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 2016-05-31

00:02:56pamauryit's hard to blame 10 year old code for triggering warnngs of new gccs
00:03:07[Saint]Oh, sure.
00:03:18pamaurybut yeah it's super annoying
00:08:20[Saint]pamaury: I guess one question worth addressing is why the hell we even actually manage the toolchains in this fashion.
00:08:45[Saint]I have some work locally that pretty much scraps all of totally.
00:09:41pamaurybecause some toolchains are broken, and we prefer to have known good versions
00:09:52pamaurymany it's less of a problem now, at least for arm and mips
00:10:07pamaurybut for ypr0 for example we are stuck with it
00:10:11[Saint]It splits out each toolchain into its own script if you do want to generate it yourself, and becomes where it just downloads the preformed toochains for that architecture and dumps them in a user specified location.
00:10:14[Saint]Much nicer, IMO.
00:11:02[Saint]There's no real reason to actually generate the toolchains on each build client or dev machine.
00:11:16pamauryyou mean downloading a prebuilt one ?
00:11:23[Saint]As long as we know they can be built, there's absolutely no problem with shipping prebuilts.
00:11:25[Saint]And, yes.
00:11:46 Quit krabador (Remote host closed the connection)
00:11:57pamaurywell sure, but is it the problem: that we don't even manage to build them ? :)
00:12:35[Saint]Well, that's why I said above "as long as we know they can be built..."
00:13:15[Saint]I think shipping prebuilts might maybe remove some room for error between build clients, also.
00:14:23[Saint]Ideally each build client would refuse to build if the prebuilt toolchain hashes were not matched.
00:14:44[Saint]That way we know each client is on the same page.
00:15:10[Saint]I would also like to discuss how we feel about tight versioning with the build clients.
00:15:41[Saint]I do not think just satisfying a dependency and "as long as it builds" is really enough, and I think it opens up a lot of room for error.
00:16:15[Saint]We need to not only satisfy the build dependency, but get every build client on the roster to agree to what whitelisted versions of the dependencies are allowable.
00:16:59[Saint]But that revolves around a protracted discussion on the mailing list that I really don't have the energy for.
00:17:37[Saint]pamaury, and gevaerts: what, if anything, are your feelings on this?
00:18:04[Saint]I want all our build clients on the same page.
00:18:20pamaurybut what "on the same page" even means for sim and android for example ?
00:18:54[Saint]for Android it is a very clear path. SIM...yeah, hence the discussion I don't really want to have.
00:19:09[Saint]But, I do, because we need to.
00:20:08[Saint]Note: I only think this should be managed for the build clients.
00:20:30[Saint]For Joe User and Joe Developer, if you want to satisfy build dependencies in any way you can - go for it.
00:21:08[Saint]But I think it is desirable to have our build clients agree on versioning for their dependencies.
00:21:28pamaurywell, our build client already enforce specific toolchain versions
00:23:45*[Saint] managed to fuck something up
00:24:05[Saint]I just noticed my latest attempt at building ypr* toolchain is stuck in a loop
00:26:25pamauryhum, ct-ng ignores my make 3.82, I don't know how to force it to use it
00:33:05 Join krabador [0] (~krabador@unaffiliated/krabador)
00:33:24pamaurynot sure if ct-ng ignores it *because* it's 3.82, or because 4.0>=3.82
00:39:06[Saint]woo - ok, mipsel-elf done.
00:40:55[Saint]is anyone interested in committing the updated mipsel binutils and gcc stuffs, and arm gcc -fgnu89-inline stuffs?
00:41:25[Saint]especially the latter as it causes precisely zero fuss for anyone not affected by it.
00:43:45pamaury[Saint]: the problem is the build client still enforce the old mips version
00:44:12__builtingevaerts: should I fix the whitespace issues you referred to in g#1323?
00:44:13fs-bluebot3Gerrit review #1323 at : 3One-Time Password client (HOTP and TOTP) by Franklin Wei
00:44:14pamaurythat means we would need to change, the buildclient script and make sure all build clients are updated at once
00:44:19__builtinsorry, g#1332
00:44:20fs-bluebot3Gerrit review #1332 at : 3Fix broken simulator build with weird sdl-config by Franklin Wei
00:44:49pamauryor that would mean provides a version of mips toolchain that no build client can use
00:45:35[Saint]well, there's no problem with the armeabi change for modern gcc, is there?
00:46:44pamaurythe diff I provided on gist changes CFLAGS for all, including mips and mips is so old it doesn't know about -gnu89-inline
00:46:56pamaurythat's a f*** nightmare
00:47:32n1ckyanyone know what's up with the gerrit github oauth?
00:47:43[Saint]n1cky: hmm?
00:47:47pamauryn1cky: no never tried it, is there a problem ?
00:48:13n1ckyI get a 404 when I try it; might just be my adblocking. google oauth works fine, so whatever
00:49:10[Saint]n1cky: using GoogleOAuth2 'Just Works' for me here.
00:49:13[Saint]Just tried it.
00:49:28[Saint]I can also auth against GitHubOAth2.
00:49:28n1ckyDo you mean github?
00:49:48pamaury[Saint]: I think we need 1) update, 2) introduce a new mips toolchain version in the buildclient list, 3) make sure all existings build client have both mips toolchains available (or none) 4) convert all mips builds to the new one
00:49:54[Saint]And, no, I didn't. I meant Google.
00:49:59 Join JdGordon_ [0] (~jonno@rockbox/developer/JdGordon)
00:50:04[Saint]But I can /also/ auth against GitHub.
00:50:19pamauryn1cky: I confirm github oauth leads to 404
00:50:36pamauryno idea why though
00:50:50[Saint]Perhaps I have some locale caching shenanigans going on.
00:50:53n1cky[Saint]: it could be that gerrit recognizes a commonality between your github and google identity (maybe email) and prevents you from seeng the bug
00:51:07n1ckyor that^
00:51:07[Saint]Perhaps, yes.
00:51:24[Saint]I would like to think that local caching wouldn;t affect OAuth login though.
00:51:44__builtindoes 404 for you?
00:52:01 Quit alexweissman (Remote host closed the connection)
00:52:06n1cky__builtin: it does here.
00:52:13[Saint]Ah, Aha. There we go.
00:52:42__builtin gives that URL
00:52:51 Join The_Prospector|2 [0] (
00:52:59n1ckyanybody want a changeset reviewed? I've got an hour
00:53:06 Quit JdGordon (Ping timeout: 272 seconds)
00:53:24[Saint]One assumes all the pending changes on Gerrit might.
00:53:28__builtinn1cky: do you have hardware to test on?
00:53:58[Saint]I mean, I don't mean to sound like a dick. That sounded shitty.
00:54:11[Saint]But, that's one of the reasons Gerrit exists.
00:55:03n1cky__builtin: I've got a clip zip
00:55:23 Quit The_Prospector (Ping timeout: 260 seconds)
00:55:24__builtincan you build rockbox?
00:55:30__builtin(if not I can supply a build)
00:55:51n1cky[Saint]: I got you. Fairly easy to be able to see that something like 1334 should be +1'd though. :)
00:56:07n1cky__builtin: I run openbsd. I've built successfully before in a vm, but I'd have to get that setup
00:56:19__builtinok, I'll build it for you then
00:56:29 Quit ender` (Quit: Inside every cynical person, there is a disappointed idealist.-- George Carlin)
00:56:43n1cky__builtin: if you send a build I can review. Can I pm you my email?
00:56:49*__builtin can host it
00:57:04__builtinit's a one-time password client I want to make sure works on hardware
00:57:15fs-bluebot3Gerrit review #1323 at : 3One-Time Password client (HOTP and TOTP) by Franklin Wei
00:58:05[Saint]To be perfectly honest, I would like to remove the debug screen entirely.
00:58:21[Saint]We have debug build targets, enable it there.
00:58:40__builtinnothing wrong with having it there, right?
00:58:42[Saint]There's no reason for it to exist in the user builds.
00:59:00__builtinif something breaks they can help debug
00:59:03n1ckyI find myself going in there to see info sometimes that isn't available / clear / detailed elsewhere.
00:59:09pamaurywell it's useful to have sometimes to just tell users "have a look here and tell me what you see"
00:59:10[Saint]__builtin: stupid questions, really. people poking at boost. etc.
00:59:26pamauryobviously you can't blame a bug if you touch it
00:59:40__builtinit says "KEEP OUT" anyhow
00:59:57[Saint]Which may as well translate to "FUN STUFF IN HERE"
01:00:33n1ckyI dunno, the subset of people who are willing to load alternative firmware onto their mp3 player are not the subset of people that you need to protect from themselves imho
01:00:58[Saint]Perhaps we could force enable it with a config option.
01:01:06[Saint]liker 'userdebug=true' or so.
01:01:20__builtinmaybe it could be hidden away like on android
01:01:35[Saint]I just don't think there's any reason at all it should be displayed to the user by default.
01:01:47__builtinclick something 5 times and it shows up
01:01:48[Saint]you're right it can be useful, though.
01:01:59 Join JdGordon [0] (
01:01:59 Quit JdGordon (Changing host)
01:01:59 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
01:02:10n1cky__builtin: I like that idea, some kind of trigger
01:02:37__builtininstall it and run the "otp" plugin under applications
01:02:56[Saint]A button trigger would be needlessly difficult to implement.
01:03:07[Saint]But we could definitely squirrel it away with a config option.
01:03:41pamauryI am not sure I see the point, is it that much of a pain that we need to hide it ?
01:03:55[Saint]Change it to a toplevel menu item, and then hide it by default using the mechanism we already have deployed for editing the main menu structure.
01:04:20[Saint]pamaury: it just avoids stupid questions and Joe User poking at things they have no need to poke at.
01:04:24*__builtin thinks that hiding something's complexity just makes it needlessly complex
01:04:48[Saint]__builtin: ...yet you just suggested a convoluted method of doing so.
01:04:52[Saint]Go figure, huh.
01:05:05__builtinremoving it completely is worse
01:05:14 Quit JdGordon_ (Ping timeout: 264 seconds)
01:05:27[Saint]I never said anything of the sort.
01:05:44__builtinalright, from the default builds then
01:06:01[Saint]My secondary suggestion even makes it arguable more accessible.
01:06:04pamaurymaybe put it in the context menu of the system meny ?
01:06:24*pamaury wants it to be easily accessible
01:06:35[Saint]I was thinking of putting it in the toplevel menu structure, as if the problem is buttons the user may not even be able to navigate to it.
01:06:44[Saint]then we could just hide it by default.
01:07:12__builtincould it be hidden behind a long-click on some item?
01:08:13[Saint]that cycles around to 'if the problem was buttons' I guess.
01:08:31pamauryif you don't have access to a button, you are dead anyway, I don't see the point
01:08:49__builtinmake it to get to if you're looking for it
01:09:12n1cky__builtin: copying over
01:09:19[Saint]if it was in the top level all you would need to function is the right nav button, and either up or down.
01:09:57n1ckyOkay but if the problem was that you couldn't press buttons, you're SOL in the current setup anyway, right?
01:10:26[Saint]if you had no buttons _at all_, yes.
01:10:27__builtinI think [Saint]'s saying that if /some/ buttons are dead, you can still get to it
01:10:35[Saint]^ right
01:11:24[Saint]ideally it would be a plugin and then you could launch it from boot with no buttons at all with a config menu entry.
01:11:35[Saint]we already have a provision for launching a plugin immediately at boot.
01:11:55pamaurybut you need buttons to navigate it
01:12:17[Saint]read what I wrote.
01:12:21__builtininside the menu
01:12:34__builtin(debug menu)
01:12:39pamaurythe debug menu is useless unless you can navigate to some screens in it
01:12:43[Saint]we can launch a plugin at boot, we've been able to do this for an eternity.
01:12:57***Saving seen data "./dancer.seen"
01:13:01pamauryread what I said...
01:13:06[Saint]no user intervention is required for that after it is setup in the config.
01:13:25__builtin[Saint]: pamaury's saying that using the debug menu still needs buttons
01:13:50[Saint]assumedly you would structure it with the most important assets first.
01:13:58[Saint]but small screen targets are fucked, yes.
01:14:11__builtinn1cky: does the plugin start properly?
01:14:13[Saint]but they would be just as fucked if you couldn;t navigate in any instance.
01:14:13pamauryI have no idea what you mean, there is no such thing as the mot important asset
01:14:32[Saint]Sure there is.
01:15:57 Quit nchambers (K-Lined)
01:16:26pamauryAnd I am not even mentionning that making the debug menu a plugin is going to be a massive task
01:16:42n1cky__builtin: sorry, got distracted. Yes, it starts
01:16:58__builtingood, that means the self-check passes
01:17:31n1cky__builtin: anything I should do to test it beyond that?
01:17:47__builtinlet me see
01:18:00__builtindo you have a github account you'd be willing to test with?
01:18:32n1ckyDoes it change something about how I can access my account in the future?
01:18:45__builtinyes, but you can easily revert it
01:18:49n1ckyOkay, yeah, let's try it.
01:19:01__builtinok, enable 2-factor authentication in your account settings
01:19:06__builtinunder "Security"
01:19:09pamauryif anything, I would vote for "in the context menu of the main screen, have a setting for showing the debug menu (in the main screen)", and on debug builds it would always be there, and it would be a proper option so it is remembered over boots
01:20:11__builtinthen choose "use an app"
01:20:36n1cky__builtin: How do I use this keyboard?
01:20:55__builtinthe rockbox keyboard?
01:21:11n1ckyyeah, I've never used it
01:21:23__builtinchoose a character with the D-pad, and select
01:21:33n1ckyhow do I hit the equivilent of <enter>?
01:21:33__builtinto backspace select the line and hit select
01:21:48__builtinlong-press the center button
01:21:54__builtin(I think)
01:22:15 Quit ZincAlloy (Quit: Leaving.)
01:22:32n1ckyOkay, well I go to manual entry, am prompted for the account name, enter "gh" for github, and then I hold the center
01:22:39pamaurykeyboard map is device depending, on some devices it is crazy (especially the ones with a small numbers of keys)
01:22:56n1ckyIt then says, "Enter base32 encoded secret" but immediately exits with invalid
01:23:13__builtinhmm, it should go to a keyboard again
01:23:27n1ckynevermind, i got it to take. Just gotta be careful to not hold it too long
01:23:51__builtinnow type the string that github gives you (case doesn't matter, just don't type it wrong)
01:24:17__builtinand when it asks, it /is/ a TOTP account
01:25:41__builtinbeyond that, go with the defaults
01:25:49n1ckyhardware fault, divided by zero
01:25:56__builtinafter doing what?
01:26:20n1ckySaid /yes/ to TOTP, timeperiod I accidentally entered a null string
01:26:35__builtinah, that makes sense :P
01:26:40*__builtin needs to put in some error-checking
01:27:05n1cky__builtin: so what should my hotp counter be?
01:27:12n1ckyand totp period
01:27:28__builtinperiod should be 30 seconds
01:27:32 Quit pamaury (Ping timeout: 240 seconds)
01:28:33__builtinwhy do you need the HOTP counter?
01:28:37n1ckyI don't neverminde
01:28:56n1ckyWhen I go to generate the code, I get '*d (6 second(s) left)'
01:29:46__builtinlooks like I've got a format string bug
01:31:11*[Saint] really wishes that GitHub (and a truckload of other services) would add a system for preferential 2FA access
01:31:24[Saint]Like, many offer a fallback email, or fallback SMS.
01:31:34__builtingive me a sec
01:32:46n1cky[Saint]: I wish that I could use ssh-agent for http auth
01:32:55[Saint]I would like to use application based Google 2FA for everything, but also allow the use of SMS based auth if for whatever reason I don't have immediate access to my application to manage this
01:32:58n1ckyerr, gpg-agent
01:33:11[Saint]But for pretty much everything it is a hard app OR SMS.
01:33:17[Saint]No app *and* SMS.
01:33:58n1cky[Saint]: The dream is to be able to have your hard drive encrypted with a key on a card, and go completely passwordless
01:34:42*[Saint] very recently started using GPG signing with github but assumes that a wide majority of people would find it impossible to give less fucks about whether or not the signature they're never going to confirm is valid or not.
01:35:22n1ckyYeah, gpg is just too far out of the normal
01:36:27*[Saint] wonders what the methodology of disallowing application based OAth and SMS based OAuth to be active at the same time on any given account.
01:36:41[Saint]Very few providers allow this, for no obvious reason at all.
01:37:31n1ckyYou mean either would work?
01:37:58[Saint]yes, like asking at the time which method you would like to auth with.
01:39:01[Saint]pretty much everything only allows either application based or SMS based auth to be active at any given time.
01:39:27n1ckyI agree.
01:39:32[Saint]falling back to recovery tokens if it is inaccessible, or _maybe_ sometimes offering an SMS recovery method.
01:39:51[Saint]But I see no good reason why both couldn't be active at any given time.
01:39:58n1ckyI think the argument against it is probably something like, 2fa gaurantees that the person logging in has every secret associated with that account
01:40:10[Saint]Perhaps there is one, but I am not aware of it. I assume there is justification for it.
01:40:29n1ckybut if you introduce a concept of, "this person needs two of three secrets", well now you don't get those guarantees and might as well be using a password
01:40:45n1ckyof course, that's faulty logic, because it's still harder to have someones cell phone and their password than it is to just have thier password.
01:40:52n1ckyBut I imagine the line of reasoning comes from there.
01:41:03[Saint]You're probably right.
01:42:03__builtindoes rockbox's printf implementation support something like %*d?
01:42:23[Saint]grep would probably tell you.
01:42:51[Saint]though looking for all printf calls might be hilarious in the entire codebase.
01:43:08[Saint]I imagine there is a non-trivial amount.
01:44:16__builtinactually it seems I can't find any
01:44:31__builtingrep -R printf | grep "%\*d"
01:45:52n1cky[Saint]: I've prodded you about this before, has anyone's opinion changed on the outlook of supporting ext2 for external media?
01:47:53[Saint]"there's no point", really.
01:48:09[Saint]and permission based filesystems for removable media is batshit.
01:48:21__builtinlarge files?
01:49:03 Join nchambers^ [0] (
01:49:11[Saint]If we were going to support any alternative filesystem, it would be exFAT.
01:49:21[Saint]But even then there's no real reasoning for this.
01:49:29[Saint]Who uses 4GB+ files on a DAP?
01:49:34__builtinmovies :P
01:49:59[Saint]__builtin: on our largest screen DAPs that would be a ~6 hour long movie, lol.
01:50:12[Saint]and wikipedia has no single files that large.
01:50:33__builtinwell, it has to work around that barrier
01:50:50__builtinn1cky: untested, but herE:
01:51:19__builtinsynchronize the clock to within 30s of real time before using the plugin
01:51:31__builtinthen start it and generate a code for github
01:51:35*[Saint] restates "permission based filesystems on removable media is batshit"
01:51:52[Saint]I very firmly believe that. There's no good reasoning for it and plenty against it.
01:55:00n1ckyI'd agree if the rockbox project was engineering their own filesystem... But that's not the case
01:55:20 Nick nchambers^ is now known as nchambers (
01:55:58n1ckyAs in, if you were starting rockbox from scratch and decided to make rockboxfs, adding a bunch of ext2 features would be ridiculous
01:56:04 Nick nchambers is now known as nchambers_ (
01:56:05 Nick nchambers_ is now known as nchambers (
01:56:20__builtin on a side note, I found a really cool game that might be cool to have on rockbox
01:56:35[Saint]I'm not sure you're getting the pint I'm making, n1cky.
01:56:54[Saint]I'm saying that permission based filesystems in any capacity on removable media is batshit crazy.
01:59:23n1ckyI disagree−− not because permissions are ever going to be used by the removable media, but because the origin of the media might be a computer that has a permission-based filesystem and it's simpler to be able to just copy the permissions and not use them than it is to deny the ability to store that information
02:00:33[Saint]but then what happens when you attempt to touch it from any other system that doesn't happen to be set up identically to the initial host of said content?
02:01:01n1cky__builtin: what do I set my time to? UTC?
02:01:07[Saint]you get it (rightfully) refusing to do so. entirely breaking the concept of removable media.
02:01:55[Saint]the idea of removable media in this context is that it 'just works' and is accessible on anything it touches capable of reading it.
02:02:04__builtinuhh, try UTC
02:02:08[Saint]permission based filesystems screw that up spectacularly.
02:02:25 Join alexweissman [0] (~alexweiss@
02:02:41[Saint]you want to be able to pop out that media and have the entire content accessible to anything else that can touch it.
02:03:10[Saint]you don't want it bouncing off user permissions.
02:03:46__builtinn1cky: it needs to be within 30 seconds of "real time"
02:04:07n1cky__builtin: what on earth does that mean?
02:04:13__builtinUTC, I guess
02:04:17*__builtin actually isn't so sure
02:04:39n1cky[Saint]: if the user wants to use ext2 and wants their media to be accessible by anything else that can touch it, he can figure out how to fix the permissions
02:04:51n1cky__builtin: I'll try UTC
02:05:00__builtinonce the time is set, generate a code and try to use it for github
02:05:05[Saint]n1cky: or you could just not support a really silly model in the first place. :)
02:05:25[Saint]IFF Rockbox was going to add another filesystem, it would be exFAT.
02:05:30__builtinand if it doesn't work, try whatever your local time is
02:06:12__builtinhmm, the RFC specifies UTC
02:06:54 Quit alexweissman (Ping timeout: 258 seconds)
02:07:05__builtinbut the weird thing is, I haven't used UTC and it still works
02:07:55n1cky__builtin: the code that is generated seems to be wrong... :\
02:08:05__builtintry local time, I guess
02:08:30n1ckydid you push this changeset yet?
02:08:44fs-bluebot3Gerrit review #1323 at : 3One-Time Password client (HOTP and TOTP) by Franklin Wei
02:08:52__builtinbut that doesn't have some new changes yet
02:09:35 Quit [7] (Ping timeout: 258 seconds)
02:09:56 Join TheSeven [0] (~quassel@rockbox/developer/TheSeven)
02:10:28n1ckySeems to me like it's something with the format string, still.
02:10:45__builtinin gerrit?
02:10:48*[Saint] now has a go at building the git head mipsel toolchain
02:11:02__builtinor the build I gave you?
02:11:04n1cky__builtin: no, in the most recent build you sent.
02:11:29n1ckyAhahaha, nevermind
02:11:39n1ckyYou switched the time remaining and the code, i tihnk.
02:12:03n1ckyI kept getting, "000006 (341234 seconds remaining)"
02:12:15*__builtin changed the format string, but didn't change the arguments
02:12:21__builtindoes GH take it?
02:14:37n1ckyYes. Works! :)
02:14:55__builtinUTC time, right?
02:15:18n1ckyYup, UTC time.
02:15:24__builtinok, I'll put that in the docs
02:15:54__builtinthough... it won't be very practical to ask the user to change the time each time the plugin runs
02:16:27n1ckyYeah, probably specifying an hourly offset is needed
02:16:52*__builtin needs to look into how rockbox deals with time
02:17:08__builtinwhere does the simulator get its time from?
02:18:09n1ckybtw, when I kept getting mixed up arguments, "000006" was static. So even if the arguments are mixed up, the time remaining seemed to be not accurate, but that could just be the formatter chomping
02:18:21__builtinyeah, I think I have it figured out
02:18:56__builtinbut you can call yourself the first person to ever use an MP3 player as an OTP token now ;)
02:20:09n1ckyhaha! Yeah this is a really cool plugin, great idea.
02:21:38n1ckyUpdate the changeset so I can +1 and we can get this merged hopefully
02:21:58*__builtin is going to test things again
02:22:10__builtinand account for timezones
02:22:41n1ckyfeel free to send another build, i'm just going to be watching the west finals and sitting on my laptop
02:24:50 Join smoke_fumus [0] (~smoke_fum@
02:26:43__builtinif I do, it'll probably break your previous save data
02:27:40__builtinso beware
02:30:19__builtinit seems that rockbox doesn't know about time zones at all
02:32:44__builtinbut the simulator uses time() which is in seconds since the epoch, which is in UTC
02:38:25__builtinDST is going to be a pain
02:44:58n1ckyYeah I disabled it in github right after; I want to wait until this is a bit more stable
02:58:30n1cky[Saint]: why are you so afraid of the user hurting themselves?
02:59:12n1ckyIf s/he fills his SD card with wrong permissions, he's dug his own grave. It's not like I'm suggesting rockbox should replace FAT with ext2.
02:59:25n1ckyJust the ability to read it, for people who want it.
03:05:37__builtinn1cky: I think it should account for time zones properly now
03:07:19__builtinpushed to gerrit
03:08:39n1ckycan you provide me with a build again?
03:13:01***Saving seen data "./dancer.seen"
03:13:40__builtinyou'll have to delete the old save data for it to work
03:14:00__builtinso remove .rockbox/rocks/apps/otp.dat before running it
03:28:25*[Saint] has mipsel gcc 4.1.2 and gcc 4.6.3 now
03:28:49n1ckyGave it a shot−− worked great. Changed my time and everything worked with the time offset.
03:28:56n1ckyFormatting is all fixed up.
03:29:20n1cky+1'd on gerrit. Cheers
03:29:24__builtinlet me add some time zone labels to make it pretty
03:29:30[Saint]but I had to resort to piecing it together from what the build system is capable of compiling currently with a modern host and the leftovers I compiled about a year ago.
03:32:25__builtinn1cky: can you re-do the +1 on the new patch set?
03:34:26n1cky__builtin: done
03:36:23__builtinthanks :)
03:37:07n1cky:) nice work on that plugin
03:45:57 Join JdGordon_ [0] (~jonno@rockbox/developer/JdGordon)
03:47:33__builting'nite o/
03:48:32 Quit JdGordon (Ping timeout: 240 seconds)
05:13:05***Saving seen data "./dancer.seen"
05:30:40 Join Stroboko1p [0] (
05:34:04 Quit Strobokopp (Ping timeout: 244 seconds)
05:37:03 Quit krabador (Quit: Take The Time)
06:27:31 Quit mc2739 (Ping timeout: 260 seconds)
06:30:44 Join alexweissman [0] (~alexweiss@
06:56:33 Quit alexweissman (Remote host closed the connection)
06:58:12 Quit [Saint] (Read error: Connection reset by peer)
06:59:05 Join [Saint] [0] (~hayden@rockbox/staff/saint)
07:04:11 Quit [Saint] (Remote host closed the connection)
07:05:02 Join [Saint] [0] (~hayden@rockbox/staff/saint)
07:05:03 Part edhelas ("Disconnected: closed")
07:07:55 Quit foolsh (Ping timeout: 244 seconds)
07:13:09***Saving seen data "./dancer.seen"
07:14:48duo8does rockbox have a log of sorts?
07:25:35 Join alexweissman [0] (~alexweiss@
07:27:08 Quit dfkt (Ping timeout: 246 seconds)
07:27:26 Join foolsh [0] (~starchase@2601:241:c200:4bc5:a408:f457:56ab:b7d7)
07:43:06 Quit amiconn (Remote host closed the connection)
07:43:06 Quit pixelma (Remote host closed the connection)
07:45:19 Quit Moarc (Quit: i znowu NADMUCHAŁ BALONA)
07:45:21 Join pixelma [0] (~pixelma@rockbox/staff/pixelma)
07:45:22 Join amiconn [0] (~amiconn@rockbox/developer/amiconn)
07:45:29 Join Moarc [0] (~chujko@2002:551c:8069:1:cc73:aeb8:2b17:990b)
07:49:17[Saint]duo8: Not by default no, but, you can build a logf build and sprinkle logging around as you desire.
07:58:56duo8btw is there a guide on the debug functions?
08:00:16[Saint]Which debug functions are you talking about specifically?
08:02:26duo8the buffer thread viewer
08:04:02[Saint]Not outside of the source itself, no.
08:11:19 Quit [Saint] (Remote host closed the connection)
08:13:17 Join einhirn [0] (
08:14:32 Quit nlogex (Ping timeout: 258 seconds)
08:18:58 Join ender` [0] (
08:24:31 Join girafe [0] (
08:37:08 Quit alexweissman (Remote host closed the connection)
08:55:28 Join petur [0] (~petur@rockbox/developer/petur)
09:13:12***Saving seen data "./dancer.seen"
09:21:55 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
09:23:06 Quit girafe (Quit: Leaving)
09:24:33 Quit JdGordon_ (Ping timeout: 250 seconds)
09:36:31gevaerts__builtin: why not fix them?
09:40:52 Join edhelas [0] (~edhelas@2001:1c02:1903:d800:78dc:423:6858:2d5f)
09:40:52 Quit edhelas (Client Quit)
09:41:17 Join elensil [0] (~edhelas@2001:1c02:1903:d800:78dc:423:6858:2d5f)
09:41:24 Quit elensil (Client Quit)
09:41:49 Join paulk-collins_ [0] (
10:04:34 Join elensil [0] (~edhelas@2001:1c02:1903:d800:78dc:423:6858:2d5f)
11:13:14***Saving seen data "./dancer.seen"
11:38:20 Join fs-bluebot_ [0] (
11:38:20 Quit bluebrother (Disconnected by services)
11:38:26 Join bluebrother [0] (~dom@rockbox/developer/bluebrother)
11:40:38 Quit fs-bluebot (Ping timeout: 260 seconds)
12:11:22 Join pamaury [0] (~pamaury@rockbox/developer/pamaury)
12:29:31 Quit pamaury (Ping timeout: 246 seconds)
12:50:51 Join pamaury [0] (~pamaury@rockbox/developer/pamaury)
13:12:43 Quit Strife89 (Ping timeout: 260 seconds)
13:12:54 Join Strife89 [0] (
13:13:16***Saving seen data "./dancer.seen"
13:51:10 Join pamaury_ [0] (~pamaury@rockbox/developer/pamaury)
13:51:15 Quit pamaury_ (Remote host closed the connection)
14:16:47 Join JanC_ [0] (~janc@lugwv/member/JanC)
14:17:49 Nick JanC is now known as Guest93383 (~janc@lugwv/member/JanC)
14:17:49 Nick JanC_ is now known as JanC (~janc@lugwv/member/JanC)
14:19:20 Quit Guest93383 (Ping timeout: 272 seconds)
14:23:02elensilpamaury: tu es français :D
14:24:02 Quit PurlingNayuki (Quit: PurlingNayuki)
14:24:34elensilI didn't knew that :)
14:40:30 Quit Guest85177 (Ping timeout: 244 seconds)
14:42:48 Join Guest85177 [0] (
15:00:08 Quit Guest85177 (Ping timeout: 244 seconds)
15:02:10 Join Guest85177 [0] (
15:03:53 Join amayer [0] (
15:12:01 Quit Guest85177 (Ping timeout: 244 seconds)
15:13:17***Saving seen data "./dancer.seen"
15:13:58 Join Guest85177 [0] (
15:18:36 Quit amayer (Quit: Leaving)
15:30:32 Join amayer [0] (
15:37:03 Join Boltermor [0] (
15:57:32 Quit idonob (Ping timeout: 240 seconds)
16:00:00 Join idonob [0] (
16:04:29 Join krabador [0] (~krabador@unaffiliated/krabador)
16:07:03 Join ZincAlloy [0] (~Adium@2a02:8108:8bc0:1664:6cca:5899:7d37:cbe4)
16:21:19 Join alexweissman [0] (~alexweiss@
16:22:01 Join nlogex [0] (
16:22:50 Quit Boltermor (Ping timeout: 272 seconds)
16:40:47 Join girafe [0] (
16:42:59 Join dfkt [0] (~dfkt@unaffiliated/dfkt)
16:51:17 Join Boltermor [0] (
17:13:21***Saving seen data "./dancer.seen"
17:26:20 Quit petur (Quit: Connection reset by beer)
17:36:28 Quit krnlyng (Ping timeout: 246 seconds)
17:39:25 Quit Boltermor (Read error: Connection reset by peer)
17:39:52 Join Boltermor [0] (
17:49:23 Join krnlyng [0] (
17:50:01 Quit Boltermor (Read error: Connection reset by peer)
17:50:27 Join Boltermor [0] (
17:58:36 Join alexweis_ [0] (~alexweiss@
17:58:36 Quit alexweissman (Read error: Connection reset by peer)
18:01:51 Quit elensil (Quit: Leaving.)
18:05:13 Quit einhirn (Quit: Miranda IM! Smaller, Faster, Easier.
18:18:10 Quit Boltermor (Read error: Connection reset by peer)
18:18:35 Join Boltermor [0] (
18:23:08 Quit Boltermor (Read error: Connection reset by peer)
18:23:35 Join Boltermor [0] (
18:24:58 Join dudemann [0] (4adf986a@gateway/web/freenode/ip.
18:25:43 Quit Boltermor (Read error: Connection reset by peer)
18:26:11 Join Boltermor [0] (
18:26:58dudemannhas there been an MP3 player released, since the Archos Studio, that holds 100+gb of data (and is capable of running rockbox)? I almost got a FiiO X3 but see it cant run rockbox.
18:36:11 Quit Boltermor (Ping timeout: 252 seconds)
18:39:45pamaurydudemann: any player with microsd slot
18:40:03amayeriPod Classic
18:40:24Horrorcatpamaury: is that true? don’t you need SDXC for > 32 GB?
18:41:13dudemannid like to have an entire library of 100+gb to randomly play from and make playlists on, rockbox allowed me to make 6000 file playlists
18:42:43pamauryas far as know, you can take an sdxc card, reformat it as fat32 and you are rolling, since sdxc is electrically compatible with sdhc (it simply provides more bandwith if the host has support for it)
18:43:02Horrorcatand all targets support sdhc?
18:44:07pamauryas long as they have a microsd slots (except maybe the oldest players when high speed cards did not even existed)
18:44:25Horrorcatshould have known all that before I got only 32 GB for my zenxfi3 :-)
18:44:29Horrorcatguess I should’ve asked.
18:44:53pamaurybasically there is no difference between sd and sdhc at the electrical level, simply sdhc mandates support for high speeds mode
18:45:44pamaurydon't expect to read/write your sdxc card at 100MB/s with rockbox though, it simply doesn't have the hardware support for it
18:46:18Horrorcatwell, that speed isn’t important anyways.
18:53:02dudemannso with a sansa fuze, you can use a 128gb card as long as its fat32 ?
18:55:14 Quit alexweis_ (Remote host closed the connection)
18:55:41 Join alexweissman [0] (~alexweiss@
19:00:03 Quit alexweissman (Ping timeout: 244 seconds)
19:09:49duo8you can use the upcoming 256GB if you want
19:10:07duo8or a 512GB with a SD -> mSD adapter
19:12:21 Join lebellium [0] (
19:12:28 Quit amayer (Ping timeout: 276 seconds)
19:13:22***Saving seen data "./dancer.seen"
19:14:19 Join JdGordon_ [0] (~jonno@rockbox/developer/JdGordon)
19:15:07 Join amayer [0] (
19:17:50 Quit JdGordon (Ping timeout: 264 seconds)
19:19:09dudemannthank you
19:24:38 Join mc2739 [0] (~mc2739@rockbox/developer/mc2739)
19:28:43duo8will probably be slow though, FAT isn't efficient afaik
19:28:58duo8f2fs support when
19:29:47Horrorcatfragmentation aside, how would the filesystem influence the read speed on a non-rotating medium?
19:30:26 Quit mc2739 (Ping timeout: 264 seconds)
19:50:58 Quit smoke_fumus (Read error: Connection reset by peer)
19:55:52 Join JdGordon [0] (
19:55:52 Quit JdGordon (Changing host)
19:55:52 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
19:59:16 Quit JdGordon_ (Ping timeout: 276 seconds)
20:00:41 Quit krabador (Quit: Take The Time)
20:03:52 Join krabador [0] (~krabador@unaffiliated/krabador)
20:04:07 Join alexweissman [0] (
20:07:46 Join smoke_fumus [0] (~smoke_fum@
20:39:50 Quit Smx (Excess Flood)
20:42:47 Join Smx [0] (Elite8556@gateway/shell/elitebnc/x-kaxclrrlxmrzdqrw)
20:47:53 Quit alexweissman (Remote host closed the connection)
20:52:32 Join JdGordon_ [0] (~jonno@rockbox/developer/JdGordon)
20:55:37 Quit JdGordon (Ping timeout: 246 seconds)
21:03:05 Quit paulk-collins_ (Ping timeout: 250 seconds)
21:13:24***Saving seen data "./dancer.seen"
21:15:06 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
21:18:15 Quit JdGordon_ (Ping timeout: 260 seconds)
21:30:13 Join paulk-collins_ [0] (
21:37:07 Quit Smx (Excess Flood)
21:37:47 Join Smx [0] (Elite8556@gateway/shell/elitebnc/x-obyrdmryxcpewzgb)
21:41:07 Quit smoke_fumus (Quit: KVIrc 4.2.0 Equilibrium
21:46:41 Quit paulk-collins_ (Quit: Leaving)
22:01:56 Join edhelas [0] (
22:03:31 Join Guest65764 [0] (~ac@2001:910:113f:1:6a05:caff:fe1c:1627)
22:05:30Guest65764I finally had an opportunity to try opus files again
22:05:47Guest65764has anyone else experienced issues with the autoseek feature on opus files ?
22:06:25Guest65764by autoseek I mean "resume play" or whatever it's called in english
22:25:28 Part edhelas
22:36:56 Quit TheSeven (Disconnected by services)
22:37:07 Join [7] [0] (~quassel@rockbox/developer/TheSeven)
23:06:05 Join JdGordon_ [0] (~jonno@rockbox/developer/JdGordon)
23:08:37 Quit JdGordon (Ping timeout: 246 seconds)
23:13:27***Saving seen data "./dancer.seen"
23:20:22 Quit lebellium (Quit: ChatZilla 0.9.92 [Firefox 46.0.1/20160502172042])
23:35:32 Join edhelas [0] (
23:46:01pamauryHorrorcat: FAT is a bad file system for that, finding where is Nth block of a file takes a take linear in N, so on huge file the file system can incur a big penalty even for read. But except for very big files, what matters most is fragmentation of course
23:46:07 Quit pamaury (Remote host closed the connection)
23:58:38 Quit ender` (Quit: /* Return code=1: generic error condition, Return code=2: all other error conditions */)

Previous day | Next day