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 2014-09-11

00:00:01[Saint]TheSeven: debian
00:00:16 Quit [Franklin] (Ping timeout: 245 seconds)
00:00:49ZincAlloyI didn't use all that many graphics. I guess I should have made CabbieV2 as simple as that :D
00:01:01[Saint]Hmmmm. Odd. I did keep some local changes in my tree that shouldn't affect this _at all_, but I'll pull in the patch set in an absolutely clean tree and try again.
00:01:12[Saint]Its quite possible I ballsed something up merging.
00:01:29 Join [Franklin] [0] (
00:01:37[Saint]ZincAlloy: are you using a backdrop? It looks to me like you could very easily not.
00:01:45ZincAlloydozens of different icon states on a gradient backdrop are a porting nightmare
00:02:08[Saint]I hated you for that backdrop for so long...
00:02:21ZincAlloyI hated myself for that backdrop
00:02:35ZincAlloybut it did look nice..
00:02:35[Saint]It looks great. But the issue you outlined.
00:02:38[Saint]Yes. Very yes.
00:03:05ZincAlloyExcept on Clip Zip. That screen makes everything look ugly.
00:03:33[Saint]Only other thing I would suggest is using a UI viewport so the menu text isn't hard up against the screen edge.
00:03:35ZincAlloyand yes, I'm using a backdrop for rockdroid
00:03:43ZincAlloyit makes everything so much more convenient
00:03:54[Saint]right, there's a metric shittonne of savings available if you get creative.
00:03:57ZincAlloyI am using a UI viewport...
00:04:28[Saint]Huh...odd. Then why is the menu text hard up against the left edge? Looks kinda weird.
00:05:11ZincAlloyyou're right. I should move it a little so taht it line up with the clock
00:06:30[Saint]It makes a scrollbar a _lot_ easier to use if its enabled by the user as well.
00:06:45[Saint]touchscreen edge detection on some screens is kinda flakey.
00:07:00ZincAlloyspeaking of scrollbars: are they themeable?
00:07:14[Saint]Not trivially.
00:07:56[Saint]IIUC, to do that, you need to reimplement the lists with that weird slightly unfinished skinned lists stuff.
00:08:14[Saint]IIUC, gevaerts is the only one to ever successfully master that.
00:08:15ZincAlloysounds like a nightmare
00:08:19[Saint]It is.
00:08:48ZincAlloynot gonna bother then. custom scroll bar would be nice, but I can live without one
00:09:03 Quit amayer (Quit: Leaving)
00:09:28[Franklin]just color change ought to be trivial
00:09:47ZincAlloyit needs more than a color change
00:09:53[Saint]You'd think so, but, no. It isn't.
00:10:06[Saint]It uses the global foreground/background colors.
00:10:12[Franklin]just find the scrollbar code and change the... oh
00:10:24[Saint]Themeing, unless you've actually been there, is extrememly non-obvious.
00:10:43*[Franklin] doesn't want to even touch the theming language
00:11:15[Franklin]<shiftplusone> It looks like regex written by a drunk cat on a keyboard.
00:11:36[Saint]Yes. He's a good friend of mine. That amused me greatly.
00:11:53[Saint]That was said after I accidentally dropped a pastebuffer full of theme in #raspberrypi
00:16:48[Saint]ZincAlloy: this may interest you, please excuse the lack of pastebin:
00:16:49[Saint]# /* Set and check for a PLAY_STATE variable to allow us to always display the
00:16:49[Saint]# * correct PAUSE/PLAY state icon when seeking. */
00:17:05[Saint]This fixes a very real problem with play state icons and seeking.
00:17:22[Saint]ZincAlloy: may I also suggest making "stop" long press on play/pause?
00:17:49ZincAlloythat's a good idea!
00:17:56[Saint]ZincAlloy: if you're not sure what's going on in the code snippet, gimme a yell, but if I know you well enough that should be within your grasp.
00:18:16[Saint]Its saving a state variable to check what icon we _should_ display when seeking.
00:18:31[Saint]otherwise if you seek, you'll get no playback state icon.
00:18:36[Saint]which sucks a giant fuck.
00:19:31ZincAlloyactually the play state doesn't bother me much. I did notice it, though.
00:19:49[Saint]It bothered the fuck out of me for _years_.
00:20:02[Saint]Until one day, I shit you not, I /dreamed/ that code snippet up.
00:20:12[Saint]Then I promptly awoke myself at 5am and started writing.
00:20:29[Saint]code-dreaming, yo. I'm sure we've all done it.
00:20:43[Franklin]not herew
00:21:37foolsh[Franklin]: It will happen if you keep programming, I dream about math
00:21:42ZincAlloyisn't it just a workaround bug that needs fixing? or am I missing something?
00:21:52[Franklin]well it has... but nothing like that cat regex :)
00:21:53[Saint]getting offtopic, but I keep my phone right next to me so that if I wake up in the middle of the night with wild inspiration I can actually document it instead of just falling asleep again.
00:22:13[Saint]ZincAlloy: its technically behaving exactly as expected.
00:22:27[Franklin]how is this offtopic? it'll be of massive use to every theme "expert" after you
00:22:29[Saint]While you're seeking, it makes perfect sense for the play state icon not to show.
00:22:33[Saint]...because you're seeking.
00:22:47[Saint]and not in playback/stopped/paused.
00:22:49ZincAlloyit's true...
00:22:53[Saint]So we need to remember this state.
00:22:56*[Franklin] suggests that you have fastforward/rewind icons for play state
00:23:05[Saint]We can only do this on touch targets, though.
00:23:24[Saint][Franklin]: that just leads to an info double-up.
00:23:34[Franklin]a what?
00:23:42[Saint]the NEXT/PREV icons should change to FF?RW when long pressed.
00:23:58[Saint]No need to show it twice.
00:23:59ZincAlloythey should.
00:24:07[Saint]...yours dont?
00:24:12ZincAlloywhen you're pressing them you can't see them
00:24:27[Saint]right, but you know what you're doing.
00:24:34ZincAlloyat least not the one you're pressing
00:24:38[Saint]and you've got the progressbar for a visual indicator.
00:24:54ZincAlloywhich you could use for scrolling as well
00:25:07[Franklin]so when you press, the play state is a FF/RW icon
00:25:29ZincAlloyI guess everyone will expect long presses on prev/next to be used for ffwding/rewinding
00:25:41[Saint]its kinda a universal thing.
00:25:47[Saint]with touch UIs, anyway.
00:26:05 Join Riviera [0] (Riviera@2a03:b0c0:1:d0::10:b001)
00:26:17ZincAlloyit's exactly like players with buttons would behave
00:26:22[Saint]I hope this criticism comes across in the positive fashion that I intend it to.
00:26:38[Saint]I'm not so good with...people stuff. Y'know.
00:26:53ZincAlloyoh, I don't see it as criticism
00:27:00ZincAlloyit's interesting input
00:27:40ZincAlloybut the issue doesn't bother me at all
00:27:52ZincAlloyquite unlike the cover art issue :D
00:28:03 Quit ender` (Quit: First things first, but not necessarily in that order.)
00:28:40ZincAlloyI might have to make another version of the sbs with buttons on the bottom
00:29:05[Saint]Yeah, I think we should be doing something like buffer=(worst case fullscreen ARGB bitmap)*4 or so.
00:29:15ZincAlloyI don't really click the tiny cover art to go back to the wps intuitively
00:29:34ZincAlloythat should be plenty.
00:29:52ZincAlloyunless you go crazy with fonts I guess
00:30:28[Saint]That would result in a much smaller buffer on the lowres targets, if I'm not mistaken.
00:30:39ZincAlloysomeone will hate you for that :D
00:30:40[Saint]8MB of a 240x320 target seems a bit extreme.
00:30:48[Saint]*for a
00:31:23ZincAlloywhen people are having issues with it you could still change it
00:31:39ZincAlloybut that doesn't seem all that likely
00:32:20[Saint]In all honesty, I don't think anyone saw this coming.
00:32:38[Saint]When RaaAoA was born, 480x800 was a very extreme resolution.
00:32:49[Saint]Now its considered laughable small.
00:33:05ZincAlloyyeah, things are changing rapidly
00:33:45[Saint]Wait til we get "RaaAoA doesn't work on my Moto360's round display"
00:33:45ZincAlloyfunny that you didn't run into this issue before
00:34:16[Saint]I've generally been fairly conservative with my themes.
00:34:17ZincAlloyoh yeah. round jpgs. far out
00:34:40ZincAlloywell.. this rockdroid is me being conservative and lazy at the same time
00:34:44[Saint]The themes where I have gone really extreme, the targets had ~60MB of buffer to play with.
00:35:15[Saint]But on a hosted target, we can't just suck up all the remaining RAM for buffer.
00:35:39 Quit bertrik (Remote host closed the connection)
00:35:43ZincAlloyat least we shouldn't
00:35:49[Saint]Well, we could, but I suspect the host would kill it off frequently.
00:35:56ZincAlloybut I guess the ones with huge screens should have plenty of ram
00:37:05[Saint]I'm hesitant to make such a sweeping change to the buffer allocation without making it resolution dependent, though.
00:37:20ZincAlloywow. I only used six bitmaps for rockdroid. Now that's efficient
00:37:29[Saint]Allocating ~32MB for a 1024x1280 target is perfectly reasonable.
00:37:38[Saint]But, on a 240x320 target, its insane.
00:38:10ZincAlloyresolution dependent buffers is exactly what is needed then
00:38:11[Saint]Targets at that resolution are likely to have less than 512MB available for the entire system.
00:38:52[Saint]And Android itself is going to want about two thirds of that.
00:39:01ZincAlloymy bitmaps for the wps are only 3.2mb
00:39:29ZincAlloyanother 2.8 for the menu backdrop
00:39:43[Saint]all mine are 900kB
00:40:01ZincAlloyyou're too efficient
00:40:12[Saint]Never had anyone say that before. ;)
00:41:21ZincAlloyor maybe not. being that efficient is more work
00:42:09ZincAlloyanyway, backdrops plus large album art are to be expected
00:47:59 Join xorly [0] (
00:50:45ZincAlloyI've changed the UI viewport and added a stop function to the play/pause button as per your suggestions
00:51:02ZincAlloythanks a lot for your help. I highly appreciate it
00:53:10[Franklin]one word: kickstarter
00:53:36[Franklin]start a campaign for a "bounty fund" for all the features that no one wants to write
00:53:48[Franklin]number one: those stupid icons
00:54:18[Saint]That would be a breach of kickstarter's protocols.
00:54:27[Saint]You need to display a functional prototype.
00:54:33[Franklin]we have that
00:54:40[Franklin]click any link
00:54:41[Saint]You can't have a functional prototype of a suggestion no one has given you.
00:54:58[Saint]you would be selling _nothing_.
00:55:16[Franklin]any other crowdfunding sites
00:55:19[Saint]That hasn't stopped people abusing KS in the past, though...
00:55:54[Saint]I really don't think there's a market for this.
00:56:06[Saint]If people want to pay for additions, they go to the forum.
00:56:25[Franklin]there's no real escrow to make sure they pay
00:56:56[Saint]99.99% of the populous is going to have no fucking idea what Rockbox is, let alone have suggestions for a feature to add (and pay for it) to the software they don't know about.
00:59:53foolshDude, cheese cake
01:00:56 Quit xorly (Ping timeout: 272 seconds)
01:01:11ZincAlloycheese cake would be an awesome kickstarter project
01:02:52***Saving seen data "./dancer.seen"
01:04:39[Franklin][Saint]: indiegogo doesn't have any requirements that could disqualify rockbox
01:05:05[Saint]That still doesn't change the other glaring issue, it being a terrible idea.
01:05:12[Saint]Virtually no one knows we exist.
01:05:20[Franklin]we could get famous
01:05:28[Saint]So they're not going to have an opinion on adding features to the OS they don't know exists.
01:05:55ZincAlloywould they "get" rockbox at all?
01:06:48[Saint]I sincerely doubt it.
01:07:12[Franklin]why not try?
01:07:13[Saint]Why use <obscure app> when <host> already offers <functionality>.
01:07:27[Franklin]we get 91% if we don't reach the goal
01:07:57[Saint][Franklin]: Rockbox has _never_ been about widespread adoption or monotization, and I hope in my time, it never will be.
01:08:08[Saint]This would be turning it into something its not.
01:08:11[Saint]Something ugly.
01:08:49[Franklin]hmm... true
01:09:02[Franklin]just like linux :)
01:09:14[Franklin]though they do try
01:15:23ZincAlloya crowdfunded rockbox player could be nice, though
01:17:43[Saint]Its just not going to happen.
01:18:20[Franklin]There is beginning to be a void in the DAP market with apple leaving
01:18:26[Saint]No one is going to pay the amount we'd need to get the hardware off the ground when there are companies that have been doing this for a long time, are established, and can produce a higher quality build for less.
01:18:36[Saint]We couldn't even afford to get the SoCs we'd need.
01:18:51*[Franklin] wonders if SanDisk could fill that void
01:19:17[Franklin]the top-of-the-line DAP market
01:19:24ZincAlloya cooperation with sandisk would be great
01:19:31[Franklin]not the cheap, plasticky, DAP market
01:19:37[Franklin]they already dominate there
01:19:39[Saint]"Lyre" is a great idea, and it has philosophical merit, and good intentions.
01:19:40ZincAlloya lot of people buy their players to use them with rockbox already
01:19:50[Franklin]ZincAlloy, why not try it?
01:19:52[Saint]In practice, though, its going to fall on its ass _every_ time.
01:19:55[Saint]Due to cost.
01:20:06[Franklin]here comes kickstarter :)
01:20:08[Saint]We just couldn't afford to fab a decent quality player.
01:20:10ZincAlloyit might be interesting for sandisk
01:20:19ZincAlloythat would get them press
01:20:26*[Franklin] spams the sandisk live support
01:20:53[Franklin]yes, true
01:21:00[Franklin]we design, have them fab it
01:21:25ZincAlloydeliver it with a dumbed down version of rockbox that won't overwhelm the average user
01:21:40[Franklin]Rockbox is fine the way it is
01:21:49[Franklin]no command prompt yet
01:21:50[Saint]No. No it is not.
01:21:51ZincAlloyfor you and me, yes
01:22:02[Saint]It is in no way ready for the consumer market.
01:22:05[Franklin]what's hard about it?
01:22:08[Saint]No way. No how.
01:22:17[Franklin]installation? no... manual!
01:22:20ZincAlloythere are too many options
01:22:34*[Franklin] figured out what they meant his first time using Rockbox
01:22:42[Saint]Labyrinthine menus and options that Joe User won't know, understand, or comprehend.
01:23:00ZincAlloyshould be an easy fix
01:23:06[Franklin]then "if it ain't broke, don't mess with it"
01:23:21[Franklin]ZincAlloy, I think that would alienate a large portion of the devs
01:23:29foolshOh god no, just fork it and fuck off with the window-ized version of rockbox, I like it the way it is now
01:23:31[Saint]ZincAlloy: Every time we've tried to change the main menu ordering, it turns into a massive clusterfuck.
01:23:48[Franklin]"NO CONFIGURABLE SPINDOWN!?!?"...
01:23:49[Saint]I have never deeply regretted championing an idea in a FOSS community so badly.
01:23:55ZincAlloyit would just be a branch
01:24:08[Franklin]and how to switch?
01:24:14[Saint]About two years ago I tried to get behind a serious push to reorder the main menu.
01:24:18[Franklin]"Dumbed-down mode"?
01:24:31[Saint]And, I repeat, I have never regretted a decision in a FOSS community so badly before, or since.
01:24:36[Franklin][Saint], what did you want it to be?
01:24:45ZincAlloyjust install a non dumbed down build
01:25:00[Franklin]and then devs will have to maintain both branches?!
01:25:10[Franklin]as if one wasn't enough
01:25:50ZincAlloyI wouldn't go that far. just get one stable release and then forget about it
01:26:06[Franklin]ah... yes
01:26:11[Franklin]"Simple mode"
01:26:30[Saint]Reordering the main menu was a classic example of too many chefs spiling the broth. Too many hammers, and not enough nails, Insert other metaphor about relationship breakdowns here. Etc.
01:26:35ZincAlloya simple mode would make sense
01:26:48[Franklin]Just cut out about 90% of the settings options and get rid of the "incomplete" plugins, etc
01:27:02[Franklin]so no doom for ipod6g to avoid people saying "it don't work!"
01:27:03[Saint]At least one other person before me tried to seriously push for a main menu re-shuffle and failed miserably, too.
01:27:12[Franklin]how so?
01:27:13[Saint]It just disseminates into madness.
01:27:14ZincAlloyYou just gotta have a vision and be the steve jobs of rockbox, [Saint]
01:27:19[Saint]A giant cluserfuck of disagree,ent.
01:28:39foolshhmmm, A design contest could keep things better organized and those who want to work on it could do so Independent of whats "going on in the kitchen"
01:28:52[Saint]You can certainly see in some places of the menu someone _tried_ to order it based on what they thought was likely to be used more often.
01:28:57*[Franklin] smells forks
01:29:13[Saint]But then you can also see lots of "ahhh, fuck it" and just tagging on new items to the ends of lists because its easier to do.
01:30:20[Saint]While you are certainly free to fork Rockbox, its also a fucking terrible idea, like most forks are. WHen you take on a fork of a large project, you also take on the support and maintenance of it alone.
01:30:48*[Franklin] didn't say he would fork it
01:30:52foolshBut can't the theme engine do most of the necessary heavy lifting for something like this already?
01:30:53[Saint]All for a fucking shitty reason like UI changes?
01:31:04[Saint]foolsh: not really, no.
01:31:08ZincAlloyit's just an idea at this point. adding a dumb mode is another great idea
01:31:36[Saint]Its possible in theory.
01:31:38[Franklin]yeah, I don't see a reason the average Joe would want a dev build
01:31:44[Saint]Fucking nightmarish in practice.
01:31:52[Franklin]maybe for next release, just make a dumb mode
01:32:17ZincAlloyjust put together a nice .cfg that will make the player work intuitively for most people, get rid off many options and let joe average be happy with it
01:32:18[Franklin]and don't think about it for a year or so
01:32:34ZincAlloywouldn't be all that hard to do, would it?
01:33:20[Franklin]so, for next release?
01:33:26[Franklin]so they get 2048 :)
01:33:53[Saint]We already do that already, basically. Rockbox is configured entirely sensibly and regularly accessed features are readily available.
01:34:09ZincAlloyyeah. everything it needs is already there.
01:34:10[Saint]Joe Average just can't seem to comprehend not touching things they don't understand.
01:34:24[Saint]But, this isn't /for/ Joe Average.
01:34:25[Franklin]but then they set spindown to max... and then complain about the battery dying in 4 hours
01:34:32ZincAlloya dumb mode would just leave things away
01:34:52[Franklin]just like windows does when you try to access C:\winblows
01:34:58ZincAlloyit hasn't been for joe average so far
01:35:03[Saint]I can't argue against it being a reasonable idea, but at the end of the day, they're not the target market.
01:35:16ZincAlloybut it wouldn't be so hard to make it work for joe average
01:35:39[Franklin]no new features
01:35:43[Franklin]just taking some away
01:35:47ZincAlloythe only feature is less options
01:35:55[Franklin]"simple" mode or something
01:36:09ZincAlloyyeah. calling it dumb mode would be bad marketing
01:36:21[Saint]you could scatter ifdefs around and make it a compile time option *really* easily.
01:36:29[Saint]and the main menu is already configurable.
01:36:41[Franklin]#ifdef STEWPID_MODE...#endif
01:36:48[Saint]making it a dynamic option on the device would be rather a bit more complicated.
01:37:07[Franklin]no... compile time so Joe doesn't freak out about those extra options
01:37:47[Franklin]So, we need to know what it will be called
01:37:52[Franklin]Here's my ideas:
01:37:54[Franklin]"Simple mode"
01:37:57[Franklin]"Easy mode"
01:38:00[Franklin]"EZ mode"
01:38:05ZincAlloy"rockbox for dummies"
01:38:11[Franklin]"Light mode"
01:38:12foolshbut calling it "Enhanced mode" would really pull in them in
01:38:26ZincAlloyRockbox 3G
01:38:57[Franklin]No, Rockbox Lite will cause them to want to "upgrade" to rockbox full
01:39:13[Franklin]So something that makes it sound better than normal rockbox
01:39:22foolshIU mode
01:39:22[Franklin]Enhanced then
01:39:33[Franklin]Rockbox XS
01:39:37ZincAlloyRockbox Advanced
01:39:50[Franklin]And say it stands for Xtra-simple
01:40:08ZincAlloyRockbox 2.0
01:40:11foolshRockbox - Enhanced - Edition
01:40:20[Franklin]2.0?.... no
01:40:20[Saint]Pretty much nothing is going to work I don't think. CHanging Rockbox for the sake of fucktard users is a mistake I would think.
01:40:26[Franklin]there is a 2.0 already
01:40:53ZincAlloyoh, right, we're on 3.13
01:41:03ZincAlloy2.0 just has a nice ring. it sounds so advanced
01:41:04[Franklin]Rockbox 2000
01:41:13ZincAlloy2000 is so 14 years ago
01:41:14[Franklin]THE Rockbox 2000 :)
01:41:22[Franklin]THE Rockbox 3000 then :)
01:41:49foolshRockbox The Experience
01:42:07foolshlol that was aweful
01:42:44[Franklin]I'm leaning towards Rockbox XS
01:42:47[Franklin]It needs an X
01:42:54ZincAlloyXS sounds way too small
01:43:03ZincAlloymake it Rockbox 3XL
01:43:13[Franklin]No... just X
01:43:23[Franklin]X for Xtremely simple
01:43:24ZincAlloyno, wait..
01:43:29[Franklin]Rockbox Xtreem
01:43:31[Franklin]Rockbox Xtreme
01:43:43ZincAlloylike in Xtremely simple to use
01:43:52[Franklin][Saint], sorry :)
01:44:01[Franklin]spamming up the logs again :O
01:44:13ZincAlloyhey, we're being productive here
01:44:47[Franklin]OK, so here's the list: "Rockbox 4.0X" "Rockbox 3000" "Rockbox Simple"
01:45:06ZincAlloyRockbox X doesn't sound hip enough
01:45:20[Franklin]For release 4.0
01:45:32[Franklin]Maybe an X version each *.0 release
01:45:46foolsh[Franklin]: Loves those exxes doesn't he?
01:45:57ZincAlloywe need a completly unrelated name that expresses the characteristics of this simplified rockbox
01:47:24foolshAnyway it not going to happen unless you do it yourselves
01:47:58[Franklin]I will when next release comes out
01:48:11foolshkeep waiting
01:48:38foolshHow long has it been since 3.13?
01:49:09*[Franklin] thinks its long overdue
01:49:14[Franklin]so 4.0 it is
01:49:49foolshIts an arbitrary thing to be honest, you could just pick a revision and go for it
01:50:23[Franklin]and no feature freeze?
01:50:47[Franklin]Major Changes list is huge
01:50:50[Franklin]so yeah, it's time
01:51:02[Franklin]About 20 vs 4 for 3.12-3.13
01:51:19[Franklin]Though 3.0 was massive
01:51:37[Franklin]what, like 20 new plugins
01:52:55[Franklin]so it's time
01:54:12 Quit foolsh (Quit: foolsh)
01:57:30 Join Aldem [0] (~Aldem@unaffiliated/aldem)
02:00:57 Quit ZincAlloy (Quit: Leaving.)
02:05:14 Join drvink [0] (
02:06:52 Quit [Franklin] (Quit: Leaving)
02:10:23drvinkhi, a while ago i had mentioned adding an option to disable disk spindown while connected to usb to rockbox. i added it for the ipod 6g:
02:13:51 Quit jhMikeS (Ping timeout: 240 seconds)
02:25:03[Saint]Its unlikely to be accepted anyway, but its doubly unlikely to be accepted if it targets a single device for seemingly no reason at all other than personal preference.
02:25:22[Saint]However, I'm unsure if getting it accepted is an actual end goal you're considering or not.
02:25:46drvinki would like to get it accepted so that i don't have to maintain my own fork of rockbox just for one feature
02:25:56drvinkso whatever needs to be done to it to get it accepted i am willing to do
02:26:40drvinki just don't have the means to test it on other configs
02:26:42[Saint]Then it would need to target all devices that do HAVE_SPINDOWN (or whatever the define is, can't recall offhand), and have a manual entry.
02:27:47[Saint]The manual entry should also make it clear that in some instances the device very well may be actively discharging despite possibly claiming otherwise.
02:28:46drvinkis the define ALLOW_USB_SPINDOWN?
02:29:47drvinkthere also appears to be a DISK_SPINDOWN, but it's only referenced in some plugins
02:30:43[Saint]I think, though someone else might need to chime in here, that you'd also need to make it dependent on whether or not we differ to the OF for USB or not. Or do we use the OF by default for USB on all the SanDisk players now?
02:31:56[Saint]I'm not too knowledgeable on the current state of the SanDisk devices.
02:32:43[Saint]Oh, actually, derp.
02:32:53[Saint]They're all flash players anyway aren't they, heh.
02:33:41[Saint]There might still be some targets I'm unaware of that we do use the OF for USb.
02:34:21[Saint]I don't have the 1st gen iPod, can't find my iPod 2G, or a functional iPod Mini/Mini2G.
02:34:39[Saint]*disk based targets
02:48:10 Join ploco [0] (dce9b7f9@gateway/web/freenode/ip.
02:50:43fs-bluebot3Gerrit review #956 at : 3increase memorysize for Android build. by Chiwen Chang
02:52:38 Quit bcobco ()
02:53:15plocoI think ask 11MB on 240x320 target is ok.
03:00:01 Quit AlexP (Remote host closed the connection)
03:01:52 Join AlexP [0] (~alex@rockbox/staff/AlexP)
03:02:55***Saving seen data "./dancer.seen"
03:08:43 Quit DIY-Noob (Quit: - A hand crafted IRC client)
03:11:24 Join cmhobbs [0] (~cmhobbs@fsf/member/cmhobbs)
03:14:51drvink[Saint]: should i just wait for the review to see what else i ought to do?
03:18:49[Saint]ploco: that seems like a really weird way of doing it if you ask me.
03:20:01[Saint]when I looked to see how you were coming up with the numbers for each allocation, to say I was surprised, would be an understatement.
03:22:30plocoit's from RCC. with a smaller base. initially the base was 16mb
03:27:18plocobut I change it to any pattern, such as only increase size for targets more then 720p
03:28:52ploco16mb for 720p, 32mb for 1080p and above ..8mb for the rest etc
03:29:28 Join jhMikeS [0] (~jethead71@rockbox/developer/jhMikeS)
03:33:29ploco*but I can change ....
03:57:24 Join steffengy [0] (
03:58:35JdGordon8 + width/100?
03:58:54JdGordonwhy not just bump all to 32mb?
03:59:17JdGordonwhere is the ram required?
04:00:27 Quit steffengy1 (Ping timeout: 246 seconds)
04:07:18plocoJdGordon: [Saint] is worry about 240x320 target's ram availability.
04:08:07[Saint]its just..unnecessarily large for for the low resolution devices.
04:08:31ploco8 for < 540p, 16 for 720p, 32 for 1080+. how about this?
04:09:10[Saint]that seems fairly reasonable.
04:09:11JdGordonwait, this is the audiobuffer right? why can't that be allocated dynamically?
04:09:27JdGordondo an "8mb" build and autoallocate based on avilaable ram
04:09:58[Saint]there's no reason to have the audio buffer so large when disk acecss is essentially free.
04:10:13[Saint]and its audio/skin/everything buffer.
04:10:43plocotheme is sharing the same buffer. allocate right before audio part
04:10:48JdGordonaudio should be completly splt out for android
04:10:50[Saint]audiobuf is just the leftovers after skin allocations, like native targets.
04:17:37plocoupdated. this code looks soooo ugly :P
04:19:01plocoI knew you gonna say that
04:19:43[Saint]the prior example certainly /looked/ cleaner, but it was based on essentially arbitrary magic.
04:20:12[Saint]this at least makes reasonably sane decisions based on a metric that isn't pulled from the sky.
04:22:09plocothis buffer size issue actually remind me of g#681
04:22:11fs-bluebot3Gerrit review #681 at : 3buflib: Add malloc-based backend for application builds. by Thomas Martitz
04:23:13plocoto me, 681 is the better solution
04:29:28 Quit [Saint] (Quit: Going down for system update)
04:41:02 Quit Pessimis- (Ping timeout: 245 seconds)
04:44:08 Join [Saint] [0] (~saint@rockbox/staff/saint)
04:49:03 Quit amiconn (Disconnected by services)
04:49:03 Join amiconn_ [0] (amiconn@rockbox/developer/amiconn)
04:49:04 Quit pixelma (Disconnected by services)
04:49:04 Join pixelma_ [0] (quassel@rockbox/staff/pixelma)
04:49:06 Nick pixelma_ is now known as pixelma (quassel@rockbox/staff/pixelma)
04:49:08 Nick amiconn_ is now known as amiconn (amiconn@rockbox/developer/amiconn)
04:54:22 Join CyberYoda [0] (
04:55:07 Quit krabador (Quit: Take the time.)
04:55:45 Join Pessimis- [0] (Anon@gateway/shell/elitebnc/x-vymhgwfnwbxxgyyi)
04:55:50CyberYodaHi, I found a mistake in the wiki. It says I need to be added to the WikiUsersGroup before I can update the page. Who does that?
04:57:20ploco hmmm.. 681,682 need some modification when compile against latest source. 1 dircache.c is gone. 2. talk.c move_callback() called buflib_context_relocate() , which isn't available after the patch.
04:59:31[Saint]CyberYoda: I can.
05:00:34 Join chrisb [0] (
05:02:59***Saving seen data "./dancer.seen"
05:03:52[Saint]argh. whats going on...
05:05:57[Saint]CyberYoda: you _should_ have access now
05:06:04CyberYoda[Saint]: Thanks
05:07:36[Saint]Let me know if reality doesn't meet expectation.
05:08:07CyberYoda[Saint]: I was able to make the correction.
05:08:21[Saint]Excellent. Also, welcome.
05:09:39CyberYodaThanks. I'm working on an podcast program to interface with RockBox devices.
05:09:58CyberYodaI just found a mistake in the TagcacheDBFormat wiki.
05:10:22CyberYodaDoes anybody know if that DB is going to change in the next/future release?
05:13:20 Join Diaz_ [0] (0122605d@gateway/web/freenode/ip.
05:13:39Diaz_hi all
05:13:49CyberYodaHi Mr. Diaz_
05:13:53Diaz_is asf_read_packet demux?
05:14:35Diaz_since I would like to merge libwma to ffmpeg libavcodec
05:14:43Diaz_for fix point wma decoder
05:15:19 Quit TheSeven (Ping timeout: 272 seconds)
05:15:49Diaz_any input?
05:16:09 Join TheSeven [0] (~quassel@rockbox/developer/TheSeven)
05:22:12 Join Strife89 [0] (
05:24:27 Quit advcomp2019_ (Ping timeout: 245 seconds)
05:31:37 Join advcomp2019 [0] (~advcomp20@unaffiliated/advcomp2019)
05:56:33 Quit ploco (Quit: Page closed)
06:00:52 Quit KotH (Ping timeout: 240 seconds)
06:01:52 Join KotH [0] (
07:03:02***Saving seen data "./dancer.seen"
07:09:08 Nick DormantBrain is now known as SuperBrainAK (~andy@
07:25:22 Quit soap (Ping timeout: 240 seconds)
07:27:25 Join mortalis [0] (~kvirc@
07:28:11 Join soap [0] (~soap@rockbox/staff/soap)
08:25:28 Join ender` [0] (
08:32:50 Quit chrisb (Ping timeout: 255 seconds)
08:34:19 Join ygrek [0] (~user@
08:39:20 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
08:43:02 Nick SuperBrainAK is now known as DormantBrain (~andy@
08:49:37 Quit ygrek (Ping timeout: 272 seconds)
09:03:03***Saving seen data "./dancer.seen"
09:03:54 Quit CyberYoda ()
09:09:11 Join einhirn [0] (
09:10:59 Join wodz [0] (
09:12:56 Join petur [0] (~petur@rockbox/developer/petur)
09:19:44 Join rela [0] (
09:19:49 Quit rela (Changing host)
09:19:49 Join rela [0] (~x@pdpc/supporter/active/rela)
09:22:21 Join ygrek [0] (~user@
09:22:44 Join xorly [0] (
09:23:33wodzpamaury: ping
09:31:15pamaurywodz: pong
09:32:32wodzpamaury: In case I would like to stress bulk ep in hwstub what I need to change? I guess some descriptors to indicated bulk in/out eps?
09:33:50wodzand bulk ep handling in irq of course
09:43:16 Quit rela (Read error: Connection reset by peer)
09:48:57pamaurywodz: you need at least a few endpoint descriptors to describe the EP yes
09:49:18pamauryand of course some code to some something with the data
09:49:58pamaurylet me give an example
09:50:24wodzpamaury: What is truly amazing is that OF doesn't do any cache coherency for ep0
09:50:56pamaurywith dma ? that sounds wrong ^^
09:52:36wodzfor bulk ep OF discards d cache lines separately for every chunk
09:52:44wodznot the whole buffer at once
09:57:13pamaurywodz: something along these lines
09:59:01 Quit xorly (Ping timeout: 245 seconds)
10:00:06wodzpamaury: where do I set ep number? I want ep1 and ep2 (in and out respectively)
10:00:50wodzalso shouldn't .wMaxPacketSize be 512?
10:01:01pamaurywodz: in this code the usb_drv_request_endpoint is responsible for endpoint allocation
10:01:09pamaurywMaxPacketSize depend on link speed
10:01:44wodzah, now I see
10:01:56pamauryfor full speed it can be 8, 16, 32 or 64 and high speed 512
10:02:59wodzpamaury: There is no usb_drv_request_endpoint() in hwstub
10:03:30pamauryno, you need to implement it, because it needs to configure the endpoint
10:04:06wodzpamaury: I'd go for simple hardcoded allocation. I will configure endpoints in some init()
10:04:41pamaurydon't forgot to put epnum | 0x80 for IN endpoints in bEndpointAddress
10:05:04wodzgreat, thats what I was about to ask
10:05:09pamauryso ep1 has address 0x81 and ep2 has address 0x2
10:06:09wodzSo after this changes I should be able to do simple read/write to this endpoints with some libusb program, right
10:07:27pamauryassuming there some code to handle read/write yes
10:07:58pamaurybut yeah, the descriptors are enough to make the host happy
10:29:57 Quit [Saint] (Quit: Quit.)
10:30:22 Join [Saint] [0] (~saint@rockbox/staff/saint)
10:37:14 Join rela [0] (~x@pdpc/supporter/active/rela)
10:45:04 Quit alucryd (Quit: alucryd out)
10:49:36 Join alucryd [0] (quassel@archlinux/trusteduser/alucryd)
11:03:04***Saving seen data "./dancer.seen"
11:07:52wodz[Saint]: have you tested new usb driver on n2g?
11:08:59[Saint]just firing off a round of builds now.
11:09:06[Saint]can test later this evening.
11:13:07 Join rela_ [0] (
11:13:14 Quit Aldem (Read error: Connection reset by peer)
11:13:32 Join Aldem2 [0] (~Aldem@
11:16:09 Quit rela (Ping timeout: 252 seconds)
11:44:33 Quit Diaz_ (Ping timeout: 246 seconds)
11:45:52 Join ZincAlloy [0] (
11:50:29 Quit [Saint] (Ping timeout: 246 seconds)
11:57:52 Quit ygrek (Ping timeout: 240 seconds)
12:00:28 Quit alucryd (Quit: Alucryd out.)
12:00:38 Join alucryd [0] (
12:02:39 Quit alucryd (Changing host)
12:02:39 Join alucryd [0] (alucryd@archlinux/trusteduser/alucryd)
12:05:37 Quit alucryd (Quit: Alucryd out.)
12:06:51 Join alucryd [0] (quassel@archlinux/trusteduser/alucryd)
12:14:53wodzpamaury: heh, hwstub driver for rk27xx doesn't do cache coherency as well
12:18:15 Quit rela_ (Quit: Leaving)
12:23:46pamaurywodz: maybe the dma is cache coherent ? I think the cache is external to the cpu on this soc right ? Or maybe they disable caches completely ?
12:24:20 Join DIY-Noob [0] (b8a34d2d@gateway/web/cgi-irc/
12:24:59wodzpamaury: what you mean by 'disabled caches completely'?
12:25:36wodzpamaury: disabling cache has severe effect on performance so I assume it actually works
12:27:02wodzright, hwstub disables cache in crt0.S
12:27:39pamauryI mean the OF, that would be surprising I admit
12:28:03wodzpamaury: no way, with cache turned off the OF will literally crawl
12:28:29pamaurymaybe the hardware has some very poor kind of MMU which can tell whether memory is cached or not ?
12:28:44pamaurythe code might just be completely buggy as well
12:29:38wodzno that I know of. You can specify regions which are *not* cached but the smallest chunk you can declare this way is 32MB so you cannot put part of dram as uncached
12:30:07wodzthis is basically to set i/o space as uncachable (and cache controller address space as well)
12:32:45fs-bluebotBuild Server message: 3New build round started. Revision b7e3515, 253 builds, 28 clients.
12:35:51pamauryprobably the OF is buggy and somehow relies on the bug to never actually trigger
12:36:48wodzthis udc core drives me mad
12:43:04fs-bluebotBuild Server message: 3Build round completed after 587 seconds.
12:43:04 Quit fs-bluebot (Quit: connection lost?)
12:47:43 Join fs-bluebot [0] (
12:53:51 Join [Saint] [0] (~saint@rockbox/staff/saint)
12:58:57 Join ploco [0] (dce9b7f9@gateway/web/freenode/ip.
13:00:27plocosaw a hidden touch action {"cancel", ACTION_STD_CANCEL }, in skin_parser.c
13:01:07plocoso can use "cancel" for get 1 level back
13:02:23plocoeh.. i mean touch action "cancel" work just like the back key
13:03:05***Saving seen data "./dancer.seen"
13:05:55 Quit ploco (Quit: Page closed)
13:12:28 Join ygrek [0] (~user@
13:28:31 Quit jhMikeS (Ping timeout: 245 seconds)
13:29:48 Join rela [0] (
13:29:53 Quit rela (Changing host)
13:29:53 Join rela [0] (~x@pdpc/supporter/active/rela)
13:50:51 Quit rela (Read error: Connection reset by peer)
14:05:52 Quit Pessimis- (Remote host closed the connection)
14:11:56 Quit dys (Ping timeout: 260 seconds)
14:16:37 Join dys [0] (~user@2a01:1e8:e100:8296:21a:4dff:fe4e:273a)
14:18:16 Join foolsh [0] (
14:20:51 Join Pessimis- [0] (Anon@gateway/shell/elitebnc/x-bhmshuqswadujkes)
15:03:06***Saving seen data "./dancer.seen"
15:07:06 Join chrisb [0] (
15:08:45 Join rela [0] (~x@pdpc/supporter/active/rela)
15:23:06 Quit wodz (Quit: Leaving)
15:34:02 Quit cmhobbs (Ping timeout: 255 seconds)
15:40:05 Quit mortalis (Ping timeout: 246 seconds)
15:42:22 Join rela_ [0] (
15:44:55 Quit rela (Ping timeout: 252 seconds)
15:45:02 Join amayer [0] (
15:46:30 Quit amayer (Client Quit)
15:47:43 Join amayer [0] (
15:49:51 Quit Strife89 (Ping timeout: 245 seconds)
15:49:52 Nick sobukus_ is now known as sobukus (
15:50:04 Quit sobukus (Changing host)
15:50:04 Join sobukus [0] (~thomas@sourcemage/mage/sobukus)
15:53:48 Quit rela_ (Quit: Leaving)
15:54:05 Join rela [0] (~x@pdpc/supporter/active/rela)
15:57:42 Join rela_ [0] (
16:00:19 Quit rela (Ping timeout: 252 seconds)
16:20:44 Quit copper (Quit: ZNC -
16:23:45 Join copper [0] (~copper@unaffiliated/copper)
16:34:48 Join edhelas [0] (~edhelas@2001:bf0:c001:30:863a:4bff:fe85:8a3c)
16:51:36 Join krabador [0] (~krabador_@unaffiliated/krabador)
16:51:58 Join ikeboy [0] (
16:53:02 Quit ygrek (Remote host closed the connection)
16:53:07 Join ikeboy_ [0] (
16:53:23 Join ygrek [0] (~user@
16:56:23 Quit ikeboy (Ping timeout: 246 seconds)
17:03:10***Saving seen data "./dancer.seen"
17:15:21 Quit bughunter2 (Quit: Leaving.)
17:20:28 Quit ygrek (Ping timeout: 260 seconds)
17:22:27 Quit einhirn (Quit: Miranda IM! Smaller, Faster, Easier.
17:30:52 Join Diaz_ [0] (7a74df5e@gateway/web/freenode/ip.
17:31:06Diaz_hi all experts.
17:31:08Diaz_ff_imdct_calc cause SIGSEGV error
17:31:21Diaz_when I try to merge fixed point wma decoder from Rockbox to ffmpeg libavcodec,
17:31:29Diaz_does any one have ideas?
17:31:39Diaz_where I can check in advance?
17:31:49Diaz_the OS is Linux, CPU is mips
17:33:22 Join rela__ [0] (
17:36:32 Quit rela_ (Ping timeout: 264 seconds)
17:36:55 Quit edhelas (Quit: Quitte)
17:37:41 Quit mc2739 (Ping timeout: 245 seconds)
17:39:33 Join mc2739 [0] (~mc2739@rockbox/developer/mc2739)
17:58:49*foolsh is not an expert by no means
17:59:56foolshbut in linux when a program tries to access memory that is not own of accessible to it, the os will kill it
18:00:12foolshis not owned by it*
18:01:03foolshit sounds like that is what is happening IMO
18:04:19foolsh(in my opinion )
18:04:58Diaz_it seems not easily to debug in imdct functions
18:05:36 Quit petur (Quit: *plop*)
18:07:22foolshSIGSEGV means it's trying to access memory that is not owned or accessible to it and the os is killing it, its a linux thing
18:08:09foolshI think
18:08:17*foolsh doesn't use anything else
18:08:20gevaertsIt's not a linux thing
18:08:28gevaertsIt's a protected memory thing :)
18:08:49foolshGood to know
18:10:15*foolsh did give a disclaimer at the start, that he was no expert ;-)
18:10:35 Quit pamaury (Ping timeout: 246 seconds)
18:31:54 Join y4n [0] (~y4n@unaffiliated/y4ndexx)
18:35:50foolshgevaerts: g#954 sorry it's just a boring addition to the manual
18:35:53fs-bluebot3Gerrit review #954 at : 3Rockbox Manual - Gigabeat FX and Fuze+ Touchpad Config Appendix by Benjamin Brown
18:36:26 Quit krabador (Quit: Take the time.)
18:44:54 Quit Diaz_ (Ping timeout: 246 seconds)
18:46:30 Join bertrik [0] (~quassel@rockbox/developer/bertrik)
18:47:44foolshgevaerts: or if you want a really hard problem there is always g#656 I'll even ship you some hardware too if you like ;-)
18:47:47fs-bluebot3Gerrit review #656 at : 3WinCE SDL port using Toolchain CeGCC 0.55 & SDL 1.2 by Benjamin Brown
18:48:01gevaertsfoolsh: is g#954 ready from your p.o.v.?
18:48:03fs-bluebot3Gerrit review #954 at : 3Rockbox Manual - Gigabeat FX and Fuze+ Touchpad Config Appendix by Benjamin Brown
18:48:19foolshgevaerts: yes
18:48:24fs-bluebotBuild Server message: 3New build round started. Revision 12f5c63, 253 builds, 30 clients.
18:49:23 Quit rela__ (Quit: Leaving)
18:49:41 Join rela [0] (~x@pdpc/supporter/active/rela)
18:56:25fs-bluebotBuild Server message: 3Build round completed after 480 seconds.
18:57:04foolshphew!!, that was exciting, right gevaerts?
18:57:29gevaertsNot very :)
19:03:14***Saving seen data "./dancer.seen"
19:10:19*foolsh blows the dust off #656 "I haven't messed with this in a long long while"
19:10:24foolshit was sort of a joke gevaerts ;-)
19:10:46foolshnot the patch the you helping part
19:10:51gevaertsfoolsh: I noticed :)
19:11:04gevaertsNone of the URLs you mention actually work!
19:12:33foolshYeah, there was some financial trouble last winter for me, and I shut down for a long time, and they went stale
19:12:46foolshI'll see what is what
19:14:48 Quit chrisb (Ping timeout: 260 seconds)
19:20:01 Join lebellium [0] (
19:21:16foolshThat patch is long gone now, but if I remember right it was only patching the makefile to prefix=/usr/local/ like the rest of the rockbox toolchain
19:27:13 Join chrisb [0] (
19:28:45foolshgevaerts: g#957
19:28:48fs-bluebot3Gerrit review #957 at : 3Clean up - Removes whitespaces from tools/ by Benjamin Brown
19:44:00 Join ygrek [0] (~user@
19:56:55 Quit fs-bluebot (Quit: connection lost?)
20:01:36 Join fs-bluebot [0] (
20:02:44 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
20:04:42 Quit AlexP (Remote host closed the connection)
20:06:21 Quit chrisb (Ping timeout: 240 seconds)
20:07:38 Join AlexP [0] (~alex@rockbox/staff/AlexP)
20:20:29 Join krabador [0] (~krabador_@unaffiliated/krabador)
20:34:51 Quit ygrek (Ping timeout: 245 seconds)
21:03:18***Saving seen data "./dancer.seen"
21:06:18 Join chrisb [0] (
21:09:22foolshHas anyone witnessed spontaneous e200 or fuze+ player reboots while plugged into a usb charger
21:13:35foolshOk weird the battery was drained, but what was keeping it from charging in rockbox?
21:18:53ikeboy_I have on fuze+
21:19:33ikeboy_Try just getting a few minutes charging to a computer in bootloader, then the rest should work
21:21:40foolshYeah, but I had it plugged into a usb charger and running in rockbox for a day, but it ran the battery out completely, wasn't charging at all, is there a setting I might have enabled that does this?
21:21:53foolshIt charging on the PC usb now though
21:26:03pamaurythat's a known bug
21:26:21pamaurythe charging code is dumb, once it reaches 100%, it stops charging and the battery slowly drains
21:26:42pamauryit's on my todo list, just not very high
21:37:27 Join markun_ [0] (~markun@rockbox/developer/markun)
21:37:41 Join K1773R_ [0] (~K1773R@unaffiliated/k1773r)
21:38:39 Join Cultist_ [0] (
21:38:42 Join ygrek [0] (~user@
21:38:47 Join ikeboy__ [0] (
21:40:13 Quit K1773R (Ping timeout: 252 seconds)
21:40:13 Quit markun (Ping timeout: 252 seconds)
21:40:14 Quit Cultist (Ping timeout: 252 seconds)
21:40:14 Quit ikeboy_ (Ping timeout: 252 seconds)
21:40:14 Quit Cinos (Ping timeout: 252 seconds)
21:40:16 Nick K1773R_ is now known as K1773R (~K1773R@unaffiliated/k1773r)
21:42:17 Join Cinos [0] (
21:44:49 Quit ikeboy__ (Quit: Leaving)
21:44:51 Quit ygrek (Ping timeout: 240 seconds)
21:47:08 Quit Jack87 (Quit: Jack has left the building.)
21:52:25 Join Jack87 [0] (Jack87@nasadmin/admin/jack87)
21:53:44 Join ygrek [0] (~user@
22:04:07 Join xorly [0] (
22:04:24 Quit y4n (Quit: 6,000,000 ways to die — choose one.)
22:14:26 Quit Provel (Ping timeout: 245 seconds)
22:15:08 Join Provel [0] (
22:24:15 Quit ygrek (Ping timeout: 260 seconds)
22:51:37 Quit pamaury (Ping timeout: 246 seconds)
23:03:22***Saving seen data "./dancer.seen"
23:06:39 Join ygrek [0] (~user@
23:14:56 Quit foolsh (Remote host closed the connection)
23:17:26 Join foolsh [0] (
23:29:30 Quit amayer (Quit: Leaving)
23:38:07 Quit ygrek (Ping timeout: 276 seconds)
23:45:48 Quit bertrik (Remote host closed the connection)

Previous day | Next day