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

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

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

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

#rockbox log for 2010-03-30

00:00:24archivatordomonoky: well, in the case of festival, for voicing as well (festival forks for each client)
00:01:01domonokyi think making voicing in parallel is risky, there are many different tts in use.
00:01:44archivatorwell, the TTSExec ones can be parallelized
00:01:52archivatorFestival too (I think)
00:02:10archivatorthe rest are out of my expertise :)
00:02:56notlisteningarchivator, are the festival voices your using free?
00:03:12domonokyhm, if we give the tts/encs a list of work, instead of only one entry, they could decide themself todo it paralell or not.
00:03:22archivatornotlistening: all festvox voices are free, yes
00:03:49notlisteningand they are better you say? ;)
00:03:52archivatordomonoky: error handling would become a mess then
00:04:07domonokyarchivator: yes, thats a problem.
00:04:24 Join Blue_Dude [0] (
00:04:33archivatornotlistening: depends on what you're currently using. Festival is primarily a research platform so all the flashy new research is available for it
00:04:39archivatorlook up multisyn
00:04:50domonokythe other big thing the voicing in rbutil needs is: voice correction as done in :-)
00:05:03notlisteningwill give it a go thans archivator
00:05:59notlisteningdomonoky, I am planning to have an online database backend to open-sapi app/task based vocab if you want to feed from that eventually?
00:06:23archivatordomonoky: you know what's funny? Qt doesn't provide a way to query the # of available cores :(
00:06:56domonokyarchivator: it has to know it somehow, QtConcurrent creates threads based on # of cores.
00:06:56archivatorwait, my bad
00:07:11domonokynotlistening: sounds interesting.
00:08:29notlisteningdomonoky, when i get it done I will ping you :D
00:08:47notlisteningtill then sleep is need as i work tomorrow end of holidays :D
00:09:07*domonoky needs to sleep to..
00:09:37 Quit domonoky (Read error: Connection reset by peer)
00:14:25 Quit notlistening (Quit: Leaving)
00:17:40 Quit DV (Ping timeout: 276 seconds)
00:18:48 Join S_a_i_n_t [0] (S_a_i_n_t@
00:21:57 Nick fxb is now known as fxb__ (
00:24:35kugelfunman: the of reads D6 twice because both power and hold are on it ;;
00:24:47 Join Schmogel [0] (
00:25:20funmanhm :/ how do you differentiate ?
00:25:23 Quit robin0800 (Remote host closed the connection)
00:25:56RadicalRjust a quick report. funman, your latest update fixes the LCD. Just wanted to confirm that for you.
00:27:01CIA-5New commit by 03funman (r25390): as3525v2 : recording (FM only, no microphone yet) ...
00:27:11funmanRadicalR: on fuzev2?
00:27:27RadicalRNo buttons yet, but you haven't worked on that yet.
00:27:31kugelfunman: if it's set on the first, but not on the second read, it's hold
00:28:27funmani'd swear it would read as set while hold was off and power unpressed
00:30:34funman -> Clip+ has less red than Clipv1 \o/ (but still more yellow)
00:35:04 Part Llorean
00:40:36kugelthere needs to be a way to read 16bits via dbop din;;
00:41:09funmanOF uses ldrh?
00:42:23*archivator fixed 1 half of the festival integration
00:42:35archivatorother half when I get my sleep *yawn* :)
00:42:55 Quit archivator (Quit: Leaving)
00:43:28kugelit looks a bit like CCU_IO bit 12 is the old afsel switching
00:44:36 Quit tmzt (Ping timeout: 276 seconds)
00:44:49 Quit ender` (Quit: Female software engineers become sexually irresistible at the age of consent, and remain that way until about thirty minutes after clinical death. Longer if it's a warm day.-- Scott Adams)
00:49:05 Quit stripwax (Quit:
00:49:48 Join S_a_i_n_t_ [0] (S_a_i_n_t@
00:51:38 Quit petur (Quit: Zzzz)
00:52:21CIA-5New commit by 03funman (r25391): as3525v2: microphone recording
00:52:27 Quit S_a_i_n_t (Ping timeout: 245 seconds)
00:53:16 Nick S_a_i_n_t_ is now known as S_a_i_n_t (S_a_i_n_t@
00:53:19funmani can't change the volume when recording, i believe it's a keymap proble
00:57:02 Quit anubisfremen (Quit: Page closed)
00:57:02 Join tmzt [0] (
00:59:32 Join krabador [0] (darkham@
00:59:36Tornehow is stuff in firmware supposed to access settings?
00:59:40***Saving seen data "./dancer.seen"
00:59:42Torneor is it not? :)
01:00:03 Join CaptainKewl [0] (
01:01:04 Part toffe82
01:02:43funmancould the rockbox fund sponsor me with a Clipv2 ?
01:03:07CIA-5New commit by 03kugel (r25392): Fuzev2: Preliminary button support. Scrollwheel does not work yet.
01:04:28Tornefunman: sounds like a plan :)
01:04:46kugelfunman: I'm sure it could :)
01:05:31funmani'll talk with zagor tomorrow
01:05:52*funman made some red
01:06:11 Quit komputes (Remote host closed the connection)
01:06:43kugelis gcc bad at armv5 or why is the binsize/ram usage higher on clipv2 than on clipv1?
01:06:50funmanah only for fuzev2 though
01:07:47funmankugel: it's not much different: only 10k on binsize, 20k on ram (before recording was added)
01:08:20 Quit Schmogel (Read error: Connection reset by peer)
01:08:51kugelI'd expect it lower, though
01:10:02CIA-5New commit by 03funman (r25393): fuzev2 FM doesn't work (yet), don't try to record from it
01:10:45funmankugel: red in bootloader
01:10:56kugelfunman: where does the of use ldrh? I assume all the buttons can be read from dbop in again (most of the ones I found can already, but I only get 8bits of dbop_din)
01:11:04funmanah fuzev2 is in the binsize table, but there's no label for it
01:12:20funmankugel: dunno, i erased all my fuzev2 disassembly the other day :/
01:13:56kugelmine isn't very helpful?
01:14:04funmancouldn't read it
01:14:15kugelah, yes
01:14:46CIA-5New commit by 03kugel (r25394): Fix bootloader red, it doesn't have backlight handling.
01:15:19CIA-5New commit by 03moos (r25395): Update the norwegian (norsk) translation. ...
01:17:12 Quit Stephen__ (Quit: Leaving)
01:18:17CIA-5New commit by 03moos (r25396): Update the norwegian (nynorsk) translation. ...
01:18:22 Quit krabador (Ping timeout: 276 seconds)
01:18:53 Quit HellDragon_ (Read error: Connection reset by peer)
01:19:11 Join jd [0] (
01:19:12 Quit jd (Changing host)
01:19:12 Join jd [0] (~jd@Wikipedia/HellDragon)
01:21:25 Join webguest95 [0] (
01:22:19 Quit Adubb (Read error: Connection reset by peer)
01:22:42 Join Adubb [0] (~aldubuc@
01:23:53Blue_DudeGoing to commit FS #11081 (hotkey patch) in a few minutes. Speak now or forever hold your peace.
01:24:45 Quit webguest95 (Client Quit)
01:27:22moosBlue_Dude: Hi, you don't want to work on more keymaps before to commit it?
01:28:19Blue_DudeI've got e200 and c200 in the bag. I just don't know enough about others to change their keymaps. Even so, this patch doesn't change them at all.
01:28:54Blue_DudeWhich ones would you like to work on?
01:29:29kugelthe home button on the fuze is free I think
01:29:34moosyeah I understand it, but then the targets in question canot benefit yet of the feature... probably not too crucial but anyway
01:30:03kugelfunman: I thought sd card works on the fuzev2?
01:30:32moosBlue_Dude: Nice to see a NoDo kicked off btw
01:30:45funmankugel: they should? I just listened to a mp3 on µSD
01:30:55kugelmine is not detected
01:31:04funmanyou plug it at boot ?
01:31:08kugeli.e. doesn't show up on the browser
01:31:13kugelI tried both
01:31:36Blue_Dudekugel: I looked at the other Sansa targets, but some were still in development. And their existing keymaps were inconsistent at best. I think they need a more systematic look than just a new key.
01:31:40funmanperhaps it gets confused with B5 ?
01:32:18Blue_Dudemoos: That wasn't the plan. I guess I'm just stiubborn.
01:32:23kugelBlue_Dude: which sansa are you talking about? the fuze is mostly finished
01:33:05Blue_DudeTheir keymaps were confused. I can't figure out the big picture.
01:33:29*kugel never saw a big picture in any keymap
01:33:35funmanhow could I generate musical tunes from a computer (without an instrument) to test pitch_detector ?
01:33:41Blue_DudeYeah I know. :)
01:34:00moosBlue_Dude: at least that made us revise our NoDo concept and made it more formal
01:34:16Blue_DudeWhich is always a good thing. :)
01:34:40 Join Llorean [0] (~DarkkOne@rockbox/user/Llorean)
01:34:51funmannevermind, random noise in the microphone does the trick :p
01:35:52LloreanAnyone else have an opinion re: committing features without finishing keymaps for them for all targets they could work on?
01:36:24moosfunman,FlynDice,kugel: you guys works so fast yet another congratulations for you
01:36:59kugelLlorean: I think it's a pain
01:37:30funmanthere are too much keymaps
01:37:32kugel"could work on" isn't immediately clear as well
01:37:36Lloreankugel: I'm of the opinion that if a feature could work on multiple targets, it should be working on them before the first commit
01:37:47linuxstbBlue_Dude: Looking at your patch quickly, the first thing is your #ifdef for the onplay function - that's ugly... Couldn't you use something like the "IF_MV" macro?
01:37:49LloreanEspecially if all it needs is a keymap for each of them
01:37:53moosLlorean: I guess that once that goes in, people could help submitting patches... no blocker here
01:37:55kugelwhile I do appreaciate if it works on many targets from the beginning I don't find it's a requirement
01:38:15moossame here
01:38:49kugellinuxstb: I absolutely hate the IF_MV macro
01:39:04moosand judging the speed you are porting new sansas, lot more targets/keymapson the near futur
01:39:10Blue_Dudelinuxstb: I changed it to make it cleaner in the onplay.c file. I could easily make it cleaner in the onplay.h file as well. Or are you referring to the function calls?
01:39:35linuxstbBlue_Dude: The function calls - surrounding a parameter with #ifdef (and therefore newlines)
01:39:40kugelbut yea, that one in the patch is ugly as well
01:40:22Blue_Dudelinuxstb: Yeah, they're ugly. I could macro it to make it cleaner looking, but it's still not going to be pretty.
01:40:24kugelI'd just pass it unconditionally and (void) it in onplay. a single parameter isn't worth making code ugly and hard to work with IMO
01:41:02Blue_Dudekugel: that was the original implementation. I made it uglier to save bytes.
01:41:43LloreanBlue_Dude: Seriously, why aren't there keymaps for other targets with a free short press?
01:42:13kugelBlue_Dude: do you think that few bytes is that important?
01:43:20 Join Hillshum [0] (
01:43:28LloreanThere's several targets that have a record button exactly like on the e200
01:43:57LloreanIt's just very frustrating that an entirely apps-related feature is going to be disabled on targets because the author doesn't feel like doing a little research and fill in the keymaps.
01:44:17kugeldoing the work is very frustrating as well
01:44:48LloreanYes, but the commit could be held back until the work is done instead of enabling apps code for only one or two targets.
01:44:50Blue_DudeLlorean: I would to keep the hotkey consistent across different menus. Maybe the same free key isn't available for all menus? Or maybe even if it was free, it would be too easy to hit unintentionally. I *don't know* enough about those targets to make changes of that sort so I *didn't make them*. *You* are perfectly free to make those changes if you think they're worthwhile. It will take...
01:44:52Blue_Dude...about 15 minutes of your time and lots of experience with that target.
01:45:03kugelif you do it, you usually do it to keep the crowd quiet, not to seriously add this enhancement to other targets
01:46:03LloreanI just don't see why this is so important that it can't wait until volunteers are found for the remaining targets with a free key
01:46:26Blue_DudeI am trying very hard to *do no harm* above all else. I can't do that and simultaneously gain experience on dozens of targets.
01:46:46LloreanYou could just enable it and expect it to be fixed if it does harm
01:46:49LloreanIt's a devel build
01:47:16HillshumIsn't that what he's been asking to do?
01:47:38LloreanHe wants to only enable it for two targets and leave everyone else out until *hopefully* someone else does the work for him
01:48:01LloreanI just don't like the idea of creating an artificial disparity between offered features.
01:48:11LloreanJust because there's a rush to commit it rather than waiting for the patch to be complete.
01:48:14funmanif nobody uses this feature, then it can be left as that. If someone does, he can add the keymap
01:48:32S_a_i_n_tHmmm... Yes, I think it should wait until there are keymaps for all the targets. It just makes sense.
01:49:50S_a_i_n_tIt seems like, "No, sorry. You can't have this feature because I haven't bothered implementing it for the targets I don't have, so don't care about"
01:49:58Blue_DudeI've gone as far as I can go with the knowledge I have. Asking me to recruit volunteers to fill out the remaining (very minimal) target specific code for all targets is asking too much.
01:50:09CIA-5New commit by 03funman (r25397): pitch_detector: fix builds on monochrome targets, remove an unused variable ...
01:50:14S_a_i_n_tNo its not.
01:50:27LloreanBlue_Dude: Have you even *tried* looking at the keymaps and other sims and making a best guess?
01:50:32Blue_DudeS-a_i_n_t: have you seen the patch?
01:50:48Blue_DudeLlorean: I spent a good part of this afternoon doing exactly that.
01:50:54LloreanSo what did you think of the H100?
01:51:00LloreanWhy didn't you enable it for that one?
01:51:16Hillshumoverall, getting others who have more experience with given targets seems a better use of time
01:51:24S_a_i_n_tWhat does seeing the patch have to do with it, my point was merely "asking on IRC/Forums for help with keymaps is not "asking too much"
01:51:41S_a_i_n_tIts common sense.
01:51:46Blue_DudeLlorean: I don't have one, have never held one, and don't know if I I could make matters worse by fooling around with its code.
01:52:00LloreanBlue_Dude: So if you're going to reject them all because you don't have them, why did you even look?
01:52:06S_a_i_n_tThats what SIMs are for ;)
01:52:26LloreanThe sims are excellent for telling if a target is technically capable
01:52:33Blue_DudeS_a_i_n_t: If you'd seen the patch, you'd realize how little work is required to enable it for a given target. I am not lazy, I just lack target specific knowledge.
01:52:44S_a_i_n_tThats what SIMs are for ;)
01:53:02linuxstbBlue_Dude: Can I just check I'm understanding the feature? Firstly, there's a "file browser" hotkey, which works in the file/database browser (mapped to short-rec)? Then you've added a second "WPS hotkey" which works in the WPS (also short-rec) ? Each hotkey can be mapped to one of about 3 or 4 actions, as defined in hotkey_assignment hotkey_items[] ?
01:53:05HillshumBlue_Dude: How about taking a guess based on the sim, and if somebody who has actually used the target has a better idea, you use that?
01:53:38Blue_DudeLlorean: to see if there was a common pattern. Some are internally consistent and I wouldn't have to do any remapping, but most need a harder look at the keymap scheme.
01:53:53LloreanBlue_Dude: So which category was the h100 series in?
01:54:21Blue_DudeLlorean: I looked at more than 2 dozen files. I don't remember.
01:54:43LloreanSo for the ones where you wouldn't have to do any remapping, why didn't you try enabling it for users to give feedback on?
01:55:01LloreanRather than hoping someone with the target will compile it, enable the feature for users and find out if it works - we don't even necessarily have regular devs with all targets these days
01:55:24LloreanGet it working in the sim, get it out to users, fix problems if it has them.
01:55:33LloreanRelease builds are where things need to be extra safe.
01:56:03LloreanBut an artificial disparity like this is just going to last for who knows how long. If it were going to take a short time to be fixed, you would be able to wait on the commit until it's fixed.
01:56:04Blue_Dudelinuxstb: Pretty much. The assignment struct can be very easily expended too. Most context menu items aren't very useful for quick access or they're *too* context dependent.
01:57:34 Join komputes [0] (~komputes@ubuntu/member/komputes)
01:57:52Blue_DudeOK, how about this. I'll commit the thing without *any* targets being active. I'll comment out the defines that enable the build. In the meantime, we can work on the keymaps that need help and we'll switch them all on at a later date.
01:58:06LloreanI wouldn't object to that
01:58:09Blue_DudeIf we wait for 100% compliance, it'll never happen.
01:58:15LloreanI'd rather see it enabled everywhere that there's a record button
01:58:39LloreanBut the main problem I have is having targets that do different things just because it was enabled before it was wholly finished
01:58:48Blue_DudeRecord was just a free key already being used in a similar role. Nothing magic there.
01:59:03LloreanBut there's several targets that also have record as a free key
01:59:14LloreanAnd, in fact, I think those are the only targets that really *have* an obvious free key to use.
01:59:27Blue_DudeIf the HOTKEY define is commented out, the code will not compile.
01:59:34 Quit funman (Quit: free(random());)
02:00:08HillshumIt will do so silently right?
02:00:38Blue_DudeThat is, the code will remain in svn, but it'll seem to the compiler as a bunch of #if 0 lines.
02:00:49LloreanAs I said, I don't really object to that
02:00:52LloreanI'm sorry to be such a hassle
02:01:02LloreanI just really don't like the idea of artificial disparity unless it's actually necessary
02:01:07LloreanIn which case, I guess it's not artificial
02:01:30Blue_DudeAnd if someone wants to work on a keymap, all he has to do is uncomment the line in the config file and work on the target keymap.
02:01:59 Quit User67703 (Ping timeout: 240 seconds)
02:02:10S_a_i_n_tHmmm, any WPS guru know if there is a conditional for <time is set to 12 hour format|time is set to 24 hour format>? I'd rather not "force" the time to be displayed in one way or another in this WPS.
02:02:25kugelfor that he's got to look at the config file to even notice the hotkey feature :)
02:02:40kugelS_a_i_n_t: %St at the very least
02:02:58S_a_i_n_tAHA! thanks kugel.
02:03:05kugelCustomWPS should list a separate tag if there's one
02:03:15S_a_i_n_tIt needs to match the setting as it appears in the .cfg right?
02:03:28S_a_i_n_t*for %St I mean.
02:04:12Blue_Dudekugel: well, the ifdef's will still be all over the code, but it will compile as though it doesn't exist.
02:04:46linuxstbBlue_Dude: My view is that this is a useful feature, and one which we should try and get working on as many targets as possible. My only worry is that this _may_ need more than simply adding keymaps - i.e. the feature may need implementing differently. (although I don't know how...)
02:04:56 Join RoronoaZoro [0] (~vijayss@
02:05:06kugel#ifdefs aren't that bad, except in paramter lists of functions
02:05:22Blue_Dudekugel: meaning someone might notice that *something* is there, it won't be active.
02:05:49kugelit makes the code a pain to work with, and we don't want that
02:05:52linuxstbBlue_Dude: And I'm not sure what the point is to commit it with it not enabled...
02:05:59Blue_Dudelinuxstb: worst case: different keypresses for different menus.
02:06:41Blue_Dudekugel: hard to argue with that, but we've already got conditional builds all over the place.
02:06:46*linuxstb wonders about all the targets where "select" and "right" do the same thing in the file browser
02:07:05S_a_i_n_tDammit...%St doesn't work that way.
02:07:16kugelBlue_Dude: did you realize that my only concern is the #ifdef in the paramter list, not the otehr #ifdefs?
02:07:18S_a_i_n_tAnd there doesn't * appear* to be another way to do it.
02:08:04Blue_Dudelinuxstb: I'm not 100% sure of the point either, other than making sure that it's more widely available before switching it on.
02:08:07kugel%?St|<config string>|<true|false> doesn't work?
02:08:18Blue_Dudekugel: you mean in feature.txt?
02:08:33kugelBlue_Dude: I mean the #ifdefs linuxstb mentioned
02:09:15Blue_DudeOh, OK. That's about what? 5 lines? I guess I can fix them but there's no go way to make them pretty.
02:09:38kugelwhy does "error accessing the playlist control file" prevent playback to be started?
02:09:41S_a_i_n_tkugel: it can't be true/ its 12hour/24hour
02:10:08kugelBlue_Dude: I don't understand
02:10:23kugelwhat about to simply remove the ifdef there and pass the parameter for all targets?
02:10:36Blue_Dudekugel: Let me guess. You're offended that wps.c isn't nice looking. :)
02:11:03kugelthat too
02:11:13Blue_DudeHah, knew it. :)
02:11:13kugelbut I'm generally against that style
02:11:55Blue_DudeI don't have a problem with voiding the parameter. It will cost every target a few bytes but the code will be cleaner.
02:12:15S_a_i_n_tAs far as I can tell, %St only returns the value of the appropriate setting, not whether its true/false etc.
02:13:16kugelBlue_Dude: I would prefer that
02:13:18linuxstbkugel: And you object to IF_HOTKEY(,hotkey) ?
02:13:22S_a_i_n_tHmmm..actually, it *may* work.
02:13:29S_a_i_n_tI'll give at a try at least.
02:13:49kugelthat's no better in my eyes
02:14:04kugelall places where we have that only confuse me
02:14:19linuxstbIt's far better in my eyes.
02:14:32linuxstbSo three people, three different preferences...
02:15:26Blue_DudeWell, those specific function calls are all in switch/case instances, where the case is invalid without the HOTKEY define. It will still need some #ifdef's, just not within a function call.
02:15:46kugelan additional annoyance is that this is hard to parse for IDEs, leading to wrong call tips being shown and frustration
02:16:30kugelBlue_Dude: you #ifdef'd the action as well..
02:17:09Blue_DudeThey're case fall-throughs. That's even uglier.
02:17:20kugelwhich is uncommon practise as well
02:17:26saratogai dislike the idea of people who don't have a target changing its keymap . . .
02:17:46Blue_Dudesaratoga: thank you. That's what I've been thinking all afternoon.
02:18:02kugelI didn't do that with my playlist shortcut patch, probably halving the #ifdef count at very very low costs
02:19:12Blue_Dudekugel: I didn't put in the switch/case instances, just added onto them.
02:19:51S_a_i_n_t*duh*....I found the 12/24 hour conditional...
02:20:38kugelBlue_Dude: you don't necessarily need to #ifdef the action (ACTION_HOTKEY)
02:21:52kugelit's an enum, so free, that'd cleanup 1 #ifdef in wps.c
02:22:12Blue_Dudekugel: I fonld something interesting in the bloat-o-meter. Just changing the key assignment - and nothing else - cost several dozen bytes for no apparent reason. I ifdef'ed the action item to avoid reassignments.
02:22:41Blue_DudeBut maybe it was just a weird glitch.
02:22:48kugelwhat key assignment?
02:24:45Blue_DudeI originally had ACTION_HOTKEY a lot further up the list in the enum. During the build of bookmark.c, when ACTION_HOTKEY replaced BMARK_DELETE (or whatever), the bloat count went up for no reason whatsoever. But moving them adjacent and defining HOTKEY out fixed it. Dunno why.
02:25:59kugelI assume it's because it makes it harder for gcc to setup the jump table
02:26:47Blue_DudeThat's way beyond my area. Your guess is probably much better than mine.
02:27:15kugelthe cases in switches should be as least out-of-order as possible, since the value is multiplied with a constant to get the jump location
02:27:17Blue_DudeBut yeah, I can take the conditionals easily enough.
02:27:37Blue_Dude...conditionals out...
02:27:39kugelbad order means more complexity in fixing up the the preparation
02:27:53Blue_DudeHm. I didn't know that.
02:28:50Blue_DudeSo given that, still think it's worth it to clean up the ifdefs?
02:28:59kugelyou could for example see what happens if you make the case in wps.c uncontitionally, and moving ACTION_HOTKEY near to ACTION_WPS_CONTEXT
02:30:02Blue_DudeWhich means moving it away from BMS_DELETE. Pick your poison.
02:30:21Blue_DudeThe case is much simpler in bookmark though.
02:30:48kugelI mean just for experimenting if my guess is correct
02:31:58Blue_DudeOh wow, look at my wrist, got to go. :)
02:32:44Blue_DudeAll right, I'll fool with it some more. Experimenting with builds in Cygwin really blows though.
02:33:46kugelso get a real linux ;)
02:33:57kugelevery VM is faster than cygwin
02:35:40 Quit Strife89 (Read error: Connection reset by peer)
02:35:57Blue_DudeNo kidding.
02:36:04 Join Strife89 [0] (
02:36:25kugelit's true
02:36:27 Quit kugel (Remote host closed the connection)
02:40:42CIA-5New commit by 03moos (r25398): Oops, forgot to update the header with the added name for the nyorsk translation too.
02:54:36saratogaBlue_Due: FWIW if you just want fast builds on a modern windows machine, virtualbox supports an unlimited number of CPU cores free and you can get preconfigured ubuntu images on line that only take a few minutes apt-get gcc and then run
02:59:42***Saving seen data "./dancer.seen"
03:05:10Blue_Dudesaratoga: I might on another machine, but this one is work related and I can't do it. Thanks though.
03:05:57RoronoaZoroI would like to continue the discussion that was going on 26 on the irc....I have been looking for various decoders and papers related to decoding ....i have collected various papers, sites for my reference
03:07:42RoronoaZoroI would like to implement the MPEG-4 ASP codec in rockbox i am looking at 2 decoders particularly ffmpeg and xvid.
03:08:34 Quit Hillshum (Ping timeout: 246 seconds)
03:12:47 Join LambdaCalculus37 [0] (~rmenes@rockbox/staff/LambdaCalculus37)
03:15:05 Quit RoronoaZoro (Remote host closed the connection)
03:15:26Blue_DudeAll right, I think I have a solution for several problems. I changed the keymap to respond to several different key defines. In essence, each menu can have its own hotkey button. In a perfect world, they should be the same for each menu, but now they don't *have* to be. It also makes the code a lot cleaner this way.
03:16:13 Join Ag3nT [0] (~a464aa04@gateway/web/freenode/x-mffohsewcopjowwd)
03:19:05 Part Ag3nT
03:19:17 Join Ag3nT [0] (~a464aa04@gateway/web/freenode/x-mffohsewcopjowwd)
03:25:52 Part Ag3nT
03:30:16 Join planetbeing_ [0] (
03:33:07 Quit Adubb (Read error: Connection reset by peer)
03:33:18 Join Adubb [0] (~aldubuc@
03:36:34 Quit Blue_Dude (Quit: ChatZilla 0.9.86 [Firefox 3.6.2/20100316074819])
03:39:44 Join Hillshum [0] (
03:41:02 Quit CaptainKewl (Quit: ( :: NoNameScript 4.22 :: ))
03:44:23 Join CaptainKewl [0] (
03:56:30 Quit Topy44 (Quit: No Ping reply in 180 seconds.)
03:57:06 Join Topy44 [0] (
04:03:12 Quit TheSeven (Disconnected by services)
04:03:26 Join The_Seven [0] (~theseven@rockbox/developer/TheSeven)
04:03:36 Nick The_Seven is now known as TheSeven (~theseven@rockbox/developer/TheSeven)
04:04:13 Quit Tomis (Ping timeout: 264 seconds)
04:04:28 Quit adnyxo (Remote host closed the connection)
04:06:45 Quit Darkknight512 (Quit: ChatZilla 0.9.86 [Firefox 3.5.8/20100202165920])
04:07:06 Join Darkknight512 [0] (
04:08:09 Join adnyxo [0] (
04:10:39 Quit adnyxo (Remote host closed the connection)
04:12:42 Quit slck (Ping timeout: 276 seconds)
04:20:04 Join BHSPitMonkey [0] (~stephen@unaffiliated/bhspitmonkey)
04:22:14 Quit LambdaCalculus37 (Ping timeout: 258 seconds)
04:23:13 Join DV [0] (~DV@
04:24:55 Quit DV (Read error: Connection reset by peer)
04:27:13 Join LambdaCalculus37 [0] (~rmenes@rockbox/staff/LambdaCalculus37)
04:36:08 Join Barahir_ [0] (
04:36:35 Join Rob2223 [0] (
04:36:50 Join Tomis [0] (~Tomis@
04:37:35 Quit CaptainKewl (Read error: Connection reset by peer)
04:38:13 Quit Tomis (Read error: Connection reset by peer)
04:38:17 Join Tomis [0] (~Tomis@
04:38:50 Quit komputes (Ping timeout: 245 seconds)
04:40:02 Quit Barahir (Ping timeout: 276 seconds)
04:40:22 Quit CGL (Ping timeout: 246 seconds)
04:40:32 Quit Rob2222 (Ping timeout: 265 seconds)
04:41:55 Join CaptainKewl [0] (
04:47:37 Join planetbeing__ [0] (
04:50:48 Quit planetbeing_ (Ping timeout: 248 seconds)
04:52:26 Quit panni_ (Read error: Connection reset by peer)
04:59:43***Saving seen data "./dancer.seen"
05:03:30 Quit Darkknight512 (Remote host closed the connection)
05:05:05 Quit tmzt_ (Ping timeout: 276 seconds)
05:06:32 Join slck [0] (Venci@Slackware.SlackPix.Com)
05:16:00 Quit Tomis (Quit: Tomis)
05:18:59 Quit anewuser ()
05:27:19 Join tmzt__ [0] (
05:29:12 Quit Strife89 (Quit: Bed.)
05:31:47 Quit LambdaCalculus37 (Quit: Fwump)
05:38:06 Quit Horscht (Quit: Verlassend)
05:46:49 Quit n17ikh ()
05:47:38 Join planetbeing_ [0] (
05:50:08 Join n17ikh [0] (
05:50:51 Quit planetbeing__ (Ping timeout: 276 seconds)
06:04:03 Quit n17ikh (Remote host closed the connection)
06:18:21 Join frogyman [0] (
06:27:55 Join n17ikh [0] (
06:41:05 Join CGL [0] (~CGL@
06:48:00 Quit YPSY (Ping timeout: 264 seconds)
06:50:42 Join Ypsy [0] (
06:51:00 Nick Ypsy is now known as YPSY (
06:59:47***Saving seen data "./dancer.seen"
07:15:56pixelmaLlorean: I don't think the sim is a very good way to test keymap (changes). You can see the current mapping sure and maybe find conflicts but - you don't see mechanical or electrical restrictions the real target has and don't have the same feel because the actual placement of buttons differ (especially things like side buttons and scrollwheels or touch pads)
07:16:50Lloreanpixelma: But since this button isn't used in a combo, and is a single button to perform a single function, the sim should be fine for it.
07:17:35LloreanAt least for preliminary maps.
07:18:13pixelmathe "every target with a rec button" is way too simple - the Iaudios have a "Rec/A-B" button too but don't really have free buttons because of allowing no combos
07:19:11LloreanEvery target with a free rec short press, though.
07:19:37LloreanMy point is that we should at least try for preliminary button maps so that *users* can test it.
07:20:47 Join BlehTM [0] (
07:21:18pixelmaLlorean: as an example, without trying though: I'm a bit sceptical about using a short "Rec" on the c200 because it's easily pressed accidentally. I remember mapping "Go to rec screen" from the WPS to "Long Rec" om purpose, because it was quite easy to accidentally end up there (and for consistency)
07:22:47LloreanIsn't the c200 one of the two targets he's already mapping it on?
07:22:56LloreanI thought he said he was mapping it on c200 and e200.
07:23:12LloreanAlso, I believe if you don't explicitly assign it, the button will continue to do what it did before.
07:23:28pixelmayes, that's why I state my little concern here
07:23:37LloreanAnd since assignment includes a confirmation screen, even an accidental press won't be significant unless you are one one of the places you can change its functions *and* choose to confirm the assignment
07:23:49LloreanSo it seems pretty safe from accidental presses.
07:24:02*ThomasAH gets an error when building svn for clip+: make: *** No rule to make target `/home/thomas/rockbox/rockbox.svn/build/pluginbitmaps/pitch_notes.h', needed by `/home/thomas/rockbox/rockbox.svn/build/apps/plugins/pitch_detector.o'. Stop.
07:24:20Lloreanpixelma: The side button on e200 is about as easy to accidentally press as the one on c200 in my opinion
07:24:37pixelmaI am still wondering if there really is a target with a free button on all screens
07:25:07LloreanAren't the only screens that matter filetree and WPS though?
07:27:11pixelmadepends on what you can (or want to) map to it, no. And as I remember, the WPS has one of the most crowded mappings - pitch screen, ID3 viewer, context menu, menu, file browser, the playlist viewer shortcut added relatively recently
07:27:53 Quit frogyman (Quit: CGI:IRC)
07:30:40pixelmait's just my doubts about the patch in general I guess
07:32:25pixelmaThomasAH: did you try a "make clean"?
07:33:04pixelmaah no, reds in the build table too
07:33:40ThomasAHpixelma: I even did "rm -rf build" ... but I found the problem, && defined(HAVE_LCD_COLOR) in apps/plugins/bitmaps/native/SOURCES line 847 has to be removed
07:33:53Lloreanpixelma: Well, I have doubts about the patch too.
07:34:00 Quit CaptainKewl (Remote host closed the connection)
07:34:11 Quit Minataku (Ping timeout: 252 seconds)
07:34:13LloreanBut if it has to go in, I'd like to see it have a keymap on every target that has the necessary free mappings.
07:35:54ThomasAHpixelma: I don't have commit access, can you chage this? funman forgot to remove this again after r25397
07:36:18ThomasAHpixelma: (he introduced it temporarily in r25387)
07:37:24pixelmahow long do build rounds take these days?
07:48:27pixelmaI wonder why the completely same line as in plugins/SOURCES need to be there. Would bitmaps be included if the plugin itself doesn't even build?
07:48:58pixelmaor isn't built for a specific target
07:49:15pixelmathere goes my plan for a "quick fix" :\
07:50:11 Join Buschel [0] (
07:51:05pixelmaThomasAH: unfortunately I have to go in a bit, so I have to leave it to someone else - I'd be interested in an answer to my question for this fix (to do it "correctly")
07:55:36 Part BlehTM
07:59:56 Join DV [0] (~DV@
08:04:01 Join DV_ [0] (~DV@
08:04:17Buschelstripwax: you mentioned that your ipod hung up due to "recent PP ide power" changes. you are talking of r25255 and r25273? are your sure this change is the reason?
08:05:01 Join LinusN [0] (~linus@rockbox/developer/LinusN)
08:05:56 Quit DV (Ping timeout: 260 seconds)
08:20:41 Join Makuseru [0] (
08:23:44 Quit Buschel (Ping timeout: 268 seconds)
08:29:01 Join ender` [0] (
08:30:14 Quit TheSeven (Ping timeout: 258 seconds)
08:30:41 Join Minataku [0] (~Ed@unaffiliated/payphoneed)
08:32:03 Quit Makuseru (Quit: CGI:IRC (EOF))
08:33:18ThomasAHpixelma: sorry, I did not notice that the "build round" question was directed to me ... it takes few minutes on an 1.8GHz Core 2
08:35:05 Join fragilematter [0] (~user@
08:35:36 Quit ivortheengine (Ping timeout: 258 seconds)
08:39:38amiconnTorne: firmware/ isn't supposed to access settings. Such settings have to be set through a function in the firmware module, called from the settings code
08:49:41 Join flydutch [0] (
08:52:42 Join mitk [0] (~mitk@
08:55:05pixelmaThomasAH: no, my question was "why do we need the same stuff in the SOURCES files inside plugin/bitmaps that's there in plugin/SOURCES". If the latter excludes the whole plugin from being built then the build shouldn't look into the former anyways, should it?
08:56:35pixelmathe whole MIC_IN etc. shouldn't be necessary in the bitmaps SOURCES if I understand correctly and should be cleaned up
08:57:20pixelmathis would also have helped avoiding this error in first place
08:58:14amiconnThe bitmaps are built independently of the plugins. If you build the plugins but not the respective bitmaps, it will cause an error at link time
08:58:34amiconnThe other way round won't cause an error, but the build will take longer than expected
08:58:37 Join Bagder [0] (~daniel@rockbox/developer/bagder)
08:59:49***Saving seen data "./dancer.seen"
09:02:10pixelmaI remember that now, but does it make sense?
09:14:18amiconnRight now we're building bitmaps mainly based on lcd resolution and colour depth. That means that if a plugin could work on the target's lcd but isn't supported otherwise (for various reasons, e.g. lack of buttons, too slow cpu, ...), the bitmaps are still built
09:14:49 Quit arbingordon (Quit: `)
09:15:07amiconnThe only solution with the current system is to duplicate the conditions
09:16:00pixelmano, for e.g. the pitch_detector plugin they weren't because plugin/bitmaps/native/SOURCES contained the same lines as plugin/SOURCES - which is why the former needs fixing too now
09:17:16pixelmamy point was that duplicating conditions seems to be error prone
09:17:59 Quit BHSPitMonkey (Ping timeout: 245 seconds)
09:19:49pixelmaand if things happened in a different order might not be necessary. I remember now seeing the bitmaps being built before the plugin and I realise that this is an advantage of the current solution though (that it's independent and probably faster if all the bitmaps can be built in one go)
09:20:24 Quit fragilematter (Remote host closed the connection)
09:20:36amiconnIt may be error prone, but it's the only solution with the current system that avoids building unneeded bitmaps
09:20:38 Join fragilematter [0] (~user@
09:21:15amiconnIt works that way since ...bitmaps/SOURCES
09:22:27amiconnThe big make system change didn't touch that. It might have changed the order because of better dependency generation, but the dependencies only decide the order in which things are built (and what needs to be rebuilt if something changes), not what gets built at all
09:26:00 Quit saratoga (Quit: Page closed)
09:26:20pixelmayes, then I ask whether it would make sense to change the system to - I may put it non-technically - look into the plugin first to see which bitmaps are needed
09:27:05 Join efyx [0] (
09:28:55pixelmaor just live with the fact that they are built and not used later
09:30:07 Quit CGL (Quit: Saliendo)
09:30:29ThomasAHpixelma: what about making the plugin builds check if the bitmap is available?
09:31:01pixelmathat's what it does right now
09:31:15ThomasAHpixelma: (though this will move the knowledge about which plugin required which feature to the wrong place)
09:33:11 Join BHSPitMonkey [0] (~stephen@unaffiliated/bhspitmonkey)
09:35:49pixelmaThomasAH: sorry, I don't understand (seems there is a misunderstanding maybe already about your first thought)
09:37:43ThomasAHpixelma: currently the bitmaps say "have this, have that, have color -> build bitmap" and the plugin says "have this, have that, have color -> build plugin". My suggestion was to make plugins check "have bitmap -> build plugin". But then I thought the knowledge of "have this, have that" should stay inside the plugin, only "have color" might be bitmap related.
09:39:40pixelmathat sounds weird - my suggestion was build plugin and see which bitmaps are needed
09:40:50 Join pamaury [0] (~pamaury@rockbox/developer/pamaury)
09:41:04ThomasAHpixelma: yes, and that's the better way. Though for now it might be more important to fix the regression
09:42:15pixelmaindeed. I'm at work now though (as you can see ;) ) and don't have svn or the Rockbox source around
09:46:32 Quit DV_ (Quit: Leaving)
09:47:14 Join petur [0] (~petur@rockbox/developer/petur)
09:47:16ThomasAHpixelma: I'm at work now, too (as you can't see :))
09:49:26fragilematterhey peeps
09:50:44fragilematterI've finally tracked down that fm problem, basically it's a simple (and honest) logic mistake
09:57:52ThomasAHfragilematter: preset mode vs. scan mode oder something different?
09:58:31fragilematterany mode, it's an initialization problem at start-up
09:58:47fragilematterI'm now adding a flyspray task describing it
10:03:51 Quit planetbeing_ (Quit: planetbeing_)
10:31:52 Join Zagor [0] (~bjst@rockbox/developer/Zagor)
10:45:03 Quit BHSPitMonkey (Ping timeout: 260 seconds)
10:49:39 Join TheSeven [0] (~theseven@rockbox/developer/TheSeven)
10:50:11 Join wodz [0] (
10:50:49 Join Barahir [0] (
10:52:03 Quit fragilematter (Quit: I'm off)
10:54:11 Quit Barahir_ (Ping timeout: 240 seconds)
10:59:50***Saving seen data "./dancer.seen"
11:08:08 Join archivator [0] (~archivato@
11:09:30 Join byondo [0] (
11:14:32 Quit byondo (Ping timeout: 276 seconds)
11:16:02wodzPlease take a look at FS #11153
11:16:19 Join byondo [0] (
11:20:38 Quit byondo (Ping timeout: 248 seconds)
11:24:36 Join byondo [0] (
11:26:09 Join blairb_ [0] (
11:28:12 Quit wodz (Quit: Leaving)
11:28:48 Part byondo
11:30:00 Quit blairb (Ping timeout: 258 seconds)
11:41:43archivatorI have a feeling that if I dive any deeper into rbutil, I'll find out it's just one big memory leak and my computer just happens to have the right data at the right addresses :(
11:47:23 Join Buschel [0] (
11:50:06 Join byondo [0] (
11:50:19 Quit byondo (Client Quit)
11:57:18 Join dfkt [0] (dfkt@unaffiliated/dfkt)
11:58:01 Join fragilematter [0] (~user@
12:00:48 Quit TheSeven (Quit: ChatZilla 0.9.86 [Firefox 3.6.2/20100316074819])
12:04:46 Nick fxb__ is now known as fxb (
12:07:42 Join kugel [0] (~kugel@rockbox/developer/kugel)
12:10:10 Join fragilematter_ [0] (~fragilema@
12:11:42 Quit fragilematter (Ping timeout: 276 seconds)
12:12:23 Nick fragilematter_ is now known as fragilematter (~fragilema@
12:13:02 Quit Buschel (Ping timeout: 276 seconds)
12:15:08 Join Guest62701F [0] (
12:15:14 Quit Status (Read error: Connection reset by peer)
12:15:54 Quit fragilematter (Quit: leaving)
12:16:56 Quit blairb_ (Quit: Leaving)
12:21:59 Join ivortheengine [0] (~ivortheen@
12:30:28archivatorbluebroth3r: check out fs#11155 when you have the chance :)
12:36:29 Join byondo [0] (
12:41:07 Quit byondo (Ping timeout: 264 seconds)
12:44:46 Join byondo [0] (
12:49:16 Quit byondo (Ping timeout: 260 seconds)
12:53:04 Join byondo [0] (
12:57:49 Quit byondo (Ping timeout: 265 seconds)
12:59:52***Saving seen data "./dancer.seen"
13:01:21 Join byondo [0] (
13:02:30 Join junkY_San [0] (
13:02:36Bagderis there any specific plans for friday the 4th for devcon ?
13:03:05Bagderlike at what time does it make sense to be there?
13:03:55junkY_Sanhey, i got a sansa clip v2 and would like to use rockbox. is it possible to use the simulator to change and save my configs and write it to the player via usb?
13:04:01Bagderthere's one flight landing in BRU at 14:05 that seems fine
13:04:29BagderjunkY_San: yes, the config is just a text file
13:04:39linuxstbjunkY_San: configs are saved in the "config.cfg" file inside the .rockbox folder. And yes, there shouldn't be a problem copying it between the simulator and the real device.
13:07:11junkY_Sani have searched the whole wiki but maybe i'm blind. is it planned to enable write support for the sansaAMS platform?
13:08:50linuxstbI would assume it's just something that isn't done yet. It's not simply a matter of flicking a switch.
13:09:42linuxstbBut don't lots of "SansaAMS" targets already have write support? It's just missing on the very new ports?
13:09:58junkY_Sani haven't tried yet to be honest
13:10:23junkY_Sani saw this page: it says read-only for clipv2
13:10:53Bagderyes but clipv2 is not the only sansaams target
13:11:06Bagderthat table is a bit cruptic though
13:11:29Bagder"Internal/External storage" somehow implies that it is about reading and writing to it
13:11:41BagderI think
13:12:10linuxstbBagder: It looks to me as if it should be two columns...
13:12:26Bagderyes, that would make it a lot more easy to grasp
13:13:06*linuxstb clicks edit and wonders again what the point of all that TinyMCE javascript is
13:18:05 Quit ivortheengine (Ping timeout: 258 seconds)
13:18:29 Quit byondo (Ping timeout: 248 seconds)
13:20:17 Join ivortheengine [0] (~ivortheen@
13:21:01archivatorlinuxstb: re flite license: I got a response: "That's usual BSD license. You just need to have a changelog in source files/top folder, there is no need to have comments in the code."
13:21:22archivatorNot too bad, considering that I can just use the git log :)
13:22:31 Join byondo [0] (
13:27:15 Quit byondo (Ping timeout: 252 seconds)
13:30:49 Join byondo [0] (
13:35:42 Quit byondo (Ping timeout: 264 seconds)
13:39:06 Join byondo [0] (
13:43:45 Quit byondo (Ping timeout: 264 seconds)
13:43:57 Join webguest83 [0] (
13:44:07 Join webguest99 [0] (
13:45:06 Join Schmogel [0] (
13:45:06 Quit webguest99 (Client Quit)
13:47:24 Join byondo [0] (
13:48:58 Quit webguest83 (Quit: CGI:IRC (Ping timeout))
13:55:03 Quit byondo (Ping timeout: 276 seconds)
13:58:06 Join byondo [0] (
14:02:57 Quit byondo (Ping timeout: 264 seconds)
14:05:21 Part LinusN
14:06:26 Join byondo [0] (
14:09:27 Join wodz [0] (
14:10:14 Join watto [0] (~watto@
14:10:42 Quit byondo (Ping timeout: 252 seconds)
14:14:45 Join byondo [0] (
14:16:52 Join funman [0] (~fun@rockbox/developer/funman)
14:19:05 Quit byondo (Ping timeout: 258 seconds)
14:23:00 Join byondo [0] (
14:26:13CIA-5New commit by 03funman (r25399): Build pitch_detector bitmaps for all targets, even those without color ...
14:27:43 Quit byondo (Ping timeout: 260 seconds)
14:30:10 Quit kugel (Ping timeout: 240 seconds)
14:30:55 Quit Bagder (Quit: It is time to say moo)
14:31:06 Join geertvdijk [0] (
14:31:18 Join byondo [0] (
14:31:57funmanZagor: I would like to get a Clipv2 from the rockbox fund, how would i do that? I could send an email to the dev mailing list first to be sure no people disagree
14:33:27Zagorfunman: simply buy it and tell me the cost. I already have your bank details, don't I?
14:33:56funmanI don't think so, but I'll send them to you then
14:40:45 Quit byondo (Ping timeout: 265 seconds)
14:41:15 Join byondo [0] (
14:41:32byondohey funman!
14:41:33dfktfunman, i just read that "Development hardware is available for the Clip v2 to interested developers."
14:42:56funmandfkt: yes, i checked that yesterday and this was added by mc2739. I don't know if he has some devices or if he proposed to buy players for interested devs
14:42:59byondoI'm here just to say thanks for your great work
14:43:13byondoand other developers one too :)
14:43:34Tornefunman: iirc he just has several..
14:44:04funmanmc2739: ping, do you have some spare Clipv2 ?
14:44:22funmanbyondo: thanks ^^
14:46:04byondoand I donated 20 bucks for rockbox project the other it to buy your clip v2 ;)
14:48:11byondoonly one thing, why is FM untested on clipv2 status
14:48:22byondoit works, indeed
14:48:46funmanbyondo: because i have no clipv2 to test on :o
14:49:12 Join xsteadfastx [0] (~spectrum@
14:50:07funmandoes microphone recordign work as well ? just go into recording screen and say something : you can hear the feedback in the headphones but not record on disk (no write support)
14:50:31 Join StickleBricks [0] (
14:50:47byondoyes! everything works
14:51:34byondoFM radio works, but not the sound settings with it
14:51:53 Nick StickleBricks is now known as TillW (
14:52:10funmando you have another sansa except the clipv2 ?
14:52:59byondoonly a 8GB clipv2
14:53:53mitkfunman: I have clip+. What can I check?
14:54:36funmanFM sound settings
14:54:43funmanit seems to work on Clipv1, i'm checking on the +
14:54:49 Quit wodz (Quit: Leaving)
14:55:04funmanbass works
14:55:33TillWhmm, i'm thinking of submitting an application for GSOC for the theme editor project
14:55:38funmantrebble too, but those only work in the positive range ? going below 0 doesn't change the sound
14:56:03xsteadfastxdoes the clip+ work? it looks like a neat little device
14:56:23funmanxsteadfastx: some features don't work yet (like writing to the storage)
14:56:34TorneTillW: hopefully someone who has a clue about that will be around ;)
14:56:38mitkfunman: Do I need new bootloader or 25318 will be enough?
14:56:50byondobass and treble controls just modify the volume
14:56:53funmanmitk: just a current build, there should be no need to update the bootloader
14:57:02funmanbyondo: what is sound settings then?
14:57:02xsteadfastxfunman, but that looks already great. maybe i will buy one to test it
14:57:21TillWTorne: yeah, i'm hoping that someone might be around too
14:57:56TillWrockbox is one of the few projects on the list that i actually care about, and also has something i could do on it
14:58:40Tornebluebroth3r: ping, got a gsoc student for you ;)
14:58:58Tornedomonoky isn't here right now, but is the other mentor listed for that project
14:59:25funmanare those 2 mentors a hint that the theme editor would be written in QT4 toolkit ? :)
14:59:48gevaertsWell, that depends on the proposal we get :)
14:59:55***Saving seen data "./dancer.seen"
15:00:07byondofunman, correcting myself, rising bass and treble rises volume, goin down it does nothing
15:00:57funmanbyondo: i see the same thing on Clipv1
15:01:01funman(and Clip+)
15:01:43funmanI'm not sure how it's supposed to work since the signal goes directly from the FM chip to the headphones, rockbox can't see it (unless it's recording)
15:02:00archivatorAnyone care to test FS #11155? Festival should be a bit less temperamental now :)
15:02:19funmansomeone with a e200v1 or c200v1 with FM ?
15:02:28funmanor a philips sa9200
15:03:23byondoand graphic eq. simply doesn't work
15:03:53xsteadfastxcan it be that not all e200 have a radio built in?
15:03:56byondobalance works
15:03:58gevaertsTillW: do you have some ideas about how you'd handle this project?
15:04:08dfktthe parametric eq does work on my clip+.. or am i imagining things?
15:04:49geertvdijkxsteadfastx: I can confirm, not all e200-s have a radio builtin. I have a e250 which doesn't show radio support, neither in OF or rockbox.
15:04:51funmandfkt: with FM ?
15:05:09dfktoh, sorry, misread that part
15:05:15dfktnope, didn't try FM at all
15:05:24xsteadfastxgeertvdijk, ok good. i just wondered cause there wasnt even anything on the package box about a radio
15:05:33mc2739funman: that was not added by me, it was saratoga -;rev2=140
15:05:52funmanoh right
15:06:25byondohey, I'm talking about radio sound settings, not file playing :)
15:06:50byondoin file playing every sound setting works pretty well
15:07:01TillWwell for one, i can't really find it anywhere on the site even though the project section says it exists. all i can really remember about it is when i tried to use it last (mind you a few years ago) it didn't really work well
15:07:28funmanTillW: you're talking about the existing theme editor?
15:08:18byondogoin'on with radio sound settings, channel config. and crossfeed don't work
15:09:47byondoto summarize: only volume and balance work, bass and treble do strange things
15:10:24mc2739funman: do you still need someone with e200v1 with FM?
15:10:27 Quit ender` (Ping timeout: 258 seconds)
15:11:18funmanmc2739: yep, can you check if setting trebble/bass lower than 0 have any effect ?
15:11:19gevaertsTillW: the current theme editor is unmaintained and basically abandonned. It's in the source tree, and you can use it as a base for a new editor, but you can also start from scratch
15:12:02TillWi was toying with the idea of using it as a chance to learn Python...
15:13:24TillWthough i'm not entirely sure how appropriate a goal that is. i do tend to under state my skills on the basis of not having proven them to myself yet
15:14:11 Quit jae (Quit: leaving)
15:14:18 Join jae [0] (
15:14:22 Quit jae (Client Quit)
15:14:59pixelmaI'm not sure bass or treble controls work for the radio on any software tone controls targets, should they?
15:15:03 Quit byondo (Quit: CGI:IRC (Ping timeout))
15:15:21funmanpixelma: iiuc, no they shouldn't
15:16:22mc2739funman: it seems that the treble and bass have no effect when going less than 0. Going greater than zero appears to just invrease volume for both. This was from FM->Context Menu->Sound Settings
15:17:01funmani see the same thing on my clips
15:20:01 Join byondo [0] (
15:21:15 Join jae [0] (
15:21:46gevaertsTillW: well, I guess python would be a valid option. Do you have some ideas about what sort of features you could get done during GSoC?
15:24:25byondoa curiosity, how much does pitch correction affect RB runtime?
15:26:09TillWgevaerts well, i'm not exactly sure what i could accomplish in the time. i would also be learning python in that time (decent C++ background right now) i would almost like to understate what i plan to acheive and get more done than say i want to do everything and only get half of that done. i've just never worked in an open source project (or even a software development environment) before, and don't know how sane this is sounding
15:26:29pixelmafunman: I could try on my c200 (v1) if it is still needed and some other targets with radio
15:27:01funmanbyondo: we don't try to do any "pitch correction"
15:27:34funmanpixelma: i think it's clear as3514 doesn't support bass/trebble settings on line input (FM) but I don't know how to remove the settings
15:28:53 Join adnyxo [0] (
15:29:21archivatorTillW: are you familiar with Qt? As someone with experience (albeit not much) in both Qt and PyQt4, I can assure you you'll spend more time learning Qt than Python. Just a heads up :)
15:29:48gevaertsTillW: overestimating your goals and then not reaching them is definitely not good :)
15:29:56 Join evilnick_B [0] (~0c140464@rockbox/staff/evilnick)
15:30:19byondofunman, LOL, wanted to say "pitch modification"
15:30:55TillWi'm not currently familiar with either python (well, slightly familiar with python, learning it in spare time right now) or Qt
15:32:01gevaertsAre you familiar with other crossplatform GUI libraries?
15:32:07archivatorTo be fair, no one said anything about Qt, it's just that rbutil is written in Qt and there are people here who have experience with it :)
15:32:52funmanbyondo: the feature which keeps the pitch when playing slower/faster ?
15:33:03gevaertsarchivator: if it's not C++, I'd assume that there are no real advantages for staying with Qt either
15:33:08 Join anewuser [0] (anewuser@unaffiliated/anewuser)
15:34:15byondono, just the simple "pitch" setting on wps settings, that affects both pitch and time
15:34:43TillWi'm not really familiar with any GUI libraries related to this task, all my C++ experience thus far has been terminal type stuff
15:36:16 Join ender` [0] (
15:36:40 Join CGL [0] (~CGL@
15:38:26 Join kugel [0] (~kugel@rockbox/developer/kugel)
15:39:41 Quit byondo (Quit: CGI:IRC (Ping timeout))
15:41:00 Join byondo [0] (
15:43:40byondoI was just curious to compare FS #10906 vs a simple pitch modification, maybe making it permanent
15:49:01 Quit byondo (Quit: CGI:IRC (Ping timeout))
15:50:08archivatorgevaerts: except, Qt is a well-designed and supported library :) Arguably better than wxWidgets (definitely better than wxPython, IMHO) :)
15:51:23pamauryand what about gtk ?
15:52:54funmanQt is the most cross platform friendly afaik (although it doesn't support cross compilation)
15:52:56archivatorpamaury: hard to use on windows. Well, harder than the alternatives
15:53:17pamaurytrue, but qt uses this ugly moc thing
15:54:08archivatorugly but efficient. Also, wxWidgets covers the same desktop platforms, so it's not a question of most cross-platform. Ease of use and versatility come into the equation as well
15:54:09pamaury(anyway I never used qt so perhaps the day I'll try it I love it)
15:54:28 Quit mitk (Quit: Leaving)
15:54:40pamauryugly and not necessary I would say
15:55:08TillWideally i'd like something that works in both linux and windows, as i'd prefer not to tinker and make some OS dependant code
15:55:21Tornegevaerts: (or anyone who knows usb): could you take a look at at some point and see if i'm handling usb charging reasonably?
15:55:39Torneit doesn't implement the "force" mode yet so it will only charge at full speed if there's an active host
15:55:50gevaertsTillW: let me put it this way: if you're not planning on making something that works on linux, windows, and osx, you're probably not going to be selected :)
15:56:05Tornebut it seems to work, have tested with low power hubs, and tested changing the setting while it's connected too
15:57:15Tornethe actual details of how it toggles the GPIOs on ipod is not interesting, i just want comments on the changes to the non-target-specific code ;)
15:57:33funmanTillW: you could start by having a look at existing toolkits and selecting one that fits the requirements (I also suggest looking at Qt4 but there might be others)
15:57:39TillWwell i'd like it to work on the major OSes, i just don't know how much if any code would need to be changed to make it work on all them individually
15:57:40archivatorIf the choice is between qt and wx, I would personally choose qt, if only to make future maintenance easier. As to the language of choice, that's a personal decision (though C++ is, ironically, easier to ship than Python)
15:58:49 Part archivator ("Leaving")
15:58:52 Join archivator [0] (~archivato@
15:58:53gevaertsTillW: I don't mean that you personally have to get it running on all of them, but choosing a library that doesn't work on all is definitely bad
16:00:01TillWi just don't think that C++ would be well suited to such a task personally. and i know i'd prolly be able to personally make it work on windows and linux; i couldn't do osx unless it just plain works on it, since i have no access to it (and don't want to have access to osx)
16:00:39archivatorTillW: interesting, why do you think that?
16:01:25kugelC++ is bad :p
16:01:40pamauryC++ is good :P
16:02:14*funman would like to run C++ plugins on target
16:02:49pamauryC++ is not so complicated to handle as long as you don't have exceptions, and with gcc
16:03:00TillWwell, to me, C++ is low level, and better for performance oriented tasks, or where you need to control things more tightly. something like say python is more oriented towards just getting it done. since this isn't performance oriented, and there's no details that need to be tightly controlled. i see this as purely a program to parse a WPS file, checkit (though i think there's a checkwps tool for that) display it, edit it, or create one from scratch
16:03:40 Part aCiD2
16:03:48funmanit makes perfect sense
16:03:56TillWso the program is really more oriented towards layout and file parsing than low level details of how it's doing all those functions
16:04:44TillWC++ really just seems like jamming a Merlin into a golf cart for this
16:05:16gevaertsTillW: would you have wysiwyg as an end goal (possibly out of scope for GSoC)?
16:05:25Zagor"PyQt is available for Windows, UNIX/Linux, Mac OS X and the Sharp Zaurus."
16:05:30funmankugel: "GPIOA4_cached" seems to be the selector between A0 and A4 for FM i2c sda
16:05:40Zagorsounds sensible to me
16:05:54pamauryfunman: you want to be able to have C++ plugins with rockbox ? That's clearly feasible I think, with little changes in linker scripts and loader
16:05:57kugelfunman: ah, interesting
16:06:04funmanpamaury: no i don't want :)
16:06:30Zagorpamaury: please don't mention C++ and Rockbox in the same line. it upsets the children!
16:06:35archivatorI just think PyQt4 is a bit annoying (especially with Python 2.x) but that's just my opinion. It does many things right but a few dozen more wrong :)
16:06:52pamauryhehe, for plugins it makes sense, for the rest, not really
16:07:08Zagorarchivator: I have no idea, really. I just looked it up...
16:07:18gevaertsTorne: that looks reasonable I think. We might want to extend it to retry 100mA if the host doesn't accept 500mA, but that's a local detail that can be added later I think
16:07:24TillWgevaerts: i'd kinda like wysiwyg as a goal all along. and if i get heavily into this, i could see working on it out of GSoC aswel
16:07:33 Quit n17ikh (Ping timeout: 258 seconds)
16:10:27kugelfunman: amsinfo thinks usb_functio's size is negative
16:11:27funmanah right, the end is before the start
16:11:34 Join n17ikh [0] (
16:11:38gevaertsTillW: the theme editor tends to raise one question here: do you plan to reuse the existing rockbox WPS parser (I guess not, if you propose to use python), and if not, do you have ideas on how to make sure the editor doesn't fall behind when WPS syntax changes (mostly because of new tags)?
16:11:56funmanfor "trusted_fla" it makes sense because 0x81xxxxxx aliases 0x0, but here i don't know
16:12:32gevaertsI don't really expect you to have a ready answer to that, but it's something to think about when you write a more detailed proposal
16:13:12funmankugel: you should just print the size instead of stop-start if it annoys you, the correct size is used for dumping the file
16:13:53TillWgevaerts: i would prolly have to check the state of the existing WPS parser, possibly use it in part. i assume that the tool would have to be updated when the syntax changes unless i came up with some sort of loadable file that defines the syntax and output
16:15:28funmankugel: what's GPIOA value in __dbg_ports() for you ?
16:16:17Tornegevaerts: it calls the functoin redundantly quite a bit, I think the bus reset thing happens immediately before several of the others
16:16:38Tornegevaerts: but that doesn't really hurt, i don't think. it calls it maybe ~10 times total in the process of booting up, plugging usb, and unplugging usb
16:17:20gevaertsTorne: I think it's fine. Anything we might want in addition should fit easily in there
16:17:32 Join {phoenix} [0] (
16:17:35Tornegevaerts: yah, and I made it a two-way interface for the benefit of targets with a charging thread
16:17:46Tornegevaerts: so they can poll for how much to use instead o fbeing notified if that's more convenient
16:17:46kugelah, no wonder I got only garbage when loading the libraries
16:17:53kugelI mixed up the firmware versions
16:18:11Tornegevaerts: the only missing part for the generic code is making "force" actually do something, at the moment it does the same as "on".
16:20:13funmankugel: 2A590 : I see CCU_IO bit 12 is unset, then some FM code, then it is set again
16:21:16 Quit n17ikh (Ping timeout: 265 seconds)
16:24:14mt.. and proposal submitted !
16:25:14 Quit liar (Quit: Verlassend)
16:26:01 Join n17ikh [0] (
16:27:36 Join wodz [0] (
16:28:48 Join toffe82 [0] (~chatzilla@
16:30:47kugelwhen are the proposals due, btw? I guess not before we've been assigned the slots?
16:31:36 Quit Edwerd (Read error: Connection reset by peer)
16:31:40kugel9th april, meh
16:32:05TillWi suppose it's not like evil to change some of what i'll be doing after i make the proposal right? like if i get part way in and figure out that i was approaching it in a terrible way?
16:32:33pixelmakugel: that's not a good impression you leave there ;)
16:32:47 Quit petur (Quit: bleh)
16:34:02kugelah, finally found a caller for dbop_write_data_fast
16:34:41gevaertsTillW: If you mean after the project starts, if you agree with your mentor to change some approach or goals, that's fine. Before the submission deadline, you can change anything you like
16:35:20 Join liar [0] (
16:35:29kugelit must be lcd_update_rect
16:35:47 Join scokem [0] (
16:35:53TillWalright, that sounds about right. kinda just working through how things go with gsoc
16:37:56Zagoryay, 750 GB 9.5mm drives are here
16:40:28kugelhm, ida doesn't always detect veneers
16:40:42 Part xsteadfastx
16:42:13TillWjust out of curiosity, have there been many students about in here?
16:42:24funmanyes a bit
16:43:19TillWdo you remember what they were mostly interested in, and how sane their ideas were?
16:43:49gevaertsyes :)
16:44:51TillWhow sane do you think i am in comparison to them?
16:46:31funmankugel: 2A590 is fmradio_i2c_write(address, subaddress == -1, unsigned char *buf, int count)
16:48:15gevaertsTillW: I'm not going to compare :) If you're asking if it's worth it to spend time on a proposal, I'd say yes
16:48:52*pamaury notices that usb arc driver doesn't follow the manual advice for endpoint reset$
16:49:04*gevaerts accepts all blame
16:49:05 Join hebz0rl [0] (
16:49:08funmanpamaury: "the manual" is an official USB document ?
16:49:10TillWgevaerts: i'd say that's fair to say, thanks. brb in a few minutes
16:49:15moosTillW: don't look for other students, and just works/thinks more on your ideas. bluebroth3r and domonoky are often here if you want to wait and ask them something specific
16:49:29pamauryfunman: no, it's the manual of imx31
16:49:29gevaertspamaury: You mean I should have read the datasheet before implementing this? :)
16:50:01TillWmoos: those two are the mentors for that project?
16:50:02pamaurygevaerts: don't worry, the deviation is said to happened in "rare" cases
16:50:23geertvdijkI'm wondering; is there a guide to plugin writing available? something step by step about general structure stuff like that?
16:50:24pamaurywell, "very rare"
16:50:25gevaertsTillW: possibly, but that's not decided yet
16:50:34gevaertsThey're candidates at least
16:50:40geertvdijkfunman: wow, do I feel stupid now. thanks
16:50:49TillWso, as it stands they're the best to ask, alright =)
16:51:00TillWthanks, been a big help
16:51:26gevaertsWell, just ask "the channel". It's best not to fixate on one or two people
16:51:44TillWstill helpful to keep a few notes on it
16:51:54 Quit TillW (Quit: moving classes)
16:52:23moosTillW: if not mentors, they could help you out if you ask them particular questions about Qt and rockbox....
16:52:42mooshehe :)
16:55:03 Join naag [0] (~harish@
16:56:32 Quit kugel (Read error: Operation timed out)
16:57:07 Quit wodz (Quit: Leaving)
16:59:57***Saving seen data "./dancer.seen"
17:00:38archivatorspeaking of rbutil, seeing that there are more people around now, I wouldn't mind if someone checked FS #11155
17:01:09archivatori.e. check that it works :)
17:06:11Lloreanregarding the whole "resume a finished playlist" - I think one way for it to make sense (and not require a new option) is that you could be allowed to resume the playlist if you turn repeat on first.
17:06:19LloreanSo it allows, essentially, to retroactively enable repeat.
17:06:25*pamaury thought the broken usb queueing implementation of tomers was complicated but begins to find his driver quite complicated also...
17:07:16 Join dockimble [0] (~dockimble@
17:07:53 Join TillW [0] (
17:07:53gevaertsLlorean: that might be a solution, but my impression is that the problem the patch actually tries to solve is the "Nothing to resume is ugly!"
17:07:55 Join Xerion [0] (
17:08:04scokemfunman: I've sent you email to your hotmail account - let me know if there's any custom code you want me to run now I've got it installed
17:08:08Lloreangevaerts: Yes, but that's a subjective problem
17:08:24Lloreangevaerts: Meanwhile it might be considered a more "real" problem that if you let your dynamic playlist finish on accident, you lose it.
17:08:31pamauryLlorean: add a setting then :)
17:08:41gevaertsLlorean: very true, but is that one real enough?
17:08:51funmanscokem: i can't read that mail, if there's some code to try i'll post it in the forum thread
17:09:03Lloreangevaerts: Maybe, maybe not. But it doesn't require a new setting, addresses most of *both* sides' objections, makes a sort of logical sense, etc.
17:09:16LloreanI think it's a pretty good compromise solution.
17:09:30gevaertsLlorean: yes, but it doesn't solve any of the perceived issues, so I'd say it's not needed
17:10:52LloreanI don't know. For all the people who don't want to see "nothing to resume" can't they just, after the first time it accidentally ends, turn on repeat, resume (which they'd then be able to" and be okay?
17:11:11gevaertsOnly if they want repeat
17:11:25LloreanIf you're resuming an ended playlist, you want repeat.
17:11:41LloreanYou're just doing it manually.
17:13:53funmankugel: weird, I see B1 is used in buttons code, and also as SCL for FM i2c
17:14:02funmanthat can't be right
17:14:22dockimblehow is the fuze v2 port coming
17:15:17 Join einhirn [0] (
17:15:20dockimblereal close
17:16:02dockimblei have a fuze v2 if it would be helpful at all
17:16:53funmandockimble: read and tell us if your screen works
17:18:07pixelmagevaerts, Llorean: with the patch - wouldn't "Nothing to resume" be superfluous?
17:18:33Lloreanpixelma: With the patch as it stands, I think "nothing to resume" is now gone entirely.
17:19:03pixelmathen the patch is incomplete as far as I saw it today in the morning
17:19:18LloreanWell, my understanding is that it was to get rid of that.
17:19:48pixelmayes, but it doesn't touch any language file
17:19:52dockimblerockbox boots, shows home screen
17:20:32dockimblescroll wheel does not work, but we knew that
17:20:32funmandockimble: thanks
17:20:45gevaertspixelma: ah, right. Well, I want to keep the "nothing to resume" splash, so I don't care much :)
17:21:03dockimbleawwww - sooo close :)
17:22:01dockimblelet me know if i can do anything else
17:22:23Tornepixelma: you can still have nothing to resume
17:22:28Tornepixelma: if .playlist_contrl didn't exist on boot
17:22:32Torneand nothing hs been played since.
17:22:44Tornei.e. there has never been anything played
17:22:53Torneso, the splash (and string) are still needed.
17:23:04LloreanTorne: How do you feel about allowing retroactive repeat?
17:23:08pixelmaand another thought: it's been implemented that you can still *view* a finished playlist. You should be able to start it again by selecting the first track from the playlist viewer or save playlist and select it again (a tiny bit more interaction) - but I agree with gevaerts that it feels a bit wrong,
17:23:12TorneLlorean: i don't care either way, tbh
17:23:34pixelmaTorne: that would only be the very first time (or a broken playlist control)
17:23:39TorneLlorean: as long as it doesn't restart the playlist when i power on with start screen set to "resume playback" i don't care, i don't think i've ever selected resume playback manually
17:23:51TorneLlorean: i turn off my player fairly often, and i expect it to keep playing when i turn it back on :)
17:24:03Tornebut i don't want it to start playing the same thing again if it finished.
17:24:16Tornebecause that slows down the process of picking more music (huge contention for disk bandwidth)
17:24:42LloreanI almost exclusively use "resume playback" manually on startup
17:24:55Torneyeah, i modified my build to make resuming even *more* automatic :)
17:24:55GodEaterI use the Torne method ;)
17:25:09LloreanI usually hit it sometime during booting, but not always because I may boot my player before I actually can put my headphones on.
17:25:12Torne(i disabled hold switch detection in the bootloader and for resetting settings, so i can just press power, turn on hold, drop in pocket)
17:25:23Torne(minimal interaction to carry on listening)
17:25:29GodEaterTorne: neat idea ;)
17:25:43funmanFlynDice: (logs) have you tried to use PIO instead of DMA for writing ?
17:25:46TorneGodEater: yeah. you can still boot OF by holding menu, on ipodvideo at least
17:26:43 Join jgarvey [0] (
17:27:00GodEaterI very rarely have repeat turned on. like, almost never. But I usually have my player doing the automatic folder advance, so my playlist is almost never empty.
17:27:02LloreanShouldn't we be removing hold detection anyway once we've got our own USB code working well?
17:27:04pixelmaTorne: I'd guess that you'd see a difference with this patch if your player turned off after the end of playlist (and maybe an idle power-off). With the patch I'd expect it to start the playlist again if your start screen is set to "resume at startup"
17:27:07funmansomeone with a clipv2 can confirm ?
17:27:28dockimblei'll see what i can do
17:28:31GodEaterI fail to see the point of having a DAP actually powered up and NOT playing something :D
17:28:33pixelmaI'm not sure though, but can't imagine a difference between manual resume and resume as start screen
17:28:59*funman thinks GodEater never used the lamp plugin
17:29:09GodEaterthe most pointless plugin ever
17:29:12LloreanGodEater: Sometimes if yours boots slowly (Gigabeast) you turn it on then stuff other stuff in your pockets, then put the headphones on, then hit resume.
17:29:17GodEaterbut then, I barely use any plugins at all
17:29:19LloreanSo you're not waiting for it to boot as you go down the stairs outside.
17:29:22funmanGodEater: i have had use for it
17:30:07GodEaterLlorean: I did say "I"
17:30:32GodEaterI think the beast boots faster the ipod video anyway
17:31:17LloreanI can't imagine how long the Video takes these days then
17:31:24LloreanThe beast is pretty slow recently
17:31:30 Join kugel [0] (~kugel@rockbox/developer/kugel)
17:31:37funmanwhat is slow, ten seconds?
17:31:46LloreanSlow is "twice as long as it used to"
17:31:53GodEaterI think the video is about 20 seconds, sometimes longer
17:32:23funmanhum i have had an ipod video with 3.4 in my hands, it was nothing in this range, more like 5 seconds maximum
17:32:26Lloreanfunman: About 15 seconds.
17:32:38Llorean3.4 is before the slowdown, I think
17:32:45GodEaterfunman: then your ipod has sod all music / files on it
17:32:46LloreanTry an SVN build
17:32:56GodEaterdircache takes AGES on a full 80GB
17:32:56funmani don't have it anymore
17:33:01LloreanGodEater: Mine's 15 seconds without too many files.
17:33:05LloreanIt's not dircache that's doing it.
17:33:08LloreanFor me at least
17:33:17funmancould you bisect the change that introduced the problem?.
17:33:25GodEaterfunman: no point
17:33:36GodEaterit's all the stupid eye candy that people keep insisting on
17:33:48GodEaterit's not going to get taken out
17:33:51funmanwe can have fast eye candy
17:33:52Tornepixelma: yes, that's what i'm complaining about :)
17:33:52GodEaterso I live with it
17:34:04Tornepixelma: there *isn't* a difference between manula resume and resume from start screen
17:34:11pixelmaLlorean: please test and give numbers... :\ I saw slowdown but not as much anymore (my targets are all flash now and I also wondered if the fact that my WPS doesn't use many bitmaps has to do with it)
17:34:32GodEaterfunman: besides, it's been such a slow increment in the boot time I don't think it's any single change.
17:34:33Lloreanpixelma: I just use the default cabbiev2
17:35:01Tornealso, er, my ipodvideo boots in about 8-9 seconds max
17:35:12Tornefrom pressing button to being on main menu
17:35:25*GodEater will try now
17:35:28LloreanI'll admit the first large chunk of time on my beast is that Microsoft loader.
17:35:31Torne(and i can take an extra second off that by installing rockbox to OSOS)
17:35:40LloreanBut it used to boot in a very short time after that finished.
17:35:49LloreanNow I seem to spend just as much time staring at a Rockbox logo
17:35:50evilnick_BOnce a playlist (dynamic or not) is played out, can it still be saved?
17:36:03evilnick_BOr is it immediately deleted?
17:36:09Lloreanevilnick_B: IIUC it's still viewable in the playlist viewer.
17:36:16GodEater21 seconds to playback starting
17:36:17moosLlorean: The rockbox logo used to appear and go in fraction of seconds...longer here.... :/
17:36:46evilnick_BLlorean: I'm thinking about... never mind! You already mentioned that on the dev thread. Good work!
17:37:11TorneGodEater: ok, it takes about 16-17 seconds before i get sound
17:37:22GodEaterwell that's what I count as startup time
17:37:36Torneand I don't have retailos in the firmware parittion
17:37:41Tornejust the RB bootloader in OSOS
17:37:43GodEatersee my previous comment about a DAP being essentially pointless unless it's playing something
17:37:48pixelmaLlorean: for comparison - on the Ondio which I currently have handy, cabbiev2 uses 8 bitmaps, my theme 3 - on my greyscale and colour targets there's another difference that my theme only uses monochrome bitmaps... and well yes, no HD in my targets anymore
17:37:51*moos complained few times here, and promise to not slaps jdGordon again without stats
17:37:59TorneGodEater: i disagree though, i don't want it to play the wrong thing
17:38:00Lloreanpixelma: Do you use an SBS, etc?
17:38:04LloreanAll the "newer" theme stuff.
17:38:07 Nick fxb is now known as fxb__ (
17:38:09Torneit makes it take two or three times longer to pick the right thing.
17:38:12 Join kugel_ [0] (
17:38:20GodEaterTorne: but if you have it set to resume on startup how is it playing the "wrong" thing?
17:38:28 Quit kugel (Disconnected by services)
17:38:32 Nick kugel_ is now known as kugel (
17:38:36 Quit kugel (Changing host)
17:38:37 Join kugel [0] (~kugel@rockbox/developer/kugel)
17:38:40kugelfunman: good find, already hearing sound? :p
17:38:44TorneGodEater: because someone wants to make it resume from the start of a finished playlist
17:38:48moosI remenbered that this began with the restructuration of statusbar with this new sbs system :(
17:38:49Torneisn't that the point of this discussion? :)
17:38:52pixelmaLlorean: only on the c200 - and I currently use test builds with the FMS patch applied though no advanced .fms yet (no bitmaps)
17:38:58GodEaterTorne: yeah, but I don't
17:39:07GodEaterI'm just saying how long it takes to start up *now*
17:39:14GodEaternot with / without this proposed change
17:40:02*Llorean still thinks enabling repeat to enable this feature makes an elegant sort of sense.
17:40:18funmankugel: nope, I can't get FM to work at all
17:40:31dockimblei have to say - great job on getting the clip+ rockbox out so fast - i was using it with rockbox when exercising today and it works great
17:40:33TorneLlorean: that does seem kinda logical :)
17:40:35funmani saved my wip in git stash
17:41:06TorneGodEater: for me it only takes that long to resume playback because it's loading the db to ram, though, i think
17:41:15GodEaterI don't even use the DB
17:41:28*GodEater goes to check he's not lying about this
17:41:40*Torne tests also
17:42:04funmanhm it could be a newer revision of teh chip though
17:42:04*GodEater coughs and shuffles his feet
17:42:08GodEaterI feel a new test coming on
17:42:20Tornehm, actually that makes very little difference
17:42:23Tornemaybe 1-2 seconds
17:42:38GodEaterTorne: I also just turned off DB auto update
17:42:45Tornei don't have that anyway
17:42:46*Llorean doesn't have the DB enabled.
17:43:13GodEatermade no difference at all turning it off
17:43:17 Quit kugel (Remote host closed the connection)
17:43:27Tornehm. maybe it really is wps loading then
17:43:46GodEaterI'm convinced it is.
17:43:56GodEaterEvery time an eye candy monkey comes near the code, Rockbox gets slower.
17:44:00Torneis there still a builtin wps?
17:44:14GodEaterno idea how you choose it tho
17:44:22LloreanI'm not wholly sure it is WPS
17:44:23pixelmaLlorean: that's what I think about the resume patch too. The only downside to enabling repeat generally is that your player will play on until the battery is flat
17:44:24GodEaterbut you get it if you load a broken theme
17:44:38pixelmaGodEater: chose the rockbox_default.wps
17:44:39LloreanI just turned of dircache and it shaved my start time *a lot*
17:44:49TorneLlorean: yes, dircache is some of it
17:44:53Tornebut i've always ha ddircache on
17:45:02LloreanWhich is odd, I've had dircache on for years, not changed the content on my device in almost as much time, and the significant increase in boot time just happened recently
17:45:14LloreanTorne: Maybe something has broken / slowed dircache scan time?
17:45:34pixelmamaybe some dircache changes then (there was some work by pamaury recently)
17:45:41LloreanTHAT is odd
17:45:45*Torne sets wps to rockbox_default
17:45:53LloreanThe "scanning disk" *splash* boot took less time than the boots prior to me disabling dircache
17:45:53pamauryLlorean: normally, boot time should not have changed, I recently changed the initial scan but the speed should be the same
17:45:54GodEaterI know there was a huge difference that some chance that jhMikeS made a while back (but only on the ipod video) - and I think his commit got reverted
17:45:58LloreanSomething may be wrong with teh background scan
17:46:04Torneyep,it's the wps
17:46:08Tornei assur eyou
17:46:13GodEatersee? :)
17:46:17LloreanTorne: It may involve the WPS
17:46:19Tornesetting it to rockbox_default just saved mayhbe 8-9 seconds or more
17:46:25*Torne tries again
17:46:27*GodEater proposes banning all future wps changes
17:46:34Tornemy wps is old and uses almost no features, too
17:46:34pamauryso dircache is not the cause ? :)
17:46:37Torne(escape pod)
17:46:39Tornei dont have an sbs
17:46:42LloreanBut the explicit "scanning disk" scan took less time for it to boot than the background dircache scan boots cost me
17:46:44Torneit's not even a complicated one.
17:46:48LloreanSo dircache may be related to the cuase.
17:47:18Torneyeah, 10 seconds to music with no wps
17:47:21LloreanYeah, dircache back on, boots incredibly slow again (Except that first "scanning disk" one)
17:47:21Tornefrom cold
17:47:38Tornewith load db to ram re-enabled
17:47:39LloreanTorne: Try dircache off, and then two boots with it on
17:47:49LloreanI'd like to see if you get the same behaviour as me for that with cabbiev2 at least
17:48:23pixelmadircache off could explain why I don't see it anymore
17:48:27*Llorean would apparently currently always prefer a dircache full scan.
17:48:33*Torne disables dircache
17:48:44pamauryTry with and without tagcache loaded to ram
17:48:54LloreanI'd say try with tagcache off
17:48:58LloreanSo that you can reduce the number of elements.
17:49:01TorneLlorean: disabling dircache makes no difference at all
17:49:06Tornefor me, with no wps
17:49:12LloreanWith cabbiev2 though.
17:49:36LloreanCould the WPS loader and dircache's background scan both be fighting for the disk slowing everything down?
17:49:46LloreanAs I said, the explicit foreground dircache scan was *faster*
17:49:50LloreanPossibly half or less the time
17:49:58Tornei know that i fyou load tagcache to ram and have dircache on and load a wps and resume playback all at once the entire UI grinds to a halt
17:50:02Tornebecause that's my usual case
17:50:08Torneas a bonus if you hold stop while it's doing that it normally panics
17:50:12Tornewith a directory error :)
17:50:24pamaurybut background scan is in background so it doesn't impact performance very much
17:50:38LloreanAnd it seems rather odd to me that the explicit foreground scan was the faster boot - I thought we do a background scan after the first time so that boots are faster.
17:50:43*GodEater awards pamaury the prize for "stating the bloody obvious"
17:51:07Lloreanpamaury: I think my point is that it's supposed to be faster but clearly isn't on my player right now
17:51:33Torneok, next one: cabbiev2, no dircache, no tagcache in ram
17:51:44pamauryLlorean: background should be faster ?.
17:52:04Lloreanpamaury: The background scan should have a faster boot. It does not.
17:52:18TorneShit, that's even slower than my old wps
17:52:27Torne22 seconds to boot to audio with cabbiev2, no caches
17:53:15LloreanNow 2 boots with dircache. The slow scan and the background scan.
17:53:29 Join dockimble1 [0] (~dockimble@
17:54:00 Join captainkewllllll [0] (~2669ecc2@gateway/web/freenode/x-dgavdmuzohkmwmuu)
17:54:03pamauryLlorean: ah yes of course, the background scan should allow boot to continue and the scan can take place afterwards
17:54:24TorneLlorean: 34 seconds
17:54:56Tornethen 17 seconds.
17:55:02Torneso for me it's behaving exactly how i'd expect
17:55:05LloreanThat's exactly the opposite of mine.
17:55:14Tornebooting with cold dircache is very slow, with hot dircache is faster than no cache
17:55:18Tornethat's what i'd expect it to do
17:55:27GodEaterLlorean: is the beast your only remaining disk based player?
17:55:29Torneexcept if i disable the wps this effect drops to bascially zero and *everythin* is way faster
17:55:34 Quit hebz0rl (Quit: Ex-Chat)
17:55:38LloreanGodEater: No, I also have an H100
17:55:52LloreanBut I don't have any clue where the charger is for it.
17:55:56pamauryok, have your mind pretty clear and tell me your conclusion when I'll be back in a hour or so (or before)
17:56:14 Quit captainkewllllll (Client Quit)
17:56:20GodEaterI imagine the H100's wps requirements are less that for a colour target too
17:56:28mooshere I can see difference comparing to before. When I unplug from usb, the sacn take 19s, but from boot 22s. Those added seconds came frome the culprit revision that I still have totest
17:56:36LloreanGodEater: Probably significantly so.
17:56:49 Quit pamaury (Quit: Quitte)
17:56:59*GodEater glares at the eye candy people
17:57:03moosI refer to debug infos here
17:57:14LloreanGodEater: Well, the change seems to at least relate to the WPS for sure. Whether it interferes with the disk access too isn't too important if it can be sped up there.
17:57:38GodEaterdon't get me wrong, I don't want to go back to text only wpses
17:57:39LloreanCan the WPS load be taken out of the boot process and be handled the first time you try to enter the WPS screen?
17:57:50GodEaterbut all the gubbins we have in there now is mostly sizzle I don't care about
17:58:05*moos is sure that all this slowdown came from sbs but still have to compare
17:58:07TorneLlorean: you mean resume playback without actually going to the wps?
17:58:16GodEaterLlorean: loading the WPS *is* part of the boot process to me, since I have resume on startup selected.
17:58:26LloreanGodEater: Yes, but it's not part of it to all of us.
17:58:32TorneLlorean: I guess you could hack the resume code to not actually switch to wps
17:58:41Torneand just resume playback while staying on th emenu
17:58:44Tornethat might make a difference
17:58:49LloreanI'd say even then, "Boot, Music starts, then WPS finishes loading" gets you where you want faster than "Long ass boot, then Music"
17:59:11LloreanTorne: I'm more just talking about the wps files not being loaded until they're explicitly needed, rather than during the boot process.
17:59:11GodEaterI'd be happy with that
17:59:22GodEaterI spend almost zero time actually looking at the WPS anyway
17:59:28TorneLlorean: are they loaded at boot?
17:59:38Tornemy original point was that booting to the main menu is *also* much quicker
17:59:50LloreanTorne: I don't boot with resume enabled
17:59:55Torneif i unplug my headphones to prevent resume on boot i get a usable main menu in ~9 seconds
17:59:56LloreanAll of my times were "time to menu"
18:00:02LloreanNot time to WPS
18:00:05TorneThat's a massive difference
18:00:10 Join bluebrother [0] (
18:00:10 Quit bluebrother (Changing host)
18:00:10 Join bluebrother [0] (~dom@rockbox/developer/bluebrother)
18:00:12TorneI am couting from pressing power to hearing music from autoresume
18:00:24Torneif it's not resuming it boots in about 9 seconds anyway
18:00:34 Quit guymann (Ping timeout: 252 seconds)
18:00:41Torneit is unlikely to go *much* faster no matter what because a significant chunk of that time is in the apple/rockbox bootloaders
18:00:46LloreanIf I hit "resume" while the disk is still spinning, the resume is instantl
18:01:08 Join guymann [0] (
18:01:11Torneso we are comparing two completely different things :(
18:01:17LloreanTorne: See, literally 5 seconds of my boot is looking at the Rockbox logo *after* the bootloader
18:01:24LloreanMaybe 7 seocnds
18:01:36LloreanThis is way too long for something as fast as the 'beast.
18:02:06Tornehm, this is being all very unrepeatable
18:02:16moosonce again, the rockbox logo used to show and go, not this latency we have now :/
18:02:41Tornewell, when i get home i may investigate :)
18:02:50Tornei'll bootchart it with logf and USEC_TIMER :)
18:02:51LloreanTorne: Well, if it's disk throttle rather than CPU throttled it might explain why the gigabeast "feels" more affected than the iPod
18:03:11moosand the dircache take more time to scan than before. ie it used to take same time unpluging from usb that at boot
18:03:16Tornebecause now that you mention it i've noticed
18:03:28Torneand for me it's definately *way* faster without a wps
18:04:00 Quit bluebroth3r (Ping timeout: 276 seconds)
18:04:11Tornei'll pick a bunch of points during boot and logf it, so i can make accurately timed logs of how long each stage takes
18:04:23Tornethen repeat it for various settings combinations (dircache/tagcache/wps)
18:04:28*moos would like to test his theory, but his beast is on the OF recovery mode waiting to access a 32bits computer to use the apps
18:04:28LloreanTorne: It's 1 or 2 seconds faster with the rockbox_default theme
18:04:29Tornethat should be informative :)
18:04:33LloreanTorne: It helps some.
18:05:11*Torne puts his settings back for now though ;)
18:05:16moosTorne: I would like to see if I can test before the revision I suspect to be culprit
18:05:20moosto compare with
18:05:48Tornemoos: well you can test what you like..
18:06:02TorneI have spent ages analysing the *prior* part of boot speed on ipod before
18:06:05moossure we can
18:06:10Tornefrom poweron until the end of the bootloader :)
18:06:18Torneso, i'm gonna do the same for this bit
18:06:20mooshehe :) good guy
18:06:31Tornewell that's how come you can boot rockbox from OSOS now
18:06:42Tornebecuase i measured the speed difference in detail ;)
18:06:44Torneand added support
18:07:02Tornei just stopped counting immediately before the bootloader runs the loaded app
18:07:37Torneso, now i know how to use logf, i can chart what happens from entry to the app until we're done.
18:08:45moosLlorean: can you compare the boot time given in the dircache debug menu when (normal boot) and when you unplug from usb?
18:08:51 Quit scokem (Quit: CGI:IRC)
18:09:25 Join komputes [0] (~komputes@ubuntu/member/komputes)
18:10:52 Join Buschel [0] (
18:12:30 Join mt__ [0] (~mtee@
18:12:57Lloreanmoos: 3s for a foreground scan on boot, 5s for a background scan on boot, 3s after USB logf threadsafe?
18:13:50Torneoh, wait, that's a stupid question
18:13:51 Quit mt (Ping timeout: 268 seconds)
18:13:56 Nick mt__ is now known as mt (~mtee@
18:13:59Tornebeen doing too many threads today :)
18:14:03 Join fyrestorm [0] (
18:15:40 Quit Buschel (Ping timeout: 276 seconds)
18:16:36funmanwell there are thread safety problems in rockbox, but only in the isr
18:17:30 Join Buschel [0] (
18:21:23 Join domonoky [0] (~Domonoky@rockbox/developer/domonoky)
18:23:03moosLlorean: those 2 extra sec wasn't before. That is one symptom of this slowdown, some things appears to hung at starts, that's why I first thought that's just an init seq problem... I will try to get my beast back on working, and try to find the culprit commit (I know almost where to start to test revisions)
18:28:07 Join Locke_Fireclaw [0] (
18:28:36Locke_FireclawCould someone please, pretty, PRETTY please...gimmie a hand with this Rockbox dealie? I'm running out of hair to pull...
18:29:46archivatorLocke_Fireclaw: just ask your question
18:30:37Locke_FireclawI'm getting stuck at an unsupported sector size error, the iPod is a 80gb 5.5 Video, but using a drive from a 80gb Classic.
18:32:28 Quit funman (Quit: free(random());)
18:32:53archivatorwhat's the sector size?
18:33:26Locke_Fireclaw4096, says the little black screen my iPod has been sat on for the last 5 hours. lol
18:33:53 Quit Buschel (Ping timeout: 276 seconds)
18:33:55 Quit ivortheengine (Ping timeout: 258 seconds)
18:34:34moosLlorean, Torne and others: I'm sure the problem began at r23258 if that not this revision, sure it's around it. If anyone could test revisions prior, before me (before I fix my beast to test)
18:34:35archivatoriirc you'll have to modify the source - there is a constant somewhere defining the maximum size, compile a new bootloader and binary and use those..
18:35:05moosmore than 5 months already :/ !!!
18:35:09archivatorLocke_Fireclaw: see this:
18:35:23moosand one release
18:35:58Locke_FireclawI've already read that page archivator, and several more like it, even tried to recompile...but it can't have gone right, seeing as my 'pod is still broked. :(
18:36:19archivatorLocke_Fireclaw: did you change the bootloader?
18:36:50Locke_FireclawI overwrote all the files in .rockbox on the 'pod with the new ones.
18:37:17archivatorThat won't change the bootloader
18:37:37 Join pamaury [0] (~c2c7a50a@rockbox/developer/pamaury)
18:37:45Locke_FireclawAha, so how do I do that? The only folder I have in the recompiled .zip is .rockbox.
18:40:48junkY_Sanusing the clip v1 simulator, settings shuffle to yes and repeat to all, change directory to random. it doesn't change the folder it repeats all tracks in the folder i started playing
18:41:17archivatorLocke_Fireclaw: I'm not sure what the install process is but you can build a boot loader by rerunning tools/configure and selecting a Bootloader build
18:41:56pamauryTorne, Llorean : did you reach a conclusion concerning dircache ?
18:42:04archivatorLocke_Fireclaw: see
18:43:42archivatoripodpatcher resides in rbutil/ipodpatcher btw
18:43:45Lloreanpamaury: moos seems fairly convinced something is slowing it down during boot (the background scan takes 5 s during boot according to debug, when it takes 3s after USB)
18:44:43pamauryAnd did a recent commit slow it down ?
18:44:44 Join fragilematter [0] (~fragilema@
18:45:05Lloreanpamaury: Don't know. His suspicion is one about 5 months ago, which would correspond approximately to when complaints started.
18:45:25pamaurywhich number ?
18:45:41junkY_Sanmh, it also doesn't shuffle the files in the current dir
18:47:05Locke_FireclawWell, all I got is errors, told you I suck at this. :( All I have to do with my 40g 4th gen is run the installer.
18:47:39Lloreanpamaury: 23258 "or around it"
18:47:54junkY_Sanman i'm stupid, i need to create a playlist containing all directorys
18:48:43archivatorLocke_Fireclaw: your 4g doesn't have a changed HDD :) I'm off to dinner now but I'll help you when I'm back
18:49:11Locke_FireclawThanks archivator, appreciate it...and, bring me back some scraps eh? I haven't eaten yet today. :/
18:49:36*moos just svn update -r 23257 to test on his h10 20GB
18:50:09pamauryLlorean: ok thanks, I'll have a look to see if something changed
18:50:14domonokyLocke_Fireclaw: you need to build the bootloader, but you can use the normal ipodpatcher to install the freshly built bootloader. Where do you get those errors ?
18:50:55moospamaury: The problems are not dircache related. Don't worries :)
18:51:51Locke_FireclawWell, domonoky, it's a long story... I have a full Ubuntu install, but I'm currently stuck on this side of the PC due to giant adobe updates, so I'm getting the errors in my Ubuntu x64 virtual machine, after I run the ../tools, forget the exact line, I'm on a different machine again and the VB is powered down. lol
18:52:33domonokyLocke_Fireclaw: we can only help you if you gives the the full error output (via :-)
18:52:40pamaurymoos: ah ok, so I won't have a look then :)
18:52:43Locke_FireclawOk, bare with.
18:52:49Locke_FireclawI'll power it back up again.
18:53:45 Join mitk [0] (
18:54:19 Join bmbl [0] (~Miranda@unaffiliated/bmbl)
18:56:35 Join ivortheengine [0] (~ivortheen@
18:57:00moospamaury: ggr, I don't remenber where I leave the usb cable of my H10, still searching
18:57:01Locke_FireclawOk, sorry 'bout that.
18:58:02Locke_FireclawSo I'm in the terminal on my ubuntu virtualbox.
18:58:10Locke_FireclawWith ../tools/configue
18:58:28Locke_FireclawI get the option for which device, so I go 22 for the iPod video.
18:58:57Locke_FireclawSelect 64mb ram, and B for bootloader
18:59:03domonokyLocke_Fireclaw: just paste the full terminal output after you got the error to and post the link here.
18:59:19domonokythen we also know what you have done :-)
18:59:58***Saving seen data "./dancer.seen"
19:01:08domonokyLocke_Fireclaw: looks like thid build directory isnt clean, try make clean first, or just delete and recreate the build directory.
19:01:19mooswow r19450 on his H10 (long time I didn't use it) fast boot (/me swears)
19:01:31Locke_FireclawOk, just wipe out everything domonoky?
19:01:52 Join froggyman [0] (~me@unaffiliated/froggyman)
19:02:02domonokyLocke_Fireclaw: yes, you can just delete an recreate the build directory, or just use another build dir.
19:02:13junkY_Sanis it possible to let the player create a playlist with all the files on it everytime it boots?
19:03:04domonokyyou should not try to build different builds oin one dir, without makeing sure its clean. :-) But you can just use more build dirs like "buildipod" and "buildipodboot" :-)
19:03:25domonokyjunkY_San: not automatically.
19:04:42Locke_FireclawOk, it's run through all that, with no errors.
19:04:53 Join Strife89 [0] (~michael@
19:05:39domonokyso now you can use ipodpatcher with the right parameters (try ipodpatcher -?) to install the new bootloader to your ipod.
19:06:00 Join tomers [0] (
19:06:01Locke_FireclawOk, bare with, cross fingers etc...
19:06:27Locke_FireclawOh darn,'s a virtual machine, I won't be able to move it to the 'pod from there.
19:06:49Locke_FireclawIs there a windows version I can use?
19:07:39Locke_FireclawWait, is it just the .ipod file?
19:08:08Locke_FireclawCopy paste ftw...
19:09:07CIA-5New commit by 03tomers (r25400): Viewer: Implement alignment option, to provide alignment of RTL text (Hebrew, Arabic) to the right
19:09:17junkY_Sanfrom the advanced random folder plugin i select "play shuffled" but it's still the same order like on the fs
19:10:37junkY_Sanor is it possible that it shuffles only folders but not the files in them?
19:11:33domonokythats possible, use the normal shuffle option to shuffle the files in the folders :-)
19:12:02junkY_Sanmh yes i'm just playing with it, i have to reselect the shuffle option the reshuffle the whole playlist
19:12:08Locke_FireclawNope, I'm being dim, where do I put the .ipod file manually? Or is that not possible? I don't see an original one on the iPod it's self, after using the install utility.
19:12:19junkY_Sani'm searching for an easy way to just listen to all the tracks shuffled from the player bootup
19:13:31domonokyjunkY_San: you could use the database or use insert shuffled on your music folder on startup (with recursive insert enabled)
19:14:04junkY_Sanah okey the 2nd idea sounds good. 1st is not possible since it's clipv2 and i can't save the database
19:14:25domonokyLocke_Fireclaw: you are dumb ! :-) i told you about ipodpatcher, please read what i wrote.
19:15:01Locke_FireclawWas just hoping I could do it on the fly domonoky, but yeah, just grabbed ipodpatcher.
19:15:40bluebrotherdomonoky: help for ipodpatcher is -h or −−help, not -? :)
19:15:57archivatordomonoky, bluebrother : thoughts on FS #11155? :)
19:16:15domonokybluebrother: i dont know, he should just try it ! :-)
19:16:32bluebrotherarchivator: had a short look only right now. Regarding the spinlock, wouldn't it be better to use a QTimer instead?
19:16:59Locke_FireclawWell, baring in mind I'm using the windows version...
19:17:07bluebrother(though I have to admit that I implemented a pretty similar spinlock in ttscarbon and forgot the look into changing that)
19:17:44archivatorbluebrother: didn't want to use any events - that would interrupt the flow of the function.. Plus, it's not really a spinlock, so I don't see why we can't leave it like that
19:18:19archivatorIt works, doesn't block the GUI and is easy to understand - adding Timers would add unnecessary complexity and only make things harder to follow
19:19:19domonokyarchivator: i am not sure of the leak patch. the tts object is created with the config window as parent, so it should get destroyed on closing of the config window.
19:19:19bluebrotherarchivator: you've got a point :)
19:19:29 Quit dionoea (Changing host)
19:19:29 Join dionoea [0] (~dionoea@videolan/developer/dionoea)
19:19:29archivatordomonoky: nope
19:19:43archivatordomonoky: config objects are never destroyed, they're only hidden
19:20:00bluebrotherdomonoky: IIRC the problem with the config window (and in fact various other windows) is that they aren't destroyed but only hidden
19:20:15archivatorSo you end up with ~10 TTSBase objects destroyed when you close the application
19:20:27archivatorWhich explains why the festival process was constantly running
19:20:36archivator(it's terminated in the destructor)
19:20:40domonokyyes, then we should delete them, but we have to make sure the parent doesnt do the delete also.
19:20:45bluebrotherand even if the config window would destroy the objects destroying them manually won't hurt.
19:22:04 Join punkt [0] (
19:22:16junkY_Sandomonoky: insert shuffled gives me "error updating playlist control file" and just adds one dir
19:22:26archivatorWell, I didn't want to touch the core rbutil since I haven't given it more than a cursory glance. But yes, ideally, everything would be deleted once we're done with it.
19:22:51 Join funman [0] (~fun@rockbox/developer/funman)
19:22:53domonokyjunkY_San: so that also doesnt work because of no write. so you have to be patient :-)
19:22:54Locke_FireclawOk, that's it should've installed my edited file...*crosses fingers*
19:23:06junkY_Sanyeep :)
19:23:11 Join Horscht [0] (~Horscht2@xbmc/user/horscht)
19:23:12bluebrotherI hope to address that at least partially with the GUI rework I'm planning to do. Though I guess it will take some time until I get something working there
19:23:13archivatorLeaving it to QObject is not optimal (btw QObject's destructor notifies the parent, so you won't get double destruction)
19:23:18funmansomeone with a discharged (anything between 0 and 95%) Clipv2/Clip+ willing to try a patch ?
19:23:27 Quit toffe82 (Read error: Connection reset by peer)
19:24:01*domonoky reads that QObject automatically deregister on parents if they are destroyed, so deleting manually is fine.
19:25:00bluebrotherwe're using delete on a few objects already. It definitely works correctly :)
19:25:37Locke_FireclawA little further, but still not dice. :( I get past the black screen with the error, the rockbox logo flashes up, then I get a white screen with exactly the same sector size error. heh
19:25:40 Quit ivortheengine (Ping timeout: 258 seconds)
19:27:10archivatorLocke_Fireclaw: that's good news, actually - means the bootloader works.
19:27:11domonokyso thats now probably rockbox itself, di you replace it with your own built version ?
19:27:34Locke_FireclawNope domonoky, just one from the installer utility.
19:28:23*domonoky whished people would do what they are told. We already told you replace that.
19:28:54Locke_FireclawYou said to replace the bootloader...I did. :/
19:29:11domonokyarchivator: patch looks fine on a first look.
19:30:52Locke_FireclawOk, I'll replace the lot.
19:31:00archivatordomonoky: Well, if there are no objections, you are welcome to commit it :)
19:32:02 Join phanboy4 [0] (
19:34:23junkY_Sanfunman: clipv2 92%
19:34:49funmanjunkY_San: can you compile your own build?
19:35:58Locke_FireclawWoot, ok, yes, I'm stupid then.
19:36:01Locke_Fireclaw100% dim.
19:36:15Locke_FireclawReplacing the entire thing with the recompiled version worked, it's up & runnin' again.
19:36:23Locke_FireclawThanks guys, sorry for being slow.
19:36:29junkY_Sanfunman: have not tried yet, give me 5 minutes
19:36:49 Quit fragilematter (Quit: leaving)
19:37:54funmanjunkY_San: i can send you a binary if you prefer
19:38:06 Join mischasworld [0] (
19:39:28junkY_Sansince i wanted to have a look at the code i at least have to get the svn tree anyway
19:40:25funmanthat helps :P
19:41:35domonokyarchivator: are you already in the credits ?
19:42:14archivatorI think so
19:43:05funmanhm my Clip+ has less battery now and the patch seems to be alright
19:43:26archivatorLocke_Fireclaw: you'll have to forget about rbutil for anything but the extras/accessibility. If you want to update RB, you'll need to recompile it manually. Just making sure that got through :)
19:43:51CIA-5New commit by 03Domonoky (r25401): Delete tts objects after use. ...
19:44:38funmanjunkY_San: <- just checks if battery reading is correct with charging plugged AND unplugged (go to debug > battery to check)
19:45:17bluebrotherdomonoky: you've committed part 2, not part 1 ;-)
19:45:25CIA-5New commit by 03Domonoky (r25402): Fix Festival tts engine. ...
19:45:57domonokybluebrother: no this is part2. :-)
19:46:35bluebrotherdomonoky: well, the patches were numbered the other way round ;-)
19:46:49*domonoky renumbered them :-)
19:47:25bluebrotherbut for something different: what do people think about the changes in r24778 (linking to the sansa forums from rbutil)? I'm not too fond of linking to external pages from rbutil as we don't have them under our control
19:48:20bluebrotherIMO it would be better to create some tinyurl-like links via, or even keep linking to our wiki only instead.
19:48:21 Join toffe82 [0] (~chatzilla@
19:49:27junkY_Sani'm looking for the file but don't find it
19:49:39bluebrotherjunkY_San: tools/ in svn
19:50:55 Quit anewuser ()
19:51:30 Join dfkt_ [0] (dfkt@unaffiliated/dfkt)
19:51:55 Join fragilematter [0] (~fragilema@
19:52:44 Quit bmbl (Quit: Bye!)
19:54:03 Quit tomers (Quit: ChatZilla 0.9.86 [Firefox 3.5.8/20100214235838])
19:55:04 Join Stephen__ [0] (~S@
19:55:22 Quit dfkt (Ping timeout: 265 seconds)
19:56:35 Quit fragilematter (Client Quit)
19:57:26 Quit dfkt_ (Ping timeout: 268 seconds)
20:02:58topikfunman: i have a clip+ at 90%. worth testing your patch on?
20:03:19funmantopik: sure
20:04:18 Quit TillW (Read error: Operation timed out)
20:05:00 Join dfkt [0] (~dfkt@unaffiliated/dfkt)
20:05:54topiknever figured finding my usb cable would be the hardest part
20:09:29 Quit mitk (Quit: Leaving)
20:10:08topikfunman: did you break any clip+ usb ports yet? it seems very tight
20:10:22 Quit Locke_Fireclaw (Quit: ChatZilla 0.9.86 [Firefox 3.6.2/20100316074819])
20:10:39funmanphysical properties of USB connectors are specified afaik
20:11:02topikit feels vulnerable
20:12:43topikbattery level value is still jumping
20:12:48topikafter this patch
20:13:16funmanwhen charging only ?
20:13:28funmanit is flat for me
20:13:43funmanvery slowly increasing
20:13:50topikjumping between 85 and 91
20:14:20funmanthe graph shows that ? (first screen in battery debug)
20:14:39topikno, the 4th screen
20:14:54topikwith the text 'battery level: xx%'
20:15:03funmanhm :/
20:15:12 Join panni_ [0] (
20:15:15funmancan you double check you're running the patched version?
20:15:42 Quit flydutch (Quit: /* empty */)
20:15:48 Join wodz [0] (
20:16:15topikr25402M-100330. pastie-895592.diff causing the M
20:17:04funmanthe OF seems to only check this ADC channel
20:18:42funmantopik: with charger unplugged, level is read the same as without the patch ? (i.e. read fine)
20:18:46topikOF only shows an icon does it?
20:19:05funmanyes but there are 4 or 5 icons for each level (very low, low, high..)
20:19:28topik85-91 might not make a difference to the OF's icon range
20:19:47topikit does slowly increase
20:19:50topikwhile charging
20:21:25funmanhm right but we should be able to read the correct level anyway (like what i experience now and like it is on Sansa AMS "v1")
20:22:40 Join ivortheengine [0] (~ivortheen@
20:23:09topikyeah, my fuze doesn't do this jumping. it's a "shame" you're a perfectionist :)
20:23:09pamaurygevaerts: I'm reaching 21MiB/s with isochronous transfer currently
20:23:20gevaertspamaury: with correct data? :)
20:23:22topik(fuze v1)
20:23:32 Join dfkt_ [0] (dfkt@unaffiliated/dfkt)
20:23:43pamaurygevaerts: I don't do anything with the data (not sure the processor would follow) but the controller reports not crc error
20:23:47funmantopik: well there is another problem : charging is very slow afaict
20:24:02 Quit dfkt (Disconnected by services)
20:24:05 Nick dfkt_ is now known as dfkt (dfkt@unaffiliated/dfkt)
20:24:28topikyes, OF seems quicker. would even be quicker if it didn't leave the screen on
20:24:39pamaurygevaerts: What is the maximum possible speed on one endpoint ? I can't manage to go any further, I have minimum interval, maximum packet multiplier, maximum packet size
20:26:11gevaertspamaury: I'd say 1024 bytes * 3 per microframe * 8192 microframes per second
20:26:26gevaertsi.e. 24MB/s
20:26:39gevaertsA bit weird that you only get 21
20:26:41pamauryI'd say the same
20:26:56pamauryahhh, I think I know why
20:27:02pamaurygive me one minute
20:27:31gevaertsOf course you could make two such endpounts to get 48MB/s :)
20:27:53shaiHi everyone :) I've tried to upgrade my RB to the Current Build and for the first time after many times in the past that I've done this, it won't load the image at the boot loader. It says something about "Can't load rockbox.ipod: Read failed (image)" ; it also says to reboot and go into disk mode, which I did and tried to use rbutil again and again but it keeps failing to load the image
20:27:57 Join Buschel [0] (
20:27:59shaiWhat am I missing?
20:28:13pamaurygevaerts: yep, I get 24M/s, I had inserted a 100ms delay after sending to be sure not to cancel too early
20:28:25gevaertsah, ok
20:29:02pamauryNow, I'm not sure that the processor can process data at such a rate but at least I reached the maximum :)
20:29:39gevaertsDoes it have to? Enable DMA on ATA and write it to disk immediately :)
20:29:59pamauryhehe, and to process audio ?
20:30:21pamauryusb audio protocol is not so well designed, you have to manually process the data
20:30:31shaiI also tried to load the Rockbox Stable and made sure it isn't loading from cache but still won't load.
20:30:33 Quit fyrestorm (Quit: Ur skills' fireproof like a wooden panel -- U got feds talking leet on your IRC channel!)
20:30:34shaiAny ideas?
20:30:51 Quit punkt (Remote host closed the connection)
20:32:24pamaurygevaerts: but the amount of code to have both queueing and "repeat" mode with usb arc plus all changes to code and other drivers is quite important
20:33:12 Join kramer3d [0] (~kramer@unaffiliated/kramer3d)
20:35:22gevaertsshai: I'd give my standard advice for all weird problems: "check the filesystem"
20:35:34gevaertspamaury: well, audio doesn't need 24M/s
20:35:58pamaurygevaerts: true, but it needs most of the code I have written :)
20:36:23shaigevaerts, oh .. didn't really consider that at all .. I'll try it now
20:37:46 Quit phanboy4 (Read error: Connection reset by peer)
20:41:32 Join anubisfremen [0] (~c18cf902@gateway/web/freenode/x-mwwxxuxmjldmmyxt)
20:42:01 Nick anubisfremen is now known as Guest19550 (~c18cf902@gateway/web/freenode/x-mwwxxuxmjldmmyxt)
20:42:01 Quit Guest19550 (Client Quit)
20:42:45 Part naag
20:43:42shaigevaerts, "Windows has checked the file system and found no problems."
20:44:07shaiSo is it time to buy a new iPod?
20:44:36funmanshai: is .rockbox/ folder present?
20:44:56gevaertsshai: much too early to say that...
20:45:24shaigevaerts, Thanks :)
20:45:33 Join TillW [0] (
20:45:35funmanand .rockbox/rockbox.ipod has a correct size ?
20:45:54shaifunman, 685k (not sure what the correct size is)
20:46:03gevaertswhich ipod is it?
20:46:08funmanaround this
20:46:14shaigevaerts, 5th gen 3gb
20:47:05shaiany other thoughts?
20:47:26funmanis it the 5th gen which has 2 different memory sizes?
20:47:35shaihmm... donno :(
20:47:36gevaertsyes, but that's irrelevant
20:48:42wodzI would like to ask coldfire skilled people to review FS #11153
20:49:33funmanwodz: i think that is amiconn or mcuelenaere
20:49:42shaigevaerts, disk mode works as expected. I can copy files to and from the iPod.
20:50:00funmanshai: do you have a weird partition layout?
20:50:08shaifunman, no.
20:50:14shaiJust 1 H: disk
20:50:25wodzamiconn: ping
20:50:31shaiI mean, 1 partition which is labeled H: in Windows
20:51:00gevaertsshai: I'd try removing (or moving away) the entire .rockbox directory and reinstalling
20:51:15funmantopik: how does the battery graph look like ? (1st screen)
20:51:18shaigevaerts, Ok ... I can do that with ease? Just rename it?
20:51:27gevaertsthat should be enough, yes
20:51:35shaigevaerts, Can Windows handle hidden partitions?
20:51:44shai.rockbox is hidden, right?
20:51:53gevaertsnot actually
20:51:54evilnick_BIt's a hidden folder, not a partition
20:52:06shaiI meant directory
20:52:08funmanand it's hidden by common convention on UNIX file browsers
20:52:09topikfunman: pretty much like on my fuze
20:52:13funmanno special attribute
20:52:24funmantopik: i mean when charging
20:52:35funman(after sometime)
20:52:51funmanis it slowly increasing or showing peaks (85%-96%)
20:53:09topikjust plugged it in again
20:53:10gevaertsshai: maybe it's hidden, if so either set explorer to show it anyway, or rename it from a cmd window
20:53:41topikthe sample rate might not be high enough to show the peaks
20:53:42shaiSweet! I can now go out with my friends and listen to some trance :) it is fixed!
20:53:59shaiAll my settings are gone, but who cares :)
20:54:10shaigevaerts and funman thank you guys!
20:54:12gevaertswell, they're still in the old directory
20:54:15funmanshai: what was the problem?
20:54:26gevaertsso you can copy them back if you want to (same with themes)
20:54:27shaifunman, I renamed the .rockbox and reinstalled rockbox
20:54:45shaigevaerts, when I come home :) thanks!
20:55:06shaiThey are all waiting for me in the city, so I gtg! thanks a bunch guys!!!
20:55:11shaiHave a good one!
20:55:26gevaertsfunman: "Read failed (image)" means that the bootloader can't read the entire file (i.e. it reads less than expected). It has by then correctly read the first 8 bytes, so the file is definitely there
20:55:41gevaertsSo I assume some sort of corruption that chkdsk didn't see
20:56:58topikfunman: after being plugged in for some minutes, the graph is pretty much flat (as it would be if charging would be very slow). no peaks.
20:57:20funmanand still the battery level jumping in the 4th screeN?
20:58:35 Join captainkewllll [0] (~2669ecc2@gateway/web/freenode/x-kdtoielnihcuorkq)
20:59:28topikyes, 80/86-ish
21:00:00***Saving seen data "./dancer.seen"
21:00:25topikthe voltage on the 2nd screen is a bit jumpy too. could that be related? is yours?
21:02:32 Join phanboy4 [0] (
21:03:39 Join tenfoot [0] (
21:04:43funmanno, difference is about 5mV
21:04:54funmanthe graph & percentage are derived from that reading afaik
21:05:06 Quit TillW (Remote host closed the connection)
21:05:21topiki think the graph is jumpy too, but the samplerate is much lower than the 4th screen so it's less noticable
21:05:49 Quit Battousai (Ping timeout: 245 seconds)
21:05:49 Quit ved (Ping timeout: 245 seconds)
21:06:29tenfootHi - I've just tried the fuzev2 firmware from the forum thread - it works OK.
21:06:46tenfootIs there anything else I can do to help get rockbox on the fuze v2?
21:07:59funmantenfoot: thanks for testing, there's nothing more to do atm if you can't code
21:09:44 Join Battousai [0] (~bryan@gentoo/developer/battousai)
21:09:51tenfootI can code - lots of experience with ARM & embedded SW - just starting to dig around the rockbox code now
21:10:19funmantenfoot: then you're welcome :)
21:10:24 Join spaax [0] (
21:10:32funmankugel is working on the scrollwheel atm
21:10:53funmani had a look at FM radio but without luck so far
21:11:14funmanboth things require reverse engineering on the OF (Original Firmware), where i can give plenty advices
21:12:01tenfootso the v2 is very different from the v1 then?
21:12:34 Join ved [0] (ved@
21:12:42 Part watto
21:12:57funman"very" is a strong word
21:13:13funmanit's different from fuzev1, but much more similar to fuzev1 than to ipod video for example
21:14:37funmanthe differences (so far) are: IRAM size (bigger), i2c registers (as3543 unlike the as3517 in fuzev1, but still very compatible), CGU_PROC & CGU_PERI registers, the SD controller, button (and scrollwheel) bits, how the LCD bits are accessed (although the same controller is used)
21:16:04tenfootOK. So we need to find out how the OF drives these? I guess that means disassembly?
21:16:54funmanyes, afaik fuzev2 disassembly is needed for scrollwheel and FM
21:17:54funmanthen there are the common bits between fuzev2/clipv2/clip+ : SD controller (writing doesn't work atm but flyndice is working on it), charging (I have a patch ready), and CGU_PERI/CGU_PROC for dynamic cpufreq
21:18:08funmanand writing an USB driver
21:18:15 Quit Buschel (Ping timeout: 264 seconds)
21:18:54funmanwe have linux patches for as353x SoC, which shares some similarities with the SoC inside those devices
21:19:30 Join clauwn [0] (
21:19:35funmannamely the as3543 audio/pmu chip and the synposys SD controller, else the SoC is a as3525 (with datasheet publically available from austriamicrosystems)
21:19:35clauwnhey there, i
21:19:56 Join TillW [0] (
21:19:57clauwni just installed rockbox onto my ipod mini 2gen 4GB but now i can't shut it down
21:20:19bluebrotherdefine "can't shut down".
21:20:29clauwni'm holding play/pause
21:20:33clauwnnothing happens
21:20:34bluebrotheryou need to hold Play for a while to shut it down. It's described in the manual.
21:20:49clauwni've been holding it for a minute or so
21:20:54tenfootGuess I should go and absorb the datasheet then.... I'll try to get my head around the fuze v1 first to get an idea about how it all fits together
21:21:09bluebrotherdo you have hold enabled?
21:21:33 Join xiainx [0] (
21:21:34clauwnif i just click it
21:21:39clauwnit says, nothing to resume
21:21:40CIA-5New commit by 03funman (r25403): Make sure FM Tuner is stopped at init ...
21:21:48 Quit einhirn (Read error: Connection reset by peer)
21:21:51clauwnand if i hold it, it does not shutdown
21:22:23bluebrotherdid you install a current build or the latest release?
21:22:26clauwni just pulled the power cord
21:22:31clauwnit shut down
21:22:44clauwncharging prevents shutdown?
21:22:56funmanit can't charge while powered off
21:23:12clauwnone needs to know that...
21:23:28clauwnbut doesn't that make charging slower?
21:23:45TillWdidn't he just say the key to that
21:23:52funmanno, if it's off it won't charge at all
21:24:20funmantenfoot: i think the easiest is to pick up an area you want to work on
21:24:36clauwnhow do i kill games or apps?
21:25:01funmanclauwn: everything you need to know is in the manual
21:25:21bluebrotherclauwn: the Ipod doesn't shut off while charging. That's even true for the apple firmware. It just hides that fact from you.
21:25:36 Join bmbl [0] (
21:25:36 Quit bmbl (Changing host)
21:25:36 Join bmbl [0] (~Miranda@unaffiliated/bmbl)
21:25:38clauwnbluebrother: thank you guys
21:25:48clauwnand thanks for this nice piece of software :)
21:25:49funmantenfoot: the fuzev2 port is already quite advanced, you should talk with kugel if you want to work on the scrollwheel
21:25:58funmantalk with me if you want to work on the FM
21:26:19funmantalk with flyndice if you want to work on the SD controller (which controls internal storage and the µSD slot)
21:27:30funmanI started to look at USB for the fuzev1 but i don't know much about USB
21:28:27clauwnwhat's WPS?
21:28:39funmantenfoot: i have linux patches (written by AMS) for as352x and as353x, i can send them to you if you're interested. The person who gave them to me just asked that they're not made available on the wild internet, but after all it's GPL code
21:28:43 Join planetbeing_ [0] (
21:28:44funmanclauwn: While Playing Screen
21:29:27clauwnfunman: thanks :)
21:29:38clauwni really like you guys :)
21:30:21tenfootI guess scrollwheel & SD are what interest me most so I can get music playing, but if there's already work under way then there's not much point in treading on other peoples toes
21:31:02tenfootit's going to take me a while to get my head around it all before I can do anything useful anyway....
21:31:38funmanok :/
21:33:13 Quit mischasworld (Read error: Connection reset by peer)
21:35:38 Quit dfkt (Read error: Connection reset by peer)
21:36:02clauwnthanks for your help guys
21:36:56 Quit Strife89 (Quit: Going home.)
21:39:07 Part clauwn
21:42:42 Part dockimble ("Leaving")
21:42:48 Join dockimble [0] (~dockimble@
21:43:09funmandomonoky: test somewhere else please
21:43:21bluebrotherdockimble: ok, now that you've tested if anyone is annoyed by stupid test in on-topic channels, how can one help you?
21:43:46dockimbleno problem
21:43:57dockimblemy connection is clearly no longer cutting out
21:44:28 Join clauwn [0] (
21:44:49clauwnis there any trick on getting it to connect to my pc?
21:45:00clauwni'm using Debian GNU/Linux
21:45:09bluebrotherclauwn: just use the cable?
21:45:38bluebrotherhowever, you might want to disable HID support. Its known to cause problems with Mac OS X.
21:45:47clauwnbut it sometimes starts with the apple logo, then there is rockbox, then there is "multimedia mode", "do not disconnect", "ok to disconnect"
21:46:11bluebrother"do not disconnect" and "ok to connect" come from the apple firmware.
21:46:40bluebrotherthe "multimedia mode" is Rockbox USB mode. As it defaults to present it as UMS and HID device that "multimedia mode" simply displays the current HID mode.
21:47:07clauwnbut do not disconnect and ok to disconnect should not follow each other quickly
21:47:29archivatorDo we even have a usb stack for iPod Mini?
21:47:33bluebrotherprobably, but that's not a Rockbox issue
21:47:39archivatorI thought it was Video and above..
21:47:43bluebrotherarchivator: yes. I'm using it since its first days
21:47:54archivatorsomething new every day..
21:48:36bluebrotherit also works on other PP devices. Like e200 and mrobe 100
21:48:38clauwnbluebrother: solved
21:48:40 Join planetbeing__ [0] (
21:48:46clauwni just need to start the rockbox installer utility
21:48:48junkY_Sanfunman: is it enough the just replace the rockbox.sansa after a rebuild of the patched source?
21:48:52clauwnand it will stay connected
21:49:02bluebrotheroh, and the beast ;-)
21:49:06funmanjunkY_San: yes, although the plugins might be out of date wrt rockbox.sanas
21:49:25bluebrotherclauwn: that's strange. Rockbox Utility is not related to *using* Rockbox at all.
21:49:29funmanif you want to update the plugins you can do make PREFIX=/mnt/clip install and it'll install everything in /mnt/clip/.rockbox
21:50:14clauwnbluebrother: i thought of that too :)
21:50:39clauwnbluebrother: but it seems that this works quite well
21:51:15*bluebrother not sure what "this" is
21:52:46 Quit planetbeing_ (Ping timeout: 276 seconds)
21:53:25 Join Buschel [0] (
21:53:41CIA-5New commit by 03bluebrother (r25404): Update german Rockbox Utility translation. ...
21:54:15clauwnwoo, it plays my ogg files :)
21:55:34clauwnbluebrother: but charging during playback works?
21:56:56bluebrotherclauwn: if you hold Menu during USB connect you can use the player and charge it. The problem is that charging is very slow, so you might consider that rather "not working properly". The problem is that the Ipod is drawing too less power.
21:57:18clauwnbluebrother: i've got an usb charger
21:57:38clauwna device that plugs into the wall and has an usb output
21:57:41 Quit liar (Ping timeout: 276 seconds)
21:57:46bluebrotherwell, not exactly sure about that but afaik it's the same issue here.
21:58:14clauwnso the ipod itself is not taking enough power to charge fast?
21:58:20gevaertsYes, it's the same. We hope to fix that soonish, but currently ipods don't charge well from the USB connector in rockbox
21:58:53clauwnso i should probably stop the playback and just charge it
21:59:27clauwnuuh, i like the fade out :)
21:59:52 Quit wodz (Quit: Leaving)
21:59:54clauwnhow do i completely stop playing?
21:59:55 Quit dockimble1 (Quit: WeeChat
22:00:05 Quit planetbeing__ (Read error: Connection reset by peer)
22:00:57bluebrotherclauwn: hold Play
22:01:25clauwnto shut down?
22:01:39junkY_Sanfunman: got it patched and loaded, what should i see in the debug -> battery screen?
22:01:51 Join planetbeing_ [0] (
22:01:53clauwnah, i see
22:01:58funmanjunkY_San: see if the voltage jump around 2 values or is flat when charging
22:02:12junkY_Sanit's flat, 4.117
22:02:21junkY_Sanah wait
22:02:23bluebrotherclauwn: I can also highly recomment the fine manual ;-)
22:02:32clauwnbluebrother: i'm reading it..
22:02:32junkY_Sanjust changed to 4.129
22:03:25junkY_Sandisconnecting decreases it by 0.005
22:04:04funmancan you leave it charging for some minutes and check the graph ?
22:04:30junkY_Sanit's playing in the background, that's ok?
22:05:07bluebrotherclauwn: great :)
22:05:12topikfunman: any idea how low battery voltage can go before battery level is near 0 ?
22:05:57junkY_Sanbtw what's the icon in the upper right corner with that boxed arrow? while it's active i can't press a key
22:05:59topikor does that "just" require discharge curve calibration
22:06:02funmantopik: nope, i just read that powermgmt-as3525.c is filled with "TODO"
22:06:18funmanjunkY_San: storage active
22:06:40junkY_Sanah ok, so it's just busy while reading the file
22:07:01topik3.962 represents 79% according to rb. it start at 4.084 or something
22:07:29topikso i wonder if it will shut down at 0% level or 0 voltage
22:07:40funmani think it's 0%
22:07:53 Join TheSeven [0] (~theseven@rockbox/developer/TheSeven)
22:08:06funmanpowermgmt-as3525.c says 3.4V is the "dangerous" voltage
22:08:10 Quit xiainx (Ping timeout: 268 seconds)
22:08:44topikif level is linear, it will nowhere be near 3.4V
22:09:06funmanwhat do you mean?
22:09:51topikhmm, i thinking i'm counting wrong. 3.4 will be 0%
22:10:04junkY_Sanit's been 4.132 for some minutes now
22:11:13junkY_Sanmh i think it's fully charged
22:11:26funmanis it ? what does the OF say ?
22:11:59 Join xiainx [0] (
22:12:58 Quit tenfoot (Quit: CGI:IRC (Ping timeout))
22:13:49funmancan you let it discharge a bit and try again?
22:14:10junkY_Sanyes i take it to gym now be back in 2h
22:14:30 Quit Buschel (Ping timeout: 246 seconds)
22:15:25junkY_Sanbtw it can't be turned off while plugged in
22:15:27topikdischarged some more (80%) it is jumping 74/80 while charging
22:15:57funmanperhaps we should only keep the 2nd read
22:16:13funmanjunkY_San: yep, because it can't charge while it's off
22:16:22topikit alternates two different reads?
22:16:54topikfrom two different locations i mean
22:17:26 Join dockimble1 [0] (~dockimble@
22:17:56funmancan you try ?
22:18:09funmannope i think it should only read from 1 place
22:18:25topikrevert the previous patch?
22:18:34funmanno leave it applied
22:18:39junkY_Sani'll try it when i'm back
22:18:42junkY_Sancu later
22:19:04funmanit'll do 2 reads instead of one, and discard the first value
22:20:54topikerror: conflicting types for _adc_read
22:21:35 Quit amiconn (Disconnected by services)
22:21:37 Join amiconn_ [0] (quassel@rockbox/developer/amiconn)
22:21:59 Nick amiconn_ is now known as amiconn (quassel@rockbox/developer/amiconn)
22:22:22 Quit kaniini (Ping timeout: 258 seconds)
22:22:28funmanhm add "unsigned short _adc_read(int);" before the first adc_read
22:22:46 Quit pixelma (Disconnected by services)
22:22:46 Join pixelma_ [0] (quassel@rockbox/staff/pixelma)
22:23:05 Nick pixelma_ is now known as pixelma (quassel@rockbox/staff/pixelma)
22:23:42topikcompiling now
22:26:29 Join junkY_ [0] (
22:27:04 Join kramer3d_ [0] (~kramer@unaffiliated/kramer3d)
22:28:18topikbattery debug screen 2 and 4 still do the alternating
22:28:34 Quit kramer3d (Ping timeout: 265 seconds)
22:28:45funmandamned :o
22:29:30 Quit junkY_San (Ping timeout: 252 seconds)
22:29:35topiki apologize on behalf of my clip+. it should listen to you better
22:30:04funmantopik's clip+: now tell us how you want to be charged! :P
22:32:22 Quit funman (Quit: free(random());)
22:33:15 Quit {phoenix} (Remote host closed the connection)
22:33:24topiksmoothly please
22:34:20dockimbledoes rockbox on the clip+ charge at the moment?
22:34:43dockimblebut it will charge
22:35:12Stephen__i think so yes, i jut use the OF to charge
22:35:28dockimblei'm gonna leave it overnight, we'll see i suppose
22:36:10 Part spaax
22:42:20 Quit dockimble (Quit: Leaving)
22:46:18 Part clauwn
22:48:52 Join planetbeing__ [0] (
22:49:22 Quit Zagor (Quit: Clint excited)
22:51:10 Quit kramer3d_ (Ping timeout: 258 seconds)
22:51:12 Join FlynDice [0] (
22:52:32 Quit planetbeing_ (Ping timeout: 276 seconds)
22:52:49*bluebrother just sent out a call for translators and wonders how many translations will get contributed the next days
22:56:37FlynDicefunman:(logs) I have not tried PIO yet for SD writes. I haven't been able to do much of anything the past couple of days as I was away on a trip and my laptop decided to die on me :(. I've got to take care of that and then I can dive back in.
22:57:58 Join Strife89 [0] (
23:00:03***Saving seen data "./dancer.seen"
23:00:14 Join calman [0] (
23:00:32 Part froggyman
23:01:38 Quit calman (Client Quit)
23:05:44 Quit dockimble1 (Quit: WeeChat
23:09:06 Join blairb [0] (
23:09:20 Quit efyx (Remote host closed the connection)
23:11:34 Join S_a_i_n_t_ [0] (S_a_i_n_t@
23:13:05 Quit S_a_i_n_t (Ping timeout: 276 seconds)
23:15:37 Join anewuser [0] (anewuser@unaffiliated/anewuser)
23:18:19 Quit bmbl (Quit: Bye!)
23:18:27 Quit xiainx (Ping timeout: 268 seconds)
23:19:23 Join kaniini [0] (
23:22:09 Join xiainx [0] (
23:25:44 Quit Kitr88 (Ping timeout: 260 seconds)
23:26:31 Join Kitr88 [0] (
23:29:40 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
23:33:19 Quit evilnick_B (Quit: Page closed)
23:40:41 Join Ziphilt [0] (
23:40:56 Join tenfoot [0] (
23:42:48 Quit tenfoot (Client Quit)
23:43:33 Quit domonoky (Read error: Connection reset by peer)
23:47:55pamauryFlynDice: did you have a look at the sd code ?
23:49:16 Quit Ziphilt (Ping timeout: 265 seconds)
23:50:02 Join planetbeing_ [0] (
23:52:01 Quit planetbeing__ (Ping timeout: 248 seconds)
23:52:46 Join Ziphilt [0] (
23:53:39 Nick fxb__ is now known as fxb (
23:58:29 Quit Adubb (Read error: Connection reset by peer)
23:58:54 Join Adubb [0] (~aldubuc@

Previous day | Next day