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

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

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

#rockbox log for 2009-09-02

00:00:23 Quit BlakeJohnson86 (Read error: 104 (Connection reset by peer))
00:01:11 Join BlakeJohnson86 [0] (
00:06:40 Quit domonoky1 (Read error: 104 (Connection reset by peer))
00:10:35 Quit dfkt ("-= SysReset 2.53=- Ph'nglui mglw'nafh Cthulhu R'lyeh wgah'nagl fhtagn.")
00:10:43 Quit kick_me ("grumble")
00:15:17 Join BdN3504 [0] (n=55b22760@gateway/web/cgi-irc/
00:15:54BdN3504which targets do NOT support the repeat a-b mode? are there any?
00:16:18mcuelenaereBdN3504: most touchscreen targets I guess, as there's no button mapping for them (unless there is in grid mode)
00:16:37JdGordon|those mappings should be added...
00:17:19BdN3504mcuelenaere: but there are no manuals for those atm right?
00:17:49mcuelenaereI don't know, there isn't one for the Onda's..
00:17:53***Saving seen data "./dancer.seen"
00:18:15mcuelenaereJdGordon|: that depends on the WPS I guess (if not in grid mode)
00:18:47JdGordon|mcuelenaere: sure, but adding the option is needed, then its up to themers
00:18:52 Join TopyMobile [0] (
00:18:56 Join dys` [0] (
00:19:47rasherJdGordon|: You didn't answer me why users should care how much RAM a theme uses
00:20:00 Quit Topy (Read error: 110 (Connection timed out))
00:20:13JdGordon|sorry, must have missed the message?
00:20:32JdGordon|umm... I would really like to have the buffer size user-customisable at some point in the future
00:20:36BdN3504so can i "un"opt that part in the appendix of the manual then? atm it's \opt{player,recorder,recorderv2fm} but that's definitely outdated now, right`?
00:20:40JdGordon|so knowing how much is needed is imoprtant
00:20:59rasherWhy would it be user-customisable?
00:21:30rasherThat seems like asking for trouble
00:21:34JdGordon|so people who only want text dont need to waste huge amounts so themes like yours can work on other people daps...
00:21:50rasherI think you've stayed with MS too long
00:22:18rasherWe're talking at most a few dozen KB. What problem are you trying to solve?
00:22:52JdGordon|on some targets its tiny, 2 or 3KB, on some it could be over 100KB wasted
00:23:25JdGordon|yes its not alot, but doing that makes everyone happy
00:23:47rasherExcept the users who have to piss about with setting a useful buffer size
00:24:11JdGordon|the default would be the same or larger than now anyway so I dont know what you're complaing about
00:24:23JdGordon|its not like the default would be enough for a 1 line wps
00:24:24rasherThen why add it in the first place?
00:24:47rasherSo the people who like to moan about 10KB wasted can get YET ANOTHER SETTING THAT IS ESSENTIALLY USELESS
00:25:02rasherWe don't need more memory managing settings. We need less.
00:25:13JdGordon|you want to calm down a bit?
00:25:35rasherI'm perfectly calm
00:26:31andrewrbrasher: useless to some may be useful to others, more settings doesn't cause too much trouble as long as they aren't necessary
00:26:52rasherThat's the thinking that leads to millions of settings that shouldn't exist in the first place
00:26:55JdGordon|how it works right now is the buffer size is the same (relative to the screen size/depth) for every target.... someone comes along and wants to add a very image intensive skin for say the e200, fine now he has to argue to add a a dump, say double the size... its not very much for him but a huge waste for EVERYONE ELSE
00:27:12rasherDefine "huge waste"
00:27:27JdGordon|totally unused memory
00:27:34andrewrbrasher: some people like to tweak settings, I don't see how that causes any problems for those who don't. (clearly you)
00:27:56rasherJdGordon|: unused memory is only a problem if it's needed for something else
00:28:12rasherandrewrb: Rockbox shouldn't have settings for which reasonable defaults exist.
00:28:38rasherClearly, we can live without a wps/theme buffer size setting, as has been demonstrated by the past 5 years
00:28:40JdGordon|sure, doubling the buffer on the 64mb video build isnt a big issue... but what about the clip? we have enough issues with playback already... or rec and its talk buffer?
00:28:42rasheror however much, I forget
00:29:04rasherJdGordon|: I find it very unlikely the Clip will have 200KB themes.
00:29:56rasherYou're solving a problem that doesn't exist. Point me to one person in the past that has voiced a reasonable complaint about unused RAM due to the wps buffer being larger than it needs to.
00:30:00rasherJust one.
00:30:13rasherAnd in the process of doing this, you're adding complexity
00:30:24 Quit dys (No route to host)
00:31:37BdN3504why is \opt{player,recorder,recorderv2fm} {, A-B} in the repeat mode section of the manual?
00:31:44andrewrbi know that I never use any themes with images, yet I can't view images while also playing music on my D2 due to lack of memory. If there was a setting which might allow me to shrink the wps buffer I would be grateful
00:32:00JdGordon|rasher: 1) we had no setting doesnt mean we didnt need one... I remember a few times people complaing about not enopugh image buffer
00:32:19JdGordon|2) you're missing th epoint on the clip... its playback buffer is 1MB or something... an extra 10KB is a big deal
00:32:29Lloreanandrewrb: There is no setting.
00:32:44Lloreanandrewrb: Shrinking the WPS buffer would not solve this anyway.
00:32:46*rasher groans
00:32:48LloreanYou need to increase the plugin buffer.
00:33:14rasherClearly, what we need is MSDOS style memory management where everyone has to micro-manage memory and make sure there's adequate space for the various bits and pieces
00:33:18*andrewrb shouldn't have stuck his head in with a lack of knowledge
00:33:38JdGordon|rasher: you say that like its something we dont already do?
00:33:49JdGordon|thats EXACTLY what we do
00:34:04rasher[00:25] <rasher> We don't need more memory managing settings. We need less.
00:34:37JdGordon|ok, so lets enable cuesheet,, timestreth, everything and see how it copes
00:35:36rasherOr let's at least not add more of this tomfoolery.
00:35:37LloreanClearly you should just drop the WPS buffer entirely, and use the existing playback buffering to "buffer" the WPS using only as much memory as it needs.
00:35:43JdGordon|we have a very small amount of RAM to play with and a HUGE feature set... without toggle/sizing settings there is either feature freeze (in which case i'm out of here), or huge bloat for those which dont use those features
00:36:09rasherVery small amount of RAM? Except the clip, everything has more RAM that we could ever use
00:36:41JdGordon|you must be new here.... we dont cater to the highest speced DAP.. we cater to the lowest
00:37:08 Part rasher
00:37:12JdGordon|I've been saying for years keeping an eye on the RAM usage is stupid...
00:37:56LloreanThere's a compromise between "not keeping an eye on it" and "letting it limit everything" called "trying to make sure the amount we use is reasonable relative to what we gain"
00:38:21 Quit antil33t (Read error: 60 (Operation timed out))
00:38:32JdGordon|and I've been happy with the compromise which started this argument again....
00:40:59 Join antil33t [0] (n=Mudkips@
00:41:03LloreanCould we have a minimum WPS buffer size that can be increased *if* playback is stopped when the WPS is loaded?
00:41:34JdGordon|in theory... sort of...
00:41:50LloreanSo we could normally allocate less memory, in a value good enough for say, 80%+ of themes to work fine without playback stops, but allow people their giant themes if they really wanted it?
00:41:50 Quit BdN3504 ("CGI:IRC (EOF)")
00:42:01LloreanNo settings involved.
00:42:12JdGordon|in *theory*...
00:42:29 Join PaulJam [0] (
00:42:31JdGordon|in practice its way to hard to do
00:42:32 Quit ender` (" If Klingons had invented Usenet, killfiles really would...")
00:43:25JdGordon|its a general memory issue with rockbox... restarting playback could be used to do it but you waste lots more ram each time
00:43:29JdGordon|which could be acceptable
00:43:57*JdGordon| isnt even sure where this "restart playback to enable" idea came from
00:44:12JdGordon|i'm pretty sure its never handled doing more allocs after playbacks tarts the first time
00:48:39*Llorean shrugs
00:56:57 Quit NuNuRs ()
00:58:17 Join descendent87 [0] (n=benji@
00:59:29 Quit PaulJam (".")
01:08:48 Join goffa_ [0] (n=goffa@
01:13:28 Quit robin0800 (Remote closed the connection)
01:14:22descendent87is there any progress on rockbox for creative zen vision m?
01:14:40mcuelenaeredescendent87: the project has more or less stalled
01:15:17descendent87bummer, looked on the site and didnt see any updates on it so thought it might have
01:15:41*JdGordon| decides to beat the horse once more.... figuring out the correct skin buffer size is going to be verry difficult when statusbar, fm, and wps all use the same one.. and then there is the remote to consider also
01:16:06JdGordon|maybe we just hard code it to 1MB and call it quits?
01:20:52 Quit goffa__ (Read error: 110 (Connection timed out))
01:22:37 Quit JdGordon| ("Miranda IM! Smaller, Faster, Easier.")
01:24:38 Quit soap (Read error: 110 (Connection timed out))
01:26:19Unhelpfulrasher: i've been thinking about various ways to not have to fix buffer sizes at compile time or at startup time. i don't like many of them very much. prefixing buffer_alloc buffers with a pointer to a callback that can safely move them is probably the best scheme in my opinion.
01:28:27 Quit CaptainKwel ("Page closed")
01:32:54 Part descendent87 ("Konversation terminated!")
01:34:09 Join kugel [0] (n=kugel@rockbox/developer/kugel)
01:35:19 Quit Thundercloud (Remote closed the connection)
01:37:07kugelwhat about a buffer alloc which gives you say 50k per call, it would get those by stopping playback, but would garuantee to append/prepend the new 50k to the buffer from existing calls. Within in that buffer, stuff like skin engine/dsp/whatever could freely use a real malloc
01:37:29 Join JdGordon_ [0] (i=483efeb0@gateway/web/freenode/x-ttxqsofebwsikjgu)
01:38:08 Join FOAD_ [0] (
01:38:40kugelthe "could freely use a real malloc" is not even needed I guess
01:39:36kugelthe point is, to have a single buffer which can grow, but is still always a single buffer
01:40:58 Quit Zagor ("Clint excited")
01:42:19Hillshumfroggyman: About brickmania, I think adding more options for gameplay style would be nice. Like an '
01:42:51Hillshumendless' mode that keeps on going, and the like
01:43:08 Join rasher [50] (n=rasher@rockbox/developer/rasher)
01:44:44 Quit bubsy (Read error: 54 (Connection reset by peer))
01:46:29JdGordon_I like the idea of forcing anything that does allocation after a set "time" to handle a callback which means it has to realloc
01:46:57JdGordon_but that still wont fix the problem.... there is no way to know the required skin usage untill its finished loading
01:47:18JdGordon_^ kugel, Unhelpful
01:47:40kugelhence the idea of letting it grow
01:48:10JdGordon_letting it grow doesnt work with how the skin buffer allocation works
01:49:00JdGordon_it would have to be VERY smart t handle that
01:49:21kugelwell, you could approximate the skin usage by investigating the images first (reading headers, calculating size, add it up), the rest of the skin would unlikely take more than 5% additionally
01:49:42kugelJdGordon_: because you made it complicated
01:50:26kugelthat can still be changed, paying 4bytes per token for using the ram more effeciently in the end
01:51:21JdGordon_you understand its a hell of a lot of work needed to make that change right?
01:51:22Hillshumfroggyman: The way brickmania is currently set up it becomes a quest for the most points, with no risk of not finishing
01:51:55kugelno, not really, and I think it could be worth it in the end
01:52:15JdGordon_no you dont understand its alot of work? or you dont belive me it is?
01:52:30kugeli don't understand
01:52:33 Join bubsy [0] (n=Bubsy@
01:52:44 Join bubsy_ [0] (n=Bubsy@
01:52:45JdGordon_then trust me... it is :)
01:53:43kugelI think it could be worth it
01:54:13kugelit's not just about the skin buffer, all the other stuff too (timestrech, cuesheet, etc)
01:55:18 Quit FOAD (Read error: 110 (Connection timed out))
01:55:18 Nick FOAD_ is now known as FOAD (
01:55:45kugelbut my idea probably won't work as long as each user of that buffer wants a single buffer. making it grow would still mean to potentially split the skin buffer in 2 parts for example
01:55:46 Quit JdGordon_ (Ping timeout: 180 seconds)
01:55:56 Join JdGordon_ [0] (i=483efeb0@gateway/web/freenode/x-xrpevcmbpgbpiuut)
01:56:01JdGordon_this webchat sucks
01:56:16JdGordon_those are all constant sizes so resizing/gorwing isnt needed
02:01:57 Join fdinel [0] (
02:02:00 Join TheCoolGman [0] (n=446e5e7d@gateway/web/cgi-irc/
02:02:24 Join chandoo [0] (
02:03:30TheCoolGmanHi, I'd Like to bring up patch FS #10554 Rockpaint: enable to set canvas size and FS #10555 Rockpaint: improve browsing font. The Font browsing still neds some work but the canvas size works well
02:04:01 Quit JdGordon_ (Ping timeout: 180 seconds)
02:08:56 Quit antil33t (Read error: 113 (No route to host))
02:17:55***Saving seen data "./dancer.seen"
02:18:15 Quit chandoo_ (Read error: 110 (Connection timed out))
02:20:09 Join Strife89 [0] (
02:22:49 Join MG_Man [0] (
02:22:51 Quit MG_Man (Client Quit)
02:23:36CIA-43New commit by kugel (r22597): Forgot to add keys for the quickscreens' top item for the ipods (they didn't have ACTION_QS_DOWNINV before). As that one was used as exit button, ...
02:24:51 Join MG_Man [0] (
02:28:46 Quit Hillshum ("Ex-Chat")
02:28:55 Quit mcuelenaere ()
02:32:20 Quit DataGhost (Nick collision from services.)
02:32:28 Join DataGhost [0] (i=dataghos@unaffiliated/dataghost)
02:34:04 Join Topy44 [0] (
02:37:04JdGordonit stands to reason that if something is breaking a pointer (or not initialising it correctly) that memset()ing it to 0 first shouldnt fix the problem but make it more obvious yes?
02:37:35 Nick dys` is now known as dys (
02:38:44JdGordonon the other hand... doing that would set the bad pointer to NULL so hiding the real issue :/
02:40:00JdGordonkugel: can you test a patch?
02:40:12JdGordonStrife89: you volanteered to test yesterday yeah?
02:41:36JdGordon thanks
02:41:44Strife89JdGordon: Yep. :)
02:41:49kugelwhat did you mean with the pointer stuff?
02:42:03UnhelpfulJdGordon: perhaps if it did as the album art loader does, and grabbed the remaining buffer space to handle the load, and determined the final size afterward? then it could buffer_alloc the new skin buffer, copy the loaded data there, and free the buffer handle...
02:42:35JdGordonthere is an uninitialised pointer somehwere (I dont know where) and because its not being set to NULL something thinks its legit so follows it when it shuldnt
02:43:26Unhelpfulthe AA loader can only do that readily because it's inside bufopen, though. :/
02:43:27JdGordonUnhelpful: nope... pointers everywhere...
02:43:35Strife89JdGordon: Give me a minute, I need to get on my laptop.
02:43:52JdGordontry loading different themes and make sure they all display correctlty
02:43:56JdGordonI tihnk its all good now
02:44:54kugelit doesn't apply to my statusbar branch. no luck for you :)
02:45:00*Strife89 wonders if FS #8806 is ready for WPS comission.
02:45:05 Join tvelocity [0] (
02:45:06UnhelpfulJdGordon: well, if there's a "move buffer" callback that you'd have anyway in case something force the already allocated skin buffer to be moved, surely the same callback could be used to move the data from a buffer handle to the new buffer_alloc'd location?
02:47:18 Quit Strife89 ("Switching machines.")
02:48:35 Join Strife89 [0] (
02:49:03JdGordonit could be done, but everything would need to be changed to offsets which is something I really dont want to do...
02:49:32JdGordoneven a boring text only wps will have a absolute heap of pointers which would need reworking
02:50:43kugelis ther a skin with sublines ?
02:51:13kugeldon't want to make one
02:51:24JdGordonyou mean a shipped one?
02:51:39kugelor one from the theme site
02:51:39JdGordonmost of them no?
02:52:03Strife89JdGordon: Hang on and I'll patch and compile.
02:52:15kugelmost of them? I can't remember one
02:52:39Strife89JdGordon: Do you know if anything will break if I make a copy of my rockbox folder and rename it?
02:52:45JdGordondancepuffduo definetly does
02:52:53JdGordonStrife89: no, it will be fine
02:53:13JdGordonI was hoping for coverage on themes I hadnt tested though :p
02:53:15Strife89JdGordon: Okay, just wanted to check.
02:58:16Strife89JdGordon: Okay, making. One Sansa build, and then the iPod Photo build, coming up.
02:59:49kkurbjunamiconn: good call on updating the screen backwards, for some reason it didn't occur to me. Originally I was stuck on the 32x32 pixel squares because I was trying to do the transform in hardware, but that didn't work out so well. The fps drops to 164 fps with the ARM updating the registers 2x as often, but that's not too bad. Overall the responsiveness is much better so I think the new lcd format worked out well for it.
03:01:35CIA-43New commit by kkurbjun (r22598): Update Vertical stride so that it's oriented left to right in the destination.
03:02:44*Strife89 backs up his .rockbox folders.
03:04:03kugelJdGordon: that's why I was asking
03:05:05Strife89JdGordon: About halfway through the iPod build now, I think.
03:05:13kugelkkurbjun: that's what I meant with the framebuffer trickery :)
03:05:55kkurbjun:-D, It went over my head at the time
03:06:32kugelStrife89: announcing every step is really unneeded
03:07:44Strife89kugel: Sorry. ._.
03:10:25kugelJdGordon: seems to work here, it has 4x %t3 in it
03:10:35kugelwith conditionals and stuff
03:10:48kugelalthough, the 3s don't seem to be very accurate
03:11:03JdGordonhow not accurate?
03:11:06JdGordontoo fast or slow?
03:11:38Strife89JdGordon: Anyway, installing now.
03:11:55*JdGordon echos kugel's request from about 4 min ago...
03:14:25Strife89JdGordon: So what does this do, exactly?
03:14:41Strife89Forgive my incompetence. ._.
03:14:55JdGordonit does nothing... hopefully...
03:15:56 Quit kkurbjunW (Remote closed the connection)
03:16:10kugelJdGordon: to slow, it's more like 3.5s
03:16:47kugelno, wait, faster. ~2.5s
03:16:55 Join kkurbjunW [0] (n=karlk@
03:17:58Strife89JdGordon: At least one theme raised the volume on me when I loaded them.
03:18:20Strife89On the c250.
03:18:26JdGordonkugel: split the difference and its all good :)
03:18:29JdGordonStrife89: REALLY?!
03:18:38Strife89JdGordon: No joke.
03:18:53JdGordoni dont belive you!
03:18:57Strife89It jumped from -32dB to -13dB.
03:19:13kugeldoes it by any chance contain a volume: X line?
03:19:20*Strife89 checks.
03:19:47UnhelpfulJdGordon: i managed to get it to do that when i was hacking on convttf and it was generating buggy fonts. it would jump from -25dB to +6 or whatever the beast maximum is.
03:19:50Strife89JdGordon: False alarm.
03:20:14JdGordonUnhelpful: crap, thats really wierd...
03:20:18 Quit MG_Man (Read error: 110 (Connection timed out))
03:20:18JdGordonStrife89: told you so :)
03:20:38Strife89Why someone would make a theme jump to a certain volume is beyond me, but hey.
03:20:53Unhelpfuland i'm pretty sure it was not a false alarm in my case, as the difference in volume was quite, um, obvious.
03:21:08Unhelpfuli wasn't allowed to use the beast in the family car until i fixed it.
03:21:15JdGordonStrife89: is that one from the themes site?
03:21:28Strife89JdGordon: It's called Rockblack.
03:21:36JdGordonUnhelpful: you sure it wasnt a fixed.cfg or config issue?
03:22:00JdGordonI mean sure its possible to get the volume to have a wierd value, but I'd expect a crash or two to go along with it
03:22:11Strife89JdGordon: Anyway, the c250 checks out okay. :)
03:22:49UnhelpfulJdGordon: i'm pretty sure it started suddenly after some revision in convttf, and disappeared, along with other weird problems, after i fixed that and a glyph size calculation patch.
03:25:37kugelJdGordon: scrolling seems bugged on dancing puff duo
03:27:27JdGordonkugel: which target?
03:27:59kugelthis is a normal line with %s, it begins to scroll but resets after 0.5-1 seconds
03:28:16JdGordonwhich line? or whats on the line?
03:28:50kugelsame problem on blacknblue glass also
03:29:38*Strife89 gives JdGordon's patch a thumbs up.
03:30:43kugelJdGordon: the line after the "Next Song:" one
03:31:16JdGordonkugel: it resets, but after that it scrolls forever correctly yeah?
03:31:26kugelno it keeps doing that
03:32:10kugelit looks like the usual "next track info ready" reset, but it doesn't stop
03:33:44CIA-43New commit by kkurbjun (r22599): M:Robe 500: Add low-level support for vertical strides
03:35:48*JdGordon checks the diff
03:36:00 Join CaptainKwel [0] (
03:38:50 Quit TopyMobile (Read error: 110 (Connection timed out))
03:38:57 Join antil33t [0] (n=Mudkips@
03:39:32 Join TopyMobile [0] (
03:39:37JdGordonkugel: ok, I have a fix... you want the full diff again? or just the few changed lines?
03:40:04JdGordon is just the func to replace
03:40:08JdGordonin skin_display.c
03:41:31JdGordonseems to be good here
03:42:30CIA-43New commit by kkurbjun (r22600): M:Robe 500: fix warning
03:44:02kugelJdGordon: just post-poned the reset? by 100*HZ?
03:44:13JdGordonthats what svn does
03:44:24 Quit panni_ ("( :: NoNameScript 3.81 :: )")
03:45:22kugelhm, sounds strange, but well
03:45:36JdGordonumm.. I'm not entirely sure how that is going to handle next track when it becomes avilable
03:45:41*JdGordon checks
03:46:22JdGordonyeah, should be fine
03:48:39JdGordonI think 100s is a wierd value, but it works
03:48:55JdGordonI wonder if it should be something like 3min or so?
03:49:28kugelwhy's there this value at all?
03:49:56JdGordonbecause the displayer is stupid
03:50:23JdGordonif the line hasnt been updated it shouldnt be redrawn
03:50:31JdGordonthat a bug... but a seperate issue
03:50:38kugela line should just reset scrolling when it's swapped with another subline, or its information changes
03:50:41kugelyea, probably
03:51:27JdGordonthats for later though... I have an idea how to do it...
03:51:56JdGordonbut really.. the lines tokens need to be checked to make sure none are dynamic, or it wont work
03:52:11 Quit antil33t ()
04:05:39 Quit TheSeven (Nick collision from services.)
04:05:56 Join The_Seven [0] (
04:06:02 Nick The_Seven is now known as TheSeven (
04:06:44 Quit froggyman ("ChatZilla 0.9.85 [Firefox 3.5.2/20090729225027]")
04:13:56 Join dys` [0] (n=andreas@
04:17:41 Quit kugel (Remote closed the connection)
04:17:56***Saving seen data "./dancer.seen"
04:22:17 Quit chandoo ("Leaving")
04:25:58 Quit dys (Connection timed out)
04:27:07 Quit DataGhost (Nick collision from services.)
04:27:14 Join DataGhost [0] (i=dataghos@unaffiliated/dataghost)
04:31:51 Quit andrewrb ("ChatZilla 0.9.85 [Firefox 3.5.2/20090729225027]")
04:32:11 Quit bubsy_ ("I'll be back somewhere in time...")
04:33:45 Join antil33t [0] (n=Mudkips@
04:39:35 Quit Strife89 ("Bed.")
04:41:35CIA-43New commit by kkurbjun (r22601): Fire: Add support for 256 color hardware pallete mode
04:48:00 Quit Rondom (Nick collision from services.)
04:48:10 Join Rondom [0] (n=Rondom@
04:55:36CIA-43New commit by jdgordon (r22602): Almost the last of the skin ram wastage fixing... This one moved the line/subline handling into the alloced buffer and links them more sensibly with ...
05:00:39 Quit Sajber^ (Read error: 104 (Connection reset by peer))
05:02:03CIA-43New commit by jdgordon (r22603): fix player's red
05:04:19 Join BHSPitMonkey [0] (n=stephen@unaffiliated/bhspitmonkey)
05:06:56 Quit TheCoolGman ("CGI:IRC")
05:09:13 Join TopyMobile_ [0] (
05:09:32 Quit JdGordon ("Leaving.")
05:09:59 Quit TopyMobile (Read error: 104 (Connection reset by peer))
05:10:44kkurbjunUnhelpful: do you have a pointer on where to look to add support for vertical strides for pictureflow?
05:11:53 Quit TopyMobile_ (Remote closed the connection)
05:27:18 Join ecognito [0] (n=ecognito@
05:28:59ecognitoAnybody know good stores in Melbourne, Australia for buying rockbox friendly MP3 players? Or is ebay my best bet?
05:32:52CIA-43New commit by kkurbjun (r22604): Pictureflow: Add support for vertical strides.
05:42:06 Quit fdinel ("Miranda IM! Smaller, Faster, Easier.")
05:44:06CIA-43New commit by kkurbjun (r22605): Fix collision detection for invadrox
05:44:40 Quit gevaerts (Nick collision from services.)
05:44:51 Join gevaerts [0] (n=fg@rockbox/developer/gevaerts)
05:46:52 Join n1s [0] (n=n1s@rockbox/developer/n1s)
05:52:22 Quit Horscht ("Verlassend")
06:15:24 Quit intrados_ (Read error: 104 (Connection reset by peer))
06:16:35 Join intrados_ [0] (
06:17:58***Saving seen data "./dancer.seen"
06:27:18 Join BHSPitLappy [0] (n=BHSPitLa@unaffiliated/bhspitmonkey)
06:27:31 Quit sbhsu (Read error: 60 (Operation timed out))
06:31:28 Quit BHSPitMonkey (Read error: 104 (Connection reset by peer))
06:31:28 Join BHSPitMonkey_ [0] (
06:32:33 Join sbhsu [0] (
06:38:17 Quit CaptainKwel (Remote closed the connection)
06:44:03 Join LinusN [0] (n=linus@rockbox/developer/LinusN)
06:50:02 Join sbhsu_ [0] (
06:59:17 Quit sbhsu (Read error: 110 (Connection timed out))
07:07:24 Join bmbl [0] (n=Miranda@unaffiliated/bmbl)
07:07:49 Join sbhsu [0] (
07:14:23 Quit kkurbjunW (Remote closed the connection)
07:14:45 Join kkurbjunW [0] (n=karlk@
07:17:03 Quit sbhsu_ (Read error: 110 (Connection timed out))
07:31:42 Join FOAD_ [0] (
07:35:40 Nick BHSPitMonkey_ is now known as BHSPitMonkey (
07:40:07 Quit TheSeven (Nick collision from services.)
07:40:25 Join The_Seven [0] (
07:40:31 Nick The_Seven is now known as TheSeven (
07:44:41 Quit bmbl ("Bye!")
07:48:58 Quit FOAD (Read error: 110 (Connection timed out))
07:48:58 Nick FOAD_ is now known as FOAD (
07:58:35 Quit moos ("Rockbox rules the DAP world")
08:08:56 Join JdGordon [0] (n=jonno@rockbox/developer/JdGordon)
08:18:02***Saving seen data "./dancer.seen"
08:18:49 Join ender` [0] (
08:23:06CIA-43New commit by jdgordon (r22606): rename wps_[sub]line to skin_[sub]line
08:32:23 Part safetydan ("Leaving.")
08:34:57 Join flydutch [0] (
08:36:14 Join Rob2223 [0] (
08:37:42 Join pamaury [0] (n=pamaury@
08:38:35*GodEater yawns
08:39:10JdGordongood evening :)
08:39:37pamaurypcc1: hello
08:40:11*GodEater stumbles towards the kitchen to get coffee going
08:42:54 Join bertrik [0] (
08:42:58 Quit ender` (Read error: 104 (Connection reset by peer))
08:45:36 Join ender` [0] (
08:48:06 Join decayedcell [0] (
08:51:15gibbon_good morning everybody
08:51:34GodEaterwhat's so good about it?
08:52:01gibbon_that would be _very_ offtpoic...
08:54:31 Quit Rob2222 (Read error: 110 (Connection timed out))
08:56:06pcc1pamaury: hello
08:56:47pcc1pamaury: I saw that you pushed your MTP stuff to the github repository
08:56:57 Join AndyIL [0] (n=pasha_in@
08:57:27pcc1I was thinking what the next step would be
08:58:14pamauryIf you want to code MTP, I can advise you to get some documents before coding.
08:59:50pcc1yes, I was reading some parts of the MTP spec and comparing to your work earlier today
09:00:01pamauryYou'll need the following documents:
09:00:11pamaury- MTP spec (USB-IF), you probably have that
09:00:46pamaury- USB Still Image Capture Device Definition (USB-IF): that's the USB specific part of PTP (MTP is an extension of PTP)
09:01:15pamaury- PIMA15740: the PTP spec, not so easy to find, could be useful
09:01:27pamauryI can advise you to get the libmtp source code.
09:03:55pamauryThere are still some thing in the code that I'm not satisfied with. For example, the handling of unrecognized commands on the Control Endpoint: should it stall endpoints and wait for the reset command ? I'll need to read it once more to be sure. Furthermore, the transfer code for writing (device point of view) will not work for big transfers and reading is not implemented yet
09:07:38 Join funman [0] (n=fun@rockbox/developer/funman)
09:08:30 Quit AndyI (Read error: 110 (Connection timed out))
09:12:29 Quit BHSPitLappy (Remote closed the connection)
09:12:42pcc1pamaury: ok I'll look into those
09:13:48pcc1are you sure you committed all your code? for example I didn't see usb_mtp.c added to firmware/SOURCES
09:14:29pamauryReally ? I'm sure it was modified on my svn repo. I check that immediately
09:14:41pamauryIf you don't have usb knowledge or don't want to get hands too dirty with that, you'll noticed I tried to abstract the protocol with the *pack_data* functions and send_*/recv_*
09:15:28pamauryThere are also lots of gots to be answered about the ways objects are managed
09:16:56pcc1yes. I am concerned we may need a large object->path mapping table
09:17:10pamauryIf I'm correct, the following code is in SOURCES:
09:17:11pamaury#ifdef USB_ENABLE_MTP
09:17:11DBUGSent KICK pamaury to server
09:17:11Kick(#rockbox pamaury :No flooding!) by logbot_!n=bjst@gateway/web/cgi-irc/
09:17:26 Join pamaury [0] (n=pamaury@
09:17:38pamauryHey, why the bot kicked me ?
09:17:59funmanIt is a bit sensitive about flooding, perhaps you typed too fast :/
09:18:01pcc1I guess because you pasted three lines
09:18:19pcc1pamaury: I don't have those lines
09:18:32pcc1your commit only added two files: usb_mtp.[ch]
09:18:32pamauryHum, perhaps I forgot too commit them
09:20:30pamauryI modified several other files, I tought I had commited those changes. what command I should type to do so ? (perhaps that's the reason)
09:21:01 Join einhirn [0] (
09:21:16 Part ecognito ("I read it on the Internet so it MUST be true!")
09:21:24 Join Zagor [242] (n=bjorn@rockbox/developer/Zagor)
09:21:29pcc1you need to provide the -a flag to git commit to commit changes to existing files
09:21:49pcc1alternatively you can use "git add" to add those files to your pending commit
09:21:54pamaurydoes it adds unversioned files also ?
09:22:08pcc1no, you still need to use git add for that
09:22:27funmanyou can check unversioned files with 'git status'
09:23:58pamauryThat's done
09:25:12pcc1got it
09:25:14pamauryConcerning your object->path table mapping, for now it's useless with tagcache. But the problem is that tagcache is limited to music and is not hierarchical. It's a choice I made for the first MTP code because it's simpler
09:25:30pamaury(the problem is tagcache is that's it is slow)
09:27:46pcc1yes, we may need to move away from tagcache
09:28:05GodEaterpamaury: are you mixing spaces and tabs in your code? I'm looking at the commit and the alignment is a bit wacky in places
09:28:21pamauryNormally not, only spaces.
09:28:35pamauryI check that. you mean in usb_mtp.[ch] ?
09:28:52GodEaterusb.c at the moment
09:28:57GodEaterwhere you've modified it
09:29:36pamauryAh, perhaps
09:30:22pamauryMy editor says there are no tabs in usb.c
09:30:29 Join GeekShadow [0] (n=Antoine@reactos/tester/GeekShadow)
09:30:44GodEatermaybe it's just github playing silly buggers then
09:30:56GodEaterI'm just using it's source browser, I've not actually checked your commi tout
09:31:02pamauryI'm looking on github
09:31:07GodEaters/ t/t
09:31:44pamauryWhere do you see strange things ? which line(s) ?
09:32:15pamauryI modified the usb.c code because it didn't allowed drivers to get exclusive disk access
09:32:34pamaury310 is an empty line no ?
09:32:39 Join Thundercloud [0] (
09:32:48GodEaternot according to github :)
09:32:54GodEaterit has "usb_core_enable_driver(USB_DRIVER_CHARGING_ONLY, false);" on it
09:33:06pamauryoops no storry, looking the wrong version ;)
09:33:22GodEaterand has some odd extra padding at the beginning
09:34:16pamauryI don't see the problem with this line, it's correctly indented to me
09:35:05GodEateras I say, it may well be git hub's rendering of it
09:35:18pamauryEven on github, I don't see the problem :)
09:35:45*GodEater prepares a screenshot
09:35:59funmanI see a new level of indentation but it's because of the added "else {"
09:36:34GodEatermaybe that's it
09:36:43*GodEater throws away the screenshot
09:38:56 Quit bertrik (Read error: 113 (No route to host))
09:39:22*GodEater looks up how to track a remote branch
09:45:33 Quit BHSPitMonkey ("Ex-Chat")
09:47:42 Quit Thundercloud (Remote closed the connection)
09:56:54 Join l403 [0] (n=l@
09:57:37 Join xavieran [0] (
09:57:49funman;filename=fuzev2_bottom.JPG <<< the 1st lines written on the chip is identical for Clip+ and Fuzev2 so perhaps they use the same AMS SoC ..
10:03:04 Join polobricolo_ [0] (
10:05:44 Nick polobricolo_ is now known as polobricolo (
10:06:08 Join robin0800 [0] (
10:16:56 Join moos [0] (i=mostafa@rockbox/staff/moos)
10:18:05***Saving seen data "./dancer.seen"
10:19:49 Quit pamaury ("exit(*(int *)0 / 0);")
10:30:16 Quit robin0800 (Remote closed the connection)
10:49:41 Quit moos (Read error: 145 (Connection timed out))
11:12:20 Quit elcan (Remote closed the connection)
11:12:54 Join elcan [0] (
11:14:26 Quit kkurbjunW (Remote closed the connection)
11:14:57 Join kkurbjunW [0] (n=karlk@
11:15:56gevaertsany theme site people around? I think the c200 Rockblack(2) theme should be removed/hidden. It sets the volume
11:16:00 Nick fxb__ is now known as fxb (
11:16:44gevaertsand other things actually...
11:18:08funmanI have a battery_bench.txt of YH920 (10hours runtime, battery bench stopped at 93%)
11:18:53funmanreading from ADC works, but I need to check how the value read is converted to volts (I'm not sure if it's standard on all PP devices)
11:19:52 Quit elcan (Client Quit)
11:20:04 Join elcan [0] (
11:27:40funmanI hate how PP register locations can be assigned to a register with only mov
11:28:01ZagorFYI: will be upgraded tomorrow morning european time. downtime approx 08:00 - 09:00 CET.
11:30:10 Join pamaury [0] (n=pamaury@
11:34:18 Nick fxb is now known as fxb__ (
11:35:51funmando you know how PP OFs fill IRAM ?
11:43:02funmanfunctions with variadic arguments use the registers for the variable arguments, not only the stack as I thought
11:44:42 Join robin0800 [0] (
11:47:27funmani have found a string in my yh920 PP bootloader which could print battery percentage and voltage, but I am not sure how to run the YH-920 into test mode
11:56:43 Join GeekShado_ [0] (
11:58:39 Quit GeekShadow (Read error: 104 (Connection reset by peer))
12:03:30 Quit ej0rge (Read error: 110 (Connection timed out))
12:03:51 Join bmbl [0] (n=Miranda@unaffiliated/bmbl)
12:13:24funmandamned, for some reason my YH920 enters USB mode even if usb is unplugged
12:13:57funman"charger present"
12:18:07***Saving seen data "./dancer.seen"
12:18:19gevaertsAlexP: ping
12:19:48gevaerts(or rasher, or scorche|sh, or maybe someone else. Are there other theme site admins?)
12:32:20 Quit n17ikh|Server (Read error: 104 (Connection reset by peer))
12:32:22 Join n17ikh| [0] (
12:32:30funmanthe battery level (from 0 to 1.0) seems to be (670 - adc_read()) / 110
12:32:47 Join dfkt [0] (i=dfkt@unaffiliated/dfkt)
12:33:15funmanor rather, (adc_read() - 670) / 110
12:34:18funmanso adc_read would range from 670 to 780, i need to get a DMM to measure (and calculate) the voltage
12:38:06funmanbut my battery_bench measurements show the adc_read() went from 1023 (all 10 bits set) to 916 before shutting down
12:38:30funmanPerhaps the OF bootloader code for battery test is wrong though
12:41:16funmancould have not been updated for the YH920 battery
12:41:43gevaertsAll but one of Quinlan Moll's c200 themes should be removed (11 of them). He doesn't seem to grasp the concept of only including theme-related settings in the .cfg file, so they contain any setting that he happened to be using at the time
12:43:31gevaertsonly the Zune theme looks ok to me
12:45:34AlexPgevaerts: yo
12:46:09gevaertsAlexP: can you remove eleven themes from the theme site?
12:46:23AlexPcan do, what's up?
12:46:33gevaertssee three lines up :)
12:46:36AlexPah, I see
12:46:52AlexPI wonder if hiding or removing is better
12:47:07AlexPOr getting in touch first
12:47:13AlexPOr fixing them :)
12:48:12gevaertsIt think they should be made unavailable ASAP. Most of them set things like "show files", and all of them set the volume
12:48:41gevaertsalso tagcache settings
12:49:23AlexPOK, I'll hide them all - that should send him an e-mail (or 11 rather) saying get in touch with me when he is ready to re-upload them
12:49:51gevaertshave fun, and start clamoring for a mass-admin feature :)
12:50:40gevaertshm, actually, my count may be off
12:51:12 Quit krazykit (Remote closed the connection)
12:52:56AlexPgevaerts: All except zune right?
12:53:29gevaertsone moment, I'm checking again
12:55:54gevaertszune and windowsmediacab (that one doesn't actually include a .cfg)
12:56:39AlexPOK, all the others I'll hide
12:57:17gevaertsI'm not sure if we want themes without .cfg, but at least they're not actively harmful
12:57:59AlexPgevaerts: Should be gone (hidden) now
12:58:52funmancan you give me advice on the price of a DMM I would use with rockbox targets?
12:59:02AlexPgevaerts: I told him to contact me (here or in the forums) when he has fixed them so I can remove them and he can re-add
12:59:14gevaertshm, he also uploaded ipod video themes. I guess we need a full audit
12:59:21AlexPgevaerts: I only hid them in case he doesn't reply, then I'll get round to fixing them eventually
12:59:32AlexPgevaerts: :/
13:00:20AlexP"Skylar G" doesn't sound like a proper name either, and he has a load of ipod video themes
13:00:46 Quit intrados_ (Read error: 110 (Connection timed out))
13:01:07gevaertsthe ipod video "Pod" theme is also buggy
13:01:49gevaertsand the 0800_Text c200 theme comes in a file named Not technically against the rules, but I can't say I like it
13:03:46AlexPgevaerts: I've hidden the pod one
13:05:57gevaertsit would be good if someone with easy access to the files (i.e. shell, or a bigtar file to download) would check all themes for weirdness I think
13:09:32Unhelpfulkkurbjun: look in render_slide. you should actually see a nice boost from this, since pictureflow renders by columns anyway. changing the calculation of pixel and pixelstep should do it.
13:10:01AlexPgevaerts: I guess that'd be rasher, scorche|sh or maybe mculenaere
13:10:46 Join krazykit [0] (
13:36:35 Quit l403 (Read error: 113 (No route to host))
13:52:02 Join fml [0] (n=4fd3c45a@gateway/web/cgi-irc/
13:53:05fmlIn keymap-e200.c, button_context_quickscreen, there is an entry for ACTION_NONE. What purpose does it serve? Can it be removed?
13:56:53GodEaterfml: what happens if you enter the quickscreen accidentally, and don't want to change anything ?
13:57:03GodEaterdon't you require an ACTION_NONE in that case ?
13:57:45fmlGodEater: then you press ACTION_STD_CANCEL. Other players don't have ACTION_NONE in this context.
13:57:55GodEaterah right
13:59:53fmlI think it can be safely removed (will try first)
13:59:53kkurbjunUnhelpful: thanks, I found that, I already committed the support in pictureflow
14:00:57Unhelpfulkkurbjun: ah... i see that now, i only caught the highlighted part before
14:03:45CIA-43New commit by vitja (r22607): i7: Notify the backlight driver when the HOLD button is toggled, same as D2
14:03:55amiconnACTION_NONE lines are used to suppress combos which would otherwise trigger actions in a chained context
14:04:18CIA-43New commit by kkurbjun (r22608): Pictureflow: Fix define
14:07:16fmlAh, it seems the ACTION_NONE entry in e200's quickscreen does serve a purpose. It masks the cancel action in the standard context. I don't fully understand the mechanics but without that entry, pressing LEFT quits the screen.
14:07:47fmlamiconn: where is the "standard" action defined?
14:08:06kkurbjunamiconn: another issue came up with this vertical stride stuff. This appears to mainly be a problem with the picture.c stuff in the plugin library, but it could also be a problem with the icons. Right now the stride macro makes no attempt to determine what screen it is being used on. This could be an issue for the remote since it's stride is different than the main display...
14:09:02Unhelpfulfml: a lot of times things like this happen because _STD defines an action on button press, and some other context wants to differentiate between press-and-release and press-and-hold, but wants to include other items from that context
14:09:31Unhelpfuland pretty much all of that should be in apps/keymaps :)
14:09:49kkurbjunI think the stride problem will exist for anything that uses the screens api
14:10:14fmlUnhelpful: Ok, now I understand. Thanks!
14:14:20fmlHave the USB capable bootloaders for sansa e200 been officially released?
14:18:10***Saving seen data "./dancer.seen"
14:23:39 Quit polobricolo (Read error: 110 (Connection timed out))
14:31:04 Join polobricolo [0] (
14:34:09 Join moos [0] (i=mostafa@rockbox/staff/moos)
14:34:51 Quit funman ("++")
14:35:47 Join LambdaCalculus37 [0] (i=44a0430d@rockbox/staff/LambdaCalculus37)
14:36:21kkurbjunamiconn, I'm thinking of ways to go about fixing this, at the moment, I am thinking that if there was an extra parameter in the screens api that scored either screen number or screen type then the stride macro could use that as a parameter which determines whether to return the remote stride or the main display stride. Everywhere else that used the stride macro could call it with a define that was something like LCD_MAIN or LCD_REMOTE if t
14:36:35kkurbjunI'm not sure if that's the best way to go about fixing it though
14:49:46 Quit fml ("CGI:IRC (EOF)")
14:52:59 Join petur [50] (n=petur@rockbox/developer/petur)
14:53:22 Quit SUSaiyan (Read error: 110 (Connection timed out))
14:53:30 Join kugel [0] (n=kugel@rockbox/developer/kugel)
14:54:06kugelkkurbjun: fortunately the multiscreen api has that
14:54:19kkurbjunkugel what is that?
14:54:47kugelyou can do "screens[i].screen_type == SCREEN_MAIN" etc
14:55:10kkurbjun:-D, I missed that
15:02:33 Quit antil33t (Read error: 104 (Connection reset by peer))
15:02:46 Join antil33t [0] (n=Mudkips@
15:03:53 Join Sajber^ [0] (
15:05:07 Nick polobricolo is now known as polobricolo_ (
15:06:16 Join DarkDefender [0] (
15:11:01 Nick fxb__ is now known as fxb (
15:13:05 Quit kkurbjunW (Remote closed the connection)
15:13:36 Join kkurbjunW [0] (n=karlk@
15:15:43 Quit kkurbjunW (Remote closed the connection)
15:16:12 Join kkurbjunW [0] (n=karlk@
15:23:48 Quit polobricolo_ (Read error: 110 (Connection timed out))
15:38:23 Quit bluebrother (Nick collision from services.)
15:38:27 Join bluebroth3r [0] (n=dom@rockbox/developer/bluebrother)
15:39:17 Quit petur (Remote closed the connection)
15:39:38 Join petur [50] (n=petur@rockbox/developer/petur)
15:44:03 Quit petur (Client Quit)
15:44:31 Quit antil33t (Read error: 110 (Connection timed out))
15:48:28amiconnkugel: The screen_type member is not supposed to be changed after api init. I don't see how it would help wrt kkurbjun's problem here
15:56:17 Nick fxb is now known as fxb__ (
15:57:40 Part decayedcell
16:03:41kkurbjunamiconn: the screen type would help me detect which screen the lcd operations are being done on, and I could use it as an argument to the stride macro
16:07:22 Join arohtar [0] (n=faemir@
16:07:41 Quit gibbon_ (Read error: 60 (Operation timed out))
16:07:45 Join gibbon_ [0] (
16:08:31 Quit faemir (Read error: 104 (Connection reset by peer))
16:08:37 Join bzed_ [0] (
16:17:35kugelamiconn: he didn't intent to change it, IIUC
16:18:13***Saving seen data "./dancer.seen"
16:18:17 Part LinusN
16:18:38 Join funman [0] (n=fun@rockbox/developer/funman)
16:19:21 Join Strife89 [0] (n=michael@
16:19:45 Join BlakeJohnson861 [0] (
16:20:48 Quit bzed (Read error: 104 (Connection reset by peer))
16:20:48 Nick bzed_ is now known as bzed (
16:23:59 Quit kkurbjunW ("Leaving.")
16:26:15funmankugel: i notice lcd problems with my fuze
16:27:01funmansometimes "stairs" splashes: each line goes 1 or 2 pixels to the right (from top to bottom), sometimes splashes are correctly displayed
16:27:12funmanI also notice the display going on and off on a quick rate
16:27:42funmanThe best way I found to reproduce it is to launch a demo which does intensive lcd updates (mandelbrot, pictureflow, cube to the max speed)
16:29:00 Quit Strife89 ("Gone to class.")
16:29:08amiconnTry plasma...
16:29:16 Quit BlakeJohnson86 (Read error: 110 (Connection timed out))
16:29:38kugelfunman: I also noticed that
16:30:00kugelit seems to be related to boosting also
16:30:12funmanamiconn: plasma looked fine for the whole second i ran it
16:30:45kugeltry playing bubbles with music playing, sometimes it'll go crazy, sometimes it'll be fine. If you force boosting bubbles is always crazy
16:31:03kugeladditionally, it also only seems to happen on "unaligned" or uneven transfers
16:31:59kugelplasma should be fine since it's doing lcd_update()
16:33:10funmanyou don't wait for the fifo to empty in lcd_write_data() after lcd_write_singla_data16
16:33:22funmanor rather, until the fifo is full
16:33:43funmanyou don't check if the fifo is full*
16:33:51kugelthe fifo isn't going to get full from a single transfer
16:34:03kugelI tried waiting for empty, no success
16:34:51funmanit should have been emptied when prior call to lcd_write_data() returned (or have at maximum 16 bits in it)
16:35:05 Quit bubsy (Read error: 60 (Operation timed out))
16:36:21 Quit DataGhost (Nick collision from services.)
16:36:29 Join DataGhost [0] (i=dataghos@unaffiliated/dataghost)
16:39:27 Join saratoga [0] (i=9803c6dd@gateway/web/freenode/x-vsphvxhubgycjlhw)
16:39:37saratogafunman: any basic DMM should work
16:39:45saratogain the US they might go for $10-20
16:40:19saratogaalso I saw your remarks about the YH battery voltage
16:40:30saratogai thought all was needed was a battery discharge curve?
16:42:15funmanyes but figuring out the real voltage would be nice as well
16:43:25saratogafunman: can you fully charge it in the OF and assume thats equal to 4.25 v?
16:43:32funmanwhy 4.25V?
16:43:44saratogathats how much a lipoly battery should be when fully charged
16:44:25funman : 3.7V Li-ion battery for YH-920
16:45:24saratogai think all 3.7v lipoly batteries go up to 4.25
16:45:29saratogaat least thats how it was for the sansas
16:45:33funmanthe wiki has no pictures for yh-920 battery
16:45:44kugelfunman: if you want you can revert r22574 and r22575 (I don't have time now). The blue pixels on the e200v2 and FS #10580 should be fixable by increasing BUTTON_DELAY for both in button-e200v2-fuze.c
16:45:59 Join petur [50] (n=petur@rockbox/developer/petur)
16:46:18funmankugel: i can wait ^^
16:46:20kugelalright, I can too :)
16:46:35saratogathe 3.7v on the clip says "Charging the cell initially with constant current at 0.5C and then with constant voltage at 4.2V till charge current < 0.05C"
16:46:52saratogai think its just a quirk of the chemistry in lipoly
16:47:22 Quit kugel (Remote closed the connection)
16:49:48 Quit einhirn ("Miranda IM! Smaller, Faster, Easier.")
16:50:50funmanhm when it was fully charged the adc data was maximum (10 bits set)
16:51:11funmanand the last value read before power off was 916
16:51:29funmanrockbox didn't power off because it thought the battery was 93% full
16:51:45robin0800c200 "maximum art", strange behavior after wps changes, the first alternating screen dosn't obey its "t" setting on start up just flashes
16:52:00 Join BlakeJohnson86 [0] (
16:52:10funmanso, it stopped at 3.8V, does it make sense?
16:54:50saratogano it should go much lower then that
16:55:10saratogaare you sure you have the correct channel on the ADC?
16:58:43 Quit BlakeJohnson861 (Read error: 110 (Connection timed out))
16:59:09funmanno but what else could it be ?
16:59:28funmanif it decreases almost linearly with battery
17:00:14funmanwell there was no change within the first hour
17:00:15saratogaperhaps the slope is just off? you could adjust it so that 916 is ~3.2v
17:01:25saratogacould be a floating pin, those might go up and down with the supply voltage even if they're not directly measuring it
17:01:55 Join PaulJam [0] (
17:02:59 Quit Zagor ("Don't panic")
17:03:24saratogafunman: did you try doing a battery bench? the curve from that should give us a good idea if the ADC is really measuring the voltage or not
17:06:28 Quit DarkDefender ("Leaving")
17:06:43 Join low_light [0] (i=c730180a@gateway/web/freenode/x-bzryntfhiwtsqgdw)
17:08:30dionoeayou should "set style data line"
17:08:33funmanit really starts at 1:30 (i think it was the rockbox uptime)
17:08:52funmandionoea: thanks, and i should get a title for the data ;)
17:09:01low_lightfunman: the yh9xx OF adjusts the adc readings as such
17:10:33low_lightalso, the OF calculates the battery level as: level = 100 * (ave - 775) / (924 - 775)
17:10:56funmanIs the PLL on ?
17:11:07funmanI have seen other numbers in the YH920 OF bootloader test mode
17:11:25funmanI don't know if those are right / how to enter test mode
17:11:48low_lightI don't think you can enter test mode
17:12:08dionoeafunman: If i remember correctly it's set title "bla" for the graph's title and label "bla" for a plot item's label
17:13:07funmanplot "plot.txt" using 1:2 title "discharge curve" with lines
17:13:30dionoeaah ok, so it was title too for the second one :)
17:13:40 Join FOAD_ [0] (
17:13:49funmanlow_light: the numbers i found might well be for another battery and have never been updated
17:14:01dionoeaisn't using 1:2 implicit ? (ok, i'll stop the offtopic talk)
17:14:20 Quit moos (Read error: 131 (Connection reset by peer))
17:14:47funmanlow_light: is the PLL dynamically toggled in YH?
17:15:04 Join moos [0] (i=mostafa@rockbox/staff/moos)
17:15:55low_lightfunman: it depends on the cpu freq (system-pp502x.c)
17:16:23funmanusing your numbers I see the battery went from 166% to 94% (adc[1] from 1023 to 916)
17:17:15funmanminus 0x1a would not change much
17:18:03low_lighthowever, it looks like maybe the pll is always enabled
17:18:24 Quit robin0800 (Remote closed the connection)
17:18:30funmanhave you checked if we use the correct adc channel?
17:19:51low_lightThose numbers were for the 925, let me see if the 920 is different
17:25:39 Join Creposucre [0] (n=86f69562@gateway/web/cgi-irc/
17:26:07low_lightfunman: looks like the 920 uses level = 100 * (ave - 850) / (1023 - 850)
17:27:10funmanmeans power off at 38% :/
17:28:07funmanor rather, 84% with battery full and 23% battery empty (removing 0x1a from values read)
17:28:46CreposucreDionoea: Did you found some time to test your remote?
17:28:56 Quit FOAD (Read error: 110 (Connection timed out))
17:28:56 Nick FOAD_ is now known as FOAD (
17:29:12dionoeaCreposucre: I completely forgot about it when I got home (and remembered this morning at work). I'll try not to forget tonight :)
17:31:14Creposucreno problem
17:32:16 Quit saratoga ("Page closed")
17:32:40Creposucredionoea: do you have others accessories than the remote? It should work for them too
17:33:31dionoeaI have the apple dock which features a line out, mic input, SPDIF output and an IR receiver which already kind of works when using the remote provided with a macbook
17:33:44dionoeaother than that not really
17:35:05funmanlow_light: have you seen the FM radio patch on flyspray?
17:36:20funmanmight need change in audio-pp.c
17:38:18funmankugel: the problem with fuze lcd might be related to buttons (i get invalid readings when the screen flickers)
17:38:35Creposucredionoea: What does not work correctly with the commands?
17:39:42Creposucrefunman: you mean change on the modification I made?
17:40:27 Join jgarvey [0] (
17:40:37funmanCreposucre: no, it was addressed to low_light
17:41:52 Quit petur (Nick collision from services.)
17:41:58dionoeaCreposucre: As far as I remember the only slight issue I still have (when using the radio remote for example) is the volume control which is done in hardware on the remote and not on the ipod. That might have changed recently though so I'll have to double check. I'll post all the info once I've given it a try tonight.
17:42:03 Join petur [50] (n=petur@rockbox/developer/petur)
17:44:23Creposucredionoea: ok. I added some stuffs to modify the ipod volume with a remote control, but not from the ipod to the remote yet since I had some tests to do
17:45:14 Join saratoga [0] (i=9803c6dd@rockbox/developer/saratoga)
17:45:50dionoeabasically, if I remember correctly, you had 2 serialized volume controls: the ipod's and the remote's
17:46:12dionoeawhile it seems that you only have 1 when using apple's firmware with the remote control (of course I could be mistaken)
17:46:30saratogai'm increasingly annoyed at the current "supported" and "unsupported" build system we use
17:46:43saratogai really think we should reconsider this
17:47:16funmansaratoga: what do you mean?
17:48:29Creposucredionoea: you're right, a modification of the volume on the ipod under the OF modify the volume of the remote
17:48:58saratogafunman: we have targets like the YH and AMS sitting as unsupported for ages which I don't think helps anyone
17:49:27 Join BryanJacobs [0] (
17:49:32saratogaIMO they should be supported and simply marked "unstable" since they're usable and we'd like people to use them, report bugs and perhaps develop if they're interested
17:49:41saratogabut not completely stable yet
17:49:44dionoeabut the volume of the remote and the volume on the ipod's port would be the same in the OF, while in rockbox you still have both volume filter which apply (ie volume applies on line out). Or maybe you changed that
17:49:53Creposucredionoea: I have to link the ipod volume to the remote and it is ok. I just thought that It might be a need to keep two separate level
17:50:18JdGordonsaratoga: you arn't the only one... were we going to say pretty much once the main build works it would fall under todays "supported" category and then have classes like gold/silver/bronze for how well?
17:50:19saratogasame as the Gigabeat S, having it languish as a half supported target for the last year has done nothing to encourage people to work on it, and a lot to make it harder for users and prospective developers
17:51:04saratogaIMO when a port is usable but still unstable we should encourage people to use it, since that tends to encourage people to work on it
17:51:13Creposucredionoea: yes I did, at least for the remote buttons to control the ipod volume level
17:51:40 Join ej0rge [0] (
17:51:56funmansaratoga: i'm not sure bug reports are useful at this stage (at least on AMS)
17:51:57saratogaJdGordon: I don't know what was decided, but I'd like to just have "stable", "unstable", and "unsupported"
17:52:17saratogafunman: no but they'd be nice on the YH for instance
17:52:35JdGordonstable, unstable and not working at all... supported is a bad term
17:52:47funmanI think one would just need to make a list of targets with their new status and send it to the mailing list => if no complaint, then it becomes official
17:52:47saratogastable, unstable and unusable
17:52:59low_lightsaratoga: the Philips hdd1630 should be supported too. It only lacks rbutil install and the manual.
17:53:02saratogashould be quite clear
17:54:33saratogastable means rbutil and most everything works, unstable means no rbutil but playback mostly works for day to day use, unusable means its missing playback and/or important drivers
17:54:39JdGordonstable and unstable dont work so well.... the clip is unstable plaayback but fine everything else... the fuze is definetly as stable as any other target... the mr500 would be in the unstable group because of crappy install and start procedure but playback is fine
17:55:10 Quit petur (Remote closed the connection)
17:55:23saratogathe fuze is unstable in the sense that we keep breaking drivers (well kugel) and other drivers simply don't work correctly (microsd)
17:55:44 Join petur [50] (n=petur@rockbox/developer/petur)
17:56:33saratogaits usable if you don't mind some revisions being less functional then others, and possibly overclocking your SD card
17:56:34JdGordonwe should use some rating system and then on the front page have "rockbox runs at various levels of completeness and stability on the following DAPs.... <click here> for further details and known issues" which leads to the TargetStatus wiki page
18:03:52 Quit Creposucre ("CGI:IRC (EOF)")
18:04:42GodEatersaratoga: that's essentially what was decided at DevConEurope, that a lot of targets need to be promoted from unsupported to some status inbetween fully stable and not supported at all
18:04:52GodEateralthough I think the Gigabeat S is still a special case =/
18:06:16GodEateruntil we have a confirmed working way to get a single bootloader on it at least
18:06:22LambdaCalculus37low_light: Ping
18:06:59TorneGodEater: do we actually know what the pattern is there, with which devices will take a single bootloader and which won't?
18:07:11GodEaterno more than we know what triggers the random reformating
18:07:19GodEaterunless you've made progress there ? :)
18:07:22Tornenot yet :)
18:07:34Tornesorry. recently i have been really busy and/or high. :)
18:07:43*GodEater shrugs
18:07:50GodEateryou're not on a timetable as I recall ;)
18:08:41Torneon the subject of hte beast, though
18:08:51Torneis someone willing to commit the flash dumping patch I wrote
18:09:02Torneor do we want to fix up that mess in debug_menu first
18:09:06LambdaCalculus37Torne: A flash dumping patch, you say? :)
18:09:12GodEaterFS# ?
18:09:13LambdaCalculus37Torne: What FS#?
18:09:14Torneit would probably be quite useful to hav ethe flash dumping patch in
18:09:29TorneFS #10410
18:09:42Torneit makes the huge ifdef'ed nightmare in debug_menu.c worse
18:10:04Tornethere was a discussion in here some time ago about it and we decided it would be nice to factor that crap out into firmware
18:10:19Tornethere were various methods of doing this proposed
18:10:22Torneand nothing really decided :0
18:10:37Tornebut, yah. whether that happens or not it would be nice to have flash dumping for the beast rom in
18:10:51GodEaterin that case I think it's rather silly to wait to commit this
18:10:54Tornebecause then i can ask people with beasts to dump their flash to check what version they have and what settings they have
18:11:06GodEaterif no-one has a plan for doing it any time soon, we're just hurting ourselves not having it in
18:11:08Torneand see if we can find a pattern for who experiences reformats and who can't use a single bootloader
18:11:22Tornei have 3 dumps I think
18:11:23GodEaterI should try and reassemble my first beast
18:11:28Tornewhich are all different
18:11:32GodEaterand try this on the working one I got recently
18:11:38Torneand all differ amusingly from the versions in the firmware updates as well :)
18:11:46Tornebut that might be explained now that i've found the config sectors
18:12:03Tornenot that i know what the data in the config sectors mean yet, or if it's actually changed in practise
18:12:07Tornebut it's *there*
18:12:14Torneand one of hte bits stops the reformats! :)
18:12:27Torneit also stops any method of recovering the device other than taking hte disk out, admittedly
18:12:27GodEaterwe've had similar weirdness with the Gigabeat F I think
18:12:30Tornebut it does stop it reformatting.
18:12:56GodEaterthat's a bonus
18:12:59Tornenot really
18:13:04GodEaterdid you manage to find a reliable way to trigger them then ?
18:13:08GodEaterthat you know this stops them
18:13:10Torneit means that whenever the condition that causes the reformats happens, the player bricks
18:13:17GodEaterI see
18:13:22GodEaterso it doesn't stop it very well ;)
18:13:25Torneinstead of reformatting it just crashes
18:13:31Torneand will continue to crash on boot until the problem goes away
18:13:38Tornewhich requires fixing the disk using an ipod or similar ;0
18:13:43Tornei haven't actually tried it, either
18:13:53Tornebut it's checked in all the paths i have found so far that cause reformats
18:14:00Torneit's likely it's a flag for the device type, or similar
18:14:03Torne"is this a devboard"
18:14:11Tornei.e. so that test hardware doesn't reformat
18:14:19Torneon the assumption that someone wants to debug it.
18:14:32LambdaCalculus37low_light: FS #10431 is waiting for you to check out.
18:14:39Tornethere's also a flag that controls how much memory it thinks it has :)
18:14:42GodEaterwell we could definitely do with some more dumps then I guess
18:14:42Tornewhich again is for the devboard
18:14:52Torneturns ethernet on as well i think
18:14:56funmanTorne: i have put the cleaning of debug_menu.c on my TODO list but didn't feel like doing it O:-)
18:14:59*GodEater is also suffering from lack of time atm :(
18:15:08Tornefunman: ah yes it was you i was discussing it with
18:15:16Torneanyway yeah i think for now just shove my patch in as is :)
18:15:20*GodEater votes that someone with commit access puts the dump patch in then
18:15:21Tornesomeone can clean debug_menu.c later
18:15:31GodEaterthough I realise votes don't count for squat :D
18:15:34Tornehaving an extra dozen lines in doesn't make it measurably worse :0
18:15:49Tornealso, I think FLASH_SIZE for the beast is wrong
18:15:54Torneas mentioned in the FS
18:15:56GodEateryes, I read the comment
18:16:00Tornei haven't changed it in the patch
18:16:07Tornebut all three dumps i have are just mirrored
18:16:13GodEaterit's a pity jhMikeS has fallen off the intertubes
18:16:16Torneand the available schematics on the wiki indicate the smaller size is right
18:16:49Tornethe devboard has 4x the flash, as well, according to the schematic
18:16:55Torneso the size is not right for that either :0
18:17:09LambdaCalculus37GodEater: No one's seen or heard from him?
18:17:18Torneit's not the end of the world to have mirrored dumps
18:17:19GodEaterwell I haven't heard from him for AGES
18:17:25Tornethough it means people need to punt bigger files around
18:17:25GodEaterhave you ?
18:17:45funmanlogbot has seen him 1month 10 days ago
18:17:50Torneif someone puts the flash dump patch in, then i will write a lil tool that tries to ID dumps
18:17:54GodEaterfunman: lucky logbot ;)
18:17:59Tornei'll see if i can work out what regions are supposed to be variable or not
18:18:02Torneand md5sum them seperately
18:18:13*GodEater coughs at LambdaCalculus37 and asks him to do the deed
18:18:13Tornethen we can see if everyone's dump matches up to one of hte known ones or not
18:18:14***Saving seen data "./dancer.seen"
18:18:40Tornethere's at least three things in the flash: the pmcboot_secure.bin which is the actual loader, the config page, and the recovery.bin
18:19:05Tornethere is some evidence that beasts do *not* always reflash themselves when upgraded
18:19:23Torneiirc someone had 1.3 firmware on their beast but the bootloader in their flashdump was definately older
18:19:35GodEaterthat doesn't sound like a recipe for success
18:19:44Torneit seems to work anyway, though
18:19:51GodEateryeah, but it's still weird
18:20:21Tornewell yes :)
18:20:28Tornebut there's lots of weird things in the beast rom
18:20:48GodEater"here be dragons"
18:21:01TorneI documented the disk setup ones at btw if you've not seen it
18:21:06Tornewell the ones i've found so far
18:21:19*GodEater hadn't seen it
18:21:20Tornethat's just the partition setup and disk checks, the "first round" of stuff that might cause a reboot
18:21:22*GodEater goes to read
18:21:35Tornei think there are more later when it actually tries to access the FS for real to read the firmware
18:21:46Tornethe code i've reversed only looks at MBR and BPBs
18:21:50Tornenothing in fat
18:21:56Torne(so not the TFAT stuff yet)
18:22:02GodEaterBPBs ?
18:22:10Torneboot parameter blocks
18:22:13Tornefirst sector of the partitions
18:22:17GodEaterah ok
18:22:31TorneFAT's version of superblock ;)
18:23:00GodEaterand those don't differ between FAT32 and TFAT it seems
18:23:26TorneNot sure if the volume label being TFAT is significant or not yet
18:23:35Tornenot got to the point where it might read that
18:23:56GodEaterdo you know what it is about the disk when it's "virgin" that means linux refuses to mount it?
18:24:02GodEaterI've never figured that out
18:24:08Tornelinux refuses to mount it?
18:24:23GodEateryeah, you have to open it up in fdisk and re-write the MBR
18:24:41 Join Horscht [0] (n=Horscht2@xbmc/user/horscht)
18:24:42Torneon repartition it seems to write the mbr out fairly sensibly
18:24:42 Join SUSaiyan [0] (
18:24:51Tornemakes a 150mb partitoin and a rest-of-disk
18:24:56Tornei've not actually run any of this code though
18:24:59Torneand this is only the 1.3 loader
18:25:04Tornewhich i suspect few people actually have
18:25:15Tornebut i'm reversing this one because everyone could upgrade to it if necessary
18:25:20GodEater <−− see Step 2
18:25:43Torneoh, it doesn't set the bootable field?
18:26:05funmanunsigned int lol = 3600*1024*1024; warning: integer overflow in expression
18:26:05GodEaterseems not
18:26:08TorneIt doesn't check those. :)
18:26:14funmanI don't understand how it can overflow a 32 bits integer?
18:26:20 Join robin0800 [0] (
18:26:20Tornethe bullet points i put on the wiki there is *all* it checks
18:26:33Torneso it's possible that ATACreateMBR doesn't write valid values to the status byte either
18:26:37Tornei'll check it out
18:26:47funmanespecially when 3600<<20 doesn't give a warning !
18:27:07 Quit BryanJacobs ("Java user signed off")
18:27:11Torneit should be ok to just set it to 0x00 or 0x80 though
18:27:28Tornesince it doesn't check
18:27:40Tornewhich is what the instructions say to do :)
18:27:42robin0800WPS %pt used to be track length it is now total tracks length is this correct?
18:27:56Torneanyway off for now
18:28:11Tornei will probably have a look at the dumps later and see if i can make proper notes on the wik i of which bit is which
18:28:17Torneand the md5sums of bits I know
18:28:20Torneso people can compare
18:31:07funmanQ. can rockbox FAT driver write 2GB+ files?
18:33:13 Quit Rob2223 ()
18:33:34 Join Rob2222 [0] (
18:35:10 Join JdGordon_ [0] (i=441bed28@gateway/web/freenode/x-bwkwlckzzbpfdmyn)
18:35:57 Join intrados_ [0] (
18:36:18*GodEater notices we "now have a source for the nk.bin file"
18:36:23GodEaterwhere the hell did we get that ?
18:36:36saratogatoshiba put up a firmware link last winter
18:36:57saratogalike 2 days after i finally got around to writing code for dumping the one from the disk
18:36:59 Quit intrados_ (SendQ exceeded)
18:37:03saratogawhich is why i stopped working on it
18:37:07GodEaterit's like they knew
18:37:10JdGordon_maybe they used that code?
18:37:30 Join Lear [0] (i=chatzill@rockbox/developer/lear)
18:38:44 Join intrados_ [0] (
18:40:24 Quit pamaury ("exit(*(int *)0 / 0);")
18:40:27ej0rgeyeah, there are a few different versions of the beast firware floating around
18:41:23 Quit petur (Read error: 113 (No route to host))
18:41:46ej0rgeI still think that somebody, somewhere, has the v1.1 restore utility forgotten somewhere on their computer. At least, google turns up forum posts of people who say they used it.
18:42:01ej0rgei have v1.2 and v1.3 upgrade packages
18:43:15saratogaisn't the 1.1 restore tool actually for the gigabeat V and hacked to work with the S?
18:44:23ej0rgesaratoga: I saw at least one person claim that they got a tool from toshiba to restore their S, that predated the release of v1.2
18:44:47ej0rgeanyway, bbl
18:47:11saratogais something like this better discussed on the SVN or Developers list?
18:48:29 Join bertrik [0] (
18:48:42saratogathe supported unsupported thing
18:49:03JdGordon_dev ml....
18:49:29JdGordon_it has been agreed on sort of.. so its just a matter of putting targts in the right status and fixing the site...
18:51:33 Part fish_
18:52:20saratogais there a description of what was decided?
18:55:00 Quit bertrik ("ubuntu needs a reboot")
18:55:35JdGordon_i dnt remember if it was irc or ml.. but it was decided at DCE anyway
18:55:41JdGordon_or at least that it shuld change
18:57:11 Join bertrik [0] (
18:59:28pixelmaJdGordon_: haven't tried myself yet but did you assign a default value for %t again (it was possible before to use it without number in which case something like 2 seconds was used, since I never used it the "pure" way I wouldn't notice, just wondering)
19:00:05JdGordon_%t with nothing is the same as ; isnt it?
19:00:25robin0800pixelma: I think this is now broken
19:01:16pixelmaJdGordon_: did that work before? As I said, I always used other timeouts so wouldn't know
19:01:49JdGordon_I did testing with just ; and =with %t with a timeout.... it should work
19:01:57JdGordon_robin0800: %t with no number is broken?
19:03:03robin0800pixelma: pt has changed i think and %t in alternate conditionals neither of my wps works properly now
19:04:21pixelmarobin0800: %pt wasn't touched, maybe it doesn't work correctly but I don't understand you descriptions what it does differently now
19:04:22 Nick dys` is now known as dys (n=andreas@
19:05:40robin0800pixelma: it used to show track time now its showing tracks total time
19:09:06pixelmaI can't imagine what you mean with "track time" and "tracks total time"... is your %pt on a subline by any chance?
19:09:09robin0800pixelma: perhaps that should be a question as I'm not sure what it is displaying now
19:09:09JdGordon_tracks total time?!
19:09:22gevaertsGodEater: that fdisk step should no longer be needed. The USB storage code cheats a bit on the beast
19:09:25low_lightLambdaCalculus37: Hi! I'v been really busy lately, but I hope to get to the hdd6330 port soon. I haven't tried your patch for the sa9200 plugins, but as long as it doesn't break anything, feel free to commit it.
19:10:06pixelmaI'll update and see what my WPS does
19:10:26robin0800pixelma: no its in a bunch of conditionals
19:10:58 Join stoffel [0] (
19:11:13*JdGordon_ bbs
19:12:54gevaertsAlexP: do you feel like hiding some more themes? dCleanAA sets the language to Brasilian Portuguese (and seems obsolete anyway), and dCleanAA-2 still sets lang (english this time), as well as scroll delay and scroll speed (I don't think those belong in a theme, but I'm willing to be convinced otherwise)
19:13:14 Join panni_ [0] (
19:13:23AlexPgevaerts: Sure (which target?)
19:13:46gevaertsoops, sorry. e200
19:14:14 Quit JdGordon_ (Ping timeout: 180 seconds)
19:14:31pixelmaI thought those themes would be rejected anyway? Or couldn't those settings be stripped from the cfg file?
19:15:19gevaertsI seem to vaguely remember that linuxstb told me to be nice and not do that when I proposed to upload such a cfg file
19:15:37gevaertsI think they should be either rejected or cleaned automatically, yes
19:15:59LloreanI agree.
19:17:15gevaertsI did only find two theme authors who got it wrong though
19:17:41pixelmarobin0800: from CustomWPS) %pt - Total Track Time (the length of the track). That hasn't been changed in a loooong time
19:18:26AlexPgevaerts: OK
19:18:49robin0800pixelma: I know but something has stopped it working
19:19:07AlexPI deleted the first one (as he uploaded the 2nd to replace the first), and hid the second one telling him to let me know when it was fixed.
19:19:10gevaertsAlexP: I looked at cfg files for all themes, so I think I won't bother you with this for a while now :)
19:19:27AlexPGood work, and I don't mind :)
19:19:56LloreanAlexP: So, apparently the guy working on DxVA (even posted and said he had some renderers he thought were ready to go, except that he couldn't get a directX surface to render on, since at the time the Dx build wasn't usable yet) is facing DMCA charges for something else, supposedly
19:20:03LloreanOops, wrong channel, sorry
19:21:04Rondomhello, I have just realized that my buildclient glasschale-Rondom is blocked because it disconnects every 10 seconds
19:22:30Rondomthe reason for the disconnects is the following message: "HELLO failed: error duplicate name!"
19:23:12gevaertsdo you have two clients running?
19:23:18Lloreanrobin0800: What exactly happens? "something has stopped it working" isn't a useful bit of information on its own
19:23:36robin0800pixelma: the second problem is on "maximumart" where now the first screen dosen't obey the display time
19:24:05 Join JdGordon| [0] (n=Miranda@nat/microsoft/x-usrafijcwsbafueu)
19:24:07Rondomgevaerts: no, but I do have had this message from time to time, when I got disconnected and tried to reconnect
19:24:37gevaertsRondom: I think you'd better wait for Zagor then
19:24:51robin0800Llorean: it shows a time that appears to be a sum of the playlist tracks times
19:25:28pixelmaRondom: I had that too once, killing the client and restart helped
19:26:03JdGordon|robin0800: you arnt using a custom build are you? iirc there is absoluytlyu nowhere that the count is being kept
19:26:21LloreanJdGordon|: There's nowhere to even *make* that count without opening all the files involved, right?
19:26:34Rondompixelma: hmm, maybe I can try to reproduce by disconnecting my internet connection
19:27:16pixelmaI guess it could be a random number which might be close to the complete playing time of the playlist by accident... but I update and have a look myself
19:27:25robin0800JdGordon|: no and this afternoon ive been using the sim
19:28:06robin0800pixelma: yes I'll go with a random number
19:29:16LambdaCalculus37low_light: I will, but I wanted a second opinion on the patch, and perhaps a little help getting the dimensions for a few of the plugins right.
19:36:23funmanFlynDice: can you look at FS #10507 if I forgot something and test on your side?
19:38:47funmanI can use the µSD card at a correct frequency with your patch
19:39:28 Quit funman ("free(random());")
19:50:05 Join Lss [0] (
19:50:32 Join fml [0] (n=4fd3c45a@gateway/web/cgi-irc/
19:50:32 Join petur [50] (n=petur@rockbox/developer/petur)
19:51:02fmlI'm sorry to ask again but have the USB capable bootloaders been released for sansa e200 (v1)?
19:51:24 Nick n17ikh| is now known as n17ikh|Server (
19:52:16gevaertsyes and no. They are on the download server, but not in the correct place
19:54:42pixelmaI can't see anything wrong with the %pt info in my c200 WPS. robin0800 - are you sure it's not the file(s) you tried?
19:54:47fmlgevaerts: so it's impossible to download them? Have they been tested? I.e., for example, do you use them? Do you have problems?
19:55:40gevaertsfml: you can download them, and they are tested. rbutil won't find them though
19:59:15 Join domonoky [0] (n=Domonoky@rockbox/developer/domonoky)
20:02:11fmlgevaerts: is it what's in ?
20:02:31gevaertsno, that's the old one
20:03:02 Join Hillshum [0] (
20:04:03pixelmahuh, don't know how I managed that but at the moment the WPS shows the name of the next track as title of the currently playing one - and "next track" info shows the same. Playlist position is correct though
20:04:19JdGordon|ah fooey
20:05:45domonokygevaerts: changing bootloader location on the download server is not a good idea. rbutil will install the old ones :-)
20:06:00pixelmaand I don't know if it was the case before but if you seek for a while scrolling lines stop (at "position 1")
20:06:18fmlgevaerts: should I just follow the installation instructions in the manual? Or how do I install it? Is there a wiki page?
20:07:31pixelmaJdGordon|: that next track/current track info problem did not correct itself on track change, will keep an eye on that
20:07:35fmlgevaerts: there's no mention of pp5022.mi4 there, only sansa patcher
20:07:58gevaertsdomonoky: I know that...
20:08:11gevaertsfml: the download location has both
20:08:26domonokygevaerts: good :-) so they will get moved if final ?
20:08:35pixelmaJdGordon|: not even stopping playback and resuming helped
20:09:08pixelmaa complete reboot did...
20:09:24gevaertsdomonoky: as soon as Zagor or Bagder get to it
20:10:32pixelmacan't reproduce easily though and don't know how I got there
20:10:36PaulJampixelma: did you change the theme before the next/current track problem occured? i have seen this problem too, but only after loading a theme.
20:10:50*gevaerts is one of several people who do not have write access to the download master
20:11:02pixelmaPaulJam: yes, I was testing some
20:12:19pixelmaPaulJam: do you remember about when you saw this the first time - recently or a while ago?
20:13:40fmlgevaerts: thank you! All went well. One last question: the battery charging works as before? I.e. if I just connect the player to the PC while the player is off, the battery will be charged?
20:15:12gevaertsit should
20:16:38PaulJampixelma: i think i saw it about a week ago the first time, but before that i did not update rockbox for a few weeks.
20:18:17***Saving seen data "./dancer.seen"
20:18:56fmlgevaerts: OK!
20:24:49 Join Thundercloud [0] (
20:27:06FlynDicefunman:(for the logs) re FS #10507 I'll take a look tonight sometime, very busy with real life these days....
20:27:58 Join Strife89 [0] (
20:29:19HillshumFlynDice: I'll apply it to my e200 sometime today
20:35:53 Quit petur (Remote closed the connection)
20:36:49 Join petur [50] (n=petur@rockbox/developer/petur)
20:38:00 Quit fml ("CGI:IRC (EOF)")
20:41:57 Quit petur (Remote closed the connection)
20:42:22 Join petur [50] (n=petur@rockbox/developer/petur)
20:44:07 Quit amiconn (Nick collision from services.)
20:44:09 Join amiconn_ [0] (i=quassel@rockbox/developer/amiconn)
20:44:26 Join pixelma_ [0] (i=quassel@rockbox/staff/pixelma)
20:44:26 Quit pixelma (Nick collision from services.)
20:44:31 Nick amiconn_ is now known as amiconn (i=quassel@rockbox/developer/amiconn)
20:44:45 Nick pixelma_ is now known as pixelma (i=quassel@rockbox/staff/pixelma)
20:48:57 Join GeekShadow [0] (n=Antoine@reactos/tester/GeekShadow)
20:55:14amiconnGodEater: The "linux refuses to mount beast" cause is known
20:56:37 Join Strife1989 [0] (
20:56:51 Join MDCarr [0] (
20:57:06 Quit MDCarr (Client Quit)
20:57:18 Quit Strife89 (Nick collision from services.)
20:57:23 Nick Strife1989 is now known as Strife89 (
20:57:47 Join funman [0] (n=fun@rockbox/developer/funman)
21:00:08amiconn[18:31:07] <funman> Q. can rockbox FAT driver write 2GB+ files? <= A: No, it can't, since it uses 32 bit integer for off_t
21:01:17funmanand off_t is signed
21:01:31funmanI just ran a test with a bit less than 2GB
21:02:08funmantest_disk ran fine, but rockbox couldn't play from the µSD card, elapsed time just stayed at 0 (didn't investigate further)
21:02:42amiconnThat's in fact one of the things I wanted to look into (changing off_t to 64 bits). If it doesn't cost too much, it should be enabled for all targets
21:03:15amiconnI know that at least *some* OFs do handle 2GB+ files (the iriver firmwares do)
21:03:21funmanis there a real gain to allow 4GB files?
21:03:36LloreanRecording people will probably like it.
21:04:01JdGordon|how long would a 4GB mp3 recording be?
21:04:03funmani mean we would just put the limit a bit higher, not get rid of it
21:04:08amiconnYes - not choking on perfectly legal files is certainly a plus
21:04:32Lloreanfunman: A higher limit though, means that if you need to record say, a show that caps out at 3GB, you don't have a split.
21:04:35gevaertsJdGordon|: I'd suspect people who hit the limits to record to lossless
21:04:39amiconnI tested a 2.5GB WAV file on my H300. The OF plays it, rockbox simply skips it
21:04:54JdGordon|oh right we can record to wav
21:05:46amiconnYes, and we can even do that on archos (just still nonintegrated)
21:05:57*JdGordon| forgot that
21:06:49 Quit GeekShado_ (Read error: 110 (Connection timed out))
21:09:10JohnTeddyIf I have a 30 gig iPod Video that used to have rockbox on it, and it is broken. How can I tell what part of it is broken, how can I tell if it is the hard disk?
21:11:00 Join T44 [0] (
21:22:47HillshumJohnTeddy: An iTunes restore would help
21:24:26JohnTeddyHillshum: I see, will iTunes restore tell me what is wrong?
21:24:46gevaertsHillshum: isn't that a bit heavy-handed?
21:25:01 Quit PaulJam (".")
21:25:13Hillshumgevaerts: ? Isn't that the best thing to do to misbehaving iPods?
21:25:21 Quit Strife89 ("Huzzah!")
21:25:49gevaertsHillshum: clearing everything? That may be the most thorough way, but it should never be the first thing you try...
21:26:35HillshumJohnTeddy: Will it connect? Have you run a disk test like scandisk or fsck?
21:28:51 Quit Topy44 (Read error: 110 (Connection timed out))
21:31:27JohnTeddyHillshum: not yet, my ex says it's broken.
21:31:42JohnTeddy(It was a christmas gift, I put rockbox on it)
21:32:01JohnTeddySo since rockbox players are hard to find, I thought I might be able to put a new hdd in and fix it. since she doesn't know how.
21:32:09JohnTeddyI don't know the details of the problem yet.
21:32:21 Join antil33t [0] (n=Mudkips@
21:32:25JohnTeddyShe left her car jack here, so I'm going to see if I can get the broken player.
21:32:49HillshumJohnTeddy: Has she tried resetting it?
21:33:04JohnTeddyno idea, but I don't want her to fix it. heh
21:33:09JohnTeddyI want her to give up, and give it to me.
21:33:18JohnTeddyIt's been like this for several months. I hope she didn't throw it out.
21:33:41JohnTeddyI tried to buy a new player online, but the ones that work with rockbox are too hard to find.
21:33:58JohnTeddySo I figured this one is probably repairable, as long as she didn't static shock the ICs or something
21:34:58HillshumIs there a good possibility that she overlooked something simple?
21:35:21JohnTeddyThe device is a couple years old
21:36:07rasherJohnTeddy: Be advised that this channel is logged publically
21:39:29JohnTeddyIt's fine, she can't figure out how to get on irc. I told her I wanted to try and repair it, and ask if I could have it.
21:39:44 Join froggyman [0] (
21:40:36HillshumJohnTeddy: A google search will bring up stuff from this channel.
21:41:13JohnTeddyDid I say something taboo?
21:42:02HillshumNo, just don't say anything you don't want coming up. Sounds like things are fine
21:43:01 Join p3tur [50] (n=petur@rockbox/developer/petur)
21:44:23 Quit petur (Read error: 104 (Connection reset by peer))
21:45:27JohnTeddyIs it hard to get the device 'opened' and replace the hard disk, physically that is.
21:46:23 Nick p3tur is now known as petur (n=petur@rockbox/developer/petur)
21:47:03gevaertsDepends. It's not too hard to do if you don't mind scratches and other visible traces of you doing it. I have no idea how hard it is to do it properly
21:49:57JohnTeddythe thing is keyed/scratched to hell anyway
21:58:52 Quit Thundercloud (Remote closed the connection)
21:59:10 Quit Lear ("ChatZilla 0.9.85 [Firefox 3.5.2/20090729225027]")
22:02:55 Join BdN3504 [0] (n=55b20d37@gateway/web/cgi-irc/
22:03:25 Join bubsy [0] (n=Bubsy@
22:03:49BdN3504i think i found a bug in the wps code: since we have support for realmedia files now, is that change already in the wpstag for the codec? %fc ?
22:04:31LloreanBdN3504: Did you try?
22:04:51 Nick fxb__ is now known as fxb (
22:04:54 Quit LambdaCalculus37 ()
22:05:13CIA-43New commit by rob (r22609): Make the left quickscreen item work in touchscreen grid mode.
22:08:33 Join paprica [0] (n=47ba0bf9@gateway/web/cgi-irc/
22:08:41*low_light listens to radio on the yh920
22:09:03 Quit paprica (Client Quit)
22:09:51bertriklow_light, is that a first time for radio to work on yh920 rockbox?
22:10:47low_lightit's kind of a bad hack for now
22:11:31bertrikwhat's the "bad hack" part of it?
22:12:44low_lightwell, no regular audio output (from files), just the radio
22:13:12*bertrik has some code for the c200v2 button reading ready to test on actual hardware
22:13:20BdN3504Llorean: no, i did not, but the neither the wiki nor the manual list it.
22:18:20***Saving seen data "./dancer.seen"
22:19:00funmanlow_light: cool :)
22:19:06 Join p3tur [50] (n=petur@rockbox/developer/petur)
22:19:57 Quit petur (Read error: 113 (No route to host))
22:20:01LloreanBdN3504: So try it out, then write whichever patches are necessary?
22:21:21 Join Strife89 [0] (n=michael@
22:22:39BdN3504Llorean: i can't, sorry. but i'll try. where would i have to append this/which files would i have to look at?
22:23:03funmanlow_light: 820 and 925 have no tuner?
22:23:08 Quit p3tur (Remote closed the connection)
22:23:14funmanaccording to the status page on the wiki they dont'
22:23:21 Quit tvelocity (Read error: 104 (Connection reset by peer))
22:25:50 Join p3tur [50] (n=petur@rockbox/developer/petur)
22:31:24 Quit stoffel (Remote closed the connection)
22:32:05low_lightfunman: Correct. The OF doesn't have the option, but I haven't taken them apart to verify that the hardware is not there
22:34:31funmantaking yh920 apart is simple : no plastic locking (or very easy to take off)
22:34:53 Join tvelocity [0] (
22:36:31low_lightfunman: radio patch:
22:36:57low_lightit's probably got some othere stuff there too
22:37:18funmancan you switch between playback/radio ?
22:37:59funmanide power enable > should higher a bit the runtime, I got 8/9 hours
22:40:46bertrikIIRC, fmradio_i2c_init isn't usually done inside power_init
22:41:27 Quit tvelocity ("Αποχώρησε")
22:42:00low_lightfunman: it doesn't switch back to playback
22:42:31bertrikoh, it is on the beast
22:42:45 Quit bmbl ("Bye!")
22:42:58funmanlow_light: what if you reset akcodec completely? that would give a hint
22:44:12 Quit p3tur (No route to host)
22:45:42bertrikI'm a bit surprised that fmradio_i2c-yh920 apparently works, bytes are sent as 8-bits with no ACK it seems
22:47:07low_lightbertrik: you'd have to ask someone at Samsung about that ;)
22:48:33KlownerIs there a way to get the bootloader to load the default firmware? seems as though my Sansa won't charge properly when rockbox is running
22:48:47funmanbertrik: it's possible that people implementing i2c are laxist or workaround broken software implementations
22:49:02funmanKlowner: yes, this is documented in the manual;
22:49:26Klownerfunman: ah, I must be looking at the wrong section of the manual
22:49:27 Part JohnTeddy
22:49:29bertrikfunman, but it talks to a TEA5767 chip right, so it should be doing proper i2c
22:50:10funmanKlowner: is your sansa a c200 or e200?
22:50:23Klownerfunman: e200
22:50:32Klownerhold the left button? just found it
22:50:49funmanok ^^
22:50:55Klownerthank you :)
22:53:04JdGordon|hmm... /me just got error updating playlist control file doing mass insert shuffled :(
22:53:43bertrikfunman, low_light , maybe this isn't really i2c, I'll check the tea5767 datasheet if it supports something else besides i2c
22:54:31low_lightbertrik: I think it also uses a 3-wire bus
22:55:19 Quit Hillshum ("Ex-Chat")
22:56:10bertriklow_light, yes, I'm seeing that now too, let's not call it i2c in the source then :)
22:57:13 Join Hillshum [0] (
22:57:25JdGordon|how do you stop playback on the fuze?
22:57:46LloreanUnplug the battery.
22:58:02HillshumIs there a way to make test_disk test the uSD?
22:58:15JdGordon|ok, how do normal people stop playback? power doesnt seem to do it
22:58:33HillshumJdGordon|: wait for the battery to die
22:58:43bertrikhold play maybe? or look in the source
22:58:43 Quit Zarggg (Connection reset by peer)
22:58:56gevaertsJdGordon|: set repeat to off and make it run out of playlist items
22:58:57low_lightbertrik: but tea5767.c uses fmradio_i2c_read/write
22:59:30bertrikah ok, this is probably the first time that a tea5767 is used with 3-wire instead of i2c
22:59:31JdGordon|gevaerts: the playlist is broken... i have to stop it to clear the problem
23:00:48*JdGordon| gives up shuffling his whole DAPs content and goes back to just a single folder
23:01:19 Quit BdN3504 ("CGI:IRC (EOF)")
23:02:28 Quit GeekShadow ("The cake is a lie !")
23:03:25bertriklow_light, can you put at least a big fat warning in the code about this inconsistency?
23:03:30 Join Zarggg [0] (
23:04:19low_lightbertrik: sure :)
23:16:48saratogaLlorean: what was decided at devcon regarding releases?
23:17:35Lloreansaratoga: Dunno?
23:17:57 Quit low_light ()
23:17:58*bertrik can't remember
23:18:08gevaertssaratoga: we agreed on the principle, the rest was a bit vague
23:18:21funmansaratoga: there is a brief description on the devoconeuro2009 wiki page
23:18:42funmanwhich translates to "sansa ams should get on the download page" for me
23:19:21gevaertsfunman: all of them?
23:19:24saratogawell i wrote up a proposal for that but haven't sent it to the list yet
23:19:34gevaertsThere's also the email discussion I started in july?
23:20:05funmangevaerts: i don't know what was the state of each target by that time
23:20:42saratogain principle it sounds like most people agreed on having a 2 teer system for supported targets based on what I can find
23:20:58gevaertsfunman: the status at that time doesn't matter much now :) I mean, as far as I can see fuze and e200? are pretty stable, but is clip ood enough?
23:21:26saratogaas soon as the SD issue is fixed I'd say the Fuze and e200v2 can go
23:21:29saratogaclip is not good enough
23:21:44funmangevaerts: fuze and e200 are unstable when using µSD cards, and clip/m200v4/c200v2 are equally unstable when using playback (especially the c200v2 which lacks some buttons anyway)
23:21:50saratogaits only semiusable unless you stick to MOD files or something similar that never buffer
23:22:11gevaertsfunman: "µSD cards unstable" is something that can be documented easily
23:22:29gevaertsunstable playback is slightly harder to explain away
23:22:32funmanprovided you use some unicode medium :)
23:22:35saratogathe problem is that it overclocks the cards though, which is something thats fairly dirty
23:22:56saratogain theory it could corrupt them or maybe even do damage (though that seems extremely unlikely)
23:23:11bertrikc200v2 buttons will probably work this weekend
23:23:30gevaertsbertrik: but what about after that?
23:24:00funmanbertrik: in my experience c200v2 is more unstable than clip/m200v4 because of a shorter audiobuffer (color LCD needs more space)
23:24:37funmanYou could trigger a bug in buffering.c when playing a file with album art associated (album art wouldn't fit in the buffer)
23:24:38bertrikfunman, so do you think it's a fundamental problem?
23:25:11funmanOne needs to take a look at buffering.c and playback.c, I ran logf on my clip today but was too lazy to continue ...
23:25:22bertrikgevaerts, code will be committed to SVN and the buttons will keep on working after the weekend too
23:25:31funmanI think the low audiobuffer size might trigger some corner case bug in these files
23:26:02 Part froggyman
23:26:03 Quit Hillshum (Read error: 60 (Operation timed out))
23:26:16gevaertsbertrik: ah, nice! :)
23:31:35JdGordon|funman: so maybe we shuold disable AA on low mem?
23:31:42JdGordon|to at least get around the problem?
23:31:59funmanJdGordon|: i had a get around back then, but i wanted to read more buffering.c/playback.c first
23:32:12funmanAA itself works fine
23:32:13saratogathe color screen will also be a problem, since it uses quite a bit of memory
23:32:25saratogathe underlying problem with buffering needs to be fixed
23:32:28JdGordon|funman: well if its AA we can blame Unhelpful :)
23:32:29funman2MB jpegs are a bit of a problem though :)
23:32:40bertrikc200v2 doesn't have external DRAM, like the clip?
23:33:18saratogai think the clipv1, c200v2 and m200v4 are all the same
23:34:52JdGordon|ignoring the fact AA hardly makes sense on the clip at all.... I wonder if it would be possible to convert the jpg to a preloaded bmp but on disk instead of in ram, then we could still display it on low mem targets
23:35:19funmanbertrik: clip and c200v2 have 2MB external SDRAM
23:35:25funmanlike m200v4
23:35:52funmanclipv2 has 8MB afaict (not deeply tested)
23:36:22bertrikfunman, oh, I can't remember seeing external DRAM on Clip PCB scans
23:37:38 Quit Strife89 ("Going home.")
23:38:00 Join safetydan [0] (n=deverton@rockbox/developer/safetydan)
23:38:16funmanit's not visible
23:38:21 Quit robin0800 (Remote closed the connection)
23:38:31funmanperhaps inside the SanDisk chip, next to the AS3525?
23:39:30bertrikok, I'd call that internal since it's not a separate chip, but I understand what you mean now
23:41:41 Quit xavieran (Read error: 104 (Connection reset by peer))
23:41:42saratogahuh the mailing list really eats my posts
23:41:59Lloreansaratoga: Supported v. Unsupported Builds?
23:42:00gevaertsdoes it?☕
23:42:13AlexPsaratoga: It came through fine
23:42:16saratogawell on the website the text wraps oddly and some new lines are removed
23:42:18saratogaok good
23:42:30gevaertsah yes, it does that
23:42:49gevaertsor at least, something does
23:43:49 Quit bertrik ("De groeten")
23:53:54 Quit kkurbjun (Read error: 60 (Operation timed out))
23:56:49 Join Thundercloud [0] (

Previous day | Next day