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 2005-09-22

00:03:02unL33T2firmware upgrading...
00:03:26rasherif that works, my bootloader.bin must be correct
00:03:40preglowwhere'd you get it?
00:03:57preglowi hope you're not using a cvs bootloader? :>
00:04:09rasherpreglow: I have all bootloader versions Linus put up lying around
00:04:24rasherWell, the MD5 summing would catch if I were anyway
00:06:12rasherMaybe not all, but v1 through v5
00:07:08 Quit matsl (Read error: 104 (Connection reset by peer))
00:07:39rasherOr not... I have SOME bootloaders.
00:07:44unL33T2didnt fix it :(
00:08:01preglowunL33T2: that's very improbable
00:08:10unL33T2also the iriver firmware wont allow me to do a DBscan (in the General menu)
00:08:27preglowa lot of people have done exactly what you've done, and upgrading to a newer bootloader did wonders
00:08:29unL33T2sigh stupid thing
00:08:36preglowas it should
00:08:47preglowrasher: you're sure you're using a new fwpatcher?
00:09:00unL33T2well now it says I have v5 now and still have the problem
00:09:03rasherI cvsupped a minute ago
00:09:16rasherAnd, he has v5
00:09:21rasherthat wouldn't happen unless I were.
00:09:21preglowthat doesn't matter
00:09:25preglowit's the mkboot routine that was wrong
00:09:28preglownot the bootloader itself
00:09:47rasherOh, the fwpatcher failed?
00:09:57preglowthere was a bug in mkboot.c
00:10:00rasher<unL33T2> well now it says I have v5
00:10:03preglowand fwpatcher uses that verbatim
00:10:17rasherBut but but... it md5sums?
00:10:21preglowlike i say
00:10:32preglowthat should have been picked up, yes
00:10:40preglowthat it should
00:10:54preglowwell, i'm stumped
00:10:56rasherWell, the md5sums were calculated by running mkboot
00:11:04rasherBut.. what's the bug, and has it been fixed?
00:11:07preglowunL33T2: tried just flashing vanilla iriver?
00:11:13preglowrasher: gimme a sec
00:11:30rasherI don't see any commits
00:11:52preglowrev 1.2 mkboot.c
00:11:58preglowFixed the dreaded 1.65 bootloader incompatibility
00:12:11unL33T2vanilla iriver?
00:12:22preglowrasher: and fwpatcher/main.c rev 1.3: Inserted Linus' mkboot.c fix.
00:12:35rasherSure, I have that
00:12:41preglowthat's that
00:12:51rasherNow what?
00:12:58preglowthat's what fixed the 1.65 playback bug
00:13:05preglowif all that's in place, i have no idea what's wrong
00:13:06***Saving seen data "./dancer.seen"
00:13:09unL33T2now do I throw it on the ground and stomp on it :P
00:13:12preglowunL33T2: well, yeah, there's no other options
00:13:27rasherWhat's the problem? I came in in the middle of this..
00:13:41rasherAnd there's a lot to read :-\ Quick summary?
00:14:23preglowhe's got the 1.65 playback problem
00:14:29rasherWith oggs?
00:14:31unL33T2when I select a song in my iRiver firmware in one of the sorted menus (Artitst, Album, Title, Genre, etc) then it will not play but if I use the browse by folder option then it plays fine. RockBox works fine.
00:14:35unL33T2no with everything
00:14:48rasherHave you triedrebuilding the db?
00:14:53unL33T2I have
00:14:54preglowthe 1.65 bug is the closest thing i can think of, if it's not that, i have no idea what it is
00:15:02rasherSounds like a broken db
00:15:06unL33T2but the firmware will not allow me to do a DBscan
00:15:21preglowi thought nothing played in any mode
00:15:23preglowoh well then
00:15:33preglowthe firmware doesn't dbscan
00:15:36rasher"not allow"?
00:15:39preglowyou'll have to do that from a pc, no?
00:15:41unL33T2i go General > DBscan > Yes and it just quits that and goes back to whatever screen was on it
00:15:47preglowunL33T2: that's what it does
00:15:59preglowunL33T2: the scanning itself has to be done from a computer
00:16:04unL33T2well I can do a db rebuild with moodlogic
00:16:08unL33T2but it doesnt fix it
00:16:15rasherWhat the hell is Moodlogic?
00:16:16preglowoh well
00:16:22unL33T2i dunno what came with the iriver ...
00:16:24preglowthis isn't a rockbox problem, that's for sure
00:16:45unL33T2i know i wasnt sure and was hoping someone here would have the answer
00:16:55rasherAre you not usng the DB-create program that came with your player?
00:17:02unL33T2moodlogic is what came with the iriver for creating the db
00:17:14preglowmine did not
00:17:18preglowi got the iriver program
00:17:25unL33T2mine did ... and it worked before ...
00:17:35unL33T2I think it's what iriver has listed on their site now too
00:17:42rasherNever heard of it
00:17:45unL33T2could you send me the iriver one? maybe it will fix it
00:17:46preglowhaven't checked their site for yonks
00:17:48rasherHang on
00:18:04rasherI'm pretty sure they have old versions of everything lying around
00:18:05preglowunL33T2: have you tried deleting the database file before running moodlogic?
00:18:17rasherbut their site... damn!
00:18:28preglowit's a feast for the eyes
00:18:32preglowand the patience
00:20:09unL33T2yes it is hard to find anything
00:20:23unL33T2the util on their site is "MoodLogic H Manager"
00:20:26unL33T2which is what I got
00:20:30unL33T2and it sucks
00:21:05preglowbut answer this: have you tried deleting the database file before running moodlogic?
00:21:38rasherI'm uploading my version now..
00:21:52rasherexcept nautilus is giving me hell about it
00:21:53rasherhang on
00:22:28unL33T2it is the iRivNavi.iDB file?
00:22:39rasherthat's the file it generates
00:22:51unL33T2I have tried deleting it several times ... I am noticing that it is always only 448k ... is this correct?
00:23:01rasherdoesn't sound unreasonable
00:23:10unL33T2for 3000 songs
00:23:30preglowsounds reasonable, yes
00:23:30unL33T2ok I deleted it
00:24:02unL33T2MoodLogic > Rebuild MoodLogic Service Database
00:24:12unL33T2took a whole 5 seconds
00:24:27rasherSounds wrong... I'm going to blame moodlogic
00:24:36unL33T2yeah I thought so too
00:24:55unL33T2I've tried it on two PCs and uninstalled/reinstalled it several times
00:25:00 Quit ashridah ("Leaving")
00:25:00preglowiriver's db is a friggin joke
00:25:21unL33T2i seem to recall it taking longer the first time after a firmware flash
00:25:36preglowpretty irrelevant, i think
00:25:38unL33T2but now it doesnt ... the only thing I can think of is it doesnt like one of my music files and gets stuck on it"> (not uploaded yet though)
00:26:06unL33T2it does have quite a large list of songs in the dbfail.txt file
00:26:17unL33T2apparently if they have a long filename it gets angry and rejects them
00:26:38unL33T2rasher: let me know when it's ready
00:26:59preglowunL33T2: iriver's db format is very poorly thought out
00:27:06preglowunL33T2: it's got a fixed file name limit
00:27:20rasherunL33T2: you can get"> now (if I haven't mistyped the url)
00:27:25 Quit linuxstb ("CGI:IRC (EOF)")
00:27:41unL33T2preglow: do you happen to know what the limit is so I can correct it?
00:27:50preglowrasher: you don't believe in copy&paste? :)
00:27:51 Quit unL33T2 ("CGI:IRC")
00:27:56 Join unL33T [0] (
00:28:16unL33Tbah I clicked on your link and it cancelled my sign in :P
00:28:28unL33Tcould you put it back up again :P
00:28:42rasherpreglow: not much to copy&paste from
00:28:52unL33Tit didnt work either .... but I have a javascript blocker
00:29:00unL33Tso I'll try it with that disabled
00:29:05rashersign in? javascript?
00:29:16unL33Tnvm, can you just please put the link back up?
00:29:24unL33TI am in the web client
00:29:29unL33TI dont have IRC installed
00:29:39unL33Tso when I clicked on the link it brought me out of this page
00:29:40rasherunL33T2: you can get"> now (if I haven't mistyped the url)
00:29:45rasherAh, I get it now
00:30:25unL33Ti get 404 error :(
00:30:29unL33Tfile not found on this server
00:30:30 Join linuxstb [0] (
00:30:53 Quit Strath (Read error: 104 (Connection reset by peer))
00:30:57rasherokay, hang on
00:31:23preglowlinuxstb: alac going well? no news on aac, i take it?
00:31:47rasherthat hould work
00:32:08unL33Tits downloading
00:32:37unL33Tstraining the T3 connection I'm on right now
00:32:41unL33Tlol, joking
00:34:03 Join Strath [0] (
00:34:23rasherThink it's hosted on a 2mbps dedicated line.. works well enough most of the time
00:34:33linuxstbpreglow: ALAC's fine. About AAC, I haven't done anything more apart from getting libfaad to compile.
00:35:03linuxstbBut maybe the Helix decoder would be better - but I'm not sure about the Real license.
00:35:10preglowyou're not alone in that
00:35:31preglowthe faad people seem to be going nazi on their license now anyway
00:35:55rasherthe rpsl is nowhere near gpl compatible
00:36:32linuxstbThe ipodlinux people are using the Helix decoder in podzilla - which is GPL'd
00:36:44linuxstbrasher: That was my initial understanding as well.
00:36:57rasherThey shouldn't be, tbqh
00:37:19Zagorwhich license are they using?
00:37:50preglowthe ipodlinux guys are doing much they shouldn't
00:37:56preglowi think they're using a binary mp3 decoder
00:39:25unL33Tthis other software is taking longer so probably generating a proper file
00:39:52preglowyes, scanning 3000 files is not supposed to be a five second job
00:40:02linuxstbZagor: Who are you talking about?
00:40:05preglowi'll wager this keyboard on that
00:40:11rasherunL33T: you could try using - I think it generates proper databases
00:40:14Zagorlinuxstb: faad
00:40:22preglowit did the last time i tried it, at least
00:40:35linuxstbIt's GPL'd, but they have added extra advertising conditions.
00:40:54rasherSo it's not GPL'd
00:41:00linuxstbThere's some links on the SoundCodecs page to a discussion at HydrogenAudio
00:41:05preglowfaad2 == nero codec, yes?
00:41:11linuxstbpreglow: Yes.
00:41:22linuxstbThe requirement is to advertise Nero.
00:41:28preglowbloody idiots
00:42:24linuxstbIt could be argued that it's just displaying proper copyright attribution - but a FSF lawyer was quoted as saying it was incompatible. This was a couple of months ago, I don't know the current situation.
00:42:53linuxstbThe change happened in about February this year I think - so I've been working with the last "free" version of libfaad from their CVS.
00:42:59rasherCan anyone fine the addition?
00:43:44preglowfining it sounds more sensible
00:44:05preglowand everyone responsible
00:44:48rasherCan't find the link to HA
00:45:01unL33Ti had 110 files "no db" and 103 files "fail" anyone know what that means exactly
00:45:03rasherh, there
00:45:32linuxstbComment #35 is the reply from the FSF lawyer.
00:45:47rasheroh jeez
00:45:59rasherI don't need an FSF lawyer to tell me that won't hold
00:46:35linuxstbIt seems he's not a lawyer, but "GPL compliance engineer" :)
00:46:39unL33Tinteresting, i still can run the DBscan feature on the iriver and it still wont play my songs
00:46:50unL33Tbut I gotta go get some food. I'll come harass you people some more later :P
00:47:04rasherThere's no "run the DBscan feature"
00:47:13rasheryou only need to turn that on once
00:48:24rasherunL33T: What happens when you try and play a song using the DB?
00:49:16rasher"It is not an additional restriction, it clarifies how we interprent this part of the GPL."
00:49:28rasherhow you interpret it doesn't matter, silly!
00:50:10preglowbeat them with sticks until they understand
00:50:25rasherI'd go for that solution as well
00:50:41preglowif you're going to be picky about advertising, don't bloody gpl your source
00:50:57linuxstbSo we agree that we can't use the current version of faad2, and must stick with a "pre-advertising" version?
00:51:04preglowlinuxstb: indeed
00:51:24 Join Sucka` [0] (n=NNSCRIPT@
00:51:29rasherDoes it matter?
00:51:37rasher(ie. were there any interesting changes since?
00:51:59linuxstbProblem is that it seems no-one is making a fuss about it - apart from rjamorim
00:52:12rasherFAAD2 ATE MY GPL
00:52:37preglowlinuxstb: well, he's a sensible guy, most of the time
00:52:40preglowa lot of people aren't
00:53:33preglowmost people don't think licenses matter
00:54:01Zagorthe thing is we aren't allowed to distribute faad with rockbox under that license. all of rockbox must use the GPL.
00:54:26rasherJust use the CVS version from a while ago
00:54:32rasherand ignore them
00:54:33preglowhe does, so no problem
00:54:48Zagoryeah, I just mean the issue has merit.
00:54:50preglowproblem is, they're probably going to implement a few things in the coming time
00:55:02preglowbut let's hope someone talks/beats some sense into them
00:55:14rasherThey seem to be under the misconception that people have to advertise them
00:55:31rasherI seem to recall a certain debackle about the old BSD license because of EXACTLY THIS
00:56:03CtcpVersion from Rick!i=rick@unaffiliated/Rick
00:56:03***Server message 505: 'logbot :Private messages from unregistered users are currently blocked due to spam problems, but you can always message a staffer. Please register! ( )'
00:56:08preglowpeople should care about licenses so they don't go about repeating mistakes like this
00:56:09Zagorthat's a good point actually. dig up a good summary of the old bsd debacle and post it there.
00:56:21RickWhat's logbot running?
00:56:32linuxstbDo does the fact that license X is "GPL compatible" mean that I can take that code and distribute it under the GPL?
00:56:44rasherActually, I'll just dig up the offending line from the old BSD license and point them at something KNOWN to be GPL incompatible.
00:56:55Ricklinuxstb: no, it means it does not violate the terms of the GPL
00:56:57rasherlinuxstb: I think it means absolutely nothing
00:57:06rasheroh, sorry.
00:57:06Rick(i believe)
00:57:14Zagorlinuxstb: it means you can distribute it together with GPL software
00:57:15rasherI was thinking in the context of RPSL
00:57:29 Join tvelocity [0] (
00:57:35linuxstbZagor: What's the difference between that and distributing it by itself under the GPL?
00:57:36rasherall of rockbox doesn't have to be GPL, by the way
00:57:47rasherit could be GPL+BSD
00:57:57Zagorrasher: right, it just needs to be compatible
00:58:00rasheras long as the other license doesn't add restrictions.
00:58:18rasherZagor: yup, just clarifying in case someone misunderstood
00:58:33Rickpreglow: er, I thought dancer was the ircd?
00:58:45Zagorlinuxstb: you are not allowed to change licensing terms on other peoples' property. even if those terms give you permission to redistribute it.
00:58:48rasherRick: there's a dancer ircd and a dancer irb-bot
00:58:49preglowRick: no, bot as well, i believe
00:59:13rasherZagor: you can relicense BSD stuff, can't you?
01:00:07rashernevermind me.
01:01:05rasher"You would be best using Reals HE-AAC implementation, with a less restrictive license:" NOOOOOOOOOOOOOo
01:01:07rasherhe didn't!
01:01:09 Quit Sucka (Read error: 110 (Connection timed out))
01:01:14 Quit ender` (No route to host)
01:02:05rasherlinuxstb: post #29 already did what I'm aboout to do
01:02:14rasherApparently I just turned Canadian.
01:07:54linuxstbJust read that the Helix AAC integer decoder is highly optimised for the ARM - which explains the appeal to ipodlinux
01:09:12rasherStill don't explain how they think they're allowed to combine RPSL code with gpl
01:09:28preglowok, i've degenerated to one-liner cvs commits
01:09:34rasherthe rpsl is chuck full of additional restrictions right from the get-go
01:10:13fuzziegenerally, people are happy ignoring license terms in favour of convenience, so
01:10:38rasherI don't understand this.
01:10:55rasherAt least not if you're a fairly prominent community effort
01:11:15fuzziethe whole fact they're running linux on the thing doesn't exactly speak of sanity
01:11:24Zagorhonestly, I wouldn't say ipodlinx is very prominent
01:11:28rasherIf you're breaking licensing terms in your own basement, I guess I could sortof understand it
01:11:57fuzziewell, embedded hw companies are often happy to break the gpl
01:12:25fuzziethere are still embedded devices out there with broadcom drivers statically linked to a linux kernel
01:12:28rasherYeah, that's particularly nasty.
01:12:42rasherAnd lots based on busybox
01:12:55preglowdidn't good old iriver have one of those cases?
01:13:05Zagorthere are also a heapload of linux ports for targets where linux officially does not run
01:13:17rasherMy parents have an Allnet 0186 :-\
01:15:04Zagorrasher: how about popping into #ipodlinux and ask?
01:15:14rasherI don't think any good will come of it
01:15:22Zagorprobably not
01:15:40rasherI think coob commented it at some point
01:18:17 Quit DrMoos ("Glory to Rockbox")
01:19:22rasheror not
01:23:43rasherI just asked them
01:25:20fuzzieand got a response?
01:25:29rasherNot yet
01:25:35rasherwe'll see
01:27:56rasherpreglow: what did your commit fix?
01:28:13preglowno fix like that
01:28:20preglowjusdt saved a few cpu cycles
01:28:45rasherDidn't know what an "emac stall" would do
01:29:01preglowan emac stall is what happens if you try to get a result out of the emac unit before it's done
01:29:10preglowso i just made the routine do something else while waiting
01:29:28rasherMakes sense.
01:29:36rashereven to me
01:29:36preglowahh, but this might not work at all, the mulu instruction uses the emac unit on 5249... oh well, at least i didn't make it any worse
01:31:41Zagori'm off to bed. night guys.
01:31:41preglowi was just reading through the tremor assembler and caught it
01:31:50preglowguess i should go to bed myself
01:34:34 Join bagawk [0] (n=lee@unaffiliated/bagawk)
01:36:35rasherShut up, lilo.
01:37:31bagawkrasher, Those are harsh words to lilo.. He could kline you in a heartbeat :)
01:38:03rasherHe could.
01:38:13 Quit paugh ("bbiab")
01:38:18 Quit tvelocity ("Leaving")
01:38:29rasherHis endless chattering just annoys me.
01:40:55rasher it's a Rockbox!
01:47:33 Join Febs [0] (
01:47:44rasherEvening Febs
01:47:48FebsHi there.
01:48:01FebsI see that wrecker restored the ManualMainMenu page.
01:48:10rasherYeah, an epic task well done.
01:48:39FebsDefinitely. Although he deleted the table that I was working on as a concept for how to organize that front page. But I can pull that from the earlier version.
01:48:41*rasher considers creating a list of pages with missing attachments
01:49:03rashermore lists, more work
01:49:22FebsSo true.
01:49:38FebsThis is what I was playing around with for the manual:
01:49:51FebsBut the wiki table format just doesn't achieve what I was trying to do.
01:50:00rasherBut it's better than having the pages vanish in their current state
01:50:46rasherI'm not really understanding this
01:50:56rasherBut I'm sure it was a plan
01:51:53FebsI was trying to create a pseudographic depiction of the main menu hierarchy. It would be easier to see if there were a way to join vertical cells.
01:52:04FebsI may play around with an html table.
01:53:06rasherWas "breakingthelawforrecording" deleted, or did I just falsely mark it as restored?
01:53:11FebsWhat I had in mind was to try to show something like the "User Interface" section of this page:
01:53:29Febsexcept with hyperlinks to the appropriate entry in the manual.
01:54:26FebsI couldn't figure out any way to do that sort of text outline (which requires the [verbatim] tag) and still include WikiNames.
01:54:43rasheruse <pre>
01:55:06rasherthe wiki won't mind it
01:55:35FebsHmm. I'll try it out. I thought that did the opposite of what I am trying to do.
01:56:04FebsI WANT the wiki names to create links. But I want to use fixed width spacing.
01:56:25rasherI'm pretty sure <pre> will do that
01:56:32FebsI'll play with it.
01:56:42rasher<verbatim> will tell the wiki not to touch WikiWords
01:56:58rasher<pre> will just be stuck in the html, and the wiki will continue
01:58:45FebsI just had another idea ... I could take actual screen shots and use ImageReady in Photoshop to create slices so that people could click directly on the menu option in the screen shot.
01:59:18FebsActually, though, that would be more of a gimic than a practical way to maintain the online manual, since it would be a PITA to keep up to date.
01:59:27rasherAnd blind users would kill you
01:59:33FebsExcellent point.
01:59:44rasherThey might do that with your current idea as well
02:00:04FebsAnother good point.
02:00:31FebsI like the way the current table of contents shows the menu hierarchy, but the problem is that the page gets too long.
02:01:14FebsAnd it's different to follow the hierarchy when you're far down on the page, away from the TOC.
02:02:02rasherThat's true
02:04:00*rasher sicks w3c's link checker on DataSheets
02:06:29rasherouch, 22 broken links
02:11:47preglownite, people
02:11:50 Quit preglow ("eh")
02:13:07***Saving seen data "./dancer.seen"
02:15:02 Join paugh [0] (n=kickback@2001:5c0:8fff:ffff:8000:0:3e03:6822)
02:15:37 Nick paugh is now known as AliasCoffee (n=kickback@2001:5c0:8fff:ffff:8000:0:3e03:6822)
03:05:00unL33Tbeen gone for awhile but I'm back. I reformatted my iRiver and loaded only one album onto it and rebuilt the database and it worked through the sorted menus.
03:05:15unL33Tso now I am loading all my music back onto it and hopefully it will continue to work
03:06:41unL33Tthanks to everyone that was trying to help me with it (rasher and preglow)
03:07:19rasherCould be your filesystem was slightly upsetting the database generator in some way
03:07:46unL33TI suppose I should've run chkdsk first ... i didnt even think of that
03:07:48rasherI've no idea how well the generator works, haven't ever used i
03:08:09unL33Twell I like the one you sent me better than moodlogic
03:08:24unL33Tmoodlogic runs in the background all the time and I can never tell when its doing something and when it isnt
03:08:51unL33Tand everytime you move any kind of audio file on your computer it starts scanning it and updating some mystery database it keeps of all the music on your computer
03:09:20unL33Tat least I think that's what it does ... it makes working with audio much slower and annoying because I have a stupid message pop up everytime I move, copy, or generate an audio file
03:10:43rasherWow, that sounds ghastly
03:11:43unL33Tyes horrible. so thank you very much for the older iRiver software
03:16:54rasherYou're much welcome
03:18:29FebsunL33T, have you tried TDT for building the database?
03:19:32FebsMuch better than MoodLogic.
03:21:14unL33Twow that looks pretty good
03:21:52unL33Tlooks even better than the older iRiver software because it autorenames to fix the long name limit and supports many file times
03:21:59 Quit cYmen ("zZz")
03:22:57unL33TFebs: thanks
03:23:35*rasher mumbles something about having mentioned tdt already
03:24:14 Join aga [0] (
03:25:50agato confirm the need to replace a bad IC for a charging problem, does the jb recorder 20 [rockbox 2.4, 1.8, not flashed] present with charging screen active with charger connected, yet actually results in draining the batteries?
03:26:19agatrying to rule a few things out
03:28:04unL33Tlol rasher, I didnt notice you say anything about TDT...
03:28:15unL33Tif you did then thank you too =)
03:28:56 Quit bagawk ("Leaving")
03:29:28rasherIt was sortof mashed in between a lot of other comments, don't blame you for not noticing
03:34:33unL33Ti hope my battery doesnt go dead while transferring songs... I only have one bar left and still 10 min left to transfer it back
03:34:44unL33Twhy can't usb2.0 go faster :P
03:38:29 Join koniu_ [0] (
03:47:29rasherHCl: Is the makedbj.bat and SongDB.jar on TagDatabase up to date?
03:49:37 Quit unL33T ("CGI:IRC")
03:54:28 Quit Sucka` ("a bird in the bush is worth two in your house")
03:58:27agaso noone has ever had to replace that IC for a charging problem? you guys have all upgraded
04:01:13 Part aga
04:05:46 Join QT_ [0] (i=as@madwifi/users/area51)
04:10:21 Join hd [0] (
04:13:08***Saving seen data "./dancer.seen"
04:15:59 Quit QT (Read error: 110 (Connection timed out))
04:16:06 Quit Strath (Read error: 104 (Connection reset by peer))
04:19:27 Join Strath [0] (
04:20:14 Quit goa (Read error: 110 (Connection timed out))
04:20:15 Nick hd is now known as goa (
04:29:41 Join atubbs_ [0] (
04:34:38 Quit atubbs (Read error: 110 (Connection timed out))
04:36:07 Join Mxm`Pas`Bien [0] (
04:36:07 Quit Maxime (Read error: 104 (Connection reset by peer))
05:00:16 Join bagawk [0] (
05:24:38 Quit HCl (Remote closed the connection)
05:24:41 Join HCl [0] (
05:32:12 Quit bagawk ("Leaving")
06:05:51 Quit koniu_ ("how many cans must i stack up?")
06:13:10***Saving seen data "./dancer.seen"
06:29:15 Quit AliasCoffee ("Leaving")
06:48:59 Join LinusN [0] (
06:49:25rasherMorning LinusN
06:49:49rasherI found a bunch of DataSheets
06:50:03rasherso now DataSheets is mostly back to what it was
06:52:36 Join amiconn_ [0] (
06:52:36LinusNwhat is missing?
06:53:05rasher pcf50606.pdf Philips PCF50606
06:53:05rashermas35x9f_2ds.pdf Micronas MAS3539F
06:53:05rasherSCF5250UM.pdf SCF5250 Coldfire User's Manual
06:53:20LinusNi have those
06:53:25rasher has a list
06:53:38rasherI figured they'd be around
06:55:55rasherLots of missing attachments still. At leasat 127 unless I messed up
06:56:03 Quit amiconn (Nick collision from services.)
06:56:03 Nick amiconn_ is now known as amiconn (
06:56:08rashermorning amiconn
06:56:19amiconnrasher: This DataSheets isn't the latest
06:56:25amiconnThe MAS wav codec is missing
06:56:31amiconnI have those files
06:56:33rasherWhen was that uploaded?
06:57:18amiconnI don't remember exactly; [IDC]Dragon did that
06:57:38rasherAh yes, this is one of the pages that lost a few revisions
06:57:42amiconnA list of topics with missing attachments could be helpful imho
06:58:24rasherLast one available was 1.70(2005-08-12), should be 1.74(2005-09-14)
06:58:27LinusNok, all missing attachments uploaded
06:58:38rasherso DataSheets is done, excellent
06:58:54rasherI'll make a list, hang on
07:02:31amiconnLinusN: Not all...
07:03:06rasherall the ones missing compared to r1.70
07:03:27rasherwhich is our only real reference point
07:05:06LinusNstill broken links on that page?
07:09:56rashersortof disheartening
07:13:01amiconnDisassemblePlayer, DisassemblePlayer2, the archos part of HardwareSchematics, ReassemblePlayer, Rs232Converter, SerialPort should be simple, these are straight conversions of old /docs pages
07:13:21amiconnUartBoot is my task...
07:13:47rasherAh, that's good
07:13:53amiconnThe DeviceChart H-140 pics as well...
07:14:04LinusNthe iriver schematics are recovered
07:14:11rasherI have the igp100 internals
07:14:44LinusNthe HardwareSchematics page was renamed from IriverSchematics, and the attachments in the backup was still in IriverSchematics
07:15:29rasherNo broken links on DataSheets
07:15:41rasherAccording to w3c's link checker
07:17:55amiconnLinusN: Diff topic - as we now know that the bitshifting is a MAS bug that can't be avoided, we need to work around it
07:18:02amiconn(by walking the frames)
07:18:39amiconnAlthough I think that walking the frames shouldn't be difficult, implementing that will take some days, especially as it requires testing
07:19:00LinusNi think that should be a 2.5.1 release
07:19:01amiconnI think this doesn't justify to hold back 2.5 any longer...
07:19:42rasheror even 2.6 - in two months!
07:19:45amiconnEither we do a 2.5.1 release when it's done, or that has to wait for 2.6 (which will hopefully be released roughly on schedule)
07:20:04rasherit's not like things aren't much better now
07:20:34amiconnToday is the beginning of autumn. Is that a fine day for release? ;)
07:21:23amiconnOf course we should list that as known bug, along with a comment saying it's got much better since 2.3/2.4
07:22:32amiconnLinusN: and it seems you're right, there is a performance problem, not in our transfer routine, but in the MAS
07:22:44rasherMy release note draft has a "known bugs" section - I suggest putting some stuff from ReleaseTodo there
07:22:51amiconnIf the MAS has got too do too much at once it starts making errors
07:23:21amiconn...and transferring data adds to the total load, the faster we transfer, the more load it adds
07:23:53amiconnAfter all there is a reason why the archos recorder manual says q=7 isn't recommended...
07:24:33*rasher glares at
07:24:54rasheris this referenced anywhere now?
07:26:34rasherso much fun stuff outside the wiki
07:29:28amiconnLinusN: Btw, I think the recording code size will even get smaller when adding frame walking. A number of frame searches can go away, as audiobuf_write will always point to a frame boundary.
07:30:05amiconn...and we get an exact frame count, no need to read the mas or do crude estimates
07:34:36amiconn...and finally, we won't get a partial last frame in the file when stopping the recording
07:35:28LinusNtrue again
07:38:03amiconnThe mas overload theory also explains why disabling the peakmeter (only) reduced the error probability
07:46:21 Join t0mas [0] (n=Tomas@unaffiliated/t0mas)
08:06:53 Join ender` [0] (i=ychat@
08:10:30rasherSo the release is going to happen Real Soon Now?
08:13:15***Saving seen data "./dancer.seen"
08:18:50LinusNguess so
08:29:31 Join paugh [0] (n=kickback@2001:5c0:8fff:ffff:8000:0:3e03:6822)
09:02:48 Join _nrg_ [0] (n=nrg@
09:11:56 Join B4gder [0] (
09:12:21LinusNmorning B4gder
09:12:25B4gderhey ho
09:13:09 Join preglow [0] (
09:21:25*preglow freezes to death
09:43:11 Join markun [0] (
09:43:53 Join Dan [0] (
09:44:20Danhow can i install rockboy?
09:44:36B4gderrockboy is installed when you install rockbox
09:44:45B4gderbut then you need gameboy roms
09:44:54Danoh kool
09:49:12preglowabout time someone optimised rockbox :>
09:49:18preglowrockboy, even
09:51:10preglowall that's left for a release now is for zagor to press his button?
09:51:28*Zagor wakes up
09:51:40B4gderThe Button Presser (tm)
09:51:48 Quit Seed (Nick collision from services.)
09:51:56 Join Seed [0] (
09:52:23B4gderI certainly think so, but have amiconn and LinusN given their blessings?
09:54:12preglowlinus already has, i believe, and amiconn doesn't think it's worth delaying 2.5 for the recording fix
09:54:23preglowbut i bet there's some docs left to be done
09:54:32B4gderr r r r release!
09:54:41Zagor<amiconn> I think this doesn't justify to hold back 2.5 any longer...
09:54:52Mode"#RockBox +o B4gder " by ChanServ (ChanServ@services.)
09:55:01Topic"Buckle up" by B4gder (
09:55:09*preglow does the buckling
09:55:53B4gderthen there's nothing to wait for
09:56:02Zagormassive warnings building the tools. "scramble.c:103: warning: pointer targets in assignment differ in signedness"
09:56:16B4gderZagor: gcc 4?
09:56:35B4gderwe have LOADS of gcc4 warnings to fix
09:56:41preglowmost of the legitemate
09:57:07B4gderZagor: configure disables them on purpose for the regular (sim) build
09:57:21Zagorfine, i'll just ignore them for now then
09:57:29B4gderyes, we'll deal with them post 2.5
10:00:32 Quit solexx_ ("leaving")
10:01:33 Join solexx [0] (
10:05:47linuxstbShould we add to the CREDITS file people who have unknowingly contributed code to Rockbox - such as the codec library authors and the authors of other GPL'd code that we incorporate?
10:06:15B4gdergood question
10:06:18preglowi think the info should be somewhere, at least
10:06:18B4gderI think not
10:06:24preglowbut in what form, i know not
10:06:27preglowi think CREDITS is the wrong place
10:06:29B4gderbut we should definitely have it somewhere
10:06:35linuxstbIt's in the source.
10:06:51ZagorI think we should have it in the docs and webpage too
10:07:02linuxstbWhy not the CREDITS file then?
10:07:21linuxstbWe could add (e.g.) "Josh Coalson (libFLAC)"
10:07:31preglowbecause of his ugly code
10:07:36B4gderbecause I think CREDITS should be for the persons involved in Rockbox
10:08:12B4gderwe could have a second CREDITS for people who did things in projects we depend upon
10:08:30Zagori agree, or a special section in credits
10:08:42B4gderright, could be a section as well
10:08:46preglowin a related note, the credits are unreadable on irive r:/
10:08:58B4gderI have a fix pending for that
10:09:11B4gderwith the speed as a sine
10:09:18B4gderso that it stops occationally
10:09:18preglowB4gder: good idea
10:10:15preglowi've got a feeling we'll be needing a sin()-like function some place central in rockbox soon
10:13:19***Saving seen data "./dancer.seen"
10:14:02 Join ashridah [0] (
10:16:56ZagorB4gder: did you see the faad license discussion late last night? about 00:35 in todays log.
10:18:15B4gdernow I have
10:18:19B4gderbeen reading about that before
10:19:00B4gderlicenses always cause so much confusion
10:21:42B4gderit is an obvious GPL violation
10:25:30B4gderbut as you said, people tend to ignore license quirks that would make life hard to themselves
10:26:05B4gderlike those helix/nero licenses to any GPL project
10:26:49 Join linuxstb_ [0] (
10:33:36ZagorRockbox 2.5 is now released
10:39:07Topic"Rockbox 2.5 - yes, it is true. Go get it." by B4gder (
10:39:53B4gderand now, unfreeze!
10:40:14B4gderwho's gonna be the first to commit? ;-)
10:40:49 Join linuxstb__ [0] (
10:41:33 Quit linuxstb (Read error: 110 (Connection timed out))
10:42:54 Join webguest30 [0] (
10:44:45webguest30Hi, congrats on the release. Just curious, what happens when you release a new version, do you send an announcement to some places, or just advertise on the rockbox-site/mailinglist?
10:45:03B4gderon the site, on the list and on freshmeat
10:45:29B4gderis what we usually do
10:46:11webguest30Ah, ok. Should have guessed. Interesting with this open source community.
10:46:48webguest30I wasn´t really interested a while ago, but rockbox has opened up my eyes.
10:47:20 Part webguest30
10:51:37 Quit paugh ("Leaving")
10:53:46preglowand linuxstb is the winner!
10:54:51 Quit linuxstb_ (Read error: 113 (No route to host))
10:54:55 Join gzed [0] (
10:55:16linuxstb__Sorry, forgot to implement my change to plugin.[ch] first - I added the menu functions. It's committed now.
10:56:14 Part gzed ("Parting")
10:59:21preglowlinuxstb__: now hit us with some alac :P
11:02:31 Join [IDC]Dragon [0] (
11:02:56*[IDC]Dragon cheers for v.5
11:03:14[IDC]Dragon v2.5 rather
11:03:17linuxstb__preglow: I want to commit ALAC slowly - i.e. commit the original unmodified decoder first, and then apply my changes incrementally so we have a record of them in CVS
11:03:22linuxstb__I'll do it this evening.
11:04:48[IDC]DragonI need to get gcc-sh 3.3.6 running
11:05:08[IDC]Dragonto make the proper v2.5 bootbox build
11:05:58preglowwhy don't you use 3.4 for sh?
11:06:04preglowlarger code?
11:06:20[IDC]Dragontoo buggy
11:12:22 Join am1conn [0] (
11:12:29preglowhi, amiconn
11:12:29linuxstb__For anyone wanting try out Sudoku, you need some ".ss" files - these are simple text files containing the starting Sudoku games. You can download some examples from
11:13:37HCldoes this mean the feature freeze has been lifted?
11:16:10linuxstb__HCl: Yes.
11:17:38HClgood :p
11:17:45Dananyo make rockbox play .swfs yes?
11:18:05HCli still need to know whether everyone agrees to upgrade the hash to 128bit + a hashtype field
11:18:22HClwhats swf ?
11:18:46preglownow why the frigging hell would we want that
11:18:51HCli highly, highly, highly doubt that
11:19:02Danso i can have the dancing spiderman on my mp3
11:19:32preglowbloody flash advertising has almost driven me to suicide on itself so far, i don't want to expose my poor little h120 to it
11:19:35HClwhy hasn't anyone updated the webpage about the release? lol
11:19:38preglowit'll be bricked for sure
11:19:51am1connpreglow: The primary reason for gcc 3.3.x for sh1 is code size
11:20:25preglowHCl: well, i for one think it's a good idea, but then i'm not involved with the db at all
11:20:46HCli'll prog bagder again
11:31:44 Join Moos [0] (
11:32:10rasherJesus Christ, it happened!
11:33:16MoosHello guys, the release is here
11:33:25preglowi didn't notice
11:33:26rasherthough the website is oddly silent about it
11:33:27Moosalelouya my brothers :)
11:33:59Danno video plugin for iriver
11:34:29Danthat ever gunna happen?
11:35:29*rasher updates the front page
11:35:44preglowDan: not for a long time
11:35:56preglowDan: unless you mean the rvf format, which might happen soonish
11:36:12Danwell thatl do
11:36:12linuxstb__Someone is working on the video plugin, but I think we need to give plugins access to the audio codecs in order to play video with sound.
11:36:25preglowthat we do
11:37:12linuxstb__Or perhaps video playback should be part of the core rockbox...
11:40:28linuxstb__Does anyone else have any new features waiting to commit?
11:40:59preglowlinuxstb__: yes, i believe slasheri will try to persuade zagor & co to let him commit his new dir tree caching
11:41:12Moosthe V2 of Bagder :D
11:41:20preglowwhich i think sounds like a smashing idea, heh
11:41:58Slasheriyes, i will try to commit that and also enable the recording menu on iriver (but i don't have time before the next week to prepare these commits)
11:42:52preglowthen you're commiting the wav writer? :P
11:43:23Slasherihehe :D hmm, maybe when i have done all the pending bugfixes :)
11:44:54LinusNi suggest you add the cache code to the patch tracker so we can review it
11:45:15*rasher roots for the plasma plugin
11:45:20SlasheriLinusN: hmm, i can do that :)
11:45:22preglowooh, yes
11:45:28Slasheribut not before the next week :/
11:45:34LinusNno hurry
11:46:20Slasherii have done very minimal changes to existing files and i think the cache code itself is very error proof (it will disable itself if it finds any problems)
11:46:20preglowSlasheri: code is quite clean, yes? all the new code is in one file and only minor modifications to the rest?
11:46:29preglowthen stellar
11:46:34preglowi can see no flaws
11:46:54LinusNi just want to see it before you commit it
11:48:01preglowgreat news for us compulsive browsers
11:48:43rasherGuess I can start my patched builds again
11:54:00HClas soon as everyone agrees on my proposed database changes, i'll finish and commit those
11:54:21Moosplease do :)
11:58:35linuxstb__Could someone have a look at my iRiver-presets radio patch -
11:59:27linuxstb__It's trivial, but I've changed the way the items added the menu are selected - currently it's hard-coded for the Ondio, but I changed it so that any function not assigned to a key is added to the menu.
11:59:59linuxstb__I think it's OK, but would like someone else to say that that change is the right way to go.
12:05:52 Join XavierGr [0] (
12:06:51XavierGrHi all!
12:13:21***Saving seen data "./dancer.seen"
12:19:06preglowlinuxstb__: afraid i have no idea if it's good or not
12:20:13rashert0mas (I believe it was) asked the same question, and I believe B4gder said it looked fine. Better search the logs though
12:21:06rasherHm. Maybe it wasn't him.
12:21:27rasherAh, tucoz. Started with a t.
12:22:32rasherand it was amiconn who said he "can't imagine what problems should arise from enabling fm presets on iriver this way"
12:22:54rasherThat was for only adding || IRIVER_H100_PAD though
12:24:53ashridahbloody windows. how the hell does a cdrom drive on one ide controller screw up disk access to something on a completely different controller
12:25:32preglowashridah: with great ease and elegance
12:26:59rasherHrm, wav patch doesn't apply cleanly
12:27:00 Quit am1conn ("CGI:IRC (EOF)")
12:27:38preglowrasher: i believe lear is on that
12:29:24 Part LinusN
12:33:00linuxstb__Unless anyone shouts, I'll commit that radio patch this evening (UK time).
12:33:14preglowif amiconn said it looks good, i'm all for it
12:33:21markunI changed the battery indicator to always show graphical when the charger is inserted. It there a reason why this was not done?
12:33:54rasherIt's the wrong fix
12:34:56 Join StrathAFK [0] (
12:35:01rasheryou want to switch when *charging*
12:35:11rasherBut the iriver doesn't detect this yet
12:35:17preglowi don't think it ever will
12:35:48markunHow could you detect it? With an ampere meter maybe.
12:35:50rasherHm, didn't sound like that when I asked Linus about it
12:35:57preglowthen hooray!
12:36:06rasherThere was a patch to do just this
12:36:14rasherI may have misunderstood something I guess
12:38:06*preglow prepares for acting the fool
12:38:43rasher"<LinusN> yes, you can close it, with the comment that it will be solved in a different way when the charging detection is implemented"
12:42:19markunIs it possible to detect somehow if the charger led is on?
12:44:49 Join amiconn_ [0] (
12:49:30 Quit XavierGr ()
12:51:18 Quit Strath (Read error: 110 (Connection timed out))
12:55:29rasherAny good patches I need in my build?
12:56:42 Quit Dan ()
12:58:36 Join cYmen [0] (
13:03:16 Quit amiconn (Read error: 110 (Connection timed out))
13:03:16 Nick amiconn_ is now known as amiconn (
13:05:26 Join am1conn [0] (
13:12:14*preglow vanishes
13:17:16rasherIs there any reason not to accept ?
13:18:03B4gderapart from the C++ comments, no I don't think so
13:19:01am1connlinuxstb__: There are a lot of warnings for H1x0 since your sudoku commits...
13:19:38B4gderI guess it is time to remove the gcc4 warning inhibits soon
13:20:29linuxstb__am1conn: The Daily Builds page is OK. The first compile failed because I forgot to commit my plugin.[ch] changes first.
13:20:44rasherlinuxstb__: h3x0
13:21:02linuxstb__Yes, I noticed the h3x0 - but am1conn said H1x0
13:21:15rasherAh. I'm blind.
13:21:19rasherMaybe he did mean 3x0?
13:21:43linuxstb__I'll fix the h3x0 this evening - I don't think it's urgent.
13:22:03linuxstb__Unless people start shouting at me.
13:22:57*am1conn seems to be blind...
13:24:19 Join tomylee [0] (
13:24:28tomyleehi guys
13:25:43tomyleei have some questions, and i need help, my iriver doesnt work anymore, after using it with rockbox, playing an mp3 and playing minesweeper at the same time - made it crash
13:26:05B4gdercrash how?
13:26:31B4gderand when plugged in you used usb, then rockbox isn't doing anything
13:26:32tomyleethere was an error message
13:26:54tomyleewell, after the battery was empty, i was able to get to usb mode
13:27:17tomyleewhat i did first, was to delete all the rockbox files on the disk to make it start with original firmware
13:27:39tomyleethats where i am now
13:28:02B4gderand you've made sure you've charged it a good while?
13:28:18B4gderrockbox can still start when the original can't
13:28:53tomyleeso you mean I should charge it a while, before trying this?
13:29:03B4gderlet it charge for an hour and then retry
13:29:21linuxstb__Does the bootloader automatically run the original firmware if rockbox.iriver isn't found?
13:29:27B4gderlinuxstb__: yes
13:30:17rasherB4gder: I'll add the remote-hold-fix then
13:30:34rasher(works correctly)
13:30:54B4gderme and linus will be away starting soon til the end of the weekend
13:31:15tomyleewhat about the archos players, why are there so many features implemented by rockbox, and so few in the iriver...? were the archos players first on the market, or are they the better hardware to work with?
13:31:31preglowcan't exactly say we lack many of the archos features
13:31:44B4gderarchos was first but mainly we started working with those players
13:32:17B4gderbut I agree, there aren't that many features left that archos has that iriver doesn't
13:32:39*preglow woots the langv2 commit
13:32:52B4gderit is still early though
13:33:00B4gderjust parsing the XML-like file
13:33:12preglowstill, means works is being done
13:33:19preglowbut now i need to leave
13:33:22tomyleeyeah, i cant await the recording feature for irivers, mainly the lossless mode, and the continuous recording of very big files
13:33:57rasherI'm now creating patched builds again, if anyone's interested:">
13:34:00linuxstb__Does "langv2" imply plugins will now be translated, or is that a different issue? I haven't been following the langv2 development.
13:34:14B4gderit is actually a HUGE take
13:34:20B4gderso yes, it means translated plugins
13:34:24B4gderand many other fixes
13:34:33B4gderlike target-adjusted translations
13:36:03linuxstb__Is there a Wiki page about it?
13:36:25B4gderonly my write-up
13:36:34B4gdershould of course be converted to a wiki
13:37:55*linuxstb__ notices that if he adds an iPod target to the build system it will be target #13
13:39:33linuxstb__I need to decide how we are going to split the ipod targets - there are about 7 or 8 different devices. ipodlinux just does everything at runtime but I don't think that's right for Rockbox - espeically the different LCD sizes and depths won't work.
13:40:19B4gderthen start with a target for the first model you go with
13:40:36Zagorlinuxstb__: we like many targets. witness the spectrum of archos ones :-)
13:40:44B4gderthe more the merrier!
13:41:13*rasher fiddles with FM presets
13:41:19linuxstb__I need to decide how to name the #defines - e.g. if particular features depend on processor, LCD type, keypad type etc. ipodlinux isn't clear in the distinction, so it needs a little research to do it right.
13:41:40rasherI'm expecting a long joystick press to bring up something..
13:42:03linuxstb__rasher: Are you talking about my radio patch for iriver?
13:42:34rasherJust added it to my customized build.. I'll never be able to cvsup this mess
13:43:15Zagorby the way, are you regular iriver users happy with the button bindings? As an old archos user I find them terribly confusing.
13:43:41rasherI'm mostly used to the way it is now
13:43:46markunme too
13:43:55rasherBut the lack of context-menu on the radio screen confuses me.
13:44:03linuxstb__I'm used to them as well, but I never used the iriver firmware, and I haven't used Rockbox on Archos for a long time.
13:44:04rasherSince it's available everywhere else
13:44:44rasherThe record button should do something though.
13:45:02rashernot quite sure what it should be though
13:45:15rashera long keypress could go to the record screen?
13:45:20tomyleeyes record
13:45:28rashernot a short one please
13:45:39linuxstb__I mean the record button should bring up record - everywhere.
13:45:54rasherYes, but a long keypress please?
13:46:06rasherSo you don't get yanked out of something
13:46:07tomyleeyes long, why not, it is original like that
13:46:12tomyleepeople are used to it
13:46:15tomyleejust as me
13:46:45rasherI don't think short keypresses should interrupt what you're doing (of course, STOP being an exception - I'm sure there are others like that)
13:46:49B4gderin general I'm against having a single function to a button all over
13:47:25B4gdersince it restricts the ui so much
13:47:33rasherIt makes sense though, and we have this rejected baby button that's not doing much anyway
13:47:44B4gderyes, in this case it might not be that bad
13:47:53B4gderbut I wouldn't want it to start a trend
13:48:15B4gderlike the good old "I want up/down to change volume everywhere" etc
13:48:26rasherPeople actually said that?
13:48:35B4gderlong time ago though
13:48:50B4gderbut people always tend to come up with a button they want to act the same all over
13:49:13FebsIt would be great to have dedicated volume controls, but not at the expense of being able to navigate menus!
13:49:50rasherHm, shouldn't one of the menus on the fm screen be activated like the context-menu is elsewhere?
13:49:54rasherFor consistency
13:50:02rasherCan't figure out which one though.
13:50:41tomylee(hmm, i can hear parts moving around in my iriver since the last phisical crash... :(
13:50:58rasherphysical crashes will do that I'm afraid
13:51:27tomyleeyes...but it loaded to usb mode, thats a good sign though
13:51:56tomyleei hope everything is fine after the hour of charginc, as b4gder recommended
13:53:18B4gdersee ya monday
13:53:19 Quit B4gder ("time to say moo")
13:53:39tomyleeabout interface, i would suggest sollofing: keeping the funktions froms the original as far as possible, and advancing the functions (of buttons maybe) only in submenus, somehow hidden - so the normal user will be comfortable with rockbocky anyways
13:54:18Zagortomylee: the original navigation is so horrible I'm afraid that's not an option
13:54:34Zagoralso Rockbox is not an iriver improvement firmware. it's multiplatform.
13:55:23tomyleeyes, the navigation has to be changed, but the side buttond functions are logical in some way
13:55:46rashera-b does so much more on iriver as well, I don't think I
13:55:48linuxstb__I agree - it's more important for Rockbox to be consistent across platforms than consistent with the "native" firmware of each target.
13:55:53Zagoryes, but keeping them to their labeled purpose only would severely cripple the interface
13:55:56rasherI've ever actually *used* it for a-b repeat
13:56:02 Join tucoz [0] (
13:56:43ZagorI think we should have a big button discussion before the first iriver release. I believe we can improve it.
13:57:02tucozpsst, joystick-press should be renamed to something else than play (like in the confirm delete action). But I believe that is another issue.
13:57:26Zagortucoz: it's a related, and adds to the confusion
13:57:49Zagor"select" is pretty good
13:57:56tucozBut, the langv2 will be able to handle that better i suppose
13:58:09tucozdifferent names for different platforms
13:58:53FebsBut I agree that joystick press needs to be something−−anything−−other than "Play." Having it called "Play" when there is already a play button has made it very difficult for me to write documentation.
13:59:00linuxstb__Probably my biggest wish for Rockbox is a sort of task-switching concept. So a user can "alt+TAB" between screens such as WPS, the file browser and a plugin.
13:59:02tucozSelect is good. Hmm, then "long press select"
13:59:20tucozfor long presses?
13:59:51Zagorlinuxstb__: well we have it for wps/browser already. plugins are more difficult since we don't have any control over them. it would require dual screen buffers etc.
14:00:16tucozlinuxstb: cool idea. Maybe the record-button could be used for that
14:00:26linuxstb__Zagor: Yes, I just want to take it slightly further. So each screen would need it's own lcd buffer.
14:01:06tucozfor easy access between different parts of rockbox perhaps?
14:01:42Zagortucoz: not "why the feature". why a buffer for each screen. we already do it fine with a single buffer for everything.
14:01:51tucozOh, sorry
14:02:09Zagorplugins could work too if they were modified to accept a "redraw" message
14:02:11linuxstb__Zagor: Perhaps you're right. I need to give it more thought.
14:02:58Zagorthe thing is plugins currently aren't written to be interrupted. it potentially introduces problems.
14:03:21Zagornothing we can't overcome. just things we need to consider.
14:03:23*Febs has another thought on the joystick press.
14:03:23tucozso it is only one buffer that fits the pixel-values, and when wps/browser is called, a redraw is called?
14:03:32Zagortucoz: yes
14:03:41FebsThe H300 has a "Navi" button that is the functional equivalent of a joystick press.
14:04:01FebsIf we called the joystick press "Navi" it would eliminate differences at least between those two platforms.
14:04:04ZagorFebs: I hate it when they label the buttons...
14:04:28Zagorbut it's not a very logical name, so it would suffer for other platforms
14:04:50ZagorI think "navi" sounds like "go into navigation screen", i.e. file browser
14:05:12FebsWhich, of course, is what it does in the iriver firmware.
14:05:16tucozok, hmm. well, then the plugins would have to push their current state, and pop it on when a redraw is called? Or is this an awkward way of doing it?
14:06:16*rasher wonders why he's not seeing a suduko plugin
14:06:28Zagortucoz: that's an issue too. currently the wps/browser is a forward->back switch. introducing a third screen (fourth if you count menus) could be complex
14:06:55linuxstb__rasher: It's technically a viewer, not a plugin.
14:07:05rasherlinuxstb__: ah, that'd be why
14:07:30rasherAnother need for plugins-that-are-also-viewers, I guess
14:07:49rasherat least once you create a generator
14:08:00tucozbut then again, I guess the plugin could just idle. The plugins are given their own memspace. If toggled, then the plugin "freezes" (keeping it's memspace). Upon return, it unfreezes.
14:08:09linuxstb__rasher: I agree
14:08:14 Quit bluebrother^ (Read error: 104 (Connection reset by peer))
14:08:16 Join bluebrother^ [0] (
14:08:16 Quit cYmen (Read error: 104 (Connection reset by peer))
14:08:17 Join cYmen [0] (
14:08:19rasherSo where do I find something to use it with?
14:08:57rasherAh, there
14:09:04rasherJust had to rtfs(ource)
14:09:23linuxstb__I also quoted a link in IRC earlier today.
14:09:25tucozuse the source luke ;)
14:09:39Zagortucoz: memory is not the problem, the code path is. currently the entire UI is a single thread, which means functions have to return for earlier screens to run. the plugins don't return until they quit.
14:10:08tucozah, get it. Then the plugins would have to be given their own thread then.
14:10:34Zagorso we'd either have to create some elaborate scheme where each plugin can serialize their state to disk or ram, or we'd have to run plugins in a new thread with all the problems THAT causees
14:11:26Zagorthe problem with two threads that both draw to the screen should be obvious :-) locking would be required, and the bugs emanating from that are... unfun.
14:12:02rasherlinuxstb__: that looks mighty nice
14:12:03Zagorpart of the elegance of todays system is that plugins don't have to care about the rest of the system. this makes them simply to write.
14:12:15rasherlot's of clicking though :-\
14:12:22tucozYes, then again. Plugins might not be the most important feature for toggling. Maybe only the 'core'-functions should be given this opportunity. Like, record, browser, wps and so on.
14:12:54tucozMight be a good idea to keep the plugin-interface simple.
14:12:56linuxstb__rasher: Tried the solver yet?
14:12:58 Join webguest92 [0] (
14:13:00Zagortucoz: right. which we have today already... :-) except record, but I don't quite see the use for that either.
14:13:25***Saving seen data "./dancer.seen"
14:13:29tucozMe neither. But, maybe for playback of the recorded file?
14:13:33Zagorin the end, wps/browser toggle is pretty much all you really use.
14:13:42rasherlinuxstb__: wow.. that's snap
14:13:43tucozI have almost never used record
14:13:48linuxstb__Isn't the solution to simply have a different LCD buffer for plugins - this could be implemented without having to change the plugins.
14:14:07linuxstb__And lcd_update wouldn't do anything if the plugin didn't have "focus"
14:14:37rasherlinuxstb__: I'm going to join the people wanting to use "click to enter edit mode, then use up/down/left/right to change the number, click to exit" though
14:14:57tucozZagor: then again, the coolness factor
14:15:02linuxstb__rasher: Maybe that's a feature for your builds? :)
14:15:04Zagorlinuxstb__: you still need to stop plugins. imagine having "mandelbrot" running in the background all the time.
14:15:41Zagorstopping specific threads means a more advanced scheduler, and new bugs
14:15:49rasherlinuxstb__: I guess I could do that
14:15:55Pieter__yeah, or switching a game of rockblocks to change to a new mp3 while the game keeps running.. pretty useless if it doesn't stop :p
14:16:46linuxstb__Zagor; I'm saying it's going to be easy. But from a user's point of view, I think it would be nice.
14:17:21linuxstb__rasher: Ideally I would like to incorporate both methods of navigation - if you can think of a way to do it, preferably without an option.
14:17:28tucozBut, I think that a lot of the plugins could be improved. The stop button is simply too powerful these days. Like, pressing stop always exits w/o a confirm dialog. But, that is upto the plugin iguess.
14:17:49rasherlinuxstb__: hm.. maybe just use play to enter edit mode?
14:17:53linuxstb__What about using a different key to enter "insert mode", not a joystick press.
14:17:59rasheror long joystick press
14:18:10linuxstb__Long joystick press repeats
14:18:18Zagortucoz: yes it is
14:18:40rasherplay or record.. hm
14:18:46tucozok, maybe people like it the way it is
14:20:11tucozsomeone should implement exposé for rockbox!
14:20:30rasherlinuxstb__: record, I think
14:21:45rasherlinuxstb__: want me to create a patch?
14:22:43linuxstb__rasher; Please do.
14:23:17linuxstb__Can also you also implement a random Sudoku game generator as well whilst you are there? :)
14:23:33rasherof course, this will confuse the main loop a bit
14:24:34linuxstb__I will try and do it if you want me to - but probably not until the weekend.
14:24:53rasherthink I'll have a go
14:25:21 Join DrMoos [0] (
14:25:23linuxstb__OK. The important thing is that when you leave "edit mode", you need to check that the value in that cell is legal.
14:25:37 Quit Moos (Read error: 104 (Connection reset by peer))
14:25:49rasherI'll do that
14:26:08linuxstb__The "legality" check is incremental - i.e. it assumes the board starts legal, and then only checks the effects of each change.
14:27:25 Nick DrMoos is now known as Moos (
14:27:30*rasher has no idea which button to use for the archos devices
14:29:15tucozlinuxstb_: lot's of soduku links
14:30:14tomyleeyes - my iriver is not dead... hell yeah
14:30:19tomyleeday saved
14:31:51 Join hshah [0] (
14:32:12rasherHm... should I ignore SELECT when in editmode?
14:33:41rasherehm, that'd be TOGGLE
14:34:55_nrg_hi, does rb only play withing a directory or can it also play from dir to dir without playlist?
14:35:14tucozlinuxstb_: is the solver bruteforcing, or using graph coloring as suggested by wikipedia?
14:35:59linuxstb__tucoz: I've no idea - I just stole someone else's code without looking at it.
14:37:09*rasher creates a changedigit function to minimize duplication
14:39:45rasherlinuxstb__: Why do you "ticks=*rb->current_tick;" when incrementing a digit?
14:40:07tucozI found a forum for soduku programmers, here is a link to a bf-generator.
14:40:55linuxstb__rasher; To slow down the auto-repeat.
14:41:03 Quit hshah ("Leaving")
14:41:11linuxstb__i.e. I ignore any repeats that happen within 1/3 of a second of the previous one.
14:41:51rasherAh, I see it now
14:42:47linuxstb__Gotta go now. back later.
14:43:27 Part tucoz
14:54:24 Nick QT_ is now known as QT (i=as@madwifi/users/area51)
14:58:28*Zagor eyes the #887081 patch
14:58:30am1connImho changing focus and starting/stopping threads for plugins is going to be a nightmare
14:58:40Zagoram1conn: i agree 100%
14:58:55am1connThe grayscale library is interrupt based. How would we handle that?
15:05:44 Quit am1conn ("CGI:IRC (EOF)")
15:09:44 Join am1conn [0] (
15:12:12markun_nrg_: general settings -> playback -> move to next folder
15:16:05 Quit am1conn ("CGI:IRC (EOF)")
15:16:12_nrg_markun, its working thx!
15:17:01rasherlinuxstb__: I have a patch now.. polishing now
15:20:27rasherlinuxstb__: I'm not doing any additional legality check though - only the one when moving the field.
15:24:33 Join Sucka [0] (n=NNSCRIPT@
15:34:28 Join linuxstb_ [0] (
15:36:46 Join linuxstb [0] (
15:39:04linuxstbrasher: I think the legality check should be done when you leave "edit mode", not when you leave the cell. But I can have a look at the patch (no time at the moment though)
15:42:16rasherI think we're out of buttons for the ondio
15:42:24rasherWondering how to handle that
15:42:42rasherright now I just set the BUTTON_ value to something not obtainable on ondio, but that's pretty ugly
15:42:57Zagorondio is a problem child :-)
15:49:25linuxstbrasher: Have a look at the minesweeper plugin - I think that defines combinations of keys for the Ondio. Maybe you will find an idea there.
15:50:17 Quit linuxstb__ (Read error: 110 (Connection timed out))
15:52:25 Quit linuxstb_ (Remote closed the connection)
15:52:48rasherwell, I'm thinking we might as well leave the option out, if it's going to require weird key-combos
15:54:52rasherthe "known bugs" part of the release notes weren't updated..
15:55:25 Join tucoz [0] (
15:55:54tucozinteresting to read the old release notes. Note however that the link to v2.4 is missing there
15:57:32 Join leftright [0] (
15:57:55leftrightHi, could i interest anyone in patch 1297992
15:58:20leftrightit adds quick access to .cfg files to the quick menu
15:59:07tucozleftright: is that really needed? What quickmenu btw. WPS?
15:59:34tucozshould learn to read. wps it is.
16:00:37leftrightin my opinion yes, lots of folks have several configs, and navigating to them is tediuos, this way there would be quixk acces to the different configs
16:01:13rasherit's only two buttons to get there through the main menu
16:01:22leftrightno its more
16:01:37leftrightits at least four clicks
16:01:38tucozOk, Imho, wps-context menu should only be populated by playback/wps entries
16:01:51rasherokay, four clicks.
16:02:02rasherStill don't warrant putting it where it doesn't belong imho
16:02:53leftrightdo you use different configs frequently ?
16:03:13 Quit tucoz ("CGI:IRC")
16:03:14leftrightthen please dont shoot it dwon because you dont
16:03:21 Join tucoz [0] (
16:03:22rasherI'm not
16:03:33rasherI'm shooting it down because the menu entry doesn't belong there
16:03:42rasherand the real location is not *that* far away
16:04:01leftrightgues i'll learn how to do my own patches
16:04:31rasherThese are not quickmenus by the way.
16:04:36rasherThey are *context* menus
16:05:17tucozOn the other hand, I think the a-b-menu could be arranged in a different way. Not that I have a clue of what that would be, but I have a feeling it is improvable
16:05:23linuxstbDon't shoot me for saying this, but IMO, a nice feature of the iPod's UI is the ability to choose the items that appear in the root of the main menu.
16:05:43leftrightlots of guys have car cfg's, shuffle-track, cfg's, etc, it would be nice ti have a quick path to this menu
16:06:10tucozah, that is true, But you could have the configs in a directory and just play them
16:06:11rasherI don't see how "menu, down, right right" is too far away
16:06:22tucozOr am I wrong?
16:06:25rasherare you changing configs every other song?
16:06:53tucozjust press the joystick and browse to the config directory
16:07:01leftrightyes i change cfg's frequently
16:07:04 Join dpassen1 [0] (n=dpassen1@resnet-233-61.resnet.UMBC.EDU)
16:07:27tucozleftright: I get your point with the car/etc
16:07:36rasherI don't.
16:07:46leftrightand there is NO quick path to the cfg menu
16:08:00tucozrasher, why not?
16:08:07rasherIf you have a car config, you change to it when you get in the car, is saving 2-3 clicks that important for that?
16:08:16leftrightjeez rasher, just create a few different cfg's and sse what a pain it is to access them
16:08:38tucozhehe, not that it should be in the wps-context, but the fact that some people change configs often
16:08:48 Join einhirn [0] (
16:09:18rasherleftright: I don't have any reason for having different cfgs
16:09:47 Quit leftright ("CGI:IRC (EOF)")
16:09:51 Join leftright [0] (
16:10:08rasherI guess consistency means more to me than saving 2 clicks once in a while at the cost of cluttering menus with random entries
16:10:08leftrightrasher, I have two cfg's which I use frequently
16:10:35tucozleftright: why don't you create a root/configs/ directory, and have the configs there?
16:10:42rasherI play a lot of solitaire, can I interest you in a patch that adds solitaire to the wps context menu?
16:10:46linuxstbMaybe the solution would be to think about how the main menu is structured - to add faster paths to frequently used functions.
16:10:51tucozand just play them?
16:10:57leftrightforget it rasher
16:11:02linuxstbtucoz: What if the browser is already very deep in the file tree?
16:11:14leftrightdont worry about it really
16:11:17Moosrasher: me too ;)
16:11:17rasherI don't see how y ou could get faster access from the main menu than is now
16:11:31rasherThey can't *all* be the top choice
16:11:49rasherand "browse .cfg files" is already the top choice in the 2nd level menu
16:12:28tucozlinuxstb: so I noticed. The current placement is not too hard too find.
16:13:00rasherI agree that some of the 3rd or 4th level entries might benefit from being moved, but browse cfg files is not one of them
16:13:01 Join fogcat [0] (
16:13:27***Saving seen data "./dancer.seen"
16:13:39tucozI thought it was buried deep in some a-b-menu, but it wasn't.
16:13:48linuxstbrasher: Yes, I was talking more generally, not just the "browse cfgs" option.
16:13:58rasherlinuxstb: I agree with that
16:14:27rasherZagor: around?
16:14:51tucozlinuxstb: do you mean like a dynamic menu?
16:15:07rasherI know, automatic restructuring, Microsoft style!
16:15:25tucozor a simple poll on what features are accessed most frequently, and rearrange the menus accordingly?
16:15:33leftrightbesides the feature which I'm really after, you guys certainly arent interested in anyway, which is to enable Track mode if shuffle is selected
16:15:34linuxstbrasher: I wouldn't go that far :)
16:15:47leftrightas an option ofcourse
16:16:08rashertucoz: or better yet, create builds that record menu usage in a file
16:16:21leftrightshuffle and track mode are a match made in heaven
16:16:45rasherSo we've heard.
16:16:50rasherAbout a million times by now.
16:16:50leftrightbut I'll use a cfg file as rasher insistes
16:16:53fogcata lot of the "solutions" given to requests are "create a config to change multiple settings at once" - I only have 3 but I do find it irritating navigating to change them
16:16:56fogcatIt's menu,down,down,down,click,click - 6 actions
16:17:12tucozyes, that could be a good way of gathering data.
16:17:57leftrightfogcat, glad you find it irritating as well
16:17:58rasheror, menu, dooooown, click-click, in my eyes it's more like 3½ action
16:18:26rasherBut really, adding menu entries at random isn't going to solve this
16:18:35 Quit [IDC]Dragon ("CGI:IRC")
16:18:41fogcatif you go dowwwwn you're likely to whizz past it and then click back to get to the right place
16:19:09rasherPeople will be complaining about the next thing that they're missing, and all menus will include all options eventually
16:19:25linuxstbrasher: That's why the user should be able to config their menus (to some extent)
16:19:48rasherSure, if they want
16:20:06leftrightthats exactly the standard reply, create a config, great, so give us easy acces to the config menu at least
16:20:22tucozAs the menus are now, they seem to be sorted quite arbitrarily(?).
16:20:37rasherleftright: you have as easy access as you're likely to get without user-configuarable menus
16:20:58*leftright sighs
16:21:12rashertucoz: Yeah, some seem a bit weird
16:21:14linuxstbleftright: I'm suggesting that a user could (for example) add "browse config files" to the top-level of the main menu. This would be a one-time action needed by the user.
16:21:45leftrightanything to improve and make it more user friendly
16:21:52rasherNot anything.
16:21:58rasherthe right things.
16:22:00tucozand a way better solution than adding it to wps-context
16:22:23leftrightbut that I mean I will be gratefull for any improvement made
16:22:34leftrightas will others to be sure
16:22:46rasherDepends what it is.
16:23:06rasherI wouldn't be happy about browse cfg being added to the wps-context because it doesn't belong there.
16:23:35leftrightrasher as i've said, any improvement is welcome
16:23:36tucozthe context menus should give easy access to menu items in their specific _context_
16:23:51tucozand only that, imho
16:24:51fogcatwhy is show files in the quick menu? or am I thinking of something different?
16:25:10rasherQuick menu != context menu
16:25:21fogcatI'm talking about thelong press of A-B (on iriver)
16:25:29rasherThat's the quickmenu.
16:25:36fogcatahhh - sorry terminology
16:25:43fogcatso could it go in quick menu?
16:25:45tucozI came to think of something. When I enter a empty directory, the screen is blank. I think an indication that the directory is empty would be nice.
16:25:58leftrightthe 12 o' clock position is unpopulated in that quick menu
16:26:10rasherI wish it wasn't.
16:26:18fogcatthat's what I though we wer talking about :-o
16:26:28rasherThere's no way in hell anyone will be able to agree what that should do
16:26:31linuxstbWhy is "open with" in the context menu (long joystick press in WPS)
16:27:03rasherI was wondering about that as well
16:27:06fogcatwell if you make 12 o'clock cahnge configs ... it can do anything ;-)
16:27:18tucozrasher: was that an answer to the empty dir thing?
16:27:36linuxstbI've now looked at the menus again, and am agreeing that "browse configs" should go in the context menu - "sound settings" appears there, so why not the preset sound settings (i.e. config files) ?
16:27:57rashertucoz: no, it was to linuxstb
16:28:03tucozah, ok
16:29:33 Quit ashridah (Read error: 110 (Connection timed out))
16:29:47tucozlinuxstb: well, it is a way better option than open with..
16:30:16linuxstbYes, I think "open with" being there is a bug. Unless someone can think why it was put there.
16:30:59leftrightwhy have show all files there as well ?
16:31:11leftrightits not wps related
16:31:17tucozdon't you sometimes feel the urge to open the mp3 your playing with the chip8 viewer?
16:32:02tucozmaybe the current menu is there as a proof-of-concept
16:32:10tucoznot really thought through
16:32:18fogcatOK - you guys are talking about the context menu now? Right?
16:32:22linuxstbI would be surprised if "open with" worked with the currently playing track - I'm guessing it uses the file currently selected in the file browser.
16:32:56leftrightthe quick menu is a great idea, the menus though need to be rethought
16:33:10fogcatYou have less objection to change config file being in the QUICK menu?
16:33:45tucozopen with should be remove at least.
16:33:46rashertucoz: well, vbrfix and favourites make sense to "open with"
16:33:55rasherand splitedit
16:34:40tucozcome to think of it, I think I asked this when the context menu was enabled on iriver. And I think I got the same answer then
16:34:49tucozwhich I settled with
16:35:04tucozas a good reason why to have it there :D
16:36:49fogcatleftright:: does your patch currently put it in the context menu?
16:37:30linuxstbLooking at the code (onplay.c), I think the "open with" applies to the currently selected file in the browser - can someone test? If so, then I think it's a bug that it's added to that menu.
16:37:31rasherI still don't think that loading config files has any business being in the WPS context menu
16:37:53leftrightyes it puts it there, but its not my patch, some kind coder wrote it and added it to the tracker list
16:38:15leftrightrasher, and show all files belongs there ?
16:38:24linuxstbrasher: If you think of the config files as a set of presets for the sound settings (which do appear in that menu), then I think it's logical.
16:38:24tucozlinuxstb: I just opened the file with viewer and got a I04
16:38:33fogcatrasher is talking about the context menu (i think)
16:38:53fogcatteh one you get to with a long A-B press (on iriver) is QUICK menu
16:38:59rasherI am. As witnessed by me saying "WPS context menu"
16:39:00tucozlinuxstb: how can I verify what I test?
16:39:12leftrightah yes, my apologies
16:39:32linuxstbtucoz: Start some music, then browse to a text file in the file browser, then go back to WPS, then select "open with"
16:39:42leftrightI use config files to change sound settings and the way music is played back
16:39:56rasherlinuxstb: they are not really presets for sound settings though
16:40:04tucozNo, it opens the while playing track
16:40:43tucozI got the id3 header, when I opened with the viewer. And the browser was pointing at a textfile
16:42:49leftrightfogcat, sorry, that patch adds access to the config files to the Quik menu
16:44:06 Quit tucoz ("CGI:IRC 0.5.4 (2004/01/29)")
16:44:14fogcatleftright, so *I* would say that in that case a number of the objections made vanish?
16:44:31fogcatWhile I'm here.... ocasionally I get distorted playback
16:44:31fogcat(not easily repeatable - well at least not yet)
16:44:31fogcatIt seems to happen when I cahnage to my "car config"
16:44:35fogcatWhile I'm here.... ocasionally I get distorted playback
16:44:35fogcat(not easily repeatable - well at least not yet)
16:44:35fogcatIt seems to happen when I cahnage to my "car config"
16:44:52fogcat"car config" has voice menus on but sometimes after switching they don't play
16:44:52fogcatand if that happens MP3 playback is distorted with bursts of "Static"
16:44:52fogcatTurning off and back on clears the problem
16:44:52DBUGEnqueued KICK fogcat
16:44:52fogcatSomething not initialising correctly? Or over writing something else?
16:46:20leftrightI find that sometimes I have to switch it off and on for a config file to be stored and loaded correctly
16:48:32fogcatoops - gotta go... before i do - ta to all you devs for the work and 2.5 release and hack recovery - it IS appreciated
16:48:35 Quit fogcat ("CGI:IRC")
16:50:40amiconnlinuxstb: "Open with.." is in the wps context for a reason
16:50:53amiconnYou can open the currently playing file with a plugin
16:51:12amiconnMolto importante for the splitedit plugin
16:52:31linuxstbamiconn: OK. I was waiting for someone to say why.
16:52:49leftrightamiconn, what are your thoughts to making the config files more accessable ?
16:53:10amiconnI don't need that at all, although I do use .cfg files for changing settings
16:53:26amiconnI have my .cfg files in the root - quickly accessible
16:54:23leftrightI think it would be more accessable in the Quick menu, nut thats just me
16:54:33amiconnThere's also a reason why the quick menu(s) have no 12 o'clock item, and why they weren't changed on iriver to have one
16:54:45rasherThere are loads of things people want in the Quick Menu.
16:55:14amiconnThe quick menu(s) are of archos recorder origin, and there are 2 of them, F2 and F3 quick menu
16:55:44amiconnLeft, Down and Right can be detected together with F2 or F3, but not Up
16:55:55amiconnThat's the explanation for the missing 12 o'clock
16:56:01leftrightbut the iriver could detect up ?
16:56:49amiconnThe iriver can't detect A-B and any other button at once except play, but by chance the quick menu still works as expected because of the button priorities in the hardware
16:57:41amiconnThe quick menu was added for iriver by Christi because she missed it, but it's an interim thing
16:57:53leftrightoh please dont remove it
16:58:16amiconnThere's consensus for a long time that the quick menus are ugly and should be replaced by something better
16:58:59amiconnOn the archos recorder the plan was to use F3 for context menu (which is now implemented as long-Play, equal long-Select on iriver)
16:59:15amiconnF2 was planned to be used for a user-configurable menu
17:00:22rasherSlasheri: are you around?
17:00:57rasherSlasheri: can you confirm that patch #1298779 is right?
17:13:43 Quit DangerousDan (Read error: 104 (Connection reset by peer))
17:28:08 Join DangerousDan [0] (
17:30:23 Part leftright
17:36:09amiconnrasher: I've restored a number of wiki image attachments. However, I found a topic is missing completely - WhiteBacklight or whatever it was called
17:37:55rasheramiconn: that image?
17:38:01rasherBelongs on CaseModding
17:38:35amiconnAh, nm
17:40:00amiconnI can't find the recorder uart boot images. They were made by [IDC]Dragon...
17:43:32 Join koniu_ [0] (
18:04:31 Join webguest85 [0] (
18:08:53 Join paugh [0] (n=kickback@2001:5c0:8fff:ffff:8000:0:3e03:6822)
18:11:28 Quit Moos ("Glory to Rockbox")
18:13:29***Saving seen data "./dancer.seen"
18:13:57 Quit paugh ("Leaving")
18:14:37tomyleei was just wondering...can the iriver accept 24 bit wave signals?
18:14:42tomyleei mean, record them?
18:14:50tomyleevia optical input?
18:15:02tomyleeI´d like to use it as a ols school 2 Track recorder
18:15:07tomyleeand not my sequencer
18:18:07amiconnrasher: The remote_button_hold() fix introduces a bug...
18:19:35 Quit webguest85 ("CGI:IRC (Ping timeout)")
18:21:15rasheramiconn: Oh dear, what?
18:21:18 Join Sucka` [0] (n=NNSCRIPT@
18:21:39amiconnrasher: See lines 418 and 457
18:21:58amiconnremote_button_hold() is used to check whether the remote button hardware should be read
18:22:16amiconn...which of course shouldn't happen when the remote is unplugged
18:22:45amiconn...because the pins might have arbitrary values
18:22:50rasherUsing remote_button_hold() for that sounds weird
18:23:08rashera remote_plugged() would be sensible?
18:23:09amiconnSo remote_button_hold() should return true if the remote is unplugged
18:24:04rasherokay, whatever - it seemed sensible.. I don't have time to revert now
18:24:31amiconn...or button_read() changed to not use it, but that would partially jeopardize remote_button_hold(). The UI is another thing...
18:32:45 Quit Sucka (Read error: 110 (Connection timed out))
18:36:52 Join Lear [0] (
18:39:27linuxstbLear: Have you got any metadata.c patches ready?
18:39:50LearYes, I think so. I've been using it some time now... :)
18:40:33linuxstbI've added ".m4a" support (for ALAC files, not AAC), which I'm ready to commit as well. This adds a fair amout of code to metadata.c
18:41:01linuxstbI don't know if we are going to conflict badly with each other.
18:42:14LearI know it will. :)
18:42:26linuxstbI had better hurry up and commit first then :)
18:42:52linuxstbWhat does your patch do?
18:42:53LearI've restructured it quite a bit, but if you've only added a new case, and a function or two, then it shouldn't be too bad.
18:43:09linuxstbYes - that's all it is.
18:43:35LearMainly code and size cleanup. Removes a 4 kB (I think it was) static buffer and cuts binary size by 800+ bytes.
18:43:43preglowLear: you got anything queued for soonish comiting?
18:43:43LearTag parsing is a bit generalized too.
18:45:08LearWell, the metadata thing first, then I'll look at the wave patch, or possibly that "track gain during shuffle" thing that seems to be in high demand. Darned if I can figure out a good "Replaygain type" name for it though. :)
18:45:32linuxstbThe WAV patch would be nice - more boxes to tick on the SoundCodecs page.
18:46:01linuxstbHow does "track gain during shuffle" work if you're shuffling an album?
18:46:12LearYep, should be quite easy to apply, I think, even after my changes... :)
18:46:45LearShuffle would "force" track gain, plain and simple. No other way to do it really...
18:47:06linuxstbIsn't that what the replaygain menu is for? To choose between the modes?
18:47:14linuxstbSorry - I've missed this debate.
18:47:50LearYes, but this would be another mode, "Use track gain if shuffle is on, otherwise use album gain".
18:48:55linuxstbHowabout: Replaygain type = (track,album,auto) where auto is the new option.
18:49:44LearAuto doesn't say much. Or, it implies using gain even without tags...
18:50:57linuxstbOK - I'll go find the discussion and read the reasoning.
18:51:56 Join paugh [0] (n=kickback@2001:5c0:8fff:ffff:8000:0:3e03:6822)
18:52:49 Quit rasher ("Ex-Chat")
18:58:16LearThere, metadata.c committed.
18:58:26linuxstbLear: I don't think this has anything to do with shuffling. IMO, you want "album" mode when your playlist consists of tracks from the same album, and you want "track" mode when your playlist consists of tracks from different albums.
19:00:22LearNot necessarily. A playlist can contain a number of albums in order too. And besides, without a tag database, how would Rockbox know if a playlist only contains tracks from one album (and even with a tag database that might not always work)
19:01:05markunlinuxstb, preglow: If you are thinking of optimizing Tremor, here is a patch for a TI dsp implementing imdct with ifft:
19:01:11LearThe proposed thing is simple to do, and is logical - though not the only logical way to do it.
19:01:28LearLast I tried there was no response from You mean it is up now?
19:02:11markunYes, it's up again.
19:02:44markunI was very surprised the link worked this time, don't even know why I bothered to try.
19:03:13preglowmarkun: thanks
19:03:17linuxstbLear: I'll leave the replaygain issue to others.... I now have some work to do to use your new metadata.c
19:04:37LearStill, I think we need a fast FFT before that patch is really useful...
19:04:39linuxstbYour diff is completely meaningless :(
19:04:53Lear"My" diff?
19:05:16linuxstbThe changes you committed:
19:05:46linuxstb" 889 lines, -800 lines" :)
19:06:42LearAh, well, I did touch pretty much everything... :)
19:06:44markunLear: freescale has a asm optimized fft, but I don't know what license they use.
19:07:08LearMe neither, and I've looked a bit for it too. But it is for 1024 points only, I think...
19:07:18markunpreglow: the patch is from the low-mem branch I think.
19:08:46LearThat too...
19:09:35 Quit paugh ("reboot")
19:10:00 Quit linuxstb ("CGI:IRC")
19:13:45 Quit einhirn ("Miranda IM! Smaller, Faster, Easier.")
19:17:20LearHm... not even the wav.c part of the patch applied cleanly...
19:27:34LearHm... Should a codec leave most parsing to metadata.c?
19:31:13 Join gerd|koks [0] (
19:43:34preglowmarkun: the freescale fft is just one size
19:43:42preglowLear: i think so
19:44:13preglowwe're probably just as well off stealing djbfft or something
19:44:58LearWell, the wav patch should probably move some stuff from the codec then, but I'll leave that for after the initial commit, I think...
19:58:16preglowno biggie
19:59:42pregloweventually we'll split metadata loading out to codecs as well anyway :>
20:00:01 Join muesli- [0] (
20:00:06 Join Moos [0] (
20:02:50Learbut not to the "main decoding loop" function...
20:03:03preglowthat we wont
20:03:35Learthat's a new one: int16_samples causes a section type conflict
20:04:25amiconnDid you try to move some const to iram?
20:05:01amiconnThat causes such a conflict; see my comment in firmware/drivers/lcd-h100.c
20:05:24LearAh, yes, saw some static tables that I thought could be const... Thanks.
20:06:09amiconnTwo options for now: (1) Don't use iram for these (2) Don't make them const
20:06:23amiconnThe proper solution would be an additional section .iconst
20:06:58amiconnIn conjunction with that, I renew my proposal for .ibss
20:07:24amiconnToday, all .idata content is part of the binary, even uninitialised data
20:07:34amiconn(unlike standard .bss, which is not)
20:08:12amiconn.ibss will only save a bit of binary size on archos, but way more on iriver
20:08:16amiconn(40-ish KB)
20:09:41LearLike all stacks?
20:09:51amiconnThe startup code would have to clear that area instead of copying it from dram
20:10:05amiconnNot all stacks, but main/codec/voice codec
20:10:27amiconn(for iriver) All others are in dram.
20:13:33***Saving seen data "./dancer.seen"
20:19:00preglowmarkun: got confirmation that code is working?
20:20:09 Quit gerd|koks (Read error: 110 (Connection timed out))
20:20:22preglowmarkun: what downfall is that it requires a set of friggin huge tables
20:26:38LearOuch... Flac doesn't work well in the simulator... :)
20:26:46 Join linuxstb [0] (
20:27:21linuxstb19.27.34 # <Lear> Hm... Should a codec leave most parsing to metadata.c?
20:28:33Learlinuxstb: What about it?
20:29:16linuxstbI'm not sure :) I thought I knew what the split should be, but I've realised that I don't.
20:29:35linuxstbI think at the moment the codecs duplicate a lot of the parsing already done in metadata.c
20:31:11LearIndirectly yes, as a part of the decoding libraries...
20:31:11 Quit solexx (Read error: 104 (Connection reset by peer))
20:31:12 Join solexx [0] (
20:44:28LearHe, the resampler/dsp.c couldn't handle 96 kHz files. ;)
20:47:13amiconnHmm, we're doing something concerning sections that isn't supposed to work according to the gcc docs, but it works...
20:47:40amiconnAccording to gcc docs, it isn't possible to put uninitialised data in specific sections...
20:48:49amiconnHmm, I think that's because we use linker scripts
20:49:06preglowLear: it can't, the delta calc overflows
20:49:26preglowLear: we need to make that a 64 bit mul if we want anything above 65khz
20:49:34preglowLear: that, or decrease the accuracy
20:49:51Learpreglow: buffer size calculations fails too; but it was just a test file I created without thinking too much about it.
20:50:20LearHm... Maybe the wav metadata code should check that... And possibly others too?
20:50:28preglowcheck what?
20:50:41LearThat the frequency is within acceptable limits.
20:50:44preglowwhy not fix dsp code instead?
20:50:56preglowlet the dsp layer decide if it's too much, not the codecs
20:51:21preglowand besides, we _should_ support 96khz files
20:51:46LearWell, it would be nice if unplayable contents was detected early on...
20:52:23preglowwhy? so they can be silently ignored or something?
20:52:26LearAnd the dsp code do detect it's too much and handles...
20:52:47amiconnWhy would we need 64 bit ari for 96 kHz support?
20:52:51LearYes, something like that, no point in opening and decoding part of a file that is known to not work.
20:53:01amiconn96000 fits pefectly into 32 bits
20:53:34preglowdelta = dsp->frequency * 65536 / NATIVE_FREQUENCY;
20:53:46preglowupper limit before overflow is 65536
20:54:28amiconnHmm, that doesn't look too well. Roundoff errors...
20:54:34LearWell, that is easily fixed...
20:54:47Lear(using 64-bit muls that is)
20:54:47preglowwhat doesn't look well? that particular expression?
20:55:12amiconnIf we'd use a bresenham-like algorithm, we would avoid the overflow and roundoff errors at the same time...
20:55:25preglowsure, feel free to implement
20:55:36preglowi don't remember entirely how to do it
20:55:56LearWould it change the actual resampling code too?
20:55:57preglowbut bresenham requires a branch per delta, yes?
20:56:03preglowLear: no, just the delta calc
20:56:20preglowLear: it would just replace the pos += delta part
20:58:28amiconnpreglow: A branch should still be faster than a division
20:58:40LearHm.. Wav patch looks good, with the limited set of test files I have...
20:59:00preglowamiconn: the delta calc will be faster with bresenham, it's the delta add that'll be slower, and by quite a bit
20:59:01amiconnAnyway, I wonder whether it would be worth it.
20:59:21amiconnDidn't you intend to implement some better resampling?
20:59:26preglowamiconn: the delta add will be four-five times more expensive than the simple add i do now
20:59:34preglowamiconn: yes, yes i did, but that'll still require this code
20:59:41preglowamiconn: that is, the delta handling
20:59:41linuxstbMaybe it's worth mentioning that the ipod supports 96KHz playback, and the new Neuros player probably will as well. So hopefully it won't remain a theoretical problem for long.
21:00:09CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
21:00:09*amiconn wonders what 96kHz support should be good for
21:00:21preglowplaying 96khz files, for one
21:00:26amiconnAnyway, if it exists and we can, we should support it
21:00:38preglowit shouldn't be a problem for us either
21:00:40amiconnI'd never use it though
21:00:43preglowme neither
21:01:10preglowbut like i said, it shouldn't be something we'd need to work on, it should just work
21:01:14preglowprovided the codecs are fast enough
21:01:35preglowfor 96khz you'd need a much better resampler than we have now, though
21:01:37preglowthat's for sure
21:01:40preglowso rockbox will be hard pressed
21:05:36amiconnHmm, how is that delta supposed to work? 16.16 fixed point I'd guess?
21:06:12preglowand you're right, exactly
21:06:16preglowit has too low precision now
21:06:29preglowbresenham would be very nice for precision, if it can be done fast enough
21:06:33preglowand i think it should
21:06:55 Join bbad [0] (n=bbad@
21:07:00preglowamiconn: i think you yourself noted that it sometimes makes a clicking noise when resampling, that's because of 16 frac bits not being enough
21:09:39 Join tucoz [0] (
21:10:05tucozamiconn: still in the bit shift department?
21:13:19amiconnpreglow: The disadvantage with bresenham is that the phase value wouldn't vary from 0..65536 like now, but from 0..NATIVE_FREQUENCY (upsampling) or 0..frequency (downsampling)
21:13:56amiconnHowever, that shouldn't be much of a problem because we need to normalise to S.31 fixed point format anyway
21:14:14amiconnThe bitshift would be replaced by a multiplication - not that much slower
21:15:44tucozI mean with the recording issues on archos
21:16:13amiconntucoz: Yes I know. I plan to implement Luke Framewalker (as [IDC]Dragon called it)
21:16:17LearOn the 5249, doesn't the emac do a 32-it mac in 1 cycle?
21:16:43amiconnI'm a bit undecided what to do first though
21:16:58tucozhehe, Luke Framewalker sounds cool enough. I guess he will use the source for all it is worth.
21:17:00preglow1 cycle throughput
21:17:01preglownot latency
21:17:07amiconntucoz: hehe
21:17:09preglowthey're 1 cycle a piece if you string a lot of them together
21:17:46LearOkay, the docs I read weren't that clear anyway...
21:18:16preglowLear: no shit
21:18:28LearHe, lots of commits today, and more to come... :)
21:19:49tucozlinuxstb: if you are here, the soduku plugin looks really nice on iriver. Do you have any plans to move the toggle-number from joy-press to some other button? It is nice as it is now, except for the loud click from the joystick.
21:20:27amiconnImho the joystick press is the only convenient button on the iriver
21:20:36amiconnThe side buttons are cumbersome
21:20:43tucozYes, too bad it emits the click from hell
21:21:55amiconnThe click doesn't bother me
21:22:45preglowbut i gotta go for a wghile
21:22:58tucoznot me either, but I am mostly thinking of people around me. i.e. if I am playing soduku in some quite place (school for instance)
21:23:12tucozor on a boring lecture
21:29:17Leartucoz: use the source: "press SELECT/ON/F2 to increment the number under the cursor." :) That ought to include the on/play button on the iriver...
21:30:08tucozahh, haven't really looked at the source. Good thing
21:30:37tucozThe source is really powerfull
21:32:38tucozhave to admit I didn't expect such a good documentation there. I do not usually write such good comments myself
21:33:00 Quit t0mas (Read error: 110 (Connection timed out))
21:35:13tucozLear: "#define SUDOKU_BUTTON_TOGGLE BUTTON_SELECT" no || BUTTON_ON there. :(
21:35:27tucozor whatever it is called
21:35:47LearAh, well, I only read a comment in the beginning of the file...
21:36:35LearWav patch in...
21:37:50LearHm.. When setting a patch to resolution: accepted, should it be closed too?
21:38:10amiconn#define SUDOKY_BUTTON_TOGGLE2 BUTTON_ON for iriver, then add a second case: for that (#ifdefed properly)
21:38:20amiconn...and you can use both buttons alternatively
21:41:49*amiconn didn't try sudoku yet (on rockbox)
21:41:50tucozok, thanks. So in a #define one could not simply #define THIS_THING BUTTON_ON || BUTTON_OFF ?
21:41:57amiconnlinuxstb: Does it work on Ondio?
21:43:18tucozamiconn: the keys are defined for Ondio, so I guess it should work
21:43:45linuxstbamiconn: Yes, it should work on Ondio, but I haven't tested it.
21:44:02amiconnOk, I'll try
21:44:28linuxstbI also agree with both tucoz and amiconn - the joystick click is annoying, but it's also the most convenient button.
21:45:04amiconnJust allow both buttons, as long as Play isn't needed for other things...
21:45:40linuxstbI personally play it by holding the joystick down and letting it auto-repeat. It's slow, but then again, Sudoku is a slow game.
21:46:30 Quit tomylee (Read error: 110 (Connection timed out))
21:47:02tucozyes it is very convenient, but if it is possible I think amiconns suggestion is good. When used in a silent space with people around you for instance.
21:47:29linuxstbOK - I'll implement amiconn's suggestion now.
21:47:54tucozFeels like eating crisps at the cinema
21:49:09tucozlinuxstb: good, I also like that it is perhaps the first game on the iriver that looks really good.
21:49:43LearHe, sudoku increased the number of h3x0 compiler errors... :p
21:50:43muesli-how far is a h3xx away from being released?
21:50:54LearNo bootloader yet...
21:51:18muesli-yeah i know..
21:51:27linuxstbLear: Yes, it's because Sudoku has fonts optimised for each LCD size - and I haven't implemented the H300's LCD yet.
21:51:56linuxstbI'm not sure whether to do something quick now to kill the warnings, or just leave it and fix it properly when the H300 is closer.
21:55:40amiconnlinuxstb: Any .ss files to try?
21:56:16linuxstb - download "puzzle pack 1" and "puzzle pack 2"
21:56:16 Quit bbad ("KVIrc 3.2.0 'Realia'")
22:02:44amiconnlinuxstb: Works on Ondio, and looks really nice :)
22:03:15amiconnOnly disadvantage is that it's impossible to distinguish numbers you set yourself from fxed ones
22:05:09 Join XavierGr [0] (
22:05:18XavierGrHi again!
22:05:39XavierGrJust a quick question about the FM radio....
22:06:03XavierGrWhy only 32 presets? Is there any hardware limitation for it?
22:06:03 Join linuxstb_ [0] (
22:06:26XavierGrWhy no use a special file format that will contain all the Frequencies with the Titles?
22:06:58XavierGrand then be able to select which one you want, according to your location?
22:07:14ZagorXavierGr: we must have a limit somewhere. our memory is not infinite.
22:07:33Zagoror you mean "don't load all in memory, search file"?
22:07:55amiconnThat would be a bad idea...
22:08:45XavierGrOk have a limit but I don't think that a file that contains 50-60 stations with names in a saperate file is a problem.
22:08:50linuxstb_Having a choice of presets files (e.g. for different locations) would be useful though. I can't see anyone needing more than 32 presets in the same location.
22:09:22XavierGrLets say that I have a file that contains all the stations for my area. I will just load that file and then select the station I want.
22:09:59XavierGrBut sometimes (especially travelers) can have multiple files that will contain stations for various areas.
22:10:39XavierGrWhat do you think? Though I am not of a Radio guy it will be pretty interesting for me to program it.
22:12:59ZagorI don't think expanding the list is valuable. loading different lists is much more useful.
22:13:37***Saving seen data "./dancer.seen"
22:14:33linuxstb_It gets my vote. Should be easy to just add load/save options to the file menu.
22:14:57muesli-hey XavierGr
22:15:34XavierGrhi muesli.
22:15:49XavierGrGreat then I will see what I can do...
22:16:11amiconnlinuxstb: Is there a special reason why the inverse numbers are defined separately?
22:16:50amiconn(just curious)
22:20:15 Quit linuxstb_ (Read error: 104 (Connection reset by peer))
22:21:06 Join linuxstb_ [0] (
22:21:12amiconnlinuxstb: Two other remarks: (1) For recorder, F1 is menu by convention, so I'd expect the same in a plugin with a menu (2) The menu combo on Ondio is really tricky to get, as it interferes with normal operation
22:21:41amiconn(and involved buttons on different sides of the Ondio)
22:22:27 Quit linuxstb (Read error: 113 (No route to host))
22:22:51linuxstb_Please go ahead and change them - you know the Archos devices much better than I do.
22:22:56amiconnOn Ondio, long MODE is usually menu, but that forbids button repeat for the number increment
22:24:36XavierGroh amiconn I just finished (solved all previous bugs) with the filescrolling jpeg viewer. Are you interested?
22:24:44linuxstb_Also, there is no particular reason why the inverse numbers needed to be defined - it just seemed simpler that way, and more flexible for different displays.
22:25:18amiconnAh okay. Thinking coloured here? ;)
22:25:23linuxstb_e.g. the iRiver cursor is not a true inverse - it is always white on black, even for highlighting "fixed" numbers.
22:25:31linuxstb_But yes - mainly the colour LCDs are in my mind.
22:25:50linuxstb_"inverse" is probably the wrong word - "selected" would be better.
22:26:56linuxstb_BTW, I've committed the "alternative toggle" idea for the iriver. So the ON button now also works.
22:27:29XavierGralternate toggle?
22:27:48linuxstb_Check the IRC logs from about 1 hour ago.
22:28:59 Quit muesli- (Read error: 104 (Connection reset by peer))
22:30:22LearAh, as I hoped, pluggning the USB did not stop the radio. :)
22:33:42*amiconn has an idea for the Ondio toggle<->menu clash
22:34:06amiconnShort Mode alone will toggle once (via _pre/release mechanism)
22:34:34amiconnLong Mode will bring up the menu, but only if no other button is pressed before the repeat kicks in
22:34:42amiconn(_pre check as well)
22:34:59amiconnMode+Down (+Repeat) will be AltToggle
22:35:18amiconnMode+Down is easy to press with the thumb, the buttons are adjacent...
22:36:01tucozit almost sound as one of the emacs key combos
22:38:55linuxstb_amiconn: I trust you :) Not sure if there are any other Ondio users here.
22:39:15amiconnI have two of them :)
22:39:22amiconnOne FM and one SP
22:39:41linuxstb_amiconn and am1conn are the Ondio users then...
22:40:03 Part tucoz
22:44:25 Join tvelocity [0] (
22:47:34amiconnMeh, sudoku already uses case: fall-through... I trapped on that
22:48:21amiconnNow it works as intended...
22:50:00amiconnThis implementation leaves one unreachable case: value on Ondio, but I'll ignore that
22:50:22 Nick StrathAFK is now known as Strath (
22:50:25amiconn(SUDOKU_BUTTON_TOGGLE | BUTTON_REPEAT will never happen)
22:51:01linuxstb_That's obviously not a problem though.
22:52:26amiconnThat's because SUDOKU_BUTTON_TOGGLE contains BUTTON_REL, which will never repeat
23:02:24 Quit Lear ("Chatzilla [Firefox 1.4/undefined]")
23:06:58linuxstb_Anyone fancy creating a PluginSudoku Wiki page?
23:08:51XavierGrWhy I cant find this plugin in my iriver?
23:10:01linuxstb_It's a viewer.
23:10:44linuxstb_... for ".ss" files. Download the "advanced puzzle packs" from for some games
23:11:24Zagoroh, are the sudoku boards readymade?
23:11:54XavierGrAlso another matter that just came to my mind.
23:11:57Zagori thought it created the boards itself
23:12:33XavierGrWhy do we keep the rockbox.river file in the root? And if it is needed in the root why rockbox accepts its deletion.
23:12:47linuxstb_Zagor: No - they are readymade. Creating good Sudoku games is not easy - there are humans who make a living from devising the games.
23:12:54XavierGrImagine that you are away from home and suddenly you delete it by mistake.
23:13:08XavierGrBAM no rockbox until you get that file back
23:13:08linuxstb_But yes, a game generator is an obvious enhancement.
23:13:09amiconnYou won't
23:13:24amiconnThere's even an "Are you sure?" request
23:13:46 Join paugh [0] (n=kickback@2001:5c0:8fff:ffff:8000:0:3e03:6822)
23:14:09XavierGrI just deleted it and I cant boot no more! :P
23:14:10amiconnYou wouldn't be able to use rockbox if you'd delete the .rockbox folder, or some crucial components therein (e.g. the codecs)
23:14:32XavierGryes but the rockbox folder is not on the supported list.
23:15:11XavierGr.iriver file is a firmware file and it is shown.
23:15:16ZagorI kinda agree with xavier, if nothing else than aestetics. we only have it in the root for historical reasons.
23:15:18amiconnI wonder why this should be an issue on iriver.
23:15:47amiconnOn archos it's the same, ajbrec.ajz (or archos.mod) is in the root and is shown
23:15:56XavierGrI now it is rather stupid to delete it. I am just saying that maybe someone will make a bad weekend by this.
23:15:57Zagorthe boot loader could look in both places
23:15:59amiconnIf you delete it on a non-flashed box - no rockbox
23:16:10 Join rasher [0] (
23:16:32amiconnOn archos it has to be in the root, otherwise the archos loader wouldn't find it
23:16:35linuxstb_Have you ever actually heard of anyone actually deleting their rockbox.iriver file though?
23:16:43linuxstb_Or is this just hypothetical?
23:16:56amiconnHowever, on a flashed box I can delete it with no bad effects :P
23:17:08XavierGrI did once because I was weird. I am happy that I had my PC besides me.
23:17:33linuxstb_XavierGr: No problem then - you know never to do it again :)
23:18:46XavierGrTell me that when your kid will delete this little unknown file by accident.
23:18:57XavierGrAnyway not that it matters I just had a question.
23:22:00XavierGrok thats it for today. Good night all!
23:22:25 Quit XavierGr ()
23:30:35*rasher spots an ALAC codec
23:30:41rasherwhere does one find alac files?
23:31:13rasherAs in, where are they used?
23:31:25linuxstb_rasher: It's not fully committed yet (Lear caused me problems with his metadata.c patches)
23:31:33linuxstb_But you can find them in iTunes.
23:33:46rasheramiconn: should I reverse the remote_hold commit?
23:35:18elinenberasher: does ALAC play realtime?
23:35:53rasherelinenbe: It doesn't play at all - but it's fast enough to play realtime I think
23:36:03rasherAsk linuxstb_ instead
23:36:18elinenberasher: what's holding it back from actually playing?
23:36:18amiconnYes, imho it would be better to have a separate remote_connected() function, and then use that together with remote_button_hold() for the lock display status
23:36:21linuxstb_elinenbe: Yes, ALAC plays faster than realtime - I get about 60%-70% CPU boost.
23:36:50linuxstb_I'm in the process of committing it now.
23:37:42elinenbelinuxstb_: sounds good.
23:38:52elinenbenow we just need to play "fairplay" encrypted files (port hymn!)
23:39:28linuxstb_I hope you're not confusing AAC and ALAC.
23:39:50linuxstb_This is not an AAC decoder I'm committing.
23:42:46Zagorelinenbe: doesn't hymn permanently decrypt the files? then there's no point running it on rockbox.
23:43:21fuzziewell, using the hymn code you could build the ability to directly play the m4p files
23:43:31fuzziemeaning users wouldn't have to bother putting them through hymn manually
23:43:45Zagorfuzzie: yes, but we'd open ourselves up for a world of hurt
23:43:53fuzzieyes, i'm not suggesting it :-)
23:46:19Zagorisn't it funny how jhymn, a java application, is available in a windows and a mac version?
23:46:53fuzziei think it was fairly well established that java's crossplatformness is mostly a myth, a good while ago
23:48:44linuxstb_fuzzie: I know if I bought any encrypted music files, I would want to decrypt them permanently ASAP. And I would would to encourage anyone else to do the same.
23:48:53elinenbeyes, I was a bit confused... I was thinking you could buy alac files from apple...
23:49:01elinenbebut I guess they don't sell files in HQ.
23:49:26linuxstb_elinenbe: I wish they did - I would probably start shopping there then (assuming hymn worked with them).
23:52:57rasherCan anyone remotely familiar with the playback code check this one-line patch:
23:53:32rasherFixes broken automatic shutdown
23:56:58linuxstb_OK, ALAC is now fully committed.
23:57:44elinenbelinuxstb_: nice.
23:57:45rasherSo you can now play .m4a alac files?
23:57:54rasherOr rather, I can.
23:58:54 Quit tvelocity ("Leaving")

Previous day | Next day