and all that....
01:45:28SoapIMHO only cartoons can be played back decently with Rockbox on an iPod video- and only slow-motion ones at that.
01:45:48Soapwebguest03: you can use iTunes only to transfer movies...
01:46:00linuxstbwebguest03: Rockbox can't play videos well on the ipod video - the PluginMpegplayer page states that (and explains why). So Rockbox isn't able to replace the video playback part of Apple's firmware.
01:46:45 Join advcomp2019- [0] (n=advcomp2@
01:46:51webguest03is there a way to play the videos i have on my ipod while using rockbox?
01:46:57 Quit advcomp2019 (Nick collision from services.)
01:47:03 Nick advcomp2019- is now known as advcomp2019 (n=advcomp2@
01:47:18linuxstbwebguest03: No - reboot into the Apple firmware...
01:49:16webguest03on last question .........please wait.....typing
01:51:04webguest03i have the new iphon and i am able to use it without activating it with a phone company its some sort of you guy's know if rockbow is going to be able to be used on the iphone in the future?
01:51:21webguest03that would be awsome!!!
01:51:34markunand would limit your device a lot
01:51:58webguest03i dont care about using it as a phone
01:52:07webguest03most peaople dont either
01:52:27markunwhat's so good about it if you don't use it as a phone?
01:52:42webguest03ther like it because its 1 touchscreen and 2 it has a bigger screen
01:53:18markunwell, if someone ports rockbox to it I don't think anyone would object
01:54:10webguest03well do you know if its going to be here at rockbox or something? i mean so people can have it on there iphone?
01:54:29markunwhat's going to be here?
01:55:01webguest03the download for rockbox on an iphone?
01:55:03markunwebguest03: here's some info for you guys:
01:56:35 Join Nico_P [0] (n=nicolas@rockbox/developer/NicoP)
01:56:39webguest03oh and i got the ideo from some webpage i saw that says something like that....i think someguy was already able to put rockbox or ipod linux something like that on his iphone...he like foud a way or something.
01:56:52scorcheno one has
01:57:55webguest03i wasn't sure about that thats why i asked. i just saw it on the internet about 2 days ago...i could be a lie..who knows
01:58:18webguest03they could have been lying
01:59:08***Saving seen data "./dancer.seen"
01:59:15webguest03you guy's should even make rockbox a computer operating system like windows and linux...
01:59:39webguest03rockbox is one of the greatest thing i've ever seen
02:00:03scorchethere would be no point to having rockbox as a computer os
02:00:39webguest03you could even sell it and make some money for all the work you guy's put into creating rockbox
02:01:11scorchethere are hundreds of people that have contributed to would we divide up the money fairly?
02:01:32webguest03the point is maybe you guy's could make it the best or one of the best computer os and moake money...
02:01:42scorchethere would be no point to having rockbox as a computer os
02:02:01webguest03yes you guy's can divide the money
02:02:11webguest03just dont be greedy
02:02:32 Part toffe82
02:02:43webguest03it would pay off all the time you guy's have spent on creating rockbox
02:02:56scorchethere would be no point to having rockbox as a computer os
02:03:23webguest03why wouldn't there be no point???
02:03:27scorcheas well, we dont believe in selling rockbox...rockbox has always been free, and there is no reason for it not to be
02:03:37krazykitespecially when there's another Free operating system already established...
02:03:42scorchebecause rockbox is a replacement firmware for DAPs
02:03:48markunwebguest03: do you have a link to the guy claiming to have rockbox or ipod linux on his iphone?
02:04:15webguest03i can look for the site again and give you guys the link
02:04:19scorcheit lacks a serious number of things that would be essential in a regular computer os
02:05:06webguest03all i did was type rockbox on iphone by accedent i ment ipod then i hit search on yahoo
02:05:07 Join Febs [0] (
02:05:14markunscorche: but isn't the most important thing you can play doom? ;)
02:05:18scorcheFebs: just in time for the fun...
02:05:35scorchemarkun: can you not do that on windows or linux? =P
02:05:37krazykitwebguest03, you COULD run a linux kernel, write a target that has a resolution of your screen, make a wps, and run an x-session with only the simulator ;-)
02:05:44FebsHeh. Guess I'd better go look at the logs!
02:05:57scorcheFebs: dont will make your head hurt
02:06:14webguest03but you guy's need to make guy's have work hard for rockbox and you need to get paid it just how it works
02:06:27FebsNo we don't.
02:06:30scorchewe have a donate button on our site
02:07:01krazykitwebguest03, i don't think you quite understand open-source software.
02:07:07scorcheif you feel that we deserve it, feel free to donate however much you wish, and encourage others to do the same
02:07:10webguest03yes you do thats whhy its called wok
02:07:17scorchebut rockbox isnt work
02:07:18markunwebguest03: we're all communists
02:07:21scorcheit is a hobby for us
02:07:43webguest03no you guy's worked on creating rockbox
02:08:01webguest03well then turn your hobby into work and get paid
02:08:28scorchewe have said our reasoning for not doing so...
02:08:51scorcheand as i said, if you feel so badly that we should get money for the project, there is that donate button
02:08:56webguest03well its not like i'm trying to force you guy's or anything
02:09:44webguest03i already donated...well not me my parents often do people donate anyways?
02:09:57scorchei dont know...
02:11:39webguest03well anyways you guys are awsome i really love rockbox and if i were you i would make it a computer os and sell it then i would still keep the rockbox for DAP' free!
02:13:01webguest03well ok see you guy's and thanks alot for the help!!! :)
02:13:47 Quit z35 ("Leaving")
02:14:10scorchefor the last time, there is no point in haivng rockbox be a computer os...
02:14:42krazykitbut i so want rocklinks and rockirc
02:15:58krazykityeesh. browsing the web on any of the targets would be painful.
02:16:13 Quit webguest03 ("CGI:IRC (EOF)")
02:17:44 Join IA4004 [0] (n=chatzill@
02:19:08 Join elinenbe_ [0] (
02:20:13 Quit elinenbe (Read error: 104 (Connection reset by peer))
02:20:13 Nick elinenbe_ is now known as elinenbe (
02:20:22 Join ctaylorr [0] (
02:23:06krazykitwhy did i just buy a sansa? i don't really /need/ a flash player :-/
02:24:56 Join EbErT [0] (
02:25:36scorchebecause it was incredibly cheap?
02:26:02krazykitpretty much. i couldn't say no to a $50 player.
02:26:08*krazykit curses
02:32:42 Join Shishire [0] (
02:32:59Shishireanyone else having trouble booting today's build?
02:34:51krazykitwhich revision?
02:35:50krazykitwhich target?
02:36:02ShishireiPod video
02:37:12krazykiti'm building it now, but i don't have an ipod.
02:37:22krazykitis there any specific error, or does it just not load?
02:37:31Shishireit gets to the loading screen and then does nothing
02:37:50Shishirethe rockbox splash screen, I mean
02:40:47krazykithm. works on the gigabeat, anyway.
02:40:58krazykithave you tried a fresh .rockbox?
02:41:20Shishireok, It's probably a conflicting file. I'll wipe and reinstall
02:41:23krazykityou could move the current .rockbox to something like .rockbox-bak and unzip the latest build again.
02:41:30 Join toffe82 [0] (n=chatzill@
02:41:38krazykitjust to make sure, of course
02:44:15 Join jhulst [0] (
02:55:28 Join sssSCH [0] (n=ssssch@
02:55:34IA4004erm if someone needs help with sansa's and if I'm not online call me,0115218671189687,and if you wish take the numbers as numbers for the lotto;)
02:55:39 Quit IA4004 ("ChatZilla [Firefox]")
03:03:01 Quit Thundercloud (Remote closed the connection)
03:05:01 Join sk7 [0] (
03:05:49sk7i booted the original apple software via the dualboot option but how can i boot rockbox now again? whenever i switch my ipod on it boots apple software, running ipodcatcher again didnt help?
03:06:19Shishireif the hold switch is off during boot, it should boot to rockbox
03:06:42Shishireto reboot, hod menu and select for ~10 seconds
03:07:15sk7Shishire: thanks, that helped!
03:07:34Shishireyou're welcome
03:10:45Soapthe hold switch turned on during boot will not launch Rockbox, it will launch Apple FW, or (if done after the apple appears) it will clear your settings.
03:11:56Soapyou need to do the hard reset (Menu+Select) because Apple firmware does not shutdown the iPod fully until there have been 48 (?) hours of inactivity, otherwise it just sleeps the iPod, and doesn't use the bootloader, thus the patched bootloader you installed with ipodpatcher won't kick in and lauch Rockbox.
03:19:16 Join thegeek [0] (
03:25:50 Join IA4004 [0] (n=chatzill@
03:31:29 Quit sk7 ("foo")
03:35:32 Quit Nico_P (Remote closed the connection)
03:41:19 Quit Wofl (Remote closed the connection)
03:41:58 Join Wofl [0] (
03:47:45 Join thegeek_ [0] (
03:53:23 Join RxDx [0] (
03:56:44RxDxhow can i restore my ipod (in windows theres a software that restore it)?
03:57:14 Join JdGordon [0] (n=Miranda@
03:58:11IA4004did you installed rockbox on it?
03:59:09***Saving seen data "./dancer.seen"
04:00:01RxDxand now i want to restore it... but im using linux
04:00:07RxDxand i cant run iTunes
04:00:16IA4004Ipod patcher
04:00:22RxDxfor linux?
04:01:04IA4004I guess there is a version for linux
04:01:07IA4004hold on
04:01:14IA4004just a sec
04:01:27RxDxok.. tnx :)
04:01:38IA4004wich ipod?
04:01:45RxDxif i can restore my ipod from linux.. ill leave my windows for ever :)
04:01:48RxDxnano 1st gen
04:03:22 Join aliask [0] (
04:03:51 Quit thegeek (Read error: 110 (Connection timed out))
04:04:36JdGordonrasher: : hey, nice trasnlation site
04:04:37 Part Shishire ("bye everyone")
04:05:45 Quit IA4004 ("ChatZilla [Firefox]")
04:13:26 Quit EbErT ()
04:15:08 Join perrikwp [0] (n=chatzill@
04:19:28 Quit sssSCH ("Leaving")
04:21:12 Quit jhulst (Remote closed the connection)
04:29:48 Quit JdGordon (Read error: 110 (Connection timed out))
04:35:24aliaskWhen a translated string isn't available in a lang file are you supposed to leave it as the english in the dest, or "" or none?
04:38:27 Quit Rondom ("Ex-Chat")
04:45:12 Join IA4004 [0] (n=chatzill@
05:27:12 Join saratoga [0] (i=98039a34@gateway/web/cgi-irc/
05:27:31 Quit IA4004 ("ChatZilla [Firefox]")
05:41:41 Join nick89 [0] (
05:41:55 Quit nick89 (Read error: 104 (Connection reset by peer))
05:49:09aliaskDoes the voice work in the sim yet?
05:49:10 Quit RxDx ()
05:55:56 Join Llorean [0] (
05:56:15 Join autonemith [0] (i=45f28ba9@gateway/web/cgi-irc/
05:58:40 Quit donutman25 ("ChatZilla [Firefox]")
05:59:11***Saving seen data "./dancer.seen"
06:05:12 Quit ctaylorr (Read error: 110 (Connection timed out))
06:05:13 Quit Febs (Read error: 110 (Connection timed out))
06:12:29 Quit autonemith ("CGI:IRC (EOF)")
06:20:09 Quit perrikwp (Read error: 104 (Connection reset by peer))
06:20:39 Join jhulst [0] (
06:20:50 Join perrikwp [0] (n=chatzill@
06:28:21 Quit jhulst (Read error: 104 (Connection reset by peer))
06:43:25 Join Kulawend [0] (
06:50:05 Quit Wofl (Remote closed the connection)
06:50:08 Join nick89 [0] (
06:50:17 Quit perrikwp ("ChatZilla [Firefox]")
06:50:31 Quit nick89 (Client Quit)
06:57:34 Join vmx_ [0] (
06:58:19amiconnlinuxstb: I agree that fs #5226 is made obsolete by getting rid of sectioned compilation, removing most long calls this way. And removing *all* code from iram is in fact a bad idea, at least if someone is going to implement suspend on PP502x.
06:59:33amiconnThe suspend /wakeup code itself needs to be in IRAM because the sdram is also put to sleep, keeping its contents but making it inaccessible
07:00:41 Join Kulawend8221 [0] (
07:00:50 Quit vmx (Read error: 110 (Connection timed out))
07:07:52 Quit Kulawend8221 ("Miranda IM! Smaller, Faster, Easier.")
07:13:44 Join BobShield [0] (
07:13:44 Quit BobShiel1 (Read error: 104 (Connection reset by peer))
07:14:47 Quit saratoga ("CGI:IRC (Ping timeout)")
07:15:27 Quit Kulawend (Read error: 110 (Connection timed out))
07:21:12 Join davina [0] (
07:24:10 Quit rasher (Remote closed the connection)
07:24:45 Join rasher [0] (
07:43:39 Join perrikwp [0] (n=chatzill@
07:59:15***Saving seen data "./dancer.seen"
08:01:57 Join linuxstb_ [0] (i=5343d4aa@rockbox/developer/linuxstb)
08:02:41linuxstb_amiconn: Thanks for the reply - I'll close it.
08:03:08linuxstb_Ah, Petur already did ;)
08:03:55linuxstb_amiconn: How are the 1g/2g ports going? Is the bootloader ready for a release?
08:04:20amiconnThe bootloader itself works, but all code is still untested on 1st gen
08:04:54amiconnAnd suspend isn't in yet; I need to add sys power messaging
08:05:27linuxstb_You don't have your 1g yet?
08:06:15amiconnI don't know whether you were around when I reported it, but the "1st gen" I got from ebay turned out to be a 2nd gen as well
08:06:30linuxstb_No, I didn't see that.
08:06:53amiconnI then found another one (defective), probably just the hdd
08:07:15amiconnI bid on that one as well, but that can't be here yet. Auction just ended Friday evening...
08:08:19 Join inversion [0] (n=none@
08:37:08 Join JdGordon [0] (
08:37:26GodEateris rbutilqt not in a state that builds currently ?
08:38:32GodEaterah no - my mistake - lacking accessibility in my Qt lib
08:39:20 Join ptw419 [0] (
08:40:43 Join ender` [0] (
08:48:15 Join Rob222241 [0] (
08:48:15 Quit Rob2222 (Read error: 104 (Connection reset by peer))
08:49:12 Part toffe82
08:53:25 Join perrikwp_ [0] (n=chatzill@
08:57:34 Join Rob2222 [0] (
09:00:03linuxstb_Anyone know how the SoC projects are going? IIUC, the students are expected to be finished by 20 August...
09:02:31 Join petur [0] (n=petur@rockbox/developer/petur)
09:02:33GodEaternot long to go now!
09:03:15linuxstb_And a mentor has just arrived...
09:03:24peturoh oh
09:03:36linuxstb_petur: Thanks for closing that IRAM task - amiconn's confirmed what I thought last night.
09:03:59*petur is still in the tracker closing week :)
09:04:02linuxstb_petur: So I guess there's not going to be any surprises this week?
09:04:07linuxstb_(regarding USB...)
09:04:17 Join bdgraue [0] (
09:04:26peturI fear not. We will need some serious help to get this one through
09:06:48peturaustriancoder needs the assist of a good hacker with the correct target with him... jhMikeS, MrH, .... I do not have a pp player (my bad) and I'm not the top hacker here
09:08:36linuxstb_True, but other devs don't get interested if there's no discussion in IRC about things...
09:09:24 Join B4gder [0] (n=daniel@
09:11:02peturB4gder: would MrH be interested in finding out why his usb code doesn't work inside rockbox?
09:11:21linuxstb_aliask: I've just read your superdom patch (FS #7528) which replaces calls to rb->sleep() with a while() loop repeatedly calling rb->yield() and wondered why?
09:12:16linuxstb_aliask: IIUC, the problem is that superdom should yield whilst performing complex calculations (the AI functions?). The calls to rb->sleep() are fine as they are.
09:12:18B4gderI've been reluctant to pull MrH's sleeve since we do have his (MrH's) working code to compare with and it won't be very easy for MrH either too review why it doesn't run in the Rockbox context - remember that MrH is not really familiar with Rockbox internals
09:12:58linuxstb_petur: I thought things were working once they were moved to IRAM?
09:13:10peturreceiving was
09:13:17peturtransmit still fails
09:13:21 Quit perrikwp (Read error: 110 (Connection timed out))
09:13:23aliasklinuxstb_: Ah, thanks for pointing that out. I'll redo the patch when I get a chance
09:14:10peturit may have to do with the chip being configured differently... austriancoder refuses to run e200tool to get a dump of some registers
09:14:50GodEaterwhy ?
09:15:03peturhe fears he will ruin his player
09:15:19 Join Zagor [0] (n=bjorn@rockbox/developer/Zagor)
09:15:28B4gderit won't
09:15:29GodEateris that a legitimate fear (not being a sansa owner I've no clue) ?
09:15:37B4gderno, it's perfectly harmless
09:15:45B4gderyou can load into into its ram and execute it
09:15:46petursomebody tell him then...
09:15:49linuxstb_Why do you need to dump registers from e200tool?
09:15:54 Quit Rob222241 (Read error: 110 (Connection timed out))
09:16:03B4gderlinuxstb_: to see if there's a different init situation
09:16:04peturbecause the usb code works in e200tool
09:16:10B4gderfor manufacture mode vs "normal"
09:16:12JdGordonif you tell me how to dump the registers ill do it
09:16:25JdGordonunless thats added hacking?
09:16:29B4gderit is
09:16:36JdGordonoh ok
09:16:41B4gderthe current e200tool can upload the usb code and run it
09:16:49B4gderthat's how usb works with it
09:16:58B4gderyou'd instead upload the register dump code
09:17:04B4gderwhich isn't written
09:17:39GodEaterit sounds very likely that this is the case doesn't it ?
09:17:43JdGordonwell, my sansa is a willing gueniea pig if somone writes that code
09:18:07peturI may have time tomorrow...
09:18:36linuxstb_Could you just add it to the debug menu, and then upload rockbox.bin via e200tool?
09:18:56GodEaterwouldn't that change the init though ?
09:19:03GodEatersince it would run rockbox's init code ?
09:19:13GodEaterso defeating the object of the exercise ?
09:19:56linuxstb_It depends which registers we're interested in. If they're not touched by Rockbox, I can't see why they would change. But it would seem an easy first test.
09:20:29GodEaterbut as I understand what's been said, we think it IS registers that rockbox is touching, since otherwise the usb code would work ?
09:20:49B4gderit could be either way really
09:20:51linuxstb_Then disable the usb init in this test build.
09:21:05peturisn't there a reset between manufacturing mode and normal?
09:21:10B4gderyes there is
09:21:31peturso manufacturing mode could do some init we need
09:21:59peturloads of guesswork :/
09:22:44GodEateris all this work Sansa specific ?
09:22:51GodEateror is it supposed to work on all PP targets ?
09:22:59peturfor all pp
09:23:08GodEaterso has anyone tried it on the ipods ?
09:23:18petursame result
09:23:36peturI really must have another good look at his code...
09:24:17*linuxstb_ wonders how the Rockbox SoC students are expected to upload their code to - an entire Rockbox source tree?
09:24:31GodEatersurely just a patch ?
09:25:02linuxstb_Both seem pretty pointless...
09:25:29*GodEater is wearing his "stupid" hat today
09:25:33linuxstb_I mean a central repository of all SoC patches or projects...
09:26:11linuxstb_For Rockbox, either the code should be uploaded to flyspray, or committed to SVN. I assume all other established projects are the same.
09:26:29GodEaterbut sending it to is required is it ?
09:26:44JdGordonpetur: its easy enoughh to try linuxstb_'s idea and dump the reg's int he debug menu... which do you want?
09:26:48linuxstb_It's mentioned in the project timeline. I haven't read anything else about it.
09:27:12*linuxstb_ bbl
09:27:13 Quit linuxstb_ ("CGI:IRC")
09:27:22peturJdGordon: to start with
09:27:41peturbut better would be all usbcontroller registers
09:28:06*petur fears it may still be another one outside this range
09:28:27JdGordonok, i was expecting a few.. not 100 :p
09:29:38 Quit inversion (Read error: 110 (Connection timed out))
09:29:45peturcan we access disk (flash)? maybe dump the whole range to file?
09:30:40B4gderyes, but then we need the full flash and fat code etc in that code
09:31:01B4gderI would start with dumping them all on the screen
09:31:26JdGordonwould runing rockbox.mi4 like linuxstb sugegsted work?
09:31:38B4gderwell, not the mi4
09:31:48B4gderand you'd probably have to link it to a different address
09:31:54B4gderbut yeah, it should work
09:32:25JdGordonthe .bin?
09:32:44B4gderyes, but it may still need adjustment
09:32:49jhMikeSpetur: the derived code that I saw doesn't do what the MrH code does. if a reset is requested it longjmps back to the main loop. the udelays for timouts were also replaced with counters which surely is wrong.
09:32:55B4gderI don't remember the exact e200tool details
09:33:40peturjhMikeS: well I told him about the counters, didn't see the longjmp
09:33:48B4gderthe benefit of going the full rockbox-via-e200tool would be that it would also create a perfect foundation for futher e200R research
09:34:48B4gderif there against all odds appears someone with *both* clues and and R model in the future
09:35:29jhMikeSe200tool can be injected into the R device?
09:35:50B4gderwe got the R model bootloader code using it
09:35:59jhMikeShmmm...can't overwrite the internal sig with some other one?
09:36:09nerochiaroquick question, i see that your code is compiled with -ffrestanding and -fnostdlib. is that because rockbox is its own kernel, and if picking up some rockbox code to run on linux they can be omitted ?
09:36:20B4gderjhMikeS: no, since we don't know where the original bootloader is stored...
09:36:34B4gdernerochiaro: in short: yes
09:36:58B4gderwe have those since we provide all those functions ourselves, which you wouldn't on linux
09:37:12jhMikeSwhat mean "original bootloader"? :P not sure how you got the bootloader then.
09:37:23nerochiaroB4gder: ah, makes sense. thanks
09:37:37B4gderjhMikeS: we copied the bootloader code from RAM since it is left there when manufacture mode starts!
09:37:42 Join ddalton [0] (
09:37:45 Join LinusN [0] (i=linus@rockbox/developer/LinusN)
09:38:02B4gderon the exact same address as the normal e200 model has it
09:38:48jhMikeSso it copies it from some ROM or flash into RAM at startup
09:38:56JdGordonok, so anyone remember who to run some code with e200tool?
09:39:08B4gderJdGordon: check the tarball
09:39:39*GodEater is sure the "sansa bricked by recovery mode" thread will contain some nuggets of useful info too
09:39:50JdGordonit distincly lacks documentation....
09:39:55JdGordonwrite? or run?
09:40:25B4gderJdGordon: it wasn't written to be education I believe so you need to figure out how it works
09:42:58B4gderbtw guys, thanks all for the kudos I've received on ohloh, I'm now ranked #58 with one of them shiny "10" icons
09:43:03GodEaterand when you find out - for god's sake write a wiki page for it
09:43:55ddaltonhey I found the line { ACTION_STD_REC, BUTTON_REC|BUTTON_REPEAT, BUTTON_NONE }, in a key map file what do I need to change to make the rec button go to the radio?
09:44:28 Quit hcs ("Leaving.")
09:44:59JdGordonalrighty.. what address is rockbox.bin linked to? or more precicsly.. what address should i try writing it to to see if it will actually work?
09:47:55B4gdercheck the init code
09:48:21B4gderit send over the e200code which is the arm_code.c compiled code
09:50:23peturit's written in a comment somewhere
09:51:45JdGordonhmm.... i wrote the rockbox.bin file to DRAM_START and then ran it and i got a white screen... so...
09:52:33JdGordonyay, crashed, but hard reset and not bricked
09:53:35B4gder * arm-elf-gcc -Os -fno-unit-at-a-time -mcpu=arm7tdmi
09:53:35B4gder * arm-elf-ld -Ttext 0x40004000 -N
09:53:35B4gder * arm-elf-objcopy -Obinary
09:54:20B4gderthat's how the code is linked/addressed that is loaded to the target within the init command
09:54:20linuxstbJdGordon: You could also try the bootloader (bootloader/bootloader.bin in the build directory)
09:54:34B4gder40004000 is IRAM isn't it?
09:55:12advcomp2019do you need any info off of a r version
09:55:42B4gderadvcomp2019: yes, but someone needs to write code first to extract that info ;-)
09:55:53peturyes, the usbcode of MrH runs in iram
09:56:05advcomp2019o ok
09:56:32B4gderah right, I'm stupid, the usb code of course still needs to be there and everything
09:57:00JdGordondo you know what address to try running the bl from?
09:57:17B4gderJdGordon: it is possible that you should rather use base 0x10600000
09:57:35B4gderas we know that works for the bootloader
09:57:36JdGordonfor which?
09:57:43B4gderI mean
09:57:55B4gderpossibly start of the ram is used in manufacture mode
09:58:21B4gderand we know that you can load the sansa bl at 10600000 and run it fine
09:58:34B4gderso loading rockbox.bin linked to that address _should_ work
09:58:53*JdGordon doesnt know how to change the address its linked to
09:59:18***Saving seen data "./dancer.seen"
09:59:32B4gdercheck firmware/
09:59:44B4gderDRAMORIG is the define
10:00:21B4gderand you can verify it afterwards by checking the .map file
10:02:39JdGordonwhat am I looking for in the .map?
10:03:20B4gderthe .text line indicates where the .text starts
10:03:48B4gderwhich should then be 0x10600000 and not 0x0 I believe
10:04:02JdGordon.text 0x40000000 0xc0dc
10:04:16JdGordonwhich si iram accorindg to
10:04:33nerochiaroanother question about compile options: is there a specific reason why it was chosen -O as optimization level ?
10:04:46B4gderJdGordon: are you talking about the bootloader?
10:05:07B4gdernerochiaro: we play around with different options for different CPUs and code all the time...
10:05:16JdGordonill go back to rockbox.bin
10:05:47JdGordonstupid e2tool doesnt want to work anymore :(
10:05:50JdGordoncant find the device
10:05:57nerochiaroB4gder: ok, i was just asking to see if there was a specific reason, just curious
10:06:01B4gderJdGordon: either way, you should change the base to be 0x10600000 no matter which code you use
10:06:40B4gdernerochiaro: I bet there is, I can't remember any right now and it is also possible that it just has been left like this for now without much effort behind figuring out the best one
10:07:22 Join miepchen^schlaf [0] (
10:07:34nerochiaroB4gder: ok, thanks. i'll try with different levels to see what happens then
10:08:59 Quit BigMac (Remote closed the connection)
10:09:21GodEatersomeone tried with different levels a couple of months back
10:09:33GodEaterand we ended up back where we are now
10:10:22JdGordonB4gder: same story... rockbox.bin runs but just a white display
10:10:44JdGordonjhMikeS: is the lcd setup in rockbox? or does it rely on the bootloader to init?
10:10:53B4gderJdGordon: at 1060 ?
10:11:17JdGordonwheel and backlight work.. so i tihnk it is actually running
10:11:45peturenable voice and use it blind :)
10:12:16 Join ie [0] (i=c27f0811@gateway/web/cgi-irc/
10:12:17*JdGordon grabs a english.voice file
10:12:27iepetur: ping
10:12:58iepetur: great commit on file extensions!
10:13:03JdGordonoh, and it looks like e200tool doesnt disconnect properly... I have to boot into OF before it will work again after running
10:13:29peturie: np, at least somebody is happy with it...
10:13:51*petur wonders if amiconn is using the 4th option...
10:14:02ieOne question though: why don't you check for the directory right on the start? It could save some ms if the dir contains many subdirs.
10:14:17peturie: good point....
10:14:36amiconnpetur: Didn't upgrade yet
10:15:24JdGordonlol @ the voice
10:16:12peturamiconn: the way it worked before is default, I hope you're happy ;)
10:16:23linuxstbJdGordon: The bootloader can be run from anywhere in DRAM. The first thing it does is copy itself to IRAM - hence the reason it's linked to 0x40000000.
10:16:39JdGordonah ok
10:16:40linuxstbJdGordon: (The initial startup code is position-independent)
10:19:25jhMikeSJdGordon: I disabled the init outside the bootloader. All it does in fw is change the framebuffer address.
10:20:12JdGordonwould that explain the white screen?
10:21:25jhMikeSnot really, it explains the black screen though. the white was already there before that.
10:21:58JdGordonbootloader doesnt run at all with e2t
10:22:26JdGordonbut it eventually loads the OF, dont know if thats supposed to happen in the manfact mode or not
10:24:22 Part ie
10:29:30 Quit DMJC (Read error: 113 (No route to host))
10:29:46 Join DMJC [0] (
10:30:31amiconnB4gder: Regarding the e200r, does the USB code from e200tool allow to access flash? If so, dumping the flash would allow further investigation
10:31:05B4gderamiconn: no it doesn't, but we should be able to upload code to it that scans/investigates the flash
10:33:17 Join nick89 [0] (
10:36:04JdGordonhow do I compile arm_code.c? runing the command at the top of the file fails, says no crt0.o found
10:36:06 Part tyrion
10:38:13 Quit vmx_ (Read error: 110 (Connection timed out))
10:38:16 Join vmx_ [0] (
10:42:49 Join Entasis [0] (
10:43:50JdGordonB4gder: any ideas?.
10:44:55B4gderyou prolly need options to disable all that stuff
10:45:00B4gderlike -ffreestanding or similar
10:48:42JdGordonok, well that doesnt work, so i guess its time to cut our BL down to doing noting and see how it goes
10:56:55JdGordon:( my modified bootloader is definatly not running
10:58:47JdGordonlinuxstb: you sure i can boot the bootloader anywhere and it should run?
10:58:52aliaskThe joys of low level programming :)
10:59:10JdGordons/programming/testing random things and hoping something works
10:59:31aliask:) too true
10:59:47peturs/testing random things and hoping something works/hacking
11:01:10*aliask needs to do some "hacking" on the gigabeat button driver
11:01:15 Join Nick_Brackley [0] (i=cb57494b@gateway/web/cgi-irc/
11:02:06 Join pixelma [0] (i=pixelma@rockbox/staff/pixelma)
11:02:43JdGordontime to change tactics... hacking apps instead of bootloader...
11:03:08markunaliask: any idea how the buttons are connected?
11:03:29aliaskReasonably sure they're all connected to the KPP port, but reading the values is quite tricky
11:04:07aliaskIt's designed to be a matrix, with rows and colums. I can read the row data, but not the columns (if i understand correctly)
11:06:05aliaskI've also got a keypad driver for linux, but it doesn't seem to work. I'm probably missing something though.
11:08:49amiconnIf it's a matrix, you either drive the rows and read the columns, or vice versa
11:12:07aliaskYep, that's what it looks like the linux code does, but when I read the data back, it doesn't change
11:13:19peturio configured correctly?
11:13:57aliaskI'd say so, because the column and row data is on the same register, and the rows read fine
11:14:38aliaskWhat I have at the moment:
11:14:41peturthe ones you drive configured as output I mean?
11:14:47amiconnThe fact that reading row data works doesn't mean i/o is configured correctly<
11:15:01amiconnIt might be that you need to set the column pins as outputs
11:15:29aliaskI have.
11:15:42peturI see... open drain
11:16:29aliaskWait, I see a problem... I think. Do lines 59 and 60 cancel eachother out?
11:17:19amiconnJust wanted to point you to line 60....
11:18:24peturI don't see the problem
11:18:56peturbut the code is silly
11:19:07peturyou might as well just asign the value
11:19:33aliaskYes, I don't know why they do it like this - it's like that in a few places...
11:20:45amiconnThe code looks like there's a lot of useless stuff in
11:20:54amiconnLike configuring again in the read
11:21:12 Join bluebrother [0] (i=LHGag8Ad@rockbox/staff/bluebrother)
11:22:14 Join kubiixaka [0] (
11:26:26nerochiarohmm, reading saratoga's code is pretty funny. he seems to swear a lot in the comments ;)
11:26:26 Join pondlife [0] (
11:29:19 Quit nick89 ("Gotta Go")
11:29:24 Join nick89 [0] (
11:29:55 Quit nick89 (Client Quit)
11:30:05 Join nick89 [0] (
11:30:28 Quit nick89 (Client Quit)
11:35:00*petur is humming "should I stay or should I go"
11:35:20 Join spiorf [0] (
11:39:39 Quit kubiix (Read error: 110 (Connection timed out))
11:49:08 Quit kubiixaka (Read error: 110 (Connection timed out))
11:54:40 Join Arathis [0] (
11:59:19***Saving seen data "./dancer.seen"
12:03:49Nick_BrackleyDoes anyone know if the usb charging button has been reassigned yet on the h300?
12:07:16peturmaybe short and long press should be swapped. long-press for usb charging, short press for shortcut
12:07:29peturlong-press can also be used for backlight then
12:08:04Nick_Brackleypetur: i have short for my playlist
12:08:43peturwell on devcon we decided that the key would be configurable, not just recording
12:09:27Nick_Brackleypetur: ok
12:09:58Nick_Brackleypetur: i thought that menu was decided as the new usb charging button?
12:10:13peturoh was it?
12:10:21LinusNlong-menu is assigned too
12:10:32peturas AB marker?
12:10:36LinusNquick menu
12:10:44Nick_Brackleyoh my bad
12:10:51peturLinusN: would my idea be ok?
12:11:00LinusNi wonder if the hold button would work?
12:11:29Nick_BrackleyLinusN: i like that idea
12:12:30ddaltonI am working on a patch. an if statement has been removed and in the patch it found the if statement and added code in side it. so what should I do?
12:12:47 Join linuxstb_ [0] (n=linuxstb@rockbox/developer/linuxstb)
12:14:27linuxstb_LinusN: That could confuse some users who wonder why their device isn't being detected when they attach it to their computer. It's far from obvious that the status of the hold button changes USB behaviour...
12:14:47LinusNlinuxstb_: i agree
12:15:18Nick_BrackleyLinusN: any reason why it cant be user configurable?
12:15:30LinusNNick_Brackley: how?
12:16:45Nick_BrackleyLinusN: the user could choose which button is used for usb charging - as long as its not already assigned
12:17:16LinusNwell, as far as i know, all buttons are assigned to a long press now
12:17:34ddaltonwhat about play? in the wps?
12:17:39ddaltonas a long press
12:17:44LinusNsame there
12:17:47LinusNit is assigned
12:17:59ddaltonwhat does it do?
12:18:05LinusNcontext menu
12:18:10Nick_BrackleyLinusN: what about making the recording screen button configurable? i personally never use it and would prefer to assign it to something useful...
12:18:17LinusNddalton: hmm, i was wrong
12:18:32ddaltonisn't that the senter button navigation I think it is called?
12:18:39LinusNddalton: you are correct
12:18:44ddaltonplay the button above stop
12:18:58LinusNddalton: but it seems awkward to require the user to play music to be able to charge via usb
12:19:37Nick_BrackleyLinusN: what does long play do in the menu?
12:19:42ddaltonbut couldn't a long press from anywhere make it charge for usb?
12:19:46ddaltonnothing I think
12:19:48LinusNNick_Brackley: that would still be a problem for users who want long-rec to enter the recording screen
12:20:26ddaltonLinusN the only problem I could see with a long press of play is when the user is in the radio.
12:20:34Nick_BrackleyLinusN: well they would need to use the hold as the switch then or long play - all of which would be configurable?
12:21:01LinusNi'd like a simpler and more general solution
12:21:11LinusNthat fits all targets
12:21:17ddaltonwhy does know one like 5555 I thought it would be good because it can be configured. but it hasn't been updated for h300
12:21:22Nick_BrackleyLinusN: yeah understandable
12:21:43ddaltonLinusN what is wrong with long play?
12:21:51ddaltonisn't it only used in the radio?
12:22:16LinusNddalton: could work...
12:22:33Nick_BrackleyLinusN: what about a prompt on the player when a usb cable is inserted?
12:22:50LinusNNick_Brackley: i am thinking about that, but it may be annoying
12:23:55ddaltonwell just a suggestion but since it isn't used for much I thought it could be good. and then in the radio you might just need to change the switch between scan mode and preset mode with maybe a long press of something else. but I think as you said everything is taken.
12:24:59Nick_BrackleyLinusN: could it be added to the quick menu?
12:25:44ddaltonthat isn't accessible to blind users
12:26:03Nick_Brackleyddalton: good point
12:27:05ddaltonI am blind myself and picked it up. but I can't see what is wrong with a long press of play. It just wouldn't work on the radio.
12:27:32Nick_Brackleyddalton: yeah i like the sound of long play
12:27:59 Quit spiorf (Read error: 110 (Connection timed out))
12:28:13ddaltonI can't write you a patch because can't do buttons.
12:28:27ddaltononly voice stuff and stuff like that.
12:28:47Nick_BrackleyI can't write one either
12:28:52amiconnNick_Brackley: A prompt would be severely annoying imo
12:29:14amiconnIn >90% of all cases I *do* want usb mode when plugging the usb cable
12:29:27ddaltonwell how can I learn how to do buttons?
12:29:34ddaltonanyone know?
12:29:35Nick_Brackleyamiconn: yes i believe so too, i was just trying to explore all options
12:30:15amiconnddalton: Long play is also used in other places
12:32:36Nick_Brackleycould long rec be used, so that the recording screen is opened only once the button has been released?
12:33:20Nick_Brackleyand if a cable is inserted it doesn;t go to the recording screen
12:34:24ddaltonwhat is the point of the quick screen anyway. because I know someone else wants to remove it. or do people find it useful. Because I was going to make it be voiced and someone said it would be pointless.
12:35:04bluebrotherhow about holding menu while inserting the cable, like the Ipods do?
12:35:16ddaltonthat would be a lot harder to code wouldn't it?
12:35:24linuxstb_bluebrother: long-menu is mapped to the quick menu (even on ipods)
12:35:38bluebrotheryes, but it works ...
12:35:45ddaltonbut I really don't use the quick screen does anyone here use it?
12:35:45linuxstb_Not nicely though...
12:35:58bluebrotherI usually (unintentionally) invoke the quick menu when using it
12:36:06linuxstb_You can't avoid that.
12:36:37ddaltonbecause all the options in the quick menu are in the normal menus.
12:36:44 Join vzabalza [0] (
12:36:45Nick_Brackleyddalton: i dont use it
12:36:59ddaltonthen we would have one long press of a button free.
12:37:07bluebrotheryep. But it's better than opening the recording screen.
12:37:43linuxstb_I think the default behaviour could be configurable though - I can understand that people who listen to music on their devices whilst near a computer will want to use USB to charge far more often than to transfer files.
12:38:09bluebrotherhow about hold? If hold is enabled it doesn't make sense to open a different screen so in that case charging could be enabled
12:38:31linuxstb_bluebrother: "Why doesn't my computer recognise my ipod any more?" ....
12:38:45pixelmabtw... on Ondio to get to usb power mode, you have to hold the "mode" (menu) button - this will open the menu at the same time though as it usually does but that never bothered me (if I understood the issue correctly, otherwise ignore me)
12:38:52bluebrotherhmm ...
12:39:24Nick_Brackleyi would like to use menu instead of the quickscreen...
12:39:33bluebrotherupon insertation we could show a splash "usb charging enabled, disable hold upon connect to use usb connection"
12:40:04bluebrotherbut this could be done regardless of the button
12:40:27 Join Febs [0] (
12:40:30linuxstb_Maybe it could be configurable as follows: 1) Enter USB mode (default - same as now); 2) Charge; 3) Charge on hold
12:40:47linuxstb_That way, users would need to explicitly change the default behaviour to not enter disk mode.
12:41:07bluebrothersounds reasonable to me
12:41:26Nick_Brackleylinuxstb: how would a user enter disk mode from 2?
12:41:55linuxstb_Nick_Brackley: They couldn't...
12:42:24Nick_Brackleylinuxstb: well i dont like that then
12:42:31preglowi wonder how much bigger rockbox would be if we statically linked all codecs
12:42:36preglowsharing the bss segment
12:42:46bluebrotherNick_Brackley: why not? You can disable it and it's not default
12:42:51linuxstb_preglow: Check the size of the .codec files...
12:43:37Nick_Brackleybluebrother: true, thing about it, its not so bad
12:43:46linuxstb_Nick_Brackley: Depending on your device, there could be other ways to enter disk mode - e.g. bootloader USB, original firmware, or the emergency disk mode on ipods.
12:44:12preglowabout one meg bigger :P
12:44:19preglowthat's not too bad
12:44:27ddaltonabout the splash again most blind users aren't using the voice when connecting to there computer. yes it would probably be possible to voice the screen but they wouldn't here anything as they just connect it with ear phones on.
12:44:32Nick_Brackleylinuxstb: fair enough
12:44:45bluebrotherpreglow: how about just using the common codecs? I.e. mp3, ogg and aac?
12:44:59bluebrothermight be interesting for flashing (i.e. h100 series)
12:45:02amiconnpreglow: Not a good idea imho. They couldn't use IRAM for code...
12:45:05linuxstb_Nick_Brackley: Maybe 3) could be "disk mode on hold", rather than "charge on hold".
12:45:18preglowamiconn: why not?
12:45:29linuxstb_preglow: What's the advantage?
12:45:31ddaltonwhere is a long press of play used?
12:45:39preglowlinuxstb_: no advantage, just wondering, really
12:45:48ddaltonactually that wouldn't work on the ipods.
12:45:51amiconnpreglow: How would that be possible? You suggested static linking...
12:45:52preglowi don't think 1meg++ is bad at all for all our codecs combined
12:45:53bluebrotherlong play is stop / shutdown on Ipods
12:45:57Nick_Brackleylinuxstb: yeah that could work too
12:46:00ddaltonbecause it would just shut them down.
12:46:13preglowamiconn: all of them would reuse the same iram area and bss area
12:46:17amiconnIn fact they could, but they all had to share the little iram that's available
12:46:28amiconnpreglow: That wouldn't work for code
12:46:29linuxstb_preglow: One advantage could be that the metadata parsing could be part of the codec - at the moment we have independent metadata parsing in the core.
12:46:40preglowlinuxstb_: yeah, still annoys me, that does
12:46:58preglowamiconn: why not?
12:47:04bluebrotherlinuxstb_: if it's configureable, why not 3) disk on hold and 4) charge on hold?
12:47:13vzabalzahi all, I have updated a language file. Where can I upload it?
12:47:17preglowamiconn: we'd need to copy stuff around, sure, but it doesn't sound impossible to me
12:47:20bluebrotherI would use a charge on hold option but not a disk on hold ...
12:47:26ddaltonso my idea is: (I know noone will like this one but here it is) just remove the quick screen because from what I can tell not many people use it and make a long press of menu do this. That way it would be the same as the apple firmware. Also if you want a quick screen we could put it in some where else.
12:47:27bluebrothervzabalza: the tracker
12:47:32B4gdervzabalza: submit it on the patch tracker
12:47:38vzabalzaas a new patch?
12:47:48vzabalzaok, thanks
12:47:51amiconnpreglow: Then you would loose the advantage of being able to run them in parallel
12:48:11preglowamiconn: can we do that now?
12:48:18linuxstb_bluebrother: Possibly, but I was trying to minimise the options...
12:48:33*bluebrother is strongly against removing the quick screen
12:48:34amiconnNope, but if we can't either with statically linked codes, there's no point to statically link them....
12:48:36preglowamiconn: unless i'm wrong, we have to copy stuff around now to do so
12:48:46bluebrotherit's useful.
12:48:47Nick_Brackleylinuxstb: i would only use charge on hold
12:48:53*linuxstb_ also finds the quickscreen useful
12:49:03BigBambi_i like the quickscreen
12:49:07ddaltonI am not saying it isn't useful but how offin do people use it?
12:49:11preglowamiconn: i was pretty much thinking about this for flashing and the like
12:49:24ddaltonwe only have a limet of buttons on each target.
12:49:24linuxstb_ddalton: Often enough that I would miss it.
12:49:35bluebrotherddalton: how often do people use the settings menu?
12:49:38BigBambi_much much quicker to turn shuffle on and off
12:49:48 Join thegeek [0] (
12:50:00bluebrotherwe also could remove that (and obtain a great size delta) and use cfg files only ...
12:50:19amiconnpreglow: You would only flash the core of course. I don't see a point in static linking. It would only reduce available ram
12:50:20preglowbut no, i was just wondering, i don't really think it'd be beneficial myself
12:50:22 Quit thegeek_ (Read error: 113 (No route to host))
12:50:24ddaltonok fine but I can't see how you are going to make a long press of a button start usb charging. maybe if it can detect the cable or: if you make a short press of rec go to the recording screen.
12:50:39BigBambi_bluebrother: if a button were linked to a cfg folder you mean?
12:50:50linuxstb_preglow: We shouldn't remove the possibility of rombox for more swcodec targets though. The Nano for example only has 512KB of flash, other ipods have 1MB.
12:51:00linuxstb_(I don't know about other targets)
12:51:18bluebrotherBigBambi_: why a button? Use the file browser ;-)
12:51:28preglowif this was to happen, it'd have to easily customisable anyway
12:51:28linuxstb_(I may also be wrong about all ipods, but at least some have 1MB...)
12:51:32amiconnBut codecs could consist of 2 parts, loadable independent of each other
12:51:32preglowbut no, i'm not going to spend time on this :P
12:51:35BigBambi_might as well use the settings menu then
12:51:44Slasheripreglow: i think it could be possible to create a simple filesystem on the flash and copy some codecs there
12:51:53amiconnOne part would be the loader (including metadata parsing), the other would be the decoder core
12:51:57preglowSlasheri: definitely
12:52:02linuxstb_preglow: That would seem a big complication - I wouldn't want to see two different swcodec systems...
12:52:08B4gderjffs2 support!
12:52:20Slasherihehe :D
12:52:34bluebrothersomeone go for it!
12:52:40*linuxstb_ goes for lunch
12:53:00preglowi've got a mate with a sansa now, looking forward to testing that afterwards
12:53:10amiconnpreglow: However, there's one situation where static linking of codecs (and plugins!) will be necessary: if someone revives the gmini port
12:53:26preglowis rockbox sansa runtime very much lower than retailos?
12:53:37preglowamiconn: i really can't see that happening...
12:53:38amiconnAbout 3/4 from what I know
12:53:40linuxstb_I don't think it's as bad as the other ipods.
12:53:50preglowlet's just hope we get no other harvard archs
12:54:34preglowbut installing rockbox on sansas is just a case of running sansapatcher these days, yes?
12:54:40amiconnHarvard arch itself isn't the problem on gmini, it's that it has no ram that can be mapped for code
12:54:46 Quit GodEater_ (Read error: 113 (No route to host))
12:54:55amiconnOnly flash rom can be used for code
12:55:06 Nick BigBambi_ is now known as BigBambi (n=Alex@
12:55:08 Quit GodEater (Read error: 110 (Connection timed out))
12:55:16 Join Nico_P [0] (n=nicolas@rockbox/developer/NicoP)
12:56:22Nico_Ppetur: hi, I've added some ID3 info to the properties plugin nd I thought you might want to see the patch
12:56:31preglowyeah, but isn't that a consequence of harvard arch being used?
12:56:41preglowit's pretty logical to use flash as the code memory
12:57:11B4gdersurely they ran code in ram?
12:57:22B4gderthey never flashed anything afair
12:58:48 Quit Seed (Nick collision from services.)
12:58:55 Join Seed [0] (
13:01:54 Join spiorf [0] (
13:02:48 Quit Seed (Nick collision from services.)
13:02:56 Join Seed [0] (
13:02:59pondlifeHi Nico_P
13:03:12pondlifeI finally had a look at your code and it looks pretty good
13:03:20pondlifeDidn't run it though.
13:03:31preglowhow's mob doing?
13:03:54Nico_Ppreglow: it works in a test plugin but currently I'm crying over playback.c :)
13:03:58pondlifeYou probably already answered in IRC, but .... why does bufopen take an offset? Why not do a bufopen followed by a bufseek for that.
13:04:15pondlifeI'd like your answer to be included in comments :)
13:04:42pondlifeplayback.c is indeed a tear-fest .
13:04:56Nico_Ppondlife: the offset allows to avoid starting from the beginning of a file... think of playback resume
13:05:43pondlifeAh, to save wasting buffer
13:06:02Nico_Pyes, and buffering time
13:06:18pondlifeI was wondering if a bufopen/bufseek combo could achieve the same thing, keeping the API simpler?
13:06:19Nico_Pit was in your proposal :)
13:06:29pondlifeToo long ago :)
13:06:36 Join agm3nt [0] (
13:06:59pondlifeOK, pop a comment at the start of the bufopen() implementation then...
13:07:29 Part Nick_Brackley
13:07:58pondlifeTo help my senility...
13:09:16 Part agm3nt
13:09:58 Quit vzabalza ("Leaving")
13:11:10 Quit linuxstb_ ("Client Exiting")
13:12:33 Quit Zagor ("Client exiting")
13:12:44Nico_Ppondlife: any other remarks ?
13:12:58amiconnB4gder: They had to flash the code. No other way to run it
13:13:01pondlifeNothing majorly wrong that I could see from an eyeball.
13:13:18pondlifeWould be better if testapp.c was called buffering.c, maybe :)
13:13:19B4gderamiconn: I'm quite sure they never did
13:13:29amiconnpreglow: Not necessarily. The MAS is also harvard, and it has some ram that can be used either for data or for code
13:13:35B4gderbut it doesn't matter
13:14:00B4gderamiconn: they started their code with a buffer overflow exploit
13:14:09B4gderwhich shouldn't be possible either if your statement is true
13:14:12Nico_Ppondlife: ok... atm I was kinda wondering how to keep the list of current IDs in the playback code. maybae a static array but it feels a bit of a waste when I have a linked list
13:14:24Nico_Ppondlife: OTOH I don't think the linked list should be public
13:14:32amiconnE.g. the PCM codec is loaded into RAM, and afterwards the ram is switched from data to code
13:14:35pondlifeNo, the linked list should not be exposed.
13:15:21pondlifeCouild the IDs be stored alongside the filenames in the playlist?
13:15:27amiconnB4gder: Why not? Buffers are data in RAM, but the code that's started by the overflow doesn't need to be
13:15:51B4gderamiconn: well, they didn't flash anything and only run things put their through the exploit
13:16:03B4gderso it would have to be in ram
13:16:08pondlifeNico_P: One area where several playback bugs can be found is bookmarking - ideally this should be combined with normal resuming, but might affect your decision here.
13:16:20B4gderbut as I said, its all theoretical now and history
13:17:15pondlifeIt is a bit of a mess currently. I think you should just keep the static array for now, and that can be dealt with seperately once you have a working integration.
13:18:06pondlifeHopefully it'll simplify out and we can then work out a better way (and the array could be much larger with less members in the struct).
13:18:17Nico_Ppondlife: yes, that's true
13:18:52Nico_Pmabe it could be added to struct playlist_track_info but as you said I'd better start with a simple way
13:19:23pondlifeDon't try and rewrite too much at once.
13:19:36pondlifeBetter to evolve your work in to Rockbox.
13:20:05 Join ctaylorr [0] (
13:20:05pondlifeOne other point, could you make sure each routine has a comment to say which thread it is called from.
13:20:22 Join webguest61 [0] (i=c023110b@gateway/web/cgi-irc/
13:20:23pondlifeIf any routines are called from multiple threads, beware.
13:20:34pondlife(Or at least think about it.)
13:21:05 Quit DMJC (Read error: 110 (Connection timed out))
13:21:10 Join barrywardell [0] (
13:21:22pondlifeYou should probably end up with a buffering.c that has only 2 threads involved: the buffering thread and the API calling thread.
13:21:30 Join DMJC [0] (
13:21:52pondlifeLooks like you may be there already, but I was unsure about can_add_handle()
13:22:25pondlifeThat sort of decision probably has to be made on the buffering thread.
13:23:43Nico_Pwhat do you mean by "api calling thread" ?
13:23:58pondlifeWhichever other thread is calling the API
13:24:12*petur executes return-from-lunch instruction
13:24:27*pondlife should execute jump-to-lunch soon
13:24:41pondlifeOr non-maskable lunch may occur...
13:25:34Nico_Ppondlife: I was thinking the API calls could be accessible to an thread
13:25:46pondlifeNico_P: Actually, looks like can_add_handle() is only called from the API calls anyway, so probably fine.
13:25:55Nico_P*any... but the main user would be the playback thread
13:25:56peturNico_P: sure, maybe put i in the tracker, I won't have time until tomorrow I'm affraid
13:26:09pondlifeNico_P: Yes, exactly what I mean - any API calling thread.
13:26:22ddaltondoes anyone know if talk_file_speak has changed?
13:26:58Nico_Ppetur: maybe I can commit it, but I just wanted a quick opinion...;h=5cc053f8e11710e6c7f0a9d3344812e1723d223c
13:27:19pondlifeddalton: SVN knows.
13:27:58ddaltonhow do I look it up?
13:28:41pondlifeddalton: You can see the history of talk.c:
13:29:17ddaltonok thanks
13:29:43pondlifeThe API to any routine shouldn't change much, but the implementation may.
13:29:55pondlifeddalton: ^
13:30:11peturNico_P: looks ok, is there a way to look at the complete file, not only the diff?
13:30:27Nico_Ppetur: yes, "blob"
13:30:40peturah, thnx
13:30:48 Join kubiix [0] (
13:32:07 Join slordor [0] (
13:33:15peturNico_P: I would change "num_properties = 8" into "num_properties += 3", for the rest: commit away ;)
13:36:57Nico_Pok :)
13:38:29 Quit spiorf (Read error: 110 (Connection timed out))
13:39:14 Join spiorf [0] (
13:39:15slordorhey does anyone know about creating a voice file for rockbox using ibm viavoice tts? it uses sapi 4 but the configuration script uses sapi 5 from what I've read.
13:40:27ddaltonwell then it won't work.
13:40:46 Quit ddalton ("I was using BOFHNet IRC version 1.2 by fmillion - get your copy today from !")
13:40:50slordorthere is one in the old format on the site made with ibm tts so I thought I'd ask if anyone knew how it was done
13:41:18B4gderI don't think the necessary scripts have been adjusted to the new format yet
13:41:44slordoroh ok
13:43:11slordorI guess I will wait to upgrade then as I'm not a festival fan *grin*
13:43:23B4gdernobody is really
13:44:13LinusNhowever, we *are* fans of freely distributable voices
13:44:16B4gderand since most windows users seem to "wait until it magically fixes itself", that's what's in progress... waiting
13:44:47slordorhmmm I have ttsynth in linux I wonder if I could modify the configuration script to use its shell command
13:45:02B4gderfreely distributable voices we like
13:45:52pondlifeI thought there was a patch for SAPI4 support around...
13:45:53slordorwell espeak is good but I like the ibm tts sound or my double talk lt
13:45:54 Quit jnakibbhc (Read error: 113 (No route to host))
13:47:24pondlifeslordor: might help.
13:47:37slordoroh ok thanks will take a look
13:49:02 Join Thundercloud [0] (
13:51:45slordoroh this looks very good thanks I will give it a try later today.
13:54:45 Join PeterH [0] (i=5ab841d0@gateway/web/cgi-irc/
13:56:03 Quit PeterH (Client Quit)
13:56:37 Join Caliban_ [0] (
13:57:54 Part slordor
13:58:28*Nico_P is sorry for the 3 successive commit with slightly inappropriate commit messages...
13:58:46*petur points to the corner
13:59:22***Saving seen data "./dancer.seen"
13:59:23aliaskWhy are the commit messages inappropriate?
13:59:52*petur wonders the same
14:00:11Nico_Pthe inappropraite one is actually the one for r14326... it deals mainly with MoB considerations
14:01:02Nico_Pwell actually it's not as bad as I thought... the only out-of-place comment is "This should make it easy to use a buffer handle to put the ID3 data in. "
14:01:36pondlifeWell it'll be appropriate once MoB is committed :)
14:01:45 Join rogelio [0] (n=rogelio@
14:02:19 Part rogelio ("Kopete 0.12.4 :")
14:03:15pondlifeShould struct mp3entry be renamed to struct metadata or something? Or am I confused?
14:03:27 Join GodEater [0] (n=bryan@rockbox/staff/GodEater)
14:03:36aliaskIt probably should be - it does contain metadata
14:03:40Nico_Ppondlife: I agree
14:03:46pondlifei.e. for other formats than MP3
14:04:22B4gderleftovers from ancient history
14:04:22Nico_Ppondlife: yes, it's a generic metadata container, but its name is a leftover from the ime when rockbox only played MP3
14:05:06pondlifeHmm, HWCODEC doesn't use get_metadata() ?
14:05:16amiconnpondlife: No, because it can't
14:05:29amiconnget_metadata() is in apps, the hwcodec engine is in firmware
14:05:41*pondlife slaps self
14:05:44Nico_Phmm, time for an oops commit
14:06:15GodEaterso will MoB make it in time for the 20th ?
14:06:15pondlifeHey, it's just a plugin...
14:06:43Nico_PGodEater: why the 20th ?
14:06:56GodEaterlinuxstb alledges this is the SoC deadline
14:07:01pondlifeGodEater: Yep, in one week, we'll have MoB with Album Art, a speech synth and full USB support :)
14:07:15pondlifeGonna be a good week.
14:07:33*Nico_P thought the timeline was 31st... checking
14:08:20Nico_P"August 20: Students upload code to; mentors begin final evaluations; students begin final program evaluations", "August 31: Final evaluation deadline; Google begins issuing student and mentoring organization payments"
14:08:59pondlifeOK, so that makes it the 20th, no?
14:09:08GodEatersounds like it to me
14:09:12*pondlife turns up the pressure, again.
14:09:39*GodEater joins him at the thumbscrews
14:09:48pondlifeHaha, Rockbox *is* fun!
14:09:55pondlifeRepeat after us !!
14:10:16*Nico_P feels like running away
14:10:49pondlifeNooo, don't..
14:10:54GodEaterNico_P: is that timeline unrealistic for MoB ?
14:11:37pondlifeDon't worry about Album Art or anything more, those can happily wait (of course).
14:11:39Nico_PGodEater: I hope to have some sort of integration in the rockbox code by then, but we won't be able to release on the 21st ;)
14:11:57 Join Zagor [0] (n=bjorn@rockbox/developer/Zagor)
14:12:03 Quit Caliban (Read error: 110 (Connection timed out))
14:12:07*GodEater is fine with that :)
14:12:09 Quit webguest61 ("CGI:IRC (Ping timeout)")
14:12:14GodEaterNico_P: who is your mentor, I forget ?
14:12:51GodEaterso he's doing your evaluation then ?
14:13:16GodEater"not entirely"
14:13:24pondlife"mostly harmless"
14:13:40pondlifeor..."broke the fundamental purpose of a DAP"
14:14:07GodEaterI thought that was to play Doom III ?
14:14:12 Join webguest88 [0] (i=d558f5ba@gateway/web/cgi-irc/
14:14:23*Nico_P hopes Llorean will be less severe :)
14:14:53pondlifeNico_P: Just be happy that your new code is going to be much better than what's in there at the moment...although I'm sure we'll have some minor bugs to work out.
14:14:53 Quit webguest88 (Client Quit)
14:15:09GodEaterare you dissing his code already ?
14:15:18pondlifeNot at all, bigging it up
14:15:44GodEateryou're a hard task master...
14:16:07pondlifeEspecially when I'm actually wearing my Rockbox T-shirt.
14:16:16*GodEater found his somewhat tight...
14:16:37pondlifeHey, you can grow into it..
14:16:53scorcheor, shrink, in your case =)
14:17:07pondlifescorche: I was trying to be diplomatic
14:17:12*GodEater doesn't want to shrink
14:17:31scorchepondlife: i know...i tried the same, but i couldnt make the pieces fit that way...
14:21:26Nico_Phmm I changed the spec of get_metadata... shouldn't that warrant a plugin API bump ?
14:21:39pondlifeYes, probably.
14:21:46pondlifeDifferent structure?
14:21:54Nico_Pthough I doubt plugins actually use it
14:22:07Nico_Pdifferent parameters
14:22:42bluebrotherthen it's an incompatible change that asks for a bump
14:22:52B4gderplugin API bumps are cheap anyway
14:23:56*petur did one only yesterday :)
14:24:06pondlifeTwo for the price of one
14:27:19 Nick Caliban_ is now known as Caliban (
14:33:03 Quit barrywardell (Remote closed the connection)
14:34:17 Join ie [0] (i=c27f0814@gateway/web/cgi-irc/
14:34:36ieRed builds! Some #if's may be missing I guess
14:37:20Nico_Pie: the one that is currently building should be green
14:37:38*Nico_P should have done the commits the other way around
14:38:50ieThat's the nice thing about RB: you just don't have a chance to make a mistake :-)
14:39:58B4gderour build system is quite cool indeed
14:40:58B4gderyellow instead
14:41:20ieB4gder: yes. But I meant that if you make a mistake other will immediately point you at it!
14:41:23 Quit ctaylorr (Read error: 110 (Connection timed out))
14:41:35B4gderyes, no chance to hide!
14:41:47Nico_Pdamn... I did to a test compile !
14:42:06ieBig Bagder's watching you! :-)
14:43:43scorchehe is creepy like that
14:44:14B4gderright, there are no "build servers", that's actually all me
14:44:54B4gderif you've been nice you get green cells in return
14:45:01scorcheour honorable dictator!
14:46:14 Quit aliask ("ChatZilla [Firefox]")
14:46:14B4gderNico_P: did you test compile this?
14:46:25B4gderit looks like mixed code/declarations to me
14:46:41B4gder... which I guess causes warnings
14:46:50Nico_PBagder: I did and got nothing here
14:47:23Nico_PB4gder: I almost asked you if it was OK but thought it was accepted
14:48:19Nico_PIIRC gcc only complains about this when using -pedantic
14:50:07scorcheJdGordon: please PM me (in the forums) with your address...i dont think putting "JdGordon - Aussie Land" will get far =)
14:51:12Nico_Pgreen at last !
14:51:45amiconnWhy not use the correct function for hwcodec instead of leaving out the feature?
14:54:08Nico_Pamiconn: what is the correct function for HWCODEC ?
14:54:49JdGordonscorche: sure it will
14:56:07amiconnget_mp3_metadata() directly
14:56:17 Quit Febs (Read error: 110 (Connection timed out))
14:56:30amiconnIf it's not in the api, it could be put there for hwcodec
14:57:07Nico_Pamiconn: it's not, but mp3info is
14:57:14amiconnThat's temporary anyway; hopefully hwcodec will start using get_metadata() along with the rest of the engine in apps/
14:57:28Nico_Pamiconn: is someone working on it ?
14:57:40amiconnErm, you?
14:57:48amiconn(on the enigine in apps/ I mean)
14:57:49 Join webguest24 [0] (i=c023110b@gateway/web/cgi-irc/
14:58:34amiconnYeah, mp3info() would work, it just opens the file itself. get_mp3_metadata() takes a fd
14:58:37Nico_PI hope you're not counting on me to unify HWCODEC and SWCODEC ?
14:59:18 Join barrywardell [0] (
15:00:38amiconnI hope that the reworked engine will make it easier
15:03:32JdGordonNico_P: hows it going?
15:04:37Nico_PJdGordon: reading a lot of playback.c
15:04:58Nico_Pdrowning into it even
15:05:04 Quit jhMikeS (Read error: 104 (Connection reset by peer))
15:05:06JdGordonhehe good luck
15:05:42 Join jhMikeS [0] (n=jethead7@rockbox/developer/jhMikeS)
15:06:27Nico_Pthanks :)
15:07:30 Join Crash91 [0] (i=510a6f0d@gateway/web/cgi-irc/
15:07:31 Join Febs [0] (
15:07:54Crash91any idea when
15:08:02Crash91USB support will be on Sansa?
15:08:50bluebrothersure: when it's done :)
15:08:54Crash91lol... i wish i could help but im too lazy to learn code
15:10:18 Quit Crash91 (Client Quit)
15:15:07Nico_Pmp3info: "Note, that this returns false for successful, true for error!" => how twisted is that ?
15:15:22Nico_P"result = !get_mp3_metadata"....
15:16:33 Quit Febs ("ChatZilla [Firefox]")
15:17:16JdGordonNico_P: alot of the/our code returns 0 on success
15:18:13Nico_PJdGordon: OK for 0 but false for success ? I know false == 0 but I still think it's weird, especially when get*_metadata behaves the opposite
15:18:58Nico_Panyway... the commit is arriving and hopefully I didn't oversee anything
15:20:00*Nico_P is on a commit spree today
15:20:32peturoh admit it, you're just trying to bump your ohloh stats :p
15:23:53bluebrotherit returns the error state? false means no error i.e. success ;-)
15:23:53 Quit webguest24 ("CGI:IRC (Ping timeout)")
15:24:10rasherWow, the frontpage doesn't deal well with utf8 in commit messages.
15:24:11bluebrotherQHttp does such stuff with signals containing a bool for error state
15:24:41Nico_Pbluebrother: that makes sense... I was just grumpy because I got caught by it
15:24:42 Join Febs [0] (
15:24:42bluebrotherthe front page is still latin1 :o
15:25:27bluebrotherNico_P: I admit that it's usually the other way round −− at least as far as I've seen so far
15:32:54AveLlorean: you around?
15:33:36 Quit JdGordon ("Konversation terminated!")
15:34:36 Join norbusan [0] (
15:34:40 Part norbusan
15:34:51 Join elinenbe_ [0] (n=elinenbe@
15:35:01 Join Gibbed [0] (
15:35:05 Quit Rick (Nick collision from services.)
15:35:19 Nick Gibbed is now known as Rick (
15:35:29 Join linuxstb_ [0] (i=d556da1b@rockbox/developer/linuxstb)
15:44:02amiconnNico_P: Your code has a potential file handle leak on swcodec
15:44:54 Join DerPapst [0] (
15:45:17amiconnAh, no, but it might try to close an already closed file hande
15:45:45DerPapstNico_P almost owned the frompage... almost
15:47:29 Quit Guile`` ("KVIrc 3.2.0 'Realia'")
15:48:29peturisn't fd supposed to be > 0 ?
15:50:23LinusNis is supposed to be non-negative
15:51:07 Quit datachild ("- nbs-irc 2.35 - -")
15:53:33 Join webguest24 [0] (i=c023110b@gateway/web/cgi-irc/
15:53:50Nico_Pamiconn: close will just return an error and do nothing if I try to close a negative fd
15:54:20Nico_PI assumed it wasn't worth annother (fd >= 0) check
15:54:28Nico_Psince it's done in close() anyway
15:54:59 Join Febs_ [0] (
15:57:06 Part LinusN
15:57:59 Quit Febs (Read error: 110 (Connection timed out))
15:58:11 Nick Febs_ is now known as Febs (
15:59:25***Saving seen data "./dancer.seen"
15:59:47 Quit elinenbe_ (" Want to be different? HydraIRC -> <-")
16:02:36 Quit yjoe ()
16:05:01 Quit Zagor ("Client exiting")
16:08:03 Join elinenbe_ [0] (n=elinenbe@
16:27:16 Quit BigBambi ("Leaving")
16:28:44 Quit idnar (Nick collision from services.)
16:28:46 Join idnar_ [0] (i=mithrand@unaffiliated/idnar)
16:29:12 Nick idnar_ is now known as idnar (i=mithrand@unaffiliated/idnar)
16:31:25 Quit Febs ("ChatZilla [Firefox]")
16:33:08Nico_Pdamn... get_metadata outputs a LOT of data...
16:33:42 Join fnakihlbh [0] (i=0@
16:34:01Nico_Pof debug info I mean
16:36:48 Join toffe82 [0] (
16:42:50 Quit webguest24 ("CGI:IRC")
16:46:55 Join IA4004 [0] (n=chatzill@
16:57:33 Quit Llorean ("Leaving.")
17:13:08Nico_Pundefined instruction... what can that be for ?
17:13:52 Join saratoga [0] (i=98039a34@gateway/web/cgi-irc/
17:14:57 Nick lost|X40 is now known as lostnihilist (
17:17:05Nico_PjhMikeS: my undefined instruction seems to come from the threading code (if I read the .map files correctly)
17:20:21 Quit IA4004 (Read error: 110 (Connection timed out))
17:29:28 Join rotator [0] (n=e@rockbox/developer/rotator)
17:33:18GodEaterwhat a fantastic thing gravatar isn't
17:33:59GodEater"The image cropper doesn't seem to work in your browser. We'd love to hear about what OS, Browser and version you're using. In the meantime, you might want to try Firefox."
17:34:04GodEaterIT IS FIREFOX!!!!
17:34:34Nico_PGodEater: do you have javascript activated ?
17:35:07 Quit w1ll14m (Remote closed the connection)
17:35:09GodEaterand even if I didn't they should be able to tell I'm using firefox
17:36:56bluebrothergravatar is somewhat strange −− I uploaded a picture a while ago but I can't see it except when viewing my profile at their website ...
17:37:54*petur had no problems whatsoever
17:38:41GodEaterbluebrother: same here apparently
17:38:53GodEatermy ohloh account isn't displaying the gravatar I selected at all
17:39:56peturmaybe they block pictures that aren't for the general public? :p
17:40:24bluebrotherobviously my face isn't suited for the public. Or the public for my face :P
17:44:53GodEaterbluebrother: did you "rate" your gravatar too ?
17:45:00GodEaterI just noticed mine is set to "X"
17:45:13GodEaterand apparently ohloh wants PG or lower...
17:46:10 Quit Thundercloud (Read error: 110 (Connection timed out))
17:47:38bluebrotherI rated it G −− maybe I should try PG?
17:47:56 Quit petur ("*plop*")
17:48:21GodEaterhmm - it might be a size issue too
17:48:40bluebrotherI thought they automatically resize the icon upon upload?
17:49:15GodEaterapparently only if the image is over 80x80
17:49:25GodEaterif your image is smaller than that, it doesn't do it
17:50:11bluebrotherinteresting −− after removing the size parameter from the image query it shows
17:50:56bluebrotherbut the picture _is_ 80x80 px. hmpf.
17:51:07GodEaterstupid gravatar
17:51:22GodEatermine still isn't loading at all
17:51:34bluebrotherit does −− I just noticed it.
17:52:07*bluebrother tries clearing browser cache
17:52:12bluebrotheraaah :)
17:53:44*GodEater is still getting the blue G
17:54:28bluebrotherhmm, gone again.
17:54:33*bluebrother scratches head
17:54:37 Quit barrywardell (Remote closed the connection)
17:59:28***Saving seen data "./dancer.seen"
18:00:54 Quit ie ("CGI:IRC")
18:02:13Nico_Pcould anyone tell me what kind of things lead to an undefined instr. ? the change that made it appear is here:;h=5987a139e16aef731b58372e83947b70e0ea10f7
18:07:02 Join Domonoky [0] (
18:16:28 Quit ptw419 ()
18:19:32 Quit kubiix (Read error: 110 (Connection timed out))
18:20:34 Join ptw419 [0] (
18:23:49 Join merbanan [0] (n=banan@
18:24:45 Join Arathis_ [0] (
18:26:48 Quit Arathis (Nick collision from services.)
18:26:53 Nick Arathis_ is now known as Arathis (
18:27:27 Join webguest47 [0] (i=54b3e12e@gateway/web/cgi-irc/
18:31:43 Join webguest49 [0] (i=50d81efc@gateway/web/cgi-irc/
18:32:41webguest47ok hi, i'm here again to ask for a bug before i file a bug report. if you have a folder with files in it named: 1,10,11,100,101,102 they will get sorted this way: 1,10,100,101,102,11. so you see when there is no leading zero the filenamesorting fails. or am i intended to change my filenames? but the same goes for tag info and so on. suggestions?
18:33:25 Join webguest48 [0] (i=c27f0811@gateway/web/cgi-irc/
18:33:38webguest48webguest47: file names are strings so this sorting is correct.
18:33:41Domonokywebguest47: thats normal.. no bug
18:34:09 Join elinenbe__ [0] (
18:34:16 Quit elinenbe (Read error: 104 (Connection reset by peer))
18:34:16 Nick elinenbe__ is now known as elinenbe (
18:34:20webguest48The year tag might be another story, but even that tag allows non-number values IIRC
18:34:47webguest47what does strings mean?
18:35:02webguest48Since all tags are strings. It's the programs that interpret them that make difference.
18:35:10Domonokyit means its treated as text not as numbers..
18:36:29webguest48Nico_P: compiler bug?
18:36:39 Join webguest33 [0] (i=50d81efc@gateway/web/cgi-irc/
18:36:41 Quit webguest49 ("CGI:IRC (EOF)")
18:37:07webguest47i don't want to nitpick but windows seems to be more "intelligent" in this way but maybe windows is also the source for this missnaming
18:37:19Nico_Pwebguest48: maybe, but I doubt it
18:37:28bluebrotherwell, there is a difference between "sorting" and the so-called "natural sorting"
18:37:56bluebrotherRockbox doesn't implement natural sorting. From a programmers view, sorting isn't the natural way.
18:38:01 Quit ptw419 ()
18:38:14bluebrotheras you start at the first character and sort, then go to the next etc.
18:38:46webguest47and do you think rockbox should also sort natural? so this is maybe more of a feature request
18:38:48webguest48Nico_P: can you try to exclude lines of c code until you don't have the ill-instr anymore? What can one else do in this situation?
18:39:00bluebrother(and btw, windows supports natural sorting only since wxp)
18:39:32webguest48bluebrother: what is "natural"?
18:39:43bluebrothersorting: 1, 10, 2, ...
18:39:50bluebrothernatural sorting: 1, 2, 10, ...
18:40:25webguest48bluebrother: how are the "number names" compare to non-numbers then?
18:40:32 Join BigBambi [0] (n=Alex@
18:40:34Domonokyand "natural" sorting isnt trivial, especially on a embedded device..
18:41:02bluebrotherwell, a digit is a char. So (thinking of the internal representation of the chars) 0 < 1.
18:41:30webguest48Domonoky: well, it's simple if you (internally) add sufficient number of zeroes to number names. E.g. so that the total length is 10.
18:41:40bluebrotheri.e. usual sorting compares the first digit. We get the group (1, 10) and (2). As we only compare the first digit (1, 10) comes first.
18:42:10bluebrotherwebguest48: that's not true −− you forgot about special characters like german umlauts
18:42:18bluebrotheror characters with accents etc.
18:42:35 Join Thundercloud [0] (
18:42:48webguest48bluebrother: padding should only be done for number names, i.e. those containing only digits
18:42:50bluebrotherin the next step we sort (1, 10). As 1 is shorter it comes first −− no character is "less" than some character. We get (1, 10).
18:42:53Domonokyand different languages.. different char systems.. think of japanese or something like that :-)
18:43:14bluebrotherwell, if you do natural sorting _properly_ you need to take extended characters into account
18:43:51bluebrotherand the locale −− depending on the country extended characters get sorted differently
18:44:26bluebrotherlike: is it a - ä - b or a - b ... - z - ä? That depends on the country.
18:44:32webguest48In what points does the natural sorting differ from the usual sorting? Only that number names are sorted as numbers and not lexically? Since the other cases require correct collation and that's really beyond the RB scope
18:45:04webguest48But I think that sorting just number names as numbers might be really useful
18:45:08bluebrotherwell, a simple natural sorting that only takes number into account should be possible.
18:45:15bluebrotheriirc there was a patch in the tracker once.
18:45:17 Part toffe82
18:45:58*amiconn *hates* this natural sorting thing
18:45:58 Quit DerPapst ("work->home yay! :D")
18:46:09webguest48bluebrother: that's what I think too. But then we'd have a setting for this. And then the change will be definitely rejected! :-)
18:46:22amiconnFiles in a long list are often not where I expect them
18:47:31bluebrotherhmm, that task only affects extended characters, not numbers.
18:47:38webguest48But I wouldn't take accents etc. into account. Only numbers. OTOH, I dont' have a single file with a number name!
18:47:46*Domonoky doesnt care which sorting method is used, as long it is consistent.. and not different in differen places.. thats annyoing..
18:47:57webguest48So, for me, this change would be essentially useless
18:48:06amiconnI'd like to see proper unicode collation, but that's rather complex too
18:48:15 Quit pondlife ("disconnected has pondlife")
18:48:36bluebrotherfound it:
18:48:44*webguest48 agress with Domonoky and leaves
18:48:47 Part webguest48
18:49:04amiconn6359 is half-baked at beast
18:49:22bluebrotherI agree.
18:49:24webguest33Managed to build e200tool's arm binary blob from arm_code.c now :)
18:49:35webguest33arm-elf-gcc -Wall -Os -fno-unit-at-a-time -mcpu=arm7tdmi -c arm_code.c -o arm_code.o
18:49:37webguest33arm-elf-ld -Ttext 0x40004000 -N arm_code.o -o arm_code.bin
18:49:40webguest33arm-elf-objcopy -Obinary arm_code.bin arm_code.raw
18:49:42webguest33python < arm_code.raw > e200_code.c
18:49:50bluebrotherwebguest33: please use a pastebin
18:49:53Nico_Pamiconn: any thoughts on where I could get an undefinded isntr. from ?
18:50:01 Join toffe82 [0] (
18:50:29webguest47ok so i think the user has to make sure the filenames are named the correct way for his own interrest especially on old computers and maybe old software too. i don't know if a simplified solution only with numbers considered is should i file a feature request if there is none?
18:50:37 Join BigMac [0] (
18:50:37amiconn(1) it only works for latin1 (2) the sorting is (nearly) correct for german, but not for other latin languages
18:50:50 Join Llorean [0] (
18:51:32amiconnNico_P: Execution jumps somewhere it shouldn't. Could be anything - code overwriting other code, dangling pointers, whatever
18:51:36bluebrotherwell, adding number-based sorting would be a start, wouldn't it?
18:51:41DomonokyNico_P: i recently got such errors, when i was overflowing the stack ..:-)
18:52:07Nico_Phmm I did get a stkov... tried increasing stack size and then got the undef instr
18:52:19Nico_PI'll try an even bigger stack size
18:52:50Domonokythe stack is small.. put a 10k array there .. and *bum* :-)
18:53:19Nico_PDomonoky: I'm not asking for quite as much as 10k :p
18:53:27amiconnNico_P: If you need more than a few KB (<10) of stack, you're doing something wrong on embedded systems
18:53:28linuxstb_bluebrother: The patch you linked to was rejected without explation by B4gder...
18:54:07bluebrotheryes −− I wonder why it got rejected.
18:55:01Nico_Pbluebrother, linuxstb_: IIRC it was rejected because the idea of natural sorting was
18:55:16linuxstb_Nico_P: Obviously... But why reject natural sorting?
18:55:34amiconnBecause it's confusing...
18:56:09linuxstb_Personally I agree, but if people want it?
18:56:10Nico_Pamiconn: my stacks are 4*DEFAULT_STACK_SIZE/sizeof(long)
18:56:20Nico_Pamiconn: how big is that ?
18:56:38Domonokylinuxstb_: if "people" want it they should code it :-)
18:56:42Lloreanlinuxstb_: Are you talking the one for filetree?
18:56:48linuxstb_Domonoky: They did.
18:57:00amiconnNico_P: 4KB
18:57:09linuxstb_Llorean: I think so.
18:57:16Nico_Pamiconn: so that's quite reasonable ?
18:57:29amiconnOr rather, because it's confusing *and* expensive to implement
18:57:42linuxstb_Llorean, Domonoky:
18:57:53Lloreanlinuxstb_: I think there was a discussion in here in which it was just decided, it's "unnecessary" since it didn't really add much at all (users can but Blah, The as the filename)
18:58:27webguest33no austriancoder today?
18:58:33LloreanOh, not "The"
18:58:49Lloreanlinuxstb_: Sorry, wrong patch in my mind.
18:59:01 Join GodEater_ [0] (n=bryan@rockbox/staff/GodEater)
18:59:04Lloreanlinuxstb_: In that one's case, I think the idea was that "Numbering the files properly is very easy, there's a lot of free utilities to do it"
18:59:23bluebrotherdoes that also work with tagcache?
18:59:35bluebrotheri.e. if you have a <track number> - <title> view?
19:00:01Lloreanbluebrother: I'm pretty sure track number - title sorted by track number uses the track numbers as integer values (if not, it should)
19:00:14LloreanI'm all for the database actually interpreting them as ints when coming from the "Track" tag.
19:01:04amiconnHmm, btw, was that ugly guessing removed?
19:01:09*amiconn needs to check
19:01:13pixelmaI think it works correctly based on tracknumber
19:01:37amiconnI think if it's for leading numbers _only_, it might be not too complex and a somewhat useful option
19:02:02amiconnBut the "natural" sorting in windows seems to try and be more clever than the user
19:02:20bluebrotherisn't that a known issue with windows? ;-)
19:03:50LloreanDoesn't the database give you considerable sorting power with Tagnavi?
19:04:08LloreanI think people who are used to a filetree view (other players, etc) are more likely to have an appropriate number of digits anyway
19:05:40Nico_Pyay! no more undef. instr. with 4KB stack instead of 3
19:07:01bluebrotherok, now for cleaning up my changes ...
19:10:03 Quit Arathis ("Verlassend")
19:17:43 Join dan_a [0] (n=dan_a@
19:17:52 Quit linuxstb_ ("CGI:IRC")
19:18:03 Join jgarvey [0] (
19:22:39 Join desowin [0] (n=desowin@
19:25:28 Quit miepchen^schlaf ("Verlassend")
19:25:57 Join Nibbier [0] (
19:31:59Nico_Pdan_a: still using git for rockbox dev ?
19:32:00 Quit webguest47 ("CGI:IRC (EOF)")
19:33:57 Join robin0800 [0] (
19:35:53dan_aNico_P: The little development I'm doing at the moment, yes. And I've switched to your central repo
19:36:01Nico_Pcool :)
19:36:09Nico_PI can give you push access if you want
19:36:44dan_aThat would be great - I'll try and script svn updates to go into the repo automatically
19:36:47 Join inversions [0] (n=none@
19:37:15Nico_Pdan_a: just give me your account name on and I'll add you ;)
19:37:50Nico_Palso, I'm curious as to how you manage your local branches... do you rebase them against svn or do you merge master into them ?
19:38:35dan_aNico_P: I'll set up an account and msg / memoserv you later. I had been rebasing, but if they're going to be shared then I understand they need to be merged with master.
19:39:07*dan_a goes out
19:39:14 Quit dan_a ("using sirc version 2.211+KSIRC/1.3.12")
19:39:17Nico_Pactually I think you can use the -f switch to force the push for rebased branches
19:39:25Nico_Poh, missed him
19:41:43GodEater_what's the difference between rebasing and mergeing ?
19:42:14Nico_Prebasing rewrites your commits on top of the ones you just fetched
19:42:49 Quit elinenbe_ (Read error: 110 (Connection timed out))
19:42:57Nico_Pwhereas merging merges the newly fetched commits to your exsisting ones
19:44:27*amiconn wonders why there is no daily manual for the 1st/2nd gen yet
19:44:40Nico_Pin the case of git-svn, rebase would take your local commits, unapply them, update the branch with the latest svn changes (that's a "fast-forward"), and finally try to commit your local commits on top of the new history
19:44:49GodEater_ah gotcha
19:45:05GodEater_is it possible to merge with git-svn or not ?
19:45:43Nico_Pyes and no... you can't type "git svn merge", but you never need to anyway
19:46:23GodEater_so how is it acheived then ?
19:46:49 Join miepchen^schlaf [0] (n=hihi@
19:47:27Nico_Pyou have a master branch wich you rebase regularly (git svn rebase) and do nothing else to, and another branch
19:47:50GodEater_yep - this is how I work locally
19:47:59Nico_Pwhen you get new commits from git-svn, either you can merge master into the other branches, or rebase the other branches with git svn rebase
19:48:31Nico_Pboth work, but I had a bit of trouble with the merging way (I was doing that)
19:48:40GodEater_what was the trouble ?
19:48:56Nico_Pgit svn dcommit complained about the merging commits
19:49:36Nico_Pso I think it's better to rebase, but I was merging because git push was complaining about not having the same commits as the remote
19:49:55Nico_Pbut forcing the push should solve that
19:50:47bluebrotherdoes git svn dcommit commit the commits you made one-by-one or as a big change?
19:51:03GodEater_big change I'd imagine
19:51:53webguest33petur: I'm running e200tool now.. do you think austriancoder would like a memory dump of the USB registers (0xc5000000 ..) ?
19:51:55GodEater_so, IIUC, forcing a push would bascially update the *code* at the far end, but would miss the history you had in your local repo ?
19:52:31Nico_Pbluebrother, GodEater_: one by one
19:52:48Nico_Pthat's why I did three commits in the same minute earlier today
19:53:16GodEater_your ohloh score is going to go through the roof ;)
19:53:21Nico_PGodEater_: no, you'd have the same history as locally
19:53:37bluebrotherhmm. So if I do small commits while working how can I push it as a big commit to svn? Create a patch and add that to master?
19:53:38Nico_PGodEater_: hehe :)
19:53:43GodEater_in that case I don't understand the complaint
19:54:59Nico_Pbluebrother: that's a bit tricky, you'd have to do a bit of histry rewriting in your repo, or yes maybe create a series of patches, but that would then invalidate your work branch once you commit the patches to the svn
19:55:06GodEater_the warning in the man page says "can cause the remote repo to lose commits"
19:55:34bluebrotherhmm. Why should this invalidate my working branch? It's the same result in the end ...
19:56:00Nico_Pbluebrother: yes, but then you would git svn rebase it and I'm not sure what would happen then
19:56:37nerochiarosaratoga: hi, if you have a minute, i'd like to ask you a couple of questions about the wma code. i've got it working, it decodes fine, but i still have some problematic files, and had to change some things
19:56:38Nico_PGodEater_: the problem is that when you rebase, the local (non-svn) commits change. the remote repo has the old versions of these and you don't have them anymore. this makes git push think you don't have the same history at all
19:56:38GodEater_yeah, it would have the big commit from your patch
19:56:42GodEater_and also your working changes
19:57:25Nico_PGodEater_: forcing makes the remote repo lose the old version, which is potentially dangerous
19:57:57GodEater_I think I understand =/
19:58:31GodEater_we're talking a three way process here then - you rebase you local git tree from svn, and then push from that to the remote git repo yes ?
19:59:02Nico_Pyes, but why is that thee way ?
19:59:16GodEater_svn -> local -> ?
19:59:23Nico_Pah, ok
19:59:32***Saving seen data "./dancer.seen"
19:59:35GodEater_three repos I mean
19:59:43GodEater_there's only two data transmissions
19:59:55Nico_PGodEater_: I think the best possible explanations about rebasing are in the git-rebase manpage:
20:00:07GodEater_hehe - I'll read that then
20:00:36Nico_Pbluebrother: with rebasing you could rewrite your local branch's history to have better commits for svn, and then dcommit those
20:00:36*bluebrother leaves for a few hours
20:00:55GodEater_you scared him away!
20:01:01Nico_Phehe :)
20:01:20*GodEater_ goes for some dinner too
20:08:07*Nico_P top
20:08:07 Join Lear [0] (i=chatzill@rockbox/developer/lear)
20:26:50LearHm, when changing languages, the menu title isn't updated as it should...
20:27:19n1sLear: I think that's an olds bug
20:27:33LearCould be, just noticed it though.
20:28:33saratoganerochiaro: I'm around now
20:31:07nerochiarosaratoga: well, the first thing i had to fix that was giving me a lot of trouble yesterday was that that change you recently made to "return the full 30 bit precision" of the decoded data instead of making them 16 bit like it is in your freestanding version
20:31:21nerochiarosaratoga: once i reverted that change, most of my problems have been solved
20:32:29nerochiarosaratoga: don't know if it's my output system fault or what, but the other way didn't work
20:33:55saratogathe other way works by leaving every sample in full q14.17 format and letting the rockbox DSP handle the rest
20:34:04saratogaif your DSP code isn't handling that, its not going to work
20:34:36nerochiarosaratoga: i think it doesn't, but it took me some time to figure out. my fault
20:35:54nerochiarosaratoga: now, after that i've still had lots of files crashing the decoder during the initialization of the decoding for the file. and it turned out it was the size of the vlcbufs being too small and overflowing
20:36:14nerochiarosaratoga: it tried to write past the end and it caused a segfault
20:38:10saratoganerochiaro: the size of the vlc bufs is determined by the defines at the top of wmadeci
20:38:19saratogatheir size should not be dependent on the input data
20:38:57 Join PaulJam [0] (n=Paul@
20:39:04saratogaassuming you have not changed those defines from the default 8 or 9, and you're still overflowing that buffer, you almost certainly have a corrupt or improperly parsed asf packet
20:39:58 Join IA4004 [0] (n=chatzill@
20:42:54 Join datachild [0] (
20:44:18n1sLear: I assume someone forgot to make the title pointer of the appropriate gui_list struct update... JdGordon?
20:45:19nerochiarosaratoga: you say this because the sizes of these buffers are tuned in a way that can contain the maximum allowed ASF packet size ?
20:45:51nerochiarosaratoga: in other words, i'm not understanding why you say these sizes should be enough
20:47:36 Join kubiix [0] (
20:48:14 Quit robin0800 (" HydraIRC -> <- Nine out of ten l33t h4x0rz prefer it")
20:50:49saratoganerochiaro: the tables are not dynamic
20:51:04saratogathey're preallocated by the init function without any knowledge of the stream they're going to be used on
20:51:51saratogaif your stream is causing the huffman coding step to fail, increasing hte buffer size should not fix that
20:52:41nerochiarowell, i'm by far no expert of coding, so i can't comment on that, but here it fixed it
20:52:59nerochiaroaudio coding, that is
20:53:42saratogawhat did you change to fix it?
20:54:59nerochiarovery bovinely i doubled the buffer sizes. probably overkill. i also pulled in other changes from the freestanding decoder that didn't fix anything. i'll try to revert them all and just increase the buffer size to see if it still fixes the issue
20:55:31saratogawhich files did this make a difference on?
20:55:41saratogaand did they decode with rockbox?
20:57:10Learn1s: I just think it needs a new call to str(), so maybe it's not worth fixing...
20:57:15nerochiarosaratoga: i don't know if they decode with rockbox, i don't have any hardware to run it on. but they did not crash your freestanding decoder, but the output wave was bleeping noise
20:57:43saratogawere they low bitrate files?
20:57:51linuxstbnerochiaro: You can build a Rockbox UI simulator to run on your PC. All you need is gcc and SDL (and the Rockbox source)
20:58:08PaulJamhi, little question: in r14180 there was a change that prevented the wps contextmenu to be drawn when exiting for example the playlist menu by pressing LEFT. Iwas wondering if the same could be done when leaving this menu by pressing play.
20:58:09n1sLear: IMO it's worth fixing, but I have no idea where to do that so if you want to... or I'll nag JdGordon :-)
20:58:12nerochiarolinuxstb: oh, nice. didn't know. is that in the manual ?
20:58:51nerochiarolinuxstb: i'll give it a shot later then
20:59:09nerochiarosaratoga: let me check the bitrate
20:59:49PaulJami have got to go, but i'll read the logs.
20:59:59 Quit PaulJam (".")
21:01:38Learn1s: Hmm, just getting language_browse make the proper gui_synclist_set_title() call could do the trick, I think. Not sure how though.
21:03:34 Join ompaul [0] (n=ompaul@freenode/staff/gnewsense.ompaul)
21:03:38nerochiarosaratoga: i think they are all vbr
21:04:13n1sLear: I'll try it
21:04:47nerochiarosaratoga: none of my players report bitrate and they have varying amount of frames in each packet
21:09:29Learn1s: He, on second thought, it should be quite easy. :)
21:10:01LearIf it wasn't for the lists argument. :/
21:10:08n1sLear: please explain, I'm quite lost in this maze of macros...
21:10:21 Join bluey- [0] (
21:11:42LearTwo of the arguments are known, so they are easy. Must know the list argument to use though...
21:12:31n1sah, you mean for calling gui_synclist_set_title() ? I was looking for that too...
21:13:14 Join robin0800 [0] (
21:18:19LearYes, but the language_browse knows nothing about the list. A return value to request a more complete refresh of the menu is probably the way to go.
21:19:04LearA bigger issue IMHO is that certain menu selections are not translated, containing stuff like "off" and "min".
21:19:33n1sLear: could you give an example?
21:20:16LearFile split settings (recording), backlight fade...
21:22:04 Join nerochia1o [0] (
21:23:00nerochia1osaratoga: has some network issues. i think i lost your last messages, if any
21:23:01 Quit nerochiaro (Read error: 104 (Connection reset by peer))
21:23:10 Nick nerochia1o is now known as nerochiaro (
21:24:42 Quit Llorean (Read error: 104 (Connection reset by peer))
21:24:49 Join Llorean [0] (
21:25:34nerochiarosaratoga: by the way, it looks like you were right. i reverted all recent changes, and now it still crashes, even if i increase the buffer size.
21:30:12Bagderamiconn: I didn't know the 1g2g manual is buildable
21:30:29pixelmait is
21:31:10Bagderok, manual build added for tomorrow
21:31:15 Join leftright [0] (i=d9e1fc0a@gateway/web/cgi-irc/
21:31:35 Quit Nico_P (Remote closed the connection)
21:31:58saratoganerochiaro: everytime i've had overflow issues in the vlc code, its been due to bad data passed into the codec
21:32:10leftrightI'm trying to assist a guy who seems to have bricked his h140 by trying to reflash his working rockbox'd player with irivers firmware, is there any way he can recover his player ?
21:32:14saratogain fact, if you pass it anything but a pristeen wma stream, you should expect vlc to blow up
21:32:40saratogathough if you're uncomfortable with our hackish system, its quite easy to switch back to the ffmpeg malloc version
21:32:44Bagderleftright: by bribing linusn
21:32:45amiconnsaratoga: Oh? I hope proper error handling will get added then...
21:33:26saratogaamiconn: most of the time its able to quit
21:33:38amiconnImho it's no problem if the codec skips, stops decoding the stream, or glitches if there are stream errors, but imo it must not crash on them
21:33:38leftrightBagder, thanks its a stranger on the forum, if linus is inclined to assist, I would like him to deal with said user
21:33:41 Quit _pill ("changing servers")
21:33:58nerochiarosaratoga: no, i don't mind the static alloc, but i want to be sure what kind of "bad data" i should not allow in. so far i'm just passing it the payload straight from the ASF packet
21:34:05BagderI figure he has his hands full already, which is why I mentioned bribing...
21:34:20leftrightbut basically he's bricked his player, am I correct in thinking that
21:34:52 Join pill [0] (
21:35:14BagderI haven't read the description, but it certainly can happen
21:35:40 Quit GodEater (Read error: 110 (Connection timed out))
21:36:43saratoganerochiaro: did you get the codec working with libasf (the real one not ours)?
21:36:55saratogai'd be interested in trying that out
21:37:03saratogawould make debugging our parser easier
21:37:47Bagderleftright: I'm not really able to tell for sure based on that, it could also be a problem with the OF, the disk or its database etc (afaik)
21:37:48nerochiaroyes, i did (althogh it still assumes 1 packet = 1 payload, but i'll fix that soon
21:38:48saratoganerochiaro: when you've got it reasonibly stable, could you send me a link?
21:39:22saratogai've already got a few files i suspect are gumming up the parser, but its hard to say for sure since it could also be the decoder
21:40:03nerochiarosaratoga: yes, no problem. but i have the decoder separated out in a library, so you may need to do some minor tweaking.
21:40:37saratogai just need to be able to get a wav file out of it
21:40:43saratogahopefulyl you've got a makefile setup?
21:40:44 Quit webguest33 ("CGI:IRC (EOF)")
21:42:52 Join jgarvey_ [0] (
21:43:22 Quit jgarvey (Read error: 110 (Connection timed out))
21:43:27nerochiarosaratoga: yes, i've got the wma library setup with a makefile to build in the neuros environment. the minor tweaking i mentioned was referred to have that build stand alone. the decoder application itself is just a single source file that uses the libwma and libasf and outputs a wav
21:44:44n1sLear: I added a bug to the tracker about the non-translated selections, as it is pretty bad IMO.
21:45:28 Join Entasis_ [0] (
21:46:12 Quit jhMikeS (Nick collision from services.)
21:46:18 Join jhMikeS [0] (n=jethead7@rockbox/developer/jhMikeS)
21:47:37 Quit DMJC (Read error: 110 (Connection timed out))
21:48:13 Join DMJC [0] (
21:49:45 Quit bluey- ("Leaving")
21:53:47 Quit nerochiaro (Read error: 104 (Connection reset by peer))
21:54:08 Join nerochiaro [0] (
21:57:23 Join rotator_ [0] (
21:59:35***Saving seen data "./dancer.seen"
22:01:12 Quit Entasis (Read error: 110 (Connection timed out))
22:02:07 Quit Nibbier (Read error: 110 (Connection timed out))
22:02:24 Part leftright
22:02:43 Join Nibbier [0] (
22:04:54 Quit thegeek ()
22:06:24 Join thegeek [0] (
22:14:25 Quit rotator (Read error: 110 (Connection timed out))
22:21:08 Nick rotator_ is now known as rotator (
22:23:35 Join webguest78 [0] (i=c730180b@gateway/web/cgi-irc/
22:23:42 Quit chrisjs169 (Remote closed the connection)
22:24:00 Quit Siku ()
22:25:01 Join chrisjs169 [0] (
22:25:14amiconnwhat the...
22:25:31 Quit ompaul ("Leaving")
22:26:37Lloreanamiconn: ?
22:27:55pixelmabuild table...
22:28:19Bagder92 fresh points!
22:28:20pixelma(I guess)
22:28:49*webguest78 's c250 now says "Rockbox boot loader" :)
22:29:04 Part webguest78
22:29:17Bagdernow that is cool
22:29:22 Join low_light [0] (i=c730180b@gateway/web/cgi-irc/
22:29:23LearGuess that "sim enabler" patch did more than it was supposed to...
22:29:34Bagderlow_light: cool!
22:30:16amiconnLear: power.c is target-only code. In fact that should be split and moved into the target tree
22:30:17pixelmaworking on 2 ports at the same time?
22:30:32Bagderlow_light: does the c200 have the two partitions like the e200?
22:30:34low_lightit hopefully won't be too hard to figure out the buttons now
22:30:43amiconnI'll try building a sim without it, and if that works, will do that...
22:31:06low_lightBagder: yes
22:31:12*amiconn wonders whether there is someone around with an fm recorder or recorder v2 to test
22:31:25Bagdershould make sansapatcher easy to adapt to the c200
22:31:34preglowc2x0 is pp too?
22:31:56preglowextremely cool
22:32:06amiconnNah, not cool (yet)
22:32:09preglowman, will we get a ton of cool targets the day the pp is figured out better
22:32:19amiconnIn fact rather hot, because of power consumption...
22:32:35preglowamiconn: which is why i have high hopes for you :D
22:32:58amiconnOn H10 this can be taken literal. It gets rather warm while running rockbox
22:33:25Bagdergood for winter time!
22:34:12*amiconn wonders whether he should do the quick fix for power.c, and return to sysevent messaging afterwards
22:34:55LearHmm, build servers use -j then?
22:36:25pixelmaLear: did you reorder the swedish lang by hand or did you use some helper script?
22:36:30 Join ssissi [0] (i=4a0c255b@gateway/web/cgi-irc/
22:36:58 Quit ssissi (Client Quit)
22:37:17Learpixelma: by hand, but parts had the correct order already, so I could sometimes more pretty large blocks.
22:37:41Bagderperhaps I should add an option for genlang to (re-)sort languages
22:38:49pixelmaI see. Yes, I know about the blocks - started to look into german lang myself too and wondered if I should reorder it too (IMO that's better too)
22:39:01LearYes, that might even be easy to do - if you know perl... :)
22:39:21pixelmawould be very helpful
22:39:27Bagderit already has most of the code to do it
22:39:29preglowis there any reason in having fwpatcher in svn anymore?
22:40:15LearI liked userfriendly's take on perl - the legalese of programming.
22:40:41preglowthat would be python, meticulous and boring
22:40:43 Quit Llorean ("Leaving.")
22:40:48Domonokypreglow: not really rbutil can do the fwpatcher job now job
22:40:55 Join Robin_0800 [0] (
22:40:55preglowDomonoky: yeah, that's what i thought too
22:41:03 Join stripwax [0] (
22:41:07preglowafter all, fwpatcher is windows and iriver only
22:41:18 Quit desowin ("use linux")
22:41:35LearNa, python is readable...
22:41:36 Quit IA4004 (Read error: 110 (Connection timed out))
22:41:43preglowLear: legalese is too :P
22:42:10*Bagder sits down with his popcorns to watch
22:42:51 Quit nerochiaro (Read error: 110 (Connection timed out))
22:43:05LearNot reading Groklaw are you? You can sometimes find fine examples of legalese there.
22:43:08 Join Accidental [0] (i=4a77eac8@gateway/web/cgi-irc/
22:43:11preglowi seldom let language wars span more than three lines these days
22:43:15preglownow, if you mention beer instead...
22:43:24 Join amiconn_ [0] (n=jens@rockbox/developer/amiconn)
22:43:40 Join foxxet [0] (i=4a77eac8@gateway/web/cgi-irc/
22:43:40 Quit amiconn (Nick collision from services.)
22:43:41 Nick amiconn_ is now known as amiconn (n=jens@rockbox/developer/amiconn)
22:43:53foxxethey guys quick question
22:44:02foxxetis it a bad idea to charge via USB and play music at the same time
22:44:15foxxetcuz its draining and charging the batter simultaniously?
22:44:16preglowplayer type?
22:44:41preglowwouldn't know, then
22:45:05foxxetplayer type really doesnt matter, does it?
22:45:15foxxeteither way your draining and charging your battery concurrently
22:45:23foxxetsounds like a bad idea, am I right?
22:46:00BagderI don't see why that is a bad idea
22:46:09Bagderdraining and charging is what batteries do
22:46:12 Join Terinjokes [0] (n=spader@wikinews/Terinjokes)
22:46:27preglowfoxxet: it does matter, for example, charging is so slow on some ipods that you shouldn't really use rockbox at the same time
22:46:33 Quit davina ("xchat on Ubuntu 7.04")
22:46:53preglowi wouldn't say it sounds like a bad idea on targets that handle charging better
22:46:53foxxetahh I see
22:47:00Terinjokesupdated Rockbox for the first time in months (starting from a new profile on the iPod, so I don't get screwed up old settings and stuff), and organising the Plugins was a nice update :P
22:47:07preglowbut i just don't know if the gigabeat is such a target
22:47:08foxxetok preglow, thanks
22:47:19foxxetshe charges like a demon, she should be o
22:47:29preglowthen i'd say go ahead
22:47:37preglowi do it all the time on my h120 with no problems ever
22:47:46foxxetthats what I needed to hear
22:47:54foxxetthanks bro, and thanks to all of you for rockbox
22:47:54Terinjokesthe only suggestion would be to run 'disktidy' in the background before auto-updating the database (a must for OSX users...)
22:48:00foxxetyou saved my F40 from mediocrity
22:48:10 Quit Accidental ("CGI:IRC (Ping timeout)")
22:49:29Terinjokesfoxxet: although I'm an OSX user, and love iTunes and stuff, I was never a fan of being locked into the OF on the iPod... so I'll agree with you :P
22:49:43 Part foxxet
22:51:08 Part low_light
22:51:11 Join petur [0] (n=petur@rockbox/developer/petur)
22:51:52preglowanyone against deleting fwpatcher from svn?
22:52:10Terinjokespreglow: what was it used for again?
22:52:21preglowTerinjokes: patching firmware files for iriver players
22:52:22preglowon windows
22:52:53preglowi wrote it a couple of years ago when there was no rbutil
22:52:53Terinjokespreglow: no need here (i still flash ipods using the command line tools, so wanted to make sure you guess don't deleted those :P )
22:53:20preglowhmm, the manual still refers to it
22:53:49preglowguess i'll wait until it stops doing that
22:53:51Domonokythe manual doesnt mention rbutil yet, as its not completely ready..
22:54:01 Join ompaul [0] (n=ompaul@freenode/staff/gnewsense.ompaul)
22:54:11Terinjokes(that said, I like how the new boot firmware on the iPods doesn't show the rockbox boot text, how should I got about updating it on my iPod?)
22:54:40preglowthe manual doesn't really mention any linux install method at all for hxx0
22:54:45preglowDomonoky: what's left to do?
22:54:53peturI wouldn't remove fwpatcher as long as the manual mentions it
22:54:58preglowpetur: i won't
22:55:05Domonokyas we switch to the qt version.. more then before.. :-)
22:55:05preglowTerinjokes: by using ipodinstaller?
22:55:12Domonokybut it gets closer..
22:55:15preglowipodpatcher, that is
22:55:17*ender` yawns
22:55:22preglowDomonoky: so you're going qt completely?
22:55:30preglowwhat was wrong with wxw?
22:55:50Domonokywe plan to replace th wx version with qt, when the qt version is on pair..
22:56:16Terinjokespreglow: it will happily update the old boot code on my iPod (which already does rockbox)?
22:56:22Domonokythere is not really something wrong with wx, but its not good to maintain two versions.
22:56:26preglowTerinjokes: it should, afaik...
22:56:39preglowDomonoky: sure no
22:56:46preglowmy gut feeling is qt is a safer bet than wxw anyway
22:56:49preglowqt looks pretty decent
22:57:13Domonokyatleast there a more devs with qt expirience :-)
22:57:47*Bagder has qt experience but not of the positive kind ;-)
22:57:50amiconnWhy does the sim try to switch tuner power for fm-enabled target simulation??
22:59:13DomonokyBagder: so which croosplatform gui toolkit would you choose ( if you have to) :-)
22:59:14preglowBagder: i can imagine it does quite a bit more abstraction than you'd prefer :>
22:59:44BagderDomonoky: I prefer not working with gui toolkits at all ;-)
22:59:50Bagderpreglow: yes that's it basically
22:59:59preglownot only is it c++, but it has moc too
23:00:10Bagderand 400 billion classes
23:00:20Bagderall inheriting each other
23:00:23preglowyeah, but i can hardly see how that is bad
23:00:28preglowlook at the mess gtk is, and that's c
23:00:38BagderI haven't really looked at gtk
23:00:41preglowthey basically ended up reinventing everything c++ is
23:00:42preglowbut in c
23:00:45BagderI had a project to write a gui in qt
23:00:53BagderI found it bloody annoying
23:01:02preglowi find all gui coding bloody annoying
23:01:15BagderI think it boils down to that...
23:01:24Terinjokespreglow: sure enough, updated (wow.... installing rockbox is now a two step process... need to start using the new installation method...)
23:01:41Bagdernot to mention that qt 3 has no alpha support
23:01:50Bagderso I had to bend it in
23:01:52preglowTerinjokes: it's lovely
23:01:59Terinjokes(then again, it kinda awsomes you have a winpod... i guess that could cause a problem for OSX users...)
23:02:01preglowwhy that?
23:02:11Bagderbecause qt4 is huuuuge monster
23:02:16Bagdertalking embedded
23:02:19preglowdeedey, but it's also a modular monster
23:02:23preglowso you can break parts off pretty easily
23:02:33Bagderwell, you can't make it really small
23:02:38preglowhmm, or is that one of the new 4.4 features
23:02:42Bagderthat's only what trolltech claims
23:02:55Domonokysure for embedded you probably want a more lightwight gui toolkit..
23:03:06Bagderbut sure, I haven't tried qt4 much
23:03:15preglowand they even target qt at embedded stuff
23:03:16preglowbut no
23:03:21preglowi can imagine it not being lightweight
23:03:40BagderI did an iptv gui with it
23:03:51Bagderlinux on sh4
23:04:02 Quit Terinjokes ()
23:04:37preglowi've pretty much only used gtk professionally yet
23:04:40preglowor more like gtkmm
23:04:42preglowand it's annoying
23:05:02preglowi've just reconciled with the fact that i hate gui coding
23:05:07Bagderdirectfb seems to be popular
23:05:11Domonokyfor embedded there is a special qt version.. but i dont know how good / lightweight it is.. at least it is seperate :-)
23:05:21 Join Arathis [0] (
23:05:28BagderDomonoky: it _was_ separate in qt3
23:06:06*Domonoky means Qtopia.. it runs on a phone.. :-) ( but phones are powerfull these days) :-)
23:11:42 Quit petur ("*plop*")
23:13:34 Join andrew_ [0] (
23:13:57 Join toffe82_ [0] (
23:18:02 Quit Lear ("ChatZilla [Firefox]")
23:20:09 Join toffe82__ [0] (
23:22:24 Join ddalton [0] (
23:26:23 Quit perrikwp_ ("ChatZilla [Firefox]")
23:28:32 Quit andrewg867 (Read error: 110 (Connection timed out))
23:28:59 Quit toffe82 (Read error: 110 (Connection timed out))
23:32:11 Quit toffe82_ (Read error: 110 (Connection timed out))
23:36:32 Quit ddalton ("I was using BOFHNet IRC version 1.2 by fmillion - get your copy today from !")
23:39:39 Quit rotator ()
23:40:46 Part toffe82__
23:48:37 Join JdGordon [0] (
23:49:59 Part Domonoky
23:55:59 Quit jhMikeS (Read error: 104 (Connection reset by peer))
23:56:06 Join alexc [0] (i=4d2c2c84@gateway/web/cgi-irc/
23:56:19 Join jhMikeS [0] (n=jethead7@rockbox/developer/jhMikeS)
23:58:09alexchey, i've just set up the cygwin environment to develop for ipod, and i'm a little unsure about how the simulator works - i've built it, but when i run it all I get is the screen - is there any way to get the ipod buttons as well? Couldn't find any docs for the simulator
23:58:31rasheralexc: −−background
23:58:38alexcah, quality, thanks!

Previous day | Next day