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 2007-08-24

00:00:19H10_007quickand does this stop anything else from functioning?
00:00:50jhMikeSno task switches
00:00:51H10_007quickdoes it halt everything for 1000 micro seconds or just the particular function?
00:01:00Bjerrkpixelma: Where did you buy yours?
00:01:16BjerrkH10_007quick: wouldn't it just halt the thread you're in?
00:01:24pixelmaBjerrk: at the same shop linuxstb linked you to earlier
00:01:25jhMikeSwe're not preemptive
00:01:25kapootlinuxstb: where can i get builds older than those on
00:01:28H10_007quicki hope so
00:01:51Bjerrkpixelma: The shop at Ebay? Okay.. So they're reliable?
00:01:52linuxstb_kapoot: You'll need to get the old code from SVN and compile yourself.
00:02:16ddaltonis anyone interested in p6331?
00:02:21kapootwow... that sucks
00:02:23pixelmaBjerrk: yes but they only accepted PayPal from me
00:02:47BjerrkAh, crud
00:02:47pixelmaother than that everything was ok and quick, can't complain
00:03:03BjerrkIn don't have PayPal, and i really didn't plan to get it.
00:03:15Bjerrk(whine, whine ;) )
00:03:33pixelmait was the only shop I found that sells them
00:03:38linuxstb_kapoot: We can't host every Rockbox build forever...
00:03:53kapootyour builds are like 4megs ...
00:04:12 Join kyuubiseal [0] (
00:04:12Bjerrkpixelma: Well, fair enough. Can i play doom on the M5L, btw? :P
00:04:32bluebrotherdoom in greyscale is somewhat ... grey :)
00:04:35pixelmaso... I took that (and only got to know about that after I bought it...)
00:04:36 Quit seablue ("life, death, life, death")
00:04:37kyuubisealhow do u use rockboy
00:04:41Bjerrkbluebrother: i believe you! :P
00:04:45***Saving seen data "./dancer.seen"
00:04:51bluebrotherbut I tried it on my h120 ;-)
00:05:14Bjerrkbluebrother: Yeah, i have it on my (sigh) iPod Video.. Pretty fun.
00:05:14bluebrotherrockboy is part of Rockbox. No separate install needed.
00:05:22pixelmaBjerrk: never tried... (not interested at all)
00:05:24bluebrotherbesides, the manual tells you how to use it ...
00:05:24 Nick _pill is now known as pill (
00:05:40kyuubisealso how do u play GBC files on it?
00:05:42pixelmaBjerrk: but in theory yes
00:05:50*bluebrother really suggests reading the manual before asking
00:06:16Merthsoftkyuubiseal: put them on your rockbox, and navigate to them in files, I think
00:06:28Bjerrkpixelma: Well, okay.
00:06:39 Quit Nibbl (Remote closed the connection)
00:06:40kapootlinuxstb: does anyone plan on fixing this bug?
00:07:03BjerrkSpeaking of rockboy, do you guys know how to get rid of the blank screen bug?
00:07:35BjerrkThe wiki describes one way to do it, but it isn't valid any more (with the current build)..
00:07:35kyuubisealthis is what did
00:07:43Bjerrkkyuubiseal: yea?
00:07:48kyuubiseali made a file on my sandisk sansa
00:08:05kyuubisealand called it gameboy color
00:08:15Bjerrkno file extension?
00:08:40kyuubisealand then i place a .gbc game into it
00:09:00Bjerrkyou placed a game into another file?
00:09:04kyuubisealand when i open rockbox and go under files, gameboy games it is empty
00:09:25linuxstb_kapoot: According to the flyspray bug report, builds earlier than 26th July should work fine -
00:09:43 Quit obo ("bye")
00:09:44kyuubisealany idea's
00:09:51linuxstb_kapoot: So if you're experiencing problems with earlier builds, then it would seem to be a different issue?
00:09:52kapootlinuxstb_: i just put 13990 on there −−- same deal
00:10:10pixelmakyuubiseal: checked your file view settings?
00:10:15pixelmamaybe it
00:10:25pixelma's set to "music" only
00:11:17kapootlinuxstb: just found one locally from like november that i was running ... this hsould work −−- would be nice to have a new build though
00:11:47pixelmaeither in the quick screen or settings > general settings > fileview
00:13:01kyuubisealok i am there
00:13:05 Quit midkay (Read error: 110 (Connection timed out))
00:13:15 Quit Domonoky_ (Read error: 104 (Connection reset by peer))
00:13:17kyuubiseal"show files"
00:13:23kyuubisealis that where i go
00:13:37kapootlinuxstb: shall i submit another bug report?
00:14:01linuxstb_kapoot: So the November build works fine?
00:14:01kyuubisealsaported or music
00:14:19kapootlinuxstb: yes
00:14:40linuxstb_kapoot: So what exactly happens when you play music with a build from 24th July?
00:14:47pixelmakyuubiseal: well for gameboy files, "music" wouldn't do...
00:14:49kapooti get a DATA ABORT
00:14:55BjerrkMan, this just ain't good. There's more than 12 hours between the battery runtime with rockbox and the battery runtime with the apple firmware.
00:15:07linuxstb_kapoot: And you're using the latest bootloader?
00:15:29kapootlinuxstb_: possibly not
00:15:36kapootlinuxstb: how to check?
00:15:53linuxstb_Does the bootloader display messages when you're booting?
00:16:00linuxstb_(if so, it's an old version)
00:16:09kyuubisealand it works
00:16:21BjerrkNow i hate you, kyuubiseal
00:16:28kyuubisealdoes it sapport any gb game
00:16:33kapootlinuxstb: so i'm now operating under the assumption that i need ot re-install my bootloader via the whole initial process again?
00:16:45kyuubisealno lag
00:17:02linuxstb_kapoot: The "whole initial process" is just running ipodpatcher.exe and selecting "i" for install.
00:17:08kapootlinuxstb: do i need to back-up or does it correctly preserve my music partition now?
00:17:30Bjerrk.exe? :P
00:18:03 Join krazykit [0] (n=kkit@
00:18:15linuxstb_kapoot: A Rockbox bootloader install has never modified your music partition.
00:18:21 Quit kyuubiseal ("ChatZilla [Firefox]")
00:19:30 Quit kapoot ()
00:20:05linuxstb_You're welcome...
00:20:51BjerrkNow, i know this question is not really rockbox related, but have any of you tried the archos 604, the 605 or anythin in that department?
00:21:07rasherI wonder why people who were subscribed to flyspray entry don't get notified if it's reopened?
00:22:18BjerrkA bug in the bug system?
00:22:18 Quit davina ("xchat on Ubuntu 7.04")
00:23:29rasherHrm, or maybe it was something that was done when reopening. I see that -sf wasn't sent a mail about it either
00:24:05Bagderthat polish one?
00:24:20 Quit Febs ("ChatZilla [Firefox]")
00:24:27Bagderhe did three re-open request for it and I denied two of them first, so perhaps it confused it somehow
00:24:35rasherAh, could be
00:24:43rasherI'm pretty sure he wasn't testing with an updated build
00:24:57 Quit ddalton ("I was using BOFHNet IRC version 1.2 by fmillion - get your copy today from !")
00:25:03Bagderquite possibly, yes
00:25:09rasherThank god for icons in the menus - navigating Rockbox in Polish is hard
00:26:09pixelmause an english voice with it ;)
00:26:25rasherNot a bad idea
00:26:49 Join jhulst [0] (
00:26:57Bjerrkpixelma: I can't find the X5L on the cowon website either. What's with these L-models? Are they a hidden and rare treat only meant for cowon's favourite customers or what?
00:27:36pixelmaBjerrk: I already told you what the L stands for...
00:28:14rasherBagder: no time to test the perl voice script?
00:28:30Bjerrkpixelma: You did.
00:28:44pixelmathey got a larger battery (to be accurate it's a second battery) which is why they are in a thicker case, other than that there is no difference to the "standard" X5s or M5s
00:28:45Bjerrkpixelma: and how does that explain why they're not there? :P
00:28:57pixelmaask Cowon...
00:29:31 Quit kk (Read error: 110 (Connection timed out))
00:30:43pixelmaBjerrk: there is a bit about them under "specifcations" on the Cowon site
00:31:02BjerrkOkay.. Cool :)
00:31:05 Quit jgarvey ("Leaving")
00:32:47BjerrkThey do look nice :-( . Sadly, i can't find 'em anywhere.
00:32:53rasherBagder: looks like I'm just not getting my mail
00:33:28rasherBagder: that would explain it
00:33:31 Join RavoxX [0] (n=RavoxX@unaffiliated/ravoxx)
00:33:53Bjerrkrasher: You're from Denmark, right?
00:34:15 Part RavoxX ("Time makes no sense")
00:34:31Bjerrkrasher: Cool :) . Can you recommend a good device for use with rockbox, that's reasonably easy to get your hands on here in DK?
00:34:49BjerrkOne with a good battery runtime, preferrably..
00:36:22rasherBjerrk: The sansa e200 is all over the place, but not so hot best when it comes to runtime. And of course the ipod video is easy to get as well.
00:36:30 Quit SoapSud ("ChatZilla [Firefox]")
00:36:53Bjerrkrasher: Yeah, but the two things that matter the most to me are battery runtime and storage capacity
00:37:19rasherBjerrk: I've seen older irivers lying around in FONA, so maybe you could get lucky there or in other similar stores. Webshops probably don't stock enough to have older players still in stock
00:37:24Bjerrkat the moment, i've got an iPod Video (30 GB). The stor. capacity is fine, but the battery time is BAD.
00:37:49 Quit My_Sic (Connection timed out)
00:37:52Bjerrkrasher: Ah, thanks :) . FONA is just around the corner, pretty much.
00:38:03BjerrkMight be worth a look.
00:38:24 Quit bluebrother ("sleep ...")
00:40:03BjerrkI even considered buying this one ( ) as it's got wifi capabilities, and is Linux based.. But Archos doesn't seem like a nice corporation. At all.
00:40:13BjerrkSeems that the device is pretty much TiVoized..
00:40:37rasherFor good battery runtime you need Iriver H100 or H300-series or a supported iAudio.
00:41:48linuxstb_Or an old ipod... (1g or 2g)
00:42:12rasherWhich leads me to your next option: buying second hand
00:42:15linuxstb_Or gigabeat...
00:42:28rasherHow's the gigabeat's runtime?
00:42:33Bjerrkseems OK
00:42:49linuxstb_I think it's around 15-20 hours.
00:43:12BjerrkSeems that it's ranging from around 16 hours to 18 hours..
00:43:51BjerrkThat's pretty good. Are they available here in Denmark (rasher?)?
00:44:10rasherBjerrk: I'm looking now, but not finding anything
00:44:23rasherPerhaps some shop has an exclusive deal
00:44:29 Join sarixe [0] (
00:44:59linuxstb_AFAIK, the gigabeats aren't available new anywhere. You'll have to buy used. Around 100 USD should buy you the F40 (40GB).
00:45:18sarixehi, where can i get m68k-elf-gcc?
00:45:30Bjerrksarixe: Linux or Windows?
00:45:32linuxstb_sarixe: Run the tools/ script in the Rockbox source.
00:45:41sarixeah right
00:45:44sarixethanks, linuxstb
00:45:48markunall unhappy Gigabeat users should come to the rockbox forums. I'm sure they could sell their players.
00:46:01linuxstb_But then they would discover Rockbox...
00:46:04rasherAnd of course, buying anything but ipods second-hand is next to impossible in Denmark.
00:46:20BjerrkI just need to get this deal closed before sunday :P
00:46:23markunrasher: I bought mine from the uk
00:46:32linuxstb_markun: The S seems to be impossible to find, even without a working Rockbox port...
00:46:46markunlinuxstb_: that surprises me
00:46:49rashermarkun: Yeah, that's probably necessary.
00:47:14Bjerrkmarkun: Hey, where did you buy it?
00:47:22markunBjerrk: ebay
00:47:23 Join Twist [0] (
00:47:37markunbut over a year ago, they are much cheaper now
00:47:38Twistis it possible to set an alarm with rockbox?
00:47:55Bjerrkmarkun: Okay :)
00:48:19scorchelinuxstb_: just got one (that will be shipped to GodEater) from here last night
00:48:25linuxstb_Twist: On which device?
00:48:34Twistipod video
00:49:52Bjerrkmarkun: Do you think you could find the place where you bought it, again?
00:49:56scorcheheh... " 320 X 240 dots"
00:50:07rasherBjerrk: I found this: But it's probably just a left-over, and if you order it they'll tell you "Sorry, we don't have that" and offer you something awful instead.
00:50:21markunBjerrk: no, it was a guy who sold his player privately, not a company
00:50:35BjerrkArgh, the world hates me :P
00:50:41Bjerrkrasher: Interesting :)
00:50:43linuxstb_scorche: Yes, I found that site, but noticed they didn't accept non-US credit cards...
00:51:00scorchelinuxstb_: hence why i bought it and he paid me back
00:51:04Twistlinuxstb_: Possible?
00:51:20linuxstb_Twist: I think so, yes.
00:51:22rasherBjerrk: expensive too. And I really doubt they'll be able to sell it
00:51:35Twistim downloading the simulator to find out...
00:51:48BjerrkThat's really, really expensive, isn't it?
00:52:13linuxstb_Twist: It's possibly not in the simulator, as it relies on a hardware feature.
00:52:16rasherYeah, they probably just forgot to remove it when they stopped selling them, is my bet
00:52:22Bjerrk$427.941 USD
00:52:47Twistwell ok here goes.
00:53:01Twistim gonna back up before hand i thnk
00:53:29Bjerrkrasher: Also, it's uncomforting that the site is half danish / half swedish
00:53:45Bjerrk"Best. vara, obekrftad leveranstid" for example..
00:54:20rasherBjerrk: Hah, yeah, there are a few of such shops that are in fact based in Sweden but sell to Denmark as well.
00:54:47BjerrkYeah, i've noticed
00:54:57Twistcan you dual boot the itunes ware with rockbox?
00:55:11BjerrkI might as well try buying the thing :P
00:55:26rasherIt's really very expensive though
00:55:29BjerrkProlly wont work, though
00:55:42linuxstb_Twist: Yes.
00:55:49rasherI'd probably try finding an iaudio second-hand somewhere
00:56:02Bjerrkrasher: gives virtually nothing there
00:56:03Twistthanks, im haveing a looky at the manual
00:56:22rasherBjerrk: Yeah, no one's selling, it seems
00:56:46Bjerrkrasher: Because the device is simply too nice to ever part with. Now i want it even more.
00:57:16 Quit spiorf_ ("Read error: 110 (Connection timed out)")
00:57:46rasherYou could get 4 sansa E250 for the price of that X5
00:58:08BjerrkOn the other hand, i don't want them :P
00:58:37rasherNo wait, 3.
00:58:48BjerrkBut i could also get a wifi capable, linux based portable multimedia player for that price..
00:59:08linuxstb_With a shorter battery life than a Rockbox'd ipod video....
00:59:15iambenBjerrk: where do you find one of those?
00:59:46Bjerrkiamben: for example
01:00:04Bjerrkbut i still think that it's some TiVoised evil piece of crap.
01:00:17iambenive got a gp2x now
01:00:27TwistTo get the alarm clock feature to work will i need to leave my ipod on all night, since i do not have a dock and this whole idea was to eleiminate haveing to leave my pc running..
01:00:30iambenits open source but they aren't so good about upgrades
01:00:44 Quit lee-qid (Read error: 110 (Connection timed out))
01:00:48Bjerrkiamben: Isn't the gp2x mostly for gamers?
01:00:52 Quit robin0800 (" HydraIRC -> <- The professional IRC Client :D")
01:00:56rasherTwist: The alarm clock should work even if the ipod is turned off. Assuming it works in the first place.
01:01:03 Nick fxb is now known as fxb__ (
01:01:08rasherAt least, that's my understanding of it.
01:01:11iambenwell it plays games/emulators, but also does xvid/divx playback
01:01:13Twistok well i can only try it and see huh
01:01:14linuxstb_Twist: No. The alarm feature turns your ipod off and uses the RTC chip to turn it on at the set time.
01:01:43iambenonly downside for video is that it has no hdd, SD only, so it won't hold a big library or anything
01:01:44rasherTwist: you could test it now (set it to go off in a few minutes)
01:01:51Twistand it will play a given mp3 or wav?
01:02:00Twistnot installed atm
01:02:06 Quit ctaylorr (Read error: 110 (Connection timed out))
01:02:18linuxstb_You'll need to enable the "resume on startup" option. I think it's now somewhere like General Settings -> System -> Start Screen
01:02:41linuxstb_(I think...) Hopefully the manual describes it.
01:02:59Twistwill have a look
01:03:26 Quit Galois (Remote closed the connection)
01:03:32 Join Galois [0] (
01:04:05 Quit jhMikeS (Nick collision from services.)
01:04:08Twistok found it thankyou once again ><
01:04:11 Join jhMikeS [0] (n=jethead7@rockbox/developer/jhMikeS)
01:05:48Bjerrkiamben: SD only? So you can't have anymore than a couple GB of files on it at once?
01:06:17iambenwell they make 8G SD now
01:06:29markuniamben: how's the runtime?
01:06:39iambenand in theory, you can carry 10 SD cards with you... but yeah, a hard drive would've been nice
01:06:39BjerrkI'd like at least 25 GBs.. :/
01:06:50iambenits got usb host functionality if you can carry an external HDD around =)
01:07:01iambenmarkun: runtime?
01:07:13Bjerrkiamben: battery rintime..
01:07:17markuniamben: how many hours before you need to recharge
01:07:30iambenBjerrk: 5 hours or so average
01:07:40iambenbut it uses AA, so you can carry extra sets of those too =)
01:08:29rasherYou'd have to carrey a backpack for all those extras
01:08:44Bjerrk5 hours.. that's not too great.
01:09:05iambenit came w/ a little velvet pouch that i can fit 4 AA & a few sd cards easy
01:09:18markuniamben: and is that with video and games? If so, I think that playing mp3's will last a lot longer.
01:09:26iambenbut yeah, for just a movie player there are better options
01:09:39rashermarkun: true, didnt think of that
01:09:43markun5 hours for movies would not be bad
01:09:44iambenyeah 5 hours is for video or gaming, for mp3 the screen would be off so it'd be a lot higher
01:09:59iambenand much lower cpu usage i suppose
01:10:43iambenits got an arm940 @ 200mhz, i think the mp3 player by default clocks it down to 100mhz or so
01:10:53 Quit ender` (" The reason they call it the American Dream is because you have to be asleep to believe it. -- George Carlin")
01:11:09iambenand for cpu-intensive games you can crank it to 250mhz or so
01:11:22Bjerrkiamben: Okay, let's cut to the bone: what does it cost? :P
01:11:48iambenmine was $200USD several months ago
01:11:59linuxstb_iamben: Is it 100% open source? i.e. including all the low-level hardware drivers?
01:12:17iambeni think they are more easily available over in europe
01:12:48iambenlinuxstb_: its linux based, and i don't think they had to write many drivers if any
01:13:04Bjerrkyeah i know at least one guy around here who has one..
01:13:11iambenbut stuff they DID write, like a highly optimized mplayer port, is open sourced
01:14:11BjerrkSounds nice, but the missing hard drive is very much a turnoff :P
01:14:56scorchewhat device is this?
01:15:17iambeni believe they market it as a gaming console, just happens to playback vids too
01:17:49 Quit linuxstb_ ("CGI:IRC")
01:18:45rasherAaand suddenly I receive mails from the last 8 hours
01:19:46iambentubes got clogged
01:22:49BjerrkIt's funny how hard it is to find i a decent mp3 player
01:24:01Twistok everythings done ivegot myself some mp3's on there. ive set a wake up time but the ipod didnt switch offf.
01:24:06Twistcan i just turn it off?
01:24:49Bjerrkrasher: What do you think of this: ?
01:25:32BjerrkActually, i meant this one:
01:26:00 Part n1s
01:26:18rasherWell it won't run Rockbox. It's also a PMP rather than a music player, but that might not be a problem. I guess it's not really on-topic for this channel
01:26:36rasherBut I'm sure it's probably a decent device
01:27:37BjerrkWell, i guess you're right about the OT-ness.
01:27:43Twistok powerd off set to 00:30
01:27:49BjerrkHm, i better get some sleep.
01:27:52Twistjust a waiting game i guess
01:28:15Bjerrkg'night all..
01:28:23 Quit Bjerrk ("Dead Men Tell No Tales")
01:30:10Twistthanks guys
01:30:13Twistim off to bed
01:32:13Twistill be back
01:32:20 Quit Twist ("Leaving")
01:39:10 Join jhulst_ [0] (
01:39:15 Quit bdgraue (Remote closed the connection)
01:39:20 Quit jhulst (Remote closed the connection)
01:43:28 Quit tvelocity ("Αποχώρησε")
01:46:37 Join jhulst__ [0] (
01:55:27 Quit Thundercloud (Read error: 104 (Connection reset by peer))
01:55:58 Quit jhulst_ (Connection timed out)
01:56:33 Join hannesd_ [0] (
01:56:44 Quit H10_007quick (Read error: 104 (Connection reset by peer))
01:57:01 Quit HellDragon (Remote closed the connection)
01:57:28 Join HellDragon [0] (n=Nocebo@unaffiliated/helldragon)
02:02:31 Quit desowin ("use linux")
02:03:04 Quit iamben (Read error: 104 (Connection reset by peer))
02:04:49***Saving seen data "./dancer.seen"
02:11:32 Quit hannesd (Read error: 110 (Connection timed out))
02:11:33 Nick hannesd_ is now known as hannesd (
02:11:47 Join H10_007quick [0] (
02:11:59 Quit H10_007quick (Client Quit)
02:12:30 Join Llorean [0] (
02:12:32 Join Shaid [0] (
02:16:10 Nick jhulst__ is now known as jhulst (
02:16:26rasherLlorean: How serious are we about the "proper English" rule in the forums? Thinking about this one:
02:18:29Lloreanrasher: The primary concern is abbreviations like that 'btw' there, yes.
02:18:45LloreanBasically anything that would confuse someone trying to look up words in a dictionary.
02:19:10LloreanTechnical abbreviations like "FPS" are, to my mind at least, different simply because they're almost a term in their own right these days.
02:19:37 Quit ompaul (Remote closed the connection)
02:20:01rasherOkay, I'll leave it to you to mention it to him if you deem it necessary.
02:26:39 Join iamben [0] (
02:28:38 Part pixelma
02:32:55 Join ctaylorr [0] (
02:43:54 Join Shaid` [0] (
02:47:09 Join Bagder_ [0] (
02:50:29 Quit jhMikeS (Read error: 110 (Connection timed out))
03:01:54 Quit Shaid (Read error: 110 (Connection timed out))
03:01:55 Nick Shaid` is now known as Shaid (
03:04:49 Quit Bagder (Read error: 110 (Connection timed out))
03:05:30 Join jac0b [0] (
03:10:58jac0bI am having a problem with my patch
03:11:24jac0bpatch: **** malformed patch at line 12: @@ -101,7 +101,7 @@
03:11:24 Quit miepchen^schlaf ("Verlassend")
03:11:37 Quit sarixe ("Peace")
03:14:09jac0bI have check the line and it is correct
03:15:19jac0b101 means the line right?
03:16:10 Join H10_007quick [0] (
03:16:15 Quit H10_007quick (Client Quit)
03:17:25rasherjac0b: I don't know much about the syntax, but how was the patch generated
03:18:30jac0bsvn diff
03:19:35rasherSurprising.. I really don't know
03:20:08rasherCan you perhaps put the patch on pastebin?
03:23:08jac0bit is a sync up on the button swap patch
03:23:50rasherCould it be a line-ending issue?
03:23:55rasherWhich platform are you on?
03:24:38 Quit x1jmp (Remote closed the connection)
03:25:31 Join Phalangees [0] (
03:26:39rasherI really have no idea, sorry.
03:27:46jac0bthanks for the help
03:30:30EnterUserNamehey all
03:30:33EnterUserNameAnyone around?
03:30:43EnterUserNameHow do you use SANSA with Linux how do you mount the damn device? :S
03:31:32jac0byou have to put your sansa in MSD and not MTP
03:31:38EnterUserNamehow do you do that
03:31:51EnterUserName in settings?
03:31:54EnterUserNamewhats MTP and MSD :S
03:32:05jac0bMTP is the windows protocol
03:32:16jac0band MSD = Mass Storage Device
03:32:34EnterUserNamefound it thank u
03:32:34jac0bwhich I think is compatiable with most OS's
03:32:41EnterUserNameya.. its compatible with windows
03:32:51 Join Chewie[] [0] (
03:32:51EnterUserNamewhy do they even have that crappy protocol :)
03:32:54EnterUserNameshould be default as storage device
03:33:03jac0byes it should
03:33:20EnterUserNamehow do you reset it if it crashes
03:33:28EnterUserNameit crashed with the original firmwaer
03:33:34EnterUserNamedidnt know how to rset it
03:33:42jac0bhold down the power button for 15 seconds
03:34:35EnterUserNamealot of weird messages coming from fdisk
03:35:00EnterUserNameabout partition places
03:35:03EnterUserNameis that normal?
03:35:21 Part jac0b
03:37:20EnterUserNamehrm ok im sorry i found it in the docuemntation
03:37:23EnterUserNamei missed the top part
03:40:00 Join jhMikeS [0] (n=jethead7@rockbox/developer/jhMikeS)
03:42:07 Join datasleep [0] (
03:42:10 Quit datachild (Nick collision from services.)
03:48:48EnterUserNamehey all
03:48:52EnterUserNamewhy doesnt the Music folder appear
03:48:53EnterUserNameon rockbox
03:49:38PhalangeesIs it possible for rockbox to detect two buttons being hit at once?
03:49:56Phalangeeslike holding up and left at the same time
03:50:02EnterUserNameMy Linux drive sees the music folder.. but not rockbox.
03:53:29saratogaset rockbox to show all folders
03:55:52EnterUserNamesensa is onkyy 15 bit color
03:57:43LloreanEnterUserName: Please don't do a lot of lines with like "hrm" and "umm", the channel is logged and it makes them much more frustrating to read through and such.
04:00:20 Quit jhulst (No route to host)
04:01:51 Part Battousai ("Time makes no sense")
04:04:50***Saving seen data "./dancer.seen"
04:08:50 Quit SirFunk (Remote closed the connection)
04:19:45 Quit BHSPitMonkey ("Leaving")
04:20:20 Join BHSPitMonkey [0] (n=stephen@
04:20:59 Quit witz (Read error: 110 (Connection timed out))
04:22:07 Join witz [0] (n=witz@unaffiliated/witz)
04:22:42 Join aliask [0] (
04:22:47 Join H10_007quick [0] (
04:22:51 Quit H10_007quick (Client Quit)
04:28:11 Join sarixe [0] (
04:32:09EnterUserNamecan anyone let me know how rockbox album art works
04:32:11EnterUserNameif you have the patch
04:32:17 Quit qweru ("KVIrc 3.2.0 'Realia'")
04:32:38 Quit aliask ("ChatZilla [Firefox]")
04:34:57EnterUserNamenevermidn foudn wicki :)
04:36:47 Quit BHSPitMonkey (Connection timed out)
04:38:47 Quit witz (Connection timed out)
04:41:04 Join n17ikh|Lappy [0] (
04:45:34 Join SirFunk [0] (
04:48:00 Quit sarixe ("Peace")
04:56:44 Join jhulst [0] (
04:59:01 Join jrbil [0] (i=18567866@gateway/web/cgi-irc/
05:00:52 Join H10_007quick [0] (
05:01:00 Quit H10_007quick (Client Quit)
05:03:57EnterUserNamehi all
05:04:00EnterUserNamewhat is the commandl ine
05:04:01EnterUserNameto use ffmpeg
05:04:07EnterUserNamefor linux for mpegplayer
05:05:06jrbilanyone with e200 experience online?
05:05:14 Quit mbr (Read error: 110 (Connection timed out))
05:05:47EnterUserNameWell i just installed
05:05:56EnterUserNameso i have tiny little experience :)
05:06:33jrbili'm having trouble with building a working sansapatcher from SVN
05:06:50jrbildo you have 2 garbled lines at the top of your screen?
05:06:51EnterUserNameOh i didnt do it from scratch
05:06:57EnterUserNamei used custom build
05:07:02EnterUserNameadn they have fixes for it..
05:07:13EnterUserNameI think there's a patch to fix that
05:07:17jrbilwhere did you find the fixes?
05:07:30jrbili'm trying to patch a patch that i found.
05:07:48jrbili'm curious where you found your custom build.
05:08:03jrbilmaybe somebodies already got a better patch.
05:09:25EnterUserNameits on The Forums
05:09:29EnterUserNameunder unsupported builds
05:10:04EnterUserNamedo you know a good bitrate
05:10:08EnterUserNamefor mpegs?
05:10:39 Join Numb3rs [0] (
05:11:19Numb3rscan someone please help me how to put the plugins into my ipod?
05:11:39jhulstEnterUserName: This may be helpful if you are trying to get video onto your sansa
05:12:07 Quit Merthsoft ("Leaving")
05:12:25jhulstNumb3rs: What plugins? Most of them are included by default
05:12:35EnterUserNameeven album art?
05:12:46LloreanThat's a patch, not a plugin
05:13:31 Join webguest62 [0] (i=3fe6e0fa@gateway/web/cgi-irc/
05:13:47webguest62Hi guys, I need some help
05:13:56webguest62I can't find ipod_fw.exe
05:14:50Lloreanwebguest62: None of our instructions should say to use it...
05:15:18 Quit webguest62 (Client Quit)
05:15:33 Join xNumb3rsx [0] (
05:15:54 Join Pamacii [0] (
05:17:56EnterUserNamewoo hoo its working
05:20:55 Quit GodEater (Read error: 110 (Connection timed out))
05:21:45EnterUserNameoh my gosh
05:21:48EnterUserNamemyc over art is working
05:21:49EnterUserNamewoo hoo
05:22:31LloreanEnterUserName: Please, read the channel guidelines.
05:23:45 Quit Phalangees ("ZZZzzzzz")
05:26:21 Quit saratoga ("CGI:IRC (Ping timeout)")
05:30:13 Join H10_007quick [0] (
05:30:13 Quit jrbil ("CGI:IRC (EOF)")
05:30:22 Quit H10_007quick (Client Quit)
05:32:28xNumb3rsxanyone know how to install games into a Generation 5.5 ipod video?
05:33:00 Quit Numb3rs (Read error: 110 (Connection timed out))
05:34:39jhulstxNumb3rsx: what games? they should be installed by default
05:37:13 Join korndub [0] (
05:37:32korndubanyone have any tips on improving battery life on winpod 30g rockbox current build?
05:38:49BigMackorndub: turn of database if you are using it
05:39:06BigMacand what codec are you using primarily?
05:40:06korndubi use mp3
05:40:24korndubif i turn off database howw ill it work?
05:43:01korndubBigMac: i use Mp3 what do you mean turn off database?
05:43:45BigMacSorry, korndub , my kid is sick brb
05:44:06DogBoyit's explained in the docs
05:44:18DogBoyI don't remember the details but it will work without database
05:45:41BigMackorndub: Ok back. Do you sync your files via iTunes?
05:46:33xNumb3rsxJhulst: i mean how can i install more games for free?
05:46:50 Join ApooMaha [0] (i=4b596eb2@gateway/web/cgi-irc/
05:47:52ApooMahahi, im over on the irc channel because i need to be granted permission to edit the wiki?
05:47:54EnterUserNamewhich is quicker ffmpeg or mencoder?
05:48:02EnterUserNamefor rockbox..
05:48:07ApooMahai use winff
05:48:12EnterUserNameim a linux user..
05:48:24EnterUserNamei cant use winff..
05:48:39EnterUserNamei can
05:48:40ApooMahaso i have ported a theme from the video to the nano and i need to put it in the wiki
05:48:45korndubBigMac: i dont sync with itunes....i manually update music...
05:48:56ApooMahasome sort of permission thing?
05:49:03BigMackorndub: then turn it off
05:49:13ApooMahawho, me?
05:49:46ApooMahaas youve probably noticed, im new on the irc channels
05:50:25ApooMahahi, im over on the irc channel because i need to be granted permission to edit the wiki?
05:50:39jhulstApooMaha: Should have access
05:51:43jhulstxNumb3rsx: As far as I know, there aren't any other games that you can put on, unless there are some in the patch tracker
05:52:46 Quit ApooMaha ("CGI:IRC (EOF)")
05:54:13 Join webguest36 [0] (i=4b596eb2@gateway/web/cgi-irc/
05:54:22 Quit webguest36 (Client Quit)
05:54:37 Join ApooMaha [0] (i=4b596eb2@gateway/web/cgi-irc/
05:54:44 Quit ApooMaha (Client Quit)
05:54:58korndubwhat kind of battery life is expected for 30 gb ipod?
05:54:59 Join perrikwp [0] (n=chatzill@
06:00:44korndubjhulst: thanks
06:01:02 Quit HellDragon (Client Quit)
06:03:38 Join HellDragon [0] (n=Nocebo@unaffiliated/helldragon)
06:04:16 Quit BigMac (Read error: 104 (Connection reset by peer))
06:04:53***Saving seen data "./dancer.seen"
06:05:28 Join BigMac [0] (
06:07:37 Quit korndub ("Changing servers")
06:13:46 Join RudMan [0] (
06:19:44RudManI'm registered at - can someone enable me for write permission in the wiki?
06:23:37jhulstRudMan: What's your wiki name?
06:24:19RudMananyone know about the maintainer of rockbox-themes? the upload is busted and there is no response to email
06:24:47scorcheyup...blame me
06:26:07jhulstRudMan: Did you register for wiki access? I don't see your name on it
06:28:03RudMangrrr - I'm such an idiot. I asked the question before I realized that registering for the wiki is different than registering for the forums.....I've registered for the wiki and am awaiting the activation email. Sorry
06:28:56RudManscorche: the upload errors out saying the email is not valid
06:28:57jhulstRudMan: No problem, once you get activated I think you should have write permission, if not, come back and ask
06:29:11scorchei know exactly what is wrong I'll be able to upload in- what - 5 minutes?
06:38:00 Quit xNumb3rsx (Read error: 110 (Connection timed out))
06:58:28scorcheno...i dont currently have the time to fix it, and i am waiting on a file from the admin of the site's local copy
06:58:41 Join witz [0] (n=witz@unaffiliated/witz)
07:03:19RudManyour site is really useful
07:04:36scorcheit isnt my site, and i am making the new official rockbox theme respository which will pretty much do away with the purpose of that site
07:04:38RudManI hope you can find the time soon - I have a few themes to upload
07:05:51 Quit DogBoy ("Leaving")
07:06:13 Quit Chewie[] ("z")
07:13:58 Quit RudMan ("Leaving")
07:25:03 Quit midgey ()
07:25:23 Quit jhulst (No route to host)
07:43:48 Quit amiconn (" HydraIRC -> <- Like it? Visit #hydrairc on EFNet")
07:44:30 Join amiconn [0] (
07:54:03 Join toffe82 [0] (n=chatzill@
07:59:00 Part toffe82
07:59:19 Join toffe82 [0] (n=chatzill@
08:01:01 Join ddalton [0] (
08:01:41 Join webguest07 [0] (i=c023110b@gateway/web/cgi-irc/
08:04:16 Join hidden212 [0] (i=d15edb04@gateway/web/cgi-irc/
08:04:57***Saving seen data "./dancer.seen"
08:05:06JdGordonhas anyone actally done a battery bench on the e200?
08:05:47hidden212the e200 needs a lot of work for the battery runtime
08:06:14scorche(he knows...he is a dev...)
08:06:59hidden212is there any interests in making the Rockbox environment more iconic and 3D
08:07:05JdGordonyeah, I just saw a comment in the forums saying its up to 12h now, and i realised mine will not be used for the next 27 and its got full batt so may as weel do a proper test
08:07:34scorchehidden212: not really, but that isnt saying a patch wouldnt be welcomed by some
08:12:37 Join GodEater [0] (n=bryan@rockbox/staff/GodEater)
08:13:14scorchei really wasnt thinking that the cup would get there before the others...
08:13:34LloreanYou can only really do that with the main menu anyway, but a start would be simply using the existing iconset and displaying it on a novel way.
08:14:16hidden212hmmm that's true
08:14:28*GodEater thinks he may have wandered into an interesting conversation
08:15:11hidden212it is possible to also make them glossy and stylistic
08:15:45scorchethey will be displayed however they are drawn...
08:16:00Lloreanhidden212: The existing icons are customizable... they can look however the themer wants them to look.
08:16:06hidden212but which is better to study if i want to fully develop Rockbox: Computer science or information technology
08:16:09LloreanWhat you'd be doing is creating alternative methods of displaying the already themable icons.
08:16:30scorchewell, those are broad terms, really
08:16:34LloreanAs it is you can display the main menu as text-only or text-with-icons, a third option would be icon-only.
08:17:02hidden212that's true
08:20:00hidden212it might work well with scrollwheel targets
08:20:46*amiconn really doesn't get it...
08:21:12JdGordonicon only wouldnt work...
08:21:17LloreanJdGordon: It would for the main menu
08:21:21JdGordonbut showing them in a grid instead of a list might be nice
08:21:34 Join LinusN [0] (i=linus@rockbox/developer/LinusN)
08:21:43LloreanI'm not talking about a list of icons, I did mention that they'd have to be displayed in a different way.
08:22:38JdGordonwell... seen as almost all the gui uses the same list wigdet there is no rea reason why it couldnt be prettied up a fair bt.. if osmeone cold be bothered
08:22:38amiconnJdGordon: I think it would reduce usability, except *maybe* on targets with cursor cross + separate select
08:23:01JdGordonyeah, maybe
08:23:17hidden212not really
08:23:20Lloreanamiconn: Unfortunately, a lot of our users favour "pretty" over "usable"
08:23:30hidden212that's true
08:23:54*scorche wonders why so many people dislike text
08:23:57amiconnLlorean: Yeah, and that's one thing I don't get. If they want it pretty but useless, they should use the original firmware
08:23:59hidden212it would attract more attention and users
08:24:01ddaltonhow do I zip something on cygwin?
08:24:04ddaltonlike a dir?
08:24:05LloreanAnd making our interface at least look a little more "modern" by their standards, since how it looks is *obviously* more important than what it can do in judging it, might be nice.
08:24:23Lloreanddalton: Everything in cygwin is accessible through explorer too.
08:24:41ddaltonok what about on linux how would you do it?
08:24:50ddaltonjust out of quriosity.
08:24:51Lloreanamiconn: One simple step we could do toward making it usable is on color targets making the default icons colored, so it doesn't look B&W any more (to the people who don't notice the light blue)
08:24:58JdGordonscorche: icons are good because they are recognised faster than text.. if they are big enough
08:24:59GodEaterddalton: but if you really want to do it "zip dir1/"
08:25:12Lloreanddalton: Use the zip command
08:25:55 Part toffe82
08:26:14hidden212users will welcome a more "iconic and graphical" environment
08:26:17amiconnMy impression about how many products are made today is that they're designed to death, disregarding functionality. And imho rockbox is about functionality
08:27:05scorchehidden212: is that your guess, or have you taken a large poll? ;)
08:27:11Lloreanamiconn: Doesn't mean we can't add some glitz in if someone submits a patch for it, and it doesn't hurt much.
08:27:24Lloreanhidden212: Ah, but the people we want to attract most are developers. :-P
08:27:25hidden212but rockbox will definitely not loose functionality anytime soon
08:27:52Lloreanamiconn: Would you have a problem with color icons by default on color targets, like these:
08:28:13amiconnhidden212: A fairly large group of rockbox users will be in strong opposition to your statement - the blind users
08:28:30hidden212you are right
08:29:03Lloreanamiconn: If it were an alternate display of the same widget, it could still use the old voice strings.
08:29:04JdGordonLlorean: we cant really d that bcause the default icons need to work for all background colours... which a coloured set might not
08:29:07GodEaterand they're a pretty vocal bunch too
08:29:20LloreanJdGordon: The default icons already don't work for black backdrops.
08:29:27amiconnAnd I mostly dislike this talk about design because it distracts developers from working on the basics
08:29:32hidden212but Rockbox has always been about functionality... it will never loose scope if a side trip to developing a stylistic environ is taken
08:29:37LloreanJdGordon: We could simply include a theme for the mono-icons, but default to colored ones.
08:29:40JdGordonLlorean: the do unless the forground is black also
08:29:42amiconnAnd many of the basics aren't working right yet on swcodec
08:29:46 Quit Shaid (Read error: 104 (Connection reset by peer))
08:30:04ddaltonGodEater: How do I make it include the sub dirs and the files inside the directory?
08:30:08*JdGordon goned
08:30:17ddaltonwith the zip command?
08:30:21GodEaterddalton: it should do that with the command I gave you
08:30:44GodEaterah no
08:30:48GodEateradd a * at the end
08:30:53amiconnLlorean: An icon grid changes navigation, and a list is easier to navigate (regarding icon menu and blind users)
08:31:05ddaltonso I am in my home dir and I type the command and get it to run on submit dir but only submit is in the zip file.
08:31:09GodEaterddalton: so "zip dir1/*"
08:31:14ddaltonand no files in that
08:31:22Lloreanamiconn: Ah, I wasn't envisioning a grid. I was still thinking linear navigation of some sort.
08:31:26ddaltonok then
08:31:40LloreanThough with a grid, you could still do next-column or next-line wrapping.
08:31:57amiconnTrue, but what advantage would it have then?
08:32:11hidden212perhaps it could be like the e200 menu system?
08:32:14GodEateris anyone actually thinking of implementing this, or are we just blowing hot air again ?
08:32:41*GodEater is siding with amiconn, and doesn't want the menu changed.
08:32:41*scorche votes for the latter
08:32:49amiconnThe list has the big pro that if you don't recognise what the icon means (and it *is* difficult to make icons with a clear meaning) you have the text
08:33:13GodEaterwhich is a very good point
08:33:21LloreanI don't think the graphical menu should ever be on by default.
08:33:29LloreanBut I don't see why it couldn't be an option.
08:33:43LloreanI think the main-menu icons are mostly recognizable, and could be learned in about one go.
08:33:54*scorche laments for binary size
08:33:58hidden212nothing is going to change
08:34:12hidden212we are adding graphic outlike as an option
08:34:23GodEater*who* is adding it ?
08:34:32hidden212Rockbox is always about providing customizable choice features
08:34:39scorcheand adding things increase the binary size
08:34:46GodEaterhidden212: that's not what it's about to me
08:34:48Lloreanhidden212: No, it's really about audio functionality.
08:35:02Lloreanscorche: I think a basic icon-only view could be added with a minimal binary size change.
08:35:07LloreanAs long as nobody got too fancy
08:35:39amiconnLlorean: I have icons enabled, but in fact I always use the text to navigate, seeing the icons just as a kind of list bullets. When looking closely, several of the icons are misleading to me
08:35:45hidden212rockbox has good audio functionality compared to many OF
08:35:55ddaltonis anyone interested in committing p6138 and it requires p7647?
08:36:09amiconnWell, 2 to b eprecise
08:36:10Lloreanamiconn: I agree, I'm just saying I don't see it as too problematic to cater a little to the graphical interface crowd, if it can be done somewhat efficiently.
08:36:28amiconn"Settings" and "System"
08:36:36LloreanThat one is problematic.
08:36:42LloreanI think with larger icons it wouldn't be.
08:36:50LloreanGears for Settings, a picture of a player for System, perhaps
08:37:01amiconnIt doesn't have to do with icon sizes
08:37:15LloreanSystem is slightly misleading, as a name, anyway
08:37:18hidden212a picture of a Radio for Radio (obviously)
08:37:24LinusNan icon grid could actually be faster to navigate for sighted users
08:37:26LloreanSince there's a system menu under Settings too.
08:37:34amiconnThe best association I saw so far for settings are two slider knobs in different positions
08:37:49amiconnThe current icon means "menu" to me, but not settings
08:37:56LinusNsettings are often depicted as a spanner
08:37:58amiconnLinusN: How?
08:38:04LinusNlike "tools"
08:38:24LinusNamiconn: because you can reach an item with less keypresses
08:38:58amiconnI doubt that...
08:39:05LinusNsimple math really
08:39:06LloreanIt's mathematically guaranteed.
08:39:42amiconnIt depends on how many buttons are available, and how you can map them to a grid in an intuitive way
08:40:10amiconnAnd a grid has another disadvantage: it only looks good with a square number of items. But the number of items in the main menu is variable
08:40:11Lloreanamiconn: As long as it's done properly, you'll still never need *more* buttons
08:40:17LinusNamiconn: of course, it's not suitable for all targets
08:40:18LloreanErr, more presses
08:40:23amiconnIt varies by target, and on some targets it even varies dynamically
08:40:42LinusNamiconn: you said you didn't care about the looks :-)
08:40:43amiconn(radio detection)
08:40:52hidden212rockbox with a graphic environ option would be no different from.... say... a nokia 6230
08:41:14scorchei think that is kind of our point...
08:41:19amiconnLinusN: I didn't say that
08:41:29LloreanIf you use next-line wrapping on left/right, you can just make the last "row" of icons the uneven one, and space it as best you can.
08:41:36hidden212what about a slideshow-type menu?
08:41:40*amiconn hates the Nokia menus
08:41:48 Join hcs [0] (n=agashlin@rockbox/contributor/hcs)
08:43:22LloreanOf course, you'd also have to deal with icons too large to fit the grid to the screen.
08:43:23GodEateris it possible to return to the default iconset without resetting everything at start up? And if so - how ?
08:43:25amiconnLinusN: I surely have nothing against a nice look, but currently it annoys me because many of the basics are incomplete or broken on swcodec
08:43:30LloreanSuch as on an H10 5gb with large icons.
08:43:58LloreanGodEater: There should be a theme for it, I thought
08:44:20LinusNi firmly believe it is possible (on the targets with large color displays and 4-direction navigation) to implement a both good-looking and efficient gui
08:44:50LinusNbut i agree with amiconn that we need to fix the basics before working on "rubbish" ;-)
08:45:00hidden212this is a good conversation... unfortunately i g2g
08:45:06amiconnLlorean: That would be the fault of the theme designer then
08:45:33Lloreanamiconn: Yeah, but it'd still be nice if it could accept larger than necessary icons. Or adjust the grid safely based on icon size.
08:45:48LloreanEither way, 's something for the future.
08:46:03hidden212the next time around, i would like to discuss ways in which rockbox can attract more devs to contribute
08:46:28*JdGordon knows how to kill he topic... (although i may be too late)
08:46:39JdGordonthe current scroll code wont work for any other "menu" system
08:46:57*amiconn has the impression that the things-to-fix list is constantly growing instead of shrinking
08:47:24LinusNamiconn: you are probably right
08:47:36amiconnOkay, maybe partially my fault, because I added the 1st/2nd gen port...
08:47:58*amiconn reminds LinusN of radio detection for X5/X5V
08:48:08amiconn...and the oldlcd player thing
08:48:09LinusNamiconn: do you have a clue regarding the apparent ipod instability that's being reported?
08:48:51amiconnIt's only the nano afaik. I suspect it's an ata issue, but I can't fix that myself as I would need a nano for experimentation
08:48:58*LinusN has very little rockbox time nowadays
08:49:01amiconn(preferably one that shows the problem)
08:49:24 Join gtkspert_ [0] (
08:49:26LinusNamiconn: i hear about freezing on video 80GB too
08:49:56hidden212speaking of the nano...
08:50:02amiconnWell, it doesn't freeze completely. I observed that myself. Music continues, but the UI is frozen
08:50:11LinusNamiconn: nasty
08:50:13 Join Siku [0] (
08:50:18hidden212is rockbox on the nano 2 gen impossible?
08:50:28scorchenothing is impossible, per say
08:51:00amiconnYeah, but it is most probably unrelated to the clock setup changes, because a few cases were already reported in May, shortly after adding support for the 5.5/80
08:51:09hidden212i noticed on the ipodstatus twiki it says it requires a new port
08:51:28scorcheand it is correct
08:51:32 Quit JdGordon ("Konversation terminated!")
08:52:41LinusNamiconn: so it could still be an ata issue?
08:54:34amiconnI doubt that it's ata on the 5.5
08:54:55amiconnI'd rather think it might have to do with the broadcom chip
08:55:19 Quit Rick ("I don't need to be here.")
08:55:23amiconnI need to ask pixelma, iirc she even read about problems on the 5.5/30, before we supported the 80
08:56:36amiconnLinusN: Do you use your 5.0/60 much these days?
08:56:37 Quit hidden212 ("CGI:IRC (EOF)")
08:56:39 Quit hcs ("Leaving.")
08:56:56LinusNamiconn: occasionally
08:57:41amiconnHmm. Did you experience this kind of freezing on the 5.0?
08:57:54LinusNnot yet
08:58:02amiconnRecent build?
08:58:48LinusNhmmm, 2-3 weeks or so
08:58:56*amiconn tries to update his targets at least once per week
08:59:00amiconnLinusN: Btw, if it were ata, I doubt the music would continue...
08:59:16LinusNok, so it continues to fill the buffer as well?
08:59:41LloreanOn the Nano, when there aren't crashes, you get corrupted audio even.
09:00:04amiconnLlorean: Yes, the nano issue most probably is ata
09:00:16LloreanI was just pointing out, the symptoms are really quite different.
09:00:48 Join Rick [0] (
09:01:21ddaltonwhat is wrong with the indenting at
09:01:49ddaltonI have written a simple c program because I am trying to get my text editor emacs to do the indenting properly.
09:02:04 Quit gtkspert (Read error: 110 (Connection timed out))
09:02:26GodEaterddalton: the brace indenting is a little odd
09:02:44ddaltonhmmm I will have another look at this email.
09:02:52ddaltonso is everything else ok?
09:03:00GodEaterddalton: otherwise it looks fine to me
09:03:12ddaltonwhat about the line int x, y
09:03:23ddaltonint x, y; I mean.'
09:03:24GodEaterddalton: yep, looks fine too
09:03:35ddaltonok good just got to figure this out.
09:04:25*GodEater is amazed ddalton takes so much time over something which basically offers him no help at all, but is really good for sighted programmers only
09:04:29amiconnLinusN: Not sure, but I think it does
09:04:40 Join ender` [0] (
09:05:02*amiconn needs to flip the whole powermgmt thread logic around :/
09:05:23ddaltonwell I have to do it to have any chance of getting my code excepted in to rockbox. And I really need to stop getting other people to fix it for me.
09:05:38ddaltonActually it might help me later when I use a braille display.
09:05:43ddaltonbut yeah your right.
09:07:06ddaltonis it possible to right some sort of script that could fix my braces?
09:07:16ddaltonon linux?
09:09:34GodEaterddalton: such scripts already exist
09:09:45GodEaterddalton: there is a program called simply "indent" which does the job
09:09:56GodEaterddalton: however I've not used it personally, so I don't know how good it is
09:10:06 Join davina [0] (
09:10:27amiconnrasher: around?
09:10:36 Join B4gder [0] (
09:12:00 Nick gtkspert_ is now known as gtkspert (
09:14:08 Join freqmod [0] (
09:14:50LinusNamiconn: what about the power management?
09:15:19 Join petur [0] (n=petur@rockbox/developer/petur)
09:16:12amiconnCurrently the powermgmt thread has most of the battery calculation in the outer loop (the thread function itself), and calls a "sleep" function which is only left once perminute (or twice for software charging)
09:16:23amiconnThis inner loop function does event processing etc
09:17:17amiconnFor target treeifying I need to flip this, so the thread function does the event processing etc, and calls the calculation stuff once or twice per minute
09:17:37amiconnThen the calculation for software charging can be moved to the recorder tree
09:18:01amiconnI also found that the poweroff handling is kinda strange
09:18:10GodEaterddalton: using this command line "indent -bl -bli0 -br -i4 -nut ddaltons_program.c" generates what I call acceptable indenting
09:18:41GodEateralthough perhaps one of our senior devs might like to fiddle with it a bit more to get it to conform to strict "Rockbox style"
09:18:46amiconnThe powermgmt thread checks playback and radio status itself (but in 2 different ways), but recording is handled elsewhere (probably in recording)
09:18:49GodEaterit seems extremely flexible
09:19:13amiconnI think the power "locking" could work in a similar way as cpu boosting
09:19:29 Join Zagor [0] (n=bjorn@rockbox/developer/Zagor)
09:19:44amiconnThe powermgmt thread would maintain a counter for the number of locks, and if that counter is zero, it starts counting down the shutdown timeout
09:20:13 Join avis [0] (n=avis@pdpc/supporter/student/avis)
09:20:15amiconnAs soon as the lock count isn't zero, the countdown would be reset and put on hold
09:20:29avisdoes anyone know of a open source firmware for a ipod nano second generation ?
09:20:38avisi realize rockbox isn't supported for that ipod
09:20:41B4gderavis: there is none
09:20:44GodEateravis: no - there is none
09:20:47amiconnThen recording, playback, radio... would just lock/unlock depending on their state
09:21:07amiconnThis mechanism could be easily reused for plugins: mpegplayer, video.rock...
09:22:14amiconnAnd it would be more modular than checking audio_status in powermgmt....
09:22:15ddaltonwhere do I type that command?
09:22:46GodEaterddalton: assuming you have "indent" installed, you type it in the directory where your .c file is that you'd like to have indented properly
09:22:56GodEateronly replace ddaltons_program.c with your actual file name of course
09:23:04ddaltonso from the command line in cygwin? not in emacs?
09:23:27GodEaterddalton: cygwin command line yes
09:23:33B4gderemacs can re-indent too
09:23:34GodEaterddalton: assuming your cygwin install includes it
09:23:48GodEaterB4gder: it looks like emacs uses that self same program
09:24:39ddaltonGodEater: I don't have indent installed. What would it be under in the package selection in cygwin? Or do you know how I can install it?
09:25:06GodEaterddalton: sorry - not a cygwin user here :(
09:25:20ddaltonok so I type indent -bl -bli0 -br -i4 -nut my_test.c
09:25:26ddaltonfrom the command line?
09:25:32GodEaterddalton: yes
09:27:32ddaltonI wonder if google might be able to help me here.
09:27:40GodEatergoogle often can
09:28:15GodEater"cygwin indent" returns an indication that it *is* available for cygwin
09:28:25ddaltonGodEater: what do you use?
09:31:19amiconnLinusN: Have such an UI freeze on 5.5/80 now. Music continues, will see whether it rebuffers
09:31:38amiconnButtons also react in that they make the backlight light up, but nothing more
09:31:59amiconnFortunately I am playing flac, so it has to rebuffer....
09:32:05ddaltonon google?
09:32:10ddaltonor in cygwin
09:32:37 Quit avis (Remote closed the connection)
09:34:18amiconn...and it just rebuffered. Music is still playing. So it can't be ATA.
09:35:06 Join kubiixaka [0] (
09:35:16GodEaterddalton: that was my google query
09:35:20 Nick midkay_ is now known as midkay (
09:35:25 Join ie [0] (i=c27f0811@gateway/web/cgi-irc/
09:35:28 Join hcs [0] (n=agashlin@rockbox/contributor/hcs)
09:36:17ieHello. I have two questions about sansa. (1) Is the cable that sansa uses a standard one? I.e. one of its ends is just normal USB but the other? Is it also a standard?
09:36:25ddaltonnow I just need to figure out how to install it. I found it.
09:36:39GodEaterddalton: I'm sure you'll manage - you're a bright guy :)
09:37:07ie(2) Is it possible to disable wheel light with RB? It brings me nothing (one can hardly miss it) and only consumes power.
09:37:30 Join BHSPitMonkey [0] (n=stephen@
09:37:52Lloreanie: It's nonstandard, and I believe you can disable the wheel light.
09:38:03LloreanAt least, it didn't used to be enabled. So you should be able to. :)
09:39:17ieLlorean: do you mean that RB has a setting for this?
09:39:48LloreanI mean that I expect it does, I haven't looked.
09:40:06 Join Rob222241 [0] (
09:41:51 Join haemmy [0] (i=sternsch@
09:43:17 Quit freqmod (Remote closed the connection)
09:43:36ieLlorean: hmm, so, on your player the wheel doesn't light up? I'll look for a setting then.
09:45:04Lloreanie: No, it lights up on mine.
09:45:26LloreanIt didn't used to. I haven't looked, but I'm pretty sure there's now a setting for it.
09:46:07LinusNamiconn: so it could be a main thread/button action issue?
09:46:17amiconnI don't think so
09:46:31amiconnI'm not pretty sure the main thread freezes in lcd_update()
09:46:53amiconnBacklight is handled by the button tick task + backlight thread, so it continues to function
09:48:33amiconnEven baclight fading works properly
09:49:22 Join ninjai`primax [0] (
09:49:32 Quit MournBlade ("Her lips were wet with venom")
09:50:17ninjai`primaxi think i may be blind
09:50:31LinusNamiconn: you think it freezes in lcd_update?
09:50:38ninjai`primaxbecause i couldn't find any info on what directory to store the music in
09:50:47amiconnLinusN: yes
09:50:54LinusNninjai`primax: you can store it anywhere you like
09:51:56LinusNamiconn: interesting, i didn't know that was possible
09:52:01amiconnAnd since the ipod video lcd update yields, all other threads continue to function
09:52:24amiconnIt yields in the finishup
09:52:28LinusNamiconn: does it hang spontaneously, or do i have to provoke it?
09:52:43amiconnSpontaneously, but it may take a while
09:52:50LinusNi see
09:53:53GodEateris it time for me to chime in and say "This has still never happened to me on my 80GB" ?
09:55:07 Quit ddalton ("I was using BOFHNet IRC version 1.2 by fmillion - get your copy today from !")
09:55:11 Join ddalton [0] (
09:55:22 Join miepchen^schlaf [0] (
09:55:29GodEaterforums are in "wading through treacle" mode again I see
09:56:36GodEaterLlorean: you still here ?
09:56:43amiconnNo wonder video is so slow on ipod video....
09:56:56amiconnlcd_yuv_blit() isn't optimised
09:57:14LloreanGodEater: Yes.
09:57:42 Quit Rob2222 (Read error: 110 (Connection timed out))
09:57:46GodEaterwhat's your personal opinion on my last post in the "Tact and Diplomacy" thread? Do you have a maximum number of Experts you want in the forums ?
09:58:41 Join Domonoky [0] (
09:59:16LinusNGodEater: we don't need more experts, we need less idiots :-P
09:59:22LloreanGodEater: I figured you were asking that, and I've been typing a response.
09:59:32GodEaterah ok - I shall await it then
09:59:36LloreanIt should be there now
09:59:44GodEaterit's just I think we have at least two other people in the forums who seem clueful
09:59:54LloreanHm, which two?
10:00:02GodEaterand LamdaCalculus
10:00:13LloreanThere's also BigBambi.
10:00:16GodEateryeah him too
10:00:38GodEaterI was going to say "he can be a bit abrupt though" and then thought, "pot, kettle"
10:00:42LloreanI'm a little bit uncertain about LambdaCalculus. I'll admit I'm not the most professional acting at times, but some of his comments seem to criticize the poster, rather than what they've said.
10:00:44Domonokylinuxstb: would it be possible to change the "targetname" strings in ipodpatcher ? they dont match the targennames in rbutil.. :-)
10:00:44LinusNi like bigbambis response to the f40 guy: "Yes, oddly enough the instructions in the manual to restore the OF are correct."
10:01:10GodEaterLlorean: well if you thought about handing ANY of them a badge upgrade
10:01:17GodEateryou could also explain you expect a bit more "professionalism"
10:01:49LloreanI have sent PMs with several of the Expert badges.
10:01:59GodEatermy main point for bringing it up, is that those three seem to spend ALL their time in the forums
10:02:10GodEaterrather than in here
10:02:10 Quit ie ("CGI:IRC (EOF)")
10:02:19 Quit ninjai`primax ("( :: NoNameScript 4.03 :: )")
10:02:22GodEaterso they're not dividing their concentration if you see what I mean
10:02:33ddaltonis there anyone here who knows about ini files?
10:02:53*Domonoky knows a bit about ini files.. :-)
10:03:01LloreanI don't really have any objection to handing them out Expert badges. I'll probably hand them out in the next day or two.
10:03:27ddaltonok If I upload my cygwin ini file can you edit it so it installs another package?
10:03:35LloreanI could do it in the next 30 seconds, but I'd like to check out each of their posting styles again so I can write personalized PMs.
10:03:39LinusNi absolutely *love* the Pen&Paper theme for 5G
10:03:45GodEaterLinusN: me too :)
10:03:52GodEaterhe really went to town on it
10:03:53LloreanIt's one of my favorites.
10:04:00LloreanI'm tempted to port it to a few of my targets.
10:04:02Domonokyddalton: sorry thats cygwin knowledge, i dont have.. :-)
10:04:06GodEaterit actually made someone stop and ask me about it in the lift at work the other day
10:04:12ddaltonok then.
10:04:25ddaltonI can't work out how to install a tar file.
10:04:27amiconnNo opinions on the powerdown locking?
10:04:38GodEaterI like the animated man waving at you when you're paused too
10:04:40GodEaternice touch
10:04:43LinusNamiconn: the what?
10:04:45Lloreanamiconn: Powerdown locking?
10:04:58***Saving seen data "./dancer.seen"
10:05:01ddaltonmaybe I should just get a linux machine.
10:05:06LinusNGodEater: haha, i hadn't noticed! :-)
10:05:11amiconn09:18 ff
10:05:43Lloreanamiconn: For idle powerdown?
10:05:57linuxstbDomonoky: Yes - grep is telling me that ipodpatcher isn't actually using targetname for anything... So feel free to change them if it will help rbutil.
10:06:15Domonokylinuxstb: good, i will do..
10:06:20LinusNamiconn: i think i would prefer a keep-alive mechanism instead of a lock
10:06:42amiconnIt would need to be called in regular intervals....
10:06:59LinusNthen it isn't a lock, is it?
10:07:15amiconnNo, I mean a keepalive mechanism
10:07:27LinusNthen we agree :-)
10:07:39amiconnYes, but why is that better than a lock?
10:07:49LloreanWouldn't a keepalive mean several things have to call the keepalive, or at least know that something else is doing so?
10:08:29LinusNi can only compare with how complicated the cpu boost locking has been
10:08:50LinusNa keepalive is more robust, since you don't have to remember to unlock
10:08:56linuxstbamiconn: Maybe it would be sensible to add a timeout to that yield() loop in the ipod video lcd_update?
10:09:18 Join bascule [0] (i=86920006@gateway/web/cgi-irc/
10:09:26amiconnLinusN: hmm...
10:09:41amiconnlinuxstb: Maybe, but I think it's more likely that we're doing something wrong
10:10:00linuxstbamiconn: Something wrong in the lcd_update?
10:10:18amiconnI just searched the 5.5 OF disassembly for "characteristic" bcm addresses, and there's also a 0x30050000 which we don't use
10:10:40amiconnMaybe a timeout will fix the symptoms for now
10:10:56*amiconn thinks that finishup_needed should be a bool
10:11:08LinusNbut do we know that the lcd can be refreshed again without a reset?
10:11:21amiconnNeeds testing
10:11:33amiconnHnce the 'maybe'
10:11:46ddaltonCan someone tell me if this will be able to run linux and compile rockbox?
10:11:48linuxstbIIUC, there are two firmware images used on the broadcom chip. There's one in the flash ROM, which I assume is transferred to the broadcom chip on boot, and another on the disk, which I assume is used by the main firmware. So it's possible they communicate in different ways.
10:11:52LinusNi have run my 5G for an hour now without hanging
10:12:54LinusN"the perfect POS machine" does that mean Piece Of Shit?
10:13:19Lloreanddalton: I can't seen any reason why it wouldn't be able to.
10:13:21linuxstbddalton: Almost anything can run Linux. But a 866MHz CPU and 256MB RAM will be quite slow when building Rockbox.
10:13:22LinusNi think that machine will run linux and compile rockbox quite well
10:13:34GodEaterLinusN: that's no good then, we need "Bag Of Shit", not a mere Piece
10:13:39LinusNnot very fast, but it will work
10:13:46LinusNGodEater: hahahahaha
10:13:49amiconnlinuxstb: In theory we should be able to use the on-disk broadcom image, shouldn't we?
10:14:19linuxstbamiconn: I don't see why not.
10:15:30amiconnMaybe the problem is that the in-flash image is buggy on some G5.5s?
10:16:01GodEateris it time for another "What version OF are you running?" question on the forums ?
10:16:47amiconnMight also be a newer chip revision that's more sensitive to timing/command order/whatever
10:18:12GodEaterwell my 80GB was bought *very* soon after they brought them out
10:18:22GodEaterso it's likely the earliest version available
10:18:45 Join pixelma [0] (i=pixelma@rockbox/staff/pixelma)
10:19:58B4gderDomonoky: do note that ipodpatcher was using the target names as found in configure...
10:20:43B4gderthere are some targets that appear with different short names all over :-)
10:20:54DomonokyB4gder: the problem is, there are to much different names for the targets :-)
10:21:49B4gderyes, we should punish the wrong doers harder!
10:22:08linuxstbThe broadcom image on disk is a FAT image, and one of the files is "vmcs.bin" which is about 256KB. The flash ROM contains an image called "vmcs" which is about 95KB. So there are definitely two different firmwares.
10:22:31linuxstbB4gder: There are also the model identifiers used for scramble (numbers and four character strings), the target number in configure...
10:22:44 Join aliask [0] (
10:22:52 Quit Domonoky (Read error: 104 (Connection reset by peer))
10:23:17B4gderyeah, we should start referring to targets by that number instead ;-P
10:23:22Lloreanlinuxstb: Have you checked that the vmcs.bin isn't padded to 256kb (for some unknowable reason)?
10:23:31amiconnlinuxstb: The main firmware uses a lot more addresses for bcm communtication. 0x30050000, 0x30060000, 0x30070000 are a few of them
10:23:38GodEaterah, a true paranoid :)
10:23:47B4gderI think my 23 player is nicer than my 2
10:23:57LloreanI'm quite fond of my 21 but wish it had better battery life.
10:24:03LloreanBut my 40 sees all the use day to day.
10:24:13GodEaterI prefer 40, but somehow use my 22 more
10:24:20B4gder40 is not a legal target_id!
10:24:33B4gder40 is just the menu number
10:24:44B4gderit uses 20 as target_id
10:24:45Lloreantarget_ids are sequential?
10:24:52B4gderpretty much
10:25:05*Llorean is gonna need a checklist.
10:25:14B4gdermeant to remain unique, while as we can change the ones shown in the configure menu
10:25:21LloreanMakes sense
10:25:28linuxstbLlorean: Checking again, it's actually around 261KB, and no, it's not padded.
10:25:42Lloreanlinuxstb: Okay, 256 just sounded like too convenient a number. :)
10:27:27GodEaterso 20 > 15, but 15 sees more use due to it's compact size, and 9 only ever goes out in the car
10:28:00amiconnlinuxstb: Maybe the big image contains the video codecs, and the rest is identical?
10:29:08*GodEater wonders that if that's true, whether some clever ipod hacker could replace the codecs with different ones
10:29:25pixelmaLinusN: got a few minutes?
10:29:26ddaltonso would that machine I mentioned be faster than cygwin?
10:30:21Lloreanddalton: Depends an awful lot on how fast a machine you're running cygwin on.
10:30:26linuxstbamiconn: There seem to be a set of libraries for decoding - .vll files in the FAT images. I've compared a hexdump of the two firmwares, and they start to differ after the first 4 bytes...
10:30:43GodEaterddalton: why not just try running linux on your current machine ?
10:30:57LinusNpixelma: sure
10:31:13linuxstbamiconn: Files like "h264dec_mw.vll", "aacdec_mw.vll" etc
10:31:18*amiconn has 1, 2, 7 (half working), 8, 9, 10, 12, 21, 24, 29 (2x) ;)
10:31:40amiconnMostly using 2, 8 and 10
10:31:49ddaltonbecause I need windows. How could I?
10:31:53ddaltonrun it
10:32:04linuxstbYou can dual-boot.
10:32:16ddaltonis that hard to setup?
10:32:19linuxstbOr if your PC has a reasonable amount of RAM, you can use vmware.
10:32:51linuxstbddalton: I would expect all Linux installers to set up dual-boot by default.
10:32:51B4gderddalton: dual-boot is often setup automatically for you when you install linux on a box that already has a windows partition
10:32:54GodEateror just try a livecd in the meantime
10:33:14*amiconn doubts that live cds come with tts
10:33:30*GodEater is willing to bet there's at least one out there which does
10:33:33ddaltononly one does but the command line isn't accessible.
10:33:35GodEaterthere's soooo many of them
10:33:51 Join bluebrother [0] (i=C4absD67@rockbox/staff/bluebrother)
10:33:54*GodEater wins his bet
10:34:00 Join obo [0] (n=obo@rockbox/developer/obo)
10:34:35 Join tvelocity [0] (
10:34:49ddaltonIf I setup dual booting will it slow windows down or linux?
10:34:54GodEaterddalton: no
10:35:01bluebrothernone of both
10:35:10GodEaterddalton: the only thing that will happen is that you will reduce the amount of disk space available to windows
10:35:14ddaltonwhat can go wrong in the install.
10:35:23GodEaterddalton: so if you're short of that, it might be worth just buying another hard drive
10:35:29bluebrotherdepends on what you consider "wrong" ;-)
10:35:43ddaltonwell where it is broken and won't boot.
10:35:59bluebrotherdual-booting requires installing a bootloader that allows you to choose between both OSes
10:35:59GodEaterddalton: I *personally* haven't seen a linux install go that wrong in years
10:36:10GodEaterddalton: but it remains a remote possibility
10:36:26bluebrotherthat's the only thing that might be critical (but is usually absolutely safe)
10:36:35ddaltonSo how do you actually install linux?
10:36:41ddaltonjust what file do you run?
10:36:47bluebrotherand of course you need to make sure to not overwrite the windows installation when installing linux
10:36:48ddaltonwould you do the install in windows
10:37:01linuxstbYou normally boot from an installation CD or DVD
10:37:09*amiconn already saw non-booting linux quite often
10:37:15 Join webmind [0] (
10:37:21webmindgood morning :)
10:37:30GodEateramiconn: you seem plagued with linux problems
10:37:31linuxstbamiconn: Yes, but that's different to making the original OS unusable
10:38:11webmindhmm, how do I find out what gen/type ipod I've got ?
10:38:21linuxstbwebmind: Run ipodpatcher and it will tell you
10:38:31ddaltonok there you go
10:38:38GodEateronce you have it connected to your PC anyway
10:38:41GodEaterit's not magic ;)
10:38:47linuxstbGodEater: That's for 2.0
10:38:55GodEaterlinuxstb: I'll hold you to that
10:39:06*linuxstb makes a mental note that he can now never release 2.0
10:39:10ddaltonso you guys think I should just dual boot.
10:39:10 Join spiorf [0] (
10:40:02bluebrotherI use dual-boot on my box too.
10:40:05linuxstbddalton: It depends how often you want to use Windows. You may find it easier to just buy a second PC, then you can use both at the same time.
10:40:37bluebrotheryou can also use a virtual machine −− I have a VirtualBox running windows
10:40:37linuxstbI would find dual-booting a pain - I only ever use one operating system on any computer.
10:40:52ddaltonwell that's what I thought because I will probably need access to irc ie thunderbird and then thre rockbox compilers.
10:41:01ddaltonthe rockbox compilers
10:41:09linuxstbYou can run thunderbird in Linux, and there are also many IRC clients...
10:41:19amiconnDoes putty work with tts?
10:41:28linuxstbBut I don't know how well the speech options work...
10:41:43ddaltonok well I think I should just a basic pc to run linux and then when I start programming a lot more I will update.
10:41:46 Join desowin [0] (n=desowin@
10:41:56ddaltonbecause I really need to learn linux properly.
10:42:11linuxstbddalton: The other advantage of a second PC of course is that there is zero risk to your existing setup...
10:42:29ddaltonyeah well that's what I thought.
10:42:55ddaltonalso if I get a pc running win 98 will the install of linux just over write it?
10:42:57amiconnIf putty can be used with tts then it would be possible to use the vmware image, running it in the background
10:43:10ddaltonI don't have enough ram.
10:43:16ddaltononly 1 gb for vista
10:43:35linuxstbddalton: Yes - the installer should let you completely overwrite the existing operating system.
10:43:37ddaltonand I could never get SSH to work.
10:43:40linuxstb^linux installer
10:43:42amiconnThe non-graphical vm needs only 128MB of ram
10:44:23ddaltonwell I wouldn't be using the grapical version of linux. So what would be a good amount of ram to have?
10:45:08ddaltonvista requires 1 gb of ram apparently but I find that hard to believe.
10:45:24 Join ie [0] (i=c27f0811@gateway/web/cgi-irc/
10:45:40*linuxstb doesn't
10:45:42webmindhmm, how do I find out if it's a 32mb or a 64mb
10:45:46webmindipod ?
10:45:52webmind(it's a 5th gen ipod 30gb)
10:45:58B4gderwebmind: 30gb is 32mb
10:46:02bluebrotherit depends on the programs you intend to run. I ran a graphical linux installation with 256MB a couple of years ago
10:46:13ddaltonhow did it go?
10:46:57bluebrotherit worked well back that time. Updating to current software and abandoning the gui stuff should work just fine
10:47:02linuxstbwebmind: Does the manual still refer to 32MB and 64MB versions of Rockbox?
10:47:05ddaltonlinuxstb: You think vista requires 1 gb of ram?
10:47:14 Join pondlife [0] (
10:47:29ddaltondid you compile rockbox?
10:47:38bluebrotherin fact, I use my old box as server sometimes −− no gui, and not much RAM
10:47:50linuxstbddalton: I've seen vista run on a 3GHz AMD64 and it was about the slowest windows install I've ever seen.
10:47:56bluebrotherand I used that box compiling Rockbox. It's just a bit slow (K6-2 450MHz)
10:48:05linuxstb(and that was with 1GB of RAM I think, possibly even more)
10:48:33ieLlorean: the only settings (in setting_list.c) that could be related to wheel light are "button light timeout" and "button light brightness". I haven't find anything special for the wheel.
10:48:35ddaltonwell I can't say I am a fan of vista
10:48:36Lloreanlinuxstb: I installed Vista on a 2.0 GHZ Dual core athlon lapto with about 2gb of RAM, and it was actually faster than most of my XP installs I've done. I think there's some fishiness in their whole process.
10:48:59 Part bascule
10:49:03amiconnFor vista, many computer magazines recommend 2GB of RAM nowadays
10:49:17ddaltonso if I get a 500 mhz cpu and 256 mb ram how would compiling go?
10:49:22ddaltonusing linux
10:49:30ddaltonwith out the graphical interface.
10:49:54ddaltonI have only got one. And it is much the same as my old computer.
10:49:59pondlifeAny mpegplayer gurus around? Looks like might be a useful inclusion...
10:50:08bluebrotherit will work. Just it'll be slow.
10:50:17ddaltonhow long do you think?
10:50:40bluebrotherI can try a clean build on my old box. Wait a couple of minutes ... ;-)
10:50:42amiconnI would think a 500MHz linux box would still compile rockbox faster than cygwin on a 2GHz windows box
10:50:47ddaltonok thanks.
10:51:00bluebrotherwow −− is cygwin really _that_ slow?
10:51:04ddaltonwell cygwin takes about 20 minutes here.
10:51:09ddaltonmaybe 25
10:51:10Lloreanpondlife: I believe that patch is still somewhat glitchy
10:51:45amiconnddalton: Wow. Do you have a virus scanner running during compilation
10:51:48pondlifeLlorean: I think I'll suggest it's broken down into smaller patches. Good idea?
10:52:02ddaltonwell it is in my tray.
10:52:03Lloreanpondlife: Absolutely
10:52:04amiconnI found that it helps a lot to disable that while compiling
10:52:19ddaltonok I will try that next time.
10:52:44amiconnMy laptop also needs ~30 minutes for a build with cygwin and virus scanner running
10:52:44ddaltonbut I might even go up to a 1ghz cpu and 256 ram how would that go.
10:52:49ddaltonwith compiling
10:53:01amiconnDisabling the virus scanner brings it down to ~10 minutes
10:53:19ddaltonthat's a lot isn't it?
10:54:21 Quit cmvjk (Read error: 110 (Connection timed out))
10:54:35ddaltonso in a machine with 256 mb of ram and 500 mhz cpu is the cpu slowing things down?
10:54:41ddaltonor the amount of ram.
10:54:59linuxstbpondlife: I would like jhMikeS to comment on that patch, as he's been working on the buffering code in mpegplayer with a view to implementing seeking.
10:55:10ddaltonso your saying 256 mb of ram is fine but maybe go up to a 1ghz cpu?
10:55:21 Join cmvjk [0] (
10:55:55bluebrotherddalton: building now ...
10:58:28Lloreanie: There's a "Wheel Light Timeout" option in the menu, I've gone and checked
11:00:30webmindlinuxstb, looks like it
11:00:43webmindlinuxstb, Warning: If your player has 32MB RAM (typically if it has a 30GB harddisk), do not download the Ipod Video 64MB build as it will not work.
11:00:55webmindrockbox looks as wonderful as ever btw...
11:00:56linuxstbwebmind: OK... The builds are now called "30GB" and "60/80GB"
11:01:04webmindthanks :)
11:01:09webmindI kinda noticed yes :
11:01:21ddaltonok how would this go with compiling?
11:02:02LinusNddalton: looks ok, but more memory will certainly help
11:02:17ieAh, yes! This is called "button timeout" in the code but is spelled as "wheel timeout" in the menu (via special sansa entry in english.lang). Thanks Llorean! Also for your patience!
11:02:52Lloreanie: I assume then that it's the same thing as the button-light timeout on Gigabeats, perhaps
11:02:55ddaltonas in ram or hdd?
11:03:08bluebrotherboth :)
11:03:33amiconnFor compiling on linux without graphical desktop, even 128MB RAM are enough
11:03:44dionoea10 gb hd is more than enough to compile rockbox IMO
11:03:51bluebrothermy linux installation (about one year old, including GUI programs) requires about 13GB on hard disk
11:04:05amiconndionoea: Depends on how many builds you want to keep around...
11:04:15bluebrotherbut I also have an experimental installation around with about 4 GB system size
11:04:38linuxstbAnyone have a view on the 32MB/64MB line in the manual? It currently says "Warning: If your player has 32MB RAM (typically if it has a 30GB harddisk), do not download the Ipod Video 64MB build as it will not work." but the builds are now called "30GB" and "60/80GB". Should we just delete that warning?
11:04:41Lloreanbluebrother: I think Soap is running an Ubuntu install, and compiling RB, on an 8gb thumb drive.
11:05:26Lloreanlinuxstb: I'd replace it with "While the 60/80gb iPods can run the 30gb version, if you listen to music without skipping often the 60/80gb version may offer you better battery life"
11:05:31LloreanOr something similar
11:05:39LloreanSo it suggests to users the intent of the version.
11:05:48bluebrothersounds reasonable.
11:05:55linuxstbLlorean: Hmm... That may just confuse users.
11:06:08preglowwhy not just tell them to get the right build?
11:06:18preglowthere's no reason to tell them they can run an inferior build
11:06:33Lloreanpreglow: The 32mb build may not be inferior for some peoples' playback habits.
11:06:39linuxstbpreglow: I guess Llorean's point is that if the user skips tracks often, then it's not inferior.
11:06:40LloreanApparently filling the 64mb buffer can take quite some time.
11:07:11LloreanOne thing that could be done...
11:07:24LinusNstill no freeze on my 5G...
11:07:27LloreanJust direct them to the "right" build, and maybe we could add a section on "How to improve your battery life" with tips for various players
11:07:43preglowi still say we don't try to confuse people here
11:07:52dionoeaLinusN: how are you trying to make it freeze?
11:07:58preglowpeople should get the build for which they have a player
11:08:02ddaltoncould I update the ram if I need to later on?
11:08:03LinusNjust letting it run
11:08:14preglowand if some people MIGHT want to only use 32 megs of ram, that sounds like a setting to me
11:08:20LinusNddalton: yes, most likely
11:08:23preglownot a reason to use another build
11:08:29dionoeawith modified source code ? or just svn trunk ?
11:08:42ddaltonok then I will look at something like that machine. Just got to make sure the sound card will be ok.
11:08:43LinusNdionoea: svn trunk
11:08:45ieLlorean: yes, code wise button light and wheel light are the same (wheel is a button in that sense)
11:09:13dionoeaLinusN: it never froze using that ... and i use it all day long here (30GB version)
11:09:30dionoeamaybe i'm not using the right features to make it freeze :)
11:09:31linuxstbpreglow: So something like "WARNING: There are two ipod video builds - one for 30GB models, and one for 60/80GB models. You must ensure you use the correct one for your player."
11:09:40preglowthat's what i'd go for
11:09:43amiconndionoea: It seems to depend on the individual unit
11:10:25dionoeakeyboard freezes sometimes but it always comes back a few secs later
11:10:33preglowideally, we should have different buffering schemes in the playback engine anyway, but i can't see that happening anytime soon...
11:11:07ddaltonhas anyone noticed that iriver h300 sometimes doesn't boot rockbox?
11:11:16ddaltonand requires the settings to be reset
11:11:26Lloreanpreglow: Such as, 80% of the buffer being sequential music, then the first 5 seconds of several other songs for "more responsive" skipping, etc?
11:11:52linuxstbpreglow: Yes - it would be nice for any rewrite of the buffering code to be prepared for that.
11:11:53preglowLlorean: like that, for example
11:12:14linuxstbBut it sounds like simply abstracting the buffering code from the playback engine would do that job.
11:13:04preglowabstracting is good
11:13:16linuxstbTo some degree...
11:13:28preglowyou mean like everything else? :P
11:13:39preglowsure, not saying you should overdo it
11:13:46 Nick datasleep is now known as datachild (
11:13:46preglowbut rockbox does too little of it in places
11:14:02linuxstbYes, Rockbox can't be accused of over-abstraction...
11:14:33GodEaterI'd go with "WARNING: There are two ipod video builds - one for 30GB models, and one for 60/80GB models. If in doubt, use the 30GB one"
11:14:40preglowwhy would you be in doubt?
11:14:45*linuxstb was about to ask that
11:14:48preglowthe size is etched on the damned thing
11:14:57ddaltonwould the 30 gb one work for the 80 gb?
11:15:01linuxstbAnd on your receipt, and the box, ....
11:15:27GodEaterpreglow: we've seen at least one example of someone getting their 80GB recon'd, and getting a 30GB board back with it
11:15:51preglowi'd say that is fairly rare
11:16:12preglowand not something we should worry about anyway
11:16:32preglowor should we start providing builds for archos records with h120 mainboards too? :P
11:16:57linuxstbGodEater: One solution to that would be to make the bootloader detect the RAM size, and refuse to load a 64MB build on 32MB ipods (and display a message).
11:17:06preglowis this a big problem?
11:17:15linuxstbpreglow: I doubt it...
11:17:18preglowthen i'd say the manual notice should do
11:17:19GodEaterI doubt it too
11:17:26GodEaterI'm just saying it's happened is all :)
11:17:30rasherIt happened once in a while when they were labelled 64mb build
11:17:30preglowuntil someone goes ahead and codes proper ram detection support in rockbox
11:18:04linuxstbrasher: People installing the wrong build is different to Apple installing the wrong motherboard...
11:18:10rasherI really doubt it'll be a problem now that it's labelled 60/80GB ipod video
11:18:41rasherHas that even happened more than once?
11:18:41preglowme too
11:18:52preglowit's a freak occurence, don't worry about it
11:19:28 Quit scorche (Connection reset by peer)
11:20:11Lloreanrasher: I *think* it's happened twice.
11:20:29preglowanywho, what is the scheme for a rockbox build with ram autodetection? move plugin stuff to a fixed offset at the start, then detect end of ram before setting buffer size?
11:20:39rasherStill rare enough to pretty much ignore, I'd say
11:20:48bluebrotherI just stumbled upon FS #2960 −− shouldn't the script go to svn (and the task closed)?
11:20:57 Quit webguest07 ("CGI:IRC (Ping timeout)")
11:21:12rasherbluebrother: what, mine?
11:21:14preglowbluebrother: forgot about that
11:21:22linuxstbpreglow: Yes, basically put the codec and plugin buffers at the start of RAM (before the main Rockbox binary).
11:21:24preglowis it important now that rbutil is coming up?
11:21:31bluebrotheryes. Or simply forget about it as rbutil does the patching now.
11:21:39preglowi'd say just close it
11:21:49preglowno commit, time to let fwpatcher rot
11:22:12rasherSome people may not want to use a huge beast like rbutil
11:22:21preglowwhy? :P
11:22:34bluebrotherit's a nice beast ;)
11:22:36rasherThey might not want to install qt, or even X
11:22:52preglowstatic binaries
11:23:05bluebrotherthat's the reason why I wondered about committing the script
11:23:21preglowdoes that thing compile for unix too, then?
11:23:42*preglow reads the entire patch text
11:23:48bluebrothererr −− it's a shell script around the scramble tools ...
11:23:49preglowokay then :P
11:24:22preglowi mistook it for an old command line patch for fwpatcher, but that's already in, i see
11:24:55preglowfrontends for scramble tools wouldn't hurt, i guess
11:25:54 Part Llorean
11:27:30 Join Thundercloud [0] (
11:27:55 Part ie
11:30:36 Quit spiorf (Read error: 110 (Connection timed out))
11:31:30 Part pixelma
11:35:18linuxstbbluebrother: Thinking about ipod detection in rbutil, it seems this method could be used to determine the RAM size of an ipod -
11:36:11linuxstbWhich is possibly something we should implement in the Rockbox USB driver...
11:38:09bluebrotherddalton: a clean build of the mini 2G needed 15 minutes −− machine is K6-2 450MHz w/ 128MiB RAM
11:38:40ddaltonthat is a long time.
11:38:54bluebrotherwell, rebuilding usually is way faster.
11:38:57ddaltonso if I got 256 mb ram and 1 ghz cpu how faster do you think it would go
11:39:08ddaltonso how long would it take to compile?
11:39:32bluebrotherwell, the limiting factor is the cpu in my case.
11:39:47bluebrotherccache also might help −− I don't have it installed on that box
11:40:21ddaltonwhat is "ccache "?
11:40:40bluebrothercompiler cache. It caches parts that haven't changed.
11:42:43 Join webguest07 [0] (i=c023110b@gateway/web/cgi-irc/
11:44:56preglowpetur: any news from ac yet?
11:45:21 Join P3at [0] (
11:45:42bluebrotherlinuxstb: interesting ... but can we do such an inquiry easily on windows?
11:45:50bluebrotheror would this require a device driver?
11:48:13markunpreglow: what do you think about committing FS #7286 ?
11:48:49preglowmarkun: let me have a quick look at the latest patch first
11:49:16preglowyeah, it still has the unnecessary bias adjustment in dithering
11:49:44preglowi could remove that and commit it now, but i have no means of testing it
11:50:32 Quit pondlife (Read error: 110 (Connection timed out))
11:51:55preglowi'm Cfairly certain it works, but does anyone have a cf target and the possibility to test after i commit?
11:52:11preglowbest placed 'C' ever
11:53:06 Join pondlife [0] (
11:53:16linuxstbbluebrother: Yes, it will be a standard part of the generic scsi driver
11:54:21linuxstbbluebrother: I did make a start on integrating it into ipodpatcher, but didn't finish it. That's probably the best approach, rather than making it rbutil specific.
11:55:34markunpreglow: I have one, but no time to test
11:57:04peturpreglow: I've set austriancoder a deadline (sunday evening), if I do not see his code by then, I start the evaluation with a negative result for him. I'm getting tired of his way of working...
11:57:06rasherLet me just make some more noise about FS #7646 - perl voice building. If anyone could test it, or possibly even have a look at the code (perhaps implementing SAPI building), that'd be nice. (B4gder?)
11:57:48markunpetur: will he hold the code hostage in that case? ;)
11:58:13peturmarkun: that is his choice
11:58:20markunof course
11:58:22linuxstbpetur: Do I understand correctly that there are two versions of the low-level code - one written by AC, and one which is a port of MrH's code?
11:58:50peturhe swithced to MrH to find out why it wasn't working
11:59:20linuxstbSo we don't actually have any working code yet?
11:59:20peturmarkun: I told him I have been _very_ patient with him, I have to draw a line somewhere
11:59:59linuxstbI think we've definitely as a mentoring organisation learnt a lot for next year...
12:00:53preglownot a surprise, really
12:00:55 Nick fxb__ is now known as fxb (
12:00:58preglowwe've had no guidelines
12:01:06peturyes, we have been way too relaxed about this
12:01:08preglowand we should be tons stricter by next year
12:01:20preglowbut that's the way it goes anyway, you need experience to do aything
12:01:24preglowanything, and now we have experience
12:01:36peturand I should have gotten myself a suitable pp target _before_ the start
12:01:58ddaltonI am about to try and port espeak to rockbox. Any tips on how I should go about this?
12:02:04ddaltonit is in c++
12:02:15preglowwell, an obvious start would be porting it to c
12:02:25ddaltonso the whole code?
12:02:35pondlifeYes, and any libraries it depends on
12:02:41ddaltonso don't make it compadible with rockbox until it is in c?
12:02:42preglowwell, saying espeak is going to end up in the core is not wholly unrealistic
12:02:48preglowso you're going to have to port the whole thing to c
12:02:51preglowwe can't use c++ code in the core
12:03:16preglowyou COULD start developing it as a plugin, in which case it could remain c++
12:03:23peturddalton: how about finding a programming task that is your size? I think porting espeak will a bit too heavy for you...
12:03:26linuxstbddalton: I would forget Rockbox for now, and make it compile as a standalone C program without any dependencies on external libraries, or system library functions such as malloc (which don't exist in Rockbox).
12:03:27preglowbut you'd be the first to do such a thing (making a c++ plugin)
12:03:32ddaltonwould that ever be excepted?
12:03:45ddaltonok well I think I will stick to c
12:03:59ddaltonso first port it to c and then as a plugin?
12:04:04*petur walks away for a while
12:04:24preglowwhen it's c, you can integrate it to the core, make a plugin, whatever you like
12:04:27ddaltonwill I need to know c++?
12:04:39preglowwell, if you are to translate something from c++ to c, then sure...
12:04:46markunddalton: as far as I can tell the only c++ thing is the "Translator" class.
12:04:56ddaltonso I need to finish learning c and then learn c++?
12:05:01***Saving seen data "./dancer.seen"
12:05:07markunso, we should replace it by a struct I think
12:05:09linuxstbddalton: I don't know C++, but I've converted programs from C++ to C...
12:05:34bluebrotherddalton: did another test ... building h120 took nearly 13 minutes. Rebuilding with FS #7627 applied required less than 30 seconds
12:05:43ddaltonok then so I should make it run on my windows machine all written in c?
12:05:51ddaltonok then
12:06:03bluebrotherI assume there's something fishy with my arm compiler −− it threw a couple of warnings while building.
12:06:13linuxstbddalton: Yes, but without any windows-specific functions. i.e. make it read a text file from disk and write the output to a WAV file (or similar).
12:06:31linuxstbOr just hard-code a sentence to speak in the .c program...
12:06:59ddaltonmarkun: So what files need converting?
12:07:26markunddalton: just look for "Translator::"
12:07:30markunthere are quite a few files
12:07:40ddaltonok then so search them all?
12:07:44linuxstbmarkun: Is there floating point as well?
12:07:51markunlinuxstb: not in the main loop
12:08:09ddaltonok and how good will I need to be with c?
12:08:29markunddalton: it might also be a good idea to talk to the espeak developer about it, he's a nice guy
12:08:39ddaltonok then.
12:08:48ddaltonso will I need to know pointers and arrays?
12:09:17bluebrotherI bet so
12:09:38 Join My_Sic [0] (
12:09:40linuxstbmarkun: Have you spoken to the espeak developer about C? i.e. would he be against converting the official code to C?
12:10:00markunlinuxstb: I've talked about it, I'll search my emails..
12:11:00markunI think he's not against it: "Yes, I think it's only the Translator class.
12:11:00markunI think he's not against it: "Yes, I think it's only the Translator claMy original idea was that you could write replacement functions within
12:11:03 Quit kubiixaka ("Miranda IM! Smaller, Faster, Easier.")
12:11:03markunthe Translator class for different languages, where necessary. But I
12:11:06markunhaven't used that much.
12:11:13 Join kk_ [0] (n=kkit@
12:11:14markunthat didn't come out so well..
12:11:18ddaltonso how would it voice the screen through the speakers?
12:11:26ddaltonwould any of that need to be changed?
12:11:27linuxstbmarkun: That could still be done via function pointers in a struct anyway.
12:11:37markunlinuxstb: yes, exactly what I wanted to do
12:11:50markunbut I haven't done much programming lately
12:13:22linuxstbmarkun: I see espeak is now GPLv3...
12:13:31markunis that a problem?
12:13:34*linuxstb doesn't know the implication of that (if any)
12:13:44 Join Domonoky [0] (
12:13:53bluebrotherwill that make it incompatible to v2?
12:14:03linuxstbbluebrother: That was my initial thought.
12:14:11B4gderI'm quite sure that it is
12:14:25linuxstbespeak says "gplv3 or later"
12:14:27bluebrotherthat means we would need to use an older version
12:14:29B4gderv3 is far more restrictions than v2
12:14:39markunbluebrother: or ask jonathan nicely
12:14:53bluebrotherright ...
12:15:16amiconnmarkun: Any news regarding your font caching work?
12:15:22markunamiconn: nope
12:15:46markunI'll not ask about the viewports :)
12:17:28amiconnB4gder: couldn't we go gplv3?
12:17:45B4gderwe could, if all copyright owners agree
12:18:05B4gderpersonally I'm not a convinced gplv3 devotee
12:18:17rasherI rather agree with Torvalds on gplv3
12:18:30linuxstbI guess the question is if we would be happy if someone "tivo-ised" rockbox?
12:18:39*amiconn has nothing against gplv3
12:18:48*linuxstb wouldn't be happy if someone did that
12:19:16amiconnrasher: Regarding your perl script for voice, what do I need to do to use it?
12:19:35markunlinuxstb: me neither
12:19:38rasherlinuxstb: I personally wouldn't mind. It'd still bring Rockbox to more people, and we'd get any improvements into the main Rockbox branch
12:19:48amiconnI would like to try it for building english and german at&t voices on cygwin
12:19:54markunlinuxstb: in that case we could just as well have chosen the BSD license
12:19:55preglowhow would someone tivoize rockbox?
12:20:07B4gderI agree with rasher
12:20:14rasheramiconn: I haven't re-implemented SAPI5 building yet, so you can't do that
12:20:15markunpreglow: use it for their DAP and making it impossible to modify it
12:20:23B4gderI rather have a company tivoize rockbox than not use it at all
12:20:31*linuxstb wouldn't
12:20:38preglowbut they'd have to not modify it at all, yes?
12:20:49B4gderif they modified it, we'd get the changes back
12:21:03 Quit My_Sic (Read error: 104 (Connection reset by peer))
12:21:06B4gderpreglow: the tivoize part is most likely done in the bootloader part
12:21:17B4gdersignatures, checksums etc
12:21:18linuxstbpreglow: They can modify it, and distribute the source, but they don't have to give users the ability to run modified copies.
12:21:34preglowcan't say i really care too much about it, but i'd rather they didn't
12:21:35linuxstbWhich IMO is what the GPL is all about.
12:21:48 Quit krazykit (Read error: 110 (Connection timed out))
12:22:07B4gderthat's exactly where the FSF vs torvalds discussion starts ;-)
12:22:15ddaltonso what is the main function in a plugin?
12:23:07preglowgetting all the copyright holders to agree might be a problem...
12:23:13linuxstbI guess torvalds thinks of the gpl as being a license for developer freedom, rather than a license for user freedom.
12:23:24GodEaterddalton: plugin_start
12:23:27preglowlinuxstb: pretty much
12:23:31rasherlinuxstb: A license for code freedom, to be more exact, probably
12:23:32 Join Llorean [0] (
12:23:41ddaltonso is that like main in a normal c program?
12:23:49preglowhe's a couple of notches too pragmatic even for my taste
12:23:51GodEaterddalton: pretty much
12:24:15rasherpreglow: in some cases I agree, but not in the gplv3 one
12:24:24 Join My_Sic [0] (
12:24:35ddaltonok so if I have a program with lots of functions I can just put them in a plugin and rename main to the plugin start function?
12:24:47GodEaterddalton: not quite, see the wiki page I just linked
12:24:51ddaltonfor example a simple c program I might write on my computer.
12:25:04GodEaterplugin_start does not return an int for a start
12:25:38GodEaterit also doesn't take argc and argv as arguments
12:25:47ddaltonso can't plugins use scanf and printf?
12:26:03GodEaterddalton: where would they read from, and where would they write to
12:26:12bluebrotherrasher: I get an Error 141 when trying to build the voice using the perl script
12:26:46ddaltonwell the scanf function reads from the key boared normally and the printf writes to the screen.
12:26:46amiconnrasher: Hmm, so the perl script is useless for me atm
12:26:50rasherbluebrother: What's spewing that error? And under which conditions
12:27:13ddaltonso Would I have to change printf to the function to display text in plugins?
12:27:20rasheramiconn: True, I need someone (pondlife again, perhaps?) to reimplement the sapi building
12:27:21ddaltonand what would scanf be?
12:28:21ddaltonso if I have a program that use argc and argv how would I get around this when porting it to a plugin?
12:28:55preglowsounds to me like you're in over your head, my good sir
12:29:38linuxstbddalton: Plugins don't take any parameters, unless you make it a viewer, in which case it takes exactly one parameter - the filename to be viewed.
12:30:20bluebrotherhmm −− I reconfigured and now it seems stuck. Or doesn't it print the current phrase anymore?
12:31:11ddaltonok so how would I port a program that use argc and argv?
12:31:34rasherbluebrother: It doesn't print much, but there's a debug flag to
12:31:56bluebrotherrasher: ok, reconfigureing helped. But it's a bit surprising it's totally silent
12:31:59rasherbluebrother: add -v in the call to in the Makefile (should probably react to V=1 being set)
12:32:02linuxstbddalton: By not using argc and argv... You either hard-code the options, or make them available via a settings menu.
12:32:23rasherbluebrother: I found that printing everything was in fact slowing things down (at least for me, over ssh)
12:32:29ddaltonok then can plugins play music?
12:32:38ddaltonjust out of quriosity
12:33:01Domonokyyes, but only wav i think..
12:33:17ddaltonso they can play sound?
12:33:19bluebrotherhmm −− maybe just print a dot for each phrase? Or is it the printing itself?
12:33:32ddaltonif someone wrote a wma plugin could it play wma?
12:33:37preglowthey can both play and record sound
12:33:37bluebrotherddalton: look at doom −− it has sound
12:33:42rasherbluebrother: Probably the amount of text is the problem.. some sort of progress indication should be possible
12:33:46linuxstbddalton: Yes, lots of plugins play sound. But for now, forget about sound, just write the output to a wav file.
12:33:48preglowbut you need to provide the sound yourself, it doesn't do files, but that shouldn't a problem for you
12:33:54ddaltondoes it I am blind so never bothered but I will.
12:34:12ddaltonand then play the wav file or what?
12:34:52GodEaterlinuxstb: re: number of arguments, the wiki is wrong in that case
12:35:02preglowjhMikeS: yo?
12:36:37 Quit My_Sic ("Miranda IM! Smaller, Faster, Easier.")
12:36:54ddaltonso if I can write the text on the screen to a wav file I could then play the wav file and then delete it ?
12:37:53rasherddalton: you could, and should, play the data directly. Writing it to a file is not needed. But it's easier for a first step
12:38:07preglowrasher: i left a comment there, i'll apply it when it gets answered
12:38:41ddaltonso what I said before would take a lot of time?
12:39:34rasherddalton: Writing it to a file is the easiest thing for you to do as a start, but in the end you should just get the data from espeak and feed it directly to the sound output function
12:40:35rasherpreglow: What, where?
12:41:31markunddalton: did you see the PM I sent you with jonathan's email address?
12:41:41preglowrasher: the rounding patch
12:41:49linuxstbGodEater: Are you talking about parameters to plugins?
12:42:15rasherpreglow: Ah, well I'm not in any way involved in that, so if you'll excuse my surprise
12:44:05 Quit miepchen^schlaf (Read error: 104 (Connection reset by peer))
12:44:06preglowrasher: yeah, sorry, it was markun pasting it to me
12:44:19preglowyou have the same amount of letters in your nick :>
12:44:47B4gderthat should be disallowed!
12:44:48GodEaterlinuxstb: I was talking about the plugin_start() function
12:45:34 Join My_Sic [0] (
12:45:38rasherbluebrother: Did it produce a working voice file in the end?
12:46:44linuxstbGodEater: I think the wiki's fine. I wouldn't say the API struct is a parameter in the same way as argv/argc are.
12:47:37linuxstbB4gder: Going back to espeak, are you saying we can't use gplv3 code in Rockbox?
12:48:15rasherNot without going "gplv2 or later". I believe Rockbox is v2 only at the moment?
12:48:16B4gderlinuxstb: yes, that's my interpretation
12:48:22Lloreangplv2 and gplv3 aren't compatible.
12:48:44LloreanThere pretty much has to be a clause allowing the code to be used with earlier (or later) versions of the license.
12:49:07rasher"GPLv3-or-earlier" now there's a useless license
12:49:15linuxstbWhat a nightmare situation...
12:49:52Lloreanrasher: But Rockbox going GPLv2 or later wouldn't work. We'd have to go full-on v3 to be able to use v3 code, wouldn't we?
12:50:04linuxstbPresumably if Linux is staying GPLv2, anything from the IPL kernel will have to stay GPLv2 as well...
12:50:33linuxstb(unless of course we get the IPL devs to dual-license...)
12:50:34rasherI'm pretty sure v2 or later would work, since it'd then be GPLv3 in the context where used with v3-only code, but taking bits from Rockbox could be done under v2
12:50:50B4gderrasher: I don't think so
12:50:55B4gderwe use only one license
12:50:58bluebrotherrasher: voice is silent on target. No idea if I did something wrong, will recheck later
12:50:58LloreanHow did eSpeak move to v3?
12:51:05bluebrothergot to go now. bbl.
12:51:11LloreanNone of their code was licensed under v2 explicitly?
12:51:46LloreanI mean, you can't really rerelease under another license without all the copyright holders agreeing, can you?
12:52:07rasherB4gder: Yes, but I'm saying *if* Rockbox went "v2 or later", GPLv3 code could be used, afaiu and the codebase would be effectively v3 unless you picked out parts of it that were v2-or-later, in which case you could use that specific part under v2
12:52:10linuxstbLlorean: Either 1) The original license said "v2 or later", or 2) The author wrote it all himself
12:52:23rasherLlorean: A relevant quote: Linus Torvalds deciding not to take his 10,000 copyright holders and go v3 is like the driver of an 18-wheeler "deciding" not to do donuts on my lawn.
12:52:46rasherNot quite as big a task with Rockbox of course.
12:53:13linuxstbLlorean: Checking an older version of espeak, it did indeed say "v2 or later"
12:53:21B4gderrasher: I'm saying that if rockbox changed to "v2 or later" it wouldn't matter since we'd have to go to v3 before we can use code from another v3 project
12:53:44 Join Weiss [0] (
12:53:49preglowany idea how many different copyright holders we have?
12:53:55B4gderv2 or later is just an option to change license, it is still v2
12:53:56Lloreanlinuxstb: It's now explicitly only v3 it seems.
12:54:20linuxstbLlorean: "v3 or later" in the files I'm looking at
12:54:27rasherB4gder: But v2 or later means that the whole codebase *can* be used under v3 - in that case, why couldn't we use v3 code?
12:54:32LloreanThe license posted at their page doesn't include the "or later".
12:54:40ddaltonis anyone working on porting espeak yet or do I need to start the whole thing from scratch?
12:54:41linuxstbLlorean: I'm looking at the source files
12:54:41LloreanAt least, not that I spotted. But the files are what matters, I guess.
12:55:00GodEaterddalton: I don't think anyone has started on it yet
12:55:10B4gderrasher: because it is still licensed as v2 which isn't compatible
12:55:20ddaltonok I will try and make a start tomorrow
12:55:31Lloreanrasher: Because if we use V3 code, we have to distribute the whole codebase as V3, so that they can link with it. That specific distribution has to be v3.
12:55:32linuxstbddalton: Are you paying attention to this license conversation? Basically we can't port the current version of espeak - you'll need to go back to an older version.
12:56:03GodEaterlinuxstb: he can port and use it himself can't he? He just can't get it commited to the rockbox codebase and he can't redistribute it ?
12:56:06Lloreanpreglow: Well, look at Credits for a start. ;)
12:56:11rasherLlorean: Yes, but we can still distribute the non-v3 parts as v2-or-later, as far as I can tell. Although I'm having a hard time argumenting for it
12:56:22rasherMight be because I'm wrong of course.
12:56:43Lloreanpreglow: Good luck convincing id software to rerelease Doom under GPLv3 for example, I don't think they could be bothered. ;)
12:56:45B4gderrasher: since v2 and v3 is not compatible, we can't ship a mixed one
12:57:05Lloreanrasher: You could make all the "v2 or later" code separately available under v2, yes, but the actual distribution would be v3.
12:57:29linuxstbGodEater: Yes, but what's the point?
12:57:48GodEaterlinuxstb: it's of great use to him ? :)
12:57:49preglowLlorean: very good point, we have tons of other libs...
12:57:57linuxstbChecking its SVN, espeak changed the license on 26 June 2007.
12:58:23rasherLlorean: I don't see why there has to be a separate distribution. Just a distribution where some code is under both licenses, and some is strictly gplv3. But I think I'll stop here before I get frustrated.
12:58:26linuxstb(svn revision 56). So svn r55 would be the last we could include.
12:58:56B4gderrasher: because v2 and v3 are not compatible, v2 code cannot be mixed with v3
12:58:57rasher shall be my last word on this
12:59:05Lloreanrasher: The GPLv3 doesn't allow for that, though. That's the problem, License A can say "We accept License B code" but License B also has to say "We accept License A code"
12:59:14linuxstbLlorean: Doom is "v2 or later"...
12:59:15preglowhmm, if rockbox itself goes v3, we can use any codec libs that are "v2 or later", yeah?
12:59:20Lloreanlinuxstb: Ah, well bad example then.
12:59:24preglowor not...
12:59:25 Quit haemmy ()
12:59:33B4gderpreglow: yes sure, since we'd then license them as v3
12:59:44B4gderrelicense even
12:59:46linuxstbpreglow: Most codecs aren't gpl'd though...
13:00:04preglowsome are
13:00:07preglowlibmad, for example
13:00:09preglowand a52
13:00:11rasherLlorean: But GPLv3 accepts "GPLv2 or later" since it's effectively also GPLv3
13:00:14preglowbut both are v2 or later
13:00:21B4gderrasher: I disagree with that
13:00:25B4gderit isn't
13:00:35B4gderit is v2, with an option to relicense
13:00:36rasherWell, GNU doesn't, as far as I can tell
13:00:40linuxstbpreglow: So we don't have a problem with codecs at least.
13:00:48B4gderrasher: what makes you say that?
13:00:57rasherB4gder: that FAQ I linked
13:01:02preglowbut doom is a loadable plugin and not really part of rockbox itself, is it a problem to distribute that as gplv2?
13:01:04B4gderI read it now, I can't see that stated
13:01:28Lloreanrasher: "GPLv3 is one of the options it permits." means that you can distribute it *as* v3
13:01:35preglowdoom is v2 or later too
13:01:58rasherYes.. which is my entire point. Right, I'm not going to discuss this anymore. Really, this time.
13:02:48B4gderthe whole argument from FSF is that they are compatible since "v2 or later" means you can relicense it to v3 and then it works with other v3
13:02:55LloreanIs this an extension of the tivo-ised Rockbox discussion of earlier, or just simply an offshoot of espeak?
13:02:57*linuxstb can't see any (C) info in the rockboy source...
13:03:03preglowhow nice
13:03:26ddaltonso if I port something to rockbox what should I start off with. Because this espeak project looks like it is going to be hard.
13:04:07rasherB4gder: (argh) But the -or-later license is "at your option", not at the copyright owner's option, meaning the "relicensing" can happen implicitly
13:04:30amiconnHmm, I am no license expert at all. Iiuc if we incorporate gplv3 code, the whole rockbox distribution would be under gplv3, however, that should not cause problems with gplv2 code as long as it includes the "or later" clause
13:04:32B4gdereh, implicit or not, it is a relicense
13:04:42B4gdersince you can't then go back to v2
13:05:12Lloreanlinuxstb: gnuboy seems to have been v2 or later
13:05:21B4gderamiconn: I don't see how that could work
13:05:32rasherOf course you can, but only for the parts that were originally GPLv2 or later. It's effectively dual-licensing with "any license the FSF can come up with" as an option
13:05:34B4gderbut I
13:05:37B4gderwill stop
13:06:12preglowlinuxstb: is all that code in metadata/ actually written by you, or did you just paste some copyright header?
13:06:15 Quit My_Sic (Read error: 104 (Connection reset by peer))
13:06:15amiconnB4gder: Well, rockbox incorporates code which is not gpl licensed, but under various gpl compatible licenses
13:06:21preglowoh no, there's one by david bryant :>
13:06:27linuxstbCan I just ask if anyone here with (C) in any Rockbox code would object to going gplv3 ?
13:06:39B4gderamiconn: yes but they don't have any further restrictions
13:06:40XavierGrlicense talk is furstrating indeed, I didn't understand a thing of the whole discussion :P
13:06:41B4gdergplv3 has
13:06:42preglowi don't think i would, but i'd have to read a bit more about it
13:06:46LloreanI haven't read the gplv3 yet.
13:06:47amiconnThat makes the code fall under gpl when it comes with rockbox, but the individual code retains its license
13:07:05B4gderthey don't have further restrictions!
13:07:13amiconnYes, and?
13:07:20GodEaterI wouldn't be keen on going to GPLv3 - but only based on the discussions around it I've seen, not because I've actually read and grok'd it
13:07:22B4gderyou cannot ship v2 code with anythin with further restrictions
13:07:29linuxstbpreglow: I think I wrote the original metadata.c, but that was then split by markun into lots of separate .c files - so the (C) is probably inaccurate. I wrote some of them from scratch though - Ogg, FLAC and APE IIRC.
13:07:30amiconnI can't see the difference. gplv2 is a gplv3 compatible license, isn't it?
13:07:33linuxstb(plus ASF)
13:07:35B4gderamiconn: no
13:07:43Lloreanamiconn: That's the big problem, they aren't compabitle.
13:07:55LloreanExplicitly not so, in fact. They make no claims that they are.
13:07:56XavierGrI would just follow the others, if the major developers want to change license then sure
13:07:57preglowyeah, i'm pretty sure i wrote a couple of them, and they're now copyright you
13:07:57B4gder"Is GPLv3 compatible with GPLv2?"
13:08:09amiconnAny specific reasoning?
13:08:09preglowthat would be the wrong window
13:08:23amiconnThat directly contradicts the gplv2 license text....
13:08:24B4gderamiconn: yes, let me repeat
13:08:24linuxstbpreglow: You should correct the (C) if you feel the code is yours.
13:08:26rasheramiconn: v3 adds further restrictions over v2. v2 doesn't allow this.
13:09:09amiconnSo what about the "...or later" clause in gplv2?
13:09:14rasherlinuxstb: Ignoring that I don't have any substantial contributions, if there was a very good reason for it, I wouldn't object to going v3. But I'd rather that it didn't get to it. As I said already, I'm very much with Torvalds on thhis
13:09:27B4gderamiconn: that allows them to relicense to the v3 license
13:09:30rasheramiconn: There is no "or later" clause in gplv2.
13:09:31B4gderand then they are compatible
13:09:55rasherThat's something they suggest authors add in their license terms.
13:10:00 Join miepchen^schlaf [0] (
13:10:06linuxstbrasher: Ignoring the philosophical issue, it seems more practical for Rockbox to be gplv3. At least if a lot of other projects go that way.
13:10:48rasherlinuxstb: Yeah, *if* they do that. And then, that'd be a good enough reason. But I still wouldn't be happy about it.
13:11:03amiconnB4gder: Not in gplv2 itself, but in the license statement in our source headers
13:11:10preglowall i can remember writing is the mpc.c one
13:11:22rasheramiconn: Rockbox headers say gplv2 or later?
13:11:33preglowrasher: why not happy about it?
13:11:58rasherpreglow: Because I don't like the extra restrictions of v3.
13:11:59B4gderamiconn: even if it would say so, it says it can be distributed under those licenses and then it have to be v3 to get used with v3 code
13:12:03B4gderthen it isn't v2
13:12:31B4gdersince v2 isn't compatible with v3
13:12:31rasherB4gder: But it's still v2-or-later. It's a dual-license.
13:12:35amiconnHmm, indeed they don't, but they also don't specify a version. Just GPL
13:12:43rasheramiconn: badness
13:13:04amiconn...and then they refer to docs/COPYING
13:13:09rasherAh, then no badness
13:13:14amiconn...which is a verbatim copy of gplv2 atm
13:14:39*amiconn wouldn't object going gplv3
13:14:56jhMikeSpreglow: here. what's up?
13:15:37XavierGrit seems to me that to bring all copyright holders of Rockbox and agree about it is a very very long shot...
13:15:49LloreanHow does someone tivo-ize a piece of software under GPLv2 anyway?
13:16:00B4gderLlorean: signature check in the bootloader
13:16:01LloreanI would think "plus the scripts used to control compilation and installation of the executable." would include anything necessary to sign the binary so it can be installed.
13:16:07linuxstbLlorean: Install it on hardware where the bootloader will only run signed binaries... (what B4gder said)
13:16:23rasherLlorean: You can install it all you want. Doesn't matter if the hardware won't load it
13:16:28B4gderI believe that's exactly what tivo does
13:16:41linuxstbLlorean: Yes, that would be my interpretation of the gplv2 as well. But it seems it isn't...
13:16:52LloreanIt's not the same executable, then, is it?
13:16:58LloreanI dunno
13:17:15zeseems like it'd be lame if some company made signed builds of rockbox that're all their player would run
13:17:19preglowjhMikeS: you sure the rounding you do the in cf asm in the correct rounding patch is ok?
13:17:44B4gderLlorean: the bootloader of course doesn't have to be gpl
13:18:13B4gderit can still load, check and start linux/rockbox
13:18:24jhMikeSpreglow: I really didn't concern myself too much with it. I asked if that was acceptable and just quickly implemented that. I even asked the same question.
13:18:25rasherze: Certainly very lame and stupid, but in my eyes it's still a net benefit to everyone. The company gets code to build on. Customers get good code (but lack freedom). Rockbox proper gets useful any improvements back.
13:18:34LloreanB4gder: True. I suppose to me "installation" means "make ready to use"
13:18:45LloreanI wonder if there's an established legal definition of it somewhere.
13:18:52zerasher: true
13:18:55preglowjhMikeS: hmm, ok, i'll have a look before i commit it anyway
13:19:11B4gderLlorean: this was debated _at length_ on the kernel ml the other week
13:19:22LloreanI wouldn't mind Rockbox being Tivo-ized simply because there's plenty of darn good hardware it already runs on
13:19:22B4gderif private keys should be considered part of the code etc
13:19:43B4gdereven the lawyers around can't agree on that
13:19:49linuxstbrasher: Speaking as a developer, I see it as making my code non-free. i.e. it's using the code in a context where I'm not able to modify it.
13:19:54B4gderand there's no court case to follow
13:19:55LloreanB4gder: I don't think private keys necessarily are, but I do think that "installation" means that the end result must actually run.
13:20:04rasherlinuxstb: But the code is still free. The hardware is not.
13:20:07 Join djtremolo [0] (n=c1b937fd@
13:20:16B4gderLlorean: the gpl doesn't say it has to
13:20:21linuxstbrasher: But code is useless without hardware. And I bought the hardware...
13:20:21jhMikeSpreglow: I think adding a constant like that will just dc-offset in the positive direction -.5->1, .5->0 instead of -.5->-1, .5->0.
13:20:36LloreanIn regards to Rockbox, denying your customers the option to use the most current builds provided by us is, I think, more likely just going to suggest to them they should buy a player supported by us next.
13:20:55preglowjhMikeS: it's basically just round to nearest instead of round towards negative infinity
13:21:06linuxstbrasher: We're not going to agree ;)
13:21:08jhMikeS make that -.5 ->-1
13:21:15preglowjhMikeS: but your code seems to use a constant for it, which i can't figure out how works, since it needs to be calculated from the codec precision in the c code
13:21:21B4gderin the tivo case, linux has been improved by them
13:21:25rasherlinuxstb: It's a matter of preference. In my eyes the code is perfectly useful without the hardware. Rockbox proper would possibly get drivers for new hardware, etc
13:21:31*jhMikeS just got up
13:21:31XavierGrvery informative now I think I can grasp some of it
13:21:34LloreanB4gder: Yeah, the GPLv2 unfortunately uses the words "all scripts", you can include the scripts to sign it, without the key it needs to be signed with.
13:22:08jhMikeSpreglow: it's done _after_ doing the multiply to fill the 32-bit word.
13:22:36djtremolosorry for disturbing.. I was looking for current build rombox.ucl, but the zip doesn't include it. Am I looking in wrong place?
13:22:49linuxstbdjtremolo: For which device?
13:23:02djtremololinuxstb: recorder, v1
13:23:07preglowjhMikeS: anywho, i'll have a look when i get back, before commiting it
13:23:11jhMikeSaccumulator is initially 0x8000. the sample is shifted and clipped to have a range 0x80000000 to 0x7fffffff based on the dsp
13:23:15 Join Rincewind [0] (
13:23:30rasherlinuxstb: Anyway, yes, it's unlikely people will suddenly change their mind about this. Let's just agree to disagree, and wait it out to see if rockbox needs to go v3 in the future.
13:23:38amiconndjtremolo: Rockbox has grown too much to be able to run directly from ROM, so rombox is no longer available, unfortunately
13:24:09amiconnYou can still boot rockbox from flash using rockbox.ucl, which is decompressed into RAM
13:24:30amiconnJust 2 targets still have rombox :|
13:24:38djtremoloamiconn: ok.. sad, but thanks for the info!
13:25:13djtremoloamiconn: yep, that's the way i'm doing now.. I just was so impressed when I found the BootBox and Rombox stuff.
13:26:36*preglow out
13:26:52jhMikeSamiconn: testing showed you can expect 10-12 context switches per second (default settings) on most devices. the pp i2c driver inflates it because of the yield loop and being the only thread running.
13:28:37jhMikeSqueue_send/reply takes 19.55uS CPU/COP, and 27.68us CPU/CPU
13:28:45jhMikeSon e200 at 80MHz
13:29:04 Quit djtremolo ("CGI:IRC 0.5.7 (2005/06/19)")
13:32:51 Quit miepchen^schlaf (Read error: 104 (Connection reset by peer))
13:40:56 Join bdgraue [0] (
13:42:47 Join bascule [0] (i=86920006@gateway/web/cgi-irc/
13:44:41 Join My_Sic [0] (
14:00:21 Quit jhMikeS (Nick collision from services.)
14:00:25 Quit Pamacii ("Ex-Chat")
14:00:27 Join jhMikeS [0] (n=jethead7@rockbox/developer/jhMikeS)
14:02:00 Join midgey [0] (
14:02:55 Quit ddalton ("I was using BOFHNet IRC version 1.2 by fmillion - get your copy today from !")
14:03:39linuxstbjhMikeS: Have you looked at this mpegplayer work?
14:05:06***Saving seen data "./dancer.seen"
14:05:42jhMikeSlinuxstb: I think he's the one I consulted with about it
14:07:00linuxstbWhat do you mean?
14:08:00 Quit pondlife ("disconnected has pondlife")
14:08:15jhMikeSoh, no. someone else was working on the same thing (jim_h)
14:08:44 Quit pill (Nick collision from services.)
14:09:03 Join pill [0] (
14:09:40 Join pondlife [0] (
14:11:22jhMikeSI do know how to stop the prefecth aborts there too. I guess that bugfix will come with dual core. :) I really think I'll just make that so it can be enabled for a player individually. It can be switched on when someone verifies a particular device is ok.
14:11:38 Quit pondlife (Client Quit)
14:14:20 Quit aliask ("ChatZilla [Firefox]")
14:16:15 Quit pill (Nick collision from services.)
14:16:33 Join _pill [0] (
14:20:06 Join kubiixaka [0] (
14:20:10rasheramiconn: looks like I have sapi5 building working now
14:20:18rasheramiconn: Will upload a new patch in a moment
14:20:59 Quit midgey ()
14:24:30rasherStill not terribly fast, but you could in theory run it with activestate perl rather than cygwin perl provided you do some manual legwork. That should give a nice speedup
14:27:37 Part Neovanglist
14:28:08 Quit webguest07 ("CGI:IRC (EOF)")
14:31:41 Part Llorean
14:32:00 Join Llorean [0] (n=llorean@rockbox/administrator/Llorean)
14:33:45amiconnrasher: I think that implementing a client/server model for sapi voice building will give further speedup (provided you didn't already do that)
14:34:30amiconnjhMikeS: Does it even work on one of the PP5020 targets yet?
14:34:32rasheramiconn: I didn't. I'll leave that to someone else, because a) I don't really know the details and b) working with cygwin really gives me the creeps
14:35:28rasheramiconn: I did however make some improvements that should make it easier (the tts_init function returns an object that the "generate a voiceclip" function can use however it likes)
14:35:30amiconnI know that it crashes the H10/small, but does it work on H10/big, ipod G4, ipod color or mini G1?
14:36:32rasherAnd of course, being perl, everything happens in a single process, so passing stuff around is actually possible
14:36:36 Join freqmod [0] (
14:36:36 Quit bascule ("CGI:IRC (EOF)")
14:37:31amiconnrasher: Huh, does (non-activestate) perl allow handling the sapi objects directly?
14:38:44rasheramiconn: I don't know how it's possible to communicate with SAPI, so I can't answer that
14:39:00rasherBesides calling a .vbs script continually
14:39:24amiconnYeah, then the sapi cleint/server thing needs to be handled in the vbscripts, not in perl
14:39:26jhMikeSamiconn: I haven't had a clean run on those. Other 502x seem to be fine. I'm just thinking about having a way to enable the extensions per player. With it in SVN, someone might be curious to enable them and figure out the problem on their device.
14:39:33 Join ie [0] (i=c27f0814@gateway/web/cgi-irc/
14:39:52amiconnI was explicitly asking for 5020, not 502x with x != 0
14:40:00amiconnI know that it works on those
14:40:16rasheramiconn: Still, this should be easier to do with the new script than with the old one. I did also see a SAPI perl module mentioned, but I don't know if that's preferable.
14:40:18jhMikeSok, and I believe I've answer that with "no clean run" :)
14:40:42ieSpeaking about copyright, licences etc: some .bdf files (fonts) don't have a copyright statement. I thought every piece of work that is released under GPL should have one.
14:41:06rasherie: The bdfs aren't necessarily gpl. In fact, I'd bet that most aren't
14:41:10amiconnIf it works on the latter but not on PP5020, that would be an indication that there is still something fundamentally wrong, and I would like to know what that is
14:41:21Lloreanie: For example, profont has a distribute but no-modify license, I believe.
14:41:44rasherie: Check the svn history for details - and submit patches to add copyright statements
14:42:08jhMikeSamiconn: and "fundamental" meaning?
14:43:14ierasher: but are they considered part of RB? If yes, their licence's compatibility with GPL should be verified.
14:43:21linuxstbie: IIUC, the fonts are data used by Rockbox so technically don't have to be GPL'd. But they need to be at least freely redistributable, and yes, should have a copyright and license statement.
14:43:24jhMikeSI wonder if swp(b) even works right on those.
14:43:32amiconnCould be two things. Either cache handling, or some of the new code
14:44:12rasherie: They're not linked into Rockbox (except rockbox_default, or whatever its name is), so they don't have to be GPL compatible
14:44:29LinusNthey are even distributed separately
14:44:29amiconnlinuxstb: I agree that fonts are just data and don't need to be gpl'd, except probably for the built-in font
14:44:40rasherLinusN: Also a good point
14:45:15LinusNnevertheless, i think we need to be clear about which license each font is distributed under
14:45:30amiconnjhMikeS: Do you have test code that would tell?
14:45:35jhMikeSIf it were the new code, I'd expect a quick death on those queue tests. It just runs and runs.
14:46:08ieLinusN: yes, that too. Ok then. Some have the following inside: "Public domain font. Share and enjoy." Is that ok for RB?
14:46:22 Join webguest54 [0] (i=c023110b@gateway/web/cgi-irc/
14:46:28jhMikeSI even straightened out the order of operations in the cache test and that runs without incident now.
14:46:37amiconnOn 5020?
14:47:04 Quit My_Sic (Read error: 104 (Connection reset by peer))
14:47:11rasherie, LinusN: Ideally we'd have some sort of confirmation that the font is indeed public domain?
14:47:30rasherOr at least a note of where we got it from
14:47:46jhMikeSno. that really seems to be something outside an error in locking things. perhaps something similar to have to do inl in the PLL routine.
14:48:07jhMikeSmaybe it needs noops in the swp(b) loops. who knows.
14:48:21amiconnThat's why I was asking for test code...
14:49:16jhMikeSI'll will have a general update soon that simplifies thing. If you want the plugins that do queue send accross cores, I'll pastebin them,
14:49:57 Join My_Sic [0] (
14:50:40amiconnjhMikeS: Hmm, as they probably need a patched core, and that patched core crashes, they wouldn't be useful I presume
14:51:26rasheramiconn: updated with SAPI5 building
14:52:22rasher(and a bit of other stuff)
14:54:21 Quit chrisjs169 (Connection timed out)
14:54:26amiconnThe Win32::SAPI5 module looks promising...
14:54:34jhMikeSamiconn: yes they do as they are. though some local faking of it could be done. anyway:
14:55:04LinusNamiconn: there is a sapi4 module too
14:55:10 Quit Thundercloud (Remote closed the connection)
14:55:12ierasher: we could check who initially checked in the fonts and ask the people where they got that from.
14:55:32amiconnLinusN: Yes, but sapi4 is a lot less interesting than sapi5. sapi4 can only be scripted in realtime
14:55:48amiconn(unless I missed something fundamental all the time)
14:55:50rasherSome people are requesting SAPI4 though.
14:55:56 Quit webguest54 ("CGI:IRC (Ping timeout)")
14:55:59jhMikeSyes, one creates the thread on CPU but that was for the benchmarking
14:56:18amiconnrasher: Well, if they can live with one voice needing half an hour to build, it would work
14:56:19LinusNamiconn: well, sapi5 with today's script is almost slower than realtime :-)
14:56:24 Join chrisjs169 [0] (
14:57:54 Part LinusN
14:58:42jhMikeSA test with a locked variable could be: swap busy value into i. increment the local copy. unlock by writing new value. each core does while((value = swap(x)) != BUSY_VALUE; value++; x = value;
14:59:13jhMikeSswap(x, BUSY_VALUE) rather
14:59:55rasherAh, at least using a voiceclip pool does speed things up nicely (it was still rather slow on cygwin previously)
15:00:30jhMikeSblah, and check == , not !=
15:01:24LloreanIf I want to create a human-voiced .voice file, and I populate the pool myself, what merges them into the .voice? Is there a way to do this currently?
15:01:55rasherLlorean: I have a script you could use for that
15:02:33LloreanI was curious if there was anything in SVN for that at the moment. I don't really intend to speak 30 minutes of audio myself at this point. :)
15:02:59rasherIt expects a dir full of .wavs named after their contents (one named "No.wav" etc.) and will generate voicefiles for all targets, and tell you if you're missing any voice-strings
15:03:21rasherBut there's not really anything like it in SVN
15:03:42 Quit My_Sic (Read error: 104 (Connection reset by peer))
15:03:50LloreanEven just something telling you if you're missing strings would be nice.
15:03:52rasherunless you count running voicefont directly in a dir full of mp3s named after clip-ids
15:05:00rasherLlorean: In the language file?
15:05:07rasherLlorean: or in a clip pool
15:05:11Lloreanclip pool
15:05:17amiconnI know that it's posible to install cpan perl modules using perl, but I don't remember how...
15:05:23LloreanRather than printing out a copy of the .lang, and checking them off, that is.
15:05:48rasherLlorean: Well, that's the first part of my script. It generates a file of all strings to be recorded
15:06:08rasherAnd if no strings are missing, it goes on to generate the voicefiles
15:06:13Lloreanrasher: Do you think this might be useful enough to some people to go in SVN? Especially with the fact that it generates voices for all targets?
15:06:21LloreanIn /tools or such?
15:07:15rasherPerhaps, I'll put it in a tracker entry for now
15:07:20amiconnNote to self: google is your friend
15:10:38amiconnhmm :\
15:12:15 Quit SirFunk (Remote closed the connection)
15:13:38rasherLlorean: FS #7650
15:16:16amiconnEh? doesn't work...
15:16:37 Join My_Sic [0] (
15:16:48rasheramiconn: Works For Me[tm]
15:16:55 Quit My_Sic (Read error: 104 (Connection reset by peer))
15:17:09 Join My_Sic [0] (
15:17:11*amiconn just gets a timeout :(
15:17:48peturworks here too
15:18:16amiconnBah, dns caching
15:18:28 Join H10_007quick [0] (
15:18:35amiconnFlushing the cache helped. resolves to a dyndns address....
15:18:50 Quit H10_007quick (Client Quit)
15:19:46amiconnWin32::SAPI5 doesn't install on cygwin perl :(
15:19:51 Join _jz [0] (
15:20:11amiconn"Win32::SAPI5 can only be installed on Microsoft Windows Platforms" <= haha
15:20:24rasherWell that's certainly a bummer
15:20:49amiconnSo we have to stick to the vbscript if we don't want to require activestate perl
15:20:52GodEaterActivePerl only huh ?
15:21:13rasheramiconn: or we could use the current method
15:21:28rasherah right, that's the vbscript, ignore me
15:21:45linuxstbWhat about a C application? Would that be hard?
15:22:06*linuxstb has no idea what's involved
15:22:39amiconnWould be way more complicated than using a script
15:23:05amiconnOtoh, it would make sense for sapi4, as the C api allows to set higher speeds than realtime (x2, x4 and x8)
15:23:07linuxstbBut easier for users, and easy to incorporate into rbutil...
15:23:49amiconnWell, the vbscript isn't difficult. The scripting host ships standard with every half-recent windows version (2000 and up)
15:24:03amiconnAnd sapi would be needed anyway
15:24:23 Join pondlife [0] (
15:24:31rasherAdding voice-building to rbutil would be fairly involved building with sapi wouldn't be quite as bad I guess
15:25:54 Quit perrikwp ("ChatZilla [Firefox]")
15:26:01Domonokyrasher: .talk file building would be nice with sapi in rbutil.. at the moment it only supports espeak and flite.. voice building would be hard..
15:26:38rasherDomonoky: It'd make it difficult/impossible? to crosscompile rbutil though I imagine? Unless you can use standard windows apis for it
15:27:21rasherVoice building requires the langfiles, features and voicefont (the latter could of course be assimilated into rbutil)
15:27:41rasherfeatures meaning the list of features the target supports
15:27:52*Domonoky thinks the best thing to do with sapi and rbutil would be to use little scripts or exes which generate a wav file with sapi.. like it uses espeak and flite now..
15:28:27rasherDomonoky: Much like (and does it
15:28:32rasheror rather, exactly like that
15:28:42amiconnThat's what the vbscript does, but it does it in an inefficient way atm
15:28:56amiconnrasher: Is it possible to start a process asynchronously from perl?
15:29:03rasheramiconn: I'm still not entirely sure that's the bottleneck
15:29:10rasheramiconn: In which way do you mean?
15:29:19Domonokyrasher: i used this script to build the talk file generation in rbutil :-)
15:30:29amiconnWell, init_tts should start an asynchronous cscript process with the master script. voice_string() would then just pass the string to convert to that "server" script
15:30:55amiconn...and shutdown_tts() would tell the server script to terminate
15:31:20rasheramiconn: Ah yes, that shouldn't be difficult. If you can make a vbscript that reads input on stdin, you'd even only need a single process, not any client/server business
15:31:56amiconnSure vbscript can read from stdin, that's simple
15:32:00rasherStart a process in init_tts, and feed it lines of text to speak and where to save it
15:32:16rasherThe filehandles can be passed along
15:32:21rasherAnd closed in shutdown_tts()
15:32:21amiconnYes, but that process needs to be started asynchronous
15:32:34rasherAh, well that's perfectly possible
15:32:48amiconn...because the perl script needs to continue while the vbscript is waiting for input
15:32:48B4gderopen(F, "|process"), print F "data\n";
15:33:22 Join miepchen^schlaf [0] (
15:33:25B4gderwell, it can be started as a process or just a thread
15:33:30B4gderin perl I mean
15:34:09rasherThat's how we use espeak at the moment.
15:34:15rasherAlthough that gets terminated for each clip
15:34:16B4gderbut I think the open approach should work, since I bet the script will wait for the completion anyway
15:34:17 Join seablue [0] (
15:34:46amiconnHmm, it looks like that's the same method as used for festival...
15:35:07amiconnI guess a newline would be sufficient to separate clips?
15:35:21rasheramiconn: It shouldn't even need to be as complicated as festival
15:35:32rasheramiconn: alternating lines of content and filenames perhaps?
15:35:46amiconnYes, as we are able to control the server side of the messages as well
15:36:03amiconnHowever, we need to convert from utf8 to system codepage in perl...
15:36:16rasherThat should be completely possible
15:36:28rasherCouldn't the vbscript do it better though?
15:37:00rasherI'm assuming it has easier access to such information as "what is the syste codepage?"
15:37:15amiconnWScript.StdIn is the stdin TextStream object in vbscript
15:37:52amiconnThe vbscript could tell the system codepage, but there are no conversion functions we could use.
15:38:04rasheramiconn: I'll add another method that opens a process and writes alternating lines of filename and string to speak. I'll leave implementing the vbscript up to someone else
15:38:09amiconnAt least I didn't find any back when we switched to langv2 (first step) and utf8
15:38:35rasherwe could get that information when starting the script
15:38:37amiconnrasher: We could also separate the data with a tab or something
15:39:07rasheramiconn: I think a newline is better - tabs just may sneak their way into a string
15:39:18rasherBesides, isn't it easier to read a line at a time than to have to split?
15:39:24amiconnHmm, but newlines make it ambiguous
15:39:54rasherSurely you'll just read two lines at a time and know that the first is a filename and the next is a string
15:40:08rasherWe control the input, so I don't see how that could go wrong
15:40:14amiconnThat'd work - if the passing script behaves
15:40:43amiconnThe one-line-at-a-time needs the controlling script to pass <cr><lf> instead of just <lf>
15:40:56 Quit freqmod (Remote closed the connection)
15:41:03rasherWill do
15:41:41amiconnThen it's just a matter of text = WScript.StdIn.ReadLine in the vbscript
15:42:00rasherMaybe I can even do it myself
15:42:09amiconnSo, how to detect termination? Passing two empty lines?
15:42:37rasherThat should do
15:42:59rasherThe string *could* be empty during the strings, but if the filename can't be
15:43:45 Quit frederv (Remote closed the connection)
15:44:26amiconnSounds good. Iirc genlang was enabled to do the utf8->system conversion when langv2 was first introduced
15:45:13amiconnSo that mechanism could be incorporated in
15:45:27rasherIf it's possible to find the system codepage in perl, that'd be better (both reading and writing to a process is far more annoying than simply reading from one)
15:45:43linuxstbWhy is the system codepage needed?
15:45:57amiconnBecause sapi can only use that
15:45:58rasherlinuxstb: I assume SAPI5 expects input in the system codepage
15:46:43amiconnThere might be unicode interfaces to sapi, but those aren't usable from vbscript
15:47:36 Join pixelma [0] (i=d5494c8e@rockbox/staff/pixelma)
15:47:55 Join Thundercloud [0] (
15:48:06linuxstbWhat's the complication that makes a scripting language preferable to C?
15:48:53 Join Gursikh [0] (i=khalsa@tremulous/officialdevannoyer/khalsa)
15:49:08rasherlinuxstb: which part, sapi clip generation, or voice generation in general?
15:49:11 Part Gursikh
15:49:18linuxstbrasher: The SAPI parts.
15:49:39amiconnDid you ever work with the windows COM api?
15:50:15linuxstbSo that's the complication? The API is just much easier from vbscript?
15:50:41 Join tyrion_ [0] (n=SkyCon@
15:53:28tyrion_hey all, quick question, i want to the tagnavi_custom.config file to allow me to sort old podcasts acording to folder, then date, at the moment they are sorted by length
15:53:48tyrion_sorry, new, unplayed podcasts
15:56:46pixelmaI think you need to enable the gather runtime feature but I must admit that I don't know much about it as I don't use it. The DataBase page in the wiki has an example of a tagnavi_custom file with a similar option - maybe you'll find some more info there
15:58:20tyrion_pixelma: yeah, i have that enabled, and it seperates played and unplayed podcasts
15:58:38tyrion_but the sorting method for unplayed podcasts is unuseable
15:59:21 Join BobShield [0] (
16:00:52pixelmayou can modify that to your own likings, not everything is possible but a lot. The mentioned wiki page tells you how, other than that maybe you'll find more database experts in the forums, I believe
16:01:33tyrion_ok, cool, thanks
16:01:40 Part tyrion_
16:05:07***Saving seen data "./dancer.seen"
16:05:31 Quit bdgraue (Remote closed the connection)
16:05:58 Join kaaloo [0] (
16:06:10 Part kaaloo
16:07:08 Quit Zagor ("Client exiting")
16:07:30*amiconn should concentrate on one thing at a time
16:07:59 Join freqmod [0] (
16:08:18 Join frederv [0] (
16:09:59 Quit B4gder ("It is time to say MOOO")
16:12:37 Quit jhMikeS (Nick collision from services.)
16:12:43 Join jhMikeS [0] (n=jethead7@rockbox/developer/jhMikeS)
16:14:38 Join einhirn [0] (
16:25:37 Quit BigMac (Read error: 110 (Connection timed out))
16:26:47 Quit einhirn (Read error: 104 (Connection reset by peer))
16:36:58 Quit tvelocity ("Αποχώρησε")
16:43:10 Join bdgraue [0] (
16:44:16 Quit pixelma ("bbl")
16:47:10 Quit My_Sic (Read error: 104 (Connection reset by peer))
16:47:39 Join jgarvey [0] (
16:49:57 Nick _pill is now known as pill (
16:50:09 Quit pill (Nick collision from services.)
16:50:29 Join _pill [0] (
16:51:44rasheramiconn: Added a script that does what we discussed and updated to use this script. Hopefully without making any crucial mistakes
16:54:22 Join saratoga [0] (i=98039a9c@gateway/web/cgi-irc/
16:54:24 Quit chrisjs169 (Read error: 110 (Connection timed out))
16:54:40saratogais the GPLv3 compatable with the LGPL?
16:59:02 Quit freqmod (Remote closed the connection)
17:03:44 Quit petur ("Drink error: 104 (Connection reset by beer)")
17:13:46ieAny news on commit status of FS #7627? It's a good thing imho
17:16:11 Part ie
17:23:13 Join darkraven [0] (
17:24:46 Quit pondlife ("disconnected has pondlife")
17:26:53 Quit darkraven (Read error: 104 (Connection reset by peer))
17:28:16 Join darkraven [0] (
17:31:10GodEaterlord I'm bored
17:32:43rasherGodEater: test my voice-building patch! (FS #7646)
17:33:07*GodEater goes to look at it
17:33:13 Join darkraven90 [0] (
17:33:27rasherIt should work more or less like it does now
17:33:52rasherExcept better
17:34:26*GodEater thinks we should encourage people to name the files they attach to their FS entries with the FS number included
17:34:55rasherGodEater: How'd you do that when you submit a new entry?
17:35:01rasherFlySpray could do it otoh
17:35:50GodEatertrue, it's tricky for new entries
17:36:00GodEaterbut I've got a plethora of .diff and .patch files
17:36:04GodEaterand no way to tie them back to FS
17:36:14GodEaterwell configure still gets stuck
17:36:27GodEaterit displays "Voice build selected"
17:36:30GodEaterand then doesn't do anything
17:36:37rasherMaking flyspray add the tracker number to the filename would be nice
17:36:51rasherGodEater: That's by design - press enter to continue
17:37:05rasherGodEater: Or you could press s to select simulator, then enter to continue
17:37:05GodEatershould we have a prompt which says that ?
17:37:18rasherProbably - I didn't make this part though
17:37:27 Join spiorf [0] (
17:37:33GodEateralso, get a permission denied error on
17:37:54rasherProbably not set executable?
17:38:04rasherDon't know if you can do that in a patch
17:38:11GodEater'spose not
17:38:26rasher echo "Enter your developer options (press enter when done)"
17:38:32rasherWhat do you suggest it says instead?
17:38:46GodEaterI saw no such prompt
17:39:06rasherThen I do believe you weren't looking
17:39:16rasherAfter you press a, you get that prompt
17:39:52GodEaterah yes
17:40:09GodEaterso you can select both voice and simulator at that point can you ?
17:40:24GodEaterah I see
17:40:29rasherOr other combinations
17:41:02 Quit desowin ("use linux")
17:41:24GodEatervoice file generated successfully
17:41:39rasherThat's good. On linux?
17:41:48GodEatershould I run it using "time"
17:41:52GodEaterto see if there's a speed up ?
17:41:59rasherThat might be interesting
17:42:29GodEaterok - I'll revert my tree
17:42:34GodEaterdo it without the patch
17:42:36GodEaterand then again with
17:44:33 Quit datachild ("Half of the modern drugs could well be thrown out of the window, except that the birds might eat them.")
17:45:05GodEaterusing original method :
17:45:19GodEaterreal 1m20.501s
17:45:19GodEateruser 0m46.623s
17:45:19GodEatersys 0m18.770s
17:46:03rasherIt's probably not terribly significant on Linux
17:46:08rasherIf at all
17:46:52 Nick fxb is now known as fxb__ (
17:47:03rasherI just tested, and it was slightly slower for me
17:48:17GodEaterfaster here - AND I just noticed my PC had started to compile something else
17:48:30GodEaterso I'm doing it again without that to see how much faster
17:48:52GodEaterreal 1m1.498s
17:48:53GodEateruser 0m51.234s
17:48:53GodEatersys 0m8.251s
17:48:56GodEaternot bad
17:49:29rasherThere could be all sorts of thing making my test slower I guess
17:50:02 Quit darkraven (Read error: 110 (Connection timed out))
17:52:01 Join sarixe [0] (
17:53:17GodEaterwell the important thing is it works fine :)
17:53:47linuxstbsaratoga: Yes, I believe the LGPL is compatible with GPLv3 - you can use LGPL'd code with anything...
17:53:57rasherIt's certainly far faster in cygwin
17:55:36rasherRebuilding with a full clip pool (previously generated voice clips, saved in a separate dir) is an order of magnitude faster for me, on Linux..
17:55:45*GodEater spots it's 5 o'clock, and therefore time to go to the pub
17:55:49rasherThe shells cripts take 25 seconds, the perl script takes 2
17:55:54 Join perrikwp [0] (n=chatzill@
17:56:36 Join chrisjs169 [0] (
18:00:17 Join Lear [0] (i=chatzill@rockbox/developer/lear)
18:04:27 Join pixelma [0] (i=pixelma@rockbox/staff/pixelma)
18:05:08***Saving seen data "./dancer.seen"
18:05:22 Nick _pill is now known as pill (
18:17:57 Join My_Sic [0] (
18:18:11 Quit obo ("bye")
18:18:56lostlogichah, someone jtag'd the iphone... I guess maybe we should consider a port to it afterall? :-\
18:21:21 Quit chrisjs169 (Read error: 110 (Connection timed out))
18:21:28 Quit Lear ("ChatZilla [Firefox]")
18:25:55 Quit My_Sic ("Miranda IM! Smaller, Faster, Easier.")
18:29:05 Join amiconn_ [0] (n=jens@rockbox/developer/amiconn)
18:29:20 Quit amiconn (Nick collision from services.)
18:29:21 Nick amiconn_ is now known as amiconn (n=jens@rockbox/developer/amiconn)
18:29:40amiconnrasher: I applied your patch and configured for sapi, german and english. For some reason a 'make voice' the first builds all bitmap libraries
18:29:52amiconn...before starting to build the voice files
18:30:30 Join Phalangees [0] (
18:31:57amiconnrasher: Didn't you change the clip filenames to use the md5sum?
18:32:55linuxstblostlogic: I wonder if the 2nd gen Nano has any similarities...
18:36:41 Join Lear [0] (i=chatzill@rockbox/developer/lear)
18:37:10amiconnrasher: The spFileStream object could also be created once and then reused. My old script did that
18:37:28 Join petur [0] (n=petur@rockbox/developer/petur)
18:38:06amiconnIt also looks like the current voice file generation doesn't handle VOICE_PAUSE properly (both svn and your perl script)
18:43:58amiconnrasher: Urgs, sapi output doesn't handle language properly. Passing german text to an english voice sounds.... kinda funny....
18:46:39 Join mexicankiller [0] (n=chatzill@
18:47:14amiconnrasher: Hmm, and for some reason the voice file is around twice as large as the downloadable ones
18:47:39amiconn...and doesn't work
18:49:25amiconnDid you forget wavtrim?
18:54:52 Quit mexicankiller ("ChatZilla [Firefox]")
18:55:38peturamiconn: like an american trying to talk german?
18:57:15 Quit ctaylorr (Read error: 110 (Connection timed out))
18:58:20 Join linuxstb_ [0] (n=linuxstb@rockbox/developer/linuxstb)
18:58:37 Join Captain_A [0] (
18:58:50 Quit linuxstb (Nick collision from services.)
18:58:54 Nick linuxstb_ is now known as linuxstb (n=linuxstb@rockbox/developer/linuxstb)
19:03:11 Part pixelma
19:05:42 Join pixelma [0] (i=pixelma@rockbox/staff/pixelma)
19:06:44 Join whm [0] (
19:12:05 Quit sarixe ("Peace")
19:13:19whmhello. the faq keeps talking about archos players? it's valid for ipods too, right?
19:13:35bluebrotheryes, to some degree.
19:13:37linuxstbWhich FAQ?
19:13:43bluebrotherthere is a ipod-specific faq too ...
19:17:28whmok i found it, finally, very well hidden ;)
19:18:13 Join midgey [0] (
19:18:14bluebrotherwhm: what's hidden there?
19:18:28linuxstbOr "wiki -> ipod port -> ipod faq"...
19:18:57whmok, i was first looking at documentation -> index -> faq
19:18:58linuxstbor "wik" -> rockbox FAQ -> ipod FAQ"...
19:19:23bluebrotherand what's hard to find the Ipod FAQ through the documentation index?
19:19:24linuxstbWe may have too many indexes...
19:19:33*bluebrother is really puzzled
19:19:48saratogaactually, i can rarely find things in the wiki
19:19:56saratogabetter organization might be nice
19:20:01saratogabut its a lot of work
19:20:48bluebrothermaybe using the namespaces more would help ...
19:20:57whmwell my problem was that i just clicked the first "FAQ" i found. Ipod FAQ is further down..
19:21:28bluebrotheruser error :)
19:21:32whmhehe ;)
19:21:54 Join obo_ [0] (n=obo@rockbox/developer/obo)
19:22:01whmok now that i have read archos faq and some of ipod faq, i have couple of questions..
19:22:10linuxstbwhm: The second question in the main FAQ links to the other FAQs as well.. ;)
19:22:25 Nick obo_ is now known as obo (n=obo@rockbox/developer/obo)
19:22:32linuxstbwhm: What's the question
19:22:48whmok :)
19:23:01whmyes, theme keeps resetting from time to time?
19:23:25whmi can set it and it survives reboot, but then later on i get back to the default or something.
19:23:53whmto be more clear, seems it changes by itself.
19:24:10bluebrothera theme needs to be installed below the .rockbox folder
19:24:53whmyes, it's there. i dropped all the files in wps folder into .rockbox/wps and the same for the rest of the folders.
19:25:23 Quit seablue ("life, death, life, death")
19:25:31whmok now it changed again..
19:26:13whmi basicly just set "Silver", then rebooted, and it was still ok. Then left it charging, came back and now i have the default one again.
19:26:57 Quit midgey ()
19:27:13amiconnPerhaps you booted with the hold switch enabled (e.g. by plugging usb)?
19:27:33whmno, i pressed play/pause for a few seconds..
19:27:56linuxstbYou didn't turn hold on when you left it to charge?
19:28:22whmi'll try it again with other theme
19:28:42linuxstbAre you using an official Rockbox build?
19:29:03linuxstbAnd which ipod?
19:29:53whmipod nano 1st gen, and i think it's a "release" build downloaded yesterday.
19:31:09 Quit jgarvey (Read error: 110 (Connection timed out))
19:31:29 Join jgarvey [0] (
19:32:20whmanother question, charging is hardware controlled on ipods also?
19:37:01 Join SirFunk [0] (
19:37:07whmthe other theme seems to work for now at least. maybe the first one was messed up somehow (i copied the files by hand after all).
19:39:32 Join jrbil [0] (n=chatzill@
19:39:57 Join theglades00 [0] (
19:40:21 Join euz [0] (
19:40:54bluebrotherDomonoky: what do you think about splitting the wiki page for rbutil in two?
19:41:56linuxstbsaratoga: Regarding the last packet in ASF files, have you confirmed it's not being passed to the decoder?
19:42:00euzafter first failed attempts, I was wondering, what's the probability my iriver iHP120 supporting simple CF to IDE hacks? And where do I look for info?
19:42:31jrbilanyone familiar with building sansapatcher?
19:42:34linuxstbeuz: Have you searched the misticriver forums?
19:42:38linuxstbjrbil: Yes
19:42:49linuxstbjrbil: What's the problem?
19:42:53jrbilperfect... i get it to build, but it always seg faults.
19:43:02linuxstbOn what OS?
19:43:23linuxstbHave you run it in gdb?
19:43:34 Join donutman25 [0] (n=chatzill@
19:43:35jrbilyes... and i do a back trace
19:43:39linuxstb32-bit or 64-bit?
19:43:41jrbilbut it's uniformative.
19:43:44euznot yet
19:44:04jrbilhow is the actual bootloader built?
19:44:19linuxstbjrbil: I assume you've removed the call to "strip" in the Makefile?
19:44:31jrbilYou bet. And I did a make clean.
19:44:47euzI built my own adapter, tested it on pc, but iriver ignores it
19:44:50linuxstbjrbil: You build the bootloader the same way as you build Rockbox - select your target in configure, then "b" for bootloader.
19:45:02euzthink it could be a hardware limitation?
19:45:09linuxstbjrbil: But it's not recommended to use SVN bootloaders - they're not tested.
19:45:14bluebrotheradding -g to CFLAGS might be helpful too ...
19:45:15jrbilI never did that.
19:45:27jrbilWill that build sansapatcher for me as well?
19:45:38jrbilI added the "-g".
19:46:15linuxstbNo, they're independent. You don't need to build sansapatcher to install a bootloader - just do "sansapatcher -a bootloader.mi4" (or whatever it's called).
19:46:44jrbilPerfect. The bootloader I downloaded works fine.
19:47:04jrbilI've been playing around with the garbled lines on the e200.
19:47:17jrbilI've got a patch, but I haven't been able to test it with the bootloader.
19:52:58euzeveryone seems at a loss at mysticriver
19:53:29euznoone here tried it before?
19:54:11linuxstbeuz: The people that may know don't seem to be around at the moment. I would suggest either emailing the rockbox mailing list, or posting a message in the forums.
19:54:15amiconnCF to ATA will probably only work if the CF card in question supports the full ATA command set
19:54:23amiconnNot all of them do
19:54:29*linuxstb spots one of those people...
19:54:29euzI bought connectors from twaian and soldered myself, although I have found the right adapter
19:55:25euzit seems there was a update to the IDEspecifications in order to support/improve support for CF cards
19:55:52euzI have tested the CF in question on PC, it worked wonderfully
19:56:23linuxstbeuz: Are you running the iriver firmware or Rockbox in flash?
19:56:55petureuz: can you access it when connected through usb?
19:57:03linuxstbeuz: Does the bootloader give any error message?
19:57:25euzI didn format the CF the right way, was expecting the bootloader to detect it
19:57:38euzthe bootloader freezes on the third line
19:57:42amiconnThe bootloader hangs if the card isn't formatted
19:57:48 Quit jrbil ("ChatZilla [Firefox]")
19:57:52amiconnCard or disk, of course
19:57:58euzlet me try, brb
19:58:25*petur repeats himself: euz: can you access it when connected through usb?
19:58:49linuxstbpetur: If the bootloader freezes, I'm assuming he can't access it via usb?
19:59:03amiconnUSB works
19:59:09peturthought so too
19:59:34whmouch, theme got reset again and ipod also crashed.. it was stuck at the white apple screen with backlight on... any ideas? could it be that i messed up the FAT32 stuff or something?
19:59:35amiconnIt freezes when trying to boot, not when going into bootloader usb
19:59:51linuxstbamiconn: Ah, OK...
20:00:09linuxstbwhm: Do you get any other problems on your ipod? e.g. does music playback OK?
20:01:04whmusually yes, but once i got weird "rippling" sound.. sounded like a broken speaker basicly.
20:01:32linuxstbwhm: It could be a corrupt filesystem. You could try running "chkdsk /f"
20:02:27euznothing at all
20:02:36euzoriginal firmware refuses it
20:02:58euzcheck hdd connection
20:03:51linuxstbI thought you had Rockbox in flash?
20:03:56 Quit Lear ("ChatZilla [Firefox]")
20:04:02amiconnpixelma: ping.....
20:04:03euztried both :)
20:04:16euzwhat do you need in flash?
20:04:21linuxstbeuz: Ignore me...
20:04:30*linuxstb forgot how the h140 bootloader worked...
20:04:33euzheh, it ok
20:04:46whmlinuxstb is it ok to try to check it with disk utility, i only have a mac..?
20:04:49euzah, was wondering if it was possible to replace original
20:05:09***Saving seen data "./dancer.seen"
20:05:13linuxstbwhm: I don't know what disk utility can do...
20:05:27whmok i'll have a look if it can check fat32...
20:05:33euzdamm, soldering the adapter was such a bitch
20:05:38 Join pixelma_ [0] (i=pixelma@rockbox/staff/pixelma)
20:06:06 Quit pixelma (Nick collision from services.)
20:06:13 Nick pixelma_ is now known as pixelma (i=pixelma@rockbox/staff/pixelma)
20:06:38linuxstbeuz: You've partitioned and formatted it?
20:07:02whmyes, but using the manual method and i did have some problems with it...
20:07:23whmok disk utility says `..' entry in /.rockbox has non-zero start cluster
20:07:28euznot yet
20:07:54euzthink it might do the trick?
20:08:04linuxstbThat's what amiconn said 10 minutes ago ;)
20:08:18euzI think it's either a firmware or hardware limitation
20:08:35euzthe ihp120 is pretty old
20:08:57 Join p3tur [0] (n=petur@rockbox/developer/petur)
20:08:58 Quit petur ("switchin")
20:09:48 Nick p3tur is now known as petur (n=petur@rockbox/developer/petur)
20:10:02whmok i told disk thingy to fix it.. i'll be back if i have any more problems.. thanks a lot guys! (it was simple again heh ;)
20:10:12euzthe original firmware should be able to format it regardless I think, unless there was an expecific hack in the boot loader to support CF
20:10:38Phalangees when using rb->lcd_bitmap_transparent_part(data,srcX,srcY,stride,x,y): what is stride?
20:10:45 Join webguest21 [0] (i=d9e36395@gateway/web/cgi-irc/
20:11:07linuxstbPhalangees: Have you looked at the GraphicsAPI wiki page?
20:11:18 Join Nico_P [0] (n=nicolas@rockbox/developer/NicoP)
20:11:34euzI see you have a cf on your ipod mini?
20:11:36PhalangeesI am now. Thanks
20:14:22 Join pixelma_ [0] (i=pixelma@rockbox/staff/pixelma)
20:14:35 Join dandin1 [0] (
20:14:43 Quit pixelma (Nick collision from services.)
20:14:43 Nick pixelma_ is now known as pixelma (i=pixelma@rockbox/staff/pixelma)
20:14:48webguest21uhm sorry for bugging you guys but does onyone know why my ipod pops up the message "Data abort at 0030900" everytime i try to play a track? I used RIX to install rockbox on my ipod video after i resettet it... hope sb knows something
20:15:02linuxstbwebguest21: Because you used RIX to install rockbox...
20:15:10webguest21uh thanks :P
20:15:21linuxstbTry following the instructions in the manual - it's not complicated.
20:15:22webguest21so probably doing it by hand will solve the problem?
20:16:02linuxstbYes, RIX is very old and very broken.
20:16:23linuxstb(unless they've updated it in the last couple of months)
20:16:43webguest21uh thanks already
20:16:51 Quit Nico_P (Remote closed the connection)
20:17:50euzwhat time should I be back in order to find help with my CF issue?
20:17:54Domonokywebguest21: if you want to use a graphical installer, use rbutil.. :-)
20:18:39linuxstbeuz: Have you attempted to partition and format it, and see what the Rockbox bootloader reports?
20:18:40 Quit BobShield (Read error: 104 (Connection reset by peer))
20:19:10 Join desowin [0] (n=desowin@
20:19:19 Join oxygen77 [0] (
20:19:51 Quit Entasis (Read error: 104 (Connection reset by peer))
20:20:28euzI haven't
20:20:55euzlet me try
20:21:28*petur wonders why he asked two times about accessing through usb
20:21:40 Quit euz ("Leaving")
20:22:18pixelmaso that you can add one more to your action count now ;)
20:23:02 Join DC1 [0] (
20:23:04*petur checks his stats :)
20:24:29peturpixelma: I'll never catch up amiconn this month :)
20:25:40 Quit DC1 (Client Quit)
20:26:14pixelmapetur: get some practice for September then
20:26:28 Join BobShield [0] (
20:26:32*petur exercises :)
20:26:35 Quit HellDragon (Client Quit)
20:27:02amiconn"Poor his, nobody likes him/her. He/She was attacked 2 times"
20:27:22peturdoing better than forehead then :)
20:27:57pixelmawasn't me :P
20:28:48 Join midgey [0] (
20:29:28*amiconn shouldn't talk that much and code more instead :\\
20:30:07*Domonoky seems to smile to much :-)
20:33:17amiconnpetur teh attacker ;)
20:33:50 Quit SirFunk (Read error: 104 (Connection reset by peer))
20:33:55*petur slaps whoever he likes (and adds another one to the stats)
20:34:18bluebrotherpetur: checked the "all stats" page? ;-)
20:34:59amiconnThe all stats page seems to be a bit buggy. Check the "last seen"...
20:35:05 Quit midgey ()
20:35:32bluebrotherDomonoky: we should starting to advertise rbutilqt to get testers ...
20:36:11Domonokysure.. :-)
20:36:31linuxstbamiconn: Wow, you've typed 1.2 million words in IRC...
20:36:51*bluebrother goes splitting up the rbutil wiki page
20:37:00Domonokybluebrother: do we have someone who can build us a mac version ?
20:37:14bluebrotherdon't know −− linuxstb?
20:37:17 Join pixelma_ [0] (i=pixelma@rockbox/staff/pixelma)
20:37:26linuxstbbluebrother: QT wouldn't compile for me...
20:38:16 Quit pixelma (Nick collision from services.)
20:38:17 Nick pixelma_ is now known as pixelma (i=pixelma@rockbox/staff/pixelma)
20:38:42pixelmasorry for disconnects/rejoins
20:38:53*pixelma slaps router...
20:39:14webguest21uh thanks a lot it works out great if you do the installation by hand
20:39:39 Quit perrikwp (Read error: 110 (Connection timed out))
20:41:05 Join low_light [0] (i=c730180b@gateway/web/cgi-irc/
20:42:23low_lightjhMikeS: does sansa ata use dma?
20:42:41amiconnSansa isn't ata
20:43:09bluebrotheris there a way to move an attachment to a different wiki page?
20:43:28jhMikeSlow_light: no
20:44:03jhMikeSit's still called ata-e200 even though not technically ATA :)
20:44:14low_lightI found some reg writes to 0x6000a0xx & 0x6000b0xx during reads & writes and was wondering if they were dma related
20:44:22amiconnYes, same as Ondio MMC is called ata_mmc
20:44:37jhMikeSdid you check pp5022.h?
20:44:51linuxstbbluebrother: Yes, you can manage attachments and move them, but I don't think normal wiki users have the right to.
20:44:59jhMikeSpp5020.h rather
20:45:20 Join lee-qid [0] (
20:45:21linuxstbbluebrother: What do you want to attach? I thought executables were hosted on
20:45:34jhMikeSguess nothing's there for those
20:46:03low_lightjhMikeS: there's a reference in system-pp502x about 0x6000a000 turing on dma
20:46:26bluebrotherlinuxstb: I split the RockboxUtility wiki page up −− the Qt version is now RockboxUtilityQt. I want to move the screenshots I added a couple of days ago to the new page
20:47:04linuxstblow_light: MrH wrote a document (about the NAND Flash I think) documenting DMA, but I don't recall anyone succeeding with it.
20:47:36 Join thegeek_ [0] (
20:47:54 Quit thegeek (Read error: 113 (No route to host))
20:48:49 Join freqmod [0] (
20:49:20low_lightjhMikeS: here's the ata/sd_read function and the 2 functions maybe related to dma:
20:50:08 Join haemmy [0] (n=stefan@
20:50:17low_lightoh...this is from the sansa c200 (pp5022c), but maybe has the same sd controller
20:50:35 Join euz [0] (
20:51:14low_lightand there's similar functions for writing
20:51:37linuxstbeuz: You created a partition table with a single primary partition (FAT32) and formatted that partition as FAT32?
20:51:43euzwhile I like being a "pioneer", I'm way too far from civilization for it
20:51:55euzyes, type 0B
20:51:57jhMikeSlow_light: the 6000b... address do appear to be DMA. e200 does use it.
20:52:05euzmkdosfs -F 32
20:52:11linuxstbeuz: And what did the Rockbox bootloader say?
20:52:24bluebrotherwiki admins: can someone please move the images I attached to RockboxUtility to RockboxUtilityQt?
20:52:24euzsame as before, nothing
20:52:36euzhang on 3rd line
20:53:00linuxstbWhat did those three lines say? Any ATA error message?
20:53:11euzalthough the next lines when using original hdd aren't very informative
20:54:22euzno, I remember those when I had screwed my hdd, but not this time
20:56:29linuxstbSo what is the last line that's displayed? And what version is the bootloader you're using?
20:57:17amiconnIch probiere mal noch 'was
21:00:21peturanybody against me closing 7644 - doesn´t make much sense...
21:00:50euzversion 6
21:01:37 Quit webguest21 ("CGI:IRC (EOF)")
21:04:00linuxstbbluebrother: Can't you just change the URLs?
21:04:31bluebrotherhmm −− it's %ATTACHURL%/file.jpg
21:05:04bluebrotherok, using ../ works but ...
21:06:05bluebrotherok, I cheated for now. Maybe someone with the appropriate powers wants to make this nice though ...
21:06:14euzso what time should I come back?
21:06:55linuxstbeuz: You haven't said what the bootloader messages are. i.e. where in the boot process is Rockbox freezing?
21:08:37euzrockbox boot loader /n version 6 /n Batt: 4.06V
21:08:58euzBatt was a little higher when using the CF :)
21:09:55euzATA errors appear before that message, ie, damaged HDD/without HDD
21:10:20euzso when CF is plugged in, I'm actually getting past the ATA error routine
21:14:16bluebrothereuz: what are you trying to do? Use a cf card instead of the hard drive?
21:14:39euzyes, on iriver ihp120
21:14:51bluebrotherhave you seen ?
21:15:23euznot yet
21:15:40bluebrotherthe patch doesn't apply anymore iirc but you can search for that comment
21:17:11low_lightjhMikeS: I see that those registers are defined in toni's sansa emulator
21:18:47 Join einhirn [0] (
21:19:48euzinteresting, seems only CF with TrueIDE works
21:20:00euzbut patch is still required
21:20:48jhMikeSlow_light: yeah. I looked for them in reg.txt before my last posting because they seemed familiar.
21:26:16 Quit donutman25 ("ChatZilla [Firefox]")
21:28:02*linuxstb reads and sees that a file with the .ogg extension is now only supposed to contain a vorbis audio stream
21:29:55 Quit low_light ("CGI:IRC (EOF)")
21:30:30 Quit haemmy ()
21:32:21 Join agm3nt [0] (
21:32:41 Quit iamben (Read error: 104 (Connection reset by peer))
21:33:03 Part agm3nt
21:33:28 Join chrisjs169 [0] (
21:33:46 Join low_light [0] (i=c730180b@gateway/web/cgi-irc/
21:33:54 Quit spiorf (Remote closed the connection)
21:38:15 Quit Siku ()
21:44:12 Join seablue [0] (
21:49:58 Join advcomp2019- [0] (n=advcomp2@
21:50:24 Quit advcomp2019 (Nick collision from services.)
21:50:27 Nick advcomp2019- is now known as advcomp2019 (n=advcomp2@
21:51:58 Quit Phalangees ("Ex-Chat")
21:52:49 Quit jgarvey (Read error: 110 (Connection timed out))
21:52:52 Join jgarvey_ [0] (
21:52:55euzI'll post in forum, laters
21:52:59 Quit euz ("Leaving")
21:59:58 Quit dandin1 ()
22:05:12***Saving seen data "./dancer.seen"
22:07:44 Join spiorf [0] (
22:08:53preglowanyone noticed slower eq response than usual? i cant reproduce a bug report here
22:09:28peturnot an eq user....
22:09:32preglowme neither...
22:16:53amiconnsaratoga: Hmm, the comment in your code is a bit strange...
22:17:18amiconnYou changed from 17.15 precision to 17.15 precision?
22:18:17 Quit pixelma (" switching")
22:18:29 Join mirak [0] (n=mirak@
22:18:37amiconnHmm, that comment was just moved...
22:19:20 Quit davina ("xchat on Ubuntu 7.04")
22:20:16 Join pixelma [0] (i=pixelma@rockbox/staff/pixelma)
22:23:57 Join ompaul [0] (n=ompaul@freenode/staff/gnewsense.ompaul)
22:25:47 Quit low_light ("CGI:IRC (EOF)")
22:26:57amiconnjhMikeS: Should test_cache work in svn?
22:27:54amiconnAh no, doesn't compile....
22:31:10amiconnpreglow: Does software treble/bass also use the asm loop in eq_arm.S?
22:36:18 Quit gtkspert (Remote closed the connection)
22:38:03preglowamiconn: yeap
22:38:24 Join gtkspert [0] (
22:38:24amiconnOkay, so putting this loop into iram on PP5002 would prbably be a good idea...
22:38:24preglowamiconn: i actually made code to combine two first order filters to a second order filter just so i could reuse the asm loop
22:38:57preglowi'd imagine so
22:39:02peturis there a way to know the w/h that text will need (like using DT_CALCRECT with DrawText on windows if that may be a reference)
22:41:09 Join Domonoky_ [0] (
22:42:24 Join linuxstb_ [0] (n=linuxstb@rockbox/developer/linuxstb)
22:44:12 Join einhirn_ [0] (
22:45:30jhMikeSamiconn: I suppose something could be cooked up to sync the threads besides the queues
22:48:40 Join Robin0800 [0] (
22:49:20Robin0800what is the size of the wps image buffer?
22:51:45Robin0800trying to use large icons and sim cant load all the images
22:51:56 Part Robin0800
22:52:38 Quit einhirn_ (Read error: 104 (Connection reset by peer))
22:52:48 Join Robin0800 [0] (
22:54:47 Quit kubiixaka ("Miranda IM! Smaller, Faster, Easier.")
22:55:28 Quit linuxstb (Read error: 110 (Connection timed out))
22:58:23 Quit Domonoky (Read error: 110 (Connection timed out))
22:58:42 Quit spiorf (Remote closed the connection)
22:59:35amiconnpreglow: Wow, that nearly halves the cost of sw treble/bass on PP5002....
23:01:41 Quit saratoga ("CGI:IRC (Ping timeout)")
23:02:46 Quit einhirn (Read error: 110 (Connection timed out))
23:06:55 Quit desowin ("use linux")
23:07:01 Join yegods [0] (i=cf2f0e57@gateway/web/cgi-irc/
23:07:54 Join linuxstb [0] (i=5343d4aa@rockbox/developer/linuxstb)
23:07:58yegodsany GUI experts around?
23:09:19linuxstbI assume you're going to ask about ?
23:09:37yegodsthat was the intent :)
23:11:02yegodsi'm guessing there's more of a population around at night?
23:11:32amiconnpreglow: C00l, no more skipping when listening to higher bitrate oggs with treble+bass enabled :)
23:11:43amiconnBoost ratio is around 86%
23:11:53rasheramiconn: The clip filenames are named by MD5sum only if building a voiceclip pool. I have no idea about why the bitmaps are built - I was wondering that myself. Someone with a clue should fix the .vbs to be most efficient, I've basically no idea what it's doing.
23:12:06 Join BigBambi [0] (n=Alex@
23:12:16amiconnOk. What about the language thing?
23:12:16BigBambiyegods: It is night
23:12:24BigBambiAn awful lot of the devs and in Europe
23:12:26rasheramiconn: About the language, I don't know how to set the language for SAPI, so I don't know what would be a reasonable way of setting it.
23:12:31BigBambiand here it is Friday night
23:12:55amiconnrasher: Is there a language identifier of some kind available in the perl script?
23:12:59 Quit freqmod (Remote closed the connection)
23:13:02rasheramiconn: Maybe the options to tts could somehow be passed on to the sapi script?
23:13:06amiconnAnd what about wavtrim?
23:13:09rasheramiconn: Yes
23:13:15rasherAh! I forgot completely about wavtrim
23:13:45amiconnOne of the scripts (either perl or vbscript) does need a mapping from language (is that a string?) to windows languages ids
23:14:03 Join HellDragon [0] (n=Nocebo@unaffiliated/helldragon)
23:14:04amiconnThat's not a problem as such - but for several languages we have a problem
23:14:11rasherWindows language ids? Is that some sort of fixed hash?
23:14:23amiconnThese are well defined numbers
23:14:26rasherAlso, I forgot about the utf8 problem
23:14:34rasherOkay, a simple table in the perl script should do, then
23:14:42amiconnE.g. rockbox just has "english.lang" - but windows has american english and british english as separate IDs
23:15:05linuxstbyegods: I don't believe what you want to do is possible - i.e. refresh the current screen from the power thread (I'm guessing you've used the power thread).
23:15:06rasheramiconn: Shouldn't that be set in the options the user is asked for for the TTS?
23:15:13amiconnI think leaving that job for the vbscript would be better
23:15:32amiconnLess system specific stuff in the perl script this way
23:15:39rasherI can pass the language we're building to the vbscript no problem
23:15:47rasher(ie. 'english' 'polish' etc.)
23:15:55rashereh, polski, that is
23:16:02preglowamiconn: not bad
23:16:05preglowso much for cache...
23:16:09amiconnThat table would need to be extended f we add languages
23:16:17rasherNot a huge problem
23:16:25rasherBut something to be aware of, of course
23:16:31amiconnpreglow: Yes, and the function is <128 bytes...
23:17:27rasherWasn't there a patch around to make IRAM usage configurable in the target conf files?
23:18:02preglowits commited, afaik
23:18:04amiconnYes, but those I*_ATTR macros won't help in asm sources anyway
23:18:07 Join davina [0] (
23:18:11preglowthats true
23:18:20rasherI stand corrected on many accounts.
23:19:00amiconnlinuxstb: Bad suggestion...
23:19:12preglowso no skipping when playing high bitrate oggs with tone controls on pp5002?
23:19:13amiconn...unless you mean just calling lcd_update()
23:19:35amiconnpreglow: Nope, although these oggs are just ~180kbps
23:19:46preglowthat rocks!
23:19:52preglowthats way better than i would have guessed
23:19:53amiconnThey skipped a few times before putting tone controls into iram
23:20:18rasherMy, the 1..3G sounds perfectly usable
23:20:36amiconnFor mp3: linear output: ~69% boost. Tone controls in sdram: ~87% boost. Tone control in iram: ~78% boost
23:20:53preglowthats exactly the way rockbox should be, making even older players do the job
23:20:55amiconn(the usual −−preset standard stuff)
23:21:17amiconnAnd yes, mp3 performs better than ogg on pp5002
23:21:23preglowi would expect it to
23:21:32 Quit HellDragon (Client Quit)
23:21:35 Part Domonoky_
23:21:39preglowanyway, its bedtime for me now
23:21:41preglowhappy hacking!
23:21:47linuxstbamiconn: I'm not suggesting it - it's what yegods wants to do. See this thread -
23:23:33rasheramiconn: How about I pass the language as the first argument to sapi5_voice_new.vbs, and then whatever the user entered when prompted for options during configure after that?
23:23:47 Quit davina (Remote closed the connection)
23:23:51amiconnHmm, afaics that needs notifying the gui thread that config has changed, and make all screens process that message
23:24:06rasheramiconn: And then the order of options could be by convention or something. Perhaps voice-name as second option or something
23:24:35 Join HellDragon [0] (n=Nocebo@unaffiliated/helldragon)
23:25:26amiconnHmm, sounds like an idea
23:25:46amiconnDoes configure allow to select the voice(s) to use for other engines?
23:26:05rasheramiconn: only by passing flags to the executables
23:26:13rasherIn the same way it's done for LAME
23:26:18amiconnFor sapi it would be possible to list all available voices, optionally filtered by language
23:26:41rasherThat'd need some magic inside configure, but could be done - I doubt I'll be able to do it though
23:26:52rasherCertainly possible
23:26:57amiconnOnce this new way is in svn, I'll probably re-add much of the stuff from my old script
23:27:21amiconn(like the string replacement to correct pronunciation)
23:27:50rasherI think that should be done in perl, probably directly in It's relevant for all engines
23:27:59rasherAnd perl happens to be quite good at text manipulation
23:28:03amiconnIt's engine specific
23:28:16amiconnANd vbscript can do regex as well
23:28:18rasherYes I know, but it could still be done inside
23:28:29amiconnIn fact my old script does use regex for those lists
23:28:30rasherBut separate for each engine, that is
23:28:46rasherHaving it in a single location just seems to make more sense?
23:28:56rasherA big bad switch() statement
23:29:00rasherNested, perhaps
23:29:09rasherFor extra dramatic effect
23:29:10amiconnHmm, those lists might become lengthy
23:29:38rasherCould be a separate file, but I really think it makes sense to do it in the perl script in all cases
23:30:58amiconnSee here: , starting at ' Replacement list definitions
23:30:58 Join sambolina3 [0] (i=47fe6474@gateway/web/cgi-irc/
23:31:48amiconnHmm, also above that
23:32:34rasherI think I'll add my idea of string replacing to the script now. We can always change it
23:32:39amiconnrasher: Will you fix the missing wavtrim?
23:32:47rasherYes, adding that as we speak
23:32:55amiconnBtw, building one voice took ~5 minutes
23:33:09amiconnStill not great, but much better than svn according to reports
23:33:27amiconnHow do I use the pool?
23:33:39rashermake a directory, and export POOL to the path to that directory
23:33:49rashermkdir pool; POOL=pool make voice
23:34:11amiconnHow does that work if I specify more than one language?
23:34:17amiconnThe clips are named the same...
23:34:23rasherNo, not in the pool
23:34:51rasherOnly in the current build dir.. which reminds me that we need to remove *mp3 and *wav before exiting
23:34:54*amiconn thinks it's irritating that the clip naming scheme changes when using the pool
23:35:12rasherWell, we have to name the mp3s by ID, or voicefont won't accept them
23:35:17amiconnYes, and the *.wav could also be removed when using the pool
23:35:27amiconnAh, yes
23:35:46rasherAnd we have to name by md5+language+engine when using the pool, to avoid conflicts
23:35:47*amiconn is somewhat slow
23:35:51 Part oxygen77
23:36:26rasherIn fact, we should just remove the wav as soon as we've encoded it
23:36:34amiconnsaves space
23:36:54amiconnIn fact we could even pipe the .wav to the encoder (?)
23:37:17rasherTechnically yes, but it'd make a mess
23:37:36rasherI'd have to pass around a bunch of filehandles, and I don't think we'd gain much
23:38:03 Join scorche [0] (n=scorche@rockbox/administrator/scorche)
23:38:23amiconnPerhaps on windows. But my old script didn't do that either
23:40:15amiconnBtw, regarding language selection, I would think that e.g. festival language names don't match rockbox language names either, so they would need a translation table as well
23:40:35 Join ctaylorr [0] (
23:41:03rasherActually I was pretty much expecting users just to set commandline options for that
23:41:27amiconnWell, the script lists all the languages and allows multiselection...
23:41:55rasheryeah, but that just selects which strings to put in the voicefile, not how to speak them
23:42:05rasherAnd the user is asked for commandline options for each language
23:42:21rasherAlthough I've no idea how you properly configure festival. You need to write stuff in scheme, I believe, but how to actually make it execute it, I've no idea
23:42:43rasherAnd frankly, I'm not sure I want to know
23:42:58amiconnDoes the script accept selecting the same language twice?
23:43:21amiconn(e.g. if I want to build both us and british english, or male and female voices for a langae)
23:43:38rasherI think that should be possible, but they'll be named the same
23:43:44rasherThat could be changed, of course
23:44:04rasherBut one thing at a time.. I think I need to make a checklist
23:44:14 Quit sambolina3 ("CGI:IRC")
23:44:24amiconnYeah, one thing at a time...
23:44:34rasherMaybe putting something in the wiki would make sense
23:44:41amiconnThe VOICE_PAUSE issue needs work as well
23:45:02amiconnIt's a very special ID....
23:47:49 Quit chrisjs169 (Read error: 110 (Connection timed out))
23:49:53 Join midgey [0] (
23:53:15 Quit witz (Read error: 110 (Connection timed out))
23:53:39 Join witz [0] (
23:54:02 Quit jgarvey_ ("Leaving")
23:56:53 Quit ender` (" I remember being impressed with Ada because you could write an infinite loop without a faked up condition. The idea being th)
23:58:54rasheramiconn: How so? Doesn't voicing " " work?

Previous day | Next day