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 2003-12-30

00:00:34[IDC]Dragonnot if they're fixed size, or fragments of it
00:00:47***Saving seen data "./dancer.seen"
00:00:50[IDC]Dragon(here comes the malloc discussion again :)
00:00:58LinusNsize doesn't matter in this case
00:01:15BC|codeLinus: she's just being polite - lol
00:01:19LinusNthe disk reading benefits immensely from the consecutive buffer
00:01:40[IDC]Dragonyes, that's worth keeping
00:01:41LinusNreading chunk by chunk will slow it down a lot
00:02:09 Join Sebulba02 [0] (
00:02:34Laurent_waow, the cache buffer in the hitachi travelstar is 2Mb
00:04:00Laurent_with consumption going down from 1.85W (when reading) to 0.65W with a peak at 4.7W when starting
00:04:42kurzhaarrockerLets say the file source buffer is nothing else but the most part of the whole source buffer then the advantages of the current disk reading benefits could remain.
00:05:20LinusNkurzhaarrocker: and the rest of the buffer would be?
00:05:26kurzhaarrockerthe beep
00:05:40Laurent_so idle mode consumes less than a third of reading mode (because of starting peak), and adds 2Mb worth of playing data without spinning the drive
00:06:10LinusNLaurent_: if the hard drive really cahed what we want
00:06:11Laurent_might be interesting... but certainly tricky as was said
00:06:20LinusNit also caches the FAT
00:06:27LinusNmay cache
00:06:32Laurent_yup, this is the big constraint
00:06:50kurzhaarrockerLaurent_ if I recall correctly the 8 MB mod didn't help to reduce battery consumption by more then 25%
00:07:00LinusNmy point is, you never know (and aren't supposed to know) what is in the cache
00:07:21Laurent_LinusN: sure
00:07:22[IDC]Dragonthis is encapsulated ;)
00:08:25[IDC]DragonMay I summarize the (hypothetical) playback scenario:
00:08:52Laurent_kurzhaarrocker: I imagine it comes from the fact that starting the drive is eating a lot of power
00:08:59[IDC]Dragonthe source hands "a buffer" to the sink, together with a callback to query the next
00:09:31[IDC]Dragonthis gets called on IRQ context and is free to define what's next
00:10:10[IDC]Dragonand there should be an inquire function to determine the position within the current buffer
00:10:36[IDC]Dragonwhat's the necessary reaction time to setup the next DMA?
00:10:53LinusNcould be less than 1ms
00:11:10[IDC]Dragon>10000 clocks
00:11:30[IDC]Dragonbut less than our heartbeat IRQ
00:12:07[IDC]DragonI was thinking if the callback has time to post into a queue
00:12:21[IDC]Dragonso the next buffer can be passed outside of IRQ
00:12:26LinusNa context switch can take forever
00:13:47[IDC]Dragonso the next will have to be given in IRQ context
00:14:18 Join Phil [0] (
00:14:28[IDC]Dragonif the app ever wants transition into normal mode, it has to work with at least 2 buffer parts
00:14:29Phildoofer provider
00:14:48Mode"#rockbox +o LinusN " by ChanServ (ChanServ@services.)
00:14:50[IDC]DragonI thought kurzhaarrocker is Phil
00:14:52 Quit Sebulba02 ("grr")
00:14:57LinusNkick kurzhaarrocker?
00:15:03Kick(#rockbox kurzhaarrocker :LinusN) by LinusN!
00:15:14Laurent_does anyone know the current "streaming" algorithm for MP3 files ?
00:15:33LinusNLaurent_: what do you want to know?
00:16:06 Nick Phil is now known as kurzhaarrocker (
00:16:54MTLaurent_: do you mean like (shout|ice)cast?
00:16:55Laurent_at which moment the hard drive gets started again : it must be quite a while before the buffer terminates otherwise there might be problems and the user can always do a fast forward
00:17:10Laurent_MT: no, for rockbox ;)
00:17:19MTaha, nm :)
00:18:39LinusNLaurent_: we have a low watermark which is calculated from the hard drive speed)
00:18:58kurzhaarrockerand the bitrate of the mp3
00:19:37Laurent_I guess this one determines when to start prefetching in order to be ready for any fast forward
00:20:02LinusNit is currently spinup_time*3 + watermark_margin (in seconds)
00:20:14kurzhaarrockerIs there prefetchen while ffwd?
00:20:22LinusNkurzhaarrocker: no
00:20:39Laurent_* 3 ? quite a big margin
00:21:22LinusN10-15 seconds
00:21:42Laurent_what do you mean by "no prefetching while ffwding" ? that the song eventually stops if the user keeps ffwding ?
00:21:42kurzhaarrockerLaurent_ asume you're shaking your drive and it need time to find its heads again.
00:21:46LinusNwe may be able to adjust that, since we have optimized the bitswapping since then
00:22:10LinusNLaurent_: ffing in rockbox is silent
00:22:25LinusNit reloads the buffer when you release the ff key
00:22:46Laurent_ok I see
00:23:13LinusNsilent ff is one of the reasons we are discussing the buffer handling right now
00:23:29kurzhaarrockerI want it LOUD!
00:24:20[IDC]Dragonspeaking about discussing: I'd be out to bed soon
00:24:44LinusNsweet dreams (about sources and sinks)
00:24:51*Laurent_ wants a laptop to code from bed
00:24:57[IDC]Dragonnot yet
00:25:09kurzhaarrockerLaurent_: Thats what I do.
00:25:13[IDC]Dragonjust trying to focus for a few last minutes
00:25:18Laurent_kurzhaarrocker: lol ;)
00:25:46[IDC]DragonI have a webpad to read email and IRC logs from bed
00:26:15[IDC]Dragonbut no notebook
00:26:25BC|coderf keyboard and binoculars :)
00:26:49[IDC]Dragonanyway, do we have any conclusions yet?
00:27:07kurzhaarrockerYes we want the source to provide the buffer to the sink
00:27:40[IDC]DragonNr.1, OK
00:27:53kurzhaarrockerWe (I) want to be able to control where the sink reads the buffer
00:28:10[IDC]Dragonnext we don't know how to steer the sink in there
00:28:24kurzhaarrockerWe want to keep the benefits of the continuous buffer for performance reasons while reading from a file
00:29:01kurzhaarrocker[IDC]Dragon Yes, how to steer the sink is still pending
00:29:14 Join diddystar5 [0] (
00:29:22[IDC]Dragonthe buffer transition is critical
00:29:25LinusNand the "loop from here to here" instructions comes from a third party, not the sink, not the source
00:29:32diddystar5LinusN: i finished al og scale for the vu meter
00:29:39kurzhaarrockerLinusN: right
00:29:41LinusNdiddystar5: wowee!
00:29:41BC|codeheyyyyyyy diddy
00:30:01*BC|code does some kind of cheerleader dance
00:30:09[IDC]DragonLinusN: why does the source not govern the loop?
00:30:11kurzhaarrockerdiddystar5: have you been able to use peakmeter code for that?
00:30:29kurzhaarrocker[IDC]Dragon the source doesn't know what to loop
00:30:36diddystar5kurzhaarrocker: no peakmeter code in it at all
00:30:41[IDC]Dragonwho else?
00:30:51LinusNthe user
00:31:01kurzhaarrocker[IDC]Dragon the application that uses the source
00:31:09[IDC]DragonI'm viewing "the source" as what feeds the sink,
00:31:19[IDC]Dragonnot necessarily the file.
00:31:38kurzhaarrockerdiddystar5: Have you a db scaling?
00:32:02kurzhaarrocker[IDC]Dragon: Thats what I meant the source that is based on a file
00:32:21diddystar5kurzhaarrocker: yes, thats also refered to logarithmic scale
00:32:31 Quit Lowfiler ("[using [DeViL'z IRC] (build 4.0) by [ins4ne] from dIRC.scriptingBoard.")
00:32:48[IDC]DragonI'm still a bit confused. Can you explain your loop scenario?
00:33:31kurzhaarrockerdiddystar5: would you mind mailing me the code? Maybe your db scaling can be reused if it outperforms the one in peakmeter.
00:34:15BC|codestraight table lookup
00:34:19Laurent_I wonder if it's possible to convert from ogg to mp3 without decoding/recoding completely ?
00:34:23kurzhaarrocker[IDC]Dragon: The user hits a button to mark the A point, the user hits a button to mark the B point. These playback loops from a->b
00:34:29Laurent_(I guess it's not but who knows ?)
00:34:31diddystar5kurzhaarrocker: can you use dcc?
00:34:52kurzhaarrockerdiddystar5: I might be firewalled, try.
00:34:59[IDC]Dragonkurzhaarrocker: I'm not concerned about the user, I mean your source/feed model.
00:36:29kurzhaarrocker[IDC]Dragon You're right: the source must be told what to loop, too in order to provide the right data if reload is needed
00:38:07LinusNa simple loop_start=stream_offset assignment
00:39:28kurzhaarrockerthat would be sufficiant for the start point. But what happens at the end of the loop
00:39:46[IDC]Dragonwould that mean we have to take away data from the sink which has already been passed to it?
00:40:02[IDC]DragonLike, the playback reaches close to point B,
00:40:17[IDC]Dragonbuffer has been committed way past it,
00:40:31[IDC]Dragonthe the user decides to loop until there?
00:40:37[IDC]Dragonthen the
00:40:47kurzhaarrockeryes, the source may be forced to reload data
00:42:09kurzhaarrockerThat currently is a problem: When you rewind to a point the playback starts and the data played back is freed instantly. If you now rewind to the start point the data may have been overwritten and may need reloading.
00:43:00[IDC]Dragonyes, but for which "window" do you want to optimize?
00:43:25[IDC]Dragoncontinuous play is the most likely one, I'd say.
00:44:47Laurent_does 0.51307588 look like a reasonable executable compression factor to you ? looks so to me
00:44:48kurzhaarrockerLets optimize for continuous play
00:45:45kurzhaarrockerJust if we loop we should calculate wether all the loop fits into the buffer. The source knows its buffer and if it controls the loop points, too there's no problem.
00:46:24kurzhaarrockerOther applications (like speech synthesizers) are a completely different beast.
00:47:14BC|codefor spch syth, the most logical option is to store a permanent cache of phonemes
00:47:28kurzhaarrockerThats what I meant
00:47:42BC|codesorry, not following CLOSELY
00:48:02Laurent_given the nature of phonemes, they might even be MP3 coded in realtime
00:48:29BC|codefor blending?
00:48:42Laurent_(combination of simple sine waves -> should give pretty simple MP3s no ?
00:48:50kurzhaarrockerBut the source could consist of two sources: like the real mp3 source and a control source. The control source tells from where in the buffer to jump where to.
00:49:02kurzhaarrocker(mp3 source == buffer)
00:49:14Laurent_BC|code: if I understood well, the archos can only play mp3 sounds ?
00:49:25BC|codewhy encode realtime?
00:49:48Laurent_to gain memory, but it's only a thought, that might proove not feasible
00:50:29Laurent_i'm not an mp3 expert at all so consider this unrealistic ;)
00:50:34BC|codestoring wavs to be encoded or generating phoneme sounds would likely need more memory than ?43? 1/10S samples
00:51:02BC| a LOW low sample rate
00:51:31MTcall me a cynic
00:51:39Laurent_ADPCM might be used to reduce memory for a very small CPU footprint
00:51:46Laurent_memory consumption
00:51:55MTbut the quality/naturalness of the speech doesnt need to be fantastic
00:52:03MTit just needs to be understandable
00:52:22BC|codefor anything other than MP3 we need to get deeper into the MAS hardware
00:52:26MTso blending sine waves and then encoding to mp3 to play (if it were even remotely possible)
00:52:32MTseems worthless tome
00:53:04BC|codeblending would be nice ...but this will become a realistic option when we get WAV playback
00:53:07Laurent_BC|code: I meant ADPCM to store phonemes to be encoded
00:53:46BC|codeand a super-basic mp3 compressor that will run realtime at 12MHz?
00:53:52[IDC]Dragonback again
00:54:04Laurent_BC|code: seems difficult ;)
00:54:05kurzhaarrockerThe normal dma would end on a ENDPOINT. Each time a DEI occurs the sink could ask the source where to jump to a STARTPOINT
00:54:29BC|codeLaurent_: everything is difficult until you know how its done
00:54:51[IDC]Dragonthinking more about it, the bitswap is nasty, because it "destroys" or data
00:55:23[IDC]Dragonwe can play it again, if we note that it's swapped, but its more or less unreadable to the software
00:56:10Laurent_I used to read something equivalent in people's signatures : "any sufficiently advanced technology is virtually indistinguishable from magic"
00:56:34kurzhaarrockerBut for reading the bitswapped data by software realtime isn't really necessary. We could make reading functions that do the backbitswapping.
00:56:44kurzhaarrockeroh, forget it. stupid idea
00:57:26[IDC]Dragonfor audible FF/FR, this is necessary, I think
00:57:37[IDC]DragonI wonder how Archos does it
00:57:59diddystar5by everyone
00:58:18 Quit diddystar5 ("Leaving")
00:58:36kurzhaarrockerIf all our purpose is to pass the data to the mas we needn't care that it is already bitswapped.
00:59:01kurzhaarrockerThe start / end points are untouched
00:59:10LinusN[IDC]Dragon: archos doesn't care about playing exactly from the frame start
00:59:14[IDC]Dragonif we want to parse it to find the blocks, we do
00:59:25kurzhaarrockerWhy should we?
00:59:39LinusNfor better sound, i think
00:59:41kurzhaarrockerThat should be done before it is put into the buffer
00:59:55LinusNless clicking
01:00:26kurzhaarrockerYes, but the source can define the point on frame boundaried and afterwards we can bitswap whatever we want.
01:00:37[IDC]Dragonmaybe we should do early bitswap then
01:00:55[IDC]Dragonand have the parser read only such data
01:01:15kurzhaarrockerWhy do wa want to parse data in the buffer?
01:01:24LinusNfinding the frame starting points is time/cpu consuming
01:01:25kurzhaarrocker The points can be calculated before
01:01:51[IDC]Dragonto feed, say, every 1 out of 5 frames
01:02:06[IDC]Dragonfor FF/FR while listening
01:02:08LinusN[IDC]Dragon: bad move
01:02:23kurzhaarrockeryou mean for ffwd? I'd ignore frame boundaries while ffwd.
01:02:39[IDC]Dragonor every 5 out of 25
01:02:50LinusNsounds better
01:03:32LinusNactually, the mas is pretty good at resynchronizing, so we could go for an "X bytes" approach to start with
01:03:36kurzhaarrockeryes but while ffwd..
01:04:25[IDC]Dragoncan you tell it the next data is out of sync?
01:04:42kurzhaarrockerIt mutes it automatically, I think
01:05:05kurzhaarrockerIt ignores data without a frame sync header
01:05:41[IDC]Dragonbut mp3 is not safe from "start code emulation"
01:06:03[IDC]Dragonmeaning, data that looks like a header, in the payload
01:06:14[IDC]Dragonbut it's a start.
01:06:55[IDC]Dragonas LinusN says.
01:07:11kurzhaarrocker[IDC]Dragon do you mean an mp3 frame could contain data that looks like frame sync header?
01:07:36kurzhaarrockerAnd believe LinusN, he knows much more about mp3 than me
01:07:49[IDC]Dragonto resync with mor confidence, you have to checkthe interval to the next
01:08:08[IDC]Dragonbut you can never be 10% shure
01:08:40 Join DJBaz [0] (
01:08:49LinusNalso, even if we start playing at a frame boundary, the sound may still be bad because of the bit reservoir
01:08:53kurzhaarrockerBut that is only a concern while ffwd. If we have audiable ffwd on cost of occasional glitches I'd buy it. For everything else we don't really need realtime frame sync detection
01:09:20[IDC]Dragonbit reservoir also applies to the loop
01:09:23 Quit AciD (Connection timed out)
01:09:37[IDC]Dragonyou may have a click there
01:09:50kurzhaarrockerThat's true, but here too I'd say let it click.
01:09:55LinusNmy frame boundary finding code works pretty well, so we can use it
01:09:58[IDC]Dragonbut who expects it seamless?
01:10:12BC|code[IDC]Dragon: are pixels taller or wider?
01:10:31 Join AciD [0] (
01:10:44[IDC]Dragonthey are more dense in X-direction
01:11:03LinusNignore me
01:11:08[IDC]Dragonpixel aspect ratio is 0.8
01:11:10kurzhaarrockerSeamless looping isn't necessary. But i crave for a split edtitor that splits on frame boundaries.
01:11:27LinusNkurzhaarrocker: do you use my functions for that?
01:11:38kurzhaarrockerI saw something, yes.
01:11:49BC|code[IDC]Dragon: thanks, yes I recall 0.8, but wasn't sure which way w=0.8h :)
01:12:02kurzhaarrockerBut momentarily I don't use anything but the time info
01:12:03LinusNthey find the next split point when switching files when recording
01:12:21[IDC]Dragonwell, I'm out for tonight
01:12:30kurzhaarrockerOh, you mean the option to disable the bit reservoir?
01:12:30LinusNcu [IDC]Dragon
01:12:38LinusNkurzhaarrocker: no
01:12:45BC|codenite Dragon
01:12:56[IDC]Dragoncu logbot
01:13:07 Quit [IDC]Dragon ()
01:13:16LinusNkurzhaarrocker: but you should really consider using that option
01:13:37LinusNthe "independent frames"
01:13:45 Join Guest [0] (
01:13:50LinusNhi guest
01:14:10kurzhaarrockerLinusN I thought that was disabling the bitreservoir
01:14:15LinusNit is
01:14:34kurzhaarrockerThe recording option? I have it on always. :)
01:14:57kurzhaarrockerBut I saw the comment in mpeg.c 1984
01:15:21kurzhaarrockerAnd then I thought. What a pity, it's not a function.
01:15:52kurzhaarrockerstupid me: did not see mem_find_next_frame
01:15:56Guestdoes anyone have the recorder plugin for a archos Gmini120? after paying so much for it im not too fond of giving them more money so i can use a mic thats already part of the player..
01:16:43LinusNGuest: sorry
01:16:46kurzhaarrockerThere's no gmini port of rockbox yet as far as I know
01:16:56Guestoh alright
01:16:58 Quit Guest (Client Quit)
01:16:59LinusNkurzhaarrocker: he's talking about the archos code
01:17:13MTbut it got rid of him already :)
01:17:37BC|codeyou buy a serial number form archos to make your gmini work
01:17:45kurzhaarrockerI see
01:17:59BC|codeif you want "record" buy another reg key
01:18:03LinusNi assume that it is tied to the hardware serial number
01:18:20LinusNso giving away a key would be pointless
01:18:39*kurzhaarrocker is glad he doesn't have that modern stuff.
01:19:53kurzhaarrockerLinusN something else: Are you working on that prerecord thing?
01:20:13LinusNas we speak
01:21:16kurzhaarrockerWhat do you think about a special rockbox file function: trunc an mp3 file at the head and fill it with a padding header.
01:22:48kurzhaarrockerThat way we could use a prerecording file and whenever the disk spins up the head could be trimmed to the prerecord time.
01:23:03kurzhaarrocker-> no real prerecord ring file needed
01:23:43Laurent_oops did not see guest questions
01:23:56 Quit DJBaz (Read error: 104 (Connection reset by peer))
01:24:00kurzhaarrockerLaurent_ shame on you
01:24:01LinusNyou mean to delete all the prerecorded data (except for a defined margin)?
01:24:24kurzhaarrockerand pad it with a reasonable header
01:24:36LinusNcould be lots of padding data
01:24:44LinusNseveral minutes
01:24:57Laurent_kurzhaarrocker: I would have been useless
01:25:31kurzhaarrockerLinusN how big is a fat sector?
01:25:36Laurent_the plugins seem to be machine specific : you give the serial number to archos to get access to the downloading url
01:25:49kurzhaarrockerLaurent_ I don't believe that you're useless.
01:25:52LinusNkurzhaarrocker: a sector in the ATA world is 512 bytes
01:26:03BC|codepacket sniffer anyone?
01:26:46kurzhaarrockerSo 512 bytes is the granularity and in the worst case the prepended file header had 512 padding bytes.
01:27:01Laurent_kurzhaarrocker: I meant I could not have fulfilled his requests
01:27:03kurzhaarrockerThat's not minutes of mp3 data
01:28:44LinusNkurzhaarrocker: i don't get it
01:29:36kurzhaarrockerImagine you have a regular mp3 file and want to cut of the header. We don't need to rewrite the entire file. We could modify the fat so that the file starts at a different cluster.
01:30:09LinusNnow you are talking clusters all of a sidden
01:30:25kurzhaarrockerIn case that we need prepend a complete id3 tag or something we could reserver a few extra clusters and relink the file.
01:30:39LinusNa cluster can be up to 64kbytes
01:31:02kurzhaarrockerSorry my understanding of ata / fat is to limited.
01:31:43kurzhaarrockerThe smalles chunk a file has to be- that is a cluster, i now assume?
01:31:53LinusNstill, you want to prerecord into a file and then cut the beginning when recording for real
01:32:20LinusNno need to do it in realtime, is it?
01:32:37LinusNjust remember the position ion the file and trim it afterwards
01:33:04kurzhaarrockerImagine you have a trigger level and the jukebox prerecords for 2 weeks.
01:33:52kurzhaarrockerWe might want to say: In a week there may be a bang and we want to record the 5 minutes before it. Without hd overflow.
01:33:56LinusNok, so you want to continuously trim the file?
01:34:19kurzhaarrockerYes, whenever we spin up the drive anyway would be a good moment.
01:34:27 Join earHurts [0] (
01:35:20LinusNkurzhaarrocker: interesting idea
01:35:24kurzhaarrockerWhich is the same time recorded data is transferred from buffer to hd.
01:35:32Laurent_well, the cache on the hitachi travelstar seems to be a write buffer only
01:36:22kurzhaarrockerLinusN: is that the same kind of 'interesting' as in "This tastes - erm - interesting."?
01:36:38LinusNLaurent_: error 500 - unable to process request
01:36:43Laurent_so no hope of using it as an MP3 buffer
01:37:05Laurent_LinusN: bad URL ?
01:37:10LinusNkurzhaarrocker: no, more like "hmmm, i'm glad i didn't have to cook this myself" :-)
01:37:59kurzhaarrockerLinusN: so if you were Zagor you'd have said "Show me the code"..
01:38:13LinusNkurzhaarrocker: something like that :-)
01:38:17Laurent_then go to the link at the bottom
01:39:52earHurtsHey Linus
01:40:42LinusNearHurts: yes?
01:40:43 Join baz [0] (
01:40:55earHurtsjust saying hi
01:41:28LinusNoh, hi
01:41:52earHurtsMy archos is still tits-up, so I just bought a 512MB sd card for my zaurus
01:42:33kurzhaarrockerApropos sd card: Does anybody know a provider for CF -> IDE adapters?
01:43:10earHurtsthere are some cf cards that are usb masters
01:43:40kurzhaarrocker??? What will a slave do when connected to such a master?
01:44:00kurzhaarrockerand cf and usb?
01:44:27earHurtsit'll allow you cf slotted device to read a usb slave
01:45:33earHurtssuch as allowing a zaurus to read an archos drive
01:46:09MTyou could take an archos and replace the hd with a bunch of solid state memory
01:46:21MTi bet that would have kicking battery life
01:46:29kurzhaarrockerInteresting. But not what I'm searching for. For on the road recording I'd just like to exchange the harddrive of the jukebox with something without mechanics.
01:46:48MTexactly what i just said would do the trick sells them
01:47:18 Quit earHurts (Remote closed the connection)
01:47:31MTand probably almost any component store tbh
01:47:46Laurent_what would be nice would be an archos + some 64MB flash memory usable as a cache
01:48:09MTcan we use both channels on the ide interface?
01:48:29 Join earHurts [0] (
01:48:38kurzhaarrockerMT: That wouldn't work for me.
01:48:48MTwhy not?
01:49:43kurzhaarrockerAs I said: I want to record directly onto a memory device without mechanics. I don't need anything when I have to record to hd first and than transfere it to another memory medium
01:50:19MTwith no hard drive in the archos at all, there would be no mechanical elements
01:51:06kurzhaarrockerYes and nothing to store data on either -> thus exchange the hd with a cf card and some memory
01:51:20 Quit earHurts (Remote closed the connection)
01:51:25MTyou would have a CF card acting as the hard drive
01:51:57kurzhaarrockerAnd that for I need an adapter CF -> IDE. CF -> USB is useless in this case.
01:52:01 Join earHurts [0] (
01:52:14MTwe are talking at cross lengths, ive not mentioned usb once
01:53:07kurzhaarrockerMT: sorry I was mixing up with a line by earHurts
01:54:11MTalthough those are all for 3.5" drives
01:54:27Laurent_kurz : wouldn't an archos + internal flash memory be the best fit -> record on flash, backup on HD when necessary ?
01:55:03kurzhaarrockerLaurent_ Yes. That's why I'd like a cf -> ide adapter.
01:55:21kurzhaarrockerMT: what a pity
01:55:24MTthere would be no room inside the device for both the hd and a cf
01:55:43MTkurzhaarrocker: thats a mild problem!
01:55:50kurzhaarrockerMT that is'nt necessary.
01:56:09MTno, quite, that was directed at Laurent :)
01:56:35MTdamn psybnc, I hate not having the old autocomplete nick feature
01:56:47Laurent_kurzhaarrocker: what you need is a gmini ;)
01:56:57Laurent_the gmini has a cf reader
01:57:04MTI suppose I should switch to a multi server client :/
01:57:15MTand how much is the activation code? :)
01:57:15kurzhaarrockerLaurent_ but no rockbox.
01:57:26Laurent_kurz : i'm working on it ;)
01:57:58MTLaurent: surely they will just slap you with DMCA threats?
01:58:29Laurent_kurz : i got an answer from a samsung guy, he seems to be willing to help about the CPU if telechips (makers of the gmini cpu) don't want to
01:58:46Laurent_MT: those threats have no legal value in France
01:59:00MTthere are EU equivalents
01:59:06Laurent_not yet
02:00:13Laurent_and anyway, the archos protection is probably similar to the xbox one, and the xbox was hacked within full respect of the DMCA, so it can probably be done too for the gmini
02:00:29Laurent_at least, it's worth trying
02:00:50***Saving seen data "./dancer.seen"
02:00:52MTi see several key differences between xbox and gmini
02:01:01MTbut sure, try away and good luck to you :)
02:01:05Laurent_MT : which are ?
02:01:05BC|codeno Doom for gmini yet
02:01:51MTxbox-linux cannot play xbox games, so it does not affect microsofts target market
02:02:03 Part Asmotaku-neko
02:02:17MTgmini-rockbox would have the same functions that gmini would
02:02:30MTand takes away one of their revenue streams away
02:02:55MTtherefore microsoft are unlikely to pursue xbox-linux
02:03:14MTbut archos would do well to pursue you
02:03:27Laurent_I don't think it's a valid claim, this could be said also of a car in which you put your own radio without buying the constructor's one
02:03:49MTlooking at it from a strictly commercial viewpoint
02:04:03MTi dont think its a valid claim either
02:04:08MTbut ianal
02:04:27MTlawyers spend lots of time and money argueing semi valid claims
02:04:49Laurent_sure, me neither, we'll see ! Anyway, Africa looks like a nice continent to exile oneself to in case of real trouble ;)
02:05:22MTchina is best
02:05:34MTthey care little for software piracy claims
02:05:40earHurtshard to get broadband in afrique
02:05:43Laurent_yup, but I value my freedom somewhat ;)
02:05:59MTchina isnt as bad as people think
02:06:09MTits fairly democratic really
02:06:18Laurent_MT: ask the thibetans ;)
02:06:20earHurtsbut dies Laurant Kabila value your freedom?
02:06:34MThe values your $ :)
02:06:50kurzhaarrockergood night
02:07:00Laurent_not him, but two or three african countries are now real democracies
02:07:10Laurent_anyway, I'm not there yet ;)
02:07:12 Quit kurzhaarrocker ("Trillian (")
02:07:42Laurent_oops got to go too
02:07:47Laurent_it's late there
02:07:56Laurent_c u
02:08:04 Quit Laurent_ ("good bye !")
02:12:06 Quit earHurts (Remote closed the connection)
02:19:54 Quit methangas (" Like's GUI? Then try HydraIRC -> <-")
02:30:43 Quit mecraw__ ("Trillian (")
02:41:59 Quit baz (Read error: 110 (Connection timed out))
02:42:05 Quit AciD (Connection timed out)
02:42:54 Join arspy87 [0] (
02:43:39BC|codehi spy
02:43:54arspy87hey blue, what are you coding now?
02:45:21BC|codeaudio stuff
02:47:04BC|codeI'll have the stereo seperation done and dusted before I go to bed tonight ...then just MDB left :)
02:47:35BC|code2,000 lines of code and counting...
02:48:22LinusNBC|code: have you thought any more about the mdb settings?
02:48:59BC|codeI've pretty much sorted out the advanced stuff in my head, and got a few ideas rattling about the basic interface
02:49:05LinusNBC|code: also, please try to separate the new-style settings patch from the stereo separation patch
02:50:06BC|codemy code's pretty modular, but if you're more specific NOW is the right time for me to change anything you desire
02:50:38BC|codeWhat ideas do you have MDB?
02:50:47BC|code(basic user mode that is)
02:51:39 Part LinusN
02:52:12BC|codewas it something i said?
02:53:40 Join LinusN [200] (
02:54:11LinusNdamn modem
02:54:27LinusNno, it wasn't something you said :-)
02:54:29BC|codedo you not have cable near you, or is it a money issue?
02:54:50LinusNi'm on vacation at my inlaws
02:55:05LinusNmy broadband is at home
02:55:10BC|codelol - I was just joking LinusN, I've never known you to log off without saying goodnight, I had presumed it was modem trouble
02:56:20LinusNabout the separation, i want to be able to commit the stereo separation before the revamped settings screen
02:57:03LinusNi can imagine that we will discuss the look&feel a lot before we agree on it, but the stereo separation is much easier to agree on
02:57:06BC|codeokay ...understood
02:57:40BC|codeto get all the basic controls on one screen has been my main drive
02:57:53BC|codeyou're welcome to look at what I've done so far
02:58:26LinusNcan you send me an ajbrec.ajz for the fm?
02:58:39BC|codeit's a .rock atm
02:58:56LinusNsettings on the rocks
02:59:28LinusNthen send me the rock for the fm
02:59:30BC|codetried to make sure that the code can be included easily though with #ifdefs and the likes
02:59:52BC|codeAh. actually, I have updated the plugin api I will send both
03:00:53BC|code2MB of memory?
03:01:53LinusNtry dcc
03:05:44BC|codehmmm, fm does not like my changes to plugin.h....
03:06:27BC|codelmao - i have no idea how this has EVER compiled
03:08:26BC|codehmmmmm, now playlist-viewer is breaking the compile
03:09:16BC|codeignore me wittering way of saying ...gimme 5 mins to fix the build
03:13:15BC|codethere are no docs yet, so I can tell you all the features etc
03:13:33BC|codefor fractional dB adjustment use "ON" as shift
03:14:05BC|codePLAY from chans will take you to the techie screen
03:14:27BC|codeon+left, on+right do stuff in places
03:14:35BC|codef3 is default value
03:15:10BC|code+/- is not handled correctly in chan-screen ...this is where I am working now
03:16:47BC|codecode is in place for arrow-head highlighting
03:17:05BC|codethink that's about it
03:21:21LinusNi think people would want a slightly larger font
03:28:09BC|codeyeah, by moving the loudness centre frequency onto the loudness line, I can get another pixel vertically on the font :)
03:28:45BC|codebut any bigger than that, it doesn't fit on one screen
03:28:58LinusNis that the main requirement?
03:29:09LinusNi can accept a scrolling version
03:29:14BC|codefor me, yes
03:32:02BC|codeany other comments?
03:33:22LinusNi don't like the hex values in the techy screen, i'd rather have decimal
03:33:31LinusNfixed point
03:33:37LinusNlike 0.01
03:34:25BC|codethen it's back to rounding errors :(
03:34:34LinusNand i can't say that i'ts obvious for me which value is which
03:35:11BC|codehmmm, follow the Rwave to the Rspeaker
03:35:55BC|codewaveform +/- value -> speaker
03:36:34BC|codehow would you do it?
03:37:39arspy87blue: i hate those darn rounding errors!!
03:38:13BC|codethe new audio plugin is designed to make them all go away
03:38:26 Join earHurts [0] (
03:39:01LinusNsince i'm not too concerned about the precision, i would have made sliders, or discrete decimal steps
03:39:10BC|codeyou did
03:39:35BC|codei think we have VERY different goals here :(
03:40:03LinusNwhat is your goal with the techy screen?
03:40:22BC|codeprecision control over the DAC
03:41:11LinusNfor what reason?
03:41:39BC|codeI'm not sure how to answer that question
03:42:38LinusNmy point is, not many people would be interested in entering obscure hex values into the mixer registers
03:42:47BC|codebecause I am not satisfied with the guesswork an features-missing config that exists ...although I appreciate that the current system is designed to dumb it all down for the user
03:42:55LinusNthey would, however, be interested in mixing the channels
03:43:33BC|codewhat volume setting gives line-out
03:43:56BC|codesomewhere between 90...91% on the current system
03:44:21LinusNi agree that volume in percent isn't the best
03:45:01LinusNand is 0db really line-level?
03:45:33BC|codealso tallies with the scope tests that were published on the list
03:49:49BC|codethe biggest issue with the code is that I am unable to use the standard print routines because I have no way to generate the font required
03:50:29LinusNmy issue is that it shouldn't require a special font :-)
03:50:54BC|codebut you dont mind it scrolling!
03:52:37BC|codesoooooo, you may be able to save me a big fat bunch of work....
03:52:51BC|codeshall I code it exclusively as a plugin?
03:53:16LinusNnot necessary
03:53:41LinusNi mean, i was hoping to include it in the main code
03:53:57BC|codeokay - I was worried that the special font would stop that
03:54:05LinusNi was coming to that :-)
03:54:49earHurtsahem, shouldn't api and presentation (eg, font) be separated?
03:54:57LinusNi would like it in the main code if it could use a larger font
03:55:26BC|codeLinusN do you mean ANY font, or would 1 pixel taller be okay?
03:55:48BC|code1 pixel taller is already on the cards afaic
03:56:19BC|codeyou might also note that the statusbar is OFF
03:56:23BC|code(so to speak)
03:56:24LinusNi would like any font, but i can agree on 8 pixels
03:58:18BC|code8 makes it impossible to fit on a screen without a completely different interface
03:58:33LinusNthe width?
03:58:48BC|codewe've got quite a bit of width left to play with
03:59:26LinusNmy goal is not to fit it in a single screen
03:59:37LinusNwithout scrolling
04:00:01arspy87well it's been great watching all of this, but i g2g now
04:00:52***Saving seen data "./dancer.seen"
04:00:57BC|codenot fit on screen without scrolling == fit on screen with scrolling !?
04:00:58 Part arspy87
04:01:05*BC|code misunderstand
04:02:00LinusNi want to be able to use a larger font, and that would probably mean scrolling
04:02:08BC|codesorry, earHurts, not ignoring you, do you understand the problem yet?
04:02:17BC|codeyes, agreed
04:03:05BC|codebut I find the scrolling for sound config vey annoying ...too many years sat in front of a mixing desk
04:03:36LinusNagreed, but what will you do when you add mdb tweaking?
04:03:51BC|codei want to discuss that with you :)
04:04:04LinusNwhere would you fit it?
04:04:10BC|codefirst off play and see a graph of what you are doing
04:04:25BC|codetweak everything at register level
04:04:39BC|codefor the basic (main) screen...
04:04:51LinusNi don't think anyone wants to tweak at register level except you and me
04:05:13LinusNand i would only do it for debugging
04:05:33BC|codethere are a few audiophiles on the group who would play ...but not many, hence it is hidden the seperation screen
04:06:05BC|codebut what set of "limited" options would be useful to the user?
04:06:07LinusNstill, the audiophiles wouldn't want hex values and registers, they want physical units
04:06:56BC|codeThe MDB registers can be expressed nicely in terms of "f" and "db" the bass/treble settings
04:07:14BC|codeI consider my bass/treble settings to be register level editing
04:09:32LinusNthe current bass/treble is also register level :-)
04:09:49BC|codecannot set bass = 1.5dB
04:10:08LinusNah, you mean the resolution
04:10:19LinusNbut there are no rounding errors
04:10:20BC|codeyes - all possible values covered
04:11:23BC|codecurrent system limits you to 1/8th of available options
04:12:35BC|codebtw ...there is no usb code in my plugin ...careful!
04:13:59 Quit edx (Read error: 110 (Connection timed out))
04:14:05LinusNno danger
04:14:19BC|codeok, wasn't 100% sure
04:15:02LinusNit won't enter usb mode until all threads have agreed
04:15:14BC|code:) well done
04:15:38LinusNi did a lot of work on that one
04:16:00LinusNit makes the original firmware look really bad :-)
04:16:07LinusNwhich was the goal
04:16:09BC|codei've NEVER seen it screw up
04:16:44BC|codeI never really played with the original firmware - Rockbox was one of the biggest selling points to me
04:17:03BC|codesame with most of my kit
04:22:42BC|coderight... off to deal with the 2's compliment crap in the seperation screen ...Thought: f2: toggle decmial/hex mode
04:23:04LinusNsure, do you handle overflows?
04:23:36LinusNi mean, 0.5 + 0.6 = 1.1
04:23:59BC|codehmmm, no interesing thought
04:24:20 Join AciD [0] (
04:24:24BC|codecuould change the speaker icon if it happens
04:25:43LinusNsame goes for volume/bass/treble/loudness
04:26:29BC|codeyes, I considered this, but my current thinking is that reality is so far away from the maths that to display this info would be unhelpful
04:27:20BC|codesound is tweaked by-ear, and if it sounds better with +17dB of Bass then so be it! ??
04:27:53LinusNnot my point
04:28:10BC|codeenlighten me.. :)
04:28:27LinusNwith 0db volume and +1db bass, there is a risk of digital distortion
04:28:40 Quit earHurts (Read error: 104 (Connection reset by peer))
04:28:43LinusNlots of talk about this in the forums
04:29:17 Join earHurts [0] (
04:29:38BC|codeperhaps it would make more sense, if you spoke of what might fix it?
04:30:34LinusNsome kind of a) automatic adjustment (yuck) or b) warning
04:31:16LinusNthe player has auto adjustment, and i'm not sure i like it
04:31:28BC|codeit's good sometimes
04:31:43BC|code20mS setting (**now available***) is good on speech
04:32:13LinusNi didn't mean that
04:32:31BC|codeahh, some feature I have missed ! damn my eyes!
04:32:53BC|codeI also dislike "auto adjust"
04:33:41LinusNi mean, on the player, the volume is lowered when you increase the bass above the limit
04:34:38BC|codeSo if you are listening to music at +6dB (quite normal ~= 94%) and you need more Bass, does this mean that you run with a permanent *warning* on the screen?
04:36:13LinusNi dunno, maybe just a splash when you adjust the volume
04:37:58BC|codeCan we step back and think and add it later maybe? For this conversation, I think that the important issue is NOT to auto adjust
04:38:06BC|code...or not?
04:38:27LinusNi think we'll have to make it configurable some day
04:38:34LinusNbut let's leave it for no
04:38:55LinusNi'll have to get some sleep
04:38:59BC|codeok... where do you think the auto-adjust would be helpful?
04:39:18BC|codeheadphones, desktop, hifi
04:39:45LinusNfor the cl00bies that crank the settings to the top and hears the distortion without knowing what they've done
04:40:14LinusNlots of complaints about the "rockbox sound quality" in the forums
04:41:08LinusNthe archos firmware doesn't allow high volume settings
04:41:32LinusNthat's how they have "solved" it
04:41:44BC|code0db is maximum? or did they turn on this auto-adjust feature of which you speak
04:42:06LinusNthe auto-adjust is not a mas feature, it is a rockbox feature
04:42:23LinusNon the player
04:42:37BC|codeaha - that will by why I did not spot it in the mas datasheet
04:42:42BC|codewhy is it not on the recorder?
04:43:01LinusNbecause the 3587f is very different
04:43:17BC|codecould the idea be ported?
04:43:23LinusNof course
04:43:31LinusNthe thing is this:
04:43:58LinusNbass+treble+loudness+volume <= 0db
04:44:12LinusNthat's the requirement
04:44:43BC|codeconsideration to the loudness centre?
04:44:55LinusNso, when you raise the volume, which of bass/treble/loudness will you lower?
04:45:01 Quit hardeep ("[BX] Silly faggot! mIRC is for kids!")
04:45:44LinusNand when you lower the volume again, should you readjust the setting you just lowered?
04:45:53LinusNquestions, questions...
04:46:21LinusNanother approach could be to limit the volume
04:46:47LinusNcrank the bass and you can't raise the volume
04:47:07LinusNmuch easuer :-)
04:47:31LinusNand maybe more intuitive for the user
04:47:49LinusNs/intuitive/explainable/ :-)
04:48:48BC|codeHow about limit volume in wps - in limit disabled *warning"
04:48:58BC|codegimme 1 min to close...
04:49:24LinusNthat would be acceptable
04:49:36BC|codein audio settings srceen allow any volume and display error warnings next to bass & treble
04:50:11BC|codethat's my thing said :)
04:50:35LinusNthe loudness center frequency doesn't affect it, neither does the mdb
04:50:58BC|codehorrible maths
04:51:18BC|codeanyway dude, great chatting - go get some ZZ's
04:51:37LinusNnice talking to you, cu another day
04:51:48BC|codenite m8
04:51:51 Part LinusN
04:56:53 Join AciD` [0] (
05:09:33 Quit earHurts (Remote closed the connection)
05:12:16 Quit AciD (Connection timed out)
05:18:37 Quit AciD` (Connection timed out)
05:55:12BC|codenite all
05:55:21 Part BC|code
06:00:55***Saving seen data "./dancer.seen"
07:01:19 Join AciD [0] (
07:37:37 Join matsl [0] (
07:57:16 Join midknight2k3 [0] (
07:57:20midknight2k3hi all
07:59:10 Quit adi|home (Connection timed out)
08:00:59***Saving seen data "./dancer.seen"
10:01:04***No seen item changed, no save performed.
10:10:09 Join oxygen77 [0] (
10:21:17 Quit AciD ("")
10:52:36 Join methangas [0] (
11:02:16midknight2k3nite all
11:02:26 Part midknight2k3 ("parting")
12:01:08***Saving seen data "./dancer.seen"
12:15:23 Join edx [0] (
12:20:04 Join Dogger [0] (
12:20:17Doggermorning peeps
12:47:12 Join adi|home [0] (~adi|
12:56:47 Join PoiPoi [0] (
13:33:27 Join christian [0] (
13:33:29 Quit PoiPoi ("ChatZilla 0.9.35 [Mozilla rv:1.5/20031007]")
14:01:12***Saving seen data "./dancer.seen"
14:10:41edxhmmm is it possible that the mailing list doesn't work atm?
14:31:28 Quit christian ("Leaving")
14:34:14 Join Laurent_ [0] (
15:44:59 Join electronic [0] (
16:01:16***Saving seen data "./dancer.seen"
16:02:17 Quit jameson (Read error: 113 (No route to host))
16:10:31 Join jameson [0] (~reichenb@
16:13:11 Quit electronic ("Leaving")
16:43:20 Quit matsl (Remote closed the connection)
16:57:52 Join AciD [0] (
17:14:20 Join mecraw__ [0] (~mecraw@
17:15:51 Join FRiZ [0] (
17:16:11FRiZHeeeeeeeeeeeeeeeello Úvlib˘˘˘˘˘diiiiiiiii
17:16:40 Quit FRiZ (Client Quit)
17:17:01 Nick mecraw__ is now known as mecraw (~mecraw@
17:20:23 Quit BoBB ("Lost terminal")
17:21:14 Join BoBB [0] (
17:31:35Laurent_too late
17:34:47 Quit BoBB (Read error: 60 (Operation timed out))
17:45:56 Join BoBB [0] (
17:57:41 Join diddystar5 [0] (
17:58:19Laurent_hi diddystar5
17:58:30diddystar5hi Laurent_
17:58:52Laurent_nobody's speaking, god damned hollydays ;)
18:01:18***Saving seen data "./dancer.seen"
18:02:10Laurent_I think I'm gonna post my mails within one hour if I don't get any answer ;)
18:06:36 Join diddystar5_ [0] (
18:06:44diddystar5_stuiped aol
18:07:42diddystar5_diddystar5: quit please!
18:09:06Laurent_huh ?
18:10:05diddystar5_diddystar5 is dead, it will leave in a bit
18:10:13diddystar5_aol crashed.....
18:12:28 Quit adi|home (Read error: 104 (Connection reset by peer))
18:15:08 Quit diddystar5 (Read error: 60 (Operation timed out))
18:20:16 Nick diddystar5_ is now known as diddystar5 (
18:22:29 Quit diddystar5 ("Leaving")
18:38:06 Quit scott666 (Read error: 110 (Connection timed out))
18:46:11Doggerany developers here?
18:46:28Laurent_I don't think there are many
18:46:39Laurent_and you should not count me as one ;)
18:46:55Laurent_are you *the* dogger ?
18:47:05Laurent_from Avos ?
18:53:57 Join scott666 [0] (
19:00:58 Quit edx ()
19:04:27 Join edx [0] (
19:19:44 Quit oxygen77 ("ChatZilla 0.8.31 [Mozilla rv:1.4/7]")
19:21:00 Join oxygen77 [0] (
19:33:50 Join electronic [0] (
19:34:03 Quit electronic (Client Quit)
19:40:09 Quit oxygen77 ("ChatZilla 0.8.31 [Mozilla rv:1.4/7]")
19:50:42 Join kurzhaarrocker [0] (
19:51:39kurzhaarrockerc0utta: any progress with user assignable screens?
19:51:53Laurent_hi kurz
19:51:59kurzhaarrockerhi Laurent_
20:01:22***Saving seen data "./dancer.seen"
20:02:50 Join electronic [0] (
20:08:47 Join oxygen77 [0] (
20:09:54 Join top_bloke [0] (
20:23:42 Quit kurzhaarrocker (Read error: 110 (Connection timed out))
20:27:05 Quit scott666 (Read error: 110 (Connection timed out))
20:27:24 Join scott666 [0] (
20:32:44 Join baz [0] (
20:32:52 Join midknight2k3 [0] (
20:33:03midknight2k3hi all
20:34:38 Join trakk [0] (
20:34:57trakkNice clock mid
20:36:00trakkquiet in here
20:36:05trakkhello midknit2k3
20:38:15midknight2k3i'm here
20:38:48trakkhiya midknight3k4
20:39:33electronichow about a basic drawing too?
20:39:41electroniclike an Etch A Sketch machine?
20:41:08midknight2k3that'd be neat
20:41:39electronicu'd use the arrow keypad to move the cursor, and u hold Fx, ON or PLAY to draw
20:41:41electroniclike in Paint
20:42:02midknight2k3hold fx, on and play to draw what?
20:43:24 Quit jameson (Read error: 104 (Connection reset by peer))
20:43:40 Quit electronic ("Leaving")
20:43:56 Join jameson [0] (~reichenb@
20:51:50 Quit trakk ("Leaving")
21:10:32 Join BC|code [0] (
21:10:42 Nick BC|code is now known as BlueChip (
21:11:27midknight2k3hey BlueChip
21:11:44BlueChiphey hey M K
21:12:39BlueChipLaurent_: I have a routine to decompress raw zip data if you need it
21:13:00 Quit baz ("Leaving")
21:16:06BlueChipLaurent_|afk maybe?
21:16:13midknight2k3no idea
21:16:38BlueChiphey dude dcc me your clock code - i wanna look - but feelin' lazy
21:16:38Laurent_i'm there
21:16:40Laurent_ sorry
21:16:47BlueChips'okay dude, no sweat
21:16:51Laurent_BlueChip: it'd welcome ;)
21:16:58Laurent_insert "be"
21:17:18BlueChipthe brain has a nice way of grammar checking for me :)
21:17:44BlueChipby the time I come to the point of working it out, my unconciuss has already sorted out most typos etc
21:18:53BlueChip...although it has an annoying tendancy to strip out "not"s from time to time
21:19:36BlueChipLaurent_: what os?
21:20:21BlueChipif you say DJ for DOS/Win I will send you all the project files etc
21:21:16midknight2k3sorry! was afg
21:21:34midknight2k3BlueChip, it's at the main page.
21:21:44midknight2k3BlueChip, it's at the main page.
21:21:50midknight2k3there you go
21:22:06BlueChipdidn't know about that
21:23:56BlueChipactually, Laurent_: just thought ...try here:
21:27:56Laurent_bc: i can't access yahoo groups, can't sign in, it wont send me a confirmation email
21:28:47 Quit scott666 (Read error: 110 (Connection timed out))
21:30:56Laurent_let me check again
21:31:01BlueChipshould compile under linux okay ...if it does, I'd love a copy of the elf to stick on the site
21:31:22 Quit top_bloke ("The mind is a terrible thing to taste. Wasted 0 seconds online.")
21:31:23BlueChipreally must get a cross-compiler one day
21:31:40Laurent_still nothing...
21:32:11BlueChipnever had trouble myself
21:32:22BlueChipanyway, you got the source now :)
21:33:02BlueChipif you want to look at the source more closely, I think I have a copy that I cleaned up to understand it
21:34:33Laurent_BlueChip: that'd be nice ;)
21:34:44 Join scott666 [0] (
21:34:52Laurent_but it does'nt seem to be zip compression specifically
21:35:55Laurent_(i'm speaking of the archos firmware)
21:36:16BlueChipit was just one to try
21:37:33Laurent_got it thanks
21:38:12BlueChipyou might also want to chat with mediatek DVD hackers, one guy there played with a whole bunch of differet algorithms for compression and checksumming
21:38:23BlueChip...that might be a useful library to have access to
21:38:59Laurent_good idea
21:39:20BlueChipmention my name
21:41:30Laurent_a little question though : where to find them ? ;)
21:41:40BlueChiplol - yahoo groups
21:41:48BlueChip"mt1389" or something
21:42:00Laurent_shit, I need to get account
21:42:04Laurent_an account
21:42:09Laurent_ok, thanks
21:43:14BlueChipthat's it
21:43:40BlueChipyou could also try OneFirmwareForAll
21:44:02BlueChipthat's the ESS hackers, there is a logical cross over of membership between these two groups
21:44:42Laurent_got it too
21:47:05BlueChipjust looked at the MT site ...seems they never got the members to go anywhere with it :( ...shame there were a couple of real smart guys there in the early days :(
21:47:42BlueChip"lol ..OFFA"??
21:48:38midknight2k3its fnny
21:48:49BlueChipokay - it's also really cool firmware
21:50:24BlueChipDVD players
21:51:00midknight2k3and for all dvd players?
21:51:32midknight2k3i dont get it lol.. what models does it run on?
21:51:53BlueChipall ESS based players that do not have an 8052
21:52:02BlueChipall ESS 4xxx based .....
21:52:36BlueChip...which was about 80% of the market for a couple of years
21:53:11Laurent_what does it add ?
21:53:35BlueChipthree years of major hacking
21:53:40BlueChipwhat do you want?
21:53:55midknight2k3so it does basically nothing special?
21:54:16Laurent_lets you skip the stupid fbi warning ?
21:54:28BlueChipUOP disable
21:54:36BlueChip...that was one of mine :)
21:54:56midknight2k3ooh... you can skip the screen that typically displays for about 5 seconds?
21:55:01BlueChip...for the ESS players WITH the 8052
21:55:01Laurent_"you bought or rent this fucking movie but we presume you a criminal and try to make you feel guilty in advance"
21:55:02midknight2k35 seconds gained
21:55:21BlueChipdisney make you watch 10 mins of adverts!!!
21:55:36Laurent_liquid brain ;)
21:55:44BlueChiphow about ...bunch of mpegs on a dvdr
21:55:48Laurent_it flows by the ears the time you get to the movie :)
21:55:53midknight2k3"apex Reloaded"
21:56:25BlueChipthe other good stuff is the multi-standards patches
21:56:51midknight2k3ooh this one looks cool
21:56:52Laurent_this is sickening, commercial products are not even at 10% of what a free software would do
21:57:18Laurent_and they speak of technical progress thanks to concurrence...
21:57:21BlueChipmy ú30 dvd player does more than any other dvd commercailly available
21:58:24Laurent_it fits within a tower box ;)
21:58:34Laurent_it makes monster noise
21:58:44Laurent_joking of course
21:58:58midknight2k3BlueChip, what's so good about it then
21:59:11BlueChipother than the things I have already mentioned?
21:59:20midknight2k3fbi warning disabler?
21:59:46BlueChipor maybe other than the list of things at OFFA?
22:01:25***Saving seen data "./dancer.seen"
22:02:14midknight2k3i dont see much info
22:02:22midknight2k3i guess i need to sign up
22:02:30BlueChipdo you own a relevant player?
22:02:38BlueChipdont bother
22:02:43midknight2k3i just wanna see what it can do
22:02:57BlueChipwhy not write some rockbox code?
22:03:25BlueChipyou get respect for good code :)
22:03:52midknight2k3i want to see what OFFA does that's so great
22:04:02BlueChiploads of stuff you cant have!?
22:04:36BlueChip"don't read beauty magazines, they will only make you feel ugly" Baz Lurman
22:04:36midknight2k3i bet its nothing good
22:04:37midknight2k3thats why
22:05:12BlueChipnah, it's all crap dude, and if you want to feel depressed you cn just think about being dumped or turned down or something ....less typing
22:06:07BlueChipso long as it was updated since xmas it may be accurate
22:06:33midknight2k3whats new at xmas version?
22:06:47midknight2k3it has a hard drive? what?
22:07:09Laurent_sounds very nice !
22:07:40Laurent_I can't believe commercial vendors can't do that by themselves
22:07:47midknight2k3yeah neat
22:08:16BlueChipthey can ...and they will year ...when it's time for you to give them more money
22:08:37BlueChipDVD players have been doing for years what Archos have just started doing
22:09:09Doggerdoing what?
22:09:19BlueChipbut instead of a serial unmber's put the old one in a land-fill an identical one (possibly in a different case)
22:09:43BlueChipDogger: charging to add features that the machine is already capable of
22:11:12Doggerdvd players dont do that
22:11:17BlueChipif you say
22:11:22Doggernever heard of one
22:11:42Doggercant understand anyone who would pay
22:11:47Doggerits like a protection racket
22:11:47DBUGEnqueued KICK BlueChip
22:11:47BlueChip01 but instead of a serial unmber's put the old one in a land-fill an identical 01one (possibly in a different case)
22:11:49Doggerits a scam
22:12:24BlueChipyour dvd player could do 50 times what it does with full-featured firmware
22:12:36Doggerand my cable modem
22:12:45Doggerand if I got round to it I'd write firmware for them
22:13:08Doggerfirmware always lags way behind hardware
22:13:15Doggermost programmers are stupid to add a hdd to your dvd player ...(a) flash firmware; add hard-drive ...or (b) buy new player the same as the old one with a hdd and updated firmware
22:13:43BlueChipit is NOT by accident is by design
22:13:43DoggerI wouldn't add a hdd to it
22:13:50DoggerI'd get a small computer instead
22:13:53BlueChipit was an example!
22:14:02Doggersure I understand
22:14:18Doggermaybe why archos dont want open source
22:14:24Doggermaybe av500 = av300 + new firmware
22:14:51Laurent_damn why do i always make silly mistakes, such as %d to write an hex number???
22:15:07BlueChiphee hee
22:34:40midknight2k3any ideas for coding projects next?
22:36:15Laurent_yup gmini ;)
22:36:20Laurent_rockbox gmini
22:36:35midknight2k3lol no
22:39:47 Quit oxygen77 ("ChatZilla 0.8.31 [Mozilla rv:1.4/7]")
22:39:59midknight2k3as in, visuals to the music?
22:40:10Laurent_ok I think I get it
22:40:16midknight2k3i could make a fullscreen peak meter plugin lol
22:40:20midknight2k3they're all about taken
22:41:20Laurent_my god, archos are dumb
22:42:40Laurent_the file is xored (thanks to "he knows himself") and compressed, but it's compressed 16 bytes by 16 bytes intertwinned with zeros, which reveal the xor key
22:43:04Laurent_either they are dumb, either they did this to help someone trying to unscramble it
22:43:28Doggerthey are dumb
22:43:37Laurent_you can see the same sequence of letters appearing throughout the file
22:43:44Laurent_Dogger: probably ;)
22:43:51BlueChipGTA3/VC uses xor 22
22:44:10Doggergod knows where we'll be when people start using real encryption
22:44:19BlueChipbut to get around it, you need to "circumvent their encryption system"
22:44:27Laurent_sure, if they had used RSA we'd be helpless
22:44:27midknight2k3Dogger, riots are encrypted
22:44:30Doggerxor 22 != encryption IMHO
22:44:58Doggeris base64 encoding encryption?
22:45:01BlueChipwould mean employing someone who could code it
22:45:54BlueChipand as soon as they do ....along comes a hacker who understands it more!
22:46:15Doggerhackers always win
22:46:23Doggerhackers are infinitely better than programmers
22:46:32BlueChipthey'll pass laws to stop that soon
22:47:09Doggerthey won't get away with it
22:47:20Doggerthe DMCA is a perverse law
22:47:34Doggerit thinks the only people who make advances in technology are in universities
22:47:45Doggerthats bizarre
22:48:41BlueChipnoooooooo, but if they keep the pioneering in the universities then the general public will only get the information that the government deems okay ....that way they can MAINTAIN control
22:48:57Laurent_no that's logical
22:49:43Doggervery true
22:49:46Laurent_those who make discoveries are those who are interested in searching, those who are interested only in making profits try to enslave users and are not able to be productive
22:50:04Doggermost universities have technology thats 20 years out of date
22:50:13Doggerdmca is just censorship
22:50:23Doggeramazing that america thinks its a free world
22:50:30BlueChipcensorship if the fulcrum of control
22:50:32Doggersomeone should overthrow its dictator
22:50:33Laurent_that's why it always fails in the end : because there are no more advances and people revolt
22:50:45DoggerBlueChip: very true
22:50:48midknight2k3OVERTHROW THE GOVERNMENT
22:50:54Doggergood idea
22:50:59BlueChipCivil War predicted for the US in 2005
22:51:08Laurent_no, overthrow the one who seek power and money
22:51:09Doggeras more people do more online, the government will become irrelevant anyway
22:51:19Doggerpeople will merge into their computers
22:51:23Doggerand only exist online
22:51:28Doggeroh no... that would be crap
22:51:38midknight2k3overthrow the government
22:51:50Laurent_governements do what people let them do, it's just a matter of educating people to think
22:51:58BlueChipthe internet is run over the american military network ...which is why it grinds to a halt about 8 hours before each military maneouvre
22:52:33BlueChipAOL the biggest ISP on the planet filed the biggest loss of any company ever last year...
22:52:39BlueChipwho bailed them out?
22:53:50BlueChipGates has enough money to BUY the US government ...why hasn't he?
22:55:10Doggeraol is crap
22:55:23Doggeronly some of the internet is in america you know
22:56:01Doggerhow do you know gates hasnt
22:56:09Doggerhe wouldnt announce he has would he
22:56:12BlueChipyou would enjoy looking at the internet Atlas I think
22:56:14Laurent_and that part is lessening every second
22:56:18Doggerhe would just buy them, and get them to do what he wants :)
22:56:35Laurent_no, you're mixing with the oil industrials ;)
22:56:42Doggerah yeah
22:56:57Doggeramerica is the worst country in ther world I think all the world is agreed on that
22:57:02Doggerapart from america itself
22:57:03BlueChipoil is a dying industy
22:57:09BlueChip...and they know it
22:57:14Doggerwhen oil runs out, america will be fucked
22:57:16Laurent_remember, within the 10 biggest firms in the world, the majority are oil companies, they do rule
22:57:22Doggerthey cant even walk to the shops
22:57:29Doggerthey have to drive to malls
22:57:42BlueChipyou can power a car with pretty much anything
22:57:48Doggerthey should put some tax on petrol
22:57:55Laurent_they are dying but they managed bush to invade Irak
22:57:58BlueChipthe engine that runs on tap water is 180% efficient, no that isn't a typing error
22:58:15Laurent_it is a typing error, no efficiency goes above 100% ;)
22:58:23Doggerthey had to invade iraq to get the last of the oil
22:58:31Doggera dying last attempt to keep living
22:58:55Laurent_sure but for the moment they rule
22:59:05BlueChipmaybe we would have been better off if Hitler had made us all work to HIS rules
22:59:12Laurent_and economically they'll be so as long as our economy depends on oil
22:59:14Doggerfor the moment windows rules.... not in 5 years though
22:59:35Doggerat least hitler was honest
22:59:41Laurent_BlueChip: we'll let you try that by yourself if you don't mind ;)
22:59:42midknight2k3lol dogger
23:00:04Doggerbush is prolly as bad as hitler, but just lies and spins stuff so he doesnt look so bad
23:00:59Laurent_I don't think so, bush is not hitler, this is not the same degree, bush still needs manipulation to do what he wants and he probably is of good will
23:01:03BlueChipbut the cameradery of the US populous is incredible compared to many other places I've been and certainly a million fold on uk
23:01:08Laurent_Hitler did what he wanted and was plain mad
23:01:27Laurent_although bush does much harm for sure
23:01:39Laurent_but aren't we a bit off topic there ?
23:01:46Doggernot at all
23:01:54Doggerthis is 'evil dictators' channel isnt it?
23:02:02Doggerbush, blair, etc
23:02:03BlueChipWe are not allowed to like politicians, they stop *us* from being in control - lol
23:02:15Laurent_no, it's rockbox ;)
23:02:15Doggerself government
23:02:27BlueChipI've studying for my BoFH atm.
23:02:29Doggerwe should all vote on everything by email
23:02:42Laurent_self sex, self food
23:02:46Laurent_no thanks ;)
23:02:50BlueChiptodays topic: was Hitler "mad"?
23:03:00midknight2k3just angry
23:03:11BlueChipLaurent_ groosssss dude
23:03:24Doggerno, he was just terribly misunderstood
23:03:26 Join ove [0] (
23:03:43BlueChipno, vote by email!! (joking)
23:03:57Laurent_he was mad and angry because he was misunderstood, that's the story of all psychopaths but he was a psychopath
23:04:29Laurent_oh shit, got to go to bed
23:04:47Laurent_or I'll be dead stone tomorrow
23:05:07Doggerok question of the day
23:05:07BlueChip"stone dead" "dead stoned" is something else
23:05:08Laurent_stone dead
23:05:17Doggerwas hitler a genius.... can you have bad geniuses?
23:05:32BlueChiponly if they disagree with YOUR morals ;)
23:05:34Doggeris the plural of geniususes genii?
23:05:41Laurent_hitler was a genius in some very specific fields such as as a speaker
23:05:51Laurent_but for the rest he was just plain mad
23:07:03Laurent_so ladies and gentlemen, i'll let you rebuild the world while I sleep
23:07:07Laurent_see you
23:07:27Doggeranything you want where your house is?
23:07:33Doggera pyramid?
23:07:55BlueChipahhh, cone theory here we come...
23:08:11 Quit Laurent_ ("here we go")
23:08:52 Join uski [0] (
23:10:11 Nick BlueChip is now known as BC|Code (
23:10:11DBUGEnqueued KICK BC|Code
23:12:37 Join trakk [0] (
23:16:36trakknow we know the LCD can show greyscale using the video plugin, would it be possible to make .rvf files on-the-fly by plugging a Hauppage WinTV USB TV Tuner into the Archos, i..e use the Archos as a video recorder?
23:17:11trakkwhy not?
23:18:40trakkyea the USB is designed to accept data from the PC so it can store data on the HD, it would just be a case of accepting data from the tv tuner instead
23:19:09 Quit c0utta (Read error: 110 (Connection timed out))
23:19:25uskifor many reasons
23:19:29BC|Codeyou need a tv tuner with a master USB controller chip and firmware to the the hard drive
23:19:34uski1) the CPU of the Archos cannot see what comes from the PC
23:19:45uski2) the USB chip is designed ONLY to act as a IDE driver
23:19:53uski3) the usb chip is a slave only
23:20:02midknight2k3hes joking all
23:20:24uskiyes but ppl reading the logs might ask on the mailling list, so it is better to clarify everything ;)
23:20:41BC|CodeI would opt for "pointlessly difficult" rather than "impossible"
23:20:56trakkok modify the FM recorder so instead of receiving FM it receives TV broadcasts :-)
23:21:27uskitrakk impossible too
23:21:30CtcpIgnored 2 channel CTCP requests in 58 minutes and 24 seconds at the last flood
23:21:30*BC|Code starts to weep
23:21:34uskido you know the bandwidth of a TV signal ?
23:21:44midknight2k3BC|Code, are you coding
23:21:49BC|Coderadio receivers do not pick up tv pictures
23:21:51trakkabout 300Mhz?
23:21:53uskiit is impossible to work with it
23:22:01uskiwith the archos CPU
23:22:04uskitoo slow
23:22:11midknight2k3trakk: why must you always do bogus junk
23:22:24uskiif it is even 100mhz, the CPU clock is 12 mhz or so
23:22:34BC|Codealthough with the right external circuitry, you could pick up free-to-air analogue mono audio
23:22:56trakkwell mid its no different to your polyphonic ringtones or USB digital out
23:23:28trakkor your sdd/ddf even
23:23:38midknight2k3THATS NOT M SDD
23:25:30 Join mikeholden [0] (
23:25:32 Quit mikeholden (Client Quit)
23:27:49trakkah well
23:27:51trakknever mind
23:28:13midknight2k3stop it
23:28:31BC|Codei want an external data->tv converter that I can feed with the digital I/O
23:28:35midknight2k3trakk ( has joined #rockbox
23:28:41midknight2k3mikeholden ( has joined #rockbox
23:28:51midknight2k3stop stealing peoples nicks and ... all your stuff
23:29:24trakkgot your iPod yet?
23:29:41midknight2k3rush delivery
23:29:46midknight2k3should get here later afternoon
23:30:05uskiyou bought an iPod ???
23:30:25*uski takes his gun
23:30:38uskiCancel your order.
23:30:50midknight2k3i didnt order one
23:30:56midknight2k3i hate ipods
23:31:29*uski puts back his gun in the box
23:31:52midknight2k3the first comment was YOU
23:32:22midknight2k3it was posted only a few minuets after your request was posted
23:32:46trakklemme see
23:33:27*midknight2k3 has to go
23:33:30midknight2k3bye all
23:34:03 Part midknight2k3 ("parting")
23:35:46 Quit ove ("Leaving")
23:37:55 Quit trakk ("Leaving")
23:40:02 Quit Dogger ("Client Exiting")
23:49:55 Quit scott666 (Read error: 110 (Connection timed out))
23:50:18 Join scott666 [0] (
23:50:40 Join hardeep [0] (1098@

Previous day | Next day