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

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

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

#rockbox log for 2008-07-13

00:00:30 Quit fyrestorm (Read error: 104 (Connection reset by peer))
00:01:01 Join fyrestorm [0] (
00:01:56 Join cbr|w [0] (n=cbr@
00:03:27 Quit bertrik ("bye")
00:05:03 Quit d-k-t (Read error: 104 (Connection reset by peer))
00:05:34 Join d-k-t [0] (
00:06:32 Quit gregzx (Read error: 110 (Connection timed out))
00:10:19 Quit yoran ("Leaving")
00:14:39 Quit desowin ("KVIrc 4.0.0 Insomnia")
00:18:18 Quit CyBergRind|w (Read error: 110 (Connection timed out))
00:24:26 Quit davina (Remote closed the connection)
00:38:24 Join MarcGuay [0] (
00:45:16 Join tvelocity [0] (
00:49:06MarcGuayCould someone please clarify the new stance on plugin actions? "Use/create a basic five key context (up/down/left/right/select), and use custom code for everything else."
00:50:01MarcGuayIs the idea to totally phase out PLA or just to use it better?
00:58:49 Quit nplus ("Leaving")
00:59:52preglowi haven't followed the discussion at all, but got the impression it's a bad idea
01:00:57MarcGuayIt definitely makes the code a lot more complicated by trying to simplify it.
01:01:25MarcGuayDid you take a gander at that ASM->C plugin?
01:02:13MarcGuayIt didn't strike me as terribly useful outside of making life a bit easier for people trying to learn assembler.
01:03:14 Join Nico_P [50] (n=nicolas@rockbox/developer/NicoP)
01:03:31preglowme neither
01:03:47preglowthe c it made was pretty much just asm
01:04:01preglowit would be a lot more useful if it supported loop constructs
01:04:03preglowbut it just does gotos
01:04:55 Join ctaylorr [0] (
01:05:03preglowafter about a day of practice with disassemblies, you can do very well without it
01:05:27 Join lymeca [0] (n=lymeca@unaffiliated/lymeca)
01:06:11preglowi find the greatest aid is just myself transcribing the disassembly to c by hand
01:06:22preglowand i seldom bother with even that
01:09:59 Quit faemir ("Leaving")
01:11:00***Saving seen data "./dancer.seen"
01:18:13 Join m0f0x [0] (
01:24:10 Quit obo ("bye")
01:26:06 Quit lymeca (Excess Flood)
01:35:37 Quit ompaul (Client Quit)
01:44:30nixternalI picked up a Sansa e260 4gb from Tiger Direct today and it is v1. RockBox is back in my hands and I am a hampy camper!
01:45:37nixternaland finally, I have some color, some themes, and soon all of my tunes...rockbox rox, like you all didn't know that already :)
01:46:23 Join lymeca [0] (n=lymeca@unaffiliated/lymeca)
01:46:40lymecaSorry, did anyone answer my question about 4g 30gb mbr image?
01:48:25MarcGuaylymeca: You can search the IRC logs on the website.
01:53:18 Quit sitwon (Read error: 110 (Connection timed out))
01:54:49 Join Murray_ [0] (
01:55:25Nico_Pnixternal: great news :)
01:58:01 Quit Nico_P (Remote closed the connection)
01:59:18lymecaAccording to the logs my question didn't even make it.
01:59:42lymecaI have a 4th-gen 30gb ipod and I need a MBR image for it
01:59:55lymecaBut the wiki only has 20gb ang 40gb images for 4th gen ipods
02:00:15lymecaWhat should I do to get a 30gb image I can safely dd?
02:00:26Murray_Could someone help me with a problem? I'm tring to use a new theme on my Sansa e200, but when I try to use a wps file it automatically goes to the rockbox default.
02:06:59 Quit culture (Read error: 60 (Operation timed out))
02:09:48MarcGuayDoes anybody here play Doom? I stumbled across this: and can recreate the behavior on builds as far back as November 2007.
02:10:22MarcGuayI'd be surprised if Doom had been broken for that long without an outcry, though.
02:12:29 Join sitwon [0] (
02:13:39LloreanMarcGuay: Have you used Doom before?
02:14:09MarcGuayLlorean: I'd given it a test run a while ago.
02:14:14LloreanIt's a little confusing to get it to start a game, the menu navigation isn't ideal on iPods, and it has a habit of starting the demo quickly enough that it does kinda feel like the keys aren't working.
02:15:07MarcGuayLlorean: I'm testing on the e200. Could you explain how to get a game going to test this properly?
02:15:22LloreanI have no clue on an e200.
02:15:41LloreanI played with it a bit on my Nano back when I was trying to help get it working.
02:16:28LloreanAnd even that was some time ago. I just remember it would seem like keys weren't working unless you knew exactly what to do, since most keys didn't do anything until you got into the game itself.
02:17:24MarcGuayHmm. Seems like you have to choose New Game from the menu before it'll start even though it appears that play has already begun.
02:17:54MarcGuayIt would make sense to bring up the main menu when the game starts, then.
02:18:08LloreanIt's just doing exactly what the PC version of Doom does
02:18:31LloreanThe difference is, the PC version has more keys in general, and people guess the right ones sooner.
02:19:19MarcGuayNot so bright, IMO, considering the enemies are even moving, signifying to me, at least, that the game has started. Anyway, it's nice to close a bug at least.
02:19:45*MarcGuay notices he needs to vary up his lingo
02:20:45LloreanAre you using Freedoom or one of the commercial WAD files
02:21:20PaulJamon h300 doom works fine btw
02:21:30PaulJamalso all buttons
02:21:32LloreanI'm pretty sure it's just a problem with Freedoom itself: The "demo" is bad.
02:21:47MarcGuayPaulJam: Yeah it works fine here too, just that you need to know how to start the game. :)
02:22:03LloreanWhereas if you were to use a commercial Doom/Doom2 WAD file, instead of seeing that idle thing, the game would appear to play itself, making it more obvious what's going on.
02:22:06MarcGuayOr that you need to do it manually.
02:25:18 Quit shotofadds (Read error: 104 (Connection reset by peer))
02:25:35 Quit tvelocity (Remote closed the connection)
02:30:47 Quit sitwon (Read error: 113 (No route to host))
02:32:11 Join sitwon [0] (
02:33:32MarcGuayTrying to reproduce this:, but don't understand what he means by "system info reports 284 tracks", any ideas
02:34:27 Quit Murray_ ("ChatZilla 0.9.83 [Firefox 3.0/2008052906]")
02:46:14lymecaCan anyone help me with finding a MBR image to dd onto my 30GB 4th gen ipod?
02:51:00MarcGuaylymeca: Not too many people around right now, apparently. You might want to post to the forums to up your chances of a response.
02:51:30MarcGuaylymeca: The mailing list can often be helpful as well.
02:53:41 Quit PaulJam (".")
02:56:50 Join thegeek [0] (
03:04:33 Quit m0f0x (Read error: 104 (Connection reset by peer))
03:05:54 Join Thundercloud [0] (
03:09:05MarcGuayWas something committed recently that resumes playback a few seconds earlier from the stop point?
03:09:20 Quit cbr|w (SendQ exceeded)
03:09:29MarcGuayI recall a discussion about this and audio books, but not the outcome.
03:11:04***Saving seen data "./dancer.seen"
03:13:58 Join yoyoned [0] (
03:27:16 Join SirFunk [0] (
03:30:28 Nick JdGordon|zzz is now known as JdGordon (n=jonno@rockbox/developer/JdGordon)
03:39:01 Quit SirFunk_ (Read error: 110 (Connection timed out))
03:49:36 Quit SirFunk (Connection timed out)
03:50:20 Quit Thundercloud (Remote closed the connection)
03:51:53 Join m0f0x [0] (
03:52:56 Quit DerDome (Nick collision from services.)
03:52:57 Join DerDome1 [0] (
03:53:09 Nick DerDome1 is now known as DerDome (
04:09:26 Quit herrwaldo ("Konversation terminated!")
04:13:49 Join SirFunk [0] (
04:15:47 Join Riot- [0] (
04:18:43Riot-Any WPS gurus able to help tell me what I'm doing wrong with this line of code?
04:18:48Riot-%s%al%?ia<%ia - |%?d2<%d2 - |?d1<%d1 - |>>>%?it<%it|%fn>
04:21:23 Quit ctaylorr (Read error: 110 (Connection timed out))
04:23:18Riot-Ah damn sleep dep forgot the % for ?d1
04:36:32Riot-Ok here's better one. I've been modding cabbiev2 for my 5G and I put all the bottom images in viewports so I can rearrange them. I also removed the off shadow from the background for shuffle and repeat and added them to the bitmaps. The problem is now I get an annoying image flicker from whatever the last two images are in the bottom right corner everytime I skip tracks. Any suggestions?
04:36:52 Nick Zarggg is now known as anon0k3 (
04:36:54 Nick anon0k3 is now known as anon9k3 (
04:37:15 Nick anon9k3 is now known as Zarggg (
04:43:11 Join d1sturb [0] (
04:43:38d1sturbIs there anyway I can add a custom visualizer to the WPS [i.e. G-Force]?
04:46:10 Join jhulst [0] (n=jhulst@unaffiliated/jhulst)
04:49:32 Join miepchen^schlaf_ [0] (
04:53:33JdGordonRiot-: check to make sure your viewports arnt overlapping
04:53:36JdGordond1sturb: no
04:54:26Riot-JdGordon: Checked that they're not as far as I can tell
04:55:12JdGordonhave you checked just those vp's or all of them?
04:56:26 Quit d1sturb ("—I-n-v-i-s-i-o-n— 2.0 Build 3515")
04:56:53Riot-I can dcc if you want to take a look for me.
04:59:20Riot-I had that problem with the lock image cutting off text from over lap when i turned the hold on so that was my thinking too. I just can't seem to find anything overlapping
05:00:03Riot-Main reason I put them in viewports was so i could add a text battery display and volume to that line
05:00:50LloreanRiot-: And everything is in a viewport, you don't have anything in the default fullscreen one?
05:02:15Riot-Everything is in a viewport except
05:02:17Riot-#Album art
05:02:17Riot-#Progress bar
05:02:17DBUGEnqueued KICK Riot-
05:02:32Riot-Which is original cabbiev2 code
05:02:44LloreanPlease don't paste multiple lines to the channel, especially if nobody asks for it.
05:03:12LloreanTry ensuring everything is in a viewport. Generally speaking, you shouldn't mix viewports and non-viewported code.
05:04:32Riot-Isn't the code I pasted for album art and progress bar doing the same thing though?
05:04:58Llorean"Same thing"?
05:05:27LloreanWhat do you mean?
05:06:17 Quit miepchen^schlaf (Connection timed out)
05:07:04Riot-well Album art is putting it at x16 y32 w120 h120 wouldn't putting the vieport tag be just doing the same thing for that except using two lines
05:07:59LloreanYes and no. It does it within a default "fullscreen" viewport the way things are now.
05:08:19LloreanThis doesn't cause problems if you don't use any viewports, but may cause strange behaviour once you use them, which is why everything should be in explicit viewports if you're using them.
05:11:06***Saving seen data "./dancer.seen"
05:12:11Riot-I'm not sure how I would implement %CL as a viewport unless it would be V|16|32|120|120| then just %CL on the next line?
05:12:37Riot-or %Cl|-|-|-|-|
05:13:14Riot-ah wait just %C
05:16:40Riot-hmm broke it
05:19:38Riot-Not sure how to implement it into the conditional album cover code
05:22:18Riot-Ok I just removed those two lines totally and I still get the flicker so it's not those
05:28:38LloreanTry removing other bits until you see which bit causes the flicker, then.
05:28:39Riot-it has something to do with %?C<%Vda%C|%Vdb> If I don't have that before any of the viewports it overlaps the album cover to the text on the right but bottom right corner flicker is gone
05:29:20Riot-I just put my clock viewport code before it to make it do that
05:30:06LloreanIf you do have that before any of the viewports, you're displaying the cover art outside of a viewport.
05:30:50Riot-%Cl|16|32|s120|s120| is before it though
05:31:06LloreanWhy are you displaying the cover art twice?
05:31:08Riot-guess that goes back to the yes and no about viewports
05:31:17LloreanOh, wait
05:31:24Llorean%Cl just loads it.
05:31:59 Join xxbatman94xx [0] (i=4c1fa2aa@gateway/web/ajax/
05:32:01LloreanIt's the %C in the conditional there that's displaying it, outside a viewport.
05:32:14LloreanWhether you get rid of the %Cl or not, the %C is still going to be a problem in that conditional.
05:32:19 Quit xxbatman94xx (Client Quit)
05:33:39Riot-One big viewport to encompass it all the the %?C conditional then the conditonal viewports?
05:34:35LloreanI don't understand at all what you mean by that question.
05:40:08Riot-Can I dcc you my wps file or if you got the cabbiev2 theme handy? It's code from there. Might make it easier for me to explain
05:40:42 Quit dataangel (Remote closed the connection)
05:40:59LloreanI keep telling you, you need to create individual viewports for everything
05:41:10LloreanJust because "the original code" works, doesn't mean it will once you change things
05:41:27JdGordon?! a seperate vp for everything?
05:41:33LloreanYou need to create non-overlapping viewports for anything that gets displayed to help ensure things won't conflict.
05:41:44LloreanJdGordon: Not one viewport per item, one viewport per "area"
05:42:04LloreanJdGordon: As in, "don't use the base viewport for things"
05:42:41LloreanBut his two items outside of viewports are the album art, and the progress bar, and since text displays beside the AA, it probably needs its own viewport, not as wide as the PB one.
05:43:14Riot-I understand. I meant to help explain that part you you said you didn't understand what I was talking about.
05:44:01Riot-I apologise the a/c is busted here so I'm not typing clearly.
05:44:17LloreanRiot-: "One big viewport" is bad if it overlaps other viewports.
05:44:47LloreanBut you've already determined the line to focus on. Take the %C out of that conditional, and see if the flicker's there with just the %C gone. If it is, fix that.
05:44:53LloreanEr, if it *isn't*
06:15:16Riot-Ok found my problem. I didn't have the viewport correctly in my clock code. Clock's in the top right the flickering images were in the bottom right and only flickered the same amount as the clock width. I still need to get the album art and progress bar in a viewport but that's for another time
06:15:49Riot-Thank's guys
06:16:02Riot-er Thanks
06:18:31Riot-I reread the part where it says viewports share identifiers so it should be easy then %Vl|a|"album art viewport" the %Vl|a|"now playing text viewport" should still make it work with the conditional
06:24:36 Join Chronon [0] (
06:25:39Riot-I was missing that earlier when i was talking about one big viewport. i thought I'd have to do like %Vl|a|0|0|100|100| then under put %V|0|0|50|50| album art code then %V|51|51|50|50| song info code to get two viewports under one viewport conditional
06:26:05Riot-then %Vl|b|
06:39:42 Quit sitwon ("Leaving")
06:39:57 Join AndyIL [0] (n=pasha_in@
06:40:14 Quit fdinel ("Miranda IM! Smaller, Faster, Easier.")
06:53:43 Quit AndyI (Read error: 110 (Connection timed out))
07:00:27 Nick Soap_OutOfTown is now known as Soap (n=soap@rockbox/staff/soap)
07:02:51 Quit Lambduh ("Leaving")
07:05:49 Quit MarcGuay ("ChatZilla 0.9.83 [Firefox]")
07:11:09***Saving seen data "./dancer.seen"
07:21:29 Quit Chronon ("Leaving")
07:22:13 Join nplus [0] (n=nplus@141.25.Globcom.Net)
07:25:22wpyhhi JdGordon
07:25:42wpyhso, you couldn't reproduce FS #9184 on your dap?
07:26:02wpyhI can, on my 5.5g Video
07:26:35JdGordonwhen? on loading? or?
07:26:38wpyhon loading
07:26:52wpyhthe screen will just display the background
07:27:01wpyhthen the keys won't respond at all
07:27:33wpyhuh, except for holding down the select and menu
07:27:42wpyhsince that's hardware
07:28:08JdGordonok, lemme have a quick look at the code, see if there is anything suspicious
07:32:14JdGordonwpyh: i dont know how the maze save games work.. do you have any?
07:33:03CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
07:33:03*JdGordon was looking at the code for the wrong plugin
07:33:10wpyhum, no
07:33:14wpyhI've never played it before
07:33:19*wpyh feels a little guilty
07:33:25*wpyh is slow.
07:33:37 Quit Zarggg ()
07:34:35JdGordonWTF? ok yes, something is fucked
07:34:39 Join Transience [0] (
07:34:54JdGordongot a nice purple screen now
07:34:56Riot-Right here in River City
07:35:04wpyhJdGordon: you found it? :)
07:35:31Transiencedoes anyone know how to mount a MTP player as a USB hard disk in windows? (Player is Creative ZVM)
07:36:11LloreanTransience: If you are working on a Rockbox port, see the thread in the forum and associated wiki pages for all currently known info
07:36:53Transiencenobody's figured this out yet, i figured someone might have done the same thing with a different player
07:37:22TransienceI'm trying to format the device in FAT, but i can't get windows to mount it as a removable disk
07:37:40LloreanMTP is a different protocol entirely than UMS, and the software on the player controls what gets told to the computer, so you can't simply just bypass it.
07:38:31Transienceso when i mount the player in linux, the player is still controlling what i can do?
07:38:50Transienceokay... thanks
07:39:06 Part Transience
07:45:40wpyhJdGordon: the maze lockup also happens on my iPod Nano
07:45:48wpyhso this might be somehow generic
07:53:50Riot-Any idea what the "s" is in %Cl|16|34|s120|s120|? I couldn't find any documentation on it. it doesn't appear to affect anything with it there or not
07:54:15 Join miepchen^schlaf [0] (
07:55:27 Quit miepchen^schlaf_ (Connection timed out)
07:58:00JdGordonwpyh: i need to try svn on my h300, but its not happeing with whatever build is on it now.. it could be some memory aligment thing on PP, or accessing an int as bool (or vice verca)
07:58:11JdGordonRiot-: max size I think
07:59:14LloreanI believe the resize patch uses that s to denote that "if the file isn't this size it should be resized" but somehow it made it into that WPS despite the lack of the patch
07:59:20LloreanIt does mean it's forward compatible with the patch should it be included.
08:00:33JdGordonthat a bit odd...
08:00:39JdGordonis it likely that patch will get commited?
08:05:16 Join miepchen^schlaf_ [0] (
08:05:18wpyhJdGordon: ok
08:06:19wpyhJdGordon: and one more thing: the ipod just stays in that screen for ages −− it doesn't auto shutdown
08:06:25JdGordonwpyh: im busy atm, but if your keen to find the bug, commeting out big blocks of code untill it crashes is always a fun thing to do on a sunday arvo :p
08:06:33LloreanJdGordon: It's like _a_ bitmap resizing function will get committed
08:06:41LloreanIIUC the one in the patch is ugly and basic.
08:07:12LloreanThere's another patch with a faster one, that I think just uses it for PictureFlow?
08:07:15LloreanEr, not faster
08:07:20LloreanBetter looking, even.
08:07:54LloreanJdGordon: As for the s in the WPS, I imagine it was just a case of the designers doing it out of habit, and whoever committed it never looked.
08:08:07LloreanSince it didn't break parsing, they probably just assumed the S was valid syntax.
08:08:21*JdGordon thought it was used
08:09:35 Quit gevaerts (Nick collision from services.)
08:09:46 Join gevaerts [0] (n=fg@rockbox/developer/gevaerts)
08:09:53LloreanBut I'd like to see the resize patch reach a suitable point and be included.
08:09:58JdGordon says nothing about s
08:10:00LloreanBecause frankly, at that point I might actually use AA
08:10:11LloreanJdGordon: I'm well aware of that. It doesn't exist in official builds. :)
08:10:37LloreanI wouldn't mind collecting AA if I didn't have to create new files if I ever changed WPS
08:11:01JdGordonI have cover.100x100 and cover.120x120 in every folder now because of that
08:11:13scorchenot to mention with different targets
08:11:53 Quit miepchen^schlaf (Connection timed out)
08:12:05LloreanI'd kinda like to just propagate a single cover.bmp with the album when I copy it to multiple players.
08:12:11wpyhum... weird
08:12:25wpyhwhy is credits.c always compiled when I edit a single line in maze.c?
08:12:39JdGordonbecause its stupid :p
08:12:43LloreanActually, assuming the algorithm is modular enough that perhaps it's a single function, I don't see why we can't have the "bad" algorithm in (assuming it resizes on load, for more efficient buffer use) and replace it with the good one when it works well
08:12:48JdGordoncredits.raw is always rebuilt
08:13:13wpyhJdGordon: rebuilt from the header comments in source files?
08:13:33JdGordonno, from docs/CREDITS
08:13:38LloreanFrankly, vastly simplified album art wouldn't be a bad thing for 3.0
08:13:51*Llorean doesn't know why we need to be able to store our images in 40000000 places either.
08:14:28*wpyh finds it weird then, that credits.c is not recompiled if no source file is touched
08:14:36JdGordonAA integration in database would be amazingly awesome also
08:14:37SoapWhy hasn't the maximum battery capacity been upped for the iPod Minis (at least)? Is it simply that nobody has gotten around to it, or is there another reason?
08:14:55LloreanSoap: Probably just nobody's gotten around to it.
08:15:08LloreanSoap: We should probably just set it to 4x the stock for all players.
08:15:15Riot-I just use this to resize all my art. Simple but works. I have to keep two different images anyway. Large ones as folder.jpg for mediaportal and I just convert it on my dap itself
08:15:20LloreanWith the exceptions of the Archoses where it can affect charging.
08:15:20scorcheLlorean: not sure what you mean by that comment...what new features wouldnt be?
08:15:36Lloreanscorche: Which comment?
08:15:48scorche<Llorean> Frankly, vastly simplified album art wouldn't be a bad thing for 3.0
08:16:13Lloreanscorche: Oh. Well, many new features probably require some finagling before they go in. Bitmap resize is probably simple enough, and tested enough, to actually be 3.0-safe even at this late stage.
08:16:26LloreanAs well, it reduces the complexity of using Rockbox, rather than increasing it.
08:17:08scorchewell, yeah, but did you mean you would be aiming for resizing to be in 3.0 before it gets released, or...?
08:17:35LloreanWell it can't work its way into 3.0 after its released...
08:17:48JdGordonbackports :p
08:18:11LloreanOnly bugfixes get backported to point releases, and only when it's a significant enough problem. :-P
08:18:23scorchesorry...i meant having it in svn before 3.0
08:18:36LloreanYeah, I'd like to see it in SVN before 3.0, but that's just personal opinion.
08:18:42LloreanThere may be objections I don't remember.
08:19:14scorcheah...just wanted to make sure you werent considering any new features as blockers ;)
08:19:38 Join sitwon [0] (
08:19:40wpyhJdGordon: it hangs (or maybe gets into an infinite loop) at line 500-501 (FOR_NB_SCREENS(i) maze_draw(...))
08:19:43LloreanAs far as I'm personally concerned, features are never a blocker. :)
08:19:58scorcheof course :)
08:20:08LloreanAlso, I don't get to decide blockers.
08:20:10wpyhif I return from the plugin right after maze_generate, then the text on the status bar becomes black instead of white
08:20:53scorchehence the personal qualifiers
08:23:11JdGordonLlorean: just wait... you may get stuck with the RM job :D
08:23:46LloreanThat might happen, at which point I *do* get to make decisions.
08:24:02LloreanMaybe I should make some very unpopular opinions now, so nobody will risk me being RM. ;)
08:24:05*scorche thinks JdGordon will get stuck with it =P
08:26:30*JdGordon wouldnt really mind doing it... dunno how many "friends" he'll keep if he gets the job though ;p
08:27:35wpyhJdGordon: the ipod hangs inside maze_draw(), I'm looking into it
08:28:31 Join Rob2223 [0] (
08:36:01LloreanJdGordon: Apparently Nico_P has been involved with the resize patch
08:36:49*Llorean wonders how big the binsize increase is.
08:37:21*JdGordon doesnt care about the binsize increase
08:37:43*Llorean does if it's over a certain limit.
08:39:01JdGordonis there a limit on any of the targets which this would affect?
08:39:26JdGordonthe ifp is the only swcodec with a "not far away in the future" limit
08:39:34JdGordonand that port is stalled
08:41:20LloreanThe iFP port already doesn't fit.
08:41:39LloreanAt least, there's not enough room for a few of the codecs at the moment, I believe.
08:42:10LloreanBut I think there's still value in trying to keep the total ram used by Rockbox fairly low.
08:42:23LloreanThat doesn't mean "no binsize increases", but "trying to make sure increases are worth it"
08:42:52LloreanIt seems to me bitmap resizing shouldn't be *too* big.
08:43:04LloreanBut I don't know enough about it to make an educated judgement.
08:43:33*JdGordon wants to get rid of the delta table
08:43:48JdGordonits fun, but really doesnt do much good
08:44:01Soap(I swear I'm off to bed) - Has there been any thought as to releasing 3.0 in two different flavors? Lean and Mean vs Big and Bloated?
08:44:04LloreanIt does make us more aware of binsize increases over time.
08:44:12JdGordonbecause since its been there, its been harder and harder to get reed delta featuers in
08:44:15LloreanSoap: Not even been brought up.
08:44:30LloreanJdGordon: I don't think a single feature's been reverted for red bin, has it?
08:44:39 Join lafille [0] (
08:44:52wpyhsorry to interrupt, but I think maze.c hangs in maze_draw at the first call to getwidth()
08:44:57JdGordonI only remember 1 feature ever being reverted... and no it wasnt because of red delta
08:44:58LloreanBesides, the steering board will be even worse (or better) depending on your perspective, and who gets voted in.
08:45:06JdGordonI'm saying getting them commited has been harder
08:45:07wpyhmaybe the display/screen is not initialized
08:45:23LloreanJdGordon: I don't think the bin table makes it harder. It just makes more people whine. You can ignore them if you're sure enough of the feature being worth it.
08:45:34JdGordonwpyh: you can take for granted that it is setup correctly
08:45:43JdGordonwpyh: s/it/anything in the api
08:46:37 Quit Rob2222 (Read error: 110 (Connection timed out))
08:46:41wpyhthen it may be used incorrectly
08:47:19 Quit lafille (Client Quit)
08:48:00JdGordonwpyh: yes.. its not... display->getwidth/height gets the current viewports w/h which is wrong...
08:48:29*wpyh wonders whether this is related to the recent viewport changes
08:49:07JdGordonwanna test a quick patch?
08:49:10wpyhok ;)
08:52:10wpyhlet's see...
08:57:13wpyhno, it doesn't fix the lockup
09:00:58JdGordonwpyh: ok, change get_width/height to display->lcdwidth and display->lcdheight on line 179/180
09:03:48wpyhJdGordon: please wait a while :)
09:04:08JdGordonno worries
09:04:16*JdGordon still supposed to be doing uni work :p
09:10:16JdGordonyay! done for the sememster... 2 days before sem 2 starts again :p
09:11:12***Saving seen data "./dancer.seen"
09:19:39wpyhJdGordon: I have to apologize
09:19:48 Join K4rP4D [0] (n=KrPD@unaffiliated/krpd)
09:19:59wpyhbut I haven't tried the second idea you suggested, because I was playing with maze.c
09:20:23JdGordonok cool
09:20:32wpyhyeah, and I found out something
09:21:09wpyhafter maze_init() and maze_generate() are called in plugin_start(), if I call rb->screens[0]->getwidth(), then it will freeze
09:21:31wpyhif I call rb->screens[0]->getwidth() before maze_generate is called, then it won't freeze
09:21:41wpyhsounds like stack corruption to me
09:22:19wpyhif I call rb->screens[0]->getwidth() inside maze_generate(), just before the closing brace, it won't freeze (I've tried this, but will verify again)
09:24:23 Quit Acksaw (Read error: 104 (Connection reset by peer))
09:26:33 Join Lambduh [0] (n=Lambda@
09:27:31 Join Acksaw [0] (
09:28:49wpyhJdGordon: if I call getwidth() inside maze_generate, it won't freeze; if I call it after maze_generate returns, it will freeze
09:31:14JdGordonok, svn up.. shouldnt be a rproblem anymore
09:33:41wpyhnope, still same problem
09:33:55wpyhI noticed that you changed the function call to lcdwidth()
09:33:58wpyhbut it didn't help :p
09:34:13wpyhI think the problem lies inside maze_generate()
09:34:21wpyhas I said, stack corruption/overflow/whatever
09:35:26JdGordonbah, lame
09:35:39JdGordonthat change should have stopped the stack corruption
09:35:49wpyhuh.. how?
09:36:20*wpyh doesn't see anything related to a stack in the diff
09:36:41JdGordonjust because :D
09:36:56JdGordonlcdwidth() before was returning something smaller than expected
09:37:55*wpyh is confused
09:51:14wpyhJdGordon: IMHO, you didn't fix the problem, you merely moved it down to the call to clear_display()
09:51:39wpyhI've synced with the latest svn with your patch
09:51:44JdGordonarg... bugger
09:51:50JdGordonyeah, should have tested first
09:51:50wpyhit still locks up
09:52:01wpyhdon't worry, the problem is deeper than that ;)
09:52:31wpyhnow if I comment out the whole while loop inside maze_generate, the lockup doesn't happen
09:53:09wpyhthe maze is not generated, so I'm stuck in the first cell, surrounded by all 4 walls −− the game works :D and I can exit the game
09:53:36wpyhlooking at that while loop, we can see calls to stack push and pop operations
09:53:47wpyhthat means, something went wrong in those calls
09:53:51wpyhmaybe an overflow
09:54:02 Join ompaul [0] (n=ompaul@gnewsense/friend/ompaul)
09:54:03wpyhI don't know, but the problem is definitely there :p
09:54:19*JdGordon is checking the svn log for maze
09:55:26wpyhgood idea :D
09:56:57JdGordonr13996 is porbably the culrpit then
09:57:40wpyhmaybe... that's a *lot* of code
09:57:48wpyh(which I don't understand)
09:58:46*JdGordon cant really bo bothered figureing it out also
09:59:29wpyhit's really complex
09:59:58wpyhmaybe we should just rewrite the whole thing :p
10:00:19wpyhanyway, should your patch be reverted then?
10:00:33JdGordonno, its still better than the previous code
10:00:42*JdGordon added some debug splashes
10:01:12wpyhhow do we add debug splashes?
10:01:18*wpyh is a noob
10:01:44JdGordonbloody typical.... works fine again
10:01:55JdGordonrb->splash(5*HZ, "message");
10:02:06*wpyh thinks we can't just hide the bug by using splashes
10:02:09wpyhwe need to fix it
10:04:08*wpyh realizes he's just violated a rule
10:04:48JdGordononly one?
10:04:56 Quit Seed (Nick collision from services.)
10:05:00 Join Seedy [0] (
10:05:02wpyhuh... I can only spot one
10:06:53*wpyh wonders where the splash() function resides
10:06:58JdGordonhmm... could it infinite loop in the if (available_neighbours == 0){ line
10:07:00 Join stripwax [0] (
10:07:19 Quit K4rP4D (
10:07:19 Quit tarbo (
10:07:19 Quit Strogg (
10:07:19 Quit UncleRemus (
10:07:19 Quit FOAD (
10:07:19 Quit preglow (
10:07:50wpyhJdGordon: that's not the problem
10:07:53 Join tessarakt2 [0] (
10:07:55 Quit tessarakt (Read error: 110 (Connection timed out))
10:08:13wpyhI've commented out that whole if-block and replaced it with visited_cells++, but the same problem still happens
10:08:24NJoinpreglow [0] (
10:08:27NJoinFOAD [0] (
10:08:42wpyhlooking at maze_pick_random_neighbour_cell_with_walls(), I see some more pop and pushes -_-"
10:08:51NJoinUncleRemus [0] (
10:09:04NJointarbo [0] (n=me@unaffiliated/tarbo)
10:10:54*wpyh seriously thinks the maze code needs cleanup
10:11:09wpyhJdGordon: are RB devs interested in patches that only cleans up code?
10:11:18wpyhlike for example in maze_init
10:11:30wpyhI'm thinking of moving the border setup out
10:11:43JdGordonusually yes... but plugins are really low on everyone priorities
10:11:47wpyhout of the loop, and putting it in its own loop
10:12:11wpyhhow about I post the patches directly here on pastebin?
10:12:28stripwaxbetter to paste to flyspray
10:15:59 Join d-k-t-2 [0] (
10:17:49 Join Rogg [0] (n=jean@
10:17:49NJoinK4rP4D [0] (n=KrPD@unaffiliated/krpd)
10:17:49 Join Seed [0] (
10:19:57 Quit ompaul (Client Quit)
10:21:46 Quit fyrestorm (Read error: 104 (Connection reset by peer))
10:22:07 Join fyrestorm [0] (
10:22:47 Quit d-k-t (Read error: 110 (Connection timed out))
10:24:01 Quit K4rP4D (
10:24:01 Quit Seed (
10:24:01 Quit Rogg (
10:27:51 Join domonoky [0] (n=Domonoky@rockbox/developer/domonoky)
10:29:06wpyhJdGordon: where is the function rb->splash located? and what does it do?
10:29:22wpyhI can only find a function called splash() in apps/gui/splash.c, but it's static
10:29:46wpyh(and doesn't have the same parameter list)
10:31:05stripwaxwpyh - it pops up a message look in apps/plugin.h
10:31:16stripwax(and apps/plugin.c)
10:32:08stripwaxrb->splash points to the function called gui_syncsplash
10:32:14domonokythe splash uses in plugins is a "gui_syncsplash" see plugin.c
10:32:20wpyhok, thanks :)
10:34:08wpyhah, no wonder I can't find the assignment
10:35:43 Quit Horscht (Read error: 110 (Connection timed out))
10:38:20wpyhyay! I fixed the problem, although in the wrong way
10:38:44wpyhJdGordon: put set_viewport(NULL) immediately before maze_draw and the problem goes away
10:40:08 Quit d-k-t-2 (Read error: 110 (Connection timed out))
10:40:11 Join d-k-t [0] (
10:40:42 Join Horscht [0] (
10:40:42NJoinK4rP4D [0] (n=KrPD@unaffiliated/krpd)
10:40:50 Join macnoob [0] (
10:41:17 Part macnoob ("Konversation terminated!")
10:42:09NJoinRogg [0] (n=jean@
10:42:13 Quit Guest36577 (Remote closed the connection)
10:44:05 Join fml [0] (n=4fd3f80b@gateway/web/cgi-irc/
10:48:31wpyhah, i knew it
10:48:34wpyhstack overflow
10:49:00wpyhif I set the MAZE_HEIGHT and MAZE_WIDTH to something small (for example 12 and 16), then the game works fine
10:49:13wpyhset it back to the default (24 and 32) then it won't work
10:54:20 Nick miepchen^schlaf_ is now known as miepchen^schlaf (
10:55:18wpyhif I set the maze to be 32x20, it works
10:55:29wpyhif I set it to 32x23, it will error out with
10:55:39 Join davina [0] (
10:55:42wpyh"*PANIC*\nStkov main (0)"
10:55:48wpyhdoes anyone know what this means?
10:56:11domonokyyou are allocating too much on stack.
10:56:25stripwaxor recursing too deeply
10:56:37wpyhI wonder why using 32x24 doesn't generate this error
10:56:47stripwaxThreads can allocate more stack, I think..
10:57:22stripwaxMaybe it needs to be an even number
10:57:42wpyhI'll try 32x22 and 32x26
10:58:22wpyh32x22 works fine
10:59:03wpyhstripwax: 32x24 doesn't generate the error, but corrupts the stack
10:59:21stripwaxhm, curious
10:59:23wpyhor memory
10:59:30wpyhah, 32x26 works
10:59:37wpyhso what's special about 32x24?
11:01:20domonokythe buffer for the maze is allocated on stack, but if this is the problem, all bigger mazes shouldnt work, and not only a few sizes..
11:01:58wpyhmaybe for small sizes the stack is sufficient, for bigger mazes the stack is automatically extended, for the right size the stack is not automatically extended?
11:02:14 Quit fml ("CGI:IRC")
11:02:39wpyhok, confirmed that 32x25 works
11:02:46domonokystack is never autmatically extended. We have fixed sized stacks for our threads (you could change them at compile time)...
11:02:49wpyhso, only 32x23 and 32x24 generates problems
11:03:31domonokybut maybe the problem is somewhere in the calculation, so it puts more on stack depending on how big with and height are..
11:06:16*wpyh can't see it
11:07:05wpyhbtw, for 32x23 I can see the whole maze for a split second
11:07:07wpyhthen it disappears
11:07:12wpyhand I see the PANIC
11:07:42wpyhso actually 32x23 got a little farther than 32x24
11:07:53wpyhand this should explain why some people can't reproduce the bug
11:08:21 Join daurn [0] (n=daurnima@unaffiliated/daurnimator)
11:08:44domonokyyou could gradually comment out things, to find out in which function the problem is..
11:09:35 Join merbanan [0] (n=banan@
11:10:15wpyhwell, actually I've found where the problem starts
11:10:19wpyhit's in maze_generate
11:10:44 Join bertrik [0] (
11:11:16wpyhbut the problem is, I don't know why it happens
11:11:17***Saving seen data "./dancer.seen"
11:13:18 Join fml [0] (n=4fd3f80b@gateway/web/cgi-irc/
11:15:32JdGordonwpyh: moving the set_viewport call shouldnt change anythign
11:15:48wpyhbut it does :)
11:16:10JdGordonalso, im wondering why main is stkoving
11:16:40*wpyh neither
11:16:45JdGordonoh wait...
11:16:51wpyhthe stack is corrupt, I'm sure of that
11:16:59wpyhI did another experiment
11:17:00domonokyplugins run in the main thread, so if you produce a stackoverflow, i would expect this..
11:17:16JdGordonyeah, but i thought the large array was being alloced on the heap...
11:17:23JdGordonwhich it apparently isnt
11:17:38wpyhI use 32x23 and run the game
11:17:41domonokyno, maze allocates everything on stack, as far as i checked the source..
11:17:45wpyhit gives the Stkov panic
11:18:09JdGordonyeah looing now
11:18:12JdGordonlooking now
11:18:24wpyhthen if I return from plugin_start before the switch(button) near line 520, it won't panic,
11:18:47JdGordonmaze_pick_random_neighbour_cell_with_walls is going to be the problem
11:18:57*wpyh hopes JdGordon is rightr
11:19:26JdGordonactually maybe not...
11:19:34JdGordonits not called recursivly
11:20:06wpyhcommenting out the other lines below maze_pick_random_neighbour_cell in maze_generate still generates the problem
11:20:13JdGordona 736*2 byte arra shouldnt be causing a stkov
11:20:56*wpyh thought int is 4 bytes
11:21:04domonokyBut maze allocated more of this buffers (all those internal coord_stacks) but i am not sure if this is the problem..
11:21:16JdGordonso it is... i thought is was a short
11:21:50domonokythe coord_stack(s) is shorts width*height*short
11:23:03JdGordonive moved the 2 coord_stack structs out of their functions so its not on the stack anymore, testing on target
11:23:04domonokythe otherway round :-) the maze buffer is "short int", the coord_stacks are "int"
11:23:40JdGordonthey are re-init each call so dont need to be local vsriables
11:25:41JdGordonand once again it works fine on target so dunno if that really fixes it or not
11:26:15*wpyh thanks domonoky and JdGordon
11:26:56wpyhI changed the maze in struct maze to int and it fixed the problem
11:27:11wpyhbut the whole thing is still braindamaged
11:27:46wpyhlike the push and pop functions
11:27:47JdGordonmain is at 42% before running maze... 52% after
11:28:05wpyhthey shift the data by 8, while using int
11:28:26JdGordonwe welcome a patch :)
11:28:57wpyhJdGordon: I have to clean it up some more, so maybe tonight
11:29:12 Quit fml ("CGI:IRC (EOF)")
11:29:17wpyhwill you accept a patch that fixes the problem and cleans up the code?
11:29:23wpyhor should I split them into two?
11:29:36*JdGordon doesnt really care
11:32:36 Quit bertrik (Read error: 110 (Connection timed out))
11:33:34 Join bertrik [0] (
11:35:28 Join voltagex_ [0] (
11:35:53voltagex_hi, this might be the only place to ask this - what kind of soldering iron do I need to solder on the H300 board?
11:36:21JdGordonone with a hot metal tip
11:37:20*wpyh thinks he may be asking about the tip shape
11:38:06 Join Lear [0] (i=chatzill@rockbox/developer/lear)
11:38:29*domonoky thinks any solder iron with a small enough tip should work.
11:40:55 Join desowin [0] (n=desowin@atheme/member/desowin)
11:41:05voltagex_JdGordon: you know what I'm trying to repair, lol :P
11:41:12 Nick voltagex_ is now known as VoltageX (
11:41:36VoltageXdomonoky: I, err seperated the "ribbon" off the board of my H340.
11:41:52VoltageXit's not destroyed, because if i press it back into position the player works fine.
11:43:04 Quit BHSPitLappy ("Ex-Chat")
11:52:19 Join oofus [0] (
11:55:16 Quit daurn (Remote closed the connection)
11:59:40wpyhVoltageX: I've never repaired such a thing before, but I think you can just reflow the solder connecting the ribbon cable and the board
12:01:09VoltageX wpyh you mean touch it gently to melt it a bit then stick it back together?
12:02:23wpyhmelt the solder points a bit and let it cool down
12:02:48wpyhJdGordon: the problem is not entirely fixed yet, because maze_solve() still hangs with a "data abort"
12:02:51 Join herrwaldo [0] (
12:03:02wpyhand I thought Stkov was some Russian developer's name...
12:03:44VoltageXok, i can only cause it to be irrepairable, worth a shot!
12:04:06wpyhVoltageX: disclaimer: I haven't tried that procedure before
12:04:23VoltageXwpyh: neither :P
12:07:00wpyhok, maze_solve is fixed by simply using uint16_t instead of unsigned int
12:07:15wpyhall: do you prefer (unsigned short) or (uint16_t)?
12:09:30stripwaxunsigned short
12:09:44*linuxstb was about to say the opposite ;)
12:09:46wpyhmay I ask why?
12:10:09linuxstbBut I don't think it matters - afaik, short is 16-bit everywhere.
12:11:19 Quit avis (Remote closed the connection)
12:11:20stripwaxI guess it doesn't have to be. I've not looked at the code, so if it assumes the data contains exactly 16 bits, uint16_t makes more sense
12:12:26 Join avis [0] (n=ident@pdpc/supporter/student/avis)
12:12:40wpyhwell, then I'll use uint16_t
12:17:00*wpyh is guilty of not using proper context
12:28:22 Quit GodEater_ (Remote closed the connection)
12:28:37 Join moos [0] (
12:29:10 Join GodEater_ [0] (n=ge@rockbox/staff/GodEater)
12:36:55wpyhis anyone opposed to the idea of forcing square boxes for the maze?
12:37:25wpyhas in, calculate the MAZE_WIDTH and MAZE_HEIGHT using a predefined square box size
12:43:09wpyhanother question
12:43:37wpyhall: do you prefer x++ or ++x, when outside an equation? (so that both has the same effect)
12:46:06linuxstbwpyh: If you do that, be aware that the Archos LCDs don't have square pixels
12:46:50wpyhany other LCD's with non-square pixels?
12:48:18wpyhand what ratio are those LCD pixels in?
12:48:52wpyhand another question: should all of a plugin's functions be static?
12:49:56wpyh(except for plugin_start)
12:51:26bertrikwpyh, I generally prefer x++. Plugin functions don't need to be static AFAIK, but it's nice to make them static
12:51:39 Quit nplus (Read error: 104 (Connection reset by peer))
12:51:41wpyhbertrik: thanks
12:51:59wpyhI prefer ++x, but I do realize that 95% of programmers prefer x++
12:52:12wpyhI'll make the functions in maze.c static then
12:52:53stripwaxAs long as you're not using the result of x, it doesn't really matter
12:53:10wpyhstripwax: yes, it doesn't actually matter
12:53:15wpyhjust a matter of style :p
12:53:23wpyhI'll follow the general style then ;)
12:54:50bertrikI think in the main rockbox firmware, local functions should be made static to limit the scope and prevent name clashes, but plugins are not directly linked to the main firmware so it doesn't really matter.
12:55:44wpyhI think we should make them static to be consistent
13:00:03 Join nplus [0] (n=nplus@141.25.Globcom.Net)
13:02:59 Join VoltageX_ [0] (
13:03:18 Join culture [0] (
13:03:27linuxstbbertrik: Using static also means gcc can perform more optimisations - the function doesn't need to be exported, so for example, could be made inline.
13:03:31 Join robin0800 [0] (
13:04:01 Join Nico_P [50] (n=nicolas@rockbox/developer/NicoP)
13:04:24 Join einhirn_ [0] (
13:07:00bertriklinuxstb, I usually compile with an extra flag -Wmissing-prototypes to check for possible static candidates and missing #includes and it's mostly in shape for the main firmware, but if you want to fix this for plugins too there's a _lot_ of functions to be changed
13:07:16wpyhlet's fix the functions one by one
13:07:24wpyhthen it won't seem like too many functions
13:07:27 Quit Horscht ("IRC is just multiplayer notepad")
13:09:18wpyhbtw, why is the maze in the maze plugin drawn for all screens?
13:09:26wpyh(I see FOR_NB_SCREENS everywhere)
13:09:49bertrikwpyh, I think it's a lot of change with negligible impact, so I think it's not worth it
13:10:21wpyhbertrik: well, I've changed all the functions in maze.c...
13:11:05bertrikactually some plugins stopped working when I changed all of their functions to static... no idea why
13:11:21***Saving seen data "./dancer.seen"
13:11:28wpyhI'll test it before posting a patch :)
13:12:10wpyhbertrik: in your case, could it be that the functions are all inlined and the rock becomes too big?
13:13:46wpyhanyway, is it ok if I change the screen drawing functions and remove the FOR_NB_SCREENS() loops?
13:14:17bertrikwpyh, maybe I haven't looked into that. It would be nice if something like that could be detected at compile-time.
13:14:43*wpyh doesn't understand how to detect that at compile time, but probably just compare the sizes?
13:16:01wpyhI'm for removing the FOR_NB_SCREENS loops, because they are wrong when there is a remote and the remote LCD size and type is different with the main LCD
13:16:15wpyhuh, not wrong, but almost useless
13:16:50bertrikwpyh, I don't know if they should be removed, you're the expert now!
13:17:03wpyhwell, it's just a suggestion
13:17:39wpyhlike, the maze will of course be shown on the remote, but it will be small
13:18:38wpyhbut I don't have a player with a remote, so I can't just remove it, since I can't test it :p
13:19:47linuxstbbertrik,wpyh: The plugin code-size becoming too big _would_ be detected at compile-time - by the linker.
13:19:52 Quit VoltageX (Read error: 110 (Connection timed out))
13:21:08bertriklinuxstb, ok, I'll test that sometime
13:23:20 Join DrMoos [0] (
13:24:50 Quit moos (Read error: 104 (Connection reset by peer))
13:27:01 Join obo [0] (n=obo@rockbox/developer/obo)
13:27:39 Nick DrMoos is now known as moos (
13:28:12 Quit bughunter2 ("Leaving.")
13:40:02 Join fdinel [0] (
13:40:28 Join bughunter2 [0] (n=Jelle@
13:44:04 Quit DerDome (Read error: 104 (Connection reset by peer))
13:44:07 Join DerDome1 [0] (
13:44:19 Nick DerDome1 is now known as DerDome (
13:44:21wpyhmay I know who implemented the rand() function?
13:44:37wpyhI almost always get the same maze!
13:46:48linuxstbAre you seeding it?
13:47:01 Part yoyoned
13:47:22 Join [CBR]Unspoken|w [0] (n=cbr@
13:48:12wpyhlinuxstb: no
13:48:16wpyhhow do I seed it? :-\
13:49:02wpyhuh... /me thought it was a system setting / option somewhere
13:49:13wpyhapparently maze.c didn't seed it
13:49:33wpyhshouldn't it be made global?
13:49:47linuxstbShould what be made global?
13:49:51wpyhthe seed
13:50:02wpyhso all plugins share the same randomness?
13:50:04*preglow spots mixed case identifiers
13:50:27stripwaxwhat does "the same randomness" mean?
13:50:59linuxstbpreglow: The police have been slacking...
13:51:23wpyhuh... I mean initialize it once at boot, so that every plugin doesn't need to initialize it again
13:51:51wpyhbut I just realized that it would be messy, carrying a global seed around...
13:51:53linuxstbThat's missing the point of seeding it when the plugin is launched.
13:52:15*wpyh is stupid
13:52:25wpyhlet's make maze.c seed the rng then :)
13:52:58 Quit Riot- ("I have to go bathe mother.")
13:54:11 Join daurn [0] (n=daurnima@unaffiliated/daurnimator)
14:07:36 Join m0f0x_ [0] (
14:09:52 Nick Rogg is now known as Strogg (n=jean@unaffiliated/strogg)
14:09:55 Join tvelocity [0] (
14:13:19 Join PaulJam [0] (
14:13:45 Quit daurn (Remote closed the connection)
14:17:47 Quit moos (Read error: 110 (Connection timed out))
14:19:12 Quit m0f0x (Read error: 110 (Connection timed out))
14:19:49 Nick Seedy is now known as Seed (
14:24:09 Join Horscht [0] (n=Horscht@xbmc/user/horscht)
14:35:07 Join ompaul [0] (n=ompaul@gnewsense/friend/ompaul)
14:42:18 Join mf0102 [0] (n=michi@
14:45:59 Join MTee [0] (n=mtee@nat/ibm/x-a620db6219fa87df)
14:46:49 Quit MTee (Client Quit)
14:48:20 Join daurn|iphone [0] (
14:50:13 Quit daurn|iphone (Client Quit)
14:50:59avishow can i ensure i get a v1 of the sansa e280 ?
14:51:28linuxstbTake it out of the box and check the version number of the Sandisk firmware running on it.
14:52:01aviswhen i mean is, when ordering online how can i ensure it will be a v1 ?
14:52:01wpyhdoes anyone know the purpose of "if(!quit) rb->yield();" at the end of maze.c?
14:52:13wpyhthere's a comment there: /* BUG alert: always yielding causes data abort */
14:52:17wpyhwhat does that mean?
14:53:21stripwaxavis - you can ask the seller if they can open the box and check; otherwise you can't ensure it will be v1
14:54:40avisthanks stripwax
14:54:50 Quit Lear ("ChatZilla 0.9.83 [Firefox 3.0/2008052906]")
14:55:00wpyhI've tried removing the if clause, but it doesn't give me any data abort or such stuff
14:55:03stripwaxwphy - sounds suspicious. it might be a workaround to some kind of (possibly previously unreported) threading bug, does it work if you take out the if?
14:55:13*wpyh is wpyh
14:55:13stripwaxjinx. then you're probably ok to leave it out
14:55:19bertrikwpyh, must be a work-around for a bug that the original coder did not understand. Removing the if would also be my suggestion.
14:55:26stripwaxwpyh - oops! sorry
14:56:05wpyhI think the original author hit that bug we discussed earlier about stack overflow, but didn't know it was because of the stack
14:56:10wpyhI've removed it now
14:56:14stripwaxyep, sounds like it
14:57:22*bertrik looks for more BUG comments
14:58:56 Join Hobson [0] (n=Hobson@
14:59:12Hobsonhi everyone
14:59:17*wpyh waves
15:00:11*stripwax profiles libmad layer3 with some inline functions forced to not inline, just to see what happens..
15:03:05stripwaxrockbox usb mode still really messes with my laptop - renders it mostly unusable for a good thirty seconds after plugging in. any ideas why that is?
15:03:38*wpyh suggests stripwax to check his dmesg
15:03:50stripwaxoddly, my irc client works ok. but anything remotely related to Explorer hangs
15:03:57*wpyh finds that maze_draw is a huge waste of resources
15:04:01*stripwax wishes XP had a dmesg
15:04:04bertrikstripwax, initial versions of gevaerts' stack did that for me on windows, but later versions seemed to fix it completely.
15:04:15wpyhstripwax: sorry :p
15:04:15stripwaxcurrent build
15:04:41wpyhin that case, I would suggest that you log in as admin and check the event logs
15:05:25wpyhall: is rb->display() called from a plugin expensive?
15:05:34 Nick fxb__ is now known as fxb (
15:06:52stripwaxsaratoga - (for the logs) - is pretty detailed now. relative timings are probably about right, and call counts certainly are, but the instrumentation definitely slows things down quite a lot
15:07:09stripwaxwpyh - nothing in the event logs unfortunately. nothing interesting anyway
15:07:32wpyhdoes it do that to your laptop if you're on linux?
15:08:09stripwaxwpyh - given that I don't run linux on my laptop, it's hard to answer that one. I don't think it's a hardware problem with my laptop, if that's what you mean.
15:08:23wpyhon linux we can get the dmesg :)
15:08:27stripwaxha, true
15:08:30wpyhmaybe you should try a live cd sometime
15:08:42HobsonI wish there was a way to put rockbox into my mp3 thing
15:08:49Hobsonits original firmware really sucks
15:08:59stripwaxHobson - any clues what your mp3 thing is?
15:09:03bertrikHobson, please define mp3 thing :)
15:09:20Hobsonok brand is nortek
15:09:24stripwaxwpyh - I do from time to time, and I might if I think it will help diagnose what's up! :)
15:09:27avishow well does the microsd slot work in a sansa e270 ? also what is the maximum supported space ?
15:09:34Hobsonin a min I'll tell the model
15:09:41wpyhgood luck ;)
15:09:44stripwaxfor now I just use the OF for usb because it's still more stable
15:09:57bertrikavis, works very well for me, only tried an 4GB SDHC card though
15:10:06wpyhstripwax: what's your dap?
15:10:06Hobsonnortek vee live
15:10:13stripwaxwpyh - ipod 5g 60GB
15:10:18bertrikavis, SDHC can go up to 32 GB but I don't know if that's been tested
15:10:25wpyhsimilar to mine −− ipod 5.5g 8gb
15:10:25avisbertrik, does that expand the capacity beyond its native memory, or replaces it ?
15:10:39wpyhbut I'm not aware of a rb usb stack for the ipod
15:10:42avisbertrik, the e270 i'm looking at is microsd so i'm not sure if that'll work
15:10:43wpyhyeah, CF :p
15:10:59stripwaxwpyh - ah, you've been putting up those battery benches :)
15:11:05bertrikavis, it expands the capacity, you get an extra 'drive' in rockbox
15:11:11wpyhcf looks nice
15:11:13stripwaxwpyh - yeah the usb stack isn't compiled in by default, but it's there alright
15:11:25wpyhwell, maybe I should try that sometime
15:11:26***Saving seen data "./dancer.seen"
15:11:38stripwaxwpyh - let us know your dmesg too :)
15:11:39wpyhright now RB always resets and goes into the emergency disk mode
15:11:44wpyhok ;)
15:11:45avisbertrik, so i would not be able to check the whole space at once i presume ? for example to check for duplicates ?
15:11:49Hobsonmain goal would be scrobbling tracks to lastfm
15:11:59stripwaxyeah, it will do that if the usb storage isn't enabled in your build
15:12:07Hobsonbut I think I'll get an ipod and I'll solve the thing out
15:12:27bertrikavis, you can get a file view just like in windows for example, you'll probably see double entries in the database
15:12:32stripwaxHobson - careful, none of the currently-manufactured ipod models will work with rockbox.
15:12:36stripwaxonly the older models will work...
15:12:49stripwaxFifth Generation or older
15:13:09Hobsonty stirpwax
15:13:16*stripwax is stripwax :)
15:13:51*wpyh can't sue stripwax for copyright infringement because of the extra emoticon
15:14:16avisthank you bertrik. are the controls on the sansa alright with rockbox ? awhile back i got rid of a gigabeat because i didn't like the controls
15:14:37Hobsonanyway I think that's a shame that nortek player could be way better if it wasnt for the s****Y firmware
15:15:21stripwaxHobson - pictures of it look quite nice. Do you know what CPU it uses or any hardware details?
15:15:23*wpyh thinks optimizing maze_draw is too complicated for today
15:15:39 Quit desowin ("KVIrc 4.0.0 Insomnia")
15:15:42*stripwax is sorta looking for things to do instead of his Real Life work he should be doing
15:16:26Hobsonuhm I could find them, but I got to say I'm not what u call an expert at this
15:16:28bertrikavis, yes I think the controls are fine, it's my personal opinion of course
15:16:58stripwaxHobson - finding the details out will be a starting point.
15:17:25wpyhstripwax: I assume the usb stack can be enabled by simply defining USB_STORAGE?
15:17:59stripwaxHobson - If it is similar to other devices, it may be relatively easy to port rockbox onto it, but without technical details it's impossible to know
15:18:19Hobsonactually I found out it's exactly the same of an egoman mp4 player, I found that in a string between the nortek firmware files. then checking that out on the web I found out the original egoman one
15:19:09stripwaxwpyh - yeah, and USE_ROCKBOX_USB, and HAVE_USBSTACK. not sure USB_STORAGE is needed separately (I don't have it)
15:19:14stripwax(and mine works)
15:19:21stripwax(but I could be wrong)
15:19:23bertrikwpyh, the way I do it, is by adding -DUSE_ROCKBOX_USB -DUSE_HIGH_SPEED to the EXTRA_DEFINES in the master Makefile
15:19:46stripwaxI just put mine in firmware/export/config-ipodvideo.h
15:20:21bertrikthe make clean; make. The clean is required because sources are not rebuilt when changing the Makefile.
15:20:42wpyhthen I'll just put it in autoconf.h −− eliminats the make clean requirement
15:20:48stripwaxbut are when you change config-xxx.h !
15:20:59*Hobson searching the details
15:21:36wpyhwell, firmware/export/xx is supposed to be static...
15:22:14stripwaxIt's as good a place as any to put hardware specific system-wide defines..
15:22:28 Join desowin [0] (n=desowin@atheme/member/desowin)
15:22:33wpyhwhy not in autoconf.h?
15:23:01*bertrik wonders about the difference between HAVE_USBSTACK and USE_ROCKBOX_USB
15:23:02stripwaxwhen does autoconf.h get used in the build process
15:23:26wpyhI think autoconf.h is included in every header file?
15:24:12stripwaxYeah, that's probably better then!
15:24:18*stripwax learns something new
15:26:12wpyhand btw, nothing strange on my laptop when I used RB's usb stack
15:26:37wpyhthe biggest difference is the mode sense, and the device name
15:26:55wpyhand the fact that there's only 1 configuration
15:27:23 Quit VoltageX_ (Read error: 110 (Connection timed out))
15:27:35stripwaxhm, but also, you're not really using a real ipod :)
15:27:45stripwaxmine is not CF modded
15:27:48wpyhstripwax: why?
15:28:02wpyhoh, well... I don't think it should matter much..
15:28:55Hobsonstirpwax, any hint on what kind of data I should look for (sorry for asking very unspecific questions, I know it's hard to help like this)
15:29:54stripwaxHobson - well, as much as possible really. what make/model CPU it uses, also make/model of LCD controller, and the format of firmware upgrades it takes.
15:30:26stripwaxI'm suspecting it shares a lot of similarities with alternative models that are already supported by rockbox
15:30:42wpyhI've also tested the OF's usb mode:
15:31:43Hobsonstripwax, I have a txt file from the firmware rar archive, I think most of the details are described in it. Do I paste a few strings in here or what?
15:32:07HobsonI mean it's an archive I downloaded from the official site for the firmware update
15:35:28linuxstbHobson: Do you have a link to the firmware download?
15:36:36Hobsonyes but u have to be registered to download it. I can upload it anywhere if it helps
15:37:03 Quit bertrik (Remote closed the connection)
15:38:00HobsonI might have found the specifics
15:38:06Hobsonin the user manual
15:38:11Hobsonwho would expect that
15:38:19wpyhweird, in RB's usb mode, sometimes the backlight always stays on
15:39:08 Quit perrikwp (" ajax IRC Client")
15:41:58Hobson1 Capacity of Flash Memory Support Micro SD/T-Flash Card, up to 2G; Build-in
15:41:58HobsonFlash Memory, up to 4G.
15:41:58Hobson2 USB Interface USB2.0 (High Speed)
15:41:58DBUGEnqueued KICK Hobson
15:41:58Hobson3 LCD Display 1.8" TFT (128X160)
15:41:58Hobson4 Power supply Li-Polymer
15:41:58***Alert Mode level 1
15:41:58Hobson5 S/N ratio > 85dB
15:42:00Hobson6 Audio Frequency Range 20hz \endash 20khz
15:42:04Hobson8 Decoding Format MP3 (MPEG 1.0/2.0/2.5, Layer 1/2/3, ID3 display
15:42:06Hobsonsupported) and WMA(8Kbps-340Kbps), MTV and JPG
15:42:08Hobson9 FM 87.50 −−−− 108.00MHz(Normal) 30 Station preset support
15:42:10Hobson10 File System FAT16, FAT32
15:42:12Hobson11 Operating Temperature 5OC −−−−35 OC (41 OF−−-95 OF)
15:42:14Hobson12 Operating Moisture 10%\emdash 75%
15:42:52Hobsonnothing that helps?
15:43:05linuxstbHobson: Please don't paste large amounts of text here - use a site like
15:43:13linuxstbBut no, that's no help at all...
15:43:23Hobsonsorry guys
15:44:46Hobsonthis is the firmware file I was talking about before:
15:45:33Hobsonthe txt one
15:48:17Hobsonnothing there either uh
15:51:59***Alert Mode OFF
15:58:57Hobsonthis is the firmware update archive as I downloaded it from the site:
15:59:53Hobsonand this is the chinese identical thing: I guess they just changed the name to it lol
16:00:25Hobsonthanks anyway for the help guys, so kind
16:05:12preglowhow is study mode supposed to act when pressing skip around the file boundaries?
16:07:15preglowalso, do we always want to call playlist entities "tracks" ?
16:07:17preglowthey're called "song" here and there
16:08:57linuxstbHobson: I can see the string "ALiOS1.0", which rings a bell, but I can't remember what it is...
16:11:18 Join CyBergRind|w [0] (n=cbr@
16:11:24HobsonI really wouldnt know
16:11:30 Quit [CBR]Unspoken|w (Read error: 104 (Connection reset by peer))
16:12:38 Join Arathis [0] (
16:17:30 Quit desowin (Read error: 110 (Connection timed out))
16:20:49stripwaxthe usb bridge is a standard ALi chipset - not sure what ALiOS is
16:23:53 Join fml [0] (n=4fd3f80b@gateway/web/cgi-irc/
16:24:27linuxstbHobson: But the firmware doesn't look like anything I've seen before, so it would seem unlikely that your device is similar to any existing Rockbox device...
16:24:28 Quit jhulst (Read error: 113 (No route to host))
16:24:43 Quit nplus (Read error: 104 (Connection reset by peer))
16:25:13fmlNico_P: ping
16:25:27Hobsonthanks for your help anyway
16:25:44 Join nplus [0] (n=nplus@141.25.Globcom.Net)
16:26:57PaulJamHobson: if you feel like you are up to the task, then you can of course start a port of rockbox to this player. see here:
16:27:06 Join VoltageX [0] (
16:29:08Hobsonmmm I might try that to fill my next sleepless nights, but I am not sure I have the knowledge for that
16:29:18Hobsontnx PaulJam
16:30:22 Quit fml ("CGI:IRC (EOF)")
16:41:01Hobsonactually the main goal was to find a way to submit my plays to lastfm without buying a new device. but I think I'll give it up
16:43:00 Join Thundercloud [0] (n=thunderc@
16:43:11 Quit tchan (SendQ exceeded)
16:44:36 Join tchan [0] (n=tchan@lunar-linux/developer/tchan)
16:45:13preglowno thoughts on what "study mode" should do when at track boundaries?
16:46:10linuxstbIf anyone is interested, an Archos Player -
16:47:36linuxstbpreglow: Intuitively, I would expect it to skip into the next track (n-secs_remaining_in_current_track) seconds... But maybe my intuition is off ;)
16:47:49PaulJami think when skipping forward and reaching a track boundary it should start the next track from the beginning. i'm not sure what should happen when skipping back thoug.
16:49:00preglowright now it seems to just ignore the key press
16:49:03preglowwhich is damned weird, if you ask me
16:49:18preglowbut i just decided i'm not going to change that
16:49:23preglowlet people have their weird-ass feature
16:50:51preglowif in "study mode", you are apparently left without the means to change track from within the wps
16:52:30JdGordonuse the dirskip to override study mode
16:52:36 Join mcuelenaere [0] (n=mcuelena@rockbox/developer/mcuelenaere)
16:55:28 Join DerPapst [0] (
16:55:40 Nick JdGordon is now known as JdGordon|zzz (n=jonno@rockbox/developer/JdGordon)
16:56:33 Join Genre9mp3 [0] (n=yngwiejo@rockbox/contributor/Genre9mp3)
16:56:43PaulJami would say if the study mode reaches a track boundary it should act like a normal skip (but i don't use that feature anyway)
16:59:39 Quit VoltageX (Read error: 110 (Connection timed out))
17:00:40Hobsonsee ya guys
17:01:01 Quit Hobson ("scemo chi legge")
17:01:28preglowPaulJam: so would i, but i'm not about to change it any more than just making the naming slightly less retarded
17:02:05DerPapstBagder: can you commit ? a change to the irc reader css.
17:06:24 Quit CyBergRind|w (Read error: 110 (Connection timed out))
17:10:26 Join [CBR]Unspoken|w [0] (n=cbr@
17:11:28***Saving seen data "./dancer.seen"
17:15:41 Quit Rob2223 (Read error: 104 (Connection reset by peer))
17:16:23 Join Rob2222 [0] (
17:20:42 Quit Rob2222 (Read error: 104 (Connection reset by peer))
17:21:12 Join shotofadds [0] (n=rob@rockbox/developer/shotofadds)
17:22:26 Join Rob2222 [0] (
17:22:37shotofaddsCould somebody with a Gigabeat F and/or E200 try FS #9178 and check I haven't inadvertantly broken MpegPlayer playback?
17:23:58 Quit DerPapst (Read error: 110 (Connection timed out))
17:24:34 Join sarixe [0] (
17:26:14 Quit Rob2222 (Read error: 104 (Connection reset by peer))
17:26:14 Join madspin [0] (
17:27:04 Join Rob2222 [0] (
17:30:23 Quit Thundercloud (Remote closed the connection)
17:33:09shotofaddsnot bad.. the d2 benchmarks the elephant's dream mpeg (widescreen) at 47fps.
17:40:22linuxstbshotofadds: Nice... And that's without the second CPU and any specific optimisations?
17:40:48linuxstbSeems like more complex video formats are more than possible...
17:40:52preglowshotofadds: figured out the iram problem?
17:41:01preglowlinuxstb: well, retailos does mpeg4, so...
17:41:21shotofaddsyeah, that's SVN plus a small tweak to draw the YUV data the right way round. who needs hardware acceleration :g
17:41:26 Join goddel [0] (
17:41:46shotofaddspreglow: no. I'll look into that once we get the NAND driver working .... haha
17:42:03goddelhi there, iv installed rockbox, but where do i actually 'copy' the files. i mean what folder do i put them under ?
17:42:12 Join Rob2223 [0] (
17:42:29linuxstbgoddel: Any folder you like... Most people create a top-level "music" folder, and then arrange things by artist/album under there.
17:43:09CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
17:43:09*shotofadds voted for the MPEG4 GSoC project...
17:43:18goddellinuxstb, ahh ok. shall i put my MUSCI folder under the '.rockbox' folder ?shall
17:43:35linuxstbgoddel: I meant anywhere apart from .rockbox ;)
17:43:48*linuxstb too
17:43:58goddelok no worries. thanks a abunch
17:44:20goddelflac support on this is brilliant
17:44:54 Join desowin [0] (n=desowin@atheme/member/desowin)
17:45:11*linuxstb bows on behalf of the 200+ contributors to Rockbox
17:45:30shotofaddslinuxstb: did you see my last commit? there's now a (non-functional) TCC77x NAND driver if you fancy playing with it...
17:46:02shotofaddsI can't really investigate too much, I can only read half the screen on this M240 :/
17:46:03linuxstbshotofadds: Yes, but I've no time - I've been doing real work all weekend... :(
17:46:13shotofaddspah, where's the fun in that?
17:46:32 Quit miepchen^schlaf ()
17:53:04 Join perrikwp [0] (i=4aa794a0@gateway/web/ajax/
17:53:49 Join bertrik_ [0] (
17:55:36 Nick bertrik_ is now known as bertrik (
17:55:52 Part domonoky
17:57:46 Quit Rob2222 (Read error: 110 (Connection timed out))
17:59:05 Quit goddel ()
18:01:32 Join AndyI [0] (n=pasha_in@
18:09:36shotofaddsIf I update plugin.h so that lcd_yuv_set_options() is also included for d2, I need to update PLUGIN_MIN_API_VERSION, right? (and sort in the two 'waiting' functions?)
18:10:19linuxstbshotofadds: I don't think you need to if your change only affects "unreleased" targets such as the D2.
18:11:05 Join ender` [0] (
18:12:22shotofaddsyeah, that'd make sense. my patch in S#9178 forgot to update it anyway
18:12:34 Quit AndyIL (Read error: 110 (Connection timed out))
18:13:36 Quit bertrik (Remote closed the connection)
18:17:08 Join bertrik [0] (
18:20:10 Quit Lynx_ (" HydraIRC -> <- IRC has never been so good")
18:20:54 Join AndyIL [0] (n=pasha_in@
18:25:23 Join saratoga [0] (n=41becb3b@gateway/web/cgi-irc/
18:26:27 Join MarcGuay [0] (
18:27:22MarcGuaypreglow: There's a keypress to skip tracks in Study Mode. I think it's in gwps.c. I added the keypresses to the manual if they help you find it.
18:31:28 Quit AndyI (Read error: 110 (Connection timed out))
18:33:10 Join miepchen^schlaf [0] (
18:36:10wpyhgevaerts: are you the author of the usb stack?
18:36:41mcuelenaerewpyh: yes he is, but I think he's currently unavailable
18:38:01*MarcGuay helped edit the love scene at the end of chapter two.
18:42:48MarcGuayIs it possible for Rockbox to see files that were added using MTP?;topicseen#msg129792 is confusing me.
18:44:12wpyhbtw I think I'm hitting a bug here
18:44:27wpyhusing RB's usb stack, I sometimes can't get into disk mode
18:45:16 Join Zarggg [0] (
18:45:29preglowi'm starting to think "skip length" isn't a very good setting name either
18:45:38preglowbut i really can't think of any better ones
18:46:58stripwaxwpyh - as in menu+right on reboot?
18:47:10stripwaxRB's usb stack shouldn't be active at boot rime
18:47:16wpyhstripwax: hm?
18:47:35wpyhno, I mean I'm using RB's usb stack now
18:47:43stripwaxwhat is "disk mode" here?
18:47:47wpyhand RB doesn't get into disk mode when I plug the cable here
18:47:51wpyher... UMS mode :)
18:49:13stripwaxyou should get the USB logo on the rockbox device screen - and that doesn't happen?
18:49:22wpyhyep, it doesn't happen
18:49:30wpyhall I get is the battery starts charging
18:50:26stripwaxand "sometimes" means that, occasionally it does work for you, and when it doesn't you only get charging?
18:50:46stripwaxanything in dmesg?
18:51:31MarcGuaypreglow: Skip lock? Safe Skip?
18:51:43MarcGuayStudy Incrementalism?
18:57:31preglowMarcGuay: the functionality really doesn' fit with any of them
18:57:41preglowit's not a lock, it's not safe, it has nothing to do with studying
18:59:02MarcGuayMaybe re-reading the original task would help? I think it was mostly wanted by blind users who would lose their spot accidentally while listeneing to audiobooks... Although I guess anyone would find it a pain to get back to where they were...
18:59:23stripwaxwpyh - does it recognise it after a while or do you need to unplug+replug it to get it to recognise
18:59:44wpyhwell, I have to reboot for it to be recognized
18:59:56 Quit bertrik (Read error: 104 (Connection reset by peer))
19:01:43MarcGuaypreglow: If it's only useful for books maybe it could be worked into bookmarking somehow. Have a bookmark saved at regular intervals so that if the position is lost you can return to it...
19:01:58Nico_PMarcGuay: hi! did you know about FS #9137? It seems very similar to FS #8194, which you closed recently
19:02:26stripwaxwpyh - looks like one for gevaerts.. :( at a guess rb isn't identifying itself correctly/at all to the OS but can't think why
19:02:41stripwaxare you connecting via a hub? could be a signal quality thing
19:02:50wpyhno, I'm connecting straight to the port
19:03:03wpyhI think maybe one of the stack's state machines is broken
19:03:05MarcGuayNico_P: I took a look at it but didn't have the patience for the lenghty repro. recipe. :) I closed the other one simply because I tried hard to reproduce it and couldn't, along with the fact that only one person ever came across it and even they couldn't reproduce it reliably.
19:03:32wpyhI guess RB thinks it shouldn't export the storage to the host
19:03:35MarcGuayMaybe I was being hasty.
19:03:36 Join Ianto [0] (n=7960ef31@gateway/web/cgi-irc/
19:04:05Nico_PMarcGuay: PaulJam seems to be able to reliably reproduce FS #9137 so yeah I guess closing FS #8194 was the way to go
19:04:23Nico_PI plan on taking a good look at FS #9137 tomorrow
19:04:45MarcGuayNico_P: Nice. Good luck!
19:05:52Iantohello, can anyone help me?
19:06:04*MarcGuay throws a life-preserver.
19:06:49stripwaxIanto- just ask the question and we'll try and answer it. you don't need to ask to ask
19:07:51MarcGuayWhat about asking if you need to ask about asking before asking a question?
19:09:59Iantosorry i wasnt able to see anyone talking just now, maybe it takes time to refresh thr webchat
19:10:17Iantoanyway maybe i can log in thr mirc, whats this server?
19:11:23stripwaxIanto - you can look at the rockbox irc logs at
19:11:27 Join fml [0] (n=4fd3f80b@
19:11:32***Saving seen data "./dancer.seen"
19:11:34stripwaxWhat is your question? Just because we're not talking doesn't mean we're not here..
19:11:54fmlpreglow: jump length? hop length?
19:12:18fmlNico_P: hey. Have you seen my comment from yesterday?
19:12:41Nico_Pfml: I read it some time ago... could you remind me what you said?
19:13:09fmlNico_P: you forgot one case where the limits are exceeded. In viewports.
19:13:09 Join domonoky [0] (n=Domonoky@rockbox/developer/domonoky)
19:14:02Nico_Pah yes... do you know where it is in the source ?
19:14:30fmlNico_P: no, but I'd search for 'MAX_VIEWPORT'
19:15:16fmlNico_P: here: wps_data->num_viewports >= WPS_MAX_VIEWPORTS
19:15:30 Quit ompaul (Client Quit)
19:15:39 Quit Ianto ("CGI:IRC (Ping timeout)")
19:16:35 Quit saratoga ("CGI:IRC (EOF)")
19:18:56 Join Ianto [0] (n=Ianto@
19:21:25Iantohi, just started using some apps in my iphone a week ago
19:21:56Iantoand i saw rockbox in my searches, forums says that rockbox is pretty good
19:22:32Iantoive read loads of docu's but i still dont have any clue how to install
19:23:21domonokyrockbox doesnt work on a iphone ! see front page for supported players..
19:23:46Iantoso thats why i dont see a manual for it whoo
19:24:32 Join Viaken [0] (n=viaken@
19:24:36Iantothis ilounge thing, they were talking about themes in iphone and rockbox
19:24:57stripwaxIanto - rockbox doesn't run on iphones
19:25:23Iantoalright, thanks
19:26:37ViakenI hadn't updated my e260 in a couple months, and WOW, it's like lightning now. lol Everything seems much faster. Question, though. What's the point of md5summing everything? I was looking through the apps, struck me as odd. Does it help with some type of indexing or something? Or is it there just because?
19:28:16*wpyh updates every couple days
19:30:22scorcheViaken: it primarily helps to make sure that the file that you downloaded is the same as what is on the site
19:31:38 Quit vort3x (Read error: 54 (Connection reset by peer))
19:32:01 Nick m0f0x_ is now known as m0f0x (
19:32:11 Join vort3x [0] (n=vortex@unaffiliated/dfa001)
19:35:48 Join CaptainKewl [0] (
19:40:24Viakenscorche: Ah, ok. :)
19:48:45MarcGuaypreglow: Head vs. Wall?
19:48:55preglowif i add/deprecate anything in english.lang, i don't need to do anything to the other lang files, no?
19:49:06preglowMarcGuay: no, just general ignorance :)
19:49:37*MarcGuay sends out a bat-signal to rasher
19:53:05MarcGuaypreglow: If there's a translation of the thing you're adding, wouldn't it also have to be added/removed?
19:54:04preglowMarcGuay: no, i think that's the responsibility of the translater
19:54:23preglowif anyone wants to look for silly mistakes:
19:54:32preglowseems to work here
19:55:57 Quit einhirn_ (Read error: 104 (Connection reset by peer))
19:56:10MarcGuayI'm trying to remember why Party Mode's track locking wasn't good enough for studiers...
19:56:34MarcGuayI guess it does a little too much.
19:57:25wpyhmaybe the name is not suitable for studying ;)
19:59:35 Join bertrik [0] (
20:03:19 Join MTee [0] (n=mtarek16@
20:04:21 Join bluebrother [0] (n=dom@rockbox/staff/bluebrother)
20:05:13Genre9mp3MarcGuay: Party mode's concept is completely different... it assures unstoppable music playback by disabling pause/stop and skips
20:05:46MarcGuayGenre9mp3: Yep, gotcha. Reading over the original task now to remember what it's purpose and function are... :)
20:08:21MarcGuayIt confusing because it's talked about from two angles: 1) Protect people from themselves by making sure no unintentional skips happen and the place is lost and 2) Allow people to skim or browse through long files.
20:09:14MarcGuayI suppose Skip Length is generic enough to cover any reason to use it...
20:10:23MarcGuay#2 can be achieved by FF/REW, though. I guess folks don't want to hold the button for so long? But the speed of that can be adjusted to their taste...
20:11:39MarcGuayLeaving #1 sitting all alone and going by the patch's original name, Track Lock.
20:12:03MarcGuayPeople seem to like it a lot, though, from reading the flyspray comments.
20:14:58MarcGuayAnd #1 has also been discussed in other places, see
20:18:06MarcGuayHrm, only "kinda" on that last one.
20:20:24LloreanMarcGuay: In terms of #2. Take a 4 hour file that you need to be at a specific spot 2 hours into, where you have a segment ~5 minutes long that you may need to listen to more than once (language lessons?)
20:20:51LloreanGetting to that 2 hour spot, then fine-seeking to the beginning of the chapter, then repeating it is much easier with "Study mode" like functionality.
20:21:00MarcGuayLlorean: It would be really handy in that case.
20:21:12LloreanI assume that's why it had the name "Study" mode in the first place.
20:21:36MarcGuayCan you fine-seek with study mode on?
20:21:49LloreanAll it does is change "track skip" to be a coarse seek.
20:22:12LloreanAnd, you *should* still be able to track skip by coarse seeking past the end of the song, though I haven't played with our version of it to see if this is possible.
20:22:27LloreanWell, by "all it does" I mean "all it should do"
20:23:01MarcGuayIt has a track-skip key combo as well.
20:23:37LloreanThat seems odd.
20:23:43LloreanWhere do you get keys for that?
20:24:10Llorean buttons
20:24:11MarcGuayMakes sense, I can't ever see myself using it though (not an audiobook/podcast) listener. The keys are in gwps.c, it's like the short-dir/long-dir thing. Read the manual.. :)
20:26:01MarcGuayActually it's even a weird variation on the nextdir/prevdir button presses.
20:26:12LloreanPersonally I'd drop that.
20:26:22LloreanIt seems a little likely to be triggered on accident.
20:26:51LloreanIf you need to be 4 minutes back and skip is on 5 minutes, you're gonna press left, long right.
20:26:54LloreanAnd skip tracks.
20:27:23MarcGuayYou'd prefer "tabbing" through to the next track?
20:27:43LloreanSure, if I had the mode on.
20:28:05MarcGuayIt could use the regular skip-dir combos instead. Very hard to do that accidentally.
20:28:09LloreanIt's not *that* slow, and if I'm using the mode it's most likely I want to stay within one track anyway.
20:28:18MarcGuayThat's true.
20:28:33LloreanAnd if I need to skip multiple tracks, I should just use the playlist viewer or select the new track.
20:28:55 Join moos [0] (
20:28:55LloreanPersonally I don't like the skip-dir combos either.
20:28:55MarcGuayAlso very crappy, at least on the e200. The stop button is part of the combo.
20:29:03LloreanI've triggered them by accident a few times.
20:29:44MarcGuayI like the short-long myself. But it might not be as appealing if the combo one was changed to something that didn't stop the music half the time. :)
20:29:50LloreanWhen I've forgotten where I was in an audiobook, I'll resume the file I thought I was in, notice it wasn't but I'm pretty sure the next one is, skip to it, try to seek a portion in, and end up in an entirely different book by the author
20:30:06LloreanHow often do you dir-skip?
20:30:31MarcGuayLlorean: Honestly I usually only do it while testing bugs.
20:30:51LloreanI kinda think the dir skip function can go without little-to-no outcry.
20:31:08LloreanIt also doesn't really serve a function with playlist files, or with the database, anyway
20:31:13 Quit desowin ("KVIrc 4.0.0 Insomnia")
20:31:21LloreanI'm not sure what they do, but it's either nothing or something unexpected to the user.
20:32:19MarcGuayI wouldn't trash it alltogether. It's useful for random dir advance if you don't like the random album that it's chosen.
20:32:44 Quit EspeonEefi ("さよăȘら")
20:33:09 Part Viaken ("ptwang!")
20:33:42LloreanHonestly, I'm not sure I have sympathy for that point of view. You pick the random dir list in advance, so if you get one you don't like it's your own fault. :-P
20:35:50LloreanBut really, the keypress (and next folder functionality in general) seems kinda more trouble than its worth.
20:36:11LloreanIf our playlisting tools were good enough, it wouldn't be a problem.
20:37:29LloreanFor example, if you could put folders in a playlist, and when it got to that entry it would play all the songs in the folder, you could create a playlist of folders, and if you wanted to play the folders in random order, but the songs within sequentially, you could "Insert Shuffled" that playlist of folders, and you'd get shuffled folders, but since shuffle was off, sequential songs.
20:38:04LloreanAs well, insert-lasting a folder with shuffle on would put the folder last, but when played, it should logically play shuffled.
20:38:10LloreanAddressing that feature request in the forum.
20:38:26LloreanAnd negating a need for "folder advance" entirely, which iiuc would notable clean up some playback weirdness.
20:38:44MarcGuaySounds like a big project. :)
20:40:03 Nick MarcGuay is now known as MarcGuayAway (
20:40:07 Join lacrstech [0] (n=lacrstec@
20:40:42 Join dedebkgrn [0] (n=dedebkgr@
20:40:42LloreanThere are a few people in favour of changing how playlist-centric Rockbox is.
20:41:02LloreanWe disagree on some points (I like how it's just off to the side right now) but I do think the playlist system should be vastly more powerful
20:41:15LloreanEspecially since "folders as entries in the playlist" is supposedly within m3u spec.
20:50:02 Join BHSPitLappy [0] (n=BHSPitLa@unaffiliated/bhspitmonkey)
21:05:06wpyhI've almost finished the maze.c patch
21:05:28wpyhI touched a lot of things, and it's more like a revamp than a bugfix
21:05:33PaulJamLlorean: but even with the method you described it would be nice to have a buttoncombination to skip the current folder
21:05:42wpyhshould I just update FS #9184, or should I open a new task?
21:06:42 Join EspeonEefi [0] (i=espeonee@STRATTON-FIVE-TWENTY-FOUR.MIT.EDU)
21:07:05LloreanPaulJam: Yes, but that button combination happens to be possible to accidentally trigger in normal use. I think it's bad to have "short-long" style combinations if they're something you can press in normal usage.
21:07:10 Quit mcuelenaere (Read error: 113 (No route to host))
21:08:08stripwaxre folders-as-playlist-entries - how hard would that really be in fact? And a logical extension would be playlist-files-as-playlist-entries, rather than expanding inline into one playlist
21:08:25LloreanPaulJam: Not to mention if "Next Folder" were gone, there wouldn't be anything for the combination to do.
21:08:50Lloreanstripwax: Agreed that would be the logical extension. I have no idea how hard it would be to do though.
21:08:53PaulJamLlorean: ok, in that point i agree (i actually disabled the short long buttoncombo in my build because of that). but there is still the PLAY + LEFT/RIGHT butooncombo
21:09:46LloreanPaulJam: If there's no "Next Folder" functionality, there's nothing for that combo to do anyway...
21:10:03PaulJamLlorean: it could go to the next playlist entry
21:11:06 Join gregzx [0] (
21:11:34***Saving seen data "./dancer.seen"
21:14:06LloreanPaulJam: Actually, it'd make more sense to "go to the end of the current one" and let the natural behaviour go from there.
21:14:18LloreanThen "it" wouldn't need to know if there was a parent playlist.
21:14:49LloreanBut don't "Play+Left" and "Play+Right" serve as buttons for the A-B functionality?
21:15:40PaulJamLlorean: both, the idea is that it is unlikely that you want to skip a filder while in A-B repeat mode
21:16:06LloreanPersonally, I don't like keymappings that change depending on what feature is enabled.
21:16:43 Quit tecnmmag (K-lined)
21:18:18PaulJamis there a wikipage or something describing in detail how theis "folders in playlists" idea should work?
21:18:56 Quit dedebkgrn (K-lined)
21:18:57 Quit lacrstech (K-lined)
21:19:35wpyhI've created FS #9191 for the maze.c revamp patch
21:19:58LloreanAs far as I'm personally concerned, when a line in a playlist is a folder or playlist, it should play as if it had loaded that playlist, then return to the parent playlist when done.
21:20:02wpyhthe patch also fixes FS #9184, so please add that one the related tasks
21:20:03 Join Thundercloud [0] (
21:20:46LloreanAlso as far as I'm concerned, when a folder is loaded, only actual audio files should play, *not* other playlists (even if recursive insert is on)
21:20:53bertrikwpyh, ok, I'll take care of that
21:21:05wpyhbertrik: thanks :)
21:21:06LloreanIf recursive doesn't ignore other playlists, I imagine you could easily get stuck in an infinite loop.
21:22:31 Join BlakeJohnson861 [0] (
21:23:16 Quit BlakeJohnson86 (Read error: 104 (Connection reset by peer))
21:23:50PaulJami guess i could live with that concept (if it'll contain a "goto end of current playlist" button combination).
21:26:10 Join J [0] (
21:26:16LloreanPaulJam: I take it you use folder skip regularly?
21:26:38 Nick J is now known as Guest59816 (
21:27:04PaulJamyes occassionally(when i listen in random folder mode)
21:30:25 Join EspeonEfi [0] (i=espeonee@STRATTON-ONE-THIRTY-SIX.MIT.EDU)
21:30:29 Quit EspeonEefi (Read error: 110 (Connection timed out))
21:31:13 Quit culture (Read error: 60 (Operation timed out))
21:32:15 Join culture [0] (
21:33:56 Join amiconn [50] (n=jens@rockbox/developer/amiconn)
21:36:36 Join Lear [0] (i=chatzill@rockbox/developer/lear)
21:36:44 Quit MTee ()
21:43:17 Quit fml ("CGI:IRC (Ping timeout)")
21:45:09 Join PaulJam_ [0] (
21:49:11 Join PaulJam__ [0] (
21:50:56 Quit bluebrother ("leaving")
21:54:37 Part Ianto
21:55:04 Quit avis (Remote closed the connection)
22:03:11 Join webguest75 [0] (n=864c0352@gateway/web/cgi-irc/
22:03:35 Quit PaulJam (Read error: 110 (Connection timed out))
22:03:43 Quit webguest75 (Client Quit)
22:03:59 Join PaulJam [0] (
22:06:08 Join avis [0] (n=ident@pdpc/supporter/student/avis)
22:08:45 Quit gregzx (Read error: 110 (Connection timed out))
22:09:28 Quit PaulJam_ (Read error: 110 (Connection timed out))
22:10:26 Quit PaulJam__ (Read error: 110 (Connection timed out))
22:11:51 Join VoltageX [0] (
22:14:44 Join andrewbeveridge [0] (n=586f998c@gateway/web/cgi-irc/
22:16:16 Nick jfc is now known as jfc^3 (
22:16:25 Nick jfc^2 is now known as jfc (
22:17:51 Quit CaptainKewl (Read error: 110 (Connection timed out))
22:20:44andrewbeveridgehello, I just sucessfully installed the current Rockbox build on a 1st generation iPod Nano - it works great, I am really grateful to everyone involved as I can now use my iPod as I would expect an MP3 player to work. However I have a large music collection in WMA format (I hate proprietary formats generally but WMA has the best compression ratio of all losyy formats), and playback of *any* of these fil
22:21:17stripwaxandrewbeveridge - is there more to your message? it seems to have been truncated
22:21:34andrewbeveridgeah, i shall resend it on multiple lines, hold on
22:21:35PaulJamyoue message got cut after "and playback of *any* of these fi"
22:21:55stripwaxandrewbeveridge - at a guess, do your WMA files have DRM, and that is causing them to not play back?
22:22:12andrewbeveridgeplayback of *any* of these files is faulty, skipping every few seconds throughout the entire track. I was wondering if there is anything I can do to try and narrow down to find the cause of the problem. MP3 files play fine, as expected. Thanks :)
22:22:14stripwax(rockbox does not implement DRM of any sort)
22:22:20stripwaxah, forget that then ;)
22:22:27andrewbeveridgenone of the files are DRM protected, I am really anti - proprietary ;)
22:22:51stripwaxandrewbeveridge - what bitrate WMA files are you using?
22:23:00andrewbeveridgeall VBR 40 - 60 kbps
22:23:10PaulJamseems like a performance issue. do you have EQ or crossfeed enabled?
22:23:19andrewbeveridgeall disabled
22:25:29andrewbeveridgethe only thing i could think of when doing my own mental checks was that the ipod cpu might not be able to cope with wma decoding? is that even a sensible thought?
22:25:30stripwaxandrewbeveridge - did it work with prior builds of rockbox? there have been some WMAv1 fixes recently as well as a rework of the mdct code
22:25:39andrewbeveridgei havent tried any other build
22:25:49andrewbeveridgenot sure where to start
22:26:10stripwaxandrewbeveridge - see here, ipod nano is roughly equivalent to ipod video, is more than capable of decoding WMA.
22:26:25andrewbeveridgeah, ok thanks scrap my thought then
22:27:05stripwaxandrewbeveridge - try installing an older daily build of rockbox, say from 2008-07-04
22:27:06andrewbeveridgewhich build should i try?
22:27:11andrewbeveridgeok thanks
22:27:21andrewbeveridgei shall come back here after more experimentation :)
22:27:27 Quit Arathis ("Bye, bye")
22:28:20 Join ctaylorr [0] (
22:32:34 Quit merbanan (Remote closed the connection)
22:34:26 Join d1sturb [0] (
22:34:38 Quit PaulJam (Read error: 104 (Connection reset by peer))
22:36:50 Join pixelma [50] (i=pixelma@rockbox/staff/pixelma)
22:36:51 Join ctaylorr_ [0] (
22:47:58 Quit ctaylorr (Read error: 110 (Connection timed out))
22:53:26 Join PaulJam [0] (
22:54:50 Quit VoltageX (Read error: 110 (Connection timed out))
22:55:05 Join petur [50] (n=petur@rockbox/developer/petur)
22:58:10 Quit bertrik (Remote closed the connection)
23:06:59 Quit domonoky (Read error: 104 (Connection reset by peer))
23:11:36***Saving seen data "./dancer.seen"
23:13:17 Quit bughunter2 ("bye")
23:27:28preglowLlorean: i posted a patch for the study mode renaming if you want to test
23:28:33preglowandrewbeveridge: could you upload a faulty file somewhere?
23:29:23 Join desowin [0] (n=desowin@atheme/member/desowin)
23:32:47scorcheBagder: how are the midterms coming?...since you can see them and all
23:34:08BagderLinusN and JdGordon haven't submitted theirs
23:34:44scorchedo we have the whip ready?
23:34:53scorchealthough LinusN's should be easy, sadly
23:34:55BagderI'll mail a virtual one
23:35:03 Quit ender` (" On the other hand, you have different fingers.")
23:35:04 Nick MarcGuayAway is now known as MarcGuay (
23:35:26 Quit markun (Read error: 104 (Connection reset by peer))
23:36:27andrewbeveridgei'm just trying different files
23:36:35andrewbeveridgeencoded with different programs/settings etc
23:36:38 Join markun [0] (
23:36:40andrewbeveridgei shall upload when done :)
23:45:33 Join safetydan [0] (n=evertond@rockbox/developer/safetydan)
23:45:41preglowandrewbeveridge: goodie
23:46:24 Quit Thundercloud (Remote closed the connection)
23:47:46LloreanMarcGuay: Why removing "Draft" from all pages of the manual except the first? I think if it's *just* there it'll probably get missed by normal readers
23:48:14LloreanMarcGuay: And, is the ManualTodo page mostly up to date?
23:49:42MarcGuayLlorean: I added that after watching the Devcon discussion and it seemed to have been agreed upon. I just think it's rather pointless given that the entire project is a draft version of sorts.
23:50:20MarcGuayThat page is not up to date. Some things are on the FixMes page but not there, and items linger in the tracker.
23:51:35LloreanMarcGuay: Could I ask a big favour? Could we try to have one location someone can look as see a "known issues with the manual"? No rush, but before the feature freeze.
23:52:00LloreanI'm not trying to put this on you or anything
23:52:21LloreanBut just... when/if you're working on the manual again, and spot something, or if you're reorganizing one of those pages, consolidate when you're already there perhaps
23:52:30SoapMarcGuay, what program do you use to edit the manual?
23:52:53Soap(or do you do it by hand / vi / etc.
23:53:19MarcGuaySoap: Using notepad++ and I turned off the tabs today!
23:53:30MarcGuay(Assuming that's where you're going ;) )
23:53:38Soapexactly where I was going, thanks.
23:54:30MarcGuayLlorean: I've been trying to do that. And if it's small enough fix it on the spot.
23:55:40LloreanMarcGuay: Okay. Thanks.
23:55:42 Join LinuxMafia [0] (
23:55:46LinuxMafiahi all
23:55:56 Quit davina (Remote closed the connection)
23:56:06LinuxMafiacan i install rockbox on ipod nano
23:57:04SoapFirst generation Nano (plastic front, stainless steel back) only.
23:57:27stripwaxLinuxMafia - the rockbox-supported devices are listed on the front page of
23:57:34 Quit Genre9mp3 (Read error: 104 (Connection reset by peer))
23:57:40 Join bertrik [0] (
23:58:04LloreanMarcGuay: I'd personally rather leave every page marked as Draft just so there's less room for confusion. Then again the HTML manuals aren't clear about this anyway.
23:58:16 Quit midgey ()
23:58:30LinuxMafiacause i have ipod nano i connect it to computer ,nothing comes up
23:58:39LinuxMafiait wont show it on my computer
23:58:49 Join LambdaCalculus37 [0] (i=1800dac6@gateway/web/ajax/

Previous day | Next day