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

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

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

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

#rockbox log for 2008-03-04

00:00:13linuxstbmoos: It's described in the patch - there are some outstanding bug reports, plus I'm not really happy with how images were handled (not respecting the viewports, but being drawn globally).
00:01:33mooslinuxstb: sorry was lazy to check myself :), full viewports with multifonts is probably the main patches remaining "against" unofficials builds
00:02:06 Quit MethoS- (Remote closed the connection)
00:02:06gevaertsActually there's one issue left that isn't mentionned on the wiki. If a sansa is plugged in while booting, it goes to usb mode correctly, but the sd card is not initialized yet so that isn't accessible.
00:02:10linuxstbmoos: So people are using viewports rather than the older patches?
00:02:52gevaertsBut as long as there are write problems on sansa, and the bootloader still boots to OF by default, I don't think that's too important
00:03:20 Quit bttb ("leaving")
00:03:20mooslinuxstb: just a guess
00:03:35linuxstbgevaerts: It could be worth adding it to the wiki page - you never know, someone may fix it for you...
00:03:40 Quit Domonoky (Read error: 104 (Connection reset by peer))
00:04:00*gevaerts edits the wiki now
00:05:20*gevaerts tries to think if there are other minor issues
00:06:43gevaertsActually, that's feature. If the sansa is plugged in while booting, there is no write corruption on the sd card :)
00:09:33pixelmahmm... the screenshot pngs should probably be renamed to have "context-menu" instead of "file-menu" in it, too
00:10:02barrywardellhave other people reported success with UMS on H10? or is it just me so far?
00:10:26gevaertsThere is still this issue that causes Windows to be annoying when you switch from UMS to serial to charging-only. It remembers what you did last time, and ignores new information. I'll test tomorrow if that can be fixed by adding a digit to the serial number that's used as a bitmap of enabled class drivers
00:10:39 Quit moos ("Buenas noches")
00:11:09gevaertsbarrywardell: amiconn has used it, and he has no signal quality problems with it.
00:13:13amiconnWell, almost
00:14:26amiconnI saw very few resets when having it connected to the front port of my linux box (3 resets while playing a whole album from usb disk)
00:15:05gevaertsCould you try that test again with the OF and the same port ? Maybe it's just due to the port
00:15:18amiconnCompare that to the dozens... hundreds of resets I get by just browsing my mini. Mini on rear port works flawlessly though (and H10 too)
00:15:42amiconnNo problems with OF on any port, H10 and Mini
00:15:45 Quit BigBambi (Remote closed the connection)
00:16:29 Join BigBambi [0] (n=Alex@rockbox/staff/BigBambi)
00:16:39amiconnSame applies to my windows box and connecting to my hub vs. root port. Hub produces errors, root port works
00:16:50amiconnHub works for all OFs
00:17:27gevaertsAnyway, I should be able to get signal testing done tomorrow on my c250, hopefully both on FS and HS. That might provide some hints (and of course if soap manages to get these voltage measurements done, that should give even more hints)
00:18:55*gevaerts hopes to get this solved before his gigabeat f arrives, so he only has to worry about one driver at the same time
00:19:05 Quit lee-qid ("aufwiederbyebientotsayonara")
00:26:28 Quit gregzx ("ChatZilla 0.9.81 [Firefox]")
00:28:12 Join thegeek [0] (
00:30:46 Quit EspeonEefi (Read error: 104 (Connection reset by peer))
00:31:28gevaertsActually I got a mail today from someone who owns two e200s, one of which absolutely refuses to connect to his front ports, while the other does connect to one front port, but not the other; all using the OF. The one that connects has newer firmware (1.02.18P vs 1.02.15P). Back ports work for both. I'm not sure what that implies, but it is interesting
00:32:37 Join midgey [0] (
00:33:05amiconnOut-of-specs front ports, perhaps?
00:33:35amiconnI guess my front ports are noiser than the rear ports, but still within specs (it's a HP business PC...)
00:34:34*amiconn now has a working gamma test plugin
00:35:00*amiconn used it in a sim to test the display gamma of his laptop as a verification
00:35:21gevaertsDon't be too sure of that. I've heard about devices that pass electrical tests perfectly and don't connect to HP front ports, from the people who did the tests.
00:35:43amiconnHP business pc or consumer pc?
00:36:00gevaertsI think both. I'll ask
00:36:58 Join EspeonEefi [0] (i=espeonee@STRATTON-THREE-FIFTEEN.MIT.EDU)
00:38:40linuxstbscorche|sh: What server will your rockbox-themes site run on? Will you have full control over it?
00:39:01scorche|shlinuxstb: yes...full and absolute control
00:39:20scorche|shwell, short of kernel modules
00:39:40linuxstbI really just mean ssh access so we can run scripts on it?
00:40:00linuxstbAnd build things like checkwps automatically from svn checkouts.
00:40:11scorche| is my VPS (which will be others too as soon as i upgrade it)
00:40:18scorche|shno shared hosting
00:41:00gevaertsSix more clean commits and there will be this nice green rectangle with just this one vertical yellow line...
00:41:14scorche|shgevaerts: thanks for jinxing it ;)
00:41:16rasherI guess it's time I broke my buildserver again
00:42:31linuxstbgevaerts: What are you planning for the gigabeat f?
00:42:32scorche|shlinuxstb: which means i should probably edit the code to use the zip function instead of just exec() (or whichever function i used..)
00:42:46 Quit Febs (Read error: 110 (Connection timed out))
00:43:32*linuxstb wonders if uploaded themes could/should be put in SVN...
00:44:01LloreanI'd vote against it.
00:44:25linuxstbI'm not talking about the current wps directory - just an svn repository to manage the uploaded themes.
00:44:29 Join cool_walking_ [0] (
00:44:43rasherI don't think there are many advantages to it
00:44:46gevaertslinuxstb: I'd like to get USB host working, if at all possible.
00:44:56LloreanI don't think we need revision history for themes. :)
00:45:09rasherscorche|sh: Latest wpschecker script:
00:45:22scorche|shscorche: ^^
00:45:43rasherscorche: Latest wpschecker script: (TOOLSDIR) needs adjustment, obviously
00:47:05linuxstbLlorean: It's not just about a revision history, it's about allowing people to edit/maintain them in a controlled way, and also without needing server access.
00:47:59rasherlinuxstb: I think "upload a new zip on top of the current one" is fine for maintaining a theme, really
00:49:18 Join |AhIoRoS| [0] (n=ahioros@
00:50:12linuxstbrasher: For 100s of themes?
00:50:27scorche|shthere really inst *that* much traffic
00:51:46scorche|shi would be afraid of theme authors ruining others'...and it would be a pain to set up permissions each time
00:52:24 Quit tvelocity (Remote closed the connection)
00:53:25roolkupixelma: where is a good place to install unicode support for cygwin? I tried /usr/share/texmf/tex/ but it is not finding utf8x.def when I do make manual-pdf
00:57:46roolkuah, mktexlsr seems to have fixed it
00:58:28*amiconn wonders what macos has against ums devices which are not marked as removable
00:59:21 Join dandin1 [0] (
00:59:33amiconnWould plugging an iriver, or archos recorder, confuse macos too?
01:00:38pixelmaroolku: did you run mktexslr (I think it was called)? (that was necessary for me to make it find the unicode package files)
01:00:54gevaertsI don't know. It seems to ask for mode pages that should be (IIUC) optional, and fails if those are not available.
01:01:03*pixelma should have read further... :)
01:01:09roolkupixelma: yes, that seems to have fixed it :)
01:01:54amiconngevaerts: Is it possible to find out what data a certain device sends for those mode pages?
01:02:06amiconnI mean, which having control about the device side stack
01:03:17 Quit HellDragon (Client Quit)
01:03:34gevaertsamiconn: you should be able to get those using a tracer (software would be sufficient for this)
01:04:02amiconnHmm, but only if the host requests this info, correct?
01:04:14*gevaerts realizes this eject change probably broke SD-card swapping while connected
01:04:55gevaertsamiconn: most OSes ask for page code 0x3f eventually, which means "all supported pages"
01:04:58*amiconn remembers that he did some experiments with a scsi inquiry program on Amiga way back in the past
01:05:39*gevaerts remembers reading about a disk search command in an old scsi spec
01:06:22*amiconn wonders whether he'll manage to dig up that program, and whether it'll work with the usbscsi.device provided by the usb stack
01:07:50*pixelma remembers reading february's IRC stats ;)
01:08:28gevaertsamiconn: on linux you can do "scsiinfo -c /dev/sd<whatever>" to get the caching page, which is the one that macos asks for
01:08:47amiconnAh, it's called SCSIquery - and I still have it installed
01:10:19 Quit amiconn (Nick collision from services.)
01:10:25 Join amiconn [50] (n=jens@rockbox/developer/amiconn)
01:11:07***Saving seen data "./dancer.seen"
01:13:16linuxstbIsn't it sg_inq on Linux?
01:13:24 Quit barrywardell (Remote closed the connection)
01:14:46gevaertsActually, it depends on what you want. The macos problem is a mode sense page, so the correct sg_* would be sg_modes
01:16:15 Join Zarggg [0] (
01:18:33gevaertssg_dd looks interesting for performing raw performance tests
01:19:06 Quit BlakeJohnson87 ("Leaving.")
01:20:55 Join BlakeJohnson86 [0] (
01:21:43amiconngevaerts: Any sg_* output that could be interesting to compare?
01:22:39amiconn(e.g. Mini OF diskmode vs. rockbox, some output from a device with a hardware usb-ata bridge which is not marked removable, etc)
01:22:58 Quit mchua ("Holy shiny distraction, Batman!")
01:25:31gevaertsamiconn: I'm trying to work out the syntax
01:25:56 Quit jac0b ("Ex-Chat")
01:26:39roolkupixelma: could I bother you about the manual? I have created the platform file for the MR100, but I am getting an error (?)
01:27:22roolkupixelma: any idea? Undefined control sequence ? Overfull \hbox ?
01:28:54 Quit ompaul (Client Quit)
01:29:30pixelmait seems to complain about a wrong button macro ( \ButtonA ) in rockbox_interface/main.tex.
01:30:44pixelmathe overfull and underfull hbox warnings are normal, you can ignore those.
01:31:10pixelmaat least I didn't find anything speaking against that yet
01:32:17roolkuah, ButtonA would be a problem - this should be ButtonDisplay ... wonder how it got there...
01:32:24 Join Kage [0] (
01:32:56gevaertsamiconn: "sg_modes −−six −−hex /dev/sd?" and "sg_modes −−hex /dev/sd?" are interesting. Those are for MODE SENSE 6 and MODE SENSE 10. For some reason the second one doesn't cut of its data at a correct length. "sg_inq /dev/sd?" might also be useful.
01:34:48pixelmaroolku: could also be an unclosed \opt{gigabeatf} perhaps? Or does the m:robe100 also have a button "A"?
01:35:24 Quit Nico_P (Remote closed the connection)
01:36:23roolkupixelma: no, I used the gigabeatf as the basis for my platform file - the mrobe 100 uses ButtonDisplay rather than ButtonA, but I thought I had replaced them all...
01:37:17gevaertsamiconn: "sg_modes -e −−hex /dev/sd?" should also be interesting, but it doesn't seem to work properly over usb (or maybe I don't have enough patience)
01:37:43 Part TheBashar
01:37:43roolkupixelma: ah, found the culprit
01:39:18amiconngevaerts: I already have results from apple diskmode + iriver h180 (hardware bridge); compiling latest rockbox for mini now...
01:39:35amiconnThe iriver doesn't support 6 byte mode sense at all
01:39:40pixelmaroolku: ah, glad you found it :)
01:39:53 Join saratoga [0] (n=9803c6dd@gateway/web/cgi-irc/
01:41:25gevaertsamiconn: I found a byte-order bug in the rockbox mode sense 10 code. Fixing now...
01:41:44amiconn-e works here, btw
01:42:01amiconn(but apple diskmode returns strange results iiuc)
01:42:15saratogais it possible to reboot a sansa into rockbox without resetting the CPU?
01:42:42saratogai'm wondering if it would be possible to leave all the OF's register settings in place then boot rockbox
01:42:53saratogawithout clearing them
01:44:04gevaertssaratoga: I doubt it, but I don't really know
01:46:02linuxstbsaratoga: I think you would have to hack the OF somehow to do it - insert some Rockbox code at a known point in the OF to dump the registers to disk... Or the alternative would be something like jtag, but I don't know if the Sansa has that
01:46:34saratogalinuxstb: it does have jtag but thats not what i was thinking
01:46:49amiconnUmm, rockbox doesn't like sg_modes -e
01:46:59saratogasince we can resign the OF after editing it, i was thinking of maybe just branching back to the bootloader after the OF boots
01:47:03 Quit Kage (Read error: 104 (Connection reset by peer))
01:47:24saratogai guess i need to talk to bagder about this
01:47:45linuxstbThe OF will most likely overwrite the bootloader (it's in IRAM)
01:48:02gevaertsamiconn: sansa OF doesn't like it either here
01:48:06linuxstbsaratoga: Have you looked at the emulator?
01:48:15saratogalinuxstb: yes a lot last summer
01:48:20amiconnNote the loooong reply of rockbox for 10 byte mode sense as well...
01:48:29saratogathough it didn't fully boot the OF so i'm not sure how useful it would be
01:48:36gevaertsamiconn: I just fixed that one
01:49:14saratogaI'm actually really hoping we can get a GSOC student interested in improving the emulator
01:49:31 Quit Zacam (Read error: 113 (No route to host))
01:49:40 Join pabs_ [0] (
01:52:33*gevaerts found another bug
01:54:01 Join fasmaie [0] (
01:54:11 Join jac0b [0] (
01:54:27jac0bis there a way to set a them to use a inverse slector bar?
01:55:11 Quit pabs (Read error: 110 (Connection timed out))
01:55:15saratogai think you can just edit the theme file to use the inverse selector bar instead of whatever its doing now
01:56:22jac0bHOW DO YOU DO IT
01:56:29jac0bopps sorry about the caps
01:57:02saratogathe WPSLIST declares "selector type: bar (inverse)" when it wants inverse so i'd guess that would work in a theme config file too
01:57:37 Join DaCapn [0] (
01:57:53pixelmayou could also set it, then save "theme settings" and look what's in the file...
01:59:10jac0bgot it thanks
01:59:20 Quit jac0b ("Ex-Chat")
02:01:10 Quit ender` (" There are two major products that come out of Berkeley: LSD and UNIX. We don't believe this to be a coincidence. -- Jeremy")
02:02:10 Quit stripwax ("Miranda IM! Smaller, Faster, Easier.")
02:02:24 Join stripwax [0] (
02:03:13 Quit roolku ()
02:04:05stripwaxI recall someone saying (on mailing list or forum post?) that they've noticed how much louder rockbox is on disk seeking than on original firmware on ipod 5g - I just rebooted into OF and I'd have to say I agree. Are we possibly not setting the correct power options ('quiet seek'?) in ata?
02:05:10 Nick pabs_ is now known as pabs (
02:05:43stripwaxwhere quiet seek = the acoustic management mode, but I can't recall the details
02:05:57amiconnWe are setting acoustic management if the disk supports it
02:07:41stripwaxI wonder if the disk incorrectly identifies as not supporting it, or if there is a different setting used by the OF
02:12:21 Join toffe82 [0] (
02:13:28 Quit stripwax ("Miranda IM! Smaller, Faster, Easier.")
02:14:06gevaertsamiconn: I'll finish looking at this scsi info tomorrow. I'm really getting sleepy now
02:14:11 Part SSnake
02:14:45*gevaerts wonders why this build is still not finished after 16 minutes
02:14:55 Join corevette [0] (
02:15:13 Quit gevaerts ("fallen asleep on his keyboard")
02:18:05 Join JdGordon [0] (n=jonno@rockbox/developer/JdGordon)
02:18:06pixelmais it just me or is slow for someone else too?
02:18:27JdGordonits fine here
02:19:24pixelmaoh well, maybe it's just saying "it's late here, get some sleep" :)
02:22:07 Part pixelma
02:26:03 Quit andrew___ ("leaving")
02:26:21 Join andrew__ [0] (
02:30:42 Quit fasmaie (Read error: 110 (Connection timed out))
02:35:52 Quit JdGordon ("Konversation terminated!")
02:40:32 Join b__ [0] (
02:46:05 Quit argumentD (Read error: 104 (Connection reset by peer))
02:46:16 Join JdGordon [0] (n=jonno@rockbox/developer/JdGordon)
02:53:38 Join HellDragon [0] (n=Nocebo@unaffiliated/helldragon)
02:59:38 Quit b__ (Read error: 110 (Connection timed out))
03:04:22 Quit perrikwp (" ajax IRC Client")
03:11:11***Saving seen data "./dancer.seen"
03:12:10 Join countrymonkey [0] (n=4b05639a@gateway/web/cgi-irc/
03:13:01 Quit ol_schoola ()
03:13:12countrymonkeyIs there a chance fs#8627 could get a critique?
03:16:52linuxstbcountrymonkey: I don't think "on track change" is descriptive - as that (to me) implies a beep on _every_ track change, not just when skipping. But I've never used that feature at all (I never even knew it existed...)
03:18:36countrymonkeygood point. I'll change it.
03:19:47 Quit dandin1 ()
03:21:15 Join LambdaCalculus37 [0] (
03:29:21 Join mchua [0] (
03:29:32 Part zedd_D1abl0
03:33:40countrymonkeyNow it is fixed. Could it be checked again?
03:37:26 Quit LambdaCalculus37 (Read error: 113 (No route to host))
03:37:40 Join shrimants [0] (
03:38:50shrimantsIs it possible to completely remove the ipod stock firmware from the ipod before/after installing rockbox?
03:39:29 Join keanu [0] (n=keanu@unaffiliated/keanu)
03:39:39 Part keanu ("Leaving")
03:41:26countrymonkeyNo. You would have to reflash the ipod rom and that is dangerous. I see no reason you would want to do such thing, though.
03:42:04shrimantsno reason, really, I was just wondering how to make it load only the rockbox firmware instead of using the bootloader that first has the apple logo and then the rockbox logos.
03:43:15countrymonkeyYou mean you don't want the rockbox logo?
03:43:40shrimantsno i mean i dont want the apple logo. It wasnt really a question of how, it was just a question of if it was possible or not
03:43:44cool_walking_Yeah, it's not possible currently. Rockbox doesn't have it's own UMS mode yet (you can't transfer files from the computer), so you still need the Apple firmware anyway.
03:43:54shrimantsOh, alright
03:44:27shrimantsBut then how does rockbox transfer? Are you supposed to reboot into disk mode in order to transfer music through syncing?
03:44:58countrymonkeyreboot into disk mode and reboot back again.
03:45:15shrimantsah ok, thanks for clearing that up
03:45:17linuxstbThere is working code in SVN for Rockbox's own UMS mode, but it's not enabled yet.
03:45:45linuxstbRockbox currently automatically reboots to Apple's emergency disk mode when you attach to USB, and then it will reboot back to rockbox when you detach.
03:46:01shrimantsIs it not possible to implement something into rockbox that makes it use sometehing like PlaysForSure or simple USB device mode transfering?
03:46:49linuxstb"PlaysForSure" won't happen - no-one wants to implement support for DRM in Rockbox.
03:46:50JdGordoncountrymonkey: only native speakers should be trying to update languages.... youve been told that a fwe times
03:47:01JdGordonand wasnt beep volume removed anyway?
03:47:19shrimantsI just used PlaysForSure as an example, i didnt mean that protocol specificall
03:47:40shrimantsmy Zen Vision M taught me enough hatred for that protocol
03:47:57countrymonkeyno. A native speaker bleepin updated this file. He gave me the translation because we were bored together today and I put it in.
03:48:18cool_walking_PlaysForSure isn't a protocol.
03:48:34linuxstbshrimants: But yes, other USB modes are possible - and I've just mentioned that we have a (almost) working UMS mode already
03:48:36cool_walking_It's just a certification that the player works with Windows Media Player
03:49:02shrimantsoh, I did not know that. What was the transfer protocol that windows devices used then?
03:49:11JdGordon"A helpful friend (carlos tores) translated this patch to spanish for me. Here is the new patch. Beware that even though he is a native spanish speaker, he doesn't know very much about the jukebox world. Additionally, he is a mexican spanish speaker. Here's the patch." i.e not nesacerily the person we want doing the translatino
03:49:12shrimantsyes! that one
03:49:34linuxstbJdGordon: translatino? ;)
03:49:54JdGordonyeah, for some reason my fingures always do the ion wrong :p
03:50:04 Join Alonea [0] (
03:50:45linuxstbshrimants: The transfer protocol is MTP
03:51:00Aloneajust to double check, the wall chargers don't work with the sansa? We have tried and so far no luck. It does not charge the battery.
03:51:25JdGordonit shuold... worst case is you need to use the OF
03:51:35linuxstbAlonea: In Rockbox or the original firmware?
03:51:37shrimantsright, Wouldnt it be possible to implement MTP into rockbox is what my question was. for instance, I dont know how you transfer your files over, but I just copy my "my music" folder over to the HD and rockbox tags it
03:52:01JdGordonshrimants: its possible, but unlikely
03:52:06Alonealinuxstb: rockbox. and if you plug in the charger while the player is off, it loads rockbox in a loop
03:52:13linuxstbshrimants: Yes, it's been mentioned a few times. It's even on the list of possible projects for this year's Google Summer Of Code
03:52:20JdGordonit is?
03:53:10shrimantsDo any of you use an ipod video 30gb with rockbox? if so, how do you manage/transfer your music over?
03:53:24linuxstbI just copy the files over...
03:53:29countrymonkeyWhy does it do that? I noticed that on Jan. 15, 2007, my first ever day of rockbox.
03:54:18Alonealinuxstb: I was looking through the logs and it looks like there is work being done on the usb issues. Does it look like this problem will be fixed sometime soon?
03:54:50countrymonkeyMe too. I have an ipod4g20gb and it is esentially the same.
03:55:02AloneaDad pretty much wants to be able to listen to music, etc. with the charger in like I could on my gigabeat.
03:55:03linuxstbAlonea: I think the effort is concentrated on getting the UMS driver working. I never even knew the Sansas had that problem (I don't own one)
03:55:11JdGordonAlonea: wierd.. are you using a unsupported bootloader? I dont think we ever rebooted on usb on the sansa
03:55:36AloneaJdGordon: not that I know of. Unless a new bootloader has been implemented in the last month.
03:55:49JdGordonnot to my knowledge...
03:55:58countrymonkeybbl. I am having a problem talking on the phone and talking here.
03:55:59JdGordonbut ok, just boot into the OF (hold left)
03:56:04shrimantsHow would you recommend installing rockbox, using the automated installer? or command line?
03:56:17linuxstbAlonea: When does Rockbox reboot? Immediately on boot, or after while?
03:56:29shrimantsalright then, automated it is lol
03:57:19AloneaJdGordon: now, if you put in the usb cable to computer while off, it loads the OF like its supposed to. If you plug in wall charger, it loads the rockbox boot screen, goes black, reboots back to rocobox boot screen in a loop
03:58:06AloneaJdGordon: if you plug in wall charger while in rockbox, nothing happens and battery does not charge. if you plug in usb in rockbox, it usually reboots into OF.
03:58:32AloneaJdGordon: yes. 280 to be exact.
03:58:42JdGordonand your definatly using a supported build?
03:58:51JdGordonI'm sure re otting was never added for sansa
03:58:59JdGordonbut ill go find my sansa which isnt busy and try it out
03:59:18AloneaI always use rockbox builds. haven't built my own in a lonnnnnng time.
04:00:24JdGordonoh wow! it did reboot
04:00:27AloneaI meant to mention this a month ago, but I got busy. though I don't know if the loop thing was happening before. its not my player. ^__^
04:00:44AloneaJdGordon: when you plug in the usb cable to compy?
04:00:59*JdGordon checks with wall charger
04:01:04AloneaJdGordon: I was told it was supposed to.
04:01:47Aloneasince there isn't usb in rockbox support yet. (*sniffles*)
04:03:05JdGordonnup, wall charger appears to work fine
04:03:09JdGordonnot rebooting and looping
04:03:26Aloneahmm...using latest version?
04:04:21*JdGordon will try svn shortly.. running between 3 comps atm
04:04:26Aloneaits ok.
04:04:50Aloneaso your charger actually charges your player? while in rockbox? or only in OF?
04:05:15Aloneawhen you plug in your charger, does it load OF or rockbox?
04:10:40 Join sarixe [0] (
04:11:18JdGordonit loads rockbox and says its charging
04:12:53shrimantswoah the automated installer has come a LONG way since i installed rockbox last
04:13:30Aloneaodd. I will try again in a sec. I bought the charger that the reviews said didn't break within a week and didn't have any problems. give me a sec to tell you the brand.
04:14:33JdGordonI'm using some generic usb charger
04:14:53countrymonkeyI am wondering: I have an aa quickcharger and a minib-to-ipod tip adapter. How does rb fair with those?
04:14:56Aloneaok, its a belkin.
04:16:44AloneaJdGordon: where would it say "chargin" for rockbox?
04:16:56countrymonkey,In the rockbox info screen.
04:16:58JdGordonin the statusbar
04:17:08Aloneawhat theme?
04:17:31JdGordonany theeme.. just enable the statusbar
04:17:37JdGordonthe battery icno will change
04:18:07Aloneaeh, its hard to tell since the battery is currently full.
04:20:37Aloneaok, i am using cabbie, and the battery thing has no animation or anything
04:20:45Aloneaneither did cassette.
04:22:03Aloneathere it goes!
04:22:39AloneaI gues the update fixed it (as for why it did not work a second ago I am using one of the plugs on a box and I forgot that one was bad
04:23:25AloneaI guess I will see if after a while, the green is all the way and double check in OF.
04:23:30 Quit sarixe ("(EE) Failed to load "quit" module")
04:23:38 Join simonrvn_ [0] (i=simon@unaffiliated/simonrvn)
04:23:55 Join Davide-NYC [0] (
04:24:41IsolinearSo... I have a bug to report regarding the iPod5G where it crashes and makes strange static noises and I took an audio recording of the symptom... Anyone know how/where I could put this audio file somewhere for all to access to see if others are having the same problem?
04:26:52 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
04:27:18 Join miepchen^schlaf [0] (
04:27:35Davide-NYCCan you consistently reproduce the bug?
04:27:52IsolinearWell it consistently happens over and over again.
04:28:23IsolinearBut it occurs at seemingly random intervals, no telling when or how often it'll happen, nor what triggers it.
04:28:36Davide-NYCBug reports go on the tracker. I think you can attach files there.
04:28:49IsolinearThat's why I took the recording, as the noises it makes are quite distinctive.
04:29:02IsolinearAh really? So a 3MB .wav file is no problem then?
04:29:12Davide-NYCYup, on the tracker it goes. (attachment max 2MB)
04:29:22Davide-NYCcut it down to just the relevant portion. it up.
04:29:40AloneaJdGordon: so I will check some more on this charging thing and double check it works. lemme see if the loop thing still happens
04:29:48Davide-NYCI know it is not my place to question the devs, but I am wondering why the unfinished projects from last years GSoC aren't on this years wiki page?
04:30:04saratogaDavide-NYC: you're welcome to add them
04:30:12Davide-NYCI am?
04:30:17saratogathe list there is mostly just things i was interested in
04:30:21 Join perrikwp [0] (i=982175a9@gateway/web/ajax/
04:30:39Davide-NYCMy opinion is GSoc2008 should focum on a 3.0 release
04:30:48saratogathats not really possible
04:31:01IsolinearBtw, to whomever is responsible for homepage site design, the little icons on the left next to all the links, as well as including the revision number for each source code update are very welcom improvements. :)
04:31:01 Join argumentD [0] (
04:31:13*Davide-NYC agreed
04:31:17saratogareleases take basically everyone working on them
04:31:29Davide-NYCsaratoga: I know, but pieces could be targeted, no?
04:31:35Aloneaok, JdGordon, I misread the loop thing. This is what is happening. When the player is OFF and I plug in the wall charger, the sandisk logo comes up, the rockbox screen comes up, and it just stops. the black is the backlight going off.
04:31:50saratogawell yes you could pick a piece and work on it, but i doubt fixing any one thing is going to move much closer to a 3.0 release
04:32:09Davide-NYCFor the iRiver H1x0 series we are extremely close IMO./
04:32:15 Part shrimants
04:32:49 Join jurrie [0] (
04:33:07AloneaJdGordon: if I then unplug the charger, it finishes loading.
04:34:06 Join alienbiker99 [0] (
04:34:45*alienbiker99 just fixed my h320 battery
04:34:46 Quit simonrvn (Read error: 113 (No route to host))
04:34:46 Nick simonrvn_ is now known as simonrvn (i=simon@unaffiliated/simonrvn)
04:35:19IsolinearAh, never mind, I found someone else has posted a recording in FS #8651.
04:36:46Davide-NYCsaratoga: Is it appropriate to put viewports on there?
04:40:29 Join ol_schoola [0] (
04:41:09saratogaDavide-NYC: isn't that already being worked on by core developers?
04:42:10Davide-NYCYes, which is why I'm asking. I may be unclear as to what should and shoudl not be included in GSoC.
04:42:55saratogawell it has to be something that could reasonably be done by a student interested in the project
04:43:16saratogathings that are already in an advanced state of development and likely to go into rockbox before GSOC begins are not good choices
04:43:35Davide-NYCwhat about APE optimization?
04:43:42saratogaideally it should be things that would take a students roughly the summer to complete and do not involve learning massive amounts about other people's code
04:45:09countrymonkeyDoint tts, I guess?
04:45:25countrymonkeytext-to-speech, sorry.
04:46:35 Nick fxb is now known as fxb__ (
04:48:02saratogapersonally i'm hoping for the ARM emulator and Rockbox as an App projects
04:48:29 Quit JdGordon (Read error: 110 (Connection timed out))
04:48:37Davide-NYCI have never used MonkeysAudio, but I find one of Rockboxes greatest strengths is its ability to play just about anything. Currently MonkeysAudio is the worst performing 'supported' codec.
04:49:15Davide-NYCsaratoga: could you explain to me what the attraction is the RB as application project is?
04:49:17 Join jurrie_ [0] (
04:49:17 Nick Klowner_ is now known as Klowner (
04:49:50saratogait would basically allow rockbox to run on smart phones or pdas
04:50:10saratogadevices which allow 3rd party applications without much or any hacking
04:50:28Davide-NYCmainly for Rockbock style playback...
04:50:33KlownerI've been using rockbox for two days now, it's so neat, thank you developer dudes
04:50:41countrymonkeyYou mean, turning these smartphones (if they aren't already) into ogg, sid, shn, mpc, flac and other formats' players?
04:50:41saratogathings like iphones, google's android project, linux smartphones, etc
04:51:05countrymonkeyand would plugins be included in the port?
04:51:14 Join TheBashar [0] (
04:51:53Davide-NYCCool for playback, but I'm one of the resident recording enthusiast Rockbox users and as such need to look out for that niche. How about WRS support in GSoC? (while recording screen)
04:52:20Davide-NYCThis would make more sense if viewports was in place.
04:52:21soapDavide-NYC, I'm not an expert on where Rockbox's current APE implementation stands as far as optimization - but it is a _very_ CPU hungry codec.
04:52:30countrymonkeyYou mean like .wps filews we would have .wrs and .rwrs files?
04:52:46soapI think it is reasonably to believe it might never make it realtime on a vast majority of current targets.
04:52:47Davide-NYCcountrymonkey: exactly
04:53:02countrymonkeyI tried ape. I couldn't get the thing to play (using a sept. build).
04:53:34Davide-NYCsoap: even gigabeat?
04:53:50countrymonkeygigabeat is realtime, iirc.
04:54:10soapIIRC most compression levels are realtime on Gigabeat - need to look at the wiki.
04:54:31soapalso, IIRC WavPack hybrid support isn't 100% - that would be cool.
04:54:42Davide-NYCno ape info on gigabeat in the wiki
04:54:54Davide-NYCsoap: Now you're talking!
04:55:18countrymonkeyThat's right. -c 1000 to -c 3000. I used -c 1000 on a pp target.
04:56:00countrymonkeyI remember from the initial commit message when the codec was first committed.
04:56:13 Join jurrie__ [0] (
04:56:25soaplol - ape-insane @ sub 2% on coldfire
04:57:37Davide-NYC(for the logs) Depending on whether Viewports is implemented in time or not I would like to vote for some WRS (while recording screen) support as a GSoC project. Or at the very least a rework of the recording screen even if static. (I have some ideas)
04:58:01LloreanI don't know if simply reorganizing it is "big enough" for GSoC
04:58:12LloreanAnd I don't think it needs token based themability in the sense the WPS does, probably.
04:58:16Davide-NYCBut WRS would be.
04:58:22countrymonkeygute nacht
04:59:03countrymonkeyawf widersehen.
04:59:11soapfixed recording on PP would be a task
04:59:12LloreanPlease don't start with that.
04:59:19 Quit countrymonkey ("CGI:IRC (EOF)")
04:59:32Davide-NYCLlorean: I agree, but I've been itching for a couple of changes to the WRS for years now.
04:59:33 Quit perrikwp (K-lined)
04:59:46*pikhq is looking for a GSoC project
05:00:01LloreanFixed recording on PP might be. I think perhaps a general themability work for the UI, tokenizing as many elements as possible might be better. If the whole UI was more abstractly themable it might be nice.
05:00:15 Join perrikwp [0] (i=982175a9@gateway/web/ajax/
05:00:15pikhqJust thought I'd pipe in, since I saw you talking about it.
05:00:21*Llorean goes back to being away.
05:00:46soapfixed recording on PP would open so many doors.
05:01:17*pikhq tries to remember: is the Sansa a PP target?
05:01:23Davide-NYCsoap: That's definitely a great one. Fixed PP recording.
05:01:27soapthough the ARM emulator would probably be a door towards that (and many other) goals. I gotta believe the ARM emulator would catch google's attention as well./
05:01:32 Join Monga [0] (
05:01:35soapyes, pikhq
05:01:42MongaBroke my Gigabeat S today =\
05:01:45pikhqw00ts: I have hardware.
05:01:48 Quit jurrie__ ("Leaving")
05:01:58alienbiker99Monga how did you do that
05:02:09MongaShithead threw me onto the ground and kicked me.
05:02:17MongaScreens screwed.
05:02:56soapI think I'll offer to donate a Apple Accessory Protocol debugger to anyone who gets that as a GSoC project.
05:03:06 Join JdGordon [0] (
05:03:25 Quit HellDragon (Client Quit)
05:04:24soapOh no! the product is no longer sold!
05:04:26 Quit DaCapn (Read error: 104 (Connection reset by peer))
05:05:07saratogapikhq: do you have much cpu arch background?
05:05:21pikhqsaratoga: Not a gigantic one.
05:05:51saratogaMonga: looks like my gigabeat S
05:05:55saratogathanks to poor shipping
05:06:07 Quit jurrie (Read error: 110 (Connection timed out))
05:06:09pikhqI can offer skills in x86 assembly, C, C++, Tcl, and Brainfuck, but not a whole lot of CPU architectures.
05:06:17MongaNote to everyone, don't have your headphones wrapped around your mp4.
05:06:18pikhqObviously, C would be the useful one there. ;)
05:06:35*pikhq has recently gotten bored, and started writing an OS kernel, for what it's worth
05:06:36saratogapikhq: well a good knowledge of assembly certainly helps
05:07:06saratogaif you use one of the already working ARM SOC emulators, you might not need much CPU design experience
05:07:31pikhqHmm. Sansa e200 in there?
05:08:03 Join jurrie [0] (
05:08:28pikhqYeah, although it only emulates one core. . .
05:08:34 Join JdGordon_ [0] (
05:08:46saratogapikhq: the current e200 emulator may not be the best place to start
05:09:11saratogadebugging the emulation could be quite difficult, and there are already several fully working and arm compliant emulators under GPL compatable license
05:10:00pikhqMeh; it's not like this is the only task Rockbox could ever offer. ;)
05:10:20Davide-NYCThis feature request should be closed IMO
05:10:23pikhq(or, for that matter, the only mentoring organization I could be with. . . Doing stuff for Tcl is also quite tempting)
05:10:28Davide-NYCFS #5917
05:10:36saratogaskyeye looks very promising since it already runs linux on the arm7tdmi core, which is by far the most popular rockbox cpu
05:11:14***Saving seen data "./dancer.seen"
05:11:26saratogaadding emulation for the PP LCD bridge and one of the wolfson or as DACs would not be particularly difficult for a student with the right background i think
05:12:42pikhqWhich, of course, is the trick.
05:12:44 Quit jurrie_ (Read error: 110 (Connection timed out))
05:13:06saratogawell you learn
05:13:19saratogai'd certainly never writtena codec when i picked up the WMA project last year
05:14:51 Quit JdGordon (Nick collision from services.)
05:14:54 Nick JdGordon_ is now known as JdGordon (n=jonno@rockbox/developer/JdGordon)
05:17:24soapand thus busted the best of the Golden Quotes.
05:22:24 Quit Thundercloud_ (Remote closed the connection)
05:22:29saratogathough perhaps someone with a clue will someday appear to finish the last couple problems with the codec...
05:25:11Davide-NYCUI suggestion for the H1x0: From FM, long press Rec goes to the Rec screen.
05:25:35Davide-NYCFrom Rec, long press A-B goes back to FM.
05:26:02LloreanDavide-NYC: Generally speaking, this sort of thing belongs as a feature request.
05:26:03Davide-NYCIt is currently awkward to switch between the recording screen and FM.
05:26:22 Quit jurrie (Read error: 104 (Connection reset by peer))
05:26:47 Join jurrie [0] (
05:27:02Davide-NYCLlorean: do me a favor and close FS #5917.
05:28:29Davide-NYCI'm trying to close requests, not make more!
05:29:10Davide-NYCJust wondering what the devs think. It's really not that important. I just think the UI is a little wierd when navigating between REC and FM.
05:30:32LloreanI don't think there should be a shortcut to the FM screen unless it works from everywhere, but I don't mind the recording screen shortcut being enabled in the FM screen as well.
05:33:27JdGordonDavide-NYC: how is it awkward? fm -> menu -> rec
05:34:14 Join ReymiX [0] (n=admin@
05:34:59 Quit Alonea (Read error: 110 (Connection timed out))
05:35:36 Quit saratoga ("CGI:IRC")
05:35:49Davide-NYCDiffernt button presses to get back to FM.
05:36:51LloreanAh, so perhaps the real problem is inconsistent buttons for getting to the Rockbox Menu?
05:37:00Davide-NYCI think I have it. Long A-B could leave the REC screen without stoping the recording! and long press A-B could leave the FM screen without stopping playback!
05:37:03*Llorean may have complained of this in the past.
05:37:33 Quit ReymiX (Client Quit)
05:37:35*Davide-NYC as an aside I think the quickscreen menu should be abolished.
05:38:29 Quit Horscht ("electromagnetic radiation from satellite debris")
05:39:07JdGordonits not a menu!
05:39:18JdGordonif it was it would be called the quickmenu
05:40:48LloreanWhether or not it's a menu is an important semantic point in another discussion. ;)
05:42:15 Join kushal_12_27_200 [0] (n=kushal@
05:43:30Davide-NYCJdgordon: FYI: It's called Quick Menu in the manual. LOL
05:43:40 Quit |AhIoRoS| (Read error: 110 (Connection timed out))
05:43:40 Join Alonea [0] (
05:43:52JdGordonmeh, the manual is wrong :p
05:44:09Davide-NYCIs the quick screen / menu only available from the file menu?
05:44:50JdGordonbrowsers and wps
05:44:52 Quit Alonea (Client Quit)
05:46:40Davide-NYCBut not the menu itself?
05:46:50JdGordonno, but thats fixable
05:46:55Davide-NYCI find that a bit wierd as it looks like a browser.
05:47:17Davide-NYCOh, that's a bug or is that up for discussion?
05:49:13JdGordonits something which was never fixed.... and hasn been discussed
05:49:59Davide-NYCIMO if the quickscreen exists the it should be gotten to from the menu as well.
05:50:16JdGordoneveryone knows how I feel about that screen :)
05:51:22Davide-NYCHow about this as a small UI tweak? REC A-B long should go to the REC settings menu and REC A-B short should go to the regular menu. That way getting to the menu is always the same button press: A-B short.
05:51:24 Join ReymiX [0] (n=admin@
05:51:34Davide-NYCJdGordon: you must be an Archos user of some sort.
05:51:47ReymiXanybody can help me ?
05:53:24ReymiXdude i need a little help... i uninstall complety rockbox firm from my 5th ipod video
05:53:40ReymiXso i want reinstall rockbox
05:54:09ReymiXso what its the best way to install ? manual installation or automatic ?
05:54:38ReymiXn i want multifont patch too
05:56:10JdGordonDavide-NYC: no, why?
05:56:54Davide-NYCReymiX; Hmm. I don't own a 5g iPod, but I'd say that as long as you know which version of it (it comes in two flavors) I would do a manual install.
05:57:31Davide-NYCJdGordon: It just seems like such a vestige of tiny screen players.
06:04:56 Part TheBashar
06:12:32 Quit blithe (
06:14:36 Quit Monga ()
06:15:02 Join |framo| [0] (
06:19:28 Quit wisp_sh (Read error: 110 (Connection timed out))
06:27:58 Quit mchua ("Holy shiny distraction, Batman!")
06:46:11Davide-NYCDoes anybody know where I can find a list of all of the encoding options for Nero AAC Codec? It doesn't specifiy the bitrate ranges for the different encoding types./
06:46:20 Join hcs [0] (n=agashlin@rockbox/contributor/hcs)
06:47:25 Join DaCapn [0] (
06:57:45 Quit jhMikeS (Read error: 104 (Connection reset by peer))
06:59:58 Join jhMikeS [50] (n=jethead7@rockbox/developer/jhMikeS)
07:02:16 Join blithe [0] (
07:09:55 Quit perrikwp (" ajax IRC Client")
07:11:18***Saving seen data "./dancer.seen"
07:11:49 Join perrikwp [0] (i=982175a9@gateway/web/ajax/
07:15:02 Join ompaul [0] (n=ompaul@gnewsense/friend/ompaul)
07:15:32 Part toffe82
07:24:58 Quit Zarggg ("Even electronic brain pancake crystal elderly have mass naked child events lately.")
07:27:05 Join wisp_sh [0] (
07:27:28 Quit ReymiX ()
07:31:15 Join HellDragon [0] (n=Nocebo@unaffiliated/helldragon)
07:36:28 Quit JdGordon (Remote closed the connection)
07:36:52 Join JdGordon [0] (
07:44:19 Quit midgey ()
07:57:16 Quit ompaul (Client Quit)
07:57:16 Quit Davide-NYC ("ChatZilla 0.9.81 [Firefox]")
07:57:43 Join lee-qid [0] (
08:08:24 Quit BigBambi (Remote closed the connection)
08:19:52 Quit lee-qid (Read error: 110 (Connection timed out))
08:20:35 Join Rob2222 [0] (
08:22:36 Join LinusN [0] (n=linus@rockbox/developer/LinusN)
08:25:41 Join aliask [0] (n=chatzill@rockbox/developer/aliask)
08:33:09advcomp2019i just updated my rockbox build on my e200R series and the cabbiev2 theme is taking over the menus only of the other theme that i have been using since before cabbiev2 was added but the WPS is still the theme that i been using.. is this still a bug
08:38:01amiconnNo it's not
08:38:24 Quit Rob222241 (Read error: 110 (Connection timed out))
08:38:25amiconnIf you have used non-default settings before, cabbiev2 won't replace these
08:38:49advcomp2019by the way the theme that i have been using was "plain text"
08:38:53amiconnIf you want the complete thing, either reset your settings, or load the theme explicitly
08:39:24amiconnOn a fresh installation, the whole theme will be active from the start
08:40:17advcomp2019odd.. because i replaced the files and then cabbiev2 took over
08:41:26 Quit HellDragon (Client Quit)
08:43:40amiconnAgain, not odd at all
08:44:32amiconnThe defaults were changed, so you got the cabbiev2 elements for all settings where you used the old default values before
08:45:26amiconnThis is due to the way rockbox stores the current config. All settings which are set to their default aren't stored in config.cfg
08:46:11advcomp2019after updating tho the wps was cabbiev2 till i tried replied the "plain text" theme but the menus did not change tho
08:46:51amiconnThen that theme doesn't specify menu settings
08:47:06amiconnThat's not rockbox' fault, but a bug (incompleteness) in the theme
08:48:02advcomp2019o ok.. is there a way to change it back or not
08:54:15 Join ch4os_ [0] (n=ch4os@unaffiliated/ch4os/x-059673)
08:57:21 Quit cool_walking_ (Remote closed the connection)
08:58:00 Join petur [50] (n=petur@rockbox/developer/petur)
08:59:47 Join Administrator_ [0] (n=chatzill@
09:00:43 Join DerDome [0] (
09:01:14 Join ender` [0] (
09:02:26 Join webguest94 [0] (n=dabc076e@gateway/web/cgi-irc/
09:05:50 Quit webguest94 (Client Quit)
09:06:31 Nick Administrator_ is now known as convil (n=chatzill@
09:11:22***Saving seen data "./dancer.seen"
09:22:47 Join mrkiko [0] (
09:22:56mrkikoHi all...
09:24:28peturHi universe
09:32:46 Quit JdGordon (Read error: 104 (Connection reset by peer))
09:33:00 Join Abhinav [0] (n=bijolian@
09:33:29 Join JdGordon [0] (
09:34:42Abhinavis the rockbox particiapting in summer of code 2008
09:34:59peturwe hope we will be selected
09:35:03LinusNyes, we will apply
09:35:14GodEaterare any of us besides me hanging out in #gsoc ?
09:35:28peturnot any more
09:35:31aliaskNot any more
09:35:37peturanything interesting?
09:35:40aliask... :)
09:35:42GodEaternothing recently
09:36:31Abhinavi am hoping rockbox be there
09:37:02aliaskAbhinav: Are you interested in applying as a student?
09:37:13 Quit kushal_12_27_200 ("This computer has gone to sleep")
09:37:20Abhinavyes i am interested as a student
09:37:39Abhinavcan u guide me through
09:38:08aliaskGuide you through what? Applying?
09:38:48Abhinavi have just started in looking at rockbox
09:39:34 Quit JdGordon (Read error: 104 (Connection reset by peer))
09:39:39Abhinavi am really loving this project, i find it most cool project to work on
09:40:22 Join JdGordon [0] (
09:40:32 Join pondlife [50] (n=Steve@rockbox/developer/pondlife)
09:40:46 Quit joshin (Read error: 110 (Connection timed out))
09:41:39*GodEater celebrates moving into third place on the "Top Posters" stats for the forums.
09:41:49Abhinavaliask : so u guide me through
09:42:04aliaskAbhinav: Take a look at our wiki page about the GSOC:
09:42:51 Join pixelma [50] (i=pixelma@rockbox/staff/pixelma)
09:44:26 Quit JdGordon (Read error: 104 (Connection reset by peer))
09:45:44 Join JdGordon [0] (
09:50:01 Join vipintm [0] (n=vipintm@
09:51:49 Quit JdGordon (Read error: 104 (Connection reset by peer))
09:51:54markunGodEater: where's this list?
09:52:08GodEatermarkun: which one ?
09:52:16markunTop Posters
09:52:19 Join JdGordon [0] (
09:53:22 Quit JdGordon (Read error: 104 (Connection reset by peer))
09:53:38markunmost time online.. :)
09:53:40pondlifeWow, look at how much life Llorean has wasted
09:53:57pondlifeHe must spend more time online than sleeping
09:54:07advcomp2019amiconn, i am slowly getting it fixed but it is working
09:54:31 Join JdGordon [0] (
09:55:33pondlifeGodEater: You could increase your post count by updating the DevConPub3 thread
09:55:48GodEaterwhat more needs to be said ?
09:55:48pondlifeIf I did it that would just be a waste ;)
09:55:53pondlifeLocation and time?
09:55:59GodEaterhm ok
09:56:08pondlife"We didn't decide a location yet, no."
09:57:00pondlifeAlso good to bump, now we're nearly there... but don't break any rules.
09:57:17*GodEater is a good poster and uses the modify button instead.
09:58:21 Part vipintm ("Leaving.")
10:01:18linuxstbIMO, that's the kind of situation where a new post is more appropriate - it's a distinct new post, rather than adding information you forgot the first time...
10:02:48 Join BigBambi [0] (n=86ceaf37@rockbox/staff/BigBambi)
10:03:24 Join Zagor [242] (
10:03:39pondlifeYes, and it didn't bump
10:03:51pondlifeI think this time a bump is required...!
10:04:36BigBambiLlorean: Are we close to revising that rule?
10:04:36pondlifeand again
10:06:04*linuxstb wonders if laptops running the sim count in a Tower of Rockbox
10:06:29peturwe should put some ads on the forum. 800000+ page views/month...
10:06:29JdGordonthe rule which seems to work in another forum is you cant add a new post <24h after a post
10:06:58JdGordonlinuxstb: we had the sim going on the projector which made a ncie background to our tower a few weeks ago :p
10:07:29BigBambiI think it just ought to be significqnt new info is OK in a new post
10:08:14pondlifelinuxstb: hehe, how about a large tower case PC or two?
10:08:43linuxstbpondlife: Tower of Rockbox Build Servers?
10:08:46pondlifeI've got an old 486 thing somewhere, it's probably powerful enough to run the sim and it's about 3 feet tall
10:08:56pondlifeIt might break the table though
10:08:57BigBambiI had one of those
10:09:10BigBambipondlife: But it was a cyrix 166!
10:09:22pondlifeHmm, might not run realtime
10:09:34BigBambiNothing ever did. It was crap
10:09:48pixelmathe twiki main page (WebHome) states some 25th/26th October 2007 changes as "10 most recent changes" /for me?). If I go to "more" everything is alright and the latest changes to the wiki stated on the frontpage look correctly too...
10:09:49pondlifeGodEater: I'll pop an open invite up on the ML, if you're ok with that
10:10:34GodEatermore the merrier
10:10:38Zagorpixelma: the oldest change I see is 3 march
10:10:45BigBambipixelma: Twiki webhome recent updates look OK to me
10:10:45 Quit wisp_sh (Read error: 110 (Connection timed out))
10:10:56peturZagor: 4 march
10:11:02BigBambipixelma: 04 Mar 2008 - 08:38 IpodFAQ is the latest I see
10:11:27peturpixelma: hit 'Refresh cache' and ctrl-F5
10:11:38pixelmaI did several times :\
10:11:38Zagorpetur: don't you see "03 Mar 2008 - 23:15 RockboxPlayerPrototype JorgePinto" as the last line?
10:12:10peturyou mean here:
10:12:18convilI can't compile the inline assembler in libdemac's vector_math16_arm7.h with gcc 2.95.3 and gcc 3.4.5. Does anyone has a patch?
10:12:28BigBambipixelma: Maybe 'tis your ISP caching, but is definately OK here
10:12:33peturZagor: it is the last one at the bottom
10:12:44Zagorright, good
10:12:49*pondlife probably just offended some people
10:12:49 Quit JdGordon (Read error: 104 (Connection reset by peer))
10:12:54pixelmasame happens when I look at it in another browser (not used for a while)
10:13:07peturthe top one is IpodFaq - SpencerHarston
10:13:32pondlifeme too
10:13:57Zagorpixelma: that's really odd
10:13:59 Join JdGordon [0] (
10:14:35GodEaterI see the icons have gone live :)
10:15:03peturGodEater: now you're really slow... :p
10:15:40peturThey are live since yesterday
10:15:55Zagoryeah I thought we'd try something new. mixed reactions though.
10:16:10GodEaterI saw the test page yesterday
10:16:18pondlifeZagor: Any chance you could increase the vertical spacing on the menu to be like the old one? See"> ...
10:16:18GodEaterdidn't know they'd gone live till just now
10:16:34pondlifeZagor: Also right justified text is a bit nicer
10:16:34 Quit JdGordon (Read error: 104 (Connection reset by peer))
10:16:43linuxstbconvil: No, but a patch would be welcome... What target are you trying to compile it for?
10:16:49 Quit mrkiko ("leaving")
10:17:01Zagorpondlife: I saw you mention that in the log yesterday. which browser is that? the new menu actually has *more* vertical spacing than the old.
10:17:14Zagorreally odd
10:17:18pondlifeDo you see the image?
10:17:29pondlifeAnd you see something different?
10:17:35peturto be honest, I like the new spacing better
10:17:58Zagoryes, try firefox and you'll see the line spacing is identical, but the section top-margin has increased slightly
10:18:11convillinuxstb: An arm7 board with uClinux 2.4.26. And I can't even read ARM assembler, sorry.
10:18:16*pondlife has no FireFox at work.
10:18:31Zagori wonder what makes ie7 shrink it
10:18:34peturmaybe if the icons are put to the right side and text is also right-aligned some people will no longer complain about ythe left-alignment...
10:18:38pixelmaok, I'll wait a bit then. I found the icons confusing, couldn't find the IRC link that easy as before, it felt like I had to go through the list twice
10:18:54Zagorpetur: I did that at first, but it got terribly jumbled
10:20:20 Join JdGordon [0] (
10:20:48pixelmathe twiki webhome I see does not have them though (yet) :P
10:21:06Zagorpixelma: yes it does, just not your ancient cache of it
10:21:19linuxstbconvil: This is off-topic for this channel - do you want to take this to PMs?
10:21:49convillinuxstb: sure
10:22:15ZagorI'm not very attached to the icons. I just thought it could be nice to add some eye candy.
10:22:38pixelmaZagor: I figured (because the "recent changes > more" have it, hence the ;P
10:23:49Zagorpixelma: ah
10:25:42pondlifeZagor: I just noticed that the menu spacing is ok on, but not on the front page
10:26:20peturpondlife: ie7 is just being silly (as usual)
10:26:20pondlifeIn fact, only the wiki and extras pages are ok. The rest are squished up
10:27:05pondlifeoops, and the manual
10:27:24Zagoris the manual ok?
10:27:30pondlifeNo, squished
10:27:46 Join a6 [0] (
10:28:06pondlifeTake all that back.
10:28:10pixelmaI think I read that IE does something different with some <div> attributes but can't remember exactly
10:28:20pondlifeIt's the wiki pages that are fine, all others are squished
10:29:53JdGordonlinuxstb: hey, whats the story with the vp patches?
10:30:04Zagorthe only difference is that the wiki menu has all newlines removed
10:30:42linuxstbJdGordon: Still no real time to work on them...
10:30:42Zagor is used for twiki, for all other pages
10:30:57*petur confirms the oddness with ie7 and quickly shuts it down again
10:31:05JdGordonlinuxstb: what work needs to be done?
10:32:17 Quit Abhinav (Read error: 110 (Connection timed out))
10:32:35linuxstbJdGordon: The bugs mentioned in the flyspray task, the TODOs mentioned in the patch (mainly cleaning up the %V parser, and adding bounds checking to it IIRC), plus thinking about how images are handled - they are currently handled globally by the WPS code (after the lines of text are drawn), so are drawn in the default viewport, not the individual viewports, which doesn't seem right...
10:33:11JdGordonhmm, ok
10:33:20 Quit sbhsu (Read error: 110 (Connection timed out))
10:33:41linuxstbHow about your patches? ;)
10:33:55JdGordonthe list one needs testers
10:34:12JdGordonand the same bounds checking and parsing
10:34:41linuxstbThat could obviously be combined...
10:34:53JdGordonyeah, course
10:35:24 Join gevaerts [0] (n=fg@rockbox/developer/gevaerts)
10:40:57 Join TheBashar [0] (
10:41:40*GodEater reads pondlife posting to the mailling list
10:42:01GodEateryou should have said you'd fix a bug for someone if they can best you at arm wrestling ;)
10:42:11pondlifeNo way
10:42:24pondlifeIt's bad enough trying to find time to fix my own bugs
10:43:12*gevaerts was just wondering if building a Tower of Rockbox counts a bricking the players involved
10:43:26pondlifeIf you use mortar, yes
10:43:39pondlifeAnd a hod, of course
10:44:47*gevaerts prefers dry-stone towers
10:44:58pondlifeYou're a traditionalist then?
10:45:04JdGordonlinuxstb: oh, had any thoughts on what to do about scrolling lines not owned by the vp being drawn?
10:45:29pondlifeJdGordon: They should still be scrolling away, surely
10:46:02gevaertspondlife: I want to "brick" my players without ruining them
10:47:05JdGordonpondlife: and the ones which scroll overlaps the vp we are drawing?
10:47:22pondlifeHmm, viewports don't overlap
10:47:31pondlifeI thought that was fundamental
10:48:04pondlifeviewports != windows
10:48:20JdGordonthats the point... they shouldnt
10:48:29JdGordonthink screen transitions
10:48:34linuxstbJdGordon: There are cases where you don't want to stop all scrolling lines on the screen, just some?
10:48:50JdGordonwe dont want to have to kill scrolling lines in a screen at every exit point
10:49:01pondlifeI'd think that whilst a vp is displayed, any contained scrolling lines should scroll.
10:49:14JdGordonyeah, I want to add a function which just forcibly stops any lines which would be in the vp's area
10:49:33linuxstbpondlife: The case in my mind is "pop-up" menus - stop the scrolling lines underneath, but not clear the display.
10:49:45pondlifeAh, that involves overlap
10:49:52JdGordonand going wps->lists
10:50:04linuxstbpondlife: Yes, but not with "active" viewports...
10:50:15pondlifeI thought we were excluding that sort of thing
10:50:15linuxstbJdGordon: When wouldn't you want to stop all lines underneath?
10:50:26linuxstbs/underneath/on the screen/
10:51:07pondlifeHide the WPS viewports, then display the menu viewport
10:51:10linuxstbJdGordon: If you go wps->lists, don't you just call lcd_clear_display() ?
10:51:39JdGordonnot unless we want to force the lists to use the whole screen
10:52:00pondlifeOtherwise you end up with overlap and Z-order
10:52:10pondlifeWhich is out of scope, no?
10:52:24pondlifeThis isn't for pop-ups, is it?
10:52:26linuxstbJdGordon: We _do_ want that - even if the list itself doesn't cover the whole screen, it should clear the previous contents (as it does now).
10:52:41pondlifeIt's more for a standardised status bar and better WPS margin control.
10:52:47*JdGordon thought we dont want that
10:52:51linuxstbOr just call the function to stop _all_ scrolling lines.
10:52:59JdGordonthats what I;m doing now
10:53:06pondlifeNo, keep any lines that are displayed scrolling, surely
10:53:11pondlifeWhy do you want to stop them?
10:53:29linuxstbBecause the screen underneath is no longer being updated
10:53:33pondlifee.g. the status bar becomes customisable and contains the track name.
10:53:38JdGordonthats what I'm saying/..... we only want to stop lines which would get in the way of the currently active vp
10:53:39linuxstbi.e. track info isn't being updated.
10:53:42 Join MethoS- [0] (
10:53:54pondlifeBut you are displaying menus/WPS on the rest of the display... should not stop the status bar scrolling
10:54:01linuxstbJdGordon: I would disagree - unless the wps is being updated, we should "freeze" it.
10:54:29JdGordonwps isnt the only case here... its just the most obvious
10:54:37linuxstbWhat are the other cases?
10:54:57pondlifeLike I said above
10:55:02pondlifeStatus bar with scrolling
10:55:18*JdGordon doesnt tihnk statusbar will ever go wps-like
10:55:22pondlifeWhy not
10:55:32pondlifeThat would be the #1 benefit of viewports to me
10:56:09pondlifeWithout a standardised (and hence customisable) status bar, I wouldn't even worry about viewports.
10:56:14linuxstbpondlife: I agree.
10:56:17*linuxstb needs to run
10:56:20 Quit linuxstb ("Leaving")
10:56:40pondlifeThe biggest problem with cabbiev2 is that sucky non-standard status icons
10:56:55pondlifeI actually like them, but I want them standard.
10:57:42pondlifeImagine the menus etc having the status bar at the bottom, for example, using exactly the same code as on the WPS.
10:57:57 Join mchua [0] (
10:58:08JdGordonwe dont actually need vp for that... but yes, that would be cool
10:58:16 Quit convil ("ChatZilla 0.9.81 [Firefox]")
10:58:32pondlifeWell, I see vps as a way to carve up the available screen space (main and remote). That's about it.
10:58:48pondlifeThey are not windows and do not allow overlap or Z-order.
10:59:30pixelmathen I'd like the possibility for the volume that you have in the current menu status bar - that it can switch to numeric while adjusting and when you leave it alone showing the icon again
10:59:42pondlifeOf course
11:00:10pondlifeI'd like all the current status bar info to be available as "WPS" tags. Stuff like disk indicator.
11:00:12JdGordonthat _may_ not be possible with a wps-like statusbar
11:00:21pondlifeWhy not?
11:00:23JdGordonto pixelma
11:00:39pixelmait's currently not possible in the WPS but it should be
11:00:53pondlifeA conditional to say when you are changing volume...
11:00:56*petur dreams of using the vp to split the recording screen into a fixed part with peakmeters and info, and a second scrollable part for volume/gain settings
11:01:09pondlifepetur: Good idea
11:01:20*JdGordon has dreams of shattering petur's dreams :D
11:01:35pondlifeBut you';d want the menu text to continue scrolling (if required) at all times, right?
11:02:20pondlifeI don't see why vp work should need to stop/start scrolling. It operates at a higher level than the scroll code, no?
11:02:36JdGordonyes and no...
11:02:49pondlifeHmm, it must be above or below.
11:03:18pondlifeEither the vp tells the scroll code to manage a line (like we do now). Or we redo it somehow.
11:03:22JdGordonscrolling lines take the vp pointer, you can only stop all lines, or just lines which you know the vp pointer... the list code doesnt know the wps' (and vice verca)
11:03:23*petur sees vp as lower level, almost as displays within a display
11:03:27JdGordonthats where the problem is
11:04:03pondlifeWhen do we need to stop scrolling a line?
11:04:04JdGordonscrolling <-> vp has a nasty interdependancy so you cant really say which is higher level
11:04:12pondlifeYou need to decide
11:04:18pondlife(Or somebody does.)
11:04:28JdGordonwhen the wps has a line scrolling which is in an area which we now want to use for the list
11:04:50 Quit mchua ("Holy shiny distraction, Batman!")
11:05:07pondlifeOK, so you remove the WPS vp, which cancels that scroll. Then you display the list vp.
11:05:18pondlifeNo need to artificially stop scrolling
11:05:23 Join joshin [0] (n=joshin@VDSL-130-13-243-154.PHNX.QWEST.NET)
11:06:04 Quit ender` (Read error: 104 (Connection reset by peer))
11:06:06JdGordonah, but 1) we arnt definatly using the whole area (by using i mean stealing, not that its actually being drawn on), 2) stopping the wps scrolling at every exit point would be massive PITA
11:06:10JdGordonand buggy
11:06:12JdGordonand shite
11:06:27pondlifeSo you want to overlap vps
11:06:53JdGordonyou have a top one which is being updated, and a bottom one which can still scroll but nothing else
11:07:04JdGordona nice statusbar would have the same problem
11:07:04pondlifetop/bottom... thus Z-order!
11:07:18pondlifeThey are in their own screen areas
11:07:23pondlifeNever overlapping
11:07:26JdGordonsays who?
11:07:30JdGordonthey can move around
11:07:31pondlifeIn my model
11:07:40JdGordonyour model is wrong :D
11:07:42pondlifeYes , but not overlapping.
11:07:46pondlifeI mean, movable
11:08:06pondlifeI agree with petur too.. scrolling should be inside a viewport
11:08:09JdGordonbeing able to set the x,y forces it to be movable
11:08:27 Quit MethoS- (Remote closed the connection)
11:08:39*JdGordon points pondlife to scroll_engine.c to reomvve the interdepenacy
11:08:46pondlifeAfter you!
11:08:57JdGordonthat still wouldnt fix the problem anyway
11:09:00pondlifeBut it's design notes we're lacking here, not code
11:09:31JdGordonit would still need either the list vp to know about all other vp's, or have each one tell the engine when it exits
11:10:37pondlifeThe scroll engine is used from the VP code. When a VP wants a line to scroll, it asks the engine to handle it. When the VP is hidden, it tells the scroll engine to stop.
11:11:14pondlifeNo VP needs to know about other VPs
11:11:23***Saving seen data "./dancer.seen"
11:11:30pondlifeBut they must be positioned so they don't overlap
11:12:00JdGordonok, if this doesnt get my point home nothing will.... (wall of text coming....)
11:12:05pondlifeOR if you must overlap, we have a call to tell a VP to stop refreshing
11:12:17pondlife(and another to resume)
11:12:39JdGordonwe have the list code done correctly (drawing and logic in 2 seperate functions)... the list is taking up the whole screen
11:12:56JdGordonone or more items are scrolling
11:13:50JdGordonnow, the user hits play so we need to load the wps. we need to tell the scroll engine to stop scrolling our lines, but we cant.. they are local to another function/file/hacky to get to them
11:14:00JdGordonwe also need to do this every time the display changes
11:14:05JdGordonits just not nice
11:14:25JdGordonthe alternative is in the new screen say "dont let anything scroll in my margin"
11:14:33JdGordonboom, no stray lines
11:15:05pondlifeYes, but the transition from menu to WPS needs to be handled by the outer code anyway. It just tells the menu to hide, then displays the WPS. I don't think that's complicated or not nice.
11:15:20pondlifeThe menu hides, and hence scrolling is stopped
11:15:35pondlifeEach bit of display code is self-contained
11:15:39JdGordonit doesnt hide... the display gets erased
11:15:48pondlifeNo, it hides
11:15:55pondlifeThe status bar remains, for example
11:16:34pondlifeThink of each vp as a separate display
11:16:51JdGordonno it doesnt... scrolling is the only thing on the display which happens "automatically" everything else is static unless something changes the display
11:17:47pondlifeThe scrolling thread will be told to stop a particular bit of scrolling as part of the "hides".
11:17:56JdGordonby who?
11:18:01pondlifeThe vp
11:18:17pondlifeWhichever! The menu, the WPS, the status bar
11:18:39pondlifeWhichever vp asked for the scrolling text in the first place.
11:18:56JdGordonok, how does it know to stop the lines?
11:18:57pondlifeBear in mind, this is nothing to do with the current code
11:19:04pondlifeBecause it's been told to hide.
11:19:22pondlifeSo, as part of that, it tidies up after itself
11:19:23JdGordonarg... i cant do this over irc
11:19:43peturwhat? hit him?
11:20:02pondlifeHe Can do that over IRC :)
11:20:07*petur defends pondlife
11:20:55JdGordonok, so in your idea the display would be split up... so who is to say that pressing menu from the wps would actually need to hide any of the wps?
11:20:59pondlifeI'm trying to say - we need a clean design first. One in which each vp is self-contained... and without worrying about the current code too much.
11:20:59peturif the vp is told to hide, can't it notify the scrolling thread to stop all lines involved with the vp
11:21:19JdGordonthere is no notion of hiding though
11:21:41pondlifeJdGordon: Or, if you want a static WPS behind the menu, the hide doesn't need to erase.
11:21:44JdGordononly one vp can ever be being drawn at once so thats the active one, all others are disabled/hidden/static
11:22:17JdGordonbecause the gui code is all handled by the one thread
11:22:18pondlifeI'd like to see the user using a menu while the status bar updates.
11:22:42JdGordonthe status bar only updates because we call statusbar_draw() in every button loop
11:22:52*JdGordon has already coded a simple WM to handle just that
11:23:00pondlifeWell that will continue.. i.e. it'll loop through the VPs
11:23:19pondlifeBut "hidden" ones won't draw
11:23:22JdGordonyes, which means either z-order comes into play, or something else
11:23:29pondlifeSomething else
11:23:35pondlifei.e. no overlap
11:23:49JdGordonno, your right, vp's will be marked hidden there, but the WM cant stop the scrolling
11:24:00*gevaerts rotates his coordinate system to get rid of z-order
11:24:18pondlifeWhy not?
11:24:26pondlifeThe existing scroll code will need reworking
11:24:44*JdGordon has a sneeking feeling you are agreeing with me
11:25:06JdGordonyou are, you just dont know it :D
11:25:18pondlifeI don't see why anything OTHER than a vp would manage scrolling within itself.
11:25:44JdGordonbecause the scrolling lines could be in 15 different vp's under a "parent"
11:25:53JdGordonwhich is how the lists in my patch are done
11:26:15pondlifeI'm not interested in code yet. Just design.
11:26:16JdGordonand ONLY the drawing code for that vp knows about it
11:26:52JdGordon... you gonna be round in 2 hours?
11:27:04JdGordonmind if we pick this up later?
11:27:08*JdGordon has very impatient fridns
11:27:09pondlifeJdGordon: You should head over to London on Thursday :)
11:28:13JdGordonbassically the user says draw the list in some rectangle. the list itself is made up of 4 or 5 vp's of which 2 have scrolling content, only gui/bitmap/list.c has access to those vps, but the drawing code is not renentered to tell it to stop drawing... it jusst doesnt get called
11:28:17*JdGordon goned
11:28:48*pixelma is somehow reminded of DevCon when amiconn tried to explain his vp ideas
11:29:22 Join webguest94 [0] (n=8602720d@gateway/web/cgi-irc/
11:29:30pondlifeI think I need amiconn to help me....
11:30:18amiconnViewports should never overlap, so I don't see a problem
11:30:23pixelmaI think part of it also was about this scrolling problem but we all had a hard time understanding it (I didn't understand which is why I'm not sure)
11:30:33pondlifeI just want to end up with a well structure (and API'd) bit of code.
11:32:39amiconnActive viewports, that is. So the z-order and scrolling problem simply does not exist
11:33:35pondlifeamiconn: So I think we agree... When I said "hide" I probably meant "disable" or "become inactive".
11:36:02amiconnI proposed viewports as a means to simplify screen layout. One element (e.g. the list) shouldn't have to care whether other elements like the statusbar or scrollbar are visible, and how much space these elements take away
11:37:03pondlifeThere would be an top layer that divides the physical screens up into viewports, but once that is done, they are independant?
11:37:06amiconnOf course viewports can be subdivided further, e.g. the list could subdivide its viewport into one for the title, one for scrollbar,and another one for the main part of the list
11:38:06amiconnThis makes it easier to truly support rtl languages - just swap the scrollbar and main list viewports horizontally for proper scrollbar placement
11:38:12 Join n1s [0] (
11:38:22 Join ender` [0] (
11:38:37pondlifeYes, so a vp can be withiin another vp, but the parent one doesn't draw anything?
11:38:40amiconnOf course the text alignment would also have to be swapped, but that becomes easier as well
11:39:10amiconnExactly. If there are any "sub"-Viewports, the parent must not be used for drawing.
11:39:33pondlifeSo no contention.
11:40:10amiconnIt will also make colouring, theming etc easier
11:41:02amiconnEach viewport stores a set of attributes (foreground colour, background colour, draw mode, font, and perhaps later also font style, whether the backdrop should be used etc)
11:46:42 Join wisp_sh [0] (
11:46:50 Quit JdGordon (Read error: 110 (Connection timed out))
11:54:34 Join roolku [0] (
12:08:19 Join bruges [0] (
12:12:50 Join actionshrimp [0] (
12:14:43pondlifeHmm, in the sim a database update appears to open database_idx.tcd about 20 times for each track...
12:14:52 Quit actionshrimp (Client Quit)
12:15:23 Join actionshrimp [0] (
12:16:17pondlifeSorry, I think it's for every track. e.g.
12:16:44 Part actionshrimp
12:20:06LinusNwhy do we have to reopen the file for every track?
12:23:34markunBagder, LinusN: maybe there guys could help us get a place to hold the DevconEuro2008 (if it's in Berlin)
12:24:54pixelmamarkun: you could also nag amiconn a bit to ask... ;)
12:26:10markunpixelma: do you guys know some other places where we could do it?
12:26:40markunpixelma: he's not one of those 300 members, is he?
12:28:45pixelmano. But from what I know amiconn's work place would be a good location too (maybe not meeting on friday though)
12:29:58pondlifeI'm not sure I want to visit a space station for a DevCon...
12:30:25pondlifeMind you, I've already been up the Fersehturm.
12:30:32pixelmahaha, this c-base even seems to not be far away from it..
12:30:38pondlifeFernsehturm, even
12:30:50 Join seablue [0] (
12:31:01pixelmapondlife: that's also not far away ;)
12:31:32 Join Nico_P [50] (n=nicolas@rockbox/developer/NicoP)
12:32:03Nico_Ppondlife: I missed the HD by 15 minutes :/
12:32:11Nico_Pnow I need to go get it at the post office
12:32:16pondlifeJoy joy joy
12:32:36pondlifeLet's hope the post office isn't heavy-handed, or full of hammers
12:32:51Nico_Pplus, I need to wait till tomorrow
12:33:11pondlifeI await your report...
12:34:57Nico_Pyou'll have it on your desk by tomorrow afternoon ;)
12:37:54 Join fasmaie [0] (
12:40:26pondlifeA large number of my MP3s give the warning "Warning: skipping 417 bytes of garbage" too.
12:40:31 Join desowin [0] (n=desowin@atheme/developer/desowin)
12:40:43pondlifeThey all play fine, so I wonder what the "garbage" is?
12:41:15pondlifeShould all be ID3v2 only...
12:47:02pixelmapondlife: you could try running something like mp3val on them and see what that has to say
12:48:02*pondlife downloads
12:49:58 Join ctaylorr [0] (
12:51:28pondlifeWARNING: "m:\I'm Sorry I Haven't A Clue\In Search Of Mornington Crescent\01 Opening.mp3" (offset 0x40a): MPEG stream error, resynchronized successfully
12:51:41pondlifeSo, something is wrong...
12:51:51pondlifeBut OT
12:52:50 Join barrywardell [0] (
12:54:20*gevaerts seems to have some of the same tracks as pondlife
12:54:52GodEatersounds like a good track to own
12:54:55pondlifeBut presumably not with the same corruption
12:55:06*pondlife will re-rip
12:57:16gevaertsActually, I seem to have misplaced those files so I'll have to re-rip as well
13:00:02 Join ze_ [0] (
13:01:07 Quit aliask ("ChatZilla 0.9.81 [Firefox]")
13:01:57 Quit ze (Read error: 110 (Connection timed out))
13:01:58 Nick ze_ is now known as ze (
13:04:22 Join tvelocity [0] (n=tony@
13:05:20Zagoroff-topic: does anyone have experience with mod_perl and can answer a few questions in pm?
13:11:25***Saving seen data "./dancer.seen"
13:22:37 Quit gevaerts (Read error: 110 (Connection timed out))
13:23:50LinusNpondlife: the "skipping garbage" warning is often the result of wrongly padded ID3 tags, where the padding is placed after the tag, before the audio
13:24:33pondlifeHmm, I only use EAC, The Godfather and Foobar2000... I'd hope they all get it right.
13:24:57pondlifeRockbox doesn't ever write tags does it (for ratings)?
13:25:58pixelmaI thought Rockbox's ratings are stored in a database file
13:28:01pondlifeI thought so too
13:28:04pondlifeJust checking
13:28:09 Join gevaerts [0] (
13:28:38*gevaerts senses more need for infrastructure work...
13:29:20 Join Thundercloud [0] (
13:31:26gevaertsTo handl removable drives (like the sansa microSD slot) properly from the USB stack, the sd driver needs to let the usb storage driver know about plugging and unplugging. This is probably best done using the existing thread-message queues. We'll probably need some general way to pass through events to usb class drivers.
13:31:40gevaertsThis needs more thinking...
13:31:47 Quit ctaylorr (Read error: 110 (Connection timed out))
13:34:59amiconnThere should be a broadcast event sent on plug and unplug
13:35:23 Join kushal_12_27_200 [0] (n=kushal@
13:36:14 Join whoever_ [0] (
13:36:22whoever_hello together
13:36:35whoever_I have a question in a WSP-File
13:36:45whoever_(sorry, WPS:))
13:36:59BigBambiJust ask
13:37:11whoever_I want to write the first line in Blake font, the other lines in white
13:37:20whoever_how is this possible? (sorry for my enlish;))
13:37:54BigBambiwhoever_: It currently isn't, but will be when an improvement known as viewports is included
13:38:20whoever_how long could this take:)?
13:38:24gevaertsamiconn: That's probably the easy part (although it seems that those are sent only if the fs is mounted successfully, so maybe that will need some work). The "problem" is getting this event to the storage driver (which doesn't have its own event loop). It's easy to get an ad-hoc solution for this of course, but I'd like to do it cleanly
13:38:46BigBambiwhoever_ Are both lines information (i.e. song title or artist), and not just fixed, i.e. "Now Playing"?
13:39:14whoever_No, because then I would create a picture:)! I want to write the time down:)
13:39:18BigBambiwhoever_: And viewports will be done when it is done :), no estimate is available I'm afraid
13:39:40BigBambiwhoever_: OK, just asking in case you hadn't thought of the picture route
13:39:52*gevaerts sometimes feels that this usb work will never be done
13:39:59BigBambiwhoever_: There is a patch for viewports in the WPS you could try
13:40:05amiconngevaerts: If they are only sent on fs mount, the implementation is buggy
13:40:37whoever_BigBambi, where can I found it?
13:40:45amiconnMentioned event *should* be sent on physical plug/unplug (on Ondio they are). When the filesystem is mounted/unmounted, the SYS_FS_CHANGED event should be fired
13:40:48whoever_Sorry, I'm not so good in rockbox and co:)
13:41:29BigBambiYou will need to patch the source code then compile rockbox
13:41:52cgdoes anyone know when the get_more callback of pcm_play_data is called? can the same buffer be overwritten on next callback or could it still be playing?
13:42:15whoever_BigBambi, oh puh - I think I don't touch my rockbox;), then I wait:)
13:42:23BigBambiOK :)
13:42:42whoever_but thanks a lot;), I found the possabilitiy with %V, but on my rockbox it didn't work, so I came to this IRC:)
13:42:53BigBambiNo problem.
13:43:33gevaertsamiconn: it calls disk_mount(), and broadcasts SYS_HOTSWAP_* if that returns non-zero.
13:43:42amiconnMeh, why are the 2 hotswap implementations so different :(
13:43:48 Join JdGordon [0] (
13:43:56whoever_BigBambi, But with the viewport: do you think that takes Weeks, Months or Years:)?
13:44:18amiconngevaerts: That's clearly *not* how it was meant to be. Compare this with the ata_mmc.c driver (used on Ondio)
13:44:43BigBambiwhoever_: I really don't know - rockbox is written by volunteers in their spare time, so whenever someone both wants to do it and has the time, it will be done
13:45:27BigBambiwhoever_: I suspect if one of our gurus sat down to do it, it would take days, but that relies on a) one of them wanting to do it, and b) having the spare time
13:45:40BigBambiwhoever_: No problem, keep your fingers crossed :)
13:45:54BigBambidays as in not long, not litterally days of work
13:46:26gevaertsamiconn: I'll look into it tonight. I need to get back to work now.
13:46:39*BigBambi too has to do some work :/
13:47:51*gevaerts really starts to suspect a conspiracy to make him clean up lots of code
13:48:04 Quit fasmaie (Read error: 110 (Connection timed out))
13:48:18 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
13:49:19 Quit JdGordon ("Konversation terminated!")
13:49:50LinusNamiconn: damn, gevaerts is on to us ;-)
13:51:32 Join knittl [0] (n=knittl@
13:51:52*preglow performs the gevaerts appreciation dance
13:52:05Bagderah, that's a rare beauty
13:52:17preglowyeah, been practicing for days
14:05:43linuxstbIs anyone intending to review and commit the MOD codec?
14:06:19*linuxstb excuses himself as he doesn't have any mod files
14:06:48 Join webguest18 [0] (n=5b4ca593@gateway/web/cgi-irc/
14:07:10 Join vcf [0] (n=5b4ca593@gateway/web/cgi-irc/
14:07:10 Quit webguest18 (Client Quit)
14:08:03vcfmarkun hi
14:09:09vcflets wait rika enters
14:10:13 Join rika [0] (n=5767b95d@gateway/web/cgi-irc/
14:10:26rikahi markun
14:10:50vcfrika hi!but its not markun
14:11:30rikathan who? O_O
14:11:35 Join kslater [0] (
14:11:35vcfrika wait a bit
14:11:53*rika out to smoke, 10min
14:12:00vcfрика это vcf.мракун ща придёт
14:12:53markunvcf: please no russion and no repeating
14:13:41vcfrika where are you?
14:14:41 Quit vcf ("CGI:IRC (EOF)")
14:15:34 Join vcf [0] (n=5b4ca593@gateway/web/cgi-irc/
14:19:33 Quit perrikwp (" ajax IRC Client")
14:19:44rikahere now
14:20:50vcf<rika>that is great)))
14:20:56markunrika: hi, did you see my PM?
14:21:44rikano i havent notice about it(
14:22:29markunrika: there is a little tab at the bottom of your screen with "markun" in it
14:22:52rikaoh yes :) (i never used web irc)
14:22:57 Join midgey [0] (
14:25:06 Quit nplus (Remote closed the connection)
14:29:22 Quit rika ("CGI:IRC")
14:29:24 Join LambdaCalculus37 [0] (n=44a0430f@gateway/web/cgi-irc/
14:32:44pondlifeIs CanonInteger a real name?
14:32:58pondlifeOr is it IntegerCanon?
14:33:11*LambdaCalculus37 shrugs
14:33:18LambdaCalculus37Sounds more like a printer to me.
14:33:21amiconnEpsonFractional ;)
14:33:46LambdaCalculus37DotMatrix. :)
14:34:02barrywardellgevaerts: so, is there anything else holding us back from enabling UMS on H10 by default?
14:34:23 Join nplus [0] (n=nplus@141.25.Globcom.Net)
14:35:10pondlifeLinusN: I see a delay on the H300 backlight too. If you set it always on, and off on hold, it's very obvious
14:35:32gevaertsbarrywardell: probably not much, but I'd like to wait a few days (maybe while providing some unofficial builds) to let the code stabilize a bit
14:36:26barrywardellgevaerts: sure. I'll keep testing myself and see if I come across any remaining problems
14:36:27amiconnpondlife: Hold doesn't send events, so you'll see half the polling delay on average. That's normal
14:36:44pondlifeI thought it used to be much faster though
14:36:52pondlifeIt takes 1-2 seconds
14:36:55*gevaerts started on this to get his c250 working, and with this sansa ata driver bug it looks like that will be the last to be enabled...
14:37:13amiconnThere's still the signal quality problem
14:37:58gevaertsThat too. I'll probably do the tests in about half an hour. Hopefully they tell us something useful (like if full speed is clean)
14:38:09*pondlife wonders if the sim should support upside down displays...
14:39:03whoever_I have a last question. In my WPS, I want to display the picture from the album (from the TAG)! this is %C|0|25|75|75| now, but there is e error, but why? when is there a error?
14:39:07n1samiconn: for me it can take up to 3 seconds after pushing a button until backlight comes on
14:41:09linuxstbwhoever_: What do you mean by "error" ? What happens?
14:41:12whoever_or could help, if I do a %?C ?
14:41:20pondlifen1s: Did anyone FS it yet?
14:41:23whoever_it changes to the default theme
14:41:28whoever_(to linuxstb )
14:41:48n1spondlife: don't think so
14:41:51amiconnn1s: Any special options enabled? That seems to be yet another of those bugs I'm not experiencing at all...
14:42:21whoever_linuxstb, should I paste the full WPS-File?
14:42:33linuxstbwhoever_: Yes, that may help - that line looks fine.
14:42:43GodEaterdon't paste it here though
14:42:59linuxstbwhoever_: Paste it to
14:43:20whoever_oh sorry:)
14:43:26GodEaterthat'll do :)
14:43:29n1samiconn, pondlife: the reason I haven't FS:ed it yet is that it's not consistent and I haven't had time to check what makes it happen etc..
14:44:25whoever_linuxstb, there are no empty lines between the pictures in the beginning.
14:44:37whoever_and - when I clean the line with the picture, it works:)
14:44:44n1sLinusN: maybe I replied a little too fast in the forum, I guess I should have said I knew about the bug...
14:45:33 Join Arathis [0] (
14:47:35whoever_linuxstb, do you see an error? (I have the iAudio X5)! perhaps with the coordinates?
14:49:59linuxstbwhoever_: Nothing obvious, although does your WPS have all those blank lines? Are you saying it works if you remove the line?
14:50:32whoever_yes, it works when i remove the line %C....
14:50:53 Quit nplus (Remote closed the connection)
14:51:10linuxstbAre you using the UI simulator?
14:51:31whoever_linuxstb, for me there are two possabilitys: first: the coorinates are incompatibel, perhaps a Text-line has the error with the picture. the second: what happend, when there is no picture in the tag? is there an error?
14:51:41whoever_no, i wrote this in my text-editor
14:51:59linuxstbI mean testing it in the UI simulator?
14:52:11whoever_should i:)?
14:53:38whoever_But i have ubuntu
14:55:33 Quit DaCapn (Remote closed the connection)
14:56:02preglowlinuxstb: i'm planning on checking the codec out, but don't know when i'll do so
14:57:19whoever_linuxstb, how can I test there my Theme? because I don't know, how to install the theme under this emulation
14:58:13linuxstbwhoever_: Do you compile Rockbox yourself, or download pre-built binaries?
14:58:47whoever_I download it from
14:59:00whoever_(to linuxstb )
14:59:19 Part LinusN
15:00:09 Join nplus [0] (n=nplus@141.25.Globcom.Net)
15:00:12linuxstbwhoever_: Ah, I think you need to use l|0|0|.... - to load the image, and then use "" by itself later in the WPS
15:00:27*linuxstb wonders where his % C went
15:00:47linuxstbThat should be % C l
15:00:50linuxstb(no spaces)
15:00:58whoever_oh sorry linuxstb , i don't understand everything
15:01:17whoever_i shout use there %Cl|0|25|75|75|?
15:01:41whoever_and you say " and then use "" by itself later in the WPS" - i don't understand it (I'm german:))
15:02:40whoever_what have I to do?
15:03:13linuxstbAdd a line to the end of your .wps file, just containing ""
15:03:23 Quit kushal_12_27_200 ("This computer has gone to sleep")
15:03:56whoever_ok, with this "", oder just a blank line?
15:03:58linuxstbI meant "% C" (no space
15:05:03 Join nicktastic [0] (n=nick@unaffiliated/nicktastic)
15:05:03 Quit BigBambi ("CGI:IRC (EOF)")
15:05:13 Join BigBambi [0] (n=86ceaf37@rockbox/staff/BigBambi)
15:05:20whoever_ok, i try:)
15:05:33 Quit webguest94 ("CGI:IRC")
15:06:36amiconnbarrywardell: Odd...
15:07:01barrywardellamiconn: are you referring to my commit?
15:07:35barrywardellI remember that being a problem with power_off() in the past too
15:07:48whoever_hmm... linuxstb , there are no pictures:( - but now the design works
15:08:42 Quit vcf ("CGI:IRC")
15:10:11whoever_just the picture is not there:(
15:10:12amiconnbarrywardell: Yes. It's odd that this problem simply doesn't exist on PP5002
15:10:46amiconnAnd poweroff is a different situation - the power doesn't vanish immediately
15:11:27***Saving seen data "./dancer.seen"
15:15:28whoever_linuxstb, have you an idea, why there is no picture? everything is okay, but there are no pictures:(
15:16:04whoever_linuxstb, perhaps because the most are JPEGS, not BMP, I'll try it :)
15:16:11whoever_thank you:)
15:17:24 Quit crope` (Read error: 110 (Connection timed out))
15:17:38barrywardellamiconn: the problem is probably mostly related to the as3514, which only the sansas have
15:18:05 Quit maddler (Remote closed the connection)
15:18:26 Join maddler [0] (
15:19:34 Join crope` [0] (
15:20:59amiconnbarrywardell: When a reset signal is asserted, the cpu shouldn't continue processing the old instruction stream. That's not AS3514 specific
15:21:21amiconnBut there *is* that reboot inconsistency on PP502x - and it's not sansa specific
15:23:43linuxstbwhoever_: Yes, you can only use bmps...
15:23:48barrywardellamiconn: ah, ok. I thought the main problem was with the i2c write going wrong and the while after reboot is just an extra precaution
15:24:38whoever_linuxstb, and i have to copy them in the direcotrie, thanks alot
15:24:49 Quit nplus (Remote closed the connection)
15:25:10 Join nplus [0] (n=nplus@141.25.Globcom.Net)
15:31:13 Join nicktastique [0] (n=nick@unaffiliated/nicktastic)
15:31:32 Quit nplus (Remote closed the connection)
15:31:56gevaertsI have some signal quality results. I couldn't test high speed properly, because for some reason the test mode still doesn't work, and the alternative plan of getting a device descriptor and looking at that didn't work because it comes bac too late (can be improved by moving this back to the interrupt handler for test purposes)
15:31:59 Join nicktast1que [0] (
15:32:08gevaertsThe good news is that full speed signal quality looks fine
15:32:15 Join nplus [0] (n=nplus@141.25.Globcom.Net)
15:33:41 Join oldbull [0] (n=d4f23d75@gateway/web/cgi-irc/
15:36:31pixelmalinuxstb: I at least plan to test the new mod codec (and comment back in the tracker) with the few files I have here. Don't think I'll commit it though
15:37:02barrywardelloops, can i cancel a commit? or commit a reverted version?
15:37:10oldbullQ (scanned the site for answer to this a while ago, so please have mercy if this is answered somewhere else): why is it that I cannot find a partition table for my 20gb 3rd gen ipod?
15:37:21linuxstbbarrywardell: I think you can only commit the reverse...
15:37:48 Join parafin|away [0] (
15:37:55linuxstboldbull: Simply because no-one has volunteered one...
15:38:16linuxstboldbull: Which OS are you running on your computer?
15:38:37oldbulllinuxstb: linux (ubuntu feisty)
15:38:47*gevaerts notices that the build screen is all green with just a tiny yellow line
15:39:45linuxstboldbull: Then you can manually partition according to this page -
15:41:01 Join daurnimator [0] (i=daurn@unaffiliated/daurnimator)
15:41:02linuxstboldbull: I would use the command-line fdisk program, rather than anything graphical (which may not be able to handle the "empty" partition needed)
15:41:13pixelman1s: about the backlight taking so long sometimes - IIRC jhMikeS said here that there is a priority problem now or something like that, but for me it's not obvious at all.
15:41:34oldbulllinuxstb: thanks for the tip, I'll give it a try
15:41:38barrywardellgevaerts: at least some good came of my messed up commit ;)
15:41:43n1spixelma: it's not consistent for me either...
15:42:29n1sbut sometimes it's really noticeable, like I push a button and it takes so long to turn on that I think I have enabled hold...
15:44:11 Join perrikwp [0] (i=982175a9@gateway/web/ajax/
15:44:22 Join perrikp [0] (i=982175a9@gateway/web/ajax/
15:45:11 Quit perrikwp (Client Quit)
15:47:27 Quit parafin (Read error: 113 (No route to host))
15:47:37 Quit daurn (Read error: 110 (Connection timed out))
15:48:22 Quit nicktastic (Read error: 110 (Connection timed out))
15:48:36 Nick nicktast1que is now known as nicktastic (
15:49:05oldbulllinuxstb: noob as I am, how big should I make the firmware partition?
15:50:48linuxstboldbull: I think it comes from the factory at about 40MB on the earlier ipods - see the example partition layouts near the end of that page I linked to - the 4G 20GB layout should work for you.
15:51:26oldbulllinuxstb: ok, thanks
15:53:26*gevaerts seems to have bought an ipod video now
15:54:07linuxstbgevaerts: Do you know which type (5th or 5.5th gen)?
15:54:22gevaertslinuxstb: no. It is 30G and black
15:54:35linuxstbThen I think it could be either.
15:54:48LambdaCalculus37gevaerts: Can you read the serial number on the back of it?
15:55:11gevaertsLambdaCalculus37: not yet. I expect I'll get it in my hands sometime this week
15:55:37LambdaCalculus37gevaerts: Okay. When you get it, we can identify if it's a 5G or a 5.5G by the serial number.
15:56:10oldbulllinuxstb: fdisk tells me that the ipod doesn't contain any known partition table (it was formatted with itunes+os x).. hence I don
15:57:29gevaertsbarrywardell: at least now we know what the expected final delta will be when usb will be enabled
15:58:18oldbulllinuxstb: continued :) : and it has created a new partition table for me, alas I cannot delete any partions... would it be safe to continue the steps anyway? just want to be 100% sure, as I have no (easy) way to restore my ipod if this doesn't work
15:59:14oldbulllinuxstb: there ARE stupid questions.. please ignore me, I fell in an irrational hole
15:59:29barrywardellgevaerts: yes, for H10 and e200 anyway. It's an impressive delta too!
15:59:48 Join DaCapn [0] (
16:00:05 Quit knittl (Read error: 110 (Connection timed out))
16:00:18linuxstboldbull: Can I just check, if you type "fdisk -v", what is displayed? (there are different fdisks around...)
16:00:49oldbulllinuxstb: fdisk (util-linux-ng 2.13)
16:01:15linuxstboldbull: That's the right answer :) So yes, you can create a new partition table, and then create the two required partitions. This should work...
16:01:31oldbulllinuxstb: I'll give it a try...
16:01:42linuxstboldbull: I'm assuming it was HFS (Macpod) before?
16:03:16*amiconn doesn't see a delta...
16:03:32*pixelma guesses that was the joke...
16:05:30gevaertsIndeed. Basically the only effect of USE_ROCKBOX_USB right now is to reboot instead of switching to UMS mode, but all code is already there. If we want to, it would be trivial to replace that #ifdef by a debug menu setting.
16:06:14*gevaerts reversed something in that line...
16:06:37oldbulllinuxstb: yeah
16:08:56 Join mf0102 [0] (n=michi@
16:12:13 Nick perrikp is now known as perrikwp (i=982175a9@gateway/web/ajax/
16:12:32 Join gregzx [0] (
16:16:32oldbulllinuxstb: worked well, but somehow my ipod restarted, and now only shows the "disconnect cable" symbol, before it goes blank again.. cannot seem to enter disk mode.
16:18:04oldbulllinuxstb: pointed the firmware directly to the sdc1 (in my situation) partition
16:20:30*oldbull researches possible solutions
16:25:54whoever_hi together, can somebody explain me, what "%CfFFFFFF%s%t5%al%?ia" in the WPS-file dos? the question ist just about the "%CfFFFFFF"?
16:26:08whoever_what does it means?
16:26:51oldbullwell, my ipod is in fact diplaying "connect cable" symbol, and then goes blank... no matter if my cable is connected or not.. weird
16:28:39linuxstboldbull: I think it's trying to update the firmware in flash, and telling you to attach the wall charger.
16:31:43 Quit Zagor ("Client exiting")
16:33:09oldbulllinuxstb: that worked, but I had to replace "wall charger" (which I for some pure logical reason which I cannot explain in detail) with my powerbook, booting from an installation cd as the hd has crashed.. this linux box has no firewire port... :-)
16:33:33oldbull*add point that the wall charger was thrown away*
16:33:46 Quit nplus (Remote closed the connection)
16:34:09oldbulllinuxstb: anyway, thanks a million.. now, how can I upload my partition table for others to use?
16:38:15 Quit gregzx (Connection timed out)
16:41:19linuxstboldbull: "dd if=/dev/sdc1 of=mbr-2g20gb.bin count=1"
16:41:50linuxstbSorry, I meant /dev/sdc (not /dev/sdc1)
16:43:31 Quit wisp_sh (Read error: 110 (Connection timed out))
16:44:18oldbulllinuxstb: and 3g20gb.bin I presume...
16:44:30n1sany objections against moving bitswap.S and descramble.S into target/sh? amiconn?
16:45:06linuxstboldbull: Oops, thought you said you had a 2nd gen...
16:45:26 Join knittl [0] (n=knittl@
16:45:36linuxstbActually mbr-3g-20gb.bin would be consistent with the other filenames.
16:47:48 Quit knittl (Remote closed the connection)
16:47:49oldbulllinuxstb: aight, it is done.. should I send it to you?
16:48:06linuxstbDo you have any web space to host it? Otherwise you can email to me.
16:48:33oldbulllinuxstb: I would prefer email
16:50:13linuxstboldbull: Check your PMs...
16:50:52oldbulllinuxstb: check your email ;-)
16:52:38linuxstbThanks ;). Bagder - could you put this file on the download server?
16:55:04oldbullah, finally, wisdom once again conquers technology
16:55:31oldbullwell, I get an "undefined instruction" error upon rebooting into rockbox...
16:56:12pondlifejhMikeS: Did you see this claim?:;topicseen#msg117784
16:57:06 Quit desowin (Excess Flood)
16:57:15linuxstboldbull: At what stage?
16:57:26 Join desowin [0] (n=desowin@atheme/developer/desowin)
16:57:41pixelmaguess that's the current 1st-3rd gen Ipod problem...
16:57:53amiconnn1s: If you do that, descramble.S should go one level deeper, into target/sh/archos
16:58:09n1samiconn: ok, will commit with that change
16:58:13linuxstbpixelma: I thought that was always a freeze?
16:58:14oldbulllinuxstb: "undefined instruction at 00045AA0 (0)" with the rockbox logo in the downright corner... it also comes with (1), without the logo
16:58:15*linuxstb spots amiconn...
16:58:31oldbulllinuxstb: that's the only thing I see..
16:58:40linuxstbamiconn: oldbull has a 3rd gen - is this the "usual" problem?
16:58:44 Join FOAD_ [0] (
16:59:49linuxstboldbull: OK, so you've hit a known issue - the only solution is to keep trying older builds until you find one that works.
17:00:02linuxstb(no need to change the bootloader - just the
17:00:25oldbulllinuxstb: allright! thanks :-)
17:00:35pondlifelinuxstb: Is that an alignment issue?
17:00:36pixelmalinuxstb: no, it shows different symptoms from freeze at Rockbox logo, data aborts etc. -here's an example list (a bit old though)
17:00:50amiconnOr build yourself and put a number of 'nop's into main.c. The correct number is determined by experimentaion (0 to 3)
17:01:08pondlifeamiconn: Hmm, any idea what's misaligned?
17:01:54amiconnNot yet. I need to perform a binsearch where the 'nop' placement stops having an effect
17:01:58 Join jgarvey [0] (
17:02:01 Quit argumentD (Read error: 113 (No route to host))
17:02:35pixelmapondlife: seen Lear's new patch in the tracker (the 2 seconds problem)?
17:02:44pondlifepixelma: Not yet :)
17:02:56pondlifeAh, you mean the end of playlist one
17:03:12pondlifeI thought you meant the LCD backlight takes 2 seconds one
17:03:19pondlifeYes, I've tried it and it works
17:03:36pondlifeI could commit, but thought it might be nicer to let Lear do that
17:04:28pixelmayes, I meant the end of playlist one. I only tested if it stops correctly now, nothing else and it works correctly for me too
17:04:54pondlifeMe too. I don't use bookmarks, so can't be sure about that
17:05:12pixelmasame here
17:05:51pondlifeamiconn: Do you think it likely that the odd noises some users are hearing on crossfade ( could be caused by PortalPlayer CPU frequency switching? Or is that unlikely?
17:06:53 Join Horscht [0] (n=Horscht@xbmc/user/horscht)
17:07:39amiconnI don't think so. It rather sounds like the same issue as keyclick on pp
17:07:54oldbulllinuxstb: working now.. thanks a lot for the help :-)
17:07:57amiconnThere must be something wrong in pcm handling (dual core related??)
17:08:28rasherIs scrobbling broken recently? Or will I need to investigate what's going on?
17:08:56pondlifeamiconn: Ah, yes - dual-core...
17:09:10pondliferasher: Probably worth you checking.
17:09:19peturrasher: it worked last time I used it (yesterday)
17:10:04peturhmmm build might have been a few days old
17:10:30pondliferasher: If you do check, please could you try with the patch on too
17:10:49pondlifeThat will be committed soon, I expect
17:11:28pondlifeAlso, anyone know if gather runtime data is still broken? (
17:11:29***Saving seen data "./dancer.seen"
17:12:03 Join toffe82 [0] (
17:12:08LambdaCalculus37pondlife: The link to FS #8529 404'd.
17:12:24pixelmanot here
17:12:29pondlifeAh. the closing bracket got caught
17:13:12pixelmadepends on the IRC client then
17:13:19rasherCould logf break scrobbling perhaps?
17:14:55 Quit FOAD (Read error: 110 (Connection timed out))
17:14:55 Nick FOAD_ is now known as FOAD (
17:16:41 Quit XavierGr (Nick collision from services.)
17:16:54 Join XavierGr [0] (n=xavier@rockbox/staff/XavierGr)
17:20:38Tuplanollaoh yeah, icons in rockbox site navigation bar o/
17:24:01 Quit linuxstb ("Client Exiting")
17:30:30 Nick hcs is now known as hcslogbo1 (n=agashlin@rockbox/contributor/hcs)
17:30:52 Nick hcslogbo1 is now known as hcs (
17:31:20pondliferasher: I would hope logf would not affect anything much
17:31:44ArathisIs it a normal behaviour on H10 20gb, that when pressing a key while the backlight is off there's a delay of one or two seconds before the backlight turns on again?
17:32:21rasherpondlife: things seem to work fine now. With or without logf. Can't seem to reproduce, and I *think* I may have been running a modified build apart from the logf as well
17:37:07rasherpondlife: I take that back. It doesn't seem to work with logf
17:37:40pondlifeI'd suspect something else. logf doesn't even spin up the disk...
17:38:00pondlifeMight be just intermittent?
17:38:22rasherWell I can't say that for sure of course
17:40:30 Join webguest33 [0] (n=d9c70784@gateway/web/cgi-irc/
17:41:13 Quit gevaerts ("work->home")
17:41:13 Quit petur ("sport time")
17:42:48 Quit whoever_ ("Bin dann mal weg - allen noch 'nen schönen Tag")
17:43:41 Join faemir [0] (
17:43:54 Quit tchan (SendQ exceeded)
17:44:15 Join wisp_sh [0] (
17:44:20 Quit hcs ("Leaving.")
17:44:29*GodEater looks for the "delete chat lines in the wrong channel" key
17:44:36 Join hcs [0] (n=agashlin@rockbox/contributor/hcs)
17:45:30 Join tchan [0] (n=tchan@lunar-linux/developer/tchan)
17:53:09 Quit webguest33 ("CGI:IRC")
17:55:13 Join Mathiasdm [0] (
17:58:32 Quit BigBambi ("CGI:IRC")
18:00:14rasherpondlife: It seems pretty consistent - switching back and forth between logf and non-logf builds breaks and unbreaks scrobbling
18:00:39pondlifeGood luck in working that one out
18:01:58 Quit ender` (" Q: What's another name for the "Intel Inside" sticker they put on Pentiums 4? A: Warning label.")
18:09:30 Join mchua [0] (
18:12:04 Join TMM [0] (
18:13:07 Join hannesd_ [0] (
18:14:30 Nick mchua is now known as mchua|stairs (
18:16:05 Quit courtc (Read error: 113 (No route to host))
18:16:44 Join courtc [0] (
18:17:59 Join argumentD [0] (
18:18:31 Nick mchua|stairs is now known as mchua (
18:19:43 Quit TMM (Read error: 104 (Connection reset by peer))
18:20:02 Join TMM [0] (
18:20:10 Nick hcs is now known as hcslogbot (n=agashlin@rockbox/contributor/hcs)
18:21:18 Nick hcslogbot is now known as hcs (n=agashlin@rockbox/contributor/hcs)
18:29:50 Quit hannesd (Read error: 110 (Connection timed out))
18:29:51 Nick hannesd_ is now known as hannesd (
18:31:03 Quit TMM ("Ex-Chat")
18:31:22 Join TMM [0] (
18:33:20 Join gevaerts [0] (n=fg@rockbox/developer/gevaerts)
18:33:35 Nick mchua is now known as mchua|testing_aw (
18:33:46 Nick mchua|testing_aw is now known as mchua_test_away (
18:37:46 Join OlivierBorowski_ [0] (
18:41:20 Join waldo [0] (
18:41:50 Join Domonoky [0] (
18:43:00 Join Davide-NYC [0] (
18:45:05ArathisIs there a XCF, PSD, SVG or whatever available for the tango iconset which is now standard in rockbox?
18:45:35 Join ANNO [0] (
18:45:37Davide-NYCWhat is the current state of recording on PortalPlayer targets? (I'm posting fixing recording on PP to the GSoc wiki page)
18:46:21LambdaCalculus37Davide-NYC: Hey there!
18:46:25ArathisI tried to resize it, but I get problems with transparency.
18:46:51 Join BigBambi [0] (n=Alex@rockbox/staff/BigBambi)
18:47:14LambdaCalculus37Davide-NYC: Recording seems to work on the Sansa (my c240 makes for a good voice dictator), but I don't think it's working very well on the iPods or the H10.
18:47:43ArathisLast time I checked it worked pretty well on my H10/20GB
18:47:44Davide-NYCCAn you change bitrate, samplerate, channels? Can you encode to a variety of formats?
18:48:14LambdaCalculus37Let me dig my c240 out.
18:48:57Davide-NYCSaratoga said there are still issues and I don;t own a PP target.
18:49:19LambdaCalculus37My Sansa allows me to record to MP3, WavPack, AIFF, and WAV.
18:49:19 Join PaulJam [0] (
18:49:21ArathisI can adjust all these settings on my H10
18:49:40LambdaCalculus37I can as well on my Sansa.
18:49:54Davide-NYCbitrate, samplerate, channels? What about AGC?
18:50:08LambdaCalculus37Arathis: AGC on your H10?
18:50:34LambdaCalculus37No sign of AGC on the c240.
18:51:24Davide-NYCCan you guys confirm or deny the statement on the wiki (iPodStatus) about the status of recording on these targets?
18:51:41DomonokyAGC depends on the gain steps the recording hardware can give, so its not possible on all players..
18:51:47pixelmaLambdaCalculus37: the sansas also have a hardware limitation (sample rate) the other PP don't have
18:52:17LambdaCalculus37pixelma: Ah, thanks for pointing that out for me. :)
18:52:59LambdaCalculus37Davide-NYC: Going into the recording screen on the iPod does break playback, requiring a reset with MENU+SELECT.
18:53:11Davide-NYCBasically, all of the recording work done for the coldfire targets should be done for PP targets. (within hardware limitations of course)
18:53:25n1sDavide-NYC: IMHO fixing bugs isn't suited as a gsoc project (iiuc recording works on at least some ipods but they need to be restarted afterwards)
18:53:25pixelmadue to the AMS chip, jhMikeS said that with lots of optimisation it *might* be possible to get 32kHZ (i think it was). Currently only 22.05kHz is possible
18:53:27LambdaCalculus37Davide-NYC: That's a very good proposal.
18:54:31PaulJamhi, quick question: i haven't yet tried with an official build, but is there currently a known issue with codec changes that playback stops shortly after the start of the track when going from ogg/vorbis to mp3 or the other way around?
18:54:53pixelmathat's already better than the OF (at least on my c200 it only recorded 16kHz wav files)
18:55:06LambdaCalculus37Davide-NYC: I was wondering if you can take a look at an iriver H120 for me. It belongs to a friend of mine, and he's getting ATA -32 errors on it as of late.
18:55:15pondlifePaulJam: I've heard reports of stopping about 5 seconds in..
18:55:31Davide-NYCReally busy. Did he swap out the drive?
18:55:34pondlifePaulJam: FS #8455 maybe?
18:56:06LambdaCalculus37The drive works. I placed it in an enclosure, and my Linux machine and a Windows 2000 laptop both picked it up.
18:56:15LambdaCalculus37I think it's something wrong with the HD connector.
18:57:16Davide-NYCn1s: point well taken. I am myopic when it comes to rockbox (all I see is the recording part)
18:57:22 Quit OlivierBorowski (Read error: 110 (Connection timed out))
18:57:52PaulJampondlife: doesn't seem to be the same issue, playback just stopped in the WPS. so i had to stop and reatert playback. i'll see if i can reproduce with the current build.
18:58:05Davide-NYCLambdaCalculus37: Oh boy, those pins are teeny tiny. If that is the problem I can't fix that. Have them try a different drive regardless.
18:58:17Davide-NYCGotta run!
18:58:22LambdaCalculus37Davide-NYC: See ya!
18:58:25 Quit Davide-NYC ("ChatZilla 0.9.81 [Firefox]")
18:59:54ArathisAnybody knows Christoph Reiter's nick or whether he is around here at all? Wanted to ask him something about his iconset.
19:02:42 Join ompaul [0] (n=ompaul@gnewsense/friend/ompaul)
19:02:50 Quit midgey (Read error: 110 (Connection timed out))
19:06:26pixelmaArathis: rasher posted a link to the tango icon set "design guidelines" a few days ago. IIRC it suggested that the original icons should be made as svg so maybe those are available somewhere from the project's site
19:08:50 Join Arathis2 [0] (
19:09:25 Quit Arathis (Nick collision from services.)
19:09:58 Nick Arathis2 is now known as Arathis (
19:11:33***Saving seen data "./dancer.seen"
19:11:40Arathispixelma: got my answer to your suggestion?
19:12:28Arathis"which leads to catching them one by one and manage them as an iconset :/"
19:13:01*LambdaCalculus37 fiddles with the H120 a little further
19:13:17LambdaCalculus37I think the hard drive itself *is* the problem.
19:13:30LambdaCalculus37Even just copying a new Rockbox build to it is taking forever. =/
19:17:34pixelmaArathis: I think you can do the resizing and still have the transparency right, just don't resize the complete bmp - explaining here would be a bit off-topic though as it is a graphic editing question...
19:19:23 Nick fxb__ is now known as fxb (
19:27:00 Quit Siku (Read error: 131 (Connection reset by peer))
19:27:18 Join Siku [0] (
19:29:51 Quit PaulJam (".")
19:31:07 Part pondlife
19:36:16*LambdaCalculus37 kicks Windows for stalling with writing data
19:37:57 Join Lear [0] (i=chatzill@rockbox/developer/lear)
19:38:47*LambdaCalculus37 kicks Windows again
19:39:18*gevaerts is trying to understand the USB CDC spec enough to make usb serial work on Windows
19:39:30LambdaCalculus37Back in a bit.
19:39:32 Quit LambdaCalculus37 ("CGI:IRC 0.5.9 (2006/06/06)")
19:42:16 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
19:42:46 Quit XavierGr (Read error: 110 (Connection timed out))
19:50:48 Quit ANNO ("Miranda IM! Smaller, Faster, Easier.")
19:50:58 Quit oldbull ("CGI:IRC")
19:51:29Bagderlinuxstb: mbr-3g-20gb.bin is now on the download server!
19:52:14linuxstbBagder: Thanks - hope the servers can take the bandwidth ;)
19:52:54*linuxstb wonders what the http overhead is
19:53:48LearNico_P: Seems like that old "track skips too early" bug is back again. At least if I seeks back a bit right after resume, it can skip up 30 seconds too early.
19:56:07 Join PaulJam [0] (
19:56:56 Join LambdaCalculus37 [0] (n=44a0430f@gateway/web/cgi-irc/
19:57:51 Join austrian_guest [0] (n=3e2fb224@gateway/web/cgi-irc/
19:59:35Nico_Pfix one bug an three others come crawling back :/
20:00:23austrian_guestHi! does anyone of you know how pcm_play_data works?
20:00:36 Quit EspeonEefi ("さよなら")
20:00:52austrian_guestIt works partly, but eventually it crashes
20:01:26austrian_guesteven though I've made a valid callback funktion...
20:02:44 Quit perrikwp (" ajax IRC Client")
20:04:32*gevaerts kicks Windows
20:04:49n1saustrian_guest: the midi plugin uses it, take a look there
20:05:17*LambdaCalculus37 kicks Windows
20:05:44gevaertsLambdaCalculus37: did you kick too hard last time ?
20:05:57LambdaCalculus37Stupid system stalled while I was trying to back up an iriver hard drive.
20:06:23LambdaCalculus37I'm going to have to break out the heavy weaponry later on tonight.
20:07:11austrian_guest@ nls: I looked at all plugins that use it, but no luck.
20:07:53austrian_guestThe problem is that I want to play the same buffer again and again - pcm_play_data doesn't seem to like that.
20:08:01cgaustrian_guest: try smaller buffer?
20:08:40austrian_guestIt plays fine the first time, but when my callback function tells pcm_play_data to play the same buffer again, it crashes.
20:08:52cgi'm right now writing plugin that uses pcm_play_data and use buffer of 256 to 1024
20:09:05n1saustrian_guest: the buffer is unchanged?
20:09:24 Join bluebrother [0] (n=Dom@rockbox/staff/bluebrother)
20:09:32gevaertsLambdaCalculus37: try an emergency hammer
20:09:40cgmake sure to return proper size and start within callback
20:10:15austrian_guestOf course I return proper size and start - it works the first time, but not the second time!
20:10:30austrian_guestalthough I return the same start address and size!
20:11:29n1smaybe you could pastebin some code?
20:11:58*LambdaCalculus37 pulls out an enourmous mallet
20:12:26LambdaCalculus37Actually, that isn't needed.
20:12:30*LambdaCalculus37 throws it aside
20:13:19LambdaCalculus37gevaerts: I may have an idea what's wrong with the drive, anyway. I think the filesystem on the hard drive is badly corrupted.
20:13:31austrian_guestsndbuf[] contains the audio data.
20:13:40LambdaCalculus37That may be what was also causing some of the ATA -32 errors I was seeing.
20:13:57 Join perrikwp [0] (i=98213d71@gateway/web/ajax/
20:14:09n1saustrian_guest: why the if(count<2) ?
20:14:56n1sthat makes the assignements in the callback only happen the first time
20:15:04austrian_guestthis is some debugging code
20:15:05gevaertsLambdaCalculus37: something like trying to access past the end of the device ?
20:15:15austrian_guestso it only plays it one time
20:15:28austrian_guestwhen I don't write that, it crashes
20:15:56 Join XavierGr [0] (n=xavier@rockbox/staff/XavierGr)
20:16:47LambdaCalculus37gevaerts: I'm not 100% sure. What happened is that this iriver H120 belongs to a friend of mine, and he told me that Rockbox refuses to boot, instead giving him an ATA -32 error.
20:17:20LambdaCalculus37Windows and Linux can still access the filesystem, but Windows choked when I tried writing to the volume.
20:18:32n1saustrian_guest: in that code sample the buffer is changed, if this happens while playing that buffer you might get something nasty
20:19:00n1s(the callback is executed on another thread)
20:19:20amiconnaustrian_guest: The splash() in your callback is the problem
20:19:26gevaertsLambdaCalculus37: chkdsk would probably be a good idea, with possibly a surface scan I guess
20:19:35austrian_guestwhere is the buffer changed?
20:19:50 Join Zom_ [0] (
20:19:50n1sah right, single threaded gui thingy...
20:20:09 Part Zom_
20:20:20amiconnThe callback is called in isr context, and splash() with a timeout yields() - which *will* cause havoc when called from isr
20:20:45n1saustrian_guest: I was just saying that if you did that in some kind of loop it could be bad but amiconn spotted the real problem
20:20:55LambdaCalculus37gevaerts: Bad news... I don't have administrative access on my work PC, so I can't run chkdsk.
20:21:34austrian_guestI think I've solved part of the problem - appearantly the plugin ended while pcm was still playing. While the splash() at the end of the plugin was shown (for 2 secs) it worked, but everything longer than that crashed it...
20:22:06n1syes pcm_play_data returns immediately
20:22:11amiconnAlso, calling stuff that draws to the display from anywhere else than the main thread must be used with extreme caution. Avoid it if at all possible
20:22:44amiconnIf you need to debug an isr, use logf() - much cleaner
20:23:05 Quit TMM (Remote closed the connection)
20:23:28austrian_guestI'm not a very good coder (yet), so forgive me ;)
20:24:12LambdaCalculus37austrian_guest: Everyone has to start somewhere, right? ;)
20:25:17 Join olejorgenb [0] (
20:25:19austrian_guesthow does logf() work?
20:25:21amiconnAs a last resort, you *can* use slpash() for debugging - but when calling it from an ISR, *always* set the timeout to 0 (which will make the splash non-blocking and not yield, but even then you need to be extremely careful regarding timing
20:25:55austrian_gueste.g. where does logf write to?
20:26:23amiconnWith a 256 sample buffer and 44100Hz sample rate, you callback will be called almost every 5 milliseconds - adding a splash() might slow it down too much
20:26:32n1saustrian_guest: it needs to be enabled in the configure and with some define in your file and then it dumps strings into a buffer with you can dump from the debug menu basically
20:28:04amiconnIf your target has a supported lcd remote (i.e. is an iriver H1x0, H300, or an iAudio X5 or M5), you can also watch logf() output "live" on the remote lcd
20:28:12n1sUsing debugf in the simulator is a good deal more comfortable but might not work in all cases
20:28:42*gevaerts gives up his attempts to get the usb serial support to work on windows.
20:28:52*amiconn thinks we should probably unify debugf() and logf()
20:28:58n1swho uses windows anyways? ;)
20:29:21amiconnThat is, have only one, and depending on whether you build a logf build or a debug build, output one or the other way
20:29:44amiconn(or even both, if that makes sense)
20:30:17amiconnSims would always output "debugf way"
20:30:29n1ssounds reasonable
20:31:26amiconnDEBUGF() is a macro, so nothing stops it from resolving into 2 function calls
20:32:26amiconnHmm, there might be side effects if arguments are calculated twice, but that can be solved
20:32:44austrian_guestOk, now it works, but I have a new problem: there's always a pause (>100 millisecs) between the sounds.
20:33:14austrian_guesti.e [sound] pause [sound] pause ...
20:33:38n1stry a larger buffer
20:33:55gevaertsamiconn: Aren't callers that give those side effects broken anyway ? After all they can just as well not be called at all.
20:33:58 Nick mchua_test_away is now known as mchua (
20:34:10amiconnDEBUGF() could resolve to logf() if only logf is enabled, and if debugf is enabled, resolve to debugf(). Then debugf() would additionally call logf() if that's enabled too
20:34:17 Join nplus [0] (
20:34:49amiconngevaerts: In this case, true.
20:35:06 Join sinsch [0] (
20:35:22amiconnThat is, not entirely true. The DEBUGF() calls could rely on side effects for debug-only variables
20:35:25sinschhello everyone
20:35:27austrian_guestMy buffer is 88200 bytes large
20:35:40austrian_guestI doubt I should make it bigger
20:35:59amiconnStill fishy, but better implement a reasonable workaround that causing potentially hard to find problems
20:36:02gevaertsamiconn: the problem is of course also that you don't want to risk breakage on reworking _debug_ output.
20:36:50n1saustrian_guest: so what did you change in the code? this is quite hard without seeing it...
20:37:49austrian_guestI put a "while (rb->pcm_is_playing()); " at the end of the plugin
20:38:06austrian_guestSo it does not exit while it's still playing
20:38:24 Join ojb [0] (
20:38:26 Quit olejorgenb (Read error: 104 (Connection reset by peer))
20:38:29 Join EspeonEefi [0] (i=espeonee@CERULEANCITY.MIT.EDU)
20:38:57n1saustrian_guest: did you remove the splash in the callback?
20:40:46gevaertsamiconn: I've been looking at this SYS_HOTSWAP_EXTRACTED issue. It seems that ata-c200_e200.c actually tries to mount the disk to see if it's there, in order to decide if it's an insertion or extraction event.
20:41:04amiconnOdd system...
20:41:24*gevaerts agrees
20:41:25amiconnDid you compare that with how the MMC driver handles things?
20:42:52n1saustrian_guest: you might have hit one of the nice cooperative mutithreading thingies :) try putting a rb->yield() inside your while() loop
20:42:53gevaertsamiconn: yes. That one looks much saner.
20:43:25amiconnGuess who wrote that driver? ;)
20:43:45*gevaerts doesn't have the slightest idea ;)
20:44:20gevaertsAlthough svn annotate gives some hints...
20:44:31*BigBambi can guess
20:45:50gevaertsata-c200_e200.c seems to debounce its insert-detection by removing swap notifications from its queue when it detects a change
20:53:29 Join Kage [0] (
20:54:29 Quit barrywardell (Remote closed the connection)
21:00:42*gevaerts just wanted to make usb work, not look at every single other driver
21:03:17 Quit sinsch ("Leaving")
21:05:49 Join Benjamin_L [0] (
21:09:14 Nick parafin|away is now known as parafin (
21:09:49Benjamin_Lgevaerts: this is what i get on my desktop pc when connecting my e260: don't know if this is due to the newly introduced gvfs in ubuntu hardy, but my usb 2 high speed card reader works flawlessly
21:10:11Benjamin_Lgevaerts: the message is shown in a popup
21:11:27gevaertsBenjamin_L: can you also paste the dmesg output ?
21:11:34***Saving seen data "./dancer.seen"
21:11:38Benjamin_Lits above the message
21:12:09gevaertsBenjamin_L: there should be something after that.
21:12:37Benjamin_Lgevaerts: i swear i don't have more, when i disconnect the sansa i get a usb disconnect
21:13:04gevaertsso no 'usb-storage: device scan complete' or error ?
21:13:26 Quit ch4os_ (Read error: 104 (Connection reset by peer))
21:13:29Benjamin_Lon my usb 1.1 notebook it works flawlessly
21:13:32*gevaerts doesn't understand that
21:13:54 Quit mf0102 ("Verlassend")
21:14:21Benjamin_Lwell, maybe we can blame this on the ubuntu devs, i don't know if the changed much due to the switch to gvfs and gio, can't say fot sure
21:14:46 Quit courtc (Read error: 113 (No route to host))
21:14:57gevaertsBenjamin_L: any chance of trying this at the console, without gnome running ?
21:18:39 Join AceNik [0] (n=AceNik@
21:18:49 Part AceNik
21:19:00Benjamin_Lgevaerts: I'll try, just checked my brothers windows laptop, it recognizes the devide, rockbox internal... but it doesn't show the player either
21:19:07 Quit desowin (Read error: 113 (No route to host))
21:19:39 Join AceNik [0] (n=AceNik@
21:20:32 Part AceNik
21:20:36gevaertsBenjamin_L: what svn revision are you running ? And do you have high speed enabled ?
21:20:48Benjamin_Lgevaerts: forgot to say its r16509 with high speed
21:21:31 Quit perrikwp (" ajax IRC Client")
21:22:54 Quit nplus (Remote closed the connection)
21:23:09 Join jcollie [0] (n=jcollie@
21:23:29gevaertsThat should be recent enough. Later changes are mostly cosmetic
21:23:33 Join LinusN [0] (n=linus@rockbox/developer/LinusN)
21:25:04 Join Buschel [0] (
21:25:34 Part LinusN
21:25:39Benjamin_LI'll test console now
21:25:41 Quit Benjamin_L ("Ex-Chat")
21:25:52 Join LinusN [0] (n=linus@rockbox/developer/LinusN)
21:27:10 Quit corevette (Read error: 110 (Connection timed out))
21:27:57 Join Benjamin_L [0] (
21:29:09Benjamin_Lgevaerts: seems to be a gnome problem:
21:30:05gevaertsBenjamin_L: does the OF get recognizedproperly while gnome is running ?
21:31:07Benjamin_Lgevaerts: obviously not, didn't try yet, but it doesn't work
21:31:25Benjamin_Lsome error about not exclusive access
21:31:37Benjamin_LI'll file a bug
21:31:57 Quit EspeonEefi ("さよなら")
21:33:46 Quit PaulJam (".")
21:33:50*amiconn wonders whether the macos crashes where actually caused by the mode sense bugs, and not due to the removable flag
21:33:55gevaertsBenjamin_L: ok
21:34:40Benjamin_Lgevaerts: I already suspected the problem not being on the rockbox side, now we can be sure
21:35:34gevaertsamiconn: the crashes, probably (since that could easily cause buffer overflows if not handled correctly). The removable flag merely caused it not to mount (because it wants a caching mode page on non-removable devices)
21:35:36 Quit austrian_guest ("CGI:IRC (EOF)")
21:36:16amiconnAnd we don't provide that?
21:36:36amiconnAlso, do you already know a solution for the sg_modes -e errors?
21:36:58gevaertsNo, we don't (I'll probably add it though)
21:38:11gevaertsAnd no, I don't know what exactly goes wrong. I did see the same errors from the sansa OF here.
21:38:45gevaertsI'll trace it tomorrow
21:38:46 Join perrikwp [0] (i=982175a9@gateway/web/ajax/
21:39:53 Quit perrikwp (Client Quit)
21:41:09 Quit jcollie ("Ex-Chat")
21:41:17 Join perrikwp [0] (i=982175a9@gateway/web/ajax/
21:41:25 Join gregzx [0] (
21:42:37Benjamin_Lgevaerts: My brother just tried it again after rebooting his wondows laptop, it still doesn't show, but when disconnecting the usb cable, he get a removable device icon in explorer for a fraction of a second
21:44:32gevaertsBenjamin_L: Can you go to the device manager while it is plugged in, and see if it appears there ? If it does, try uninstalling it, and then unplugging and plugging back in.
21:44:33 Quit Buschel (Read error: 104 (Connection reset by peer))
21:45:10gevaertsBenjamin_L: That might also be caused by the bad signalling, so maybe try a full speed build to see if it improves things
21:47:47Benjamin_Lgevaerts: windows tells us that the device could reach higher speed when connected to a high speed usb port on which it is already connected..
21:48:13Benjamin_Lgevaerts: and it shows in the device manager, but not in explorer
21:49:08gevaertsBenjamin_L: That means it connected at full speed. Are you still running a high speed build ?
21:49:36*bluebrother wonders if FS #1777 should get closed as out of date
21:49:54Benjamin_Lgevaerts: yes
21:50:15 Join desowin [0] (n=desowin@atheme/developer/desowin)
21:50:18gevaertsThen I'd say you have really bad signalling.
21:50:35Benjamin_Lgevaerts: windows popups tell us its installing ubs mass storage device but then nothing, no device it now ready to be used etc.
21:50:39 Join Buschel [0] (
21:51:26gevaertsBenjamin_L: Is that a "normal" pc or a laptop ?
21:51:34Benjamin_Lgevaerts: laptop
21:52:25gevaertsSo probably no other ports to try (at least not really different ones)... Can you do a full speed build ?
21:52:38Benjamin_Lgevaerts: just a second
21:52:49*gevaerts suspects Benjamin_L's brother of running gnome on windows
21:55:01*Benjamin_L wishes my old T23 had usb 2.0 ports ^^
21:55:23 Quit LambdaCalculus37 ("CGI:IRC 0.5.9 (2006/06/06)")
21:55:32amiconngevaerts: Do you already have an idea what might be wrong with test mode?
21:55:43amiconnWhat's that test mode anyway, btw?
21:58:26gevaertsamiconn: there are actually several test modes. Some of them just put the lines high or low (to test voltages), and others specify a specific response to every incoming SOF (start of frame). They are specifically meant for electrical tests
21:58:58gevaertsSee 7.1.20 in the USB 2.0 spec
21:59:09 Quit Kage (Read error: 113 (No route to host))
22:01:24gevaertsAnd no, I haven't made progress yet. I'm starting to suspect that it doesn't work for the same reason that the signals are not good : the transceiver (which actually implements them) isn't initialized correctly. That's just speculation for now though
22:01:51 Join yule [0] (n=user@
22:03:47*gevaerts doesn't like reading 'Contact ARC International for port test mode capabilities.' in the iMX31 spec
22:04:14 Quit Lear ("ChatZilla 0.9.81 [Firefox 3.0b4pre/2008030206]")
22:05:28 Quit linuxstb (Read error: 104 (Connection reset by peer))
22:08:49Benjamin_Lgevaerts: full speed build doesn't show up either
22:10:17gevaertsBenjamin_L: did you connect with a non-USE_ROCKBOX_USB build before ? If so, windows might have kept that configuration. If that's the problem, the easiest solution is trying r16511 or later
22:10:29 Quit yule ("Leaving")
22:10:51Benjamin_Lgevaerts: this was the first time we tried it on his laptop
22:11:11*gevaerts doesn't know then
22:11:34gevaertsBenjamin_L: Is that xp or vista ?
22:11:40Benjamin_Lgevaerts: xp
22:11:57gevaertsIt works on xp here...
22:13:08Benjamin_Lgevaerts: strange, maybe i did something wrong, i put the rockbox usb and high speed in config-e200.h, don't know if this is the preferred way
22:14:12 Quit Buschel ()
22:14:14gevaertsBenjamin_L: I put it in the makefile, but both ways work.
22:14:50 Join petur [50] (n=petur@rockbox/developer/petur)
22:14:57gevaertsBenjamin_L: does the OF connect correctly to that laptop ?
22:16:32Benjamin_Lgevaerts: yes
22:17:28gevaertsBenjamin_L: is there something in the event log ?
22:17:34 Quit bluebrother ("Verlassend")
22:17:53*gevaerts is really nearly out of ideas now. Full speed should just work
22:17:58Benjamin_Lgevaerts: uhm, where do i find this
22:18:54gevaertsBenjamin_L: not sure. Somewhere in the control panel, system tools or something
22:21:10Benjamin_Lgevaerts: says something about controler error
22:22:11gevaertsAny windows specialists around ?
22:22:43peturevent log:
22:22:51gevaertsBenjamin_L: can you put the exact message somewhere ?
22:23:01peturright-click 'my computer' and select 'manage'
22:24:20 Quit nicktastique ("Leaving")
22:26:35 Quit hannesd ("Client suicide")
22:28:27 Nick mchua is now known as mchua|away (
22:28:57gevaertsBenjamin_L: I really don't know what is going on there.
22:29:13Benjamin_Lwe tried on his desktop pc too, exactly the same error
22:30:37gevaertsDid you try different usb ports there ?
22:31:54Benjamin_Lthis one only has back ports
22:35:36 Part LinusN
22:36:05gevaertsThose should be ok. I just retried on windows here, and it still works (latest revision). It does get error 11 as well, though. Maybe try waiting longer (5 minutes ?)
22:37:07amiconnSometimes explorer must be poked a bit to update its list(s) (by pressing F5)
22:37:40amiconnHappens from time to time, even with hardware usb-ata bridges
22:40:01*amiconn had even more nasty effects happen in explorer in a few cases
22:40:25*gevaerts remember reading about the sinking of MS explorer a few months ago
22:41:39 Join bruges_ [0] (
22:47:26amiconnSometimes explorer refuses to display a certain drive, even though it mounted properly. Entering the drive letter into the address bar opens it. The drive icon then often has a red ! next to it, but accessing it works...
22:47:33 Join JdGordon [0] (
22:48:08amiconnSometimes explorer claims that the drive wouldn't support long filenames when trying to write something, even though it shows long filenames which are already present (!)
22:48:16*XavierGr could never guess that the USB protocol is so error prone and reacts so different and unexpectedly on various OSes
22:50:41*n1s would blame buggy drivers on the Oses rather than the protocol...
22:50:56*gevaerts blames everything
22:51:25 Quit ojb ("Don't follow me")
22:52:37 Quit OlivierBorowski_ (Read error: 104 (Connection reset by peer))
22:55:57n1sIt's actually quite interesting that a buggy usb device can BSOD windows, crash OSX and kernel oops linux...
22:56:55*gevaerts hasn't done the linux kernel oops directly yet (pulling out while the kernel is swapping to it doesn't count)
22:57:36XavierGrn1s: indeed, I am amazed that a dodgy driver can bring down all OSes
22:57:41 Quit bruges (Read error: 110 (Connection timed out))
22:58:28gevaertsShould we have a config setting that enables all known crash-inducers ?
22:58:56n1sof course!
22:59:16n1sin a plugin with some witty name
22:59:40 Join shotofadds [0] (n=51016716@gateway/web/cgi-irc/
22:59:56gevaertsthe term "plugin" is strangele appropriate
23:00:05*amiconn thinks that's not very surprising
23:00:56*shotofadds 's D2 is currently displaying the Rockbox boot logo :D
23:01:22shotofaddsit crashes immediately afterwards, of course
23:02:15 Join linuxstb [0] (
23:05:19*gevaerts has to go now
23:06:02 Quit gevaerts ("will the rockbox usb host crash when a rockbox device is plugged in ?")
23:06:44Nico_Pgevaerts is working on usb host ?
23:11:27Nico_Prasher: the explanation to the logf-breaks-scrobbling bug is stupid yet simple: there is only one possible track changed event and logf uses it when enabled
23:11:38***Saving seen data "./dancer.seen"
23:12:00n1sNico_P: sorry to keep doing this but I think I found a bug... ;) on line 245 in gwps-common.c wps_state.wps_time_countup is checked but this member is never assigned anywhere...
23:12:55Nico_Prasher: see playback.c:2635
23:13:00Nico_Pn1s: lemme look
23:13:04 Quit jgarvey ("Leaving")
23:13:37Nico_Pn1s: AFAIK I never touched that line
23:14:25Nico_Pmaybe I'm the one who made wps_time_countup disappear though
23:14:37n1sah, ok just thought you might be familiar enough with the wps stuff to tell how it' ssupposed to work...
23:14:52Nico_PI have no idea what this one does, sorry
23:15:11 Quit shotofadds ("CGI:IRC (Ping timeout)")
23:15:49 Join fehmicans [0] (n=canavar@
23:17:00Nico_Pn1s: it's initialised at gwps.c:65 but it's strange that it's never assigned
23:17:09Nico_Pmaybe it just got forgotten about
23:17:35JdGordonNico_P: yeah, a generic callback system there would be good
23:18:13Nico_PJdGordon: yes, it would be much better. the gather runtime data could also be moved to that event (said Slasheri)
23:19:35 Quit Benjamin_L ("Ex-Chat")
23:24:59 Quit faemir (Read error: 104 (Connection reset by peer))
23:27:47 Join kushal_12_27_200 [0] (n=kushal@
23:27:59 Quit seablue ("life, death, life, death")
23:28:39 Quit petur ("gonne")
23:34:42 Join shotofadds [0] (n=51016716@gateway/web/cgi-irc/
23:37:55 Quit gregzx ("ChatZilla 0.9.81 [Firefox]")
23:40:39 Quit kushal_12_27_200 ("This computer has gone to sleep")
23:41:18 Join kushal_12_27_200 [0] (n=kushal@
23:44:13 Join saratoga [0] (n=9803c6dd@gateway/web/cgi-irc/
23:48:26saratogaBagder: are you working on the GSOC2008 application?
23:51:01 Quit kushal_12_27_200 ("This computer has gone to sleep")
23:52:39scorche|shsaratoga: last i checked, he was...and was going to put out a draft on the dev list or similar
23:53:53 Quit Mathiasdm ("Yuuw!")
23:54:30 Join Kage [0] (
23:57:12linuxstbIt's already there (on the dev list)
23:57:35 Quit perrikwp (" ajax IRC Client")
23:57:44*scorche|sh is at work and cant tell :)
23:58:04 Join perrikwp [0] (i=98214a43@gateway/web/ajax/

Previous day | Next day