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 2006-10-16

00:01:24SoapDST is a pox on humanity.
00:01:50*amiconn likes DST
00:02:02*JdGordon hates it..
00:02:24JdGordonthe difference is, sunset at the hieght of summer is 9pm here without DST, 10pm is just rediculous
00:02:33JdGordonin the uk its like 4pm isnt it? :D
00:03:40 Quit MarcoPolo ("Bye !")
00:05:23 Join acknabar [0] (
00:05:35 Quit acknabar (Read error: 54 (Connection reset by peer))
00:08:52 Quit t0mas ("Doei")
00:10:13 Quit lee-qid (Read error: 110 (Connection timed out))
00:17:34 Join MarcoPolo [0] (
00:20:26 Quit mirak (Read error: 110 (Connection timed out))
00:21:18 Quit ender` (" We live in an age when pizza gets to your home before the police. -- Jeff Marder")
00:23:47 Join mirak [0] (
00:24:49amiconnpixelma: ping...
00:26:32Soapthe problem I have with DST is I work outdoors, and it throws off my biological clock twice a year.
00:27:14 Quit pagefault ("Leaving")
00:27:22 Join pagefault [0] (i=pagefaul@pdpc/supporter/active/pagefault)
00:28:03 Part nave7693
00:31:33 Join mirfle [0] (i=52a680a5@gateway/web/cgi-irc/
00:31:48 Quit MarcoPolo ("Bye !")
00:33:36mirfleI just loaded the bootloader for the h120 available on the iriver flashing page, and now the original firmware won't boot.
00:37:28 Join FOAD_ [0] (
00:38:05 Join MarcoPolo [0] (
00:43:46 Join linuxstb [0] (i=3f82c736@gateway/web/cgi-irc/
00:47:51 Quit FOAD (Read error: 110 (Connection timed out))
00:48:04 Quit Id2ndR ("Parti")
00:48:50 Quit mirfle ("CGI:IRC (EOF)")
00:48:54 Join mirfle [0] (i=52a680a5@gateway/web/cgi-irc/
00:50:11mirfledoes anyone here know what might be the problem?
00:52:14linuxstbmirfle: Slasheri would probably be the person to ask - he wrote all the flashing code.
00:53:00linuxstbBut have you tried restoring the original firmware as described at the end of the IriverFlashing page?
00:53:18 Part telliott_
00:54:37mirfleI can load rockbox fine, it's just the OF that doesn't work. I wouldn't really care about that if it weren't for the few wma's I still have...
00:55:48mirflehow do I turn a .hex to a . bin?
00:55:51 Quit dan_a (Read error: 110 (Connection timed out))
01:00:42mirflethe wiki page should be changed to have the info on backing up the flash BEFORE using the iriver_flash plugin for the first time, not only when completely removing the OF..
01:01:03mirflethat would have saved me this trouble
01:04:10 Quit lodesi (Read error: 113 (No route to host))
01:08:10mirflehope somenone can help me in the forums....
01:08:12 Quit mirfle ("CGI:IRC")
01:09:54MadkissHow do I make screenshots out of rockbox?
01:11:17MadkissAnd even another question ... I loaded multiple files onto the iPod, and yet, the While Playing screen still looks the same, there are the same icons in the upper bar etc. pp.
01:11:20MadkissHow do I change that?
01:12:33Genre9mp3Madkiss: How do you expect the WPS to change?
01:13:07MadkissI want it to look like Theme X Hibar
01:13:13Madkissbut all that changes is the background
01:14:18linuxstbWhen you first installed Rockbox, did you also install the font zip file?
01:14:43 Quit Rondom ("Ex-Chat")
01:15:09 Quit dpro (Read error: 110 (Connection timed out))
01:15:20Genre9mp3Madkiss: So you just try to load that theme, right?
01:15:33Madkissthe background changes, the font changes
01:15:35Madkissbut nothing else ..
01:16:00Genre9mp3Have you copied all the files that this wps needs to your player?
01:16:22MadkissWhat am I doing wrong when I see "tag-cache is not ready"?
01:16:23JdGordonamiconn: still awake?
01:18:36Genre9mp3Madkiss: Can't think of what's wrong then (about the WPS)
01:18:59pixelmaMadkiss: do you have a folder in .rockbox/wps whith the exact same name as the actual wps-file you are trying to load?
01:19:13JdGordonanyone have an archos recorder?
01:19:52Madkissdrwxrwxrwx 1 madkiss madkiss 4096 Oct 15 22:50 Theme X HighBar 1
01:19:52Madkiss-rwxrwxrwx 1 madkiss madkiss 1370 Sep 22 17:12 Theme X HighBar 1.wps
01:21:57linuxstbMadkiss: How did you install the wps? Did you unzip it to your computer, then copy the files, or unzip it directly to your ipod?
01:22:21MadkissI unzipped them directly
01:22:40JdGordoncan sometry try getting into th pitch screen from the wps on the recorder? on+up or down) and lemme know if it works?
01:25:36amiconnJdGordon: It doesn't work
01:25:40*Madkiss laughs
01:26:02JdGordonamiconn: does naything happen? or it just plain doesnt work?
01:26:16amiconn[23:53:50] <amiconn> Pitch screen doesn't appear on recorder
01:26:29amiconni.e. nothing happens
01:27:34JdGordonwell.. thats very odd... because there is nothing in the wps keymap which should stop it from working!
01:27:59JdGordonwas the combo always on+up/down?
01:29:03 Join Elric-Kun [0] (
01:29:13Elric-KunHello guys
01:29:17 Nick Elric-Kun is now known as L (
01:30:25linuxstbHAVE_PITCHSCREEN doesn't seem to be defined for the recorder...
01:30:48JdGordon... well that would explain it :p
01:33:43JdGordon23.22.51 # <amiconn> Button actions was meant to solve problems, not create them :/ <- HA! it wasnt the actions after all and obviously hasnt worked for a while :D
01:34:17amiconnYes - in this case it wasn't
01:34:44JdGordonjust making sure botht he v1 and v2 both compile ok
01:35:00JdGordonit was only defined for the fmrecorder..
01:37:21***Saving seen data "./dancer.seen"
01:38:59 Join lukaswayne9 [0] (
01:39:46 Join Platyna [0] (
01:40:48PlatynaI have iPod Video 60 GB. Is there any way to apply skins to main menu and to change font of main menu? Because it is very, very small.
01:41:30 Join waszi [0] (
01:44:38JdGordonamiconn: ah, found the problem... i missed the 2 configs when i moved the #define from the screens file to the config.. doh!
01:44:42JdGordonall fixed now
01:45:09JdGordon2 months ago :p
01:51:13MadkissOkay. THe only problem left is that no matter what theme I use, it all looks completely foo.
01:51:26JdGordondid u download the fonts?
01:53:08 Join dan_a [0] (n=dan-mirc@
01:54:12 Join Davide-NYC [0] (
01:54:21Genre9mp3Can you be more descriptive? What exactly seems wrong?
01:55:01 Quit pixelma (" reboot")
01:55:26Davide-NYCHello all! QQ: my nephew wants to check out Rockbox, his PCs HD crashed recently and he want to get all of the music off of his iPod before taking the RB plunge.
01:55:57Davide-NYCI do not have an iPod but have used Ephod in the past to remove files (backup) from other's ipods.
01:56:04Davide-NYCIs this the best solution currently?
01:57:32 Quit L (Read error: 113 (No route to host))
02:00:08Davide-NYCdang, no one wants to chime in on this one?
02:00:22MadkissGenre9mp3: The Album-icon is not displayed, instead, I see fragments of the file listing
02:01:03SoapDavide-NYC - Sharepod is easy.
02:01:23 Quit preglow (
02:01:28Davide-NYCthanks soap, you are a scholar and gentleperson
02:01:37NJoinpreglow [0] (n=thomjoha@rockbox/developer/preglow)
02:02:17SoapDavide-NYC - I'm familiar with ephpod, but never used it. sharepod allows you to create directories/filenames based on the tags just like a good tagger.
02:02:33Davide-NYCthat is much better than ephosd
02:02:35Genre9mp3Madkiss: Album-art is not supported by rockbox (at least yet)... you need to use the Album art patch for that
02:03:54Soapif he was going to dual-boot Rockbox/Apple and use tagcache, sharepod is what I would use. Sub-1MB, and can reside on the ipod gives you /almost/ the same functionality as the drag-and-drop joy that is a RockBoxed Apple-free Ipod
02:03:54 Quit NickDe (Read error: 54 (Connection reset by peer))
02:04:10 Join NickDe [0] (
02:04:16 Join pixelma [0] (i=pixelma@
02:06:14Davide-NYCSoap: thanks again. I'll be cak Tuesday.
02:06:27 Quit Davide-NYC ("Chatzilla 0.9.75 [Firefox]")
02:06:29 Quit MarcoPolo (Remote closed the connection)
02:10:18MadkissThis "Tagcache is not ready"-stuff is getting me, dammit
02:11:54PlatynaHow to make the font bigger, please?
02:12:39Madkisshow do I fix "tagcache is not ready"?
02:14:21linuxstbPlatyna: Somewhere in the main menu (I forget exactly where), there is a "Browse fonts" option that lets you choose different fonts - assuming you downloaded and unzipped the fonts zip file when you first installed Rockbox.
02:15:09Genre9mp3Madkiss: have you initialized the Tagcache?
02:15:09linuxstbMadkiss: Have you read the manual section on tagcache and followed the instructions there?
02:15:23 Join lukaswayne9_ [0] (
02:15:53MadkissI initialized it manually
02:16:55Genre9mp3Madkiss: Have you rebooted your player after that?
02:23:41Platynalinuxstb: Well, I think it is not font problem but resolution problem...
02:24:10PlatynaHow can I change resolution? All letters are very small and that "desktop" not takes whole screen space.
02:24:43linuxstbYou choose a different sized font.
02:25:22PlatynaIf I do theme gets broken.
02:25:52PlatynaAnd I have 2 mm black "borders" by each side of screen.
02:25:57linuxstbThen you will need to use a different theme - one based on a bigger font.
02:26:08PlatynaSee above.
02:26:42PlatynaI have screen alot smaller than iPod's display.
02:27:44PlatynaAnd I don't know what do do about it.
02:28:33pixelmahave you tried another bigger font as linuxstb suggested and seen how it looks like....
02:28:40linuxstbI don't understand what you mean. The small black borders are normal - at least, I get them on my 5g.
02:29:11Platynapixelma: It is NOT a font issue.
02:29:19PlatynaWhat has font to screen resolution?
02:29:39Platynalinuxstb: But on default firmware I don't have them.
02:29:50PlatynaI mean I have them, but they are almost invisible.
02:30:48 Join Yono [0] (
02:31:01 Part jhMikeS
02:31:45Genre9mp3Platyna: There isn't a screen resolution option
02:32:35PlatynaSo, it is some fixed value in the code?
02:32:42 Join Mikachu [0] (
02:32:59pixelmaand the "desktop" not using the whole screen might result fron the fact that the lists/menus just aren't long enough to take the whole screen
02:33:10Mikachuit seems that rockbox crashes on boot if prio scheduling is disabled, at least on my ipod nano
02:33:10pixelmawith a small font
02:33:16Mikachuthis also includes the bootloader
02:33:22Mikachuwhere prio scheduling is disabled by default
02:33:41Mikachuis this known?
02:33:50Mikachups hi guys
02:34:11 Quit anathema (Read error: 113 (No route to host))
02:34:25Platynapixelma: It is not that. It has a white background right?
02:34:40PlatynaThat white background not takes as much screen as it used to do.
02:34:44PlatynaAnd that's the problem.
02:34:51Mikachulinuxstb: do you know?
02:34:51 Join Febs [0] (
02:35:26linuxstbMikachu: No idea - I haven't touched the new scheduling code.
02:35:36Mikachui'm wondering why the old code doesn't work
02:35:49Genre9mp3Platyna: It used to do? with Rockbox that is or you mean with retailOS?
02:36:20linuxstbPlatyna: I'm just reinstalling the Apple firmware on my 5g now, to compare with Rockbox. I never use it, so wouldn't have noticed if the display looks different...
02:37:06linuxstbBut in the emergency disk mode (the "Do not disconnect" screen), which is also Apple's software, the white background takes up the same area as it does in Rockbox.
02:37:07Mikachumy friends brother has an ipod video, i noticed the firmware now lets you set brightness
02:37:16amiconnMikachu: Even without priority scheduling it doesn't run the old code
02:37:27Mikachuamiconn: okay, i haven't actually looked at it
02:37:35amiconn(but it does run the same code as on archos, where it does work)
02:37:38Mikachubut i remember it worked without prio scheduling at least some time after it was merged
02:37:54Mikachubut now i haven't updated in a few weeks
02:40:48 Quit Criamos ("( :: NoNameScript 4.03 :: )")
02:41:21 Quit EspeonEefi (Read error: 145 (Connection timed out))
02:42:05Mikachuwell, the next time someone needs to compile the ipod nano bootloader, they'll have to figure it out
02:42:21PlatynaGenre9mp3: On that default firmware I had nice screen resolution.
02:42:29linuxstbPlatyna: Apple's firmware seems to use an identical area of the LCD to Rockbox - I can't see any difference. They both have a small black border.
02:43:19*Platyna scratches her head.
02:43:24 Join ton [0] (
02:43:28PlatynaI could have swore It were bigger.
02:43:32pixelmaPlatyna: what white background are you referring to... (is it in a wps)? The default theme should use light blue as background colour (I just want to understand what you mean)
02:43:34PlatynaOr maybe it is illusion...
02:43:50tonanyone working on putting rock box on creative zen vision m
02:44:03Platynapixelma: Ah apropos theme, I have problem, I found a nice one, but when I choose bigger font it gets broken.
02:44:49linuxstbton: This forum thread is the only activity:
02:45:41tonthey keep talking about it but r they?
02:46:39 Join lsajsf [0] (n=aus@
02:46:48Genre9mp3Platyna: Themes use specific fonts. Try using a theme that uses a bigger font
02:46:57lsajsfdoes rockbox work with the new ipod 5g?
02:47:14lsajsf5.5g i guess?
02:49:07PlatynaGenre9mp3: Maybe you could just point me to some theme building doc?
02:49:10linuxstblsajsf: Not yet.
02:49:29PlatynaBecause I like this theme. And there is plenty of space around elements, so maybe I could modify it.
02:49:37lsajsfok thanks
02:49:42 Quit lsajsf (Client Quit)
02:51:11PlatynaThank you.
02:51:16 Join alphakiller [0] (n=bits@
02:51:46 Quit TeaSeaLancs (Remote closed the connection)
02:52:15PlatynaHow to put cover images?
02:53:43Genre9mp3you mean album covers?
02:55:23Genre9mp3This is not officialy supported on Rockbox. You need the album art patch for that
02:55:41PlatynaMhm...Where do I obtain it?
02:56:26Genre9mp3In the tracker... but you have to apply it to the source and then compile your own build
02:57:13linuxstbton: I haven't read that thread, but I would guess not - lots of people talk about porting Rockbox, but very few actually do. And I don't think any existing Rockbox developers own a Zen, or plan to buy one.
02:59:31*Genre9mp3 wonders how many Zen users would like to have Rockbox
03:02:40linuxstbIt's probably the most asked-about port that doesn't exist.
03:03:35Genre9mp3Yes.. I guess it's the most popular player that Rockbox does not support
03:04:08Genre9mp3Maybe until the MS Zune arrives...
03:07:18coobwhat does the zen even run on?
03:07:35linuxstbYou mean the CPU?
03:07:58 Join barrywardell [0] (
03:08:33linuxstbIt's a TMS320 - an ARM9 core plus a DSP.
03:09:07Genre9mp3linuxstb: Zen Vision M that is?
03:09:16*amiconn caused some yellow...
03:09:46linuxstbGenre9mp3: Yes, IIRC.
03:10:51 Join Kalthare [0] (
03:11:58amiconn...and even red :(
03:12:17amiconnImho we need a decision concerning the iFP port and plugins
03:12:34 Quit dan_a (Read error: 110 (Connection timed out))
03:15:08KalthareI'm preparing a new patch for the 5.5. AFAIK, it fixes everything except the crash.
03:15:55linuxstbDid you get the FAT test code working?
03:16:30KalthareYeah, it just needed a few defines and a couple of source files added to the Makefile.
03:16:52linuxstbCould you post a separate patch for that? It would be nice to commit it.
03:17:52KalthareWell, one thing I didn't fix is the reams of warnings about signedness mismatches. Is that alright?
03:18:40 Part alphakiller
03:19:05KalthareAlso, I defined the build date by hand, which I'm pretty sure isn't the right way to do it.
03:19:16 Join blueworm [0] (
03:26:49KalthareStill want the patch?
03:26:54 Quit Yono ("Leaving")
03:30:20linuxstbKalthare: Sure, if you post what you've got, I'll either commit it as it is, or try to fix the problems.
03:31:41godzirraKalthare: how's it going?
03:35:26Kaltharegodzirra: I'm pretty sure I've fixed the FAT code.
03:35:44Kaltharelinuxstb: Okay, gimme a bit to put that together.
03:37:09myzarKalthare, then it will work on an 80?!
03:37:25***Saving seen data "./dancer.seen"
03:38:24Kaltharemyzar: Unfortunately, the 80 seems to have more than the FAT code wrong with it. And even on my 30 it crashes during music playback, but I'm confident at this point that the FAT code isn't responsible.
03:39:44godzirraKalthare: wow, thats awesome.
03:39:49godzirraSo you've got a working 5.5g solution?
03:40:17Kaltharegodzirra: Well, it works if you wanna play Doom. It's still not so good at music. :/
03:40:32 Join webguest30 [0] (i=47ff9014@gateway/web/cgi-irc/
03:41:20 Join midkay [0] (n=midkay@rockbox/developer/midkay)
03:41:24bluewormKalthare: I would'nt mind testing it on 80Gb ;)
03:42:27myzari'd do it, but i want to see more progress first
03:42:32myzari have an 80 =p
03:42:45 Quit pagefault (Read error: 60 (Operation timed out))
03:42:59myzarKalthare, you don't think there's a way to get in touch with an engineer from toshiba
03:43:06myzarand ask them the differences between the 30 gigs
03:43:13myzarand the new drives they made for the 80s
03:43:16myzardo you?
03:43:32myzarMK8007GAH is the model, i'm reading up on it online
03:43:41KalthareI'll put up the new patch as soon as I'm done putting it together. I'll say, tho, that I haven't done anything that's likely to fix the problem on the 80.
03:43:54Kaltharemyzar: I'm not exactly an expert on ATA programming.
03:44:00myzaroh, me neither
03:44:18myzari'm trying to figure out how to assist the coders
03:44:25myzarbut i'm afraid i wouldn't use C++
03:44:27*blueworm having happy thoughts
03:45:08myzarcould it be the perpendicular recording, Kalthare?
03:45:20Kaltharelinuxstb: , or should I post it to flyspray?
03:45:27Kaltharemyzar: I don't even know what that is.
03:45:44PaulJamaren't the 30 and 60 gb drives perpendicular too?
03:45:46linuxstbKalthare: That's fine there, thanks.
03:45:52myzardunno, PaulJam
03:45:55godzirraKalthare: I'd pr efer music to doom ;)
03:45:56myzari'm trying to figure that out
03:46:01godzirraKalthare: why doesnt the music work?
03:46:03Kaltharelinuxstb: Hope it helps.
03:46:17myzarKalthare, perpendicular recording is basically making hard-drives cheaper
03:46:19 Join pagefault [0] (i=pagefaul@pdpc/supporter/active/pagefault)
03:46:22myzarby putting 10 times more data on the same surface
03:46:30sneakumsthe drive in mine is (or claims it is) an MK8010GAH
03:46:42 Quit midkay ("Leaving")
03:46:44Kaltharegodzirra: I don't know. It freezes.
03:47:10myzarwhat, sneakums
03:47:12myzarfor your 80?
03:47:16godzirrawatch that
03:47:19godzirraits scary.
03:47:28sneakumsyes, the drive in my 80G claims to be an MK8010GAH
03:49:04myzarthat IS scary godzirra
03:49:11godzirramyzar: isnt it awesome though? :)
03:49:15godzirraGo go school house rock
03:50:00 Join midkay [0] (n=midkay@rockbox/developer/midkay)
03:50:24myzarit's scaring me
03:50:37myzarbut check, godzirra
03:50:43linuxstbDo we know where Rockbox fails on the 80GB - is it during ata_init() or later?
03:50:43myzarif the 30 and the 60 drives were perpendicular
03:50:55myzarsomebody tried to debug it
03:50:57myzarthey may be wrong
03:51:11sneakumsi tried to debug it
03:51:22sneakumssee my rockbox-dev post
03:51:43sneakumssummary: ATA_STATUS is set to STATUS_ERR during the attempt to read the MBR
03:52:27sneakumsit hangs in sleep(1) when perform_soft_reset() is called
03:52:53godzirramyzar: even after the scary movie I still dont really understand perpendicular drives.
03:53:00myzarit's so damn easy
03:53:15myzarinstead of layering the bits _ _ _ _ _
03:53:20myzarthey layer them | | | | | | | |
03:53:34myzaror something
03:53:36godzirraBut why couldnt they do that before?
03:53:41sneakumsso it hangs after ata_init(), not during
03:53:44godzirraor make stacks of them
03:53:46godzirrai.e. ======
03:53:47myzaryes, it is after
03:53:50Mikachuhow do they get the read head to different depths?
03:53:55myzarit's sleeping
03:54:12myzarwhy is it resetting, though?
03:54:19sneakumslinuxstb: here is the post:
03:54:34sneakumsmyzar: because ATA_STATUS was set to STATUS_ERR
03:54:37 Part Platyna ("Live Free Or Die!")
03:54:55myzarno, sneakums
03:54:57myzarthat just says it's an error
03:55:04myzarif it's hanging in sleep
03:55:17myzarthat probably means it's done reading (or trying to read)
03:55:18tonpeople could anyone make a port for creative zen vision m
03:55:18sneakumsyes, and then rockbox calls perform_soft_reset()
03:55:21myzarand it has slowed down
03:55:22tonso manyt people have it
03:55:27myzarthe drive physically has slowed down
03:55:29myzarwhich is sleep mode
03:55:33myzarit isn't spinning at full rpm
03:55:44godzirraKalthare: so whats the issue with music?
03:55:46myzarthe reason, godzirra
03:55:55tonso could anyone
03:55:55sneakumsthe sleep function it's hanging in is the one that's called when rockbx wants to *gasp* sleep
03:56:00myzaris because the more energy it takes to read the drive
03:56:00Mikachuton: if someone wants to do it, they'll do it.
03:56:03Mikachuton: it's a lot of work
03:56:09myzarthe more of a chance it can mess up the magnetization
03:56:26sneakumsthere seems to be an unrelated problem with using sleep in the bootloader, since udelay() works fine
03:56:26myzarand that's proportionate to the size of the magnetic region
03:56:28toni know its alot of work
03:56:32Kaltharegodzirra: I wish I knew. My guess at this point is that the longer disk accesses is making some part of the code, somewhere, unhappy.
03:56:40myzarso instead of making drives bigger
03:56:45myzarthey add a layer
03:56:49myzarand stack the data
03:57:01myzarthey wanted to do this with floppy disks years ago
03:57:06myzarand make an even bigger disk
03:57:24myzarbut 20 years later, they decided to lower costs of hard drives by applying this instead
03:57:53sneakumsthere are sheer physical limits to how fast you can seek a disk head and get it in position too
03:58:00 Join Will [0] (
03:58:01sneakumswhich is why drives have been getting smaller
03:58:12myzarthe reason for that is because of the energy i stated
03:58:13toni know its alot of work but could anyone please try to put rockbox on creative zen vision m
03:58:17myzarit has moving parts in it, remember?
03:58:23 Join webguest14 [0] (i=47566355@gateway/web/cgi-irc/
03:58:26Mikachuton: no
03:58:32myzarand the more space and the more energy in the drive, the more of a chance it can heat up and demagnetize
03:58:36tony bit
03:58:36myzarand your data will go poof at that point
03:58:40Willguys - what is the best way to boot into Apple firmware - seems a bit hit and miss to me
03:58:43linuxstbsneakums: Yes, I don't think sleep() should be used in the bootloader. We don't have interrupts enabled, so if Rockbox sleeps, it won't wake up.
03:58:55myzartry opening up a drive as it's spinning at full speed and see if you can still read the data
03:59:01MikachuWill: you'll get the hang of it after a while
03:59:14myzarlinuxstb, that's a valid hypothosis
03:59:26Willwhen is the perfect time to hit menu to get into apple firmware?
03:59:27webguest14i keep getting the -1 error message on my boot up
03:59:29linuxstbWill: It's in the FAQ - turn Rockbox off by a long press on PLAY, then press MENU to turn your ipod back on, and keep your finger pressed on MENU (never release or move it) until you see "Loading original firmware".
03:59:35myzarit shouldn't be sleeping in the bootloader, and there is no code to help it sleep if it does
03:59:40MikachuWill: just around the time the apple shows up
03:59:40sneakumslinuxstb: i replaced the sleep calls with udelay calls, and the soft reset completes, but it doesn't help
03:59:55webguest14i geuss ill wait my turn then lol
03:59:56MikachuWill: if you have it turned of and just press release and hold like linuxstb said, it should work
04:00:01myzarthe question is why is it soft resetting?
04:00:03myzarthe 30 doesn't. does it?
04:00:05Willthanks guys
04:00:10myzarthe 60 from 5gs doesn't as well
04:00:17myzarnone of the previous generations do a reset
04:01:30Willis there any way to have a menu at beginning to choose firmware to load?
04:01:40webguest14OK i have windows xp an a 30gigabit 5g ipod i downloaded everything correctly up until i tryed to install the firmware everything was fine until i downloaded the fonts build for the video ipod how an where do i extract it?
04:02:05Will <−− great program for getting album art if you are interested..
04:02:18sneakumsmyzar: actually, the question is why the status is set to STATUS_ERR after the read. i know why it's resetting: it's because that's what ATA code does in that circumstance.
04:02:21MikachuWill: you can use the ipodlinux bootloader
04:02:21PaulJamWill: you could use the ipodloader2 from ipodlinux. this has afaik a graphic menu
04:02:40myzarthat is another path as well, sneakums
04:02:42Willis it a hard one to install?
04:02:49myzarif you know why it is resetting, you can fix it from there
04:02:49 Quit lukaswayne9_ ("Ex-Chat")
04:02:56myzarif you know why it is giving an err, then you can fix it there
04:02:57webguest14does ipodlinux work on g
04:03:01 Quit lukaswayne9 ("Ex-Chat")
04:03:13myzarbut the way to check it is to compare to the 5.5g 30
04:03:17webguest14does it work on %G ipods the site said no
04:03:28godzirraKalthare: so no closer than currently?
04:03:55Kaltharegodzirra: Very much less likely to corrupt your filesystem. And you might have better luck than I'm having.
04:03:59 Quit webguest30 ("CGI:IRC")
04:06:45 Quit webguest14 ("CGI:IRC (EOF)")
04:07:37KalthareIt's a cleaner patch now, too, tho that doesn't matter too much.
04:08:15godzirraKalthare: I'll try it out. Where was the link to the patch?
04:08:42Will ipodloader2 worked a treat guys - thanks very much
04:08:55KalthareThe patch I posted earlier was for the FAT test code. I'm still putting the finishing touches to the 5.5 patch. :)
04:09:04godzirraahh ok
04:09:16godzirraif I'm not here, message me the link. I'll be happy to try it out.
04:10:39KalthareIt'll be up in a sec..
04:16:17Mikachucan't you just reformat the device?
04:16:57 Quit midkay (Read error: 104 (Connection reset by peer))
04:17:01KalthareMikachu: Yes, but you have to hex-edit the MBR, and you can only access the partition using mtools.
04:17:16KalthareAnd it still only works on the 30GB.
04:17:22myzarand none of us have any idea why it wont work on the 80
04:17:25myzaras opposed to the 30
04:17:36myzari can't find any mention of this 80 drive anywhere
04:17:42myzarno specs, differences, comparisons...
04:18:03 Part pixelma
04:19:18 Join midkay [0] (n=midkay@rockbox/developer/midkay)
04:19:53linuxstbMikachu: The Apple bootloader expects 2048-byte sectors, and the disk mode tells the host PC that the disk has 2048-byte sectors. Kaltare has worked around the first problem, but there's not much we can do with the second.
04:20:14PaulJamthis is just a wild guess, but could it have smething to do that the 30 GB drives support ATA 2 to 6 and the 80 GB drives only ATA 3 to 6 , so that the 80 GB ipods have a different hardware?
04:20:49 Quit Genre9mp3 ("I don't suffer from Rockbox psychosis. I enjoy every minute of it.")
04:20:59KalthareI don't think I can help with the 80 at all. I haven't got one, and I don't know anything about driver programming or ATA. I do hope it's made to work, tho.
04:23:55 Quit besonen_mobile (Read error: 60 (Operation timed out))
04:25:55 Part Will
04:27:06sneakumsPaulJam: if rockbox's ATA driver is using ATA-2 commands, then that might be it
04:28:04 Join besonen_mobile [0] (
04:30:01 Join EspeonEefi [0] (i=espeonee@STRATTON-SEVEN-FOURTY-SEVEN.MIT.EDU)
04:32:33 Join niskel [0] (
04:33:08KalthareHmmm. When was the new scheduler put in?
04:33:26sneakumsthat said, they do seem to be upward compatible
04:33:33*sneakums grabs the ATA-3 doc to see if there's a lit of removed stuff
04:34:22sneakumsnothing in the TOC jumping out
04:34:59bluewormOh well, still no go, on 80GB
04:35:19myzarhow can that be, sneakums?
04:35:26myzaryou mean they snuck in ATA-3 into the new 80?
04:38:20 Join Moos [0] (i=51400b8c@gateway/web/cgi-irc/
04:41:45 Join Paul_the_Nerd [0] (
04:41:51 Quit niskel (Remote closed the connection)
04:49:33KalthareHow many threads are there?
04:50:49Paul_the_NerdThere's some sort of max threads define, I believe.
04:52:21 Quit linuxstb ("CGI:IRC (Ping timeout)")
04:52:23KalthareI'm thinking it may be another thread that's freezing the system.
05:00:01 Join amiconn_ [0] (n=jens@rockbox/developer/amiconn)
05:00:22tonrockbox for creative zen vision m
05:00:24tonwe need
05:00:27Paul_the_NerdThen do it
05:00:40Paul_the_NerdNew ports happen by someone who owns the player doing it.
05:00:50 Quit midkay (Read error: 104 (Connection reset by peer))
05:00:54tonbut i dont no how
05:01:09 Join Rob2222_ [0] (
05:01:24Paul_the_NerdWell, pestering the room doesn't make it any more likely to happen
05:01:50tonk sorru
05:03:39 Join midkay [0] (n=midkay@rockbox/developer/midkay)
05:08:27 Quit ton ()
05:08:33 Quit amiconn (Read error: 60 (Operation timed out))
05:08:33 Nick amiconn_ is now known as amiconn (n=jens@rockbox/developer/amiconn)
05:09:31KalthareHow likely would it be that disc accesses taking longer would cause trouble for another thread?
05:13:08myzarthat would need to be in apple's firmware itself
05:13:23myzarthey can't put that into the drive and state that if the accesses take longer, it outputs an error
05:13:43myzarit's a possibility, but...
05:13:56myzari doubt it
05:14:07myzartake an example of reading a large large music file
05:14:12myzarit accesses longer than smaller ones
05:14:14myzarit can read it fine
05:14:45KalthareI'm talking about the playback crash on the 30GB model.
05:15:07myzaroh, i was tieing in the 80
05:15:21myzari'm not sure how the apple firmware is handled when rockbox runs physically
05:15:47KalthareThe problem with the 80 is a wholly different issue.
05:15:48myzarcould it be a caching problem?
05:15:57myzaryes, but when we fix the 80
05:16:02myzarwe'll probably get a playback crash as well
05:16:20Paul_the_Nerdmyzar: When Rockbox is running, the Apple firmware is uninvolved.
05:16:48Paul_the_NerdRockbox is a replacement firmware.
05:16:58Paul_the_NerdIt's not a modification of existing firmware.
05:16:58 Quit Moos ("CGI:IRC (EOF)")
05:17:18myzarof coarse
05:17:24myzarbut does it use any part of the original firmware?
05:17:36 Quit Rob2222 (Read error: 110 (Connection timed out))
05:17:54Kaltharemyzar: No.
05:17:55Paul_the_NerdIf it used any part of theirs, it would qualify as a modification of theirs since it was there first, wouldn't it?
05:18:02KaltharePaul_the_Nerd: Any thoughts on my question?
05:18:08myzarnot at all
05:18:18myzaryou can still get data or information from calls from something else
05:18:22myzaras an example, at least
05:18:29myzarthen it has nothing to do with the firmware
05:18:40myzarit's a change in the hardware, probably the drive
05:18:56myzarand from what we can tell from the debug, it's after ata_init
05:18:59myzaris that correct, Kalthare?
05:19:03Paul_the_Nerdmyzar: When someone says "using" a part of something, that assumes accessing or executing it, not just learning from it via reverse engineering or whatnot.
05:19:21myzarwell, not when i "use" a part of something
05:19:30myzari say this because i've used that term for almost a decade
05:19:39myzarperhaps it's because i'm not using C++
05:19:39 Quit EspeonEefi ("Leaving")
05:19:44Paul_the_NerdIf the Rockbox *devs* used the Apple firmware, it would fit your definition.
05:19:50 Join version7 [0] (
05:19:54myzarit's not my definition
05:19:58Paul_the_NerdBut if the Rockbox *software* used the Apple firmware, that would mean that it *must* be present for Rockbox to execute.
05:20:04Paul_the_NerdBecause it's a little hard to use something that's not there, right?
05:20:06myzarthank you for speaking down to me Paul_the_Nerd
05:20:31Paul_the_Nerdmyzar: I'm explaining the way a lot of terms are interpreted around here, because common ground on terms is essential for clear communication.
05:20:52myzari use terms quite differently, because of the work in the past
05:20:54myzarthat i have done
05:21:33myzari'm not exactly sure how the IPod works seperate from their firmware and the harddrive it runs
05:21:41Paul_the_NerdWhere a software can "use" something without actually using it, but rather by having its programmers learn from that something and reimplement similar or identical things?
05:22:00myzari never mentioned reverse engineering
05:22:04myzari'm not sure how the firmware works
05:22:05Paul_the_NerdEven if the terms mean something different for you overall, it's important to either use them how other people here will understand them, or clarify what is meant.
05:22:13myzarfor all i know, you can be running rockbox alongside the official firmware
05:22:17myzarand grabbing information from there
05:22:26 Quit sneakums ("taster, you're the automatic saint")
05:22:27Paul_the_NerdI said that in my first thing to you
05:22:34Paul_the_NerdYou can't *replace* something and still be dependent upon it.
05:22:55Paul_the_NerdReplacing something means substituting for it
05:23:01Paul_the_NerdMeans that the original would not be there in a true replacement
05:23:06Paul_the_NerdMeans you *can't* depend on it because it's not there
05:23:12myzarnobody said it has to be a *TRUE* replacement either
05:23:17Paul_the_NerdIt's the very definition of the word "replacement"
05:23:22version7I need a simple question answered? will the rockbox firmware allow navigation of the ipod on a external screen such as a tv?
05:23:34Paul_the_Nerdmyzar: So, you assumed that I misused the word replacement?
05:23:34myzari'm sorry Paul_the_Nerd, but we aren't speaking in definitions
05:23:41myzardid i say you misused it?
05:23:44Paul_the_NerdWe're speaking English.
05:23:53myzarno, we're actually typing english
05:23:54Paul_the_NerdYou said "nobody" said it had to be a "true" replacement
05:24:03Paul_the_NerdTrue, we are typing English.
05:24:08myzaryou mentioned that would be a TRUE replacement
05:24:16PaulJamversion7: I don't think the tv out on the g5 works yet
05:24:21Paul_the_NerdNo, I said Rockbox was a replacement firmware.
05:24:26version7this is for a ipod photo
05:24:31myzarand i mentioned that there are plenty of instances where something can replace something else and still use things from the old one
05:24:34Paul_the_NerdI used the word replacement as its true definition.
05:24:36myzari'm not arguing that
05:24:43myzari didn't know it was a full replacement
05:24:52Paul_the_NerdEven though I told you it was, completely clearly
05:25:12myzaryou can tell me whatever you want
05:25:12Paul_the_NerdWhich means you either assumed that I didn't know what I was talking about, or that I misused the word.
05:25:19myzari can still selectively choose to argue or believe it
05:25:26myzarto see how long of an argument i can pull out here
05:25:44Paul_the_NerdRockbox is a replacement firmware in the true sense of the word.
05:25:51myzarthanks for the info, Paul_the_Nerd.
05:26:11myzarbut you missed one point i made before
05:26:18Paul_the_NerdThis is why the developers are extremely hesitant to implement the current dual boot method that has been created for the X5, because it requires a dependency on the original firmware.
05:26:21myzarsince i don't know the schema of an ipod
05:26:35myzari am probably referring to the firmware improperly
05:26:45myzari am trying to figure out what can regulate how rockbox works
05:26:49myzarindependent to the firmware itself
05:27:10myzarwhat would cause the ipod to reset and crash, for instance
05:27:10Paul_the_NerdSo, you want something other than Rockbox to regulate Rockbox?
05:27:30myzarit doesn't have to only be rockbox
05:27:38myzari'm figuring out why the damn thing crashes on playback
05:27:44Paul_the_NerdMany things could cause it to crash, like unaligned memory accesses or hardware not being initialized or used properly.
05:27:46myzarand since firmware is eliminated...
05:27:56Paul_the_NerdHow is Rockbox eliminated?
05:28:18myzari meant the apple firmware.
05:28:21myzarnot rockbox.
05:28:40Paul_the_NerdWell it's a pretty important distinction to breeze over.
05:28:52myzari've been referring to apple's firmware as firmware
05:28:57myzarand rockbox as the alternate firmware
05:28:59myzarthis whole discussion
05:29:20Paul_the_NerdWell, try to use the terms everyone else does.
05:29:40Paul_the_NerdThere are many names for the original firmware that aren't vague like "firmware"
05:29:54myzari'm sorry, i rarely conform to the majority
05:30:02myzarbecause i speak for myself, not for the understanding of others
05:30:15Paul_the_NerdKalthare: Do we actually *know* that all the hardware inside the 5.5G is the same, or are we assuming it is because it works mostly the same?
05:30:41Paul_the_Nerdmyzar: Well, if you want to speak *with* someone, you have to speak in a way they understand. This is why language exists.
05:30:56Paul_the_Nerdmyzar: If you feel the desire to continue speaking solely for yourself, feel free to take it to PMs
05:31:12KaltharePaul_the_Nerd: We've at least sight-verified the PP chip to be the same.
05:32:40Paul_the_NerdKalthare: I'm just wondering about the distinction Apple is making with the Search function, and wonder if there's something else that's been missed, or if they're honestly just lying about the old ones being too slow. Because if there is a hardware difference, it could be relevant, right?
05:33:03myzarPaul_the_Nerd: I would assume you understand what I say, unless you aren't qualified to speak on the topic we are discussing
05:33:22myzarand at that point, you could ask what i mean, if you have any questions as to the validity or definition of a term or phrase i say
05:33:51KaltharePaul_the_Nerd: It might be the cause of all this 2K-sector insanity. The fps-test came out exactly the same as the 5G.
05:34:32Paul_the_Nerdmyzar: I did question it. And I've now suggested to you that you use the accepted terms for this channel
05:35:43myzarPaul_the_Nerd: Very well. Your point has been taken.
05:36:35 Quit pagefault (Read error: 54 (Connection reset by peer))
05:36:53Paul_the_NerdKalthare: I doubt longer disk accesses are the problem though
05:37:04Paul_the_NerdKalthare: But where is the extra time spent?
05:37:27***Saving seen data "./dancer.seen"
05:37:54KaltharePaul_the_Nerd: There are a lot of accesses −− particularly for directories, and the FAT itself −− that are pulling 4 sectors now, instead of just 1.
05:39:34 Join pagefault [0] (i=pagefaul@pdpc/supporter/active/pagefault)
05:39:49 Quit pagefault (Read error: 54 (Connection reset by peer))
05:39:51Paul_the_NerdKalthare: Do you have a description of the freezes somewhere?
05:39:55 Join pagefault [0] (i=pagefaul@pdpc/supporter/active/pagefault)
05:43:42Paul_the_NerdKalthare: For example, I don't know of they're consistent or semi-random.
05:44:04KaltharePaul_the_Nerd: It crashes while closing /.rockbox/.playlist_control after writing it. Specifically, it seems to be crashing partway thru the copy_read_sectors() loop reading the /.rockbox directory, so as to update the file size.
05:44:51Paul_the_NerdKalthare: I'm guessing you've tried other file writes too and gotten freezes as well?
05:45:30KaltharePaul_the_Nerd: No. It updates the tag cache with no problem. Doom can save and load without a hitch. It passes all the tests in firmware/test/fat/.
05:47:09Paul_the_NerdKalthare: So it would seem dependent about what else was happening on the same time, you're right.
05:47:28KaltharePaul_the_Nerd: I'm thinking it's playback.
05:48:12Paul_the_NerdAnd you can't get playback started to try another write during it?
05:48:51Paul_the_NerdBecause it freezes during that... =/
05:50:12KalthareMaybe I'll neuter the function that's writing .playlist_control.
05:51:00Paul_the_NerdKalthare: Maybe try that, then do another file write during playback and see if it's related to that function, playback, or some unholy combination of the two.
05:51:44 Join RoC_MM [0] (
05:56:32KalthareNot to mention seeing if removing that code makes it NOT crash.
05:56:43Paul_the_NerdYeah, that too. :-P
06:01:02KalthareI'll be a bit, tho. The old version of my patch caused nasty FS corruption −− and an intermediate version that I didn't post caused much, much worse corruption. So I need to reformat my iPod.
06:02:31KalthareWell, actually.. which, would you say, is worse? Writing FAT updates to the wrong location on the drive, or writing only the first quarter of each updated sector of the FAT?
06:03:23Paul_the_NerdThat's a tough call.
06:03:33Paul_the_NerdI think they're both bad enough.
06:04:35KalthareLuckily, the new version does neither of these things. :)
06:05:20Paul_the_NerdI think that's an important feature.
06:05:59 Join |AhIoRoS| [0] (n=ahioros@
06:07:05 Join EspeonEefi [0] (i=espeonee@STRATTON-THIRTY-EIGHT.MIT.EDU)
06:15:12 Quit |AhIoRoS| ("Abandonando, see you")
06:16:24 Quit PaulJam (".")
06:29:52pagefaultI had a request that you guys install CIA for rockbox
06:30:08pagefaultit seems to be the trend
06:31:13pagefaultit's an IRC bot and web site that tracks CVS changes
06:31:18pagefaultand posts them to the channel
06:31:24pagefaultnearly everyone is using it now
06:31:38pagefaultit's pretty cool
06:32:01pagefaultjust useless stats
06:32:09pagefaultbut you get it posted to IRC when someone makes a commit
06:33:01pagefaultthe nice thing is you can have an RSS feed for your project
06:35:48pagefaultanyway just a thought
06:36:27 Quit version7 ()
06:36:28pagefaultif you join #commits
06:36:33pagefaultyou can see it in action
06:43:25 Quit besonen_mobile (Connection timed out)
06:45:27Paul_the_Nerdpagefault: Actually, I think that's kinda neat if it could be made to return the automated build results in some reasonably readable way.
06:45:48pagefaulti'm sure the folks in #cia could do something fo ryou
06:45:56pagefaultthat is the development channel for it
06:46:27pagefaultor they may already have that but just not documented
06:46:49Paul_the_NerdWell, it mentions the possibility to do that in the section on how it works.
06:47:57Paul_the_NerdBut with how many automated builds are created, it'd either be unreadable, or require some sort of filtering, perhaps only the first failed build being reported or something. I'm just thinking anyway, use of something like that is more for those like Bagder and LinusN to consider, I think
06:52:45 Join besonen_mobile [0] (
06:55:42pagefaultI should just do the button mappings for the h10
06:55:47pagefaultsince no seems to want to do it
06:57:48pagefaultneed more 128x128 themes
06:57:54pagefaultiCatcher is getting old
07:13:59 Join MarcoPolo [0] (
07:15:22 Join BHSPitLappy [0] (
07:16:07 Quit BHSPitLappy (Read error: 104 (Connection reset by peer))
07:16:52 Quit blueworm ("Leaving")
07:17:18 Join BHSPitLappy [0] (
07:32:57 Join Sinbios [0] (
07:33:56 Join GFoux [0] (
07:37:31***Saving seen data "./dancer.seen"
07:46:05 Quit nGs|GreyFox`afk (Read error: 110 (Connection timed out))
07:50:37 Join crushjane [0] (n=jay_jay_@
07:52:17 Quit pagefault ("Leaving")
07:55:41hcshas anyone considered an encycopodia (ipod-formatted wikipedia) viewer for rockbox?
07:57:14hcsmight be fun to write a pseudo-html renderer
08:00:11scorchei believe one was started
08:00:37 Quit Sinbios ("If the definition of a klutz is someone who doesn't have eyes on their ass, then yes, I suppose I am a klutz.")
08:05:46 Join theli_ua [0] (
08:13:11 Join matsl [0] (
08:23:03MarcoPolohcs: one was started but i think the project isn't alive anymore
08:23:34Bagderthere's talk about that on misticriver only
08:23:43Bagderand there doesn't seem to be any code available for some reason
08:23:55MarcoPolowell, not completly dead :
08:24:16Bagderwell, according to posts on MR he's stopped developing it
08:24:57MarcoPoloBagder: ok
08:38:32 Join pondlife [0] (
08:43:04 Join dan_a [0] (n=dan-mirc@
08:48:07mirakis the TV out of a DVB-C card hauppauge nexus or tecnotrend premium better than TV OUT of video card ?
08:48:42scorchefor what?
08:49:09scorchei would use the graphics card out for anything besides just video honestly
08:49:39mirakvideo is not as good ?
08:49:43scorcheand you are better off using the dvi/vga signal than the tv out
08:49:58 Join rigel [0] (
08:50:07mirakon a cathodic TV I can't do that.
08:50:07rigelare there plans to port for the zune?
08:50:20rigelparticularly the wifi capabilities
08:50:33Bagderrigel: see the NewPort wiki page
08:51:22dan_aBagder: I've got a working LCD driver for the Sansa
08:51:28mirakscorche: sorry I aske in rong channel anyway ^^
08:51:34Bagderdan_a: supercool!
08:51:44rigeli dont see it
08:51:50dan_aThank MrH for his patience with me!
08:51:51amiconnNow rombox is too big on the v1 as wel :-(
08:51:58scorcherigel: then look harder ;)
08:52:10*scorche cries on amiconn's shoulder
08:52:20Bagderimho, rombox needs to become a special build with some stuff ripped out
08:52:38scorchei wouldnt mind getting rid of tagcache etc
08:52:54*amiconn disagrees
08:53:10rigelscorche: url plz
08:53:42scorcherigel: click wiki on the side, enter NewPort into the box at the top...
08:53:48Bagderamiconn: then what's your suggestion on how to keep our nose above the water?
08:54:26rigeloh, case-sensitive
08:54:34rigeli always assume otherwise
08:54:39Bagderrigel: executive summary: "we don't do port on request, people who own units do ports"
08:55:12Bagderafaik, not a single guy close to Rockbox owns a zune
08:55:20scorcheamiconn: where are you seeing that it is too big for the v1?...does it not give the error that it does on the v2?
08:55:32amiconnNo, not yet
08:55:35rigelBagder: not yet, theyre not out
08:55:45amiconnFor some reason the linker doesn't give that error immediately
08:55:47rigeli dont know if there are fcc photos yet either
08:56:05amiconnMy local build is 132 bytes too big, official is probably a bit more
08:56:20Bagderrigel: there are reasons they are somewhat similar to some gigabeat models, and we have some work on gigabeat going on
08:56:32BagderI meant - there are reasons "to believe"
08:56:54scorcheBagder: i thought it was similar to gigabeat players that we arent working on
08:57:01scorcheor were you just talking in general?
08:57:19Bagderyes, but I believe the ones they are working on are somewhat similar to the ones they aren't working on as well
08:57:22Bagderbut I might be wrong
08:57:47scorchejust taking tagcache out alone should grant us plenty of space for rombox
08:58:02 Join pagefault [0] (i=pagefaul@pdpc/supporter/active/pagefault)
08:58:09amiconnBagder: Would it be possible to collect and link the distributed build .map files somewhere? (for targets and bootloaders)
08:58:51Bagderyes, sure it could be done similar to the bleeding edge build
09:00:30Bagderwould you like it for the latest build only, or with a history?
09:01:59 Join petur [0] (i=d4efd6a6@gateway/web/cgi-irc/
09:04:33 Quit ShadowdogMU ("You know you'll miss me :P")
09:08:34 Join _FireFly_ [0] (
09:14:06 Join FireFly_ [0] (
09:14:06 Quit _FireFly_ (Connection reset by peer)
09:18:17 Quit scorche (Remote closed the connection)
09:21:32 Quit matsl (Read error: 110 (Connection timed out))
09:22:17 Join matsl [0] (
09:27:39 Join dpro [0] (
09:29:35 Quit FireFly_ (Read error: 145 (Connection timed out))
09:34:04 Join LinusN [0] (i=linus@rockbox/developer/LinusN)
09:37:33***Saving seen data "./dancer.seen"
09:41:45 Join besonen_mobile_ [0] (
09:43:59 Quit besonen_mobile (Read error: 110 (Connection timed out))
09:44:14 Join FireFly_ [0] (
09:45:22 Quit myzar (Read error: 104 (Connection reset by peer))
09:47:10 Quit dan_a (Read error: 110 (Connection timed out))
09:49:21 Quit crushjane (Read error: 110 (Connection timed out))
09:55:20Slasheriweird.. i think original firmware can corrupt the filesystem
09:55:21 Quit BigBambi ("Leaving")
09:55:29Slasherinow fsck.vfat says:
09:55:30SlasheriCurrently, only 1 or 2 FATs are supported, not 191.
09:55:59Slasheriand only booted up original fw a few times to fix the problem with flash plugin
09:56:24Slasherifortunately files are still there but even .rockbox-directory got corrupted and needed to erase it for rockbox to boot up again
09:56:51 Quit RoC_MM ("Leaving")
10:00:03pondlifeSlasheri: which device?
10:00:10Slasheripondlife: h140
10:03:41 Join spiorf [0] (
10:05:39 Join dan_a [0] (
10:05:58 Quit matsl (Remote closed the connection)
10:09:25pondlifeSlasheri: This is rather obscure, but could you look at I'm wondering why you had to change the timing at line 3221/3228..
10:09:48pondlifeI'm sure it's fine, bur I'd like to understand
10:11:00pondlifeDid this increase UI responsiveness, for example?
10:11:27Slasherijust a moment
10:11:50Slasheriah, yes, you can change it back
10:12:01pondlifeNo need particularly
10:12:19Slasherishouldn't be necessary.. i was just experimenting with something and forgot it there
10:12:35pondlifeI'm just trying to justify every change in the mid-Sept updates regarding the whitenoise issue,. I very much doubt this is relevant though
10:13:36 Quit JdGordon ("Konversation terminated!")
10:13:36 Quit FireFly_ (Read error: 131 (Connection reset by peer))
10:15:30 Part [aspect]
10:21:35 Join myzar [0] (
10:22:35 Join TeaSea [0] (n=thunderc@
10:23:58 Quit TeaSea (Remote closed the connection)
10:26:06 Join Siku [0] (
10:42:04 Join JdGordon [0] (
10:45:29amiconnSlasheri: Did you see last night's comment about the new scheduler having problems on ipod when priority scheduling is disabled (i.e. in the bootloader)?
10:48:40 Join alphakiller [0] (n=bits@
10:49:21amiconnBagder: .map files with history would be preferable, if that's possible
10:49:30 Quit barrywardell ()
10:49:36 Quit alphakiller (Client Quit)
10:49:53Bagdersure, I just need to give each a local name according to the build time stamp or similar
10:50:16amiconnI mainly want them to have a means for tracking down UIEs reported here, or on the ml/forum
10:50:25Bagdera good idea
10:50:38amiconnAnother purpose is precisely tracking size of rombox
10:50:49amiconnDon't forget that archos builds have 2 .map files
10:51:01Bagderah, right
10:51:37amiconnWould it be possible to have .map files for dailies as well, or is that system too different?
10:51:59Bagderthey're similar
10:52:01pondlifeDon't suppose Tomasz Malesinski is around under some name,,?
10:52:21Bagderpondlife: I don't think he's on IRC much
10:52:29pondlifeThere was a mod "Do not use our simple snprintf in simulator, because ALSA library requires a more advanced one. Fixes the problem with ALSA on Linux."
10:52:43 Quit JdGordon ("Konversation terminated!")
10:52:43pondlifeBut it breaks the sim logging here (Cygwin)
10:53:07pondlifeSurely the sim should be able to use our code as much as possible too.
10:53:28Bagderyeah, but alsa clearly needs a better snprintf than what we provide
10:53:49BagderI guess the proper fix would be to modify the check to only be used on linux
10:54:02pondlifeDoesn't SDL take care of the differences?
10:54:19Bagderthen you wouldn't have any problems ;-)
10:54:21pondlifeSorry, I'm lacking in cluefulness on this
10:54:39pondlifeIt seems odd that ALSA would be relying on our code.
10:54:44BagderI believe ALSA is what needs the snprintf and alsa is not used on windows/cygwin
10:54:56Bagderyeah, but it does/did
10:55:09Bagderit could of course possibly be fixed in a different way to have
10:55:15pondlifeCallbacks, I guess..
10:55:17Bagderit use the libc snprintf for alsa
10:55:37pondlifeOK - what's a good #ifdef to detect Linux vs Cygwin?
10:56:03pondlifePreferably one already used in Rockbox.
10:56:27BagderI believe "linux" and "__linux" are defined when building for linux
10:56:57pondlifeAnd Cygwin doesn't emulate that?
10:57:04Bagderno, cygwin is not linux
10:57:12pondlifeI know, but it tries to pretend
10:57:18Bagdernot really
10:57:25Bagderit emulates posix/unix
10:57:42pondlifeThanks, I'll play and fix
10:59:49amiconnpondlife: His nick is tomal, but he's indeed not often here
11:05:58 Join Lynx_ [0] (
11:05:58BagderI figure we ignore the plugin .maps, or would they ever be really useful too?
11:07:30amiconnMuch less often, and indeed they're so many that it would be a pita to collect them
11:07:39amiconnThere are codec .map files too...
11:09:10Bagderwell, I was pondering to do a 'scp' of all possible .maps and make a zip out of them locally or similar
11:10:40dan_aOr you could have the maps added in to the file by the build system?
11:11:16Bagderyes, but then everyone would download them as well
11:11:34BagderI guess I can make a separate build target and then download that zip
11:12:06pondlifeSlasheri: Another (unrelated) question..?
11:12:07BagderI think I'll try that
11:13:27amiconn'make maps'?
11:13:39amiconnor 'make mapzip'
11:13:42pondlifeIf I use Tagcache to play all tracks, while it's building the playlist (displaying Searching...), it now counts downwards, as if the playlist is getting smaller/filtered. Is this right?
11:13:52Bagdermake mapzip
11:13:54pondlifeI'm fairly sure it used to count up
11:13:58Bagderls -l
11:13:58Bagder-rw-rw-r−− 1 dast dast 119333 Oct 16 11:13
11:14:13Bagderfor a recorder build done just now
11:14:28Slasheripondlife: yep, that is a feature
11:14:47 Join barrywardell [0] (
11:14:48Slasheripondlife: at first when it searches from tagcache, it counts upwards and after that downwards
11:15:02Slasherioriginally that was a bug but i think it could be a nice feature
11:15:12*pondlife disagrees or doesn't understand
11:15:20pondlifeTo me it looks like it's started to filter
11:15:37pondlifeI actually end up with a playlist of 11000 entries, so why count down?
11:15:37Slasheribecause when it inserts tracks to playlist, you can see how many tracks there are left to be processed
11:15:48Slasheriin fact i think the playlist system is quite slow..
11:15:59Slasheritagcache is way faster than playlist could handle
11:16:11pondlifeGotcha. Maybe the prompt shouldn't be "Searching..." but "Left to queue" or something?
11:16:28 Quit barrywardell (Client Quit)
11:16:28Slasheriyep, true
11:17:03pondlifeAny idea which code is making playlist handling slow?
11:17:04Slasheriat first it's searching when tagcache finds the entries.. but that's usually quite fast that you don't even see that
11:17:16pondlifeIndeed. Or only for 2 seconds.
11:17:29pondlifeThe bulk of time is this countdown, can take quite a while
11:17:49pondlifeIs it doing lots of inserts, rather than appends or something?
11:18:03Slasheriyeah, lots of inserts
11:18:15Slasheriit needs to insert each track individually
11:18:33pondlifeInitial build could be much faster then
11:19:08Slasheritrue.. it could start playback and process in background
11:19:27pondlifeNice. Play once the first entry is queued
11:19:51Slasheribut that needs some investigation how to do it
11:20:48Slasheriamiconn: not yet, i will check from logs soon
11:20:48 Quit sublimal (Read error: 104 (Connection reset by peer))
11:20:49pondlifeCould cause a problem if the user attempts playlist maintenance whilst it's still doing the initial build
11:21:23Slasheriyep, that is one potential problem at least
11:21:45Slasherior it could start playback in background while still waiting until all entries have been added to the playlist
11:22:08pondlifeSo you'd hear music but not have the WPS?
11:22:24 Join Genre9mp3 [0] (n=yngwiejo@
11:22:27pondlifeOr have a splash over the WPS?
11:23:02Slasherimaybe.. but then we probably would have to put wps in a separate thread
11:23:51pondlifeI think we could just speed it up for now. That should be safe and there should be plenty of room for improvement by the look of it.
11:24:05Slasheribecause when wps is shown, wps likes to take over all control from main thread
11:24:44Slasheritrue.. there must be some way to optimize playlist code in performance
11:25:09pondlifeMaybe a fast Tagcache->playlist interface?
11:25:38Slasheriprobably.. i wonder if hardeep has anything in mind
11:25:40pondlifeOr are there other cases that are slow
11:26:22 Quit pondlife ("reboot time")
11:27:19 Quit darkless ("Leaving")
11:37:39***Saving seen data "./dancer.seen"
11:47:04 Join obo [0] (
11:53:40 Join apv [0] (
11:58:54 Join chendo_ [0] (
11:59:22 Quit idnar (Nick collision from services.)
11:59:25 Join idnar_ [0] (i=mithrand@unaffiliated/idnar)
12:01:43 Join midkay_ [0] (
12:04:25 Quit chendo (Read error: 60 (Operation timed out))
12:04:45 Join mirfle [0] (i=52a680a5@gateway/web/cgi-irc/
12:06:59mirfledoes anyone here know how to convert a .hex to a .bin?
12:08:36 Join chendo [0] (
12:10:24 Join blueworm [0] (
12:12:13Slasherimirfle: hi :)
12:12:58 Quit chendo_ (Read error: 60 (Operation timed out))
12:14:03Slasherimirfle: i am not sure if you got the query (might have some problems unless you are registered here)
12:14:09Slasheribut i can dcc you the fix
12:14:54Slasheriah, in fact you are using a webchat so that isn't possible. but i will provide you an url
12:15:14 Quit petur ("lunch")
12:15:15 Join _FireFly_ [0] (
12:16:08Slasherimirfle: did you got it?
12:17:04mirfleI got it
12:17:26Slasherigreat, i will remove it now
12:18:15mirflenow I just run it with the iriver_flash plugin, right?
12:19:39pagefaultI just called the BSA on my work
12:19:42Slasherithen you should be able to boot original fw again
12:19:42pagefaultthis will be fun
12:19:45 Quit midkay (Read error: 110 (Connection timed out))
12:20:22pagefaultI bet they will come in and sieze all the computers
12:20:22mirfleshould I use the new version of the plugin you just put in cvs?
12:20:54Slasherimirfle: it shouldn't matter, but i would advice using the latest version
12:21:15Slasheriit just no longer corrupts the original fw when flashing the bootloader
12:23:35mirfleit works! thanks :)
12:25:08mirflegotta go
12:25:12 Quit mirfle ("CGI:IRC")
12:28:41 Join pondlife [0] (
12:45:10 Join ShadowdogMU [0] (
12:47:35 Join Id2ndR [0] (
12:47:47 Quit psiborg ("leaving")
12:47:48 Join Criamos [0] (
12:50:28 Join lee-qid [0] (
12:59:48 Join Quazgaa [0] (
13:01:01Slasherihmm, it seems that it's possible to create a playlist of like 4000 files under 2s if there is a better tagcache <-> playlist interface and dircache is used
13:01:34LinusNwhere is the bottleneck?
13:02:27Slashericurrently it seems that update_control() -function inside playlist.c takes all power
13:03:02Slasheriand currently it's not possible to cache entries containing filenames because that data is lost
13:03:27Slasheribut if tagcache could provide direct dircache pointers to playlist, then it would be very fast and that would be cacheable
13:03:49 Join JdGordon [0] (
13:06:40Paul_the_NerdAnyone here with an H300 series?
13:07:23JdGordonwhatcha need?
13:07:27 Quit Id2ndR ("Parti")
13:07:39amiconnSlasheri: And what happens if dircache is disabled or not available at all?
13:08:10Paul_the_NerdJdGordon: Well, people are semi-reporting drastically reduced battery life. I haven't managed to talk any of them into properly testing it, but I was wondering if someone more reliable could comment. ;)
13:08:22Paul_the_NerdJdGordon: Are you running a recent build, and is battery life horrible again?
13:08:59JdGordonPaul_the_Nerd: to tell you the truth.. i never really take notice of batt life... i usually charge every 2 days or so.. but im only about 3hours per day..
13:09:17JdGordonim happy to run battery bench tonight if your really interested
13:09:32JdGordonactually ill do it anyway...
13:09:36Paul_the_NerdJdGordon: If you could, just a cleared settings disk power off enabled battery bench with a modern build.
13:09:44*JdGordon hopes it actually outputs stuff this time
13:09:46Paul_the_NerdOne person says he's getting 2 hours in Rockbox vs 10 in retail.
13:09:53Slasheriamiconn: then it would be as slow as before
13:09:55JdGordonit didnt output anything on my nano when i did it
13:09:58Paul_the_NerdBut he doesn't say when he tested retail or anything
13:10:18Paul_the_NerdJdGordon: The Nano version of the plugin had a lot of problems due to the way the flash memory was handled if I recall.
13:10:25JdGordonah ok
13:11:15JdGordonok, im charging now, batt is about 80% so i should be able to start soonish
13:12:20Paul_the_NerdI mean, the *only* place I see it discussed is Misticriver where custom builds flourish, but one of them mentioned getting his build from the Rockbox website, so... I dunno
13:13:00FebsPaul_the_Nerd: I wonder if the H300 problem could be related to this:
13:13:10 Quit _FireFly_ ("Leaving")
13:13:19JdGordonwhats the expected life of a h320?
13:13:30Febs15-16 hours
13:13:54 Join petur [0] (i=d4efd6a6@gateway/web/cgi-irc/
13:14:52JdGordonso i bassically need to start 3 hours ago to make sure i will have some batt life for uni tomorow :D
13:15:20JdGordoni doubt it will go 12 hours tho.. because the batt is a bit old now
13:15:24Paul_the_NerdJdGordon: Honestly, about 4-6 hours of test ought to provide more than enough evidence either way.
13:15:37Paul_the_NerdA full one would be nice though
13:15:51JdGordonyeah, but if im gonna do the test, a full test would be nicer anyway
13:15:56Paul_the_NerdYes. :)
13:16:05*JdGordon wonders if i can beat my sister nano's 9h
13:16:28JdGordontime remaing while chargnign is estimated till total right?
13:16:32Paul_the_NerdFebs: Honestly I probably wouldn't have even thought twice about the problem if someone hadn't blatantly misused the word "implies".
13:21:00 Quit JdGordon ("Konversation terminated!")
13:26:47 Join lodesi [0] (n=lds@
13:27:07 Join falkyre [0] (n=falkyre@
13:29:41amiconnSlasheri: To state your timing figure... I don't deem tagcache slow. Building a 300+ tracks playlist from tagcache takes just a few seconds on archos recorder
13:34:09 Join JdGordon [0] (
13:37:40***Saving seen data "./dancer.seen"
13:38:16Slasherihmm.. interesting
13:40:53 Quit JdGordon ()
13:43:07amiconnI can tell you the exact amount tonight
13:45:12 Quit Febs (Read error: 110 (Connection timed out))
13:48:50amiconnSlasheri: Could it be that priority scheduling slows tagcache->playlist transfer down on swcodec?
13:49:31 Join PaulJam [0] (
13:50:35 Join jonno [0] (
13:51:30 Nick jonno is now known as JdGordon (
13:52:49JdGordonPaul_the_Nerd: im gonna give the batt another 8min to finish chargning then start the test with default settings @ 10pm
13:53:12 Quit apv (Remote closed the connection)
13:53:55Paul_the_NerdJdGordon: Thanks.
13:56:45Slasheriamiconn: i just discovered something weird.. playlist resuming has always been _very_ slow but now it was almost instant.. not sure what caused that
13:58:53 Join matsl [0] (
14:00:52 Quit scottder (Read error: 131 (Connection reset by peer))
14:01:47Paul_the_NerdUpdating Rockbox-in-flash on AJBR is just running rockbox.ucl right?
14:07:30 Quit Kitt0s (Read error: 145 (Connection timed out))
14:09:24 Quit JdGordon ("Konversation terminated!")
14:12:44LinusNPaul_the_Nerd: yes, but you must be running the same version that you are flashing, by first running the ajz file
14:13:19Paul_the_NerdLinusN: I remembered that much. :)
14:13:49Paul_the_NerdLinusN: I just couldn't remember whether to execute Rockbox.ucl or Rombox.ucl, but managed to look it up. I was fairly certain it was Rockbox, but wasn't about to just experiment.
14:14:23PaulJamPaul_the_Nerd: i don't think there are batteryproblems in recent builds (at least not with CVS on H300). both tests were made two days ago with cvs and reset settings and one album set to repeat:
14:15:30Paul_the_NerdPaulJam: So, no noticeable difference between the 10-14 and the 09-20 build. Good to know. :)
14:17:44 Join jhMikeS [0] (
14:23:20amiconnPaul_the_Nerd: That depends on whether you want rombox or not. Both .ucl files are flash files
14:23:46amiconnbut today's daily doesn't feature rombox anymore for the v1 :-(
14:23:49Paul_the_Nerdamiconn: I already have Rombox installed, just was updating.
14:24:04amiconnYou can switch at will if both are available
14:25:01amiconnRombox gives you more free buffer at the cost of slightly slower operation (not noticeable most of the time)
14:25:32 Quit Joely ("Get MacIrssi -")
14:25:41Paul_the_NerdI don't usually spend much time "operating" anyway.
14:26:25 Quit GFoux ("Le vrai danger, c'est quand les hommes penseront comme les ordinateurs")
14:28:44pondlifePaulJam: Thanks for checking out the old builds - re white noise etc...
14:28:57pondlifeYou said you thought it was older than 2006-09-18
14:29:09jhMikeSLinusN: hello. did anything pop out at you from those logs?
14:29:29pondlifeI reckon it will be either 16th or 17th that starts it happening, please continue!
14:29:35LinusNi haven't had time to work on it
14:29:44PaulJami have had the whitenoise problem in a sim from cvs source 06-09-01
14:30:06jhMikeSLinus: ok
14:30:58LinusNjhMikeS: have you studied the logs?
14:31:00*jhMikeS has to resync the patch again :)
14:31:42jhMikeSLinusN: I gave it some going over but but can't see what is it is
14:31:56LinusNi'll see if i can take a look tonight
14:32:11jhMikeSLinusN: Tried to step away for a bit and see if something pops in my head
14:33:47jhMikeSThere must be just some wierd little thing since in principle it ought to work
14:33:56 Quit NickDe ("Leaving")
14:34:14 Join barrywardell [0] (i=892b7a86@gateway/web/cgi-irc/
14:35:12LinusNi'll make some logs with the cvs version and compare
14:36:18jhMikeSGood idea. Probably identical though! l
14:36:43jhMikeSOther than a couple bits
14:38:20LinusN"Other than a couple bits" - i'll remember that one :-)
14:39:12*jhMikeS loves making bad puns :)
14:39:46*amiconn wonders what patch this talk is about
14:41:45 Quit blueworm ("Leaving")
14:41:48 Join JdGordon [0] (
14:45:30jhMikeSmore specifically finding out why source selection goes haywire on uda devices with WSO in slave mode
14:47:22pondlifePaulJam: Aha. So maybe try going back one month at a time..?
14:47:49PaulJamcurrently i'm going back in 10 day steps
14:48:20jhMikeSTry binary search, you
14:48:34jhMikeS'll get there faster :)
14:49:00pondlifeOK. Sounds like you are finding it quite easy to repro. I have only heard it very occasionally, certainly not reliably
14:49:14pondlifereliably failing that is
14:50:47jhMikeSI had a white noise problem on track skips until I reset my settings. Is that the issue or something else?
14:51:15pondlifeSimilar. Although this will happen with reset settings (if you turn repeat Off)
14:51:19 Join Kitt0s [0] (
14:52:02 Join Febs [0] (i=2662c44b@gateway/web/cgi-irc/
14:52:13jhMikeSI don't use repeat, just EQ and the EQ gets messed up a lot but the EQ was the problem in my case.
14:53:27 Quit Kitt0s (Read error: 131 (Connection reset by peer))
14:54:31 Join Kitt0s [0] (
14:54:53jhMikeSI was wondering, if you don't need "fast" bitmap why can't we store 8 bit paletted images for some things? That would save space just from the logo.
14:57:49amiconnBitmap display is never fast enough
14:58:30amiconnRockbox generally stores native bitmaps in framebuffer format
14:58:37jhMikeSnothings ever fast enough :)
14:58:57amiconnThis is both fast and KISS
14:59:06jhMikeSI know it does. I even threw together an 18-bit version with all the trimmings to check it out
15:00:43 Quit matsl (Remote closed the connection)
15:00:44jhMikeSIt would be very simple. Color bitmaps would have just a pointer as the first data...if NULL: native format, if non-NULL: pointer to palette and 8-bit format. No new functions exposed.
15:01:19*amiconn dislikes much of the artifial complexity introduced recently :/
15:01:28BagderjhMikeS: what would the point be?
15:01:46LinusNi also fail to see the point
15:01:49*Bagder prefers simplicity
15:01:54*LinusN too
15:03:26jhMikeSSaving some ram of course when the logo and the usb connect image don't really need to be super fast but they'll be about half size. The code would be rather insignificant.
15:04:27Bagderwe don't have very many such pics
15:06:30 Quit JdGordon ("Konversation terminated!")
15:06:32jhMikeSI know that but how much complication is it really? And you get some ram back. With an optimized palette the appearance should be about the same.
15:06:40 Join _FireFly_ [0] (
15:07:26 Quit Quazgaa ("leaving")
15:07:34Bagderif we would have a second format, I'd rather see it use a compressed format
15:07:44Bagdercompressed native, that is
15:08:08Bagderbut isn't this just a few hundred bytes we're talking about to gain?
15:08:11jhMikeSAnd uncompress then drawing?
15:08:24LinusNdo we have a problem with memory shortage on the platforms with color lcd's?
15:08:31 Join JdGordon [0] (
15:09:07LinusNi don't think we have
15:09:38LinusNlet's not solve problems we don't have
15:10:23godzirraSo any word on the 5.5g?
15:10:31godzirraLast I heard Kalthare had doom working but music playback wasn't?
15:11:13jhMikeSof course not but figure 8512 bytes verses 16000 for the logo plus a few hundred code
15:11:18jhMikeSon x5
15:13:13LinusNout of 16 megs
15:14:13jhMikeSKeep it for more useful things. Was glad I was able to save some IRAM of late too but that's more precious.
15:14:43LinusNiram is a different thing, of course
15:16:45jhMikeSI was suprise that using 18-bit didn't kill things that bad. Only lost 3fps on updates even with a 32-bit frame buffer. This was just "research" of course. I ain't plugging to use it cause it actually looks worse.
15:16:48Paul_the_NerdI think it's unneeded complexity too.
15:17:22Paul_the_NerdCompressed image drawing is almost surely going to work its way in at some point, I'd guess, especially as people try for more and more graphical / skinnable UIs.
15:17:38LinusNthere is one slight advantage of a paletted format, and that would be that all color platforms would have the same image format
15:17:46jhMikeSComplexity. Not to be critical but I find it strange to call that "complex".
15:18:30LinusNjhMikeS: not complex, but *more* complex than the native format
15:18:32Paul_the_NerdjhMikeS: Anything added to Rockbox at all qualifies as "additional complexity" though. ;)
15:19:27barrywardellIs there support for Rockbox to shut a player off once the battery gets below a certain level?
15:20:15JdGordoni tinhk there is a patch which adds that for the h300
15:20:21barrywardellthere are problems on the H10 where when the battery gets below a certain level the ADC isn't read properly
15:20:38barrywardellit ends up that up on the scrollpad is constantly read
15:20:43 Quit jhMikeS (Read error: 104 (Connection reset by peer))
15:20:45LinusNbarrywardell: the adc that reads the battery level?
15:21:00barrywardellit reads the scrollpad too
15:21:22 Join jhMikeS [0] (
15:21:22LinusNso, how do we trust the battery level adc? :-)
15:21:26barrywardellit ends up constantly reading up from the scrollpad
15:21:38barrywardelland no other buttons work
15:21:44barrywardelland the volume goes up to max
15:21:46rigelbarrywardell: i've had that problem with the h10, yes
15:21:55rigelits annoying as fuck all
15:22:09LinusNoh, *that* annoying? :-)
15:22:22barrywardellLinusN: the battery level still reads kinda ok, but maybe a bit low
15:22:34LinusNbarrywardell: i'm just teasing you
15:23:10LinusNbarrywardell: i think we should look into implementing such a feature
15:23:12JdGordonPaul_the_Nerd: its been going for 90min and the batt is still @ 90%.. so the guy definatly is over-exagurating.. or has a dead batt
15:23:40Paul_the_NerdJdGordon: Gotcha.
15:23:43jhMikeSrrr Connection bit it...
15:23:46JdGordonis it possible to change the cvs shortcuts? why is maual/ being downloaded with rockbox-devel?
15:23:54barrywardellLinusN: would it be reasonable to have it shut down once the battery gets below battery_level_dangerous?
15:24:23LinusNbarrywardell: i think so
15:24:44LinusNbarrywardell: the problem is what to do with the settings and resume info
15:24:55jhMikeSI was talking about two formats simultaneously and no, all format wouldn't be the same (in principle) because the palette would still have native color. But I'll stop beating a dead horse. :P
15:25:16barrywardellonce the battery gets below battery_level_dangerous, then disk writes don't happen anyway, do they?
15:25:30barrywardellso settings haven't been saved in that case anyway?
15:25:48Paul_the_Nerdbarrywardell: In theory a user can plug in without shutting down and losing anything though, right?
15:26:31barrywardellhmm, true
15:26:52Paul_the_NerdMaybe an option for "Low Battery" with a choice like "Warn" and "Shutdown" and a note in the manual that shutting down will most likely not result in saved settings / playback info / etc.
15:27:07barrywardellplugging in would probably push the level back up above battery_level_dangerous
15:27:21LinusNobo: there?
15:27:26jhMikeSjhMikeS: smells more pcf50606 interrupt handling. :)
15:27:30barrywardellmaybe we can just save settings once the battery level gets that low, then shutdown
15:27:38barrywardelland have the warn option too?
15:28:37jhMikeSWhy should have to trust the battery level ADC on platforms where we have hardware charge management?
15:28:54barrywardellI don't feel the need for a warn option, but could add one if it was desired?
15:28:58Paul_the_Nerdbarrywardell: The problem is the disk spinup. It's hard to tell when the battery will be too low to spin up the disk. I suppose you could pick an assuredly safe voltage, and have some sort of emergency "save" occur at that point no matter what, or something.
15:29:14LinusNjhMikeS: i don't understand the question
15:29:43jhMikeSThe pcf warns about low bat and impending shutdowns and such so just watch those interrupts for that.
15:29:50barrywardellPaul_the_Nerd: that's what battery_level_dangerous is for. it's the level below which rockbox will not do a disk spinup
15:29:57LinusNjhMikeS: on the x5 yes
15:30:07jhMikeSDon't the iRivers use it too?
15:30:11LinusNjhMikeS: no
15:30:26jhMikeSx5 isn't the only one iirc
15:30:46Paul_the_Nerdbarrywardell: Doesn't that mean there's a chance that when you reach it, it's too late, or is it set high enough to be sure?
15:30:51LinusNthe irivers use the pcf50606 as an advanced voltage regulator
15:31:21LinusNand the charging is done by another circuit
15:31:33barrywardellPaul_the_nerd: I'm not too sure. I guess it depends on who calibrated it for each player
15:31:59 Join bluebrother [0] (i=810d4658@gateway/web/cgi-irc/
15:32:42bluebrotherFebs, why have you put the eq description in a table?
15:33:03Paul_the_Nerdbarrywardell: Well, assuming it's set high enough to assure one final disk write, it seems reasonable to perform the write at that time, then either shutdown or not depending on what preference a user has. Warn seems unnecessary if the write can be assured.
15:33:42bluebrotherI don't like it. It isn't a float, so it doesn't float, and it doesn't break as it's a table.
15:34:02bluebrotherand, it's completely different to all descriptions of that type in the rest of the manual
15:34:11jhMikeSAny new code should take into account the possibility of having all that hardware managed imo
15:34:21LinusNon the other hand, i don't think it's a big deal if the settings aren't saved when the batteri is too low
15:34:33LinusNwe just need to make sure it doesn
15:34:43LinusN't crash because rockbox can't save them
15:34:59LinusNjhMikeS: true
15:35:06*jhMikeS has never used his player long enough in one run to run out the battery.
15:35:22Paul_the_NerdLinusN: If you make sure to get one last write in though, you can save your playlist status, runtime data, and in the case of running a benchmark, a final battery benchmark status, all of which I think are handy.
15:36:05jhMikeSPut the "battery driver" in the target tree :)
15:36:07LinusNthe resume info should have been saved on the last successful spinup anyway
15:36:42Paul_the_NerdLinusN: Yes, but depending on buffer status that could've been as much as an hour ago, sometimes even more.
15:36:54*LinusN tries the scrobbler patch on rec-fm - doesn'
15:36:56LinusNt work
15:37:00 Join NickDe [0] (n=nicholas@
15:37:04LinusNPaul_the_Nerd: yes
15:37:09*Paul_the_Nerd hasn't tried the scrobbler patch yet.
15:37:21JdGordonLinusN: does that mean if it gets working it can be accepted?
15:37:25*JdGordon liked that patch
15:37:34LinusNi want to commit it
15:37:42***Saving seen data "./dancer.seen"
15:38:45barrywardellhaha, just hit low battery had my volume go up to max!
15:39:03Bagderis that the warning? ;-P
15:39:21Bagderthat'll teach you not to go there very often! ;-)
15:40:07barrywardellmakes me really want to get power off on low battery working
15:40:33Paul_the_NerdI think giving the user an option for graceful power down on critical battery is a good thing, is all.
15:40:45barrywardelli'll try to get something working which writes to disk, then shuts down when the battery level gets low
15:40:59Paul_the_NerdJust an option, when enabled, performs a proper shutdown (like holding Stop while in WPS on H100)
15:41:01barrywardelland I'll add a setting for it
15:41:11barrywardellI'll post a patch once I have it working
15:41:27jhMikeSCouldn't there be a broadcast with a system message re: battery events?
15:41:46LinusNi think it should be
15:43:16jhMikeSThat way they can be queued from the GPI0 as well as in other ways.
15:44:33jhMikeSNot to mention an event for charging complete.
15:46:03jhMikeSThe battery scale on the x5 isn't right cause it almost never show a full charge even when it is done charging.
15:49:50 Quit XavierGr ()
15:50:41oboLinusN: here now - how badly does it not work?
15:51:02LinusNobo: hang on
15:54:04 Quit spiorf (Remote closed the connection)
15:58:37Febsbluebrother: two reasons (1) I thought it looked better, and more importantly, (2) I needed a way to separate the general description of the EQ bands from the descriptions of the menu items themselves, for which I use item[ ].
15:59:39amiconnjhMikeS: That ins't necessarily due to the scale; charging monitoring isn't _properly_ implemented on any of the newer targets.
15:59:51amiconnEven some of the old targets need improvement...
16:00:36Febsbluebrother, I don't feel strongly about it, and I kept a copy of the prior version so that I can change that portion back to what it was, if necessary.
16:01:35FebsPerhaps the better approach would be to do two subsections: (1) Description of the EQ itself, and (2) description of each item in the "Graphical EQ" menu.
16:02:58Febs("Graphical EQ" is a bad name for that menu, considering that the submenus contain options for both graphical and non-graphical EQ interfaces.)
16:04:13Paul_the_NerdWouldn't simply "Equalizer" be enough
16:04:37Paul_the_NerdOh, right, on a related note: The custom eq .cfg files are included in the Archos builds (or at least Recorder V1) and I'm guessing this shouldn't be the case.
16:04:55Paul_the_NerdOr perhaps "Software Equalizer"
16:06:36FebsPaul_the_Nerd: those were my thoughts exactly. I think I prefer "Equalizer."
16:07:49lexhey, what's new in the 5.5G ipod? :(
16:08:01lexwhy doesn't rockbox work in it, because of the firmware change?
16:08:14Paul_the_NerdFebs: I prefer "Equalizer" too.
16:08:34bluebrotherFebs, two sections may be indeed better. But the table looks simply wrong there.
16:08:55Genre9mp3Paul_the_Nerd: Is it possible to sort search results in the forums by date posted as well? (instead of only relevance percentage)
16:09:51Febsbluebrother: OK, I'll use two sections. I probably will not be able to get to it for a couple of days though. (PaidWork has been brutal recently.)
16:10:54Paul_the_NerdGenre9mp3: Try the "Advanced Search" link on the search page
16:11:37Genre9mp3Paul_the_Nerd: Oh yes... just found it! Thanks
16:11:40bluebrotherno problem ... it's mostly an optical issue ;-)
16:11:55Paul_the_NerdGenre9mp3: No problem. :)
16:12:34LinusNobo: seems to work, i just didn't realize that the log file is not saved until you shut down
16:13:34 Quit bluebrother ("CGI:IRC 0.5.7 (2005/06/19)")
16:14:42Bagderannoyingly enough, it seems people can screw up their sansas by only changing the mi4
16:15:30Paul_the_NerdThat is Not A Good Thing.
16:15:37Bagderindeed not
16:16:00Paul_the_NerdThough I thought, from reading the thread, that in the cases I read they could still get to recovery mode, but not mount it.
16:16:59Bagderthis guy can even mount it but putting back the good images doesn't magically fix the player, it just doesn't seem to accept them
16:17:05 Quit theli_ua ("Leaving")
16:17:26Bagderto me this looks like a bug in the system
16:18:09BagderI've mailed my SanDisk friend to see if I can get some inside comments on this
16:18:12Paul_the_NerdBagder: It doesn't say that it's updating, or it says that it's updating but fails?
16:18:37Paul_the_NerdI seem to remember there being a little output text there.
16:18:47Bagder"It simply hangs with either a black or a corrupted screen."
16:18:55 Join bawb2 [0] (n=bawb2@
16:19:44Bagderthis is from Peter P, who made the wheel glow
16:19:56barrywardelland he definitely didn't update the BL too?
16:20:03Paul_the_NerdWhat version was he updating with, do we know? I know I've successfully updated / installed an MI4 via the recovery mode on mine, but not while running a non-standard MI4.
16:20:35BagderI've replaced my mi4 with crap and recovered fine every time
16:20:54Bagderso there needs to be more to it that we don't really know at this point
16:20:54dan_aSame here
16:21:17oboLinusN: yes, it will cache upto 32 tracks...
16:21:23Bagderpossibly unclean umounts or whatever
16:21:40Paul_the_NerdBagder: I'm just wondering if the problem isn't the updater but the file being used or the process.
16:22:10Bagderwell, this guy has clues so I don't think it is that simple
16:22:44peturobo: does it flush the cache when the disk spins for another reason?
16:22:56barrywardellmaybe a fluke? maybe he wrote to the bootloader flash by accident?
16:23:24Paul_the_NerdWell even a clue doesn't preclude the occasional user error.
16:23:24Bagderthat's not too likely since we only build mi4 files when we experiment
16:23:37Paul_the_NerdFor example, was he using the MI4 you've got at
16:23:40Bagderbut of course, everything is possible
16:23:56dan_aHe was playing about with accessing different registers/memory areas, so it could have corrupted something
16:24:31dan_aI've suggested reinstalling the bootloader in recovery mode
16:25:03LinusNpetur: no
16:25:36 Nick idnar_ is now known as idnar (i=mithrand@unaffiliated/idnar)
16:25:36obopetur: I just thought that since if it was active and there was something to write.... why not?
16:26:21 Join webguest45 [0] (i=c0647cda@gateway/web/cgi-irc/
16:26:32*petur is confused - so it does write on disk spinup?
16:27:14oboif there is anything in the cache it will write it if the disk is spinning, otherwise it will continue to add items until the max is reached
16:27:41obounless it's a flash based device
16:27:55LinusNobo: really? i didn't see any evidence of that when trying it out
16:28:14LinusNi played a few songs, stopped, and then looked in the file
16:28:23LinusNempty (apart from the header)
16:28:30LinusNthen i rebooted
16:28:35LinusNand the information was there
16:28:48oboit only gets a chance to write while it's being called, so if the disk isn't spinning when the track_changed() event occurs, it won't see it
16:29:23Paul_the_NerdBagder: Do you know if it did the "Reading Main Image" "Writing Main Image" for him, or if the scramble was before that?
16:29:30LinusNobo: i see
16:29:31preglowlcd driver for sansa? woot!
16:29:35BagderPaul_the_Nerd: I don't know
16:29:39preglowmrh truly is a genious
16:29:54Bagderhe's a true magician
16:30:04amiconnLinusN: That would be another case for a registered spinup callback...
16:30:10LinusNamiconn: yes
16:30:12Paul_the_NerdBagder: I think that might be useful to know. ;)
16:30:13LinusNgtg, bye all
16:30:16 Part LinusN
16:30:43dan_aBagder: Is the code I sent OK from a style perspective?
16:31:23amiconnBagder: What about a more flexible way to select alternatives in langv2?
16:31:29Bagderit doesn't use the 4 spaces indent all over though
16:31:53amiconnOne reason I didn't start going through all that stuff is that we can only filter by target
16:32:10amiconnIn case a new target gets added that means reviewing all .lang files...
16:33:19Bagderdo you have an idea how to improve then?
16:33:52Bagderwould we have named features and be able to do strings based on them?
16:34:19amiconnThat would certainly be usefu
16:34:52 Join TeaSea [0] (n=thunderc@
16:34:55dan_aBagder: I deny responsibility for that - I used someone else's LCD driver as a template!
16:34:57amiconnI guess using #ifdef in .lang files would be3 too technical for the average translator?
16:35:02Bagderdan_a: :-)
16:35:50amiconnBut perhaps another file could set those named features based on the #defines set in the config-*.h files?
16:36:05Bagderamiconn: that's indeed a good idea
16:36:55amiconnWe have features which are swcodec only, hwcodec only, MAS35xxF only (not in the core yet), RTC only...
16:37:08 Join elinenbe [0] (i=d1c4c008@gateway/web/cgi-irc/
16:38:17amiconnA thorough .lang cleanup should regain quite some space for archos (andat least some space for all others as well)
16:39:08amiconnI'm not going to break it down to every tiny feature, also because that might breakhw mods
16:39:39 Quit _FireFly_ ("Leaving")
16:40:29 Join Xerion_ [0] (
16:40:44amiconn(e.g. the alarm mod for rec v1, or backlight mod for Ondio)
16:43:46 Join _FireFly_ [0] (
16:44:21Bagdercrontab -r instead of crontab -e is not a good mistake
16:44:28 Quit Xerion (Read error: 54 (Connection reset by peer))
16:44:28 Nick Xerion_ is now known as Xerion (
16:45:06 Quit EspeonEefi ("Leaving")
16:47:19preglowno, it's not
16:47:40Bagderluckily I had just done a crontab -l so I could just copy and paste a new one
16:48:24 Quit markun (Read error: 104 (Connection reset by peer))
16:48:30 Join markun_ [0] (
16:48:37Mikachumy crontab command doesn't have a -r switch it seems
16:49:01Bagderlucky you ;-)
16:49:28dan_aI did "for i in /etc/init.d/* ; do $i status; done" on a machine 500 miles away the other day. And found that /etc/init.d/shutdown doesn't take any arguments.
16:50:13Bagdersome experiences are harder to reach than others
16:50:35lostlogicdan_a: any progress on verifying that the codec thread is indeed running on the COP and that it is also indeed running at the same time as other threads, not just linearly on two cores?
16:51:02dan_alostlogic: No, I've been playing around with the Sansa, I'm afraid.
16:51:59 Quit JdGordon ("Konversation terminated!")
16:54:17 Join psiborg [0] (
16:55:27 Join rretzbach [0] (
16:57:29 Quit Febs ("CGI:IRC (EOF)")
16:58:54 Join mra [0] (i=mra@nat/hp/x-d6272855da425bb7)
16:59:39 Join Kartun [0] (i=53a77092@gateway/web/cgi-irc/
17:00:40jhMikeSOn .lang: I haven't really observed myself but all language strings are included in a target even if not used in the c?
17:00:54Kartunhi everyone
17:00:54 Quit Kartun (Client Quit)
17:03:44amiconnjhMikeS: yes
17:04:12amiconnThat's one of the problems langv2 should (and can) solve
17:04:26amiconn...but so far nobody started the necessary rework
17:04:46jhMikeSamiconn: that's what's in use now?
17:04:49amiconn...partially because the filter options are limited
17:05:10 Join Kartun [0] (n=Kartun@
17:05:15amiconn...and partially because it's _a lot_ of work
17:05:23Kartunhi everyone
17:05:41KartunHow could I upload my Skin to ?
17:06:27 Quit rretzbach (Connection reset by peer)
17:06:30 Join rr3tzbach [0] (
17:07:01peturKartun: give us the wiki name you registered
17:07:04jhMikeSamiconn: What sort of method for filtering is there? The filtering could be done by programmers so translators don't have to worry about it if that's a concern.
17:08:40jhMikeSreading back up I guess you're considering something like that already :)
17:09:29peturKartun: ?
17:09:46pondlifelostlogic: Quick question of the day?
17:10:01pondlifeCould you have a look at
17:10:58pondlifeI'm wondering why line 953 needs to set tracks[track_ridx].buf_idx = 0.
17:10:59peturKartun: you now have wiki write access
17:11:18Kartun10x a lot ... will try :)
17:11:23 Join linuxstb [0] (i=3f82c640@gateway/web/cgi-irc/
17:11:51pondlifeJust trying to seperate the filling (track_widx/buf_widx) from the playback (track_ridx/buf_ridx)...
17:12:18lostlogicpondlife: that's just sorta a "just in case the buffer got effed, since we're rebuffering any way, let's just reset the buffer pointers entirely
17:12:35pondlifeOK, not something that is really needed then
17:12:59pondlifeRebuffer will overwrite that when it becomes the track to write.
17:13:01 Join webguest77 [0] (i=54997076@gateway/web/cgi-irc/
17:13:39lostlogicif everything is working correctly, then it doesn't actually matter to what location on the ring we start filling nor start reading as long as before any reading is done they are the same point
17:13:52webguest77where do i get the windows installer for archos, i want to instsll rockbox on my unit and it says i need to use windows installer which i cant find on the wiki
17:14:06pondlifeTrue. My current version just resets _widx = _ridx and never zeroes them
17:14:51lostlogicyep seems fine to me... unless we want talk about the small extra cost of reading over the ring boundary that will happen sooner :-P
17:15:01 Quit webguest45 ("CGI:IRC")
17:15:11pondlifeBut I don't entirely understand this rebuffer fix (mainly the overlap stuff further down), and only use MP3 myself... so am a bit concerned!
17:15:28linuxstblostlogic: When you run dan_a's COP patch, do you get both a main and a codec thread running on the COP? i.e. listed as (1) in View OS Stacks?
17:15:45lostlogiclinuxstb: yeah
17:15:59webguest77jesus, i jave RTFM and I still cant figure out how to install rockbox on my archos player, great manual you have
17:16:01 Part webguest77
17:16:02pondlifelostlogic: Tomorrow's question will be "Why do we need a guard buffer?"... :-)
17:16:13lostlogicwant the answer now, or want to wait until tomorrow?
17:16:19pondlifeNow, if you can!!
17:17:01pondlifeIn 50 words or less
17:18:16lostlogicthe guard buffer allows us to guarantee a certain minimum block size to the codecs, which is needed, because I believe some codecs will don't understand about ring buffers and if they don't get the amount of data they want contiguously the first time they will re-request, expecting that whatever reading process they are waiting on will eventually have that block available
17:18:57pondlifeOK, so it's just a mechanism by which we don't use the whole ring buffer in case that leads to that situation, right?
17:19:33pondlifeSo we just insert data from the guard buffer at wrap round if needed.
17:19:52lostlogickinda other way around
17:19:54Kartunpetur: and how do I upload my screenshot/wps ?
17:20:14pondlifeAh, I see - it's outside of the ring buffer really
17:20:21lostlogicwe never fill the guard buffer until the codec makes such a wrapping request, then we copy from the buffer start into the guard buffer to provide contiguous memory to the codec
17:21:00pondlifeOne more? audio_read_file - what is the purpose of "overlap"?
17:21:13lostlogichmm, that one I'll hafta actually go to code for, 1s
17:21:14peturKartun: look at the page when you edit it. Certain image formats are shown inline - which is what you want
17:21:21jhMikeSencoders already do that at wraparound :)
17:22:40peturKartun: click 'Attach' on the page where you want to add your screenshot, and read the notes that are at the bottom
17:23:45peturthere's also a lot of info here:
17:23:46 Join dpr0 [0] (
17:24:08lostlogicpondlife: I don't think overlap is my code, but it looks like it is the amount of data in the current read that is going to write into the currently playing track, so we can change the currently playing track's start position to not be zero, incase there is a seek backwards.
17:24:29lostlogicit also looks like the code should be refactored so that there aren't two blocks in a row that start with if (overlap > 0 &&
17:25:28 Quit Siku ()
17:25:33lostlogicyeah, this looks like someone enhanced a hack of mine which would actually update the start_pos at the start of the fill with this improved version...
17:25:45lostlogicI wonder what triggers end-of-fill now *looks*
17:26:55 Quit Kartun (" HydraIRC -> <- Try something fresh")
17:27:36lostlogichmm, would be better to not have the fill_bytesleft variable and to do a smarter end-of-fill trigger, now that we have that overlap logic in there... just check if where a full 32k block read would fill to would write over the playing position and stop if it would, otherwise do the read and then check the overlap
17:27:54lostlogicwould eliminate another file-scoped variable.
17:29:02 Join Siku [0] (
17:29:33 Quit lee-qid ("Trillian (")
17:30:04pondlifeSorry, got called away
17:30:38pondlifeIf you could add some comments to that overlap code, that would be nice. It's rather mysterious and I think it results in a bug..
17:30:45*pondlife looks for the Flyspray number
17:31:12 Quit dpro (Read error: 60 (Operation timed out))
17:33:18lostlogicpondlife: ah, sure, assign the task to me and I'll work on what all I described above over the next few days... I'm determined to get my butt back into coding rockbox this month ;)
17:33:36mramy iPod has bad blocks
17:34:03mracan I use RockBox and somehow mark sectors bad and that they should be avoided?
17:35:53pondlifeI'll carry on digging and just leave that overlap code alone...
17:35:57pondlifeThanks, bye for now
17:37:16 Quit falkyre (Read error: 104 (Connection reset by peer))
17:37:45***Saving seen data "./dancer.seen"
17:40:24 Join San2 [0] (n=Wonderbo@
17:40:43San2can the iPod Video Record audio?
17:41:20linuxstbSan2: Yes. There's a work-in-progress patch on flyspray for it.
17:42:12linuxstbYou need to build or buy line-in adapter though - the line-in pins are in the dock connector on your ipod.
17:42:19 Join SereR0kR [0] (
17:42:30San2oh, thats fine
17:42:31San2thank you
17:43:26San2And it will have the same recording capability's as the H300 has atm?
17:44:29 Join rre7zbach [0] (
17:44:29 Join Ribs [0] (n=freenode@
17:45:26linuxstbSan2: Eventually, yes. In fact it should be better than the H300 - the ipods should be able to record at a wider range of samplerates, up to 96KHz, and probably also 24-bit.
17:46:20San2Thats good. :)
17:46:25San2thanks :D
17:46:43 Quit San2 ()
17:49:13 Quit dpr0 (Read error: 110 (Connection timed out))
17:50:53linuxstbdan_a: I'm not sure yet what changed, but I just tried your COP patch again on my 5g, and it appears to work fine - I get a main thread on the COP...
17:52:03dan_aExcellent. It seems to behave... strangely, as you may have seen from the logs
17:52:27*jhMikeS has to implement encoder bit bepth now :0 :)
17:53:11linuxstbYes, it does seem odd. For example, in the audio debug screen, if I play with the wheel, the PCM buffer empties. If the codec thread is indeed running on the COP, I'm not sure why lots of wheel events and screen updates would affect it.
17:53:31Paul_the_NerdThat seems to support the linear execution theory
17:54:43dan_aYou should see what happens when you get data aborts now....
17:55:23amiconnlinuxstb: Perhaps interrupts are routed to both cores? (just a guess)
17:55:38linuxstbAh, it seems the keybindings in the audio debug screen have been changed - the wheel is skipping tracks....
17:56:31dan_aPaul_the_Nerd: The thread dies, but everything else carries on running. Then after a while, the thread sometimes comes back to life
17:57:37Paul_the_Nerddan_a: That sounds... fun.
17:58:09*linuxstb wonders why the audio debug screen doesn't use the WPS context for checking actions such as next/prev track...
17:58:11dan_aamiconn, linuxstb: Yes, it could be an interrupts thing. I wasn't sure what the correct behaviour for the interrupts should be, so I just did what felt right.
17:59:37 Join rretzbach [0] (
18:00:12amiconnI think the cop shouldn't receive any interrupts
18:00:37linuxstbHas anyone done a runtime test yet with the codec thread on the COP?
18:00:39amiconnHmm, but it should wake up from the timer tick when it's sleeping...
18:00:57 Quit petur ("*poof*")
18:01:15 Quit rr3tzbach (Read error: 110 (Connection timed out))
18:01:19dan_aThat should be set in the interrupt controller, though, not in the IRQ handler if I understand things correctly
18:01:37 Join blueworm [0] (
18:01:50dan_alinuxstb: I haven't yet.
18:01:53linuxstbI was probably wrong about the wheel events causing the PCM buffer to empty - I didn't realise it was skipping tracks...
18:01:54PaulJampondlife: a little update on the testing: with a sim from 2006-08-20 01:00 i get both problems, the stopping and the whiteniose. with a build from 2006-08-15 01:00 i get only the stopping, and not the whitenoise after rebuffering.
18:01:59 Quit bawb2 ("seacrest out!")
18:02:23 Join BigBambi [0] (
18:02:49pondlifePaulJam: Aha
18:05:05PaulJamthere was one commit on playback.c in this time. the next thing i'll try is a build from aug. 15. 18:05.
18:05:12pondlifeIf you could try the 16th, 17th, 18th - I'd really like to know which of those days it was.
18:05:26pondlifeI wonder if there is an underlying issue that the scheduler has exposed?
18:09:18pondlifeThe range 15th-18th is looking quite common
18:10:31 Quit obo ("bye")
18:11:35PaulJamthis is august, the scheduler was committed in september
18:12:09dan_aTime for me to go home, and then get this LCD driver comitted!
18:12:10pondlifeGood point! I was refering to Sept
18:15:06pondlifeHere's a thought. Can you try the current CVS build, but go into playback.c and comment out the #include PLAYBACK_VOICE line?
18:15:17pondlifeThis will remove plenty of code from playback.c..
18:15:49pondlife#define I mean, of course...
18:16:00*pondlife is too close to the code
18:17:04*jhMikeS understands that one :)
18:17:40PaulJamok, i can try that.
18:17:44pondlifeThe only playback change in 15th-20th August is
18:18:46 Join YouCeyE [0] (n=YouCeyE@unaffiliated/youceye)
18:18:47 Join bluebrother [0] (
18:18:48 Quit _FireFly_ ("Leaving")
18:19:14 Quit blueworm ("Leaving")
18:20:29lostlogicpondlife: is your RINGBUF_ADD macro safe for use in an if?
18:20:59pondlifeShould be. I put the whole thing in brackets.
18:21:15lostlogiccuz I'm going to use it in an if...
18:21:34 Join pixelma [0] (i=pixelma@
18:21:37pondlifeHmm, did I miss some arithmetic, or is this new?
18:23:24 Join Febs [0] (i=2662c44b@gateway/web/cgi-irc/
18:26:55 Quit rre7zbach (Connection timed out)
18:27:28lostlogicpondlife: herm? I haven't _used_ it, I'm _about to_ I think it'll work fine too, was just checking with you first ;)
18:27:52 Quit elinenbe ("CGI:IRC")
18:28:59bluebrotherI think we should reject this task (word viewer) because it's a strange, undocumented format.
18:29:03bluebrotherwhat do you think?
18:29:19Mikachui would agree
18:29:41hcsI think it should be rejected because of the use of the word "microoft"
18:29:42Mikachuyou can give him a link to wvware
18:31:31jhMikeSHum...I'm not sure why it would be helpful since viewing text files is next to useless enough on small screens.
18:31:54Paul_the_NerdViewing .docs would requires supporting .ttf and various other things.
18:32:12Paul_the_NerdOr would it just be extracting the text out of it, with very minimal formatting?
18:32:40PaulJamisn't it possible to show a .doc as plain text only
18:32:48jhMikeSCould extract the text but I wonder if fully formatted tables are expected
18:33:09 Quit dan_a (Read error: 110 (Connection timed out))
18:33:17bluebrotheron larger screens a html viewer would be nice. It could also help browsing the manual on the device.
18:34:13Mikachuand embedded jpg images could be a problem
18:35:27jhMikeSwe've got a jpeg decompressor already
18:35:35 Quit linuxstb ("CGI:IRC")
18:36:28bluebrotherI don't think a basic html viewer would need support for images.
18:36:44bluebrotherit would be nice, though
18:36:49 Join Lear [0] (
18:37:44 Join Arathis [0] (
18:38:09 Part Paul_the_Nerd
18:40:45 Join Davide-NYC [0] (
18:41:43jhMikeSI think we should have an image handling library and therefore a more comprehensive image viewer plugin
18:44:21Rob2222_Is here anyone in from the Unitet States?
18:45:07hcsRob2222_: I am, why?
18:45:40 Join dan_a [0] (n=dan-mirc@
18:46:22 Nick Rob2222_ is now known as Rob2222 (
18:46:31 Join TeaSeaLancs [0] (n=thunderc@
18:46:48lostlogicpondlife: the overlap logic was indeed flawed, but my adjustment to buffering termination will fix it any way.
18:47:18pondlifePaulJam: bad news!
18:47:38 Quit NickDe ("Leaving")
18:48:07pondlifeThe playback.c commit on the 15th August was indeed bad - and could result in memory corruption = white noise
18:48:25pondlifeThis was then fixed by the commit on the 23rd August
18:49:02pondlifeSo, if I understand correctly, your white noise may be from a different cause to that in the current CVS.
18:49:29 Join NickDe [0] (n=nicholas@
18:49:41 Quit TeaSea (Read error: 110 (Connection timed out))
18:49:46 Quit Rob2222 (Read error: 104 (Connection reset by peer))
18:50:28 Join Rob2222 [0] (
18:50:43Rob2222re :/
18:50:44pondlifeOf course, the slopping problem seems to have been there throughout?
18:50:47jhMikeShcs: so they can come in with a swat team on a no-knock raid...why just cause. Maybe they'll find something.
18:50:59jhMikeSslopping! :D
18:51:11pondlifeHmm, a new feature, I fear!
18:51:25pondlifeRockbox v3.1 - now with random slopping
18:51:36jhMikeSIf it can feed the cat and clean the litter too I'd be happy
18:52:19lostlogicpondlife: what do you think about a "RINGBUF_CROSS(i,j,x) (i<j?i+x>=j:i+x-filebuflen>=j)
18:52:21pondlifelostlogic: Is your update coming to CVS soon?
18:52:43lostlogicpondlife: haven't finished the code nor started testing, so you go ahead if you have some stuff ready.
18:52:45 Quit rretzbach (Read error: 104 (Connection reset by peer))
18:52:53PaulJambtw, the stopping problem is still there with the voice line commented out, but i guess this was more or less expected.
18:52:55pondlifeNo, just wanted to see it!
18:53:17pondlifePaulJam: Well, it proves it's not the voice code doing it.
18:53:40pondlifelostlogic: RINGBUF_CROSS - does this detect wrap?
18:53:43jhMikeSyeah you'd think voice would be more trouble than it really is
18:54:23lostlogicpondlife: it detects when the event "buffer pointer i moves x crosses j"
18:54:54pondlifeNope, still don't get it! What do you mean by "crosses"...?
18:55:07PaulJampondlife: so should i test a build from aug 23. 08:30 to see if this commit (temporarily) fixes the whitenoise problem?
18:55:39preglowwhitenoise problem? :>
18:55:43pondlifeYes, although it may be that the problem you're seeing is there too!
18:55:46 Join petur [0] (n=petur@rockbox/developer/petur)
18:55:53lostlogicpondlife: hmm −− so I have the question: when I advance the buffer by x, from i, will I be before or after j? and this macro answers that question
18:56:03pondlifeAha gotcha
18:56:23pondlifex = the increase... I didn't get that bit!
18:56:35pondlifeRINGBUF_ADD_CROSS maybe?
18:56:39jhMikeSwhat's j?
18:56:42 Quit lodesi (Read error: 113 (No route to host))
18:57:07pondlifej is the point you're checking against.
18:57:09lostlogicjhMikeS: that's a good question −− in my initial use of the macro, j will be the current read position of the buffer
18:57:17pondlifeThe finishing line, if you will
18:57:30 Quit godzirra (Read error: 104 (Connection reset by peer))
18:57:33pondlife#define RINGBUF_PHOTO_FINISH...
18:58:13lostlogicpondlife: I'll redefine the macro arguments to be consistent with your naming: p,v,o?
18:58:28jhMikeSi is the write pos?
18:58:36pondlifeo? or p1/p2?
18:58:37lostlogicin the initial use, yes
18:58:40lostlogicp1/p2, sure
18:58:40 Join lodesi [0] (n=lds@
18:59:07pondlifeNever was a fan of o for pointers. Have seen it many times though.
19:02:23 Quit Davide-NYC ("Chatzilla 0.9.75 [Firefox]")
19:02:25jhMikeS(wr_pos < rd_pos ? wr_pos+bytes >= rd_pos : wr_pos+bytes-filebuflen >= rd_pos) ??
19:02:38pondlifeReadable, aren't they!
19:03:22*amiconn wonders what association lets a dev use 'o' for a pointer...
19:03:50jhMikeSstands for "ouch it's pointy"?
19:03:52*pondlife won't name names...
19:03:52 Join spiorf [0] (
19:04:26jhMikeSor maybe "offset"
19:04:33pondlifeOr "overwrite randomly"
19:05:03jhMikeSheh...let's have a version of the sliding puzzle where you put ram back in order
19:05:16pondlifeActually "offset" is the reason. But they are often converted to pointers without changing the name.
19:05:23pondlifeI think that's what I've seen.
19:05:34amiconnjhMikeS: Heh, pointers are neat
19:06:41lostlogicjhMikeS: how about start, mark and bytes?
19:07:05 Quit MarcoPolo (Read error: 104 (Connection reset by peer))
19:08:06 Join dpr0 [0] (n=x@
19:09:09 Join ender [0] (i=null@
19:09:16PaulJampondlife: little update: with voice commented out, the whitenoise problem still exists (tested on target)
19:09:45pondlifeOK. I thought it might. Current code I assume?
19:09:59amiconndan_a: Do you have a scan of the iPod 3G front?
19:10:06pondlifeCould you try the 24th Aug?
19:10:15 Nick ender is now known as ender` (i=null@
19:10:23dan_aamiconn: Sorry, I don't
19:10:38dan_aAnd I don't have access to a scanner either
19:10:40pondlifeI suspect this is a long-standing issue,sadly. This could disprove that theory.
19:11:10PaulJami'm currently testing on a sim from aug 23. 08:30
19:11:42pondlifeOK. Can you repro with a short album? Or do you have to do a full buffer's worth perhaps?
19:11:52 Join rretzbach [0] (
19:11:53pondlifei.e. is it possibly linked to buffer wrap?
19:12:51pondlifeYou're certainly powering through the tests quicker than I thought possible!
19:13:03PaulJamthe first album has 4 tracks (23,4 MiB)
19:13:04 Join MarcoPolo [0] (
19:13:17jhMikeSi'm sure how that macro will detect buffer overflow reliably...maybe not trying hard enough
19:14:08PaulJamthe second album is longer, because the whitenoise problem happens after buffer refill.
19:14:32lostlogicjhMikeS: it doesn't detect overflow, actually I just modified it so it returns the number of bytes by which you will cross the mark
19:14:55 Quit _Lucretia_ (
19:14:59lostlogic'returns' I always say that, but that's not how macros work
19:15:59pondlifeI wasn't sure on the virtues of macro vs. function vs inline function on this one. Are you happy with macros?
19:16:50*pondlife wonders what isn't being reset after playback finishes naturally, and playback is then started again...
19:17:18jhMikeSlooks like it returns bool
19:23:53 Join acknabar [0] (
19:23:54 Join wooo [0] (
19:26:02dan_aSansa updates ahoy!
19:27:15jhMikeSyeah, I guess it works...have to be patient with me...I'm kinda mentally challenged :)
19:32:48barrywardellnice updates!
19:33:04barrywardellis that the same ADC driver as in the H10? does it work?
19:33:05 Quit SereR0kR (Read error: 104 (Connection reset by peer))
19:33:45NJoin_Lucretia_ [0] (
19:33:57dan_abarrywardell: I've used the ADC driver from the H10, but I don't know if anything uses it. Nothing seems to.
19:34:21barrywardellah, ok.
19:34:38barrywardelli'd imagine at least the battery should
19:35:03dan_aThe mic must be connected to something, too.
19:35:45barrywardellthe mic on the H10 isn't connected to the ADC
19:35:52dan_aIt's so much easier to figure things out when you've got an LCD screen!
19:36:05dan_aHow is it connected on the H10? Or don't you know yet?
19:36:22barrywardellit's connected to the WM8731
19:36:34barrywardellthe recording patch in the tracker works with the mic
19:36:41dan_aOh, of course.
19:37:48***Saving seen data "./dancer.seen"
19:38:55*barrywardell has a sansa on order :)
19:39:30Arathisbarrywardell: hope you don't switch to sansa_only than ;)
19:39:50barrywardellboth are quite similar i think
19:40:22Arathisgreat :)
19:41:02*Arathis hopes for even better development because of more (?) devs :D
19:43:36 Join leftright [0] (
19:44:04leftrighthi, i want to flash my recorder, but the url returns 404
19:44:59amiconnleftright: Those images are taken down on purpose, they're way outdated: Bootbox is the supported method now:
19:45:09leftrightthe diffrenece between the Original firmware and Rockbox on the recorder is Day and night, trully amazing
19:45:19leftrightthanks amiconn
19:46:33leftrightthat backfround noise with the original fimware is gone with rockbox, i dont know how you go it but its magic, my unit is a fairly old one but still sounds good, sounds different to my H140
19:48:39*amiconn thinks the sim could simulate dual-core operation...
19:49:32amiconnCurrently, the sim mimicks rockbox threading by mutexing sdl threads. For dual core, there could be 2 mutexes, one per simulated core
19:50:04PaulJampondlife: on a sim from 23 aug at 8:30 (directly after the fix) the whitenoise problem still exists.
19:50:59pondlifeGood ;-)
19:51:02pondlifeI guess
19:51:21pondlifeOK, so how about if you go back to before the Aug 15th? Or did you already do that?
19:51:49amiconndan_a: What's the 3G's backlight colour?
19:52:22dan_aamiconn: Light blue - I think it's the same as the 4G
19:52:36amiconnSo it's different from the mini
19:52:55PaulJamyes, on a sim from aug 15. 01:00 and from aug 10. 01:00 the whitnoise problem didn't happen. (but the stopping happened)
19:53:05dan_aI've not seen a mini
19:53:15amiconnThe mini has white backlight least the 2G
19:54:48PaulJamso it seems that it was a commit on aug 15. between 01:00 and 18:05
19:54:49 Join D_C [0] (
19:55:10pondlifeAh, so it must be the voice mod
19:55:16pondlifeCan you open playback.c up...
19:55:34 Join drippydonut [0] (
19:55:42PaulJamon a current cvs?
19:55:47leftrightamiconn, do i install Bootbox then Rombox, or the other way around
19:55:55pondlifeDon't rush, I need to think a bit..
19:56:31pondlifeAh, better idea
19:56:40pondlifeGo to any "bad" version
19:56:57pondlifeAnd run it with no .voice file present
19:57:27PaulJami have done all the tests with no .voice file
19:57:37pondlifeAh... ok.. Weird
19:57:53pondlifeWell, plan B then. On the current playback.c
19:58:08PaulJamon target i have a voice file, but it is located under /Daten/
19:58:26pondlifeOK. Can you try WITH a voice file and see if that helps
19:58:35pondlifeMaybe that's why I'm not able to repro it
20:00:20D_Cdo you think it would be possible for rockbox to have a pdf viewer
20:00:29D_Cor is that too intensiv
20:01:38 Join Rondom [0] (
20:02:38pondlifeI've got to go now, unfortunately, but if you'd like to try something else, go back to the CVS build and look in the current playback.c routine audio_reset_buffer(). Then remove the else clause ("Allow for 1 codec...") completely and see if that fixes it in your original test (without voice file).
20:02:40PaulJampondlife: any specific voicefile?
20:02:53pondlifeDon't think so
20:03:44pondlifeMine is 1,533,792 bytes - not sure where it's from though!
20:05:03pondlifeJust noticed you said the offending commit must have happened before 18:05 on Aug 15...
20:05:08 Join charkins [0] (n=casey@pdpc/supporter/active/charkins)
20:05:32pondlifeWell this work was after that.
20:06:28pondlifeForget me... I am in the wrong time zone
20:06:33bluebrotherD_C, I don't think it's impossible (at least on non-archos targets) but it would be quite some work. I don't think someone will pick this up soon
20:06:40 Join GreyFoux [0] (
20:06:43pondlifeIt was at 18:01!
20:07:37 Quit Ed (Nick collision from services.)
20:07:52 Join Ed___ [0] (i=ed@
20:09:42 Join GrumpyPiano [0] (n=select@
20:09:46 Join lightyear [0] (
20:11:11GrumpyPianoHi. I've a 3G iPod. And I'm wondering what would be the best audio format to use. That wouldn't have problems with buffer underruns.
20:11:41pondlifeHmm, is filebuflen -= CODEC_IRAM_SIZE + CODEC_SIZE;
20:11:46pondlifethe same as filebuflen -= (CODEC_IRAM_SIZE + CODEC_SIZE);
20:12:12leftrightamiconn: i flashed my recorder, and I get the following message on boot "File type arrayfull", is this bad ?
20:12:23pondlifeWelll, it's not that then...
20:12:33amiconnYou probably didn't flash a recent build yet
20:12:48dan_aGrumpyPiano: Anything that doesn't compress very much (eg Flac.) The best compression/performance ratio seems to be for Musepak
20:12:56amiconnThe flash images on the bootbox page contain a rockbox build from when these images were made.
20:13:16GrumpyPianodan_a: Ok thanks I'll try that :)
20:13:32amiconnYou already flashed that image, now run rombox.ucl or rockbox.ucl in order to bring in the current build
20:13:37lostlogichmm, I wonder if this will work at all.
20:14:15pondlifelostlogic: I dare you to scrap buf_ridx completely :-)
20:14:33amiconnThe description on the FlashingRockbox page is basically still correct, only that we no longer use the archos firmware as fallback firmware, but bootbox, and that there is rombox
20:14:34lostlogicpondlife: I'll leave that up to you. I _use_ buf_ridx in this code, so :-P
20:14:50amiconn(but today's daily breaks rombox for recorder v1) :-(
20:15:02pondlifeI tried and failed, but am still unsure exactly why buf_ridx is needed.
20:15:27lostlogicpondlife: only because otherwise buf_ridx = CUR_TI->buf_idx + ci.curpos and that calculation sucks
20:15:45pondlifeThe funny thing is that's not always true
20:15:51lostlogiccuz of codecs
20:16:03pondlifeDo a seek and it ends up askew, even with codecsize=0.
20:16:18pondlifeThat is they mystery I'm left with
20:16:23preglowdan_a: does musepack run well for 3g?
20:16:27pondlifeIt's probably obvious when you know about it
20:16:50pondlifeBut conceptually I don't see the need for buf_ridx. Just ci.curpos should deal with most stuff.
20:16:55lostlogicwell it plays music which is honestly better than I thought I'd do in my first shot
20:17:06pondlifeWhat have you done..?
20:17:09lostlogicpondlife: but ci.curpos is relative to the current track
20:17:18lostlogicahh, I know why it doesn't work
20:17:24leftrightamiconn; so if i rolo on the build i have it'll flash it ?
20:17:28dan_apreglow: I tested just about everything when we got sound on the 3g, and from memory, yes, musepack didn't skip
20:17:35lostlogicbecuase we move buf_idx when we overwrite part of a track, but ci.curpos is always relative to the _start_ of he track
20:17:47preglowdan_a: how nice!
20:17:48pondlifeYes, but we have .start_pos
20:18:01lostlogicso it's buf_idx + curpos - start_pos
20:18:02pondlifeOr is that a file offset.
20:18:06lostlogicthis gets prettier all the time
20:18:23pondlifeI just think it should be possible, conceptually.
20:18:38dan_aGrumpyPiano: It also helps if you go into the menu screen while you're listening to music
20:18:58pondlifeMaybe ci.curpos should be relative to the buffer, not the track anyway?
20:19:07pondlifeNo, don't change it...
20:19:14lostlogichaha, not now any way
20:19:41pondlifeOK, so what have you found and changed that is likely to break playback so much??
20:19:46lostlogicpondlife: well for the first time in a long while, it's my turn to post a patch for testing
20:20:01pondlife\o/ The master returns....
20:20:15lostlogicjust removed the use of fill_bytesleft
20:20:50pondlifeI haven't looked at the filling much yet. Just assumed that worked once it was triggered...
20:21:26lostlogicbut it was apparently cuasing bugs, so ... :)
20:23:42pondlifeSadly I have to go now. Will try it in the morning. :(
20:23:58lostlogicany other playback testing enthusiasts? PaulJam?
20:24:30 Quit dpr0 (Read error: 110 (Connection timed out))
20:24:41lostlogicsee ya pondlife
20:24:49pondlifeI've given PaulJam strict instructions not to branch from CVS until he can confirm that my voice patch was to blame :)
20:25:30*amiconn wonders why his last commit didn't trigger a rebuild
20:25:37*pondlife couldn't resist and applied the patch
20:26:30pondlifeOops, it fills forever here.
20:27:01lostlogicpondlife: as in it overwrites itself over and over?
20:27:18pondlifeNot sure, just fills and fills
20:27:56pondlifeRecipe is on
20:28:21pondlifeI had a similar problem where I had a <= rather than a < on a filebuflen check
20:28:29pondlifeWhen I was tinkering I mean
20:28:38pondlifeOK, I really do have to go now.
20:28:54 Quit pondlife ("...and is in Trouble")
20:29:39 Quit Seed (Remote closed the connection)
20:29:43amiconnAh, now it started...
20:30:10 Quit edx (Read error: 113 (No route to host))
20:30:53jhMikeSthe build server is becoming sentient and has devoloped a sense of humor
20:31:11amiconnBagder: What does the buildmaster do in the 4 extra minutes when the build itself already finished?
20:34:43 Join halfss [0] (
20:36:30 Nick dan_a is now known as dan_a_afk (n=dan-mirc@
20:37:17jhMikeSIt thinks about inventing time travel to back to the 80s and kill Sarah Connor
20:38:02 Part halfss ("Ex-Chat")
20:38:08*preglow finally realises the rockbox plot for starting to use distributed computing
20:38:45 Part GrumpyPiano
20:39:36jhMikeSIt's the endtimes and the build master will become our master and we shall gladly serve our build master
20:41:26bluebrotherserve him by doing code commits? ;-)
20:41:47jhMikeSIt is 0xfood for the build master
20:42:41 Join SereR0kR [0] (
20:44:30 Join Seed [0] (
20:44:32 Quit barrywardell ("CGI:IRC")
20:45:36amiconnAm I the only one who observes very unreliable USB connection on H1x0/H300/X5 ?
20:46:21jhMikeSx5 doesn't have a reliable usb connect no matter can startle it easily. Must move slowly...
20:47:38amiconnIt happens very often that the USB logo appears, but the PC doesn't mount the drive. Works everytime in H1x0/H300 bootloader USB, or cowon bootloader USB
20:48:19 Quit TeaSeaLancs (Read error: 110 (Connection timed out))
20:49:04jhMikeSyup. If you're using debian you must be focused to some other window program too or else debian gets the connection.
20:49:07Learamiconn: no problem whatsoever on my h140...
20:49:35amiconnThis is in windows. Never had any connection problems from the PC side
20:49:44jhMikeSbut I has probs before that when pluggin/unpluggin too fast
20:49:56 Quit Seed ("cu, Andre")
20:50:09amiconnThis seems to be a rockbox problem
20:50:13LearWindows here too. I do use the bootloader more often than not, but still...
20:50:30amiconn...only happening on swcodec (except ipod which reboots into apple disk mode)
20:50:34jhMikeSI have the same stuff with the ROM boot when pluggin
20:51:32jhMikeSSWCODEC used to swap codecs out but I think that is gone. It's set up that if all thread don't ack the connect the connect never happens?
20:52:00 Join Seed [0] (
20:52:06amiconnAll threads with a public event queue must acknowledge the connection
20:53:51jhMikeSThe codec thread can't if a codec is loaded and running. If voice is swapped, it can't either. hmmm.
20:54:46amiconnDoes the codec thread have a public event queue?
20:55:22jhMikeSyes it seems. it has codec for usb connect events or had. I haven't looked at what changed except the swapping was removed.
20:55:38amiconni.e. one that is initialised with queue_init(...., true);
20:56:52 Quit dan_a_afk (Read error: 60 (Operation timed out))
20:56:57jhMikeSyes, it seems they all are in playback.c on quick reexamination
20:57:24jhMikeSexcept codec_callback_queue...oops
21:02:45 Quit SereR0kR ()
21:03:00 Join webguest99 [0] (i=3eebdcad@gateway/web/cgi-irc/
21:04:32jhMikeSamiconn: think there should be a memswap function to swap codecs with one buffer? Could be useful elsewhere too.
21:05:09 Join ben__ [0] (
21:05:11 Quit lightyear (Read error: 110 (Connection timed out))
21:05:49 Quit Seed (Nick collision from services.)
21:05:52 Join amiconn_ [0] (n=jens@rockbox/developer/amiconn)
21:05:56 Join Seed [0] (
21:08:55 Join amiconn__ [0] (
21:10:39 Join falkyre [0] (n=falkyre@
21:10:53 Quit Rondom (Nick collision from services.)
21:11:00 Quit falkyre (Client Quit)
21:11:03 Join Rondom [0] (
21:15:05 Join Will [0] (
21:17:13 Quit acknabar ("Leaving")
21:17:27 Quit amiconn_ (Read error: 60 (Operation timed out))
21:18:23Willhey is there a way to have rockbox start in a default folder i.e. where my music is stored?
21:20:39 Part Will
21:21:09jhMikeShmmm...impatient. was gonna tell 'm
21:21:17 Join Will [0] (
21:22:27 Join drezha [0] (
21:22:28peturWill: no, that's not possible
21:22:34 Quit amiconn (Read error: 110 (Connection timed out))
21:22:35 Nick amiconn__ is now known as amiconn (
21:22:44*petur looks at jhMikeS ;)
21:22:49drezhahi all
21:23:19peturhi there
21:23:22jhMikeSwherefore, sir petur?
21:23:41peturjhMikeS: "<jhMikeS>hmmm...impatient. was gonna tell 'm"
21:24:08drezhathought you guys might be able to help...
21:24:08drezhabeen considering changing my MP3's to Ogg Vorbis now I'm mainly using linux and had a couple of questions about it
21:24:09jhMikeSWell he waited about 60s and split
21:25:31jhMikeSIf you want answers you have to wait till the season is right for the tree to bear fruit. :P
21:25:50drezhaany ideas on when that'll be? :p
21:25:56*petur gives up and minds some code
21:26:31jhMikeSpetur: never give up, never surrender!
21:26:45preglowdrezha: asking would be wise if you're after answers
21:27:44jhMikeSthe prerequesite for an affirmative or negative is an interogative =:-D
21:28:08drezha1. from a brief look at it appears that I cant run Rockbox on my IHP-100
21:28:15drezhathis true?
21:28:35markun_no, not true
21:28:52 Nick markun_ is now known as markun (
21:29:02drezhano? good. So installing Rockbox, I'll be able to convert my MP3's to Ogg's and still have them play?
21:29:17preglowdrezha: well, you can have it play your mp3s as they are too
21:29:28preglowconverting from one lossy format to another is never wise
21:29:30markundrezha: I wouldn't convert them
21:29:35jhMikeSpreglow: thanks :)
21:29:58markundrezha: which iriver do you have exactly?
21:30:19drezhathe back of it says the iHP-100
21:30:31markunhow many gigabite?
21:30:32preglowjhMikeS, the king of semantics
21:30:36jhMikeSIt's also unwise as well
21:30:48 Quit NickDe ("Leaving")
21:30:50drezhaslightly old now
21:30:59markungigabyte even :)
21:31:41 Part Will
21:31:54drezhamy second question was, will using Ogg use less space for the same quality?
21:32:01drezhaI've never used or heard an ogg play before
21:32:20markundrezha: rip a CD to ogg and try it
21:32:20*jhMikeS : The king of not noticing when someone has left or entered.
21:32:25preglowdrezha: using a good mp3 encoder, the quality is about the same, don't expect drastic savings
21:32:45preglowogg vorbis does perform better at low bitrates, though, if you use that
21:32:47*petur looks at jhMikeS ;)
21:33:30jhMikeSBut I did write that and THEN Will reentered so it wasn't that bad.
21:33:57drezharight OK.. I was looking at 192 bitrates
21:35:06markundrezha: just rip to ogg at a few bitrates and compare
21:35:49jhMikeSI've converted mp3s to OGG when they need to manually edited to be gapless...:/
21:36:28markunjhMikeS: when the MP3s need to be edited?
21:36:46peturjhMikeS: why not use mp3directcut to lossless crop mp3's?
21:37:00jhMikeSOpen is WAV editor and strip out the encoder delays to the sample and encode that to OGG.
21:37:15jhMikeSHow does it know the delays?
21:37:49***Saving seen data "./dancer.seen"
21:38:09 Quit webguest99 ("CGI:IRC (EOF)")
21:39:00jhMikeSI guess it could pick from common ones and all delays should be 1/2 frame for MPEG1...should try it thought even thought it almost never needs to be done
21:40:39drezhais there any decent rippers for Ubuntu?
21:40:45drezhaI cant seem to change the bitrate
21:40:51jhMikeSpetur: thanks, looks like a nice utility
21:50:56 Join Paul_the_Nerd [0] (
21:52:14 Quit Ribs ("eh eh ehhhh!")
21:54:37 Quit Lear ("Chatzilla 0.9.75 [Firefox 2.0/2006100319]")
21:57:51 Quit Siku ()
22:00:41 Join lee-qid [0] (
22:04:11 Quit charkins (Read error: 60 (Operation timed out))
22:06:12 Part drezha
22:06:51leftrightrockbox is trully magic on the recorder, it is streaks ahead of stock
22:08:02leftrightI hav just flashed it as well, boot is quicker,
22:10:01 Join bluey- [0] (
22:10:08leftrightthe lack of background noise with rockbox on ajr is just amazing,
22:11:06 Join nyonyai_csipcsi [0] (
22:11:49BagderI never used my recorder with Archos firmware, it is just so bad
22:11:50nyonyai_csipcsican someone tell me, how video playback is working nowdays on rockbox, especially on h300?
22:12:56leftrightits really amazing how different firmwares can be, the original is so shite in comparison
22:13:22nyonyai_csipcsidoes it plays back video with sound, or still not?
22:13:56leftrightbut, how rockbox fixes the background/white noise, or poor snr is amazing
22:14:19Bagdernyonyai_csipcsi: still no sound
22:14:39nyonyai_csipcsihm ok, thanx
22:15:29nyonyai_csipcsiwhen will the time come , when approximately be able to play with sound?:)
22:15:45Bagdernobody knows
22:16:17leftrighti in love with my Recorder, sounds different to my H140, recorders sound is "fuller" than the H140's, midrange is meatier
22:16:26 Join Rincewind [0] (
22:16:59*leftright will stop gushing now
22:17:08 Join MadDog011 [0] (
22:17:20 Quit Rincewind (Client Quit)
22:17:25 Part MadDog011
22:17:42 Join CriamosAndy [0] (
22:20:16 Join nudelyn [0] (
22:20:17leftrightnext step is to fit a 120gig HDD, :-)
22:21:28 Quit myzar ("...")
22:22:09 Join myzar [0] (
22:23:34 Part nyonyai_csipcsi
22:24:37 Join RedKennedy [0] (
22:24:41RedKennedyhello all
22:25:06*Bagder says hello on the behalf of "all"
22:28:03 Quit drippydonut ("Chatzilla 0.9.74 [Firefox]")
22:30:03 Quit Spida (Read error: 54 (Connection reset by peer))
22:31:19 Part Paul_the_Nerd
22:32:08 Quit ben__ (Remote closed the connection)
22:32:52 Quit mirak (Remote closed the connection)
22:35:24 Quit Criamos (Read error: 110 (Connection timed out))
22:38:23 Quit nudelyn2 (Read error: 110 (Connection timed out))
22:38:50 Quit CriamosAndy (Read error: 110 (Connection timed out))
22:42:27 Join godzirra [0] (
22:42:36godzirraHowdy folks.
22:45:10godzirraAny word on 5.5g?
22:45:20Kaltharehiya, godzirra.
22:45:49KalthareI've been a bit busy with other stuff for the past day, but I'm settling in now to try and debug that crash.
22:46:07godzirraAll good.
22:46:13*myzar licks Kalthare and godzirra
22:46:15godzirraLet me know if you need any help testing. Its about all I can do to help :)
22:46:24godzirramyzar: I hope you're a hot chick...
22:46:30myzaronly if you want me to...
22:46:31godzirraWait.. this is the internet.. everyone's a hot chick :)
22:46:33*myzar winks
22:46:48myzarinteresting fact
22:46:54myzarmy name in ascii totals 666
22:47:26myzarif you need me to help you guys with the 5.5s, i can help ya
22:47:36myzarbut i code in brainf**k and not C++
22:47:44myzarthough i do have an 80
22:48:17KalthareI've neutered the functions that write the file that's crashing, but I'm still setting up my iPod for testing, so I dunno yet if that'll narrow things down at all. I do note, with some surprise, that one of the functions in question steals the MP3 buffer for its own purposes.
22:50:51myzarperhaps that has something to do with either search or gapless?
22:50:51pillcan i upgrade an ipod firmware after having rockboxed it?
22:51:03pillbecause itunes seems like it hangs
22:51:03myzarbut you will need to reinstall rockbox
22:51:13myzarthat's not rockbox pill
22:51:16myzarhere's a simple test
22:51:23pillthat's itunes being sucky?
22:51:26myzarplug in your ipod and find the database file
22:51:32myzarcut it somewhere else and run itunes again
22:51:38myzarif it runs, you know some program corrupted it
22:51:46pillwell itunes runs
22:51:49myzarbut hangs.
22:51:54myzaror seems unresponsive
22:51:57 Part mra ("Leaving")
22:51:59myzaryeah, i'm in the same boat
22:52:06myzari used the winamp plugin ml_ipod once
22:52:11myzarnow anything can read the ipod but itunes
22:52:20myzarand i dont like that, because now i cant get gapless
22:52:25myzartry to do what it old ya
22:52:31pillbut say it's a friend's ipod mini
22:52:36pillif i delete the db
22:52:44myzardo not DELETE
22:52:44pillwill i be able to rebuild it
22:52:47myzarmove it temporarily
22:52:50pill(i'm a newbie at ipod)
22:53:13myzaritunes will say you have no music
22:53:19myzarbut it's there, as you can tell by the space meter
22:53:24myzarif so, it's your database
22:53:34pillokay hang on tight
22:53:37pilli'll let you know
22:53:47RedKennedyi've got a question about the doom there anything to install besides teh rockdoom wads and teh game wads?
22:54:14KalthareRedKennedy: That should be it, unless you want addons
22:54:22myzarsexy addons.
22:54:26myzarlike Kalthare in the nude.
22:54:32RedKennedyKalthare: danke
22:54:41myzarwas i right, pill?
22:54:46pillwhat is the name of the itunes db file?
22:54:50pillitunesdb? :p
22:54:51RedKennedythis is going to make study halls fun on my x5 :)
22:54:58pillthat it?
22:55:05myzardon't delete it.
22:55:14pillit's 164kB
22:55:22myzaryeah, you don't have much music
22:56:02myzarmaybe she unplugged it at a wrong time
22:56:10myzarand it killed her DB file
22:56:16myzarthat's pretty small
22:56:21myzarunless she has like... 4 songs
22:56:28pillwe'll be sure in 5 minutes
22:56:33pillwii boot
22:56:37myzarwhy are you rebooting.
22:56:40myzarugh, nevermind
22:56:49pillbecause i thought itunes was faulty
22:56:53pillso i uninstalled it
22:56:56pilland did not reboot
22:57:16pillso now it says "reboot before reinstalling itunes"
22:57:29myzargo ahead.
22:57:31 Join Id2ndR [0] (
22:57:33pillk brb
22:59:34 Quit amigan (Read error: 104 (Connection reset by peer))
23:00:08 Quit leftright (" Try HydraIRC -> <-")
23:01:23 Quit rretzbach (Read error: 104 (Connection reset by peer))
23:02:03 Join amigan [0] (i=dcp1990@unaffiliated/amigan)
23:05:39 Join blueworm [0] (
23:07:46KalthareOkay, if you destroy half of the playlist code, music playback works on the 30GB 5.5G
23:08:36pillnow itunes doesn't seem to see the ipod
23:08:42 Quit lee-qid ("Trillian (")
23:08:48pilland sometimes i wonder why i hate apple
23:10:12godzirradestroy playlist code?
23:10:28 Quit bluey- ("Leaving")
23:11:33 Join NickDe [0] (
23:12:20KalthareWhat crashes it is the playlist control file being updated.
23:12:36godzirrathats odd.
23:12:47 Quit Arathis ("[rl_bot quit]")
23:13:18KalthareNow I'm updating the tag cache during playback, and it's handling it like a champ. So, yeah, the playlist code is the problem here.
23:14:02pillhow do you get that mfing itunes to see that frigging ipod??!!
23:14:22pillis there a "scan" option or something?
23:14:40*pill dies
23:16:21KalthareWell, as I said before, one of the functions I gutted was taking over the MP3 buffer for its own purposes...
23:17:01 Join Moos [0] (i=51400b8c@gateway/web/cgi-irc/
23:17:46pillitunes still appears to hang at "upgrading "that ipod" please wait.. thing
23:17:57pillor in progress..
23:37:52***Saving seen data "./dancer.seen"
23:39:22 Quit MarcoPolo (Read error: 104 (Connection reset by peer))
23:40:54myzari think you killed the ipod
23:40:58 Join mirak [0] (
23:41:00myzarbut the DB file has nothing to do with that
23:41:18myzarif it wasn't seeing the IPod, i think that it was a short while before it would have died anyways
23:42:15*petur finds ugly ascii art in philips sample code
23:46:11Mikachuit's usually easier to fix the ipod yourself than hope itunes can do it
23:51:08 Join xamphear [0] (
23:51:15 Join charkins [0] (n=casey@pdpc/supporter/active/charkins)
23:55:19 Quit wooo (Read error: 110 (Connection timed out))
23:57:40pillit was not itunesdb
23:57:44pillit was no nothing
23:57:51pillitunes still hangs
23:57:54 Quit blueworm ("Leaving")
23:58:07pilland it's pissing me off to the point of no return
23:58:56 Quit Febs ("CGI:IRC (EOF)")

Previous day | Next day