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).

#rockbox log for 2005-11-22

00:00:11Wettdo you know if I can easily desactivate the wps ? it's... Annoying for displaying anything :)
00:00:22*TiMiD starts to crush the tablet and sniff it
00:00:28ioboundbut it is the codec that makes the playback.c fill the buffer, by calling read_filebuf?
00:00:31BagderWett: ?
00:00:36preglowiobound: no
00:01:28preglowiobound: the playback system decides when and how to load the file at its own pace, it just makes sure the codec always has data to read, the codec should not care
00:01:33WettBagder: i'm trying to blit some pixels ont the screen from a codec. I currently c/paste every needed function but i still have to desactivate the wps
00:01:51preglowiobound: this is where we need to update playback.c for mods, since a mod player needs to have the entire file in the mp3 buffer
00:02:07BagderWett: if you do weird things, you have to do it weirdly...
00:02:33preglownot only does it need to have the entire file there, it also needs to preprocess it to the internal format, so mods are right out currently
00:03:22Wett^^ lol, pasting a bunch of prototypes is already wierd. But i'm wondering if ther's a little fontion to // or something
00:03:29ioboundalright. i guess i misunderstood the stuff in codecs.h, i thought the codec itself was supposed to call read_filebuf to get data out from the buffer (drain it)
00:03:40preglowiobound: that it does
00:03:47preglowiobound: but it's not responsible for filling it
00:04:01preglowiobound: which is what i figured you ment
00:04:13 Part Jungti1234
00:04:16ioboundno, but if it never "drains" it, playback.c wouldn't overwrite it, so i thought since nobody else is touching the buffer, couldn't the codec use it as a scratch memory area
00:04:26preglowsure, but please don't
00:04:27WettBagder: cause for now, it doesn't show anything ^^
00:04:30preglowwhy'd you want to do that?
00:05:16ioboundit's the only memory area >512kb available for codecs, isn't it?
00:05:39preglowwell, yeah, you've got a malloc buffer as well, that we also would like you not to use
00:05:40ioboundso i just figured if a codec for some reason needs to work with more than 512kb data simultaneously, it would be the place to put it
00:05:43 Quit ender` (Read error: 104 (Connection reset by peer))
00:05:54preglowjust explain what you're trying to do, and i can explain how i'd rather see you do it
00:06:25ioboundi'm just trying to figure out how the missing pieces of the dumb codec could be implemented
00:06:53ioboundand to simultaneously try to grok how the various pieces of rockbox work together
00:07:06preglowthe way i'd like non-streaming formats to work is for the codec itself to load the file from the disk to the song buffer
00:07:17preglowyou can't just load an xm file and then have dumb play it
00:07:29preglowyou need to let dumb parse the xm to its internal format, then put that into the song buffer
00:07:30ioboundi know, you'd need to duh_load() it or whatever that method was called
00:07:44preglowso i don't want the playback system to load the .xm file at all
00:07:48ioboundand that method probably uses malloc() etc to build up its own internal structure
00:07:51preglowi want it to let the mod codec do the loading
00:08:01preglowiobound: probably, but that's easily fixed
00:08:13ioboundhow? :)
00:08:44preglowiobound: if you need to malloc something, you just use the current song buffer pointer, then increment it with malloc_size so that the next malloc gets free memory
00:09:01preglowand so on until the mod is completely parsed
00:09:24ioboundyes. and the song buffer you're talking about now, that's the same buffer you just previously said was to be treated read-only by the codec, or am i missing something?
00:09:43preglowiobound: yes, but that's also what i'm saying, we can't load songs into it like we do now if we are to support mods
00:09:54preglowiobound: this loading doesn't fit into the current codec architecture
00:10:23amiconnImho non-streaming format codecs should be treated more like plugins
00:10:33preglowamiconn: exactly how+
00:11:03amiconnThe playback engine would just store a marker in the buffer that tells it 'this file is non-streaming'
00:11:25amiconn...and of course not buffer tracks to be played after this one...
00:11:32preglowamiconn: and unless it fills it at once, it'll just have to stop buffering
00:11:35amiconnbecause they would have to be rebuffered anyway
00:11:40ioboundok, but the codec_api already have pointers for open(), so what's stopping the (imagined) dumb plugin from 1. not calling read_filebuf, but use open() by itself, and 2. use request_buffer() as a poor-mans-malloc?
00:12:00amiconnThen, if the previous track ends, it would fire up the non-streaming codec with the filename
00:12:02preglowiobound: because playback.c will fill the buffer for you anyway
00:12:03ioboundor maybe the buffer contains data for the "surrounding" files too ?
00:12:19amiconnThen the codec plays the .mod or whatever
00:12:41amiconnOn return, the playback engine would continue just as if resuming with the next track
00:12:42preglowiobound: this breaks buffering completely, we'll have to spin up the disk for every mod
00:13:03 Quit _FireFly_ ("Leaving")
00:13:05ioboundpreglow, ok, so playback.c fills the buffer with the .mod file, but we could choose to ignore that and just overwrite the memory? Note, i'm not saying it's necessarily the best method, but would it work at all?
00:13:08preglowamiconn: your suggestion too breaks buffering
00:13:15preglowand so none of them are acceptable
00:13:19amiconnyes I know
00:13:28amiconn...but I don't see another way
00:13:33preglowi see one
00:13:47 Quit DangerousDan ("Miranda IM! Smaller, Faster, Easier.")
00:14:06preglowwe need a codec specific loader routine for each format. when you need to buffer a nonstreaming file, you give the loader the current poitner into the filebuf, and a file name
00:14:27amiconnThat might get *very* tricky
00:14:28preglowit parses whatever it needs into the internal format, and passes the now new buffer pointer location back
00:14:40preglowamiconn: i don't see how this is tricky
00:14:43amiconnOur buffer is a ring buffer.
00:14:53amiconnWhat if the non-streaming track would wrap?
00:14:59preglowamiconn: then it wont load
00:15:16preglowit can't wrap
00:15:20preglowthat's not an option
00:15:26amiconnNow tell this to the user. Sometimes his track would play, and sometimes not
00:15:38ioboundawesome =)
00:15:47preglowoh, it will load, but not until there's enough contiguous space in the buffer for it to be loaded
00:15:57preglowwhich is just a case of delayed buffering
00:16:12ioboundwould it be possible to map memory in such a way that the ring buffer appears multiple times in memory space, continuously? then wrapping wouldn't matter
00:16:23preglowiobound: no
00:16:28preglowiobound: we haven't got an mmu
00:17:00ioboundok. does that mean it's impossible to implement some kind of "segmentation fault handler" which would "correct" the memory access?
00:17:19amiconnYes it does mean that
00:17:57ioboundwhat happens if you access memory "out of bounds"? does rbx hang, or is some kind of exception interrupt thing triggered ?
00:18:02preglowbut still, i don't see how that's a problem, if there's not enough space left at the end, fine, we check if there's enough space at the start
00:18:05preglowif not, we delay buffering
00:18:09iobound(isn't that called bus error?)
00:18:19amiconniobound: The bus will hang, and hence rockbox will hang
00:18:32ioboundno error recovery possible except for a manual reset?
00:18:57amiconnThe watchdog timer can terminate a hung bus cycle
00:19:04ioboundoh ok
00:19:26amiconn...but we don't use that during normal operation
00:19:28preglowas far as i can see, this is the best possible solution, only hampered by the fact that we'll need to split codecs into one loader part and one playback part
00:19:34ioboundvery well, thanks for answering all my newbie questions, hope I can put this info to use some day :)
00:19:50preglowiobound: start pestering slasheri to implement my suggestion :)
00:20:16amiconnWe already discussed the loader-player split for codecs
00:20:34amiconnpreglow: Wasn't it you who was against it?
00:20:41preglowsincerely doubt that
00:20:50preglowi've been hoping for that all along
00:21:05amiconnHmm, then my memory doesn't serve me well here
00:21:11preglowi think linuxstb was against it
00:21:38ioboundwith a 32mb buffer you could probably load the entire HVSC sid collection into memory and not have to hit the disk at all
00:21:41preglowbut yes, i think anything other than my suggestion will be wildly irritating
00:21:51Wettwell, time to sleep. night all
00:21:51preglowimagine skipping through a set of chiptunes
00:21:58 Quit Wett ()
00:21:58pregloweach of them very small
00:22:06preglowa big waste to spin the disk up for each of them
00:22:13FingerSoupUhm, speaking as someone who messes around with tracking, how fast is DUMB? would temporarily rendering to a readable format on the HD be too much hackery? ie: decode the entire IT/XM to disk, read the result as an MP3/WAV/whatever, Playback the resulting file, then delete the rendered file?
00:22:25amiconnpreglow: How would you handle midi with your suggestion?
00:22:42ioboundfingersoup why would you want to do that? to avoid having to fiddle with the buffering stuff in playback.c ?
00:22:51preglowamiconn: in the same way, more or less, the midi handler will load patch data into the buffer as well as note data
00:23:01amiconnmidi needs the patchset, but with this loader-player idea, the patchset would be loaded with every midi track?
00:23:04FingerSoupyeah.... Like I said, it may be too much hackery...
00:23:05preglowamiconn: sharing of patches between songs becomes very hard, but i can't see any other way
00:23:21 Join linuxstb_ [0] (
00:23:25ioboundpreglow unless we implement some kind of disk caching thing?
00:23:29ioboundfile cache, i mean
00:23:37preglowamiconn: not unless we implement some mechanism where a loader is allowed to load several files at once
00:23:49preglowamiconn: and midi is such a fringe case that i don't know if i'd bother
00:23:58amiconnHow would the loader know the next file is also midi?
00:24:09preglowamiconn: well, there's a problem
00:24:11amiconnOnly the playback engine knows the whole sequence
00:24:31preglowamiconn: if we are to handle this case efficiently, it can quite quickly become complicated
00:24:38amiconnOf course, it *might* be possible to define a global flag area for the loaders
00:24:55ioboundany loader work should perhaps take into account the possibility for sub-tunes, such as in .sid files
00:24:58FingerSoupI don't know too much about the internals of RB, or the politics of "Rockbox Purity", but if you want playback implemented fast, that'd be the most straightforward way to get tit done for the now, IMO...
00:25:18preglowiobound: oh yes, even perhaps returning a list of songs to the playback layer
00:25:32ioboundfingersoup, i tihnk dumb2wav spent like 30 seconds or so rendering a complete .mod file on my pentium4
00:25:38amiconnThe fastest way to get non-streaming playback to work would probably be an implementation as a real plugin
00:25:51FingerSoupok, so it's too slow. Scratch that idea.. :)
00:25:52amiconnOf course that wouldn't be integrated with the playback engine
00:25:57ioboundpreglow, support for .zip/.rar could probably also be added thten
00:26:12iobound(which sometimes makes sense for .sid / .mod type files)
00:26:23preglowiobound: mwelll
00:26:36preglowiobound: that might still be hard, the zip/rar loader would then also need to handle metadata
00:26:53preglowiobound: get_metadata currently assumes it reads from a file
00:26:56preglowiobound: not memory
00:27:02 Quit tim66 (Read error: 113 (No route to host))
00:27:08preglowas it should be
00:27:18linuxstb_Regarding splitting codecs into loaders and decoders, I seem to remember I was in favour of the idea, but didn't think it was necessary to make the loaders plugins.
00:27:30ioboundyeah i see a lot of challenges there. maybe it should be put closer to the file system layer or something, so you can browse a .zip
00:27:45preglowlinuxstb_: no, but not doing so will require linking rockbox with all sorts of codec libs
00:27:50ioboundlike a regular directory
00:28:07linuxstb_preglow: No - we don't use the codec libs, we do it ourselves.
00:28:12amiconniobound: I wouldn't bother implementing such things on an embedded platform. Would be a performance sucker
00:28:29ioboundamiconn, a .zip loader thingy?
00:28:39preglowlinuxstb_: porting a module reader from dumb to rockbox will be some work
00:28:52ioboundwhy? i used pkzip on a 10mhz 286 back in the day and it worked fine :)
00:28:53amiconnReading from a compressed file would decompress the same data again and again
00:28:54linuxstb_For example, libmad, liba52, the new FLAC and shorten codecs all come without "loaders".
00:29:03preglowlinuxstb_: well, yeah, but those are easy
00:29:09preglowlinuxstb_: parsing a mod or sid file is worse
00:29:26linuxstb_I admit I know nothing about the non-streaming codecs.
00:29:35amiconniobound: Taking how long for 10+ MB?
00:29:41ioboundamiconn i'm guessing you spend maybe 2-3 seconds decompressing a given file from a .zip archive, and then play it for several minutes without loading anything?
00:29:44preglowi do, and i'd very much like the codecs themselves to deal with it, all of them come with loaders
00:30:04amiconniobound: I have some XM that are >10MB
00:30:35ioboundwell that's not very common
00:30:42preglowi for one couldn't care less about zip/rar support, at least not when we'll need significant architectural chancges to make it work
00:30:49preglowiobound: it's not uncommon either
00:31:10ioboundok, i'm just throwing out ideas here
00:31:29amiconnIf someone really wants to listen to tracks he forgot to unzip on the pc, an unzip plugin would be sufficient, and much simpler to implement
00:31:48preglowit's just that if we are to support zips, we need all loader functions to read from memory
00:31:50ioboundamiconn, i was thinking about using .zip for saving drive space
00:31:59preglowwhich translates to us needing to use _HUGE_ load buffers
00:32:14amiconniobound: At what cost?
00:32:31amiconnDrive space is plenty, CPU performance, RAM and battery power are not
00:32:35ioboundi don't know
00:32:47preglowwe'll also need to hack the loaders to be interruptable
00:32:51preglowi sure as hell don't want to do that
00:33:15preglowto me it just sounds like too much work for little gain
00:33:33amiconnTo me it even sounds like work for a loss
00:33:44ioboundalright, alright :)
00:34:06preglowi can see the size reduction as a gain, especially with non-streaming formats which can be compressed
00:34:09preglowbut nah, too much work
00:35:09ioboundbut adding .gz decompression support to a specific loader (such as duh_load()) probably would be doable (when regular codec-custom-file-loaders environment is in place)
00:35:24Kyl3ive been testing with an h300 lcd remote, the "proper" one, and it can turn my h300 on, but not do anything (rockbox logo shows, too)
00:35:26linuxstb_But this is all theoretical until we get a non-streaming codec working as a plugin.
00:35:33 Join San [0] (
00:35:39FingerSoupDo you plan on supporting MO3 style files (MP3 Based samples to save space)?
00:36:02FingerSoupif so is that functionality built into DUMB?
00:36:12preglowFingerSoup: no, it's not
00:36:16ioboundi don't think dumb has support for it but i guess it could be hacked in
00:36:34linuxstb_That means running dumb and libmad together...
00:36:37preglowFingerSoup: that too would need us to substantially alter our codec architecure
00:36:59ioboundpreglow, unless one simply duplicates all the libmad code inside the libdumb code
00:37:05preglowlinuxstb also has a good point, you need to remember we'll be doing this loading while other music is playing
00:37:06linuxstb_Unless you ported dumb to the SH1. (amiconn?)
00:37:08preglowthere might already be a codec loaded
00:37:09ioboundor links it statically to the libdumb thingy
00:37:24preglowiobound: sure, you could do that...
00:37:42preglowiobound: and there'd still be the iram issue, the iram would be taken by the codec currently playing
00:37:50*LedFloyd np: prvi [00:08m/1000Kbps/44KHz]
00:37:58preglowso you can't decode the mp3 samples while another codec is playing
00:38:06linuxstb_I think you would have to do that - link libmad and libdumb together - so you can optimise their joint use of resources.
00:38:10FingerSoupJust a suggestion to look at as a space saving feature as opposed to zip...
00:38:24amiconnlinuxstb: If libdumb is as slow as mentioned, there is no point
00:38:45preglow30 seconds for how big a mod???
00:38:54ioboundi can check the timing now, hang on
00:38:56preglowhow many channels?
00:39:04preglowand how fast a pc?
00:39:17*LedFloyd np: NABJ [02:12m/608Kbps/44KHz]
00:39:47ioboundoh man, i guess i misremembered. dumb2wav spent 4.36 seconds decoding 2nd_pm.s3m into a .wav file on a p4 2.8ghz
00:40:01preglowLedFloyd: sounds like idm names...
00:40:06ioboundalmost off by an order of magnitude..
00:40:21amiconniobound: Playtime of this mod?
00:40:26preglowiobound: then i'd say we have a chance, that's an s3m of considerable size as well
00:40:27amiconnHow many channels?
00:40:34preglowit's about four minutes
00:40:37preglowif i remember corrently
00:40:43ioboundit made a 6:44 .wav file
00:40:59preglowso i'd say definitely doable, then
00:41:06linuxstb_How big is that s3m file (in bytes)?
00:41:09preglowbut someone needs to convert dumb to fixed point first
00:41:14iobound600kb, can't remember how many channels there are
00:41:15amiconnShould be possible to get it working in realtime on coldfire
00:41:24 Quit einhirn ("Miranda IM! Smaller, Faster, Easier.")
00:41:33preglowalso, i got a dumb patch from kaos54 which needs to be merged
00:41:34amiconn(rough estimation: decoding will take 2 minutes on iriver)
00:41:38preglowhe fixed a number of playback bugs
00:41:43amiconnSo no go for SH1
00:41:54ioboundhow fast is an sh1?
00:41:56preglowamiconn: i'm pretty sure sh1 can do 4 channel chip tunes
00:41:57amiconnI don't think >4 chan mods are possible anyway
00:42:05amiconnI want protracker support
00:42:07preglowamiconn: with you doing the coding, that is :)
00:42:14amiconnThat's the most common thin I have
00:42:40amiconniobound: 11 MHz on the recorders, 12 MHz on Ondios
00:43:04amiconnThe players won't get mod support as the pcm codec doesn't support MAS3507
00:44:45preglowtoo bad streaming formats will have a tough time of using the iram
00:44:47preglowat least mods
00:45:30ioboundi can play 2nd_pm.s3m in realtime on my series 60 phone.
00:45:43preglowwhat cpu?
00:45:44ioboundbut i think it maybe mixes only at 8khz and mono
00:45:58ioboundnot sure, it's a nokia 6600
00:46:10 Quit elinenbe (" HydraIRC -> <- Leading Edge IRC")
00:46:48linuxstb_It's an 104MHz ARM9
00:46:51 Join elinenbe [0] (
00:47:04ioboundindeed it is
00:47:45preglowbut yeah, coldfire should be able to play a lot of the mods
00:47:48linuxstb_Is the source available for that player?
00:47:59ioboundmaybe, let me check
00:48:01preglowi don't think you can count on being able to play the worst mellow-d tunes, though
00:48:39ioboundlinuxstb, i got it from here
00:48:48markunWhat a nice sight! no warnings and errors in the build table
00:49:01preglowmarkun: yeah, i noticed, 'tis beautiful
00:49:58linuxstb_If only all the builds actually worked...
00:50:40linuxstb_markun: Are we going to see the gigabeat added to that table?
00:51:14 Quit LedFloyd ()
00:51:44markunlinuxstb_: I'll wait till I have the unit I think
00:51:53markun23 hours remaining on ebay :)
00:52:58 Quit Moos ("Glory to Rockbox")
00:54:59markunI'm hoping I can get rockbox to run on it, because the Gigabeat is pretty useless to me without rockbox..
00:55:26markunall my music's in Ogg and I don't have windows..
00:56:18linuxstb_So you can't transfer music from a *nix machine?
00:56:27elinenbeamiconn: nice commit... i like it!
00:58:31 Quit Mxm`Pas`Bien (Read error: 110 (Connection timed out))
00:58:44 Quit FingerSoup ("Beware of programmers who carry screwdrivers.")
00:58:56 Join Mxm`Pas`Bien [0] (
00:59:22 Quit iobound ("Leaving")
00:59:30hardeeponly 29 files modified? you obviously didn't try hard enough... =)
01:01:23 Quit pinkutank ()
01:02:35***Saving seen data "./dancer.seen"
01:03:48Kyl3Would iPod Rockbox work on a grayscale iPod?
01:05:28linuxstb_If it's a 4G ipod, then it would just need a new LCD driver writing. If it's an older ipod, then it would be more work.
01:05:36linuxstb_But in principle, yes.
01:06:07linuxstb_But the first two models are the Nano and the Color/Photo - simply because those are the ones we own.
01:06:15 Join webguest10 [0] (
01:06:40 Quit webguest10 (Client Quit)
01:09:09amiconnelinenbe: Btw, my commit is in preparation of real backlight simulation and simulator grayscale lib support :)
01:10:37Kyl3I'll use iPod Linux for the time being then
01:10:57linuxstb_amiconn: Did you see the ipod scan I did the other day?
01:11:08amiconnYes, and I already downloaded it
01:11:30linuxstb_Is it OK? It was my first experience with xsane
01:12:02amiconnResolution is more than enough
01:12:21amiconnOne minor thing - the lighting comes from below instead of above
01:12:49amiconnI found that with most scanners you have to place the item upside down to get 'natural' lighting
01:13:14amiconnThe center of the wheel is also a button?
01:13:21 Quit San (Read error: 110 (Connection timed out))
01:13:41preglowyou've basically got five buttons, a hold switch and the scroll wheel
01:13:46linuxstb_And there is a hold switch on the top.
01:13:58amiconnOkay. Hold switch isn't simulated so far
01:14:49linuxstb_The scroll wheel will hopefully give us BUTTON_SCROLL_FWD and BUTTON_SCROLL_BACK events.
01:15:06 Join xameius [0] (
01:15:10linuxstb_But the real effect may be hard to simulate.
01:15:16amiconnSo there is Menu, Left, Select, Right, Scroll_forward, Scroll_backward, and ???
01:16:42amiconnOkay. I guess the scroll events won't activate BUTTON_REPEAT, but just fire repeatedly?
01:17:11linuxstb_Ask preglow. But that sounds a reasonable approach.
01:17:54amiconnThat'll be a special case in the sim button drivers then, but it's a minor thing that shouldn't hold back a preliminary sim
01:18:30amiconnI'll map these events to cursor up&down the same way as the standard BUTTON_UP and BUTTON_DOWN events on all other platforms
01:18:54preglowi think fire repeatedly is going to be the easiest
01:18:57linuxstb_I don't know how the fast scrolling will work - whether we will need to implement a sort of fast-forward event.
01:19:00amiconnMaybe someone adds the ability to use the mouse scroll wheel ...
01:19:04preglowbesides, i want some acceleration handling sooner later
01:19:08preglowwoner how we'll handle that
01:21:32linuxstb_amiconn: Fwd and Back are not always up and down. e.g. in a list browser, fwd is down, but changing the volume, foward is up.
01:21:56linuxstb_But I can't think of a key mapping that can cope with that...
01:21:58 Quit xameius ("CGI:IRC (EOF)")
01:22:39linuxstb_Or if we use the scroll wheel for seeking, it's left and right.
01:23:11preglowi'd like to use it for as much as possible, heh
01:23:13preglowespecially seeking
01:23:33preglowit really is one of the major reasons for me wanting an ipod
01:24:49 Quit markun (Read error: 113 (No route to host))
01:26:19 Join xameius [0] (
01:26:28 Part linuxstb_
01:28:05 Quit xameius ("Leaving")
01:29:39 Join xameius [0] (
01:29:55preglowlinuxstb: that GPIO chart on is a bit lacking for our targets..
01:31:58 Join JvD [0] (
01:32:33 Join markun [0] (
01:34:41 Nick xameius is now known as xameius_ (
01:34:56 Nick xameius_ is now known as xameius (
01:35:13preglowit's not hard to piece it out from the PP5020 page, however
01:35:59 Nick xameius is now known as xamieus_ (
01:36:17 Nick xamieus_ is now known as xameius_ (
01:36:42 Nick xameius_ is now known as xameius (
01:36:50preglowtime to start pp5020.h
01:38:09preglowlinuxstb: it's ok if i don't care about pp5002 at all, i hope?
01:42:13linuxstbpreglow: I think we'll support it eventually. But yes, ignore it for now.
01:44:15amiconnHow do you turn one the ipod?
01:44:27preglowamiconn: press play
01:44:41amiconnAh, so Play is in fact BUTTON_ON too?
01:44:49amiconnWould be easier this way...
01:45:24preglowwell, yes, seems so to me
01:45:45preglowthat's all i've ever used to turn it onm
01:46:58amiconnDo you think it would be good to rename it?
01:47:31linuxstbAny key will turn it on. A long press on play will send it to sleep.
01:48:02preglowlinuxstb: also when it's actually turned off?
01:48:06amiconnI mean whether to rename it in button.h
01:48:13preglowlinuxstb: only thing i've seen that's able to turn it off is podzilla
01:48:14linuxstbHow do you turn it off?
01:48:25amiconnBtw, I have to fix button.h anyway. They are clashing definitions
01:49:04linuxstbamiconn: I would prefer the raw button names to reflect what is printed on the buttons.
01:50:11 Join ghostface24 [0] (
01:53:13preglowGPIO port A has a _ton_ of refisters assigned to it
01:53:17preglowgod knows what all of them do
01:53:58amiconnlinuxstb: Sometimes you called Left 'Left', and sometimes 'Prev'
01:54:05amiconnSimilar for Right
01:54:23amiconnI need to fix that, and I would prefer to stick with Left/Right.
01:54:25preglowport a enable, output enable, output value, input value, interrupt status, interrupt enable, interrupt level, interrupt clear
01:54:31preglowall of them 32 bits..
01:56:32amiconnLeft & Right are the only true universal buttons
01:56:44amiconnThey exist for *all* targets so far
01:56:56amiconn(even the stalled gmini port)
01:57:26linuxstbWhy does it matter that there are universal buttons? I thought they were only accessed via aliases.
01:57:51amiconnIt's easier for the simulator button drivers
01:58:15amiconnFurthermore, BUTTON_LEFT and BUTTON_RIGHT are defined unconditionally in button.h
01:58:50linuxstbNo they aren't - they are universal for both the H100 and H300.
01:58:58linuxstbSorry, yes they are.
02:00:00amiconnWin32 sim for ipod color almost complete :)
02:02:23amiconnIn fact it already runs, just with wrong colours
02:02:29amiconn(missing byte swap)
02:05:35 Quit Kohlrabi ("Leaving")
02:07:08 Quit Kyl3 ()
02:12:07 Join WarrenLeyes [0] (
02:12:55WarrenLeyesthis is the hangout for the rockbox/iriver port??
02:13:44ghostface24just listen
02:15:12preglowthis is the hangout for all rockbox coding
02:15:52markunI found out how to change the endian-mode in the ARM 920T, would it be a good thing the change the mode sometimes?
02:16:32markunpreglow: what do you think?
02:17:17markunchange it to little endian in the ATA code for example
02:17:18*amiconn broke the iriver bootloaders without noticing so far
02:17:21preglowmarkun: how?
02:17:30WarrenLeyesis there a good free IRC client for winxp
02:18:17markunThe 920T has a 'internal coprocessor' CP15. Register 1 of CP15 controls all sorts of things. Bit 7 lets you set the endian-mode
02:18:39markunyou can also disable mmu and things like that
02:19:06WarrenLeyeswell the roxbox works ok on my h340, a little bit slugish at times, i see the codecs are there but i have no sound
02:19:22markunWarrenLeyes: X-Chat:
02:19:39preglowWarrenLeyes: i hope that doesn't come as a surprise
02:19:39 Quit WarrenLeyes ("CGI:IRC (EOF)")
02:19:41 Join WarrenLeyes [0] (
02:20:02preglowi suggest and you open a new windows for urls...
02:20:14preglowand yeah, i hope your not having sound on your h340 comes as a surprise
02:20:22WarrenLeyessorry can you send that link again forgot my popup blocker was on
02:22:20WarrenLeyesdidnt expect to have sound yet but hey.... can only hope ...kind of wish i could help get rockbox up to speed
02:25:15 Quit WarrenLeyes ("CGI:IRC (EOF)")
02:40:10 Join WarrenLeyes [0] (
02:41:30markunWarrenLeyes: figured out how to work with X-chat?
02:41:47WarrenLeyeswow its going to be easy to mod the look of rockbox on my h340 so far so good
02:42:01WarrenLeyeslol.... yea
02:42:26WarrenLeyesanything cool about x-chat i should know about
02:42:40amiconnmarkun: About the francais.lang update... replacing the keylock strings for player is a bad idea
02:42:41cannardmirc is free
02:42:52cannardi thought all irc clients were free?!@
02:42:52amiconnThe have to be short, really short
02:43:05markunamiconn: Sorry, I'll tell Moos
02:43:15WarrenLeyesno not all are ... and the mirc i looked at was not free
02:43:34cannardthough i cant wait for h300 rb, if markun can port to gigabeat, im jumping ship immediately!!! :p
02:43:44markunWarrenLeyes: X-chat has unicode support
02:43:51WarrenLeyesdamn this is the first time using IRC since 1993 LOL
02:44:44WarrenLeyesanyways ... after i get back inside from having a smoke i wanna talk shop here
02:44:48WarrenLeyesso brb
02:46:36amiconnBah, the build server lags again according to the estimate
02:48:07amiconnSlasheri: Something is very wrong with resume.
02:48:32amiconnAfter my playlist ended, I could still resume - it started at the beginning of the last track (!)
02:49:05preglowlinuxstb: i'm going linux now, so if you want a dump of something, i can fix it now
02:49:40 Nick CBM-away is now known as CheeseBurgerMan (n=BurgerBo@
02:50:42WarrenLeyescould that be caused by the bookmark?
02:52:13 Join San [0] (
03:02:11WarrenLeyesgrrr issues setting up sound w/ xchat
03:02:39***Saving seen data "./dancer.seen"
03:10:15 Quit WarrenLeyes (Read error: 104 (Connection reset by peer))
03:10:36 Join Vlad0man [0] (
03:18:07 Quit hardeep ("[BX] Amount of time you people were funny: -50 minutes -55 seconds")
03:18:32 Quit xmixahlx ("blah blah blah")
03:22:14 Quit San (Read error: 110 (Connection timed out))
03:24:19 Join solexx [0] (
03:28:41 Quit Vladoman (Read error: 110 (Connection timed out))
03:34:55 Quit solexx_ (Read error: 113 (No route to host))
03:45:30 Quit DreamTactix291 (Read error: 104 (Connection reset by peer))
03:46:32 Join DreamTactix291 [0] (
04:11:27 Quit xameius ("Leaving")
04:11:58 Join xameius [0] (
04:22:34 Quit Subterranean ()
04:40:09 Quit mordov ("")
04:43:08 Join mordov [0] (
04:52:24 Join San [0] (n=Test@
04:54:48 Join Drumr|doin|HW|la [0] (
05:02:46***Saving seen data "./dancer.seen"
05:06:01 Quit JvD (Read error: 110 (Connection timed out))
05:06:40 Quit DJDD_ (Read error: 110 (Connection timed out))
05:07:38 Join aliask [0] (
05:13:22 Join Paul_The_Nerd [0] (
05:13:24 Quit xameius ("Leaving")
05:32:05 Join ModernExecutive [0] (
05:36:36 Join Jungti1234 [0] (n=jungti12@
05:37:58 Join tvelocity [0] (
05:44:24 Quit tvelocity ("Leaving")
05:45:28 Join Lost-ash [0] (
05:45:33 Nick hshah_ is now known as hshah (
05:49:25 Quit San (Read error: 110 (Connection timed out))
05:51:12DBUGEnqueued KICK Jungti1234
05:51:12*Jungti1234 님은 부재중이십니다.(hey!)
05:51:32CtcpIgnored 2 channel CTCP requests in 2 seconds at the last flood
05:51:32*Jungti1234 님은 부재중이십니다.(hi)
05:51:44***Alert Mode level 1
05:51:44*Jungti1234 님이 부재중에서 돌아오셨습니다.(hi)
05:52:02*CheeseBurgerMan has not clue what that is...
05:52:43*Lost-ash notes that he should probably get around to setting up xchat to use utf16 fonts
05:53:30 Quit ashridah (Nick collision from services.)
05:53:33 Nick Lost-ash is now known as ashridah (
05:54:05 Quit stu3 (Read error: 110 (Connection timed out))
05:54:19Jungti1234I don't know well irc.
05:55:21CheeseBurgerManTo use it: Type in your message, and hit enter.
05:55:28CheeseBurgerManSimple really. :)
05:55:55ashridahCheeseBurgerMan: he's korean, using a translator, it's not that simple :)
05:56:42CheeseBurgerManIt's really not that simple then. :o
05:59:53Jungti1234Korean any user lost H340 and bought H320 but lose it.
06:00:10Jungti1234He is regrettable.
06:00:38*ghostface24 pats Jungti on the back
06:00:41ghostface24it's alright
06:01:45***Alert Mode OFF
06:02:44 Quit CheeseBurgerMan (Read error: 104 (Connection reset by peer))
06:02:59ModernExecutiveyou bought a 340 and lost it. then you bought a h320 and lost it?
06:03:43Jungti1234Not me other person.
06:04:37 Join CheeseBurgerMan [0] (n=BurgerBo@
06:06:19ModernExecutivesure... :)
06:06:50ModernExecutivei don't know of anyone that unfortunate to lose 2 h3x0s
06:08:54DreamTactix291if i lost my H140 i'd cry
06:10:14ghostface24If I lost my H320, I'd probably look for it ;)
06:10:36ashridahmy H140 is surgically attached
06:10:54Paul_The_NerdYeah, my H140 is almost always either in my pocket, or within line of site.
06:10:59Paul_The_NerdErr h120
06:11:03ashridah(well, it may as well be. i never leave the house without its familiar presence in my left pocket
06:11:09*Paul_The_Nerd wishes he had a 140.
06:11:14DreamTactix291my H140 is always at my side
06:11:16DreamTactix291i had an H120
06:11:18DreamTactix291then i got an H140
06:11:24DreamTactix291and gave my H120 to the lovely lady
06:11:26Paul_The_NerdI would, had I the available funds.
06:11:52DreamTactix291that took a while to amass
06:12:06Jungti1234I get H120 into free charge.
06:12:23Jungti1234But, its HDD was broken. :D hahaha
06:12:45CheeseBurgerManReplace the HDD. ;)
06:12:57Paul_The_NerdI'm thinking I'll just stick with the 120 anyway. I barely fill 20gigs right now, and I expect a 40gig HD to be available soon enough.
06:13:02Jungti1234Repair expense is expensive.
06:13:19DreamTactix291oh i could fill 80GB
06:13:27DreamTactix291Jung: do it yourself :)
06:13:30Jungti1234It's 170,000 wons.
06:13:50Jungti1234Is it easy?
06:13:52Paul_The_NerdI don't have that much music.
06:14:09Jungti1234Is strange in look to do.
06:14:25DreamTactix291right now at Vorbis -q6 on the H140 it's not but 22GB
06:14:34DreamTactix291i have more than i put on it
06:15:15Paul_The_NerdSee, I've got about 1400 mp3s of videogame remixes, then another 300 .wvs losslessly compressed from CDs I own, and that's it.
06:15:36DreamTactix291i have 73.5GB of music
06:15:40DreamTactix291about 50GB is wv
06:16:07Paul_The_NerdSee, I just don't have that much
06:16:20DreamTactix291large prog collection
06:16:42DreamTactix291and it keeps growing
06:18:25Jungti1234Rockbox is renewed every day.
06:18:59cannardwhen will rockbox on shuffle be released?
06:19:30Paul_The_NerdI'd say shuffle is pretty unlikely.
06:19:57DreamTactix291not much of a way to interface
06:20:00cannardyeah i figured as much
06:20:01Paul_The_NerdBut generally speaking, the answer to any "when" question is invariably "when it's done" or in the case of unstarted projects "when someone does it."
06:20:29cannardheh, you guys are too helpful, or way to used to stupid questions. either way, my joke fell over
06:20:31cannardsorry :p
06:21:07Paul_The_NerdWell, every time I come in here I see at least one "Any progress on H300?" or "When will H300 be done?" question.
06:21:36cannardyeah, i can imagine
06:21:39Paul_The_NerdMy interest actually lies in the direction of the iPod Nano. I've been thinking about getting a flash based player for a while, and it seems a suitable offering if Rockbox hits it.
06:21:50DreamTactix291we waited a long time for the H1xx port to get functional. now it's their turn to wait
06:21:57Jungti1234My interest is gigabeat.
06:22:00Jungti1234and H100
06:22:05DreamTactix291my interest is just H140
06:22:11cannardyeah, gigabeat rockbox would be cool, i'd buy one
06:22:11DreamTactix291which i'll mod to an H180 sometime
06:22:25Jungti1234If there is money, ipod nano wants to buy.
06:22:25cannardtoshiba gigabeat
06:22:34Paul_The_NerdWhat does it have that an H120/140 doesn't?
06:22:46cannardgoogle x30
06:22:51cannardbeauty for one :p
06:22:57cannardgigabeat x30
06:23:01DreamTactix291i don't like the gigabeat's looks
06:23:12Paul_The_NerdI love the H120 black's looks
06:23:15DreamTactix291me too
06:23:25DreamTactix291i don't like how a lot of DAPs look
06:23:32DreamTactix291i like how the H120/40 look in black
06:23:40Paul_The_NerdI don't really like the gigabeat's looks.
06:23:43Paul_The_NerdDoes it have an LCD remote?
06:23:51cannardi didnt like it in the pics, but when i bought mine it looks alot a better in the flesh
06:23:56Jungti1234I looks like smartest ipod.
06:24:09cannardthen i broke it and now im stuck with a big fat h340
06:24:15DreamTactix291i never liked the ipods looks either lol
06:24:22DreamTactix291hey the H340 is good and featureful
06:24:30cannardim not saying its bad
06:24:34DreamTactix291it's just weak in the aesthetic department
06:24:38DreamTactix291but worth it if you have one
06:24:46Paul_The_NerdYeah, I'm not a fan of the ipod's looks, but the Nano seems to be the best choice in flash players, and I suspect it may stay that way for a noticeable period.
06:24:50Jungti1234H300 and H100 look like so ignorant.
06:24:53cannardbut i broke my h140 about a week before rockbox was at a working point
06:24:59Paul_The_NerdAnd its form factor isn't bad, even if the style aint so great (from my perspective)
06:25:11Jungti1234It's not good design.
06:25:18DreamTactix291i'm not in teh market for a flash player really
06:26:04DreamTactix291i remember when i bought my H120 what requirements i had
06:26:05cannardthe old square gigabeats were good, minidisc sized
06:26:22DreamTactix291it had to play vorbis for one since like all my collection at that time was GT3b2 -q6
06:26:25Paul_The_NerdWhen I bought the H120, my requirements were Ogg Vorbis support, and upgradeable firmware.
06:26:37DreamTactix291yeah pretty much the same here
06:26:48Paul_The_NerdOddly enough I picked the H120 by mistake
06:26:52Paul_The_NerdWell, not really mistake
06:26:57Paul_The_NerdBut I'd made a false assumption
06:27:03DreamTactix291i picked mine after researching it and the karma for about 4 months
06:27:19cannardkarma's were never available in australia i think
06:27:30DreamTactix291the karma had gapless first
06:27:31ashridahi bought it because it was the only one that supported vorbis that was available in the stores here.
06:27:33cannardso it was irivier, ipod or xclef's
06:27:34DreamTactix291build quality was meh
06:27:38Paul_The_NerdI'd read about Rockbox for Archos, and I'd read about the upcoming (at the time) linux based iRiver media center unit, and somehow I combined them in my head to thinking Rockbox was coming to the H120 (and this was back when the H120 was fairly new)
06:27:40ashridahipods were definently out
06:27:44ghostface24I researched the H120, Karma, RipDrive, and IAudio
06:27:50Paul_The_NerdAnd now, a *long* time later, it's true.
06:28:03ghostface24H120 was my pick because it sounded better
06:28:08DreamTactix291i got my H120 back in july 2004
06:28:14ghostface24sounded like a better deal
06:28:15DreamTactix291then in july this year i bought an H140
06:28:25ghostface24got mine around the same time
06:28:27cannardyou could still buy them in july?!
06:28:34Paul_The_NerdI think mine's about two years old right now.
06:28:38Paul_The_NerdMaybe older.
06:28:40DreamTactix291one of their very last
06:28:41Jungti1234H100 is not color LCD.
06:28:53DreamTactix291the colour LCD isn't that important to me
06:28:53Paul_The_NerdI actually like the grayscale LCD
06:29:08cannardi got mine in august 2004 and rma'd it in july 2005
06:29:11DreamTactix291i have a PMP-120 i got cheap for photos and video
06:29:12Paul_The_NerdI can usually read it without having to wake up the backlight.
06:29:16 Join ModExec [0] (
06:29:39Jungti1234Iriver's PMP is not good.
06:29:39cannardbut they gave me a h300 instead :|
06:29:47DreamTactix291Jung: it was cheap
06:29:54DreamTactix291$220 when all was said and done
06:30:05Jungti1234It's bad as is cheap.
06:30:05ghostface24I had mine for I think about 6 months
06:30:16ghostface24had to return and they gave my and H320
06:30:43ghostface24*me an H320
06:30:46DreamTactix291actually i'd probably sell an H340 if i got one in an RMA and hunt down another H140
06:30:58Paul_The_NerdMy friend just RMA'd a 120 and got a 120 back
06:31:07Paul_The_NerdActually, he went through umm... three I think
06:31:10cannardthere were none left by the time i did that, except overseas
06:31:16Paul_The_NerdThe first two times the one he got back was MORE broken than the one he sent.
06:31:18DreamTactix291i got mine from england
06:31:20DreamTactix291and i'm in the US
06:31:21ashridahPaul_The_Nerd: three?! god.
06:31:30*ashridah still has his original H140 back from april last year
06:31:50Paul_The_NerdYeah, mine doesn't even have minor scratches on it.
06:31:59cannardaussie currency gets owned by everywhere else
06:31:59Paul_The_NerdHis girlfriend dropped her purse on his.
06:32:13ashridahthe only damage is a miniscule amount of damage to the paintwork where the backpanel meets the side panel containing the play button
06:32:28ashridah(and a few next to unnnoticable marks on the screen)
06:32:43Paul_The_NerdI have to say though, he should've known better than to have it in the same *room* as her. I've personally witnessed her sit at a computer and have it break after she just moves the mouse. She has terrible luck with technology.
06:32:50DreamTactix291my H120 and H140 still look new
06:32:56Jungti1234All you are different.
06:34:17ghostface24different in what ways?
06:35:32Jungti1234It's different that want.
06:36:31Jungti1234I wait that H100 comes. :)
06:36:40DreamTactix291i'm very particular in my wants/needs
06:36:45DreamTactix291and i don't like accepting less
06:36:46 Join Membrillo [0] (
06:37:03ghostface24I would buy one, but hard to find, and my pockets are empty.
06:37:19Jungti1234me yoo
06:37:30DreamTactix291i was lucky with my H140
06:37:45Jungti1234I get it into free charge.
06:38:09cannardjungti, cant you go to the iriver factory and steal one?
06:38:15cannarderr one for all of us :P
06:38:40Jungti1234I'm not a thief. :)
06:38:51cannardits ok, we will give you money
06:39:02MembrilloJungti, you work in the Iriver factory?
06:39:33Jungti1234I don't work in iriver.
06:39:35Membrillorightio, i think i came in at an awkward point in the conversation
06:39:55Membrilloplease continue...
06:41:07Jungti1234However, I want to work there later.
06:41:28cannardhmm gigabeat f20's going for 250ish on ebay
06:41:34cannardmaybe i should buy one just incase
06:43:00Jungti1234The Korea does not sell gigabeat.
06:43:35cannardjapan is across the road from you though jungti
06:44:08Jungti1234I don't have money. :(
06:44:27cannardsell stolen iriver's
06:44:31cannardbuy gigabeat
06:45:17 Quit ModernExecutive (Read error: 110 (Connection timed out))
06:45:26Jungti1234It's not good word.
06:45:38 Quit ModExec ()
06:46:49Jungti1234I steal or did not pick up.
06:46:55 Quit HCl (Remote closed the connection)
06:47:01cannardhaha i know
06:47:07Jungti1234A certain person gave simply to me
06:47:42 Join HCl [0] (
06:48:12Jungti1234Be not 'Shoju' it's 'Soju'.
06:48:30Jungti1234It's Korean wine.
06:48:44cannardahh i see
06:49:53Jungti1234Konbae means which is toast.
06:50:35cannardkonbae cause you would of bought a gigabeat :p
06:52:15 Quit Membrillo ("CGI:IRC (EOF)")
06:56:22 Nick CheeseBurgerMan is now known as CBM-away (n=BurgerBo@
07:02:47***Saving seen data "./dancer.seen"
07:21:31 Join _FireFly_ [0] (
07:22:37 Join San [0] (
07:31:36 Join KakashiKun [0] (
07:32:54 Join donkyY [0] (
07:33:19 Part donkyY
07:33:33 Join donkyY [0] (
07:33:54 Join Beav [0] (n=awgaagw@
07:40:40 Quit donkyY ()
07:44:11 Join JdGordon [0] (
07:47:09 Quit Paul_The_Nerd ("Chatzilla 0.9.68a [Firefox 1.0.7/20050915]")
07:48:23 Quit JdGordon (Client Quit)
07:48:31 Join JdGordon [0] (
07:52:05 Quit ashridah ("Leaving")
07:53:13 Join Coldtoast [0] (
07:53:39Coldtoasthey all. just noticed the remote no longer works
07:54:00Coldtoaston the h1x0
07:55:23Coldtoastit displays and updates the WPS fine. it's just that no buttons are working
07:55:49Coldtoasthaha. ignore that....
07:56:16Coldtoastupdate: it'd be quite nice if the remote TOLD you the HOLD switch was enabled :)
07:57:07SlasheriColdtoast: Hehe, common problem ;) I started once opening the remote and i thought it was broken because no buttons worked until i noticed the hold was on :D
07:58:06Coldtoasthave you been workin gon the crossfade stuff Slasheri?
07:58:12_FireFly_Coldtoast: there is a wps-tag for the remote-hold :;)
07:58:28Coldtoastoh yeah
07:58:36SlasheriColdtoast: I haven't had time for that but i try the fix the remaining bugs soon and later add the amplitude triggered crossfading
07:58:57Coldtoastoh, it's just that I've noticed over the past couple of weeks it's been much better
07:59:11Slasherihehe :)
07:59:27Coldtoastit hardyly ever fails now on my player
08:00:05Coldtoastused to be it'd be great then suddenly, would stop working for a few changes. Or randomly work and get the ott noiseburst
08:03:24 Quit San (Read error: 110 (Connection timed out))
08:04:09Coldtoastfor wps, it'd be cool if you could nest blocks of WPS code. So, for example, if emote hold is enabled, it just displays a bit fat image to tell you so and if not, it parses the rest of the .wps
08:05:19 Quit Beav ()
08:06:03ColdtoastI was going to make a remote WPS that was just a big battery indicator. that'd be kinda amusing
08:06:54_FireFly_Coldtoast: you can do it already
08:07:05_FireFly_because the remote-wps-file is parsed speratly
08:07:23_FireFly_you can have only the bat-tag and the hold-tags in your remote-wps
08:07:30Coldtoastyeah. I know
08:07:43ColdtoastI just mean I was going to do it, cos it'd be funny
08:08:12Coldtoasthey. I did have a question tho. If I use the same images in teh .wps and .rwps, are they loaded twice into mem?
08:08:54_FireFly_currently yes because the wps and remote-wps are independent
08:09:14Coldtoastcurrently=will change?
08:09:41Coldtoasttho I guess 1bit bitmaps are kinda teeny
08:09:55Coldtoastthat's all I'm using
08:10:06 Join Gibbed [0] (
08:10:08 Quit Rick (Nick collision from services.)
08:10:20 Nick Gibbed is now known as Rick (
08:10:42Coldtoasthmm. actually, you can't use 2b bitmaps anway I suppose
08:11:14Coldtoastyou've really done a great job with the remote wps I think tho
08:12:54Coldtoastthe only tag I think would be nice to add for wps is a font tag
08:13:27JdGordonhowdy ppl... anything happen in the last 2 days with the h300 port?
08:13:30 Quit KakashiKun (Read error: 110 (Connection timed out))
08:14:14_FireFly_Coldtoast: there some thoughts about seperate fonts for each screen and maybe also for each line in the wps
08:14:31Coldtoastyeah. that'd be pretty great
08:18:20_FireFly_moin amiconn
08:18:32_FireFly_just updated my remote-settings-patch ;)
08:26:59 Quit _FireFly_ ("Leaving")
08:28:11 Join Stu3 [0] (
08:28:58 Quit aliask ("Chatzilla 0.9.67 [Firefox 1.0.6/20050717]")
08:29:10 Join webguest18 [0] (
08:29:37 Join ashridah [0] (
08:31:36webguest18morning, I have problems building my own rb firmware. When I call make in the build folder (after ../tools/configure), the cygwin now tries to make the tools. And there's still the error of ipod_fw.c. How can I call the correct make command ?
08:38:53 Join KakashiKun [0] (
08:45:41ashridahwebguest18: temporarily modify tools/Makefile and remove 'ipod_fw' from the TARGETS := line
08:46:00 Join Membrillo [0] (
08:46:13ashridahthat'll stop it from trying to compile it
08:47:01 Quit curio (Remote closed the connection)
08:50:13 Quit Membrillo (Client Quit)
08:52:21Jungti1234I dozed.
08:53:21 Join Membrillo [0] (
08:56:22Jungti1234 : wow..
08:56:35Jungti1234That is Lego!
08:57:07webguest18ashridah: works fine, thanks
09:02:48***Saving seen data "./dancer.seen"
09:04:45 Quit Membrillo ()
09:06:27 Join amiconn_ [0] (
09:08:19 Quit webguest18 ("CGI:IRC (EOF)")
09:11:12 Quit amiconn (Nick collision from services.)
09:11:13 Nick amiconn_ is now known as amiconn (
09:14:45 Quit KakashiKun (Read error: 110 (Connection timed out))
09:19:28 Join ender` [0] (i=ychat@
09:20:36 Join linuxstb_ [0] (
09:21:28linuxstb_webguest18: I'm assuming you are using the devkit? If so, you should upgrade to the new version - that will compile ipod_fw.c without problems.
09:22:43 Join Membrillo [0] (
09:23:47 Join whatboutbob [0] (
09:26:22 Join San [0] (
09:27:24whatboutbobevening folks
09:27:39MembrilloYo brother
09:28:53 Part linuxstb_
09:31:25 Quit San (Read error: 104 (Connection reset by peer))
09:33:54Membrilloanyone online now started work on the H3xx?
09:36:16 Join San [0] (
09:41:52 Part ghostface24
09:46:01 Join B4gder [0] (
09:48:10amiconnTiMiD, if you read the log: Your new quickscreens behave very different from before :-( The real quick operation does no longer work.
09:53:02 Quit Membrillo ()
09:59:00 Join Kaggen [0] (
10:00:57Jungti1234My H300 is downed continuously.
10:01:25KaggenI have a question about the implementation of the lcb_blit functions, anyone here feel like diskussing it?
10:02:18 Join baobab68 [0] (
10:02:37B4gderwe're not really up-to-speed here yet ;-)
10:02:51B4gdereuro people slow morning
10:03:05*ashridah tosses B4gder a caffeinated cookie
10:03:21B4gdermore coffee!
10:03:36KaggenWell, some of us try to code full-time :-)
10:04:08Kaggenthat is what people refer to as *daytime*, not that you don't work as much :-)
10:04:59 Quit San (Read error: 110 (Connection timed out))
10:05:35B4gdernono, you just _are_ at work full-time
10:05:59KaggenYes, thats most definiatly true :-)
10:06:34Jungti1234What is %wd?
10:06:48B4gderbut I actually have to do some real work now...
10:09:27KaggenPerhaps you could just clarify that for the h300, the lcd_blit is supposed to take fb_data in 16bpp format and not 1bpp (8ppb)
10:10:16Kaggenbut the by/bheight is still given per 8 pixels
10:14:19Jungti1234Can no one answer?
10:15:58 Quit goa (Read error: 113 (No route to host))
10:20:54Jungti1234Do Europeans get up late? :P
10:21:05ashridahJungti1234: europeans have jobs :(
10:22:06linuxstbKaggen: Yes, lcd_blit should take the native framebuffer format for the lcd in question - i.e. 16bpp for the H300. But I didn't think lcd_blit was going to be used.
10:23:49ashridahJungti1234: i'm not european
10:23:59ashridahbut i don't work on WPS's
10:24:04Jungti1234i know
10:24:17Jungti1234You live in Australia.
10:24:51linuxstbJungti1234: %wd disables the status bar, %we enables the status bar.
10:25:16Jungti1234aha.. thanks
10:26:12Jungti1234By the way, it's not to website.
10:28:19KaggenThanks, I was a little confused by the fact that for iriver it takes a non-naitive 1bpp. (And that height is specified per 8 pixels even if the most compact format is either 8,4,2,1ppB or 16bpp)
10:28:41linuxstbThat's simply because it hasn't been implemented yet. The prototype was copied from the 2bpp driver.
10:28:48KaggenI think i've implemented it correcly now (for h300)
10:29:26linuxstbAre you planning on using that function for something?
10:30:20Kaggennot that I know of :-)
10:31:04linuxstbThat function is currently only used by two plugins - the chip8 emulator and the Archos video player.
10:31:18linuxstbBut maybe it would be useful for future games or video players.
10:32:23KaggenI felt like implementing it since I reimplemented the lcd_update_rect so that it uses the LCD-controllers "update window" function
10:32:42KaggenThat I posted on the list yesterday
10:34:49linuxstbDid you get any performance improvements with that function? I recall Linus saying the update was quite slow.
10:35:18KaggenWell, I don't have any hardware to test on :-)
10:35:34B4gderthat's the spirit
10:35:39linuxstbThat makes optimisation a challenge....
10:36:12*B4gder has no color mp3 player either... yet
10:36:12KaggenI'm working on a home-brew system that uses a coldfire and an identical LCD-controller
10:36:24linuxstbB4gder: Buy a nice shiny nano....
10:36:36KaggenI only have a lcd-less MUVO myself :-)
10:36:39B4gderthere are so many new targets...
10:37:05B4gderKaggen: well that was _before_ you discovered our lovely little project! ;-)
10:38:14KaggenWell, if I am going to buy one, i kinda have to use rockbox (pre-ported or not :-))
10:40:18 Quit cannard ("2.0 Build 3515")
10:40:52 Join einhirn [0] (
10:42:20linuxstbKaggen: Are you writing your own software for your home-brew system, or are you using Rockbox?
10:43:10 Join abc1 [0] (
10:43:40 Part Zagor ("Client exiting")
10:43:48 Join Zagor [0] (n=bjst@pdpc/supporter/sustaining/Zagor)
10:43:52abc1:) hello
10:44:18abc1i just been surfing to find out about iriver h140
10:44:27abc1anyone know why it was discontiued?
10:45:09abc1i notice rockbox is for iriver, thought someone might know
10:46:15 Join [1]Tomas [0] (
10:46:31B4gderabc1: no one knows
10:46:46 Join webguest09 [0] (
10:46:59linuxstbSeems the ipodlinux people have written an ARM assembler core for gnuboy... :)
10:47:07Jungti1234If edit Progress Bar, what should I do?
10:47:56abc1b4gder: thanks, i was just wondering cos i have one :) it has lots of features the other mp3 players dont
10:48:20baobab68you all can add another person to the "successfully flashed h3xx rb loader" crowd - int'l h320 here. survived ok.
10:48:21B4gderI have one too
10:48:27B4gderrunning rockbox
10:49:57abc1is rockbox free? is it worth getting?
10:50:01ashridahhmm. question. doesn't ffmpeg have an opensource wma decoder? ie, the only problem being that it uses floating point math?
10:50:02baobab68and flashed back, so i could get my custom boot screen back
10:50:14B4gderabc1: as free anything can get
10:50:30B4gderabc1: and imho, it is WAY better than the crappy original firmware
10:50:37ashridahabc1: Iriver have released several players since the H140 came out. they probably discontinued it so people would buy their newer players
10:51:05abc1but i looked at their new players and they dont have all the features which were the reason i bought it
10:51:29baobab68the h300's are ok, except for no WAV recording
10:51:46Kaggenlinuxstb: Yes, I'm porting Rockbox. My home-brew player is very similiar to the iRiver h1xx but with a h3xx similiar lcd
10:52:24abc1ashridah: i just wanted to be sure there wasnt some bad reason for it
10:53:03abc1as after i had mine fixed to remove the noise, its been great
10:54:45abc1(used to make some noise while playing but they fixed that for free)
10:55:21linuxstbashridah: Yes, the ffmpeg wma decoder is open source. Yes, it's floating point.
10:55:38amiconnKaggen: For clarification, lcd_blit() is not supposed to be implemented for high-depth LCDs (colour or 8bit greyscale)
10:56:14linuxstbWhy not? Couldn't it be useful for (for example) video playback or games?
10:56:37linuxstbi.e. the same reasons it is used now.
10:56:41amiconnIt's completely unnecessary
10:57:22amiconnWe need it for low-depth display to pump display-data *as-is* (almost as-is for 2bpp) to the lcd.
10:57:26 Quit t0mas (Connection timed out)
10:58:27amiconnWhile this is useful for video playback and greyscale engine, on high-depth LCDs it won't be useful for video, and greyscale library is completely unnecessary
10:58:52amiconnIt would be useful for *uncompressed* video playback, but that's out of question, imho
11:02:51***Saving seen data "./dancer.seen"
11:03:35ashridahlinuxstb: yeah, i just double checked the codecs page in the wiki. should have gone there first -.-
11:04:28 Quit Drumr|doin|HW|la (Read error: 110 (Connection timed out))
11:04:47abc1b4dger: which one do i download for ihp140? it only mentions archos in download section?
11:07:12linuxstbThe download section is only the official releases - there is no official release yet for iriver.
11:07:59abc1what does it mean by 50% multi-codec architecture implemented?
11:08:03amiconnlinuxstb: - check the green ticks and the footnotes
11:09:13Jungti1234How do I edit progress bar?
11:09:24abc1will it only play half the codecs>
11:09:30amiconnFor video on high colour screens, the video codec should decode into the lcd framebuffer. This is like double-buffering; lcd_update() will display the new frame
11:09:38 Quit baobab68 ("CGI:IRC")
11:10:31linuxstbabc1: The 50% probably refers to recording.
11:10:48linuxstbFor playback, it is more like 95% - there are still some bugs and stability issues to fix.
11:10:57 Join PaulJ [0] (
11:11:37abc1so recording may be limited to half the codecs?
11:12:24amiconnThere are simply no encoders yet. Just plain WAV recording
11:13:45abc1i see it mentions no codecs are fast enough for real-time decoding
11:14:00linuxstbThere is also no "multi-codec architecture" (i.e. codec plugins) for recording. We have one encoder working faster than realtime (wavpack), but that isn't integrated into the recording code.
11:14:23*Jungti1234 님은 부재중이십니다.(In meal)
11:14:37abc1ah so there was no recording codecs in the first place anyway
11:15:18abc1would the real-time decoding thing be an issue?
11:15:42linuxstbWe have real-time decoding. Do you mean encoding?
11:16:49abc1i was reading the bottom of the page
11:17:00abc1it mentions no real-time decoding
11:17:12linuxstbThat's a very old page. Read the SoundCodecs page - especially the status table at the bottom.
11:17:45linuxstbWe have about 10 real-time codecs at the moment.
11:17:51abc1oh right, i just reaced that from a link from the iriver port page
11:18:09abc1oh thats good then :)
11:18:32linuxstbCheck the date at the bottom - 8 Apr 2005. Rockbox has progressed since then...
11:18:49linuxstbBut I'll try and make that clearer.
11:19:21abc1do you run the website?
11:19:29linuxstbIt's a wiki :)
11:19:37abc1oh right :)
11:19:45abc1it all looks like very clever stuff
11:20:27abc1making new firmware that is
11:20:36linuxstbI've just looked on the iriver port page (IriverPort) and can't find a link to the RockboxAudioAPIProposal page... Just a link to the up-to-date SoundCodecs page.
11:20:46linuxstbWhich "iriver port" page are you talking about?
11:22:52linuxstbThe page only links to SoundCodecs. But I see that SoundCodecs has a link to RockboxAudioAPIProposal at the top - I'll remove that link now.
11:23:30abc1and then i clicked on rockbox table of details 'function' titled multi-codec architecture
11:23:47abc1this lefd to the API propsoal page
11:24:12linuxstbOK, thanks.
11:24:25abc1:) no prob
11:24:57linuxstbI think that table's still accurate though - there is a "Codecs" link above the "Multi-codec architecture" link. They are different things.
11:25:23abc1oh yes that link is correct
11:25:42abc1i just got confused after reading the APIproposal
11:27:34abc1the current status table explains all :)
11:29:11abc1thanks 4 your help :)
11:29:23 Quit abc1 ()
11:29:28*Jungti1234 님이 부재중에서 돌아오셨습니다.(In meal)
11:34:27TiMiDamiconn: what is different ? (I my have missed the point with the quickscreen use)
11:36:14amiconnThe quickscreens could be used in two ways before
11:36:48 Join San [0] (
11:37:44amiconn(1) Short press of the appropriate button (F2, F3 or Mode), one or more presses of Left/ Right/ Down to select options, then press the respective qickscreen button again to leave the quickscreen. This method still works
11:38:57amiconn(2) The real quick way: Press *and hold* the appropriate quickscreen button, the select options with Left / Right / Down. Leave the quickscreen by releasing the respective quickscreen button
11:39:09amiconnMethod (2) does no longer work
11:40:51 Join LinusN [0] (
11:40:55TiMiDhi LinusN
11:41:16TiMiDso for method 2 there were no drawings ?
11:42:03amiconnNo. Trying method 2 shows the quickscreen, but it doesn't work
11:42:25amiconnThere was a reason for the somewhat strange button handling in the old quickscreen code
11:42:30TiMiDit shows it ok
11:42:40ripnetukcan we read more than one button at once on remote?
11:42:54amiconnYes, buit you can't switch options, and it doesn't disappear on release
11:42:57linuxstbLinusN: I remember you saying the lcd_update() for the H300 was quite slow - is this still the case?
11:42:57ripnetukto do quicker quickscreen we need to read 2 buttons ?
11:43:14LinusNlinuxstb: yes
11:43:19TiMiDso on iriver it wasn't working "the quick way"
11:43:28TiMiDsince it can't read 2 buttons
11:43:30amiconnIt was
11:43:43LinusNi don't like the current quickscreens
11:44:06LinusNnot even before your change
11:44:30LinusNthey are too limited and are hard to translate
11:44:44linuxstbLinusN: Do you think the update speed could be improved? Or is the current speed the best the hardware can do?
11:44:47ripnetuki thought they were quite neat - very quick if you know them, and also a reminder of how they work if u do it the slow way
11:44:54amiconnThis was just by chance; the iriver button driver can't detect 2 buttons at once, but didn't send BUTTON_MODE|BUTTON_REL either when using this method
11:45:13LinusNlinuxstb: there is one optimization that can be done, but it won't improve it that much
11:45:26amiconnIt worked because Left/Right/Down have higher "priority" in hardware than Mode
11:46:07TiMiDso basically you press the quickscreen modifier (F2/F3/mode) and a button and the parameter ischanged
11:46:12TiMiDdoes I get it ? pressing one of the directions with Mode still held down 'switched' form Mode to direction events
11:46:43amiconnThis method also works in the sim, at least win32
11:46:49TiMiDeven a quick mode press + a button
11:47:26amiconnYou can also e.g. press&hold F2, press down twice, then release F2 again
11:47:57TiMiDhmmm ok I get it
11:48:19TiMiDI will try with the new code to reproduce the same behaviour
11:48:29amiconnI checked code size as well, there was some increase... (less with my const policing)
11:48:47TiMiDincreased a lot ?
11:49:04amiconnwas 300-something bytes on recorder, ~1KB on irivers
11:49:20amiconnNow 272bytes on recorder, ~600 bytes on iriver
11:49:40amiconnNot much, but no decrease either
11:49:42TiMiDwhat made those 1kb on iriver since the code is the same ?
11:50:07amiconnYou included the (unused) F3 quickscreen for iriver
11:50:21amiconnI corrected that as well
11:51:02TiMiDthe idea I had was to make a generic quickscreen handler
11:51:16TiMiDinstead of a specific function
11:51:28 Join Polo_o [0] (
11:51:44TiMiDfor 300 bytes I think it worth it
11:51:57TiMiD(wonce I will make the quick-key ethod work again)
11:52:18amiconnHowever, I agree with Linus that the quickscreens aren't very useful. But as long as they're there, they should work as intended.
11:52:47linuxstbTiMiD: Do you have much work left for your multi-screen support? I don't have a remote, so haven't been following it closely.
11:53:09TiMiDlinuxstb: yes, almost each screen is ported
11:53:10amiconnRecording screen, pitch screen
11:53:18TiMiDwps was ported by firefly
11:53:24CassandraI use the quick screen a lot. My opinion is that they're useful if less than ideal.
11:53:32amiconnwps needs rework, imho
11:53:39TiMiDimho too :)
11:53:50TiMiDwps should be redesigned from scratch
11:53:57TiMiDthe code is too mesy
11:54:20CassandraYeah, in everyone's copious free time.
11:54:25 Join webguest26 [0] (
11:54:44CassandraSometimes Rockbox is a lesson in learning to live with imperfection.
11:55:09Cassandra(I doubt there's a single piece of software out there that isn't in some way.)
11:55:11LinusNCassandra: that applies to *all* firmware
11:55:42LinusNand "imperfection" could be a subjective thing
11:55:55webguest26Is it just me or the Max files in dir browser setting and Max playlist size setting widgets not working correctly in the latest builds?
11:56:08CassandraIf we were going to rewrite WPS from the ground up, we really ought to change the appaling syntax as well.
11:56:19webguest26Max playlist size gets to 200 and stops
11:56:30webguest26Max files gets to 995 then says 100
11:57:04CassandraYou know you have a problem with syntax when your language makes look friendly.
11:57:05 Join Membrillo [0] (
11:57:10ashridahCassandra: i resent that. i wrote the most perfect hello world once.
12:01:42linuxstbpreglow: When you're next in Linux, can you let me know the start and end sectors for each partition on your Nano? (run fdisk, type "u" to change the units to sectors, then type "p")
12:04:18amiconnTiMiD: I noticed that you also added some functionality to the quickscreens - the BOTTON_INV thing
12:06:48TiMiDyes ;) (we cannot hide you anything :p)
12:06:57webguest26Something is definitely weird with the option select widgets. Try scrolling through the scroll start delay option
12:07:08TiMiDI thougth it was more natural
12:07:11webguest26Max is 250, min is 0 m
12:07:24TiMiDat least it's something I want to do each time I'm in the quickscreen
12:07:33webguest26In the middle is 500
12:07:56TiMiDI don't have time
12:08:00TiMiDI have togo
12:08:03ashridahwebguest26: rofl
12:08:20hshahTiMiD, so you gonna add the different image for remote for the usb mode thing?
12:08:27webguest26I can't see anything obviously wrong in option_select_next
12:08:43webguest26opt->option = opt->step;
12:08:47webguest26should work right?
12:08:48TiMiDhshah: I was doing the code for it
12:08:52JdGordonwhat is the plugins coded in?
12:08:54TiMiDbut I've to go
12:08:57hshahTiMiD - ok cool
12:08:59linuxstbJdGordon: C
12:09:09hshahTiMiD, chat to u later then - take care :)
12:09:19amiconnwebguest26: It works here...
12:09:20JdGordonis there a sdk around? i cant seem to find any links on the rb site
12:09:34webguest26Maybe it's my unit.
12:10:06linuxstbThere is no "sdk" as such. The plugins are compiled as part of Rockbox. You need to download the rockbox source code, install the required cross-compiler and add your plugin to the apps/plugins/ directory.
12:10:12 Quit ghode|afk (Read error: 104 (Connection reset by peer))
12:10:22 Join ghode|afk [0] (
12:10:43webguest26I just tried a settings reset and it's still doing weird things with numeric option select.
12:10:55amiconnwebguest26: Oh, you're right. Doesn't work on target, and the sim is not yet rebuilt
12:11:08 Join Febs [0] (
12:12:14*amiconn blames TiMiD
12:14:40amiconnHaha, that's a funny bug
12:14:59LinusNdid it tell a joke?
12:15:12amiconnsizeof(buffer) is always 4 if buffer is a char*
12:15:31LinusNooops :-)
12:16:13amiconnI've seen that mistake often, however I don't get why one would make that mistake. It's sooo obvious...
12:18:17linuxstbMaybe buffer was an array in the past...
12:18:56amiconnIt wasn't
12:18:58amiconnThis code is new
12:19:08B4gderh300 bleeding edge builds are now available
12:19:13amiconn...and passing arrays would be incredibly inefficient
12:19:22linuxstbI assume you're talking about option_select_get_text() ?
12:19:45linuxstbSo I guess the size needs to be passed as an extra parameter.
12:20:08MembrilloAre h300 daily builds going to be compiled?
12:20:30amiconnlinuxstb: Yes, doing that right now
12:20:40Membrilloawesome. Although testing still hasnt been done for int. right?
12:20:58Membrillointernational that is
12:21:36B4gderMembrillo: successful flashes of the int version have been reported
12:21:38CassandraH300 sim support doesn't appear to be working yet (at least under X11). I assume this is a known problem.
12:21:38amiconnWow, this adds 166 bytes to the code..
12:21:50B4gderCassandra: yes, there's no color support added in the x11 code
12:22:03JdGordonMembrillo: ive got rb on my international h320 with no probs
12:22:09linuxstbamiconn: I think that function was cut and pasted - see the end of this diff:
12:22:35linuxstb(possibly ignore me...)
12:22:50MembrilloJdGordon: ok, but im still a bit edgy. Im on my 5th h3xx due to various problems and i think ill wait for an official test
12:23:00CassandraRight, I have to dash. (And see if I can blag some sort of iPod off a friend for Rockbox. ;) )
12:23:13Jungti1234oh thanks
12:23:19amiconnB4gder/ LinusN: What's the status of the installer build?
12:23:32B4gderthere's any change?
12:23:37linuxstbCassandra: Planning on joining the porting effot?
12:24:15amiconnB4gder: The installer doesn't build for over a month now. I did already send the script to LinusN
12:24:30 Join PaulJ_ [0] (
12:24:40CassandraIf I can get hold of the hardware, sure.
12:24:49amiconn(he told me they were deleted in the wiki attack)
12:24:57LinusNamiconn: i'll fix soonish
12:25:11linuxstbCassandra: Any idea which ipod you might be able to get?
12:25:14CassandraI thought we were going to have the installer checked into CVS?
12:25:28Cassandramini or a nano. More likely a mini.
12:25:46Cassandra(My friend *found* a Nano. Some people have all the luck.)
12:25:47linuxstbThe mini will need a new lcd driver writing...
12:26:06MembrilloLinusN: is the bootloader permanent now, or will you continue to adjust and tweak it
12:26:14linuxstbWhich is not trivial - I think the ipod greyscale LCDs pack the pixels differently to the H100.
12:26:27LinusNMembrillo: it will certainly be updated
12:26:48CassandraAnyway, gotta go.
12:26:58 Quit Cassandra (" HydraIRC -> <- IRC with a difference")
12:27:16Coldtoastis the h300 bootloader itself stable? a friend of mine with a h300 wants to use it just in case he "bricks" his h300 again with a corrupt DB
12:27:34B4gdercheck the daily build page now
12:27:41B4gderthe image isn't that nice, but the links work
12:27:49SlasheriMembrillo: and there will come little different bootloader also (with more tweak & debugging options implemented) for people who wants to flash rockbox
12:27:55amiconnlinuxstb: The different pixel packing is the easiest part
12:28:44 Join NicoFR [0] (
12:28:51MembrilloColdtoast: im curious to this as well
12:28:55linuxstbBut doesn't that mean re-writing all the lcd functions? I'm not saying it's difficult work, just that we can't use very much of the existing h100 driver.
12:29:47amiconnYes it does mean that. But while it's a lot of work, it's in fact rather trivial
12:30:26amiconnThere even is a b&w driver with horizontal pixel packing already
12:30:26linuxstbOK, we have different interpretations of trivial. I meant it in terms of it not being a 5 minute job, which it would be if the format was the same.
12:30:51linuxstbThe lcd_update_rect function can simply be copied from an ipodlinux driver.
12:32:01linuxstbWould the greyscale lib be hard to implement? What will be needed for that?
12:32:55B4gderLinusN: there's one question mark for the h3x0 on the DeviceChart, do you possibly know that item now?
12:32:55 Quit ghode|afk (Read error: 110 (Connection timed out))
12:33:08amiconnThe grayscale lib needs a fast implementation of lcd_blit(9 and a programmable timer.
12:33:34amiconnAnd of course we need to know the internal refresh rate of the LCD, but that can be found by experiment
12:33:54MembrilloI just read about the H3xx bootloader, HristoKovachev: Just Tested on H340 international with 1.28K (korean) firmware...... Was this test successful???
12:34:09LinusNB4gder: will doublecheck that
12:35:05Membrillo^^^^ was the 1.28k test successful?
12:35:32linuxstbMembrillo: Good question. I wondered the same thing when reading that comment.
12:36:04MembrilloLinusN: do you have an answer? its at the bottom of the iriverboot page
12:36:06Jungti1234but there is bug.
12:36:12B4gderI think so, there was a kr h3x0 picture in the MR forums iirc
12:36:25webguest26Is the win32 simulator makefile supposed to work at the moment?
12:36:36webguest26(this is under Cygwin)
12:37:07linuxstbwebguest26: Yes. Which directory are you typing "make" in?
12:37:09webguest26make: *** /home/Dan/rockbox/uisimulator/win32: No such file or directory. Stop.
12:37:18LinusNMembrillo: you need to hold ON until the iriver firmware starts
12:37:22webguest26in my buid-simulator directory
12:37:28webguest26build even
12:37:41linuxstbwebguest26: Have you checked out the uisimulator directory from CVS? It's at the same level as apps, tools, firmware etc
12:37:42B4gderwebguest26: you don't seem to have the win32 directory!?
12:37:51webguest26good point
12:37:56MembrilloJungti1234: Thanks for that
12:38:23webguest26after a quick cvs co it works now
12:40:57 Join wehn [0] (
12:41:21MembrilloLinusN: i dont think you got what i was asking. I was asking if the test by HristoKovachev (as quotes on the iriverboot page) was successful
12:41:23 Quit PaulJ (Read error: 110 (Connection timed out))
12:42:45 Quit San (Read error: 110 (Connection timed out))
12:43:52Jungti1234Tested me.
12:44:05Jungti1234-> I tested it.
12:44:42MembrilloJungti1234: you got 1.28K?
12:46:18JdGordonMembrillo i did it on 1.28k on a non-us h320
12:48:21 Quit wehn ("ChatZilla 0.9.61 [Mozilla rv:1.7.12/20050920]")
12:49:46MembrilloJdGordon: sweet. ill most likely try it this weekend
12:50:02JdGordonreally no point
12:50:11JdGordonit just makes listenign to music a pain :D
12:50:16Membrillowhy so?
12:50:25B4gderhey you h3x0 users, get a pic of your player and upload it! ;-)
12:50:32JdGordonwell the mp3 playback isnt done
12:50:38B4gderwe need one for the daily build page
12:50:48Membrilloill do that now
12:50:49wubblaharhar, i'd have already installed rockbox on my h320 if i weren't that chicken-hearted :P
12:50:50Jungti1234yes 1.28k
12:50:58Membrilloill take a picture now
12:51:19ashridahyou people are pansies. i had rockbox on my H140 player ages before it could play music :)
12:51:26Jungti1234Do not.
12:51:30ashridahdidn't stunt me none
12:51:38thegeekhaha ashridah
12:51:55thegeekeveryone is saying dont do rockbox on the h3xx yet
12:51:57*ashridah notes that was a tense day
12:52:01thegeekand here you come
12:52:04ashridahthegeek: i'd agree with him
12:52:31amiconnB4gder: I have a rather high-res H3x0 pic
12:52:45amiconn(the one I used to make the H3x0 win32 sim background)
12:52:59B4gderanything is fine, I'll scale it to fit
12:53:14ashridahmeanwhile. i think i'm in love with eclipse.
12:53:24ashridahpity the CDT plugin isn't perfect :(
12:54:08JdGordonis it supposed to load the irvier fm instead of rb if te power is connected when its turned on?
12:54:41ashridahthe bootloader probably doesn't contain a hook in the firmware for that entry path
12:54:59ashridahmy guess is that the cpu loads at a different address (or possibly triggers an immediate interrupt) that linus hasn't caught yet
12:55:17ashridahconsider it a saftey net :)
12:56:32JdGordondoesnt help if the flash doesnt work :D
12:57:08ashridahwell, if the flashing process doesn't work, blame iriver, it's their flashing code. if the flashed image fails to load, THAT is potentially a rockbox bootloader issue tho
12:57:13JdGordonoh crap.... is there a pdf reader plugin?
12:57:19amiconnB4gder: Got it?
12:57:53Membrilloi just took a photo of my h3xx
12:58:08ashridahMembrillo: i think someone's beaten you to it
12:58:57Membrillomine has the same background colour as the rest of them lol
12:59:39Jungti1234H300's wps is small size.
13:00:20amiconnwebguest26: Try the latest bleeding edge :)
13:01:01 Join PaulJ [0] (
13:01:09webguest26will do
13:02:18 Join Drumr|doin|HW|la [0] (
13:02:38B4gderamiconn: got it, now I only need to find where the heck this xchat puts the file...
13:02:47 Nick Drumr|doin|HW|la is now known as Drumrboy|away (
13:02:52***Saving seen data "./dancer.seen"
13:02:55webguest26amiconn: works a treat, thanks
13:03:25webguest26Guess I should give up trying to fix it myself. Guess I'll need to find some other part of rockbox to poke at.
13:03:36webguest26too much guessing
13:03:36MembrilloB4gder: if you want to still see my h3xx pic i have it here
13:03:41linuxstbB4gder: Probably in ~/.xchat/
13:03:47B4gdernot on windows
13:03:58B4gderthey haven't invented the home dir yet ;-)
13:04:42B4gder(I'm forced to use windows at work)
13:05:01ashridahB4gder: uh, profiles often serve that purpose. doesn't help if the app's too stupid to put anything in there, tho :)
13:05:04dwihnoDocuments and settings! Documents and settings! :)
13:05:28dwihnoThe funny thing about "Documents and settings" is the tendency to create several folders for every user.
13:05:35dwihnoExample "user" and "user.computername"
13:05:38dwihnoreally funky stuff
13:05:49B4gderhm, nope the file is just...gone
13:06:02ashridahB4gder: search for a folder called 'dcc' i guess
13:06:17B4gderI searched my C: for the file name
13:06:24amiconnB4gder: Huh? Of course windows has a homedir
13:06:39amiconn(at least real windows has, meaning 2000/ XP/ 2003)
13:06:44ashridahB4gder: is the profile set to another location like a network drive?
13:07:26B4gderamiconn can I get that image from a URL somewhere?
13:07:47MembrilloB4gder ive got mine uploaded
13:08:13Membrilloill just grab the url
13:09:04Membrillook its....
13:09:06 Quit PaulJ_ (Read error: 110 (Connection timed out))
13:09:51Membrilloyou might have to crop it a bit sorry
13:10:48linuxstbB4gder: If you're collecting images, you may want my ipod scan:
13:10:55linuxstbIt's not perfect, but it's the one amiconn used for the sim.
13:12:26B4gderlet me know if/when you want the ipod builds to appear among the daily builds
13:13:22 Join mashalla [0] (
13:13:34linuxstbOK. I guess we'll do that when the port reaches the stage the H300 port is at now.
13:14:02MembrilloB4gder: i centered it and changed it a little is the image if you want to use it
13:18:01Membrilloi gotta run. see ya'll
13:18:17 Quit Membrillo ()
13:18:58B4gderthat image is a bit too low quality for my taste
13:20:22B4gderxchat is not too clever
13:20:39B4gderI have a non-existing directory set for storage of received dcc files
13:21:03*B4gder bounces
13:21:06B4gderno, its there
13:21:24*B4gder says no more
13:21:38LinusNgotta fly, cu later
13:22:48 Join StrathAFK [0] (
13:22:50 Quit Mark__ (Read error: 110 (Connection timed out))
13:22:54 Part LinusN
13:23:27 Quit thegeek (Read error: 104 (Connection reset by peer))
13:23:49 Join thegeek [0] (
13:25:38B4gderok, I need to edit that pic a bit before I'll use it
13:25:56B4gderhere now:
13:27:19thegeekanyone got a "valid" us postcode
13:27:24thegeekneed it for a braindead website;)
13:27:44dwihno90210 :D
13:27:47 Quit JdGordon (Read error: 110 (Connection timed out))
13:27:56dwihno(Beverly Hills) ;D
13:28:25thegeekgot the same tip from another channel
13:28:46thegeekreally nice automated streaming radio
13:29:04thegeekautogenerates playlist from your favorite artists and stuff
13:29:07thegeekseems to work rather nice too
13:29:20linuxstbWhat quality does it stream at?
13:29:50thegeekor that's what it sounds like to me
13:29:59thegeeksadly it's flash-based
13:31:22Nixsosi wanna try rockbox on my H340 with EU firmware... can i just patch the original FW with the FWpatcher from the site?
13:32:58Jungti1234 <- This is different from my H300.
13:33:24B4gderhow different?
13:33:37B4gderthey are available in different colors
13:33:55Jungti1234Pattern and Is letter of 'H340'
13:34:18Jungti1234latter? um..
13:36:46 Quit Strath (Read error: 110 (Connection timed out))
13:36:46 Quit Febs ("CGI:IRC (EOF)")
13:38:04ashridahhm. wasn't iriver forced to rename their iHP-xxx series to Hxxx during the H3xx lifecycle?
13:49:02Zagorthey renamed it, don't know about "forced to"
13:49:20preglowi think it resulted from a dispute with hp
13:49:35Jungti1234They did not force.
13:50:43Jungti1234The Korea does not call that is 'iHP series'. Call that is 'H series' only.
13:52:35ripnetuktimid - u there? is anyone looking at the fact you cant delete files from remote yet? we are missing the 'are you sure' screen, and more importantly, the check for play on the remote to say 'yes, im sure'...
14:00:02 Part webguest26
14:00:21 Join goa [0] (
14:02:37Nixsospfew, just installed rockbox on my H340 with EU firmware... booted perfectly, just when i tryed to start the original firmware, the screen went black, disk spinning but no lcd response
14:03:14ripnetuki think the ad-hoc implementation of the yes/no screen in onplay.c should be replaced with a call to yesno.c's code
14:03:28ripnetukwhich is already ported to use remote afaict
14:04:08amiconnYes, looks like an oversight
14:04:42ripnetuknever sure if I should fix stuff like this when others are working en-mass on remote support...
14:06:10 Join frederic [0] (
14:07:10amiconnIt would be cool if someone with a silver-coloured H100/H120 and someone with a H3x0 could do a scan while rockbox shows the logo
14:07:27amiconn(like the other supported units in the device hart)
14:08:24amiconnHmm, and I should probably redo the H^140 scan with the greyscale logo
14:09:53Jungti1234Is place that can buy HDD cheaply?
14:10:01wubblahmm....all iriver's use exactly the same CPU?
14:10:37B4gderthe h1x0 and h3x0 use the same cpu, yes
14:10:48wubblaand why do people always assume that the h300 series is faster than the h100 series?!
14:10:48B4gdeririver do lots of other players with different cpus
14:11:05B4gderwubbla: because they are ignorant? ;-)
14:11:24ashridahwubbla: because it displays 16000% more colour! it must be faster!
14:11:49wubblaah, now that's an argument :D
14:12:02dwihnowubbla: The main reason. People always think everything newer is faster, bigger and ... erhm... better?
14:13:20wubblabtw. to be honest i find these archos-players as ugly as sin :P
14:14:19thegeekbut they are really old too
14:14:29thegeekcan't compare them to modern players
14:14:31B4gderthey were among the first ones
14:14:57B4gderand they use 2.5" disks, which I happen to like
14:14:59wubblathat's how they look like :)
14:15:52 Quit Mxm`Pas`Bien (Read error: 110 (Connection timed out))
14:16:10*amiconn actually prefers the archos look over any ipod.
14:16:12 Join Mxm`Pas`Bien [0] (
14:16:45wubblaamiconn: you're being funny, right?
14:17:03linuxstbEven the Nano?
14:17:07amiconnThe ipods look too non-techie to me
14:17:30dwihno1.8" has it perks, but 2.5" is also nice.
14:18:38wubblaby the way, I'd like to see two versions of the bootloader: one that loads rockbox and another one that loads the original firmware by default...
14:19:43Nixsosyeah that would be nice to have
14:19:56ashridahhm. iirc, the early days of the H1x0 development had that
14:19:58preglowamiconn: i've got a silver h120, but no scanner
14:20:09B4gderwubbla: you can easily build your own
14:20:14linuxstbashridah: I don't remember that. I thought it was always the same as it is now.
14:20:18 Join San [0] (
14:20:45wubblaB4gder: yeah, i know that!
14:20:47ripnetukive always thought we needed a 'boot original firmware' option in iRiver rockbox, its very easy to implement (i implemented it as a .rock when rockbox was unusable in real life)
14:21:20ashridahlinuxstb: it changed from one to the other and back from memory.
14:21:35wubblaB4gder: but i thought that it might be a usable option for other people too...
14:21:51ashridahripnetuk: that's right, you worked on chainloading the original firmware again, or something. did that ever work?
14:22:06ripnetukdo we have a standard API call for displaying messages (like the generic yesno) that works on remote as well?
14:22:27B4gderripnetuk: we have the splash
14:22:30linuxstbashridah: OK. I ran v1 of the bootloader for months. I didn't dare try to upgrade :)
14:22:39ripnetukok cool :) thanks
14:23:19amiconnlinuxstb: Afaik all ipods need a docking adapter for USB connection. No standard USB socket. That's a blocker
14:23:26amiconn(for me)
14:24:24linuxstbIt just needs the proprietory ipod<->USB cable that comes in the box. I agree it's annoying, but I can live with it.
14:24:49linuxstbUSB charging makes up for it (compared to the H1x0).
14:25:07amiconnThere's a custom cable without a bulky docking station? That's already better
14:25:22amiconnI'm not interested in USB charging
14:25:23linuxstbYes, you don't need a docking station.
14:25:42linuxstbJust this:
14:25:53linuxstbOne comes in the box.
14:26:37amiconnQuite big plug
14:26:46linuxstbThat's a third-party one, the apple version is smaller.
14:27:05amiconnEven bigger than the old-archos non-standard (USB A - USB A) cable
14:27:36amiconnNewer archos have USB A - Mini B like iriver
14:28:26 Quit frederic ("Chatzilla [Firefox 1.5/undefined]")
14:28:33linuxstbThis is Apple's version:
14:28:44linuxstbIt looks big, but it actually very thin.
14:29:10 Join merbanan [0] (
14:30:30ender`amiconn: if you need, i could scan my iHP-120 with the RockBox logo showing when i get home (just tell me how to get it to display for about half a minte, my scanner is slow)
14:30:58amiconnI used a custom build for it that waits for a button press after displaying the logo
14:31:23ender`can you send it to me?
14:31:55amiconnHmm, I don't have it anymore, but I'll remake it
14:32:08amiconnIt's rather trivial, just a one-liner in main.c
14:33:55ender`i'd have to set up dev environment for that first :)
14:35:35merbananhow many bits do the mac unit on the iriver coldfire have ?
14:36:46preglowplus 8 extension bits
14:36:59preglowonly 32 bits are easily obtainable
14:38:08preglowwhy do you ask?
14:38:24merbananok so Q16.12 would be a good radix representation ?
14:38:45merbananpreglow: fixedpoint stuff
14:39:03preglowyou've got room for a couple of more bits
14:39:35preglowit all depends on if you're going to use fractional mode, of course
14:40:49Jungti1234Did lunch do all?
14:42:50merbananpreglow: I was thinking of just regular integer math, I thought that 16+24 would give maximum resolution without overflowing
14:43:37preglowso you're just going to use 32x32->32 muls?
14:44:24merbananyes, that was the idea
14:45:04preglowyou don't really need the emac unit for that, unless you're going to do a lot of accumulates as well
14:45:13 Join Lost-ash [0] (
14:45:29 Quit ashridah (Nick collision from services.)
14:45:35 Nick Lost-ash is now known as ashridah (
14:46:24merbananpreglow: fft will need that I guess
14:46:45linuxstbmerbanan: Are you planning on something interesting for Rockbox?
14:47:27preglowmerbanan: nah, fft doesn't do too much accumulating
14:48:46merbananlinuxstb: well maybe, first I wanna do an integer mdct transform, then port it to some codecs in ffmpeg
14:49:01linuxstbI guess WMA is on your list then?
14:49:17linuxstbHave you looked at the new Cook decoder?
14:49:30merbananlinuxstb: well you could say that
14:49:37merbananI wrote it
14:49:41linuxstbhehe :)
14:50:04linuxstbNice work.
14:50:46linuxstbDo you think Cook is feasible on the iriver?
14:50:50preglowcook decoder?
14:50:59merbananand sorry it's not well adapted to be played on portables, the buffering requirements are large
14:51:50*ashridah looks curiously on
14:51:56linuxstbMy understanding is that the frames are out of order in the input stream?
14:52:03merbananthe bitstream needs around 200-300 kb of reordering
14:52:10merbananlinuxstb: correct
14:52:14linuxstbWhy is that?
14:52:26preglowmerbanan: you mean just a genereal mdct?
14:52:30preglowmerbanan: variable size?
14:52:38linuxstbDoes the ffmpeg demuxer do the reordering now?
14:52:56linuxstb(I've been half-following the discussions on the ffmpeg-devel list)
14:53:04merbananlinuxstb: I guess it is some interleaving scheme, to reuce the impact of dropping 10 sequential frames
14:53:28merbananpreglow: yes a general one
14:54:10merbananlinuxstb: the demuxer does that now yes
14:54:31preglowmerbanan: so you're going for an fft based one, then?
14:54:48merbananpreglow: yes, kiss fft
14:55:01preglowmerbanan: well, i'd pretty much urge you to use 64 bit maths, you'll probably end up with precision problems if not
14:55:17preglowkiss fft isn't exactly fast, but it works, at least
14:55:24preglowwe need some mad person to write us a fast fft
14:55:54amiconnDoesn't fft imply 'fast' ?
14:58:12merbananpreglow: :/ I was hoping Q15.12 would be enough, then 40 bits would be enough
14:58:16 Quit NicoFR (Read error: 104 (Connection reset by peer))
14:58:23 Join NicoFR [0] (
14:58:23 Join gromit`` [0] (n=gromit`
14:58:56ripnetukhi, i have fixed 2 issues with rockbox - 1. the delete file screen didnt use the standard gui yesno and splash routine, and hence didnt work on remote and 2. yesno didnt respond correctly to remote buttons... can i submit a patch somehow??
14:59:11 Quit gromit` (Read error: 104 (Connection reset by peer))
14:59:14 Quit TiMiD (Read error: 104 (Connection reset by peer))
14:59:45NicoFRI took a picture of my H320 with the rockbox logo showing :
14:59:47ashridahripnetuk: the pack tracker on ?
14:59:58ripnetukis that he best way to submit stuff then?
14:59:59ashridahs/pack/patch/ even
15:00:05ashridahstupid 1am. making me all type no good
15:00:23merbananlinuxstb: how many bits do the ipod mac have ?
15:00:26ashridahripnetuk: well, it notifies pertinent people and lets you replace it with newer versions, etc
15:00:26preglowlinuxstb: fast, compared to the naive dct
15:00:35preglowamiconn: fast, compared to the naive dct
15:00:53ripnetukok... the bit about relavent people sounds good :) its quite a trivial patch tho
15:01:05preglowmerbanan: why not use full 32 bit coefs? the emac unit supports getting the top 32 bits of a mul
15:02:12linuxstbmerbanan: preglow probably knows better, but I think there's a 32x32->64 multiplication available in the ipod's arm7tdmi.
15:02:13 Join TiMiD [0] (
15:02:20preglowit is
15:02:40 Join speed123 [0] (
15:02:54 Quit San (Read error: 110 (Connection timed out))
15:02:56***Saving seen data "./dancer.seen"
15:02:57merbananpreglow: full 32 bit coefs? how ?
15:03:02linuxstbPlus a 32x32->64 mulitply and accumulate.
15:03:23preglowmerbanan: emac lets you do 32x32->64 bit multiply, and get the top 32 bits
15:03:51preglowmerbanan: that's how almost all our current codecs use emac
15:04:01merbananQ15.16 ?
15:04:13 Join muesli_- [0] (
15:04:28preglowwhatever you want, internally the emac unit views them as q0.31
15:04:55preglowso a 32 bit int has range from -1 to +1
15:06:03preglowas long as whatever you want fits the top 32 bits of the answer, it's all good
15:06:19ripnetukdammit do you need to be logged in to sourceforge to submit a patch? i submitted it, and its come up no fuiles :(
15:06:29merbananpreglow: ok, but how does that help if I want to represent larger numbers ?
15:06:35preglowripnetuk: the rockbox patch page isn't updated too often
15:07:06B4gderripnetuk: all subscribers of the rockbox-sf list get notified
15:07:58ripnetuki think i cokced it up... it says no files submitted :( oive registered again... but it takes 24 hours aparently...
15:08:12ripnetuki mean registered not again
15:08:35preglowmerbanan: well, you just need to keep track of that yourself with some shifting
15:08:59ripnetukooops ! should have read ""
15:09:03ripnetukDon't forget to tick the "Check to Upload and Attach a File" checkbox!
15:11:15merbananpreglow: ok
15:11:56ripnetukthat looks better... anyone care to check ive submitted it right? (ive never submitted a patch before)
15:11:58B4gderhey, its been hours since the last commit!
15:12:52ripnetukfeel free to check / commit the remote delete patch i just did :) its quite important to me personally that it works, as I listen to radio shows and delete them when done... my rockbox gets quite messy when I cant easily delete files (lazy!)
15:12:52ripnetuk Enable remote support on iRiver for delete file screen.
15:12:52ripnetuk2 fixes - 1. USe standard yesno and splash routines in
15:12:52DBUGEnqueued KICK ripnetuk
15:12:52ripnetukonplay.c for deleting files and 2. enable remote button
15:12:52ripnetuk(menu) on yesno screen.
15:12:52***Alert Mode level 1
15:12:52ripnetukMy email is I am called George
15:12:53***Alert Mode level 2
15:12:53ripnetukStyles (nick- ripnet or ripnetuk
15:13:06ripnetukdidnt mean to paste
15:13:09mordovwhay doen't the %wd tag function on todays build?
15:13:59preglowmerbanan: what will these fixed point numbers of yours contain?
15:14:34 Join gromit` [0] (n=gromit`
15:14:36merbananmdct coeffs
15:15:23 Quit gromit`` (Read error: 104 (Connection reset by peer))
15:15:29 Quit TiMiD (Read error: 104 (Connection reset by peer))
15:15:39 Quit novimon (Read error: 110 (Connection timed out))
15:19:13 Part speed123
15:22:54***Alert Mode OFF
15:23:20 Quit PaulJ (".")
15:27:27B4gderripnetuk: yes, we noticed both your submissions ;-)
15:27:51preglowmerbanan: they don't need 16 bits of integer part do they?
15:31:12merbananpreglow: no I don't think so
15:31:39preglowi think we've already got a couple of codecs that implement an imdct with fft
15:31:45preglowfaad and liba52
15:35:08 Join wraypa1 [0] (
15:36:45 Join Subterranean [0] (
15:36:45 Quit wraypa1 (Client Quit)
15:37:37merbananliba52 is not so general
15:38:57preglowi love this
15:39:02preglowa simple software interrupt hangs rockbox
15:39:06preglowwhat the hell am i doing wrong
15:40:55 Quit ashridah ("Leaving")
15:41:53 Join webguest98 [0] (
15:42:59 Join skyhawlk [0] (
15:45:07 Part skyhawlk
15:49:33 Quit ripnetuk ("Ninja IRC v1.5.8.1(#1) exiting after 4d3h30m56s of use")
15:49:39 Join ripnetuk [0] (
15:57:00 Quit webguest98 ("CGI:IRC (Ping timeout)")
16:00:24preglowlinuxstb: weee, i've got some corrupted fonts again
16:01:27 Join Kohlrabi [0] (
16:02:29linuxstbpreglow: Congratulations :)
16:03:23 Quit Kaggen (Read error: 110 (Connection timed out))
16:06:08preglowman, this is strange
16:06:24preglowi try to read back the exception vector table, and it doesn't look anything like it should at all
16:11:23 Join Febs [0] (
16:15:36preglowi understand nothing
16:17:11 Quit muesli_- (Read error: 110 (Connection timed out))
16:22:25 Quit NicoFR ()
16:25:05 Quit mashalla (Read error: 110 (Connection timed out))
16:25:23 Quit webguest09 ("CGI:IRC")
16:30:41 Join novimon [0] (
16:35:57 Quit Febs ("CGI:IRC (EOF)")
16:35:59 Join Febs_ [0] (
16:36:07 Nick Febs_ is now known as Febs (
16:37:57preglowi'm starting to think this is cop related
16:38:01preglowi completely forgot about it
16:38:07preglowbut the cop was awakened in the bootloader
16:38:11preglowand we do nothing about it
16:38:19linuxstbThat would explain it.
16:39:43preglowi just don't wake it now, and things aren't corrupted anymore
16:39:46preglowbut anyway
16:39:56preglowthe exception vectors are still right out wrong
16:40:49preglowit looks like the apple vectors are still there
16:42:20 Join muesli_- [0] (
16:42:44 Join San [0] (
16:42:57 Join _FireFly_ [0] (
16:43:12markunI hope the Gigabeat can do 44.1kHz, first I though the DAC was connected to the AC97 interface and could only do 48kHz..
16:43:19 Nick Vlad0man is now known as Vladoman (
16:43:49preglownot the worst news ever if it was true
16:44:13preglowthat cpu is able to do some pretty kick ass resampling
16:44:14markunwell, I would prefer not to resample most of the time
16:44:45markunBut the Wolfson DAC cannot be connected to a AC97 interface I think
16:45:03Jungti1234Is it digitial camera chip? :)
16:45:46markunno, for audio
16:49:02markun7 hour to go on ebay for the Gigabeat F40..
16:49:23 Join Mark_ [0] (
16:49:24 Quit Mark_ (Killed by (Nick collision))
16:49:35 Join Mark_ [0] (
16:49:41 Join Mark__ [0] (
16:49:57markunIf any of you is thinking of bidding.. please don't :)
16:50:19 Join webguest62 [0] (
16:50:57Mark_on what?
16:51:05wubblamarkun: you mean that one? :-) −−>
16:51:08preglowlooks like i'm nesting something up here...
16:51:26markunwubbla: yes
16:51:33merbananpreglow: I get a 1 bit difference on 10 samples of a 512 size mdct now
16:51:54markunYou would seriously slow down rockbox developement for it ;)
16:52:08preglowmerbanan: ok, what do you do currently, math-wise?
16:52:19wubblamarkun: don't worry about it! i already got an h320
16:52:27markunfhew.. :)
16:52:38Jungti1234I will a bidding. :P
16:53:00merbananpreglow: a Q?.12 fft aided mdct transform
16:53:28preglowmerbanan: no 64 bit muls?
16:55:01wubblai just wanted to inform you, that the win32 simulator is broken in CVS...
16:55:16wubblastubs.c: In function `simulate_usb':
16:55:18wubblastubs.c:165: error: too few arguments to function `usb_display_info'
16:55:24 Quit Mark_ ("Leaving")
16:55:31merbananpreglow: well it is a 32*32->64 mul shifted down, but I'm only using max 8 bits over 32
16:55:56preglowmerbanan: shifted down how much?
16:56:09preglowmerbanan: right
16:56:41preglowthen we'll need to fiddle with extension bytes on coldfire
16:59:37 Quit muesli_- (Read error: 104 (Connection reset by peer))
17:00:58amiconnwubbla: It's not broken both here and in the official build system. Perhaps you have an old stubs.c?
17:02:11wubblaoh...i need to update the uisimulator separately...
17:02:17wubblasorry 'bout that
17:02:32merbananpreglow: if I try to use a int32_t it doesn't work and if I use Q?.8 with int32_t the resolution of the fractional part is to low
17:02:52amiconnWonder why that is... when I do cvs up -dP in my working copy root, everything gets updated
17:02:57merbananpreglow: shifted with 12bits
17:02:58***Saving seen data "./dancer.seen"
17:05:04preglowmerbanan: then that'll be suboptimal on coldfire, at least
17:05:26preglowbut to make it optimal on coldfire you'll either need everything to be in the lower 32 bits, or in the upper 32 bits, so it's a bit of a bother
17:05:38 Join Cassandra [0] (
17:05:46_FireFly_hi cass
17:05:55 Join mashalla [0] (
17:09:24ripnetukhi FireFly - you are interested in iRiver remote arent you? i submitted a patch to make delete work on remote - fancy taking a look?
17:10:25 Join dpassen1 [0] (
17:12:12_FireFly_ripnetuk: i can have a look on it
17:12:28ripnetukthanks :) its on the sourceforge tracker
17:12:36merbananpreglow: so you have 40bits but you only see 32bit at a time? can you shift in the other 8 bits if you need them ?
17:14:39 Join TiMiD [0] (
17:14:56_FireFly_hi TiMiD
17:15:17_FireFly_i have found a little bug in your gui_quickscreen implementation
17:15:24TiMiDsounds like I've done some more mistakes from what I read right now :(
17:15:45TiMiD_FireFly_: if there was only one ...
17:15:49 Quit San (Read error: 110 (Connection timed out))
17:15:54ripnetukfirefly - its at
17:16:05 Join ghode|afk [0] (
17:16:08_FireFly_TiMiD: it's on tracker the fix
17:16:45_FireFly_gui_quickscreen fix
17:17:24TiMiDamiconn: about the sizeof(buffer), it was an old code, it worked because it was in the same area as the declaration of the buffer but I just blindly c'n'p
17:17:46TiMiD_FireFly_: I look
17:18:39_FireFly_and i just see a typo in yesno.c
17:18:53_FireFly_#ifdef TREE_RC_RUN
17:18:53_FireFly_ case YESNO_RC_OK:
17:19:35ripnetuki changed that because TREE_RC_RUN is not defined
17:19:42ripnetukprooved by deliberate error in ifdef block
17:19:53ripnetukso i changed it to ifdef YESNO_RC_OK
17:20:02_FireFly_ripnetuk: that's right
17:20:07TiMiDthe patch looks good
17:20:13TiMiDI will apply
17:20:18TiMiDit's from you ripnetuk ?
17:20:23TiMiDcan you commit ?
17:20:24ripnetukyes thanks :)
17:20:26ripnetuki cannot
17:20:35CtcpIgnored 2 channel CTCP requests in 7 seconds at the last flood
17:20:35*ripnetuk is not good enough at c to be owrthy of commit
17:20:39TiMiDok then givme your real name
17:20:44ripnetukGeorge Styles
17:20:57ripnetuk(im already on credits for some very basic changes to sim years ago in Archos days)
17:20:59TiMiDoops I'm stupid again : it's written on the patch tracker :(
17:21:15_FireFly_ripnetuk: better create a account on then you can modify your next submit to the tracker :)
17:21:16TiMiDI will apply this when
17:21:23TiMiD - I will finish the lunch
17:21:36TiMiD - I will finish the gui_logo
17:21:38ripnetukfirefly - my brain was mashed - i already have a sourceforge account for a project I released on gpl ages ago
17:21:42CassandraYou see, that's the problem with the youth of today, no commitment.
17:22:00ripnetukcass - thats my joke when people forget to commit on Oracle :)
17:22:29ripnetukthanks for looking at my patch - deleting files is important to me and I hate not being able to update to latest cvs without patches :)
17:22:37*Cassandra left her remote on the bus. :(
17:22:44ripnetukoh no :(
17:22:44_FireFly_Cassandra: ouch
17:22:59amiconnTiMiD: The problem was that you changed the datatype from array to pointer.
17:23:30CassandraAnd I didn't suceed in getting an iPod out of my friend either.
17:23:44Cassandra(which wasn't that likely, but ...)
17:24:04ripnetukwithout the remote, the iRiver is no better than the iPod IMHO
17:24:29amiconnUnfortunately the additional parameter increases code size somewhat
17:24:35CassandraI suppose I can always buy a new remorte. :(
17:25:09ripnetukhave u tried ringing the bus company? in the uK at least they sometimes reunite people with lost property
17:25:26ripnetuk(unless the police have blown it up)
17:25:28amiconnIf I wouldn't need it for development I wouldn't hesitate to give my remote away
17:25:30CassandraI sent them an email.
17:25:39CassandraThey never found my gloves though.
17:26:57CassandraI assume I must've missed when I put it in my bag.
17:27:28CassandraI need it more for development than because I use it that much (although it is kind of handy over a big coat in winter.)
17:30:52SlasheriHmm, maybe lcd_remote_update could be threaded and when wps mode with ticking reduce is activated, it would update the display on background very slowly preventing the ticking
17:31:09Slasherithat would require also a double frame buffering
17:32:35amiconnLet's not overcomplicate things
17:32:56amiconnI should try my random-interval idea soon...
17:32:59 Quit [1]Tomas (" HydraIRC -> <- The future of IRC")
17:33:06Slasheriyes, that sounds good also
17:33:38Slasherii already tried some delay compensation without any noticeable results
17:35:42 Join reset_router [0] (
17:35:47 Nick reset_router is now known as cannard (
17:37:00ripnetuki was very impressed when I accedently played a .jpg album cover and it displayed it all nice on my iRiver :)
17:37:41cannardwhat sort?
17:38:20 Join Amar [0] (
17:39:24 Join S-Eater [0] (
17:43:18 Join DangerousDan [0] (
17:44:07 Join muesli_- [0] (
17:45:05 Join LinusN [0] (
17:46:22 Quit B4gder ("time to say moo")
17:47:16TiMiDamiconn: yes, I understood just a silly c'n'p without thinking of what was inside the code, which made all numbers of more than 3 chars to get cutted hehe
17:47:51amiconnThe code size increase is rather small, I just checked
17:47:57amiconnmax. 16 bytes
17:48:23amiconnNothing to worry about
17:48:35ripnetuki reckon my patch would save around that many bytes by removing ad-hoc code and replacing with api calls... havent checked tho
17:48:57Jungti1234Grow sleepy....
17:53:12TiMiDabout the code size, I think there will be a matter for unicode : I don't think we can go a lot below where we are now (if you take in account the initial 4kb incease when applying first multi-screen onemonth ago, then all the decreases done when using the gui api, I guess we gained around 1-2kb)
17:54:51TiMiDripnetuk: yes, it will decrease the binary siez a little, but not enougth :/
17:55:05ripnetukhehe every little helps
17:55:18TiMiDI expected that to remove duplicated code would allow make it decrease more that that :)
17:55:36TiMiDyes and I think it's at least more easy to read now
17:55:56 Quit webguest62 ("CGI:IRC (Ping timeout)")
17:56:38preglowmerbanan: we have an accumulator register which holds 32 of the bits, and an extension register which holds the top eight bits
17:57:12preglowmerbanan: so if we need more than 32 bits, we need to do a an extra fetch, extract some bits and do a cross-register shift
17:58:30amiconnTiMiD: I think there's still something to be gained
17:58:45amiconnThe whole cconversion is not yet done
17:59:22 Quit muesli_- (Read error: 104 (Connection reset by peer))
17:59:31TiMiDfar from being done I must say
17:59:50TiMiDas longas there is a lcd_* remaining in the drawing code
18:00:25TiMiDkevin@harkonnen ~/rockbox-devel/apps $ grep -r lcd_ *|wc
18:00:25TiMiD 3078 12652 199447
18:00:31amiconnAnother screen that came to mind - bookmarks
18:00:33TiMiD3078 remainings :)
18:00:47TiMiDyes I wanted to go to this one after
18:00:52_FireFly_playlist uses some lcd_calls
18:00:56TiMiDbut I don't know how they works
18:01:03TiMiDplaylist ?
18:01:30TiMiDJungti1234: 0 lcd_ ? you interrest me :)
18:01:38amiconnTiMiD: Your grep -r also catches apps/gui ...
18:01:44TiMiDyes of course
18:02:02TiMiDkevin@harkonnen ~/rockbox-devel/apps $ grep -r lcd_ *|grep -v gui|wc
18:02:03TiMiD 3055 12556 197663
18:02:04_FireFly_TiMiD: display_playlist_count in playlist.c
18:02:51_FireFly_the other "big"-part is the virtual-keyboard
18:03:28TiMiDyes the virtual keyboard
18:04:17_FireFly_the "problem" is that for charcell-targets a separate keyboard-code is implemented
18:05:01TiMiDit wouldn't be too hard to merge the two into a single graphical object
18:05:32TiMiDmaybe from scratch too for this one (it doesn't seems too hard too)
18:05:37amiconnThe biggest problem with vkeyboard is (imho) how to handle the different lcd sizes in the sync_ case
18:05:47TiMiDI thought about it
18:05:59_FireFly_my remote-setting patch has maybe a soulution in it
18:06:12TiMiDjust make differents keyboards
18:06:28_FireFly_i have added a screen_id var to the screen-struct
18:06:37TiMiDsince it's not possible to have the full size keyboard on the remote
18:06:50TiMiDscreen_id ?
18:06:54_FireFly_yepp the "old" 3-page system is needed
18:06:54 Quit S-Eater ("CGI:IRC (EOF)")
18:06:58 Part LinusN
18:07:02TiMiDthere is already a screen_id or something like that
18:07:04_FireFly_which as SCREEN_MAIN or SCREEN_REMOTE
18:07:09amiconnTiMiD: Different picker areas, but a synced input line?
18:07:41TiMiDamiconn: 2 solutions
18:07:55_FireFly_that's relative new isn't it ??
18:08:01_FireFly_the screen_type var
18:08:24TiMiD1/ fast to implement : 2 pickers area, if we use the main unit, it will go outside the one on the remote
18:08:33TiMiD2/ a scrolling picker areas
18:08:46TiMiD2 seems more fun to implement ;)
18:09:07 Join Moos [0] (
18:09:12TiMiD_FireFly_: I don't remember well, but I think it was there since the beginning
18:09:15MoosHello guys!
18:09:16TiMiDhi Moos
18:09:23amiconnThe charcell picker is scrolling instead of paged (1 line only)
18:09:31TiMiDyour french language update is perfect ;)
18:09:36*preglow kicks portalplayer
18:09:38Moosthanks :)
18:09:46amiconnTiMiD, Moos: No, it's not
18:10:06amiconnMoos: You made the player keylock strings much longer than before
18:10:13amiconnThey have to be shorrrrrrrrrttt
18:10:16TiMiDoh :)
18:10:22Mooslet me see this
18:10:35linuxstbpreglow: Any progress?
18:10:36TiMiDI was talking on my iriver hehe
18:10:41TiMiDat least the sentences means more
18:10:45_FireFly_TiMiD: you are right but this var didn't tell me which screen is currently given to a function
18:10:48amiconnThat's why there are 2 sets of these, one for player (charcell lcd -> short) and one for recorder (and all others)
18:10:56MoosTiMiD: yes it was the idea ;)
18:11:20TiMiD_FireFly_: it contains SCREEN_MAIN when it's the main screen and SCREEN_REMOTE when it's ... guess what ;)
18:11:30preglowlinuxstb: i've found out what i do wrong, at least, the exception vectors aren't placed in the same spot as all the other arms in the world
18:11:38linuxstbI read your discussion.
18:11:42amiconnfrancais.lang lines 409 and 415
18:11:51_FireFly_ohm that must i overseen let me check it again ;)
18:11:52preglowproblem is now i probably need a fixed address where i can put the handlers
18:11:54Moosamiconn: I change this
18:11:57preglowand this probably needs to be iram
18:12:06amiconnpreglow: Could you please scan your nano?
18:12:16_FireFly_argh i'm blind ;)
18:12:52preglowamiconn: i have no scanner
18:13:11 Join webguest62 [0] (
18:13:16TiMiD_FireFly_: hehe this time it's not my fault :)
18:13:46_FireFly_i had implemented the same twice ;)
18:14:20Coldtoastman.... sorry to be completely off topic. but I just finished the first crossword I've evr finished! yay
18:16:03_FireFly_ok localy ereased all references to screen_id ;)
18:16:16TiMiDColdtoast: woow you should try in boring lessons, it's amazing how you will progress :p
18:16:38_FireFly_sudoku is also great ;)
18:16:52ColdtoastI did it at work
18:17:10Coldtoasthad to bring it home with me cos I was stuck on the last word. but I got it. AND without cheating
18:17:28ColdtoastI really like Bejewelled
18:17:52ripnetuki like gta:lcs on my psp :) best portable game ever IMHO
18:17:57TiMiDoh I never tried
18:18:20 Join _DangerousDan [0] (
18:18:37Moosamiconn: LANG_KEYLOCK_ON_PLAYER=Verrouillage touches ON
18:18:42Moosit's better?
18:18:47TiMiDhehe this game is addicting ;)
18:20:55TiMiDsounds like tomorrow morning's lesson will be fun ^^
18:21:23 Quit Vladoman ("Leaving")
18:22:04 Join RotAtoR [0] (
18:22:34 Quit Febs ("CGI:IRC (EOF)")
18:22:37TiMiDhshah: if you are here, givme your usb logo ^^
18:22:44TiMiDthe remote one
18:23:08ripnetukim off now... thanks for looking at my patch so quickly firefly...
18:23:09ripnetukcya later
18:23:35TiMiDripnetuk: I will commit it
18:23:45ripnetuk:) /me is pleased
18:23:48 Join Kaggen [0] (
18:23:50ripnetuktalk later
18:23:51 Quit ripnetuk ("Ninja IRC v1.5.8.1(#1) exiting after 2h34m16s of use")
18:26:42 Join muesli_- [0] (
18:28:54RotAtoRhmm, i've been looking at working with color graphics with the upcoming h300 and ipod ports and the graphics api seems not very useful for anything other than monocrome graphics
18:29:03RotAtoRit would be very useful to be able to have a draw mode where you could draw everything but a certain background color, but it doesn't look like that's possible
18:29:18Jungti1234Korea time is 2:28 now. I'm sleepy
18:29:57TiMiDJungti1234: good night !
18:30:55RotAtoRdoes anyone (amiconn?) know if there are plans to extend the graphics api?
18:31:27TiMiDRotAtoR: I discussed a little bit with amiconn about this matter
18:31:52 Join webguest07 [0] (
18:31:52 Quit Amar ("CGI:IRC (EOF)")
18:31:53RotAtoRok, so at least there are talks
18:32:25TiMiDthe problem is that you must have 3 api in the same time :
18:32:33TiMiDmonochrome, grey, color
18:33:12 Quit webguest07 (Client Quit)
18:33:14RotAtoRhmm, that would be a pain to maintain, but there must be some way to consolidate it all
18:33:35TiMiDsince players like iriver H1x0 / H3x0 have more than onescreen type
18:34:14RotAtoRahh yes
18:34:18TiMiDto the application level, the 3 api would co-exist
18:34:27 Quit DangerousDan (Read error: 110 (Connection timed out))
18:35:22TiMiDbut since most of the apps uses predefined colors, we could easily make the app call one fn on one screen with the predef color and the convert it to the native screens's format
18:35:58TiMiDof course for custom colors, whe have no choice but to let the app make special cases
18:36:31TiMiDsince we cannot easily translate a r g b to grey or monochrome without possibilities of it looking ugly
18:36:37TiMiDon the target screen
18:38:10TiMiDthat's a first draft in our heads:)
18:38:31TiMiDI'm also a lot in favor of a user space drawing library
18:38:36TiMiDrather than in kernel space
18:38:43*preglow tears hair out
18:38:46TiMiDkernel should handle specificities of each display
18:39:01TiMiDlike low level doperations
18:39:05RotAtoRwell i hope it goes somewhere, i think a lot of people will be screaming for a nice color graphics api as soon as the h300 and ipod ports are ready :)
18:39:25TiMiDbut since high level operations are the same for almost every drivers ...
18:39:58TiMiDhehe they can scream, we aren't a company, we can do it as we want ;)
18:40:09RotAtoRhehe :)
18:43:37TiMiDthis would need a lot more reflection of course :)
18:43:51TiMiDwe should write it on the wiki
18:43:56TiMiDsince it's important
18:45:24 Quit muesli_- (Read error: 104 (Connection reset by peer))
18:45:45RotAtoRit looks like there's a " Further ideas/suggestions" section already at the bottom graphics api page, but i don't know if that would be the best place for it
18:47:26hshahTiMiD, u still need it?
18:48:04TiMiDhshah: I could include it with my incoming commit if it's nice :)
18:48:05cannardwould colour display be easy to implement?
18:48:11amiconnTiMiD: The problem is that many drawing primitives need knowledge about the low-level pixel format for optimum performance
18:48:26cannardeven though it might be an afterthought/addition
18:49:34TiMiDhshah: give it to me, I will test ^^
18:49:44TiMiDamiconn: yes, I undertand that
18:49:46amiconnRotAtoR: specifying a background colour isn't really flexible. (1) it will become impossible to use the colour reserved for masking. Not a big problem with highcolour, but certainly with only few greylevels
18:49:47hshahim trying to send it here...
18:50:40hshah^^ there TiMiD
18:50:42amiconn(2) Standard bitmaps are in target format for size and performance considerations. The drawing code would have to extract a mask from that
18:50:48TiMiDbut for example do you think that all the font handling needs to be have knowledge about the low-level format ?
18:51:04amiconnSo it might be better to use a mask bitmap in the first place
18:51:40amiconnTiMiD: Yes, because the font bitmaps are also in (monochrome) target bitmap format
18:51:45_FireFly_amiconn for this could my cbmp-patch be usefull
18:51:59TiMiDhshah: way nicer that the current one, if you give it to me in hex format, I will put ti I think
18:52:09 Quit mashalla ()
18:52:14amiconnIn fact not all font handling, but the string drawing itself (lcd_putsxy)
18:52:28TiMiDhshah: don't try to send me files, put them on the web btw
18:53:23amiconnThe scroll thread only needs knowledge about the capabilities (b&w/ grey/ colour), so it could be app layer
18:53:42TiMiDthx ;)
18:53:44 Join San [0] (
18:54:30TiMiDhshah: could you make the same for the remote, but only with the usb plug for example ?
18:54:44hshahok - what r the dimensions of the remote?
18:54:54amiconnTiMiD: Why not use the old standard USB logo for the remote?
18:55:12TiMiDamiconn: it works
18:55:15amiconnShould fit quite well
18:55:19TiMiDbut if this one is nicer
18:55:22_FireFly_it fits well
18:55:48amiconnI prefer the old logo over this one
18:55:53amiconnI like simplicity
18:55:55TiMiDhshah: do it to the dimensions of the logo, since it's automatically centered
18:56:16TiMiDI think at east the pkug is nice
18:56:22Moosamiconn: you want I DCC you the new fr .lang file?
18:56:38amiconn..and btw the computer picture in this logo doesn't match my comp at all
18:56:59TiMiDhehe mine neither :)
18:57:17TiMiDwell but just the plug ?
18:57:17 Join Febs [0] (
18:57:50TiMiD(I find the actual usb logo a little ugly, that's why)
18:57:58amiconnI like the old plug more. Being slighty slanted it looks more dynamic
18:58:36TiMiDhehe guts and colors...
18:58:41hshaherm i can't remember how to convert a bmp to hex...
18:58:44TiMiDwell a compromise then
18:58:52TiMiDnew plug on the remote
18:58:58TiMiDold plug on the ain screen :)
18:59:18_FireFly_hshah: in tools there is a tool for it
18:59:19amiconnThe main screen could have a better plug logo greyscale
18:59:35TiMiDtaht would be nice indeed :)
18:59:58 Join NicoFR [0] (
19:01:11 Quit NicoFR (Client Quit)
19:01:37amiconnSpeaking about logos - imho it would make sense to put the logos in separate source files the same way as already done for the colour rockbox logo
19:01:48hshah^ there TiMiD
19:01:51HCl <- USA allows chemical testing on kids that have been abused in the past
19:02:09preglowHCl: well, sure, why not, they're scarred for life anyway, nothing to life for
19:02:12amiconnThey tend to be really long and make actual code hard to spot the same file
19:02:39TiMiDyes I agree
19:03:00***Saving seen data "./dancer.seen"
19:03:07 Join muesli_- [0] (
19:03:09thegeekpreglow : I love your logic
19:03:16thegeekit's just so.. logical;)
19:03:27preglowi'm like a vulcan
19:03:57preglowi wish i didn't recognize that
19:04:18TiMiDamiconn: with all the icons and so on ???
19:05:19amiconnYes, perhaps. Depends on how large these will become for colour LCD...
19:06:00hshahTiMiD, any chance while you are doing that, that you would be able to comit my wps?
19:06:15hshahor is that too much to ask?
19:07:18TiMiDHCl: USA frightens me when it comes to his kinds of things :(
19:07:24 Quit Febs ("CGI:IRC (EOF)")
19:07:30amiconnHmm, different question: Are there objections against replacing the simple 'backlight on when plugged' with a separate timeout setting when plugged?
19:07:40TiMiDHCl: it can't hurts anyway...
19:07:54preglowamiconn: well, more flexible, can't see why that'd hurt
19:08:19amiconnShould be possible with almost no additional code, and the old behaviour would be achieved by setting it to 'always on' when plugged
19:08:30TiMiDHadaka: there is a problem with the hex version of the logo you sent me
19:08:48TiMiDsr Hadaka
19:08:52amiconn...but you could e.g. set a timeout of 5s on battery, and 30s on mains
19:09:09 Join Acksaw [0] (
19:09:17hshahwhats wrong with it?
19:09:50amiconnpreglow: Also an important point: it's more flexible without an additional setting.
19:10:05amiconnWe already have so many settings that I am hesitant to add more
19:10:37_FireFly_amiconn: then you could have a look ar my remote-settings patch :)
19:11:48_FireFly_argh misunderstood your sentences
19:12:46 Quit RotAtoR ("CGI:IRC 0.5.4 (2004/01/29)")
19:12:57 Join Febs [0] (
19:18:02Sanhey febs
19:19:03TiMiDsry lag
19:19:15amiconn_FireFly_: Remote scrolling is a different thing; these settings are simply necessary
19:19:42 Join speed123 [0] (
19:25:25 Quit cannard (Read error: 104 (Connection reset by peer))
19:25:53 Part Polo_o
19:26:04 Quit muesli_- (Read error: 110 (Connection timed out))
19:30:19Jungti1234Korea time is 3:30 now.
19:30:36Jungti1234Much late.
19:31:10 Join Sando [0] (n=lolsteam@
19:33:28 Quit einhirn ("Miranda IM! Smaller, Faster, Easier.")
19:34:41speed123so what u guys thing about the rumors of a official firmware update for the H3xx? - if i dont interrupt someone
19:35:12preglowwe don't much care about it
19:35:50speed123yes, when rockbox is going to be well i am sure to swich
19:37:07 Quit Mark__ (Read error: 110 (Connection timed out))
19:37:36 Quit dpassen1 ()
19:39:25hshahTiMiD - the remote hex work for you in the end?
19:41:59Sanwhere is the buttons.c file in the source code?
19:42:43Sangot it
19:42:47Jungti1234Made new iPod WPS.
19:43:10TiMiDhshah: the problem is that you have both bytes and double bytes in the file you gave me
19:43:13TiMiDin hex
19:43:19TiMiDthe first part seems ok
19:43:25Jungti1234It resembles closely with iPod.
19:43:26TiMiDthe second one is wrong
19:43:29hshahi just typed the command in :s
19:43:39hshahbmp2rb a.bmp > a.txt
19:44:00Sanany new H300 builds?
19:44:28Jungti1234Night was late. good night
19:44:29hshahTiMiD - do i need to add anything to the command?
19:44:35TiMiDI don't know
19:44:46TiMiDI never used it myself
19:44:49 Quit Jungti1234 ("Bye Bye~")
19:45:45_FireFly_hshah: you need the -f parameter
19:46:04hshahahh rite - lol
19:46:05_FireFly_execute only bmp2rb without any params for the help
19:46:22 Join Wett [0] (
19:46:38hshahyeah i got all that
19:47:11 Quit Kaggen ("Download Gaim:")
19:48:12hshahbmp2rb -f 2 a.bmp > a.txt
19:48:14hshahthat rite?
19:48:35 Part speed123
19:49:47 Join xmixahlx [0] (n=xmixahlx@
19:49:51 Join arkascha [0] (
19:50:15_FireFly_hshah: for which screen is this logo ?? when it is for the remote then it's wrong :)
19:50:27hshahyeah, its for the remote
19:50:53hshahso what does the command need to be?
19:50:59_FireFly_then you need the mono param 0 Archos recorder, Ondio, Gmini 120/SP, Iriver H1x0 mono
19:51:49hshahit still outputs things like that
19:51:55hshahwhich it shouldn't do
19:52:38_FireFly_is your bmp really 1bit = monochrome ??
19:53:31hshahi just resaved it as a 1bit monochrome
19:54:39hshah_FireFly_ - try it yourself mate please
19:57:11 Join trinidad [0] (
19:57:34trinidadhow do i mount my archos in mtab to allow me to write to it?
19:57:44_FireFly_hshah: this is right i have used this command: /bmp2rb -f0 usbhshahremote.bmp
19:57:57_FireFly_trinidad: add users to the fstab-entry
19:58:11_FireFly_then you can also mount it as user
19:58:56hshah_FireFly_ - i did the same thing and even in the cmd window its got text like 0x202c
19:59:58_FireFly_me not
20:00:18linuxstbhshah, _FireFly_: I get the same problems with bmp2rb - it outputs 0xe8f1 as one of the values.
20:00:18_FireFly_this is the output of the program
20:00:29amiconn_FireFly_: For bmp2rb the .bmp format doesn't really matter
20:00:42amiconnIt should handle all depths from 1..32 correctly
20:01:00hshahhmm... then what on earth is going on...
20:01:10_FireFly_maybe a windows problem ??
20:01:14hshahTiMiD was saying something about not being able to use the values...
20:01:19_FireFly_because i'm under linux
20:01:19trinidad_FireFly_, thanks for the tip, however, there is no entry for the archos in fstab, it resides in mtab. Now, I have added users to the options, but when it is mounted i still don't have write permissions unless I run konqueror as root.
20:01:22linuxstb_FireFly_: I get identical output to you - your output also has a bug.
20:01:31hshah_FireFly_ - ive always used windows - i managed to hex my USB logo fine
20:01:42 Quit xmixahlx (Read error: 104 (Connection reset by peer))
20:02:16_FireFly_linuxstb: just saw it
20:02:54amiconnHmpf. Seems the one who extended bmp2rb for highcolour output introduced a bug
20:03:14*hshah loads shotgun ;)
20:03:42markunamiconn, preglow: any idea what a A606-0512 does? It's located close to backlight connector, maybe a DC/DC convertor?
20:03:46TiMiDnext time maybe for the usb screen ;)
20:04:12 Join xmixahlx [0] (n=xmixahlx@
20:04:14hshahTiMiD - as another request - any chance for my wps?
20:04:36TiMiDhshah: when my connexion will be more stable, I will commit first what I have to commit
20:04:49hshahok cool :)
20:05:44TiMiDthen I will see for your wps and Moos's french update
20:06:52Acksawany foobar users here?
20:07:06 Join tucoz [0] (
20:07:24tucozhshah, where can I have a look at your wps?
20:07:39tucozon the tracker perhaps?
20:08:04tucozok, thanks
20:12:18tucozI am not sure how to say this without being rude to all the wps-makers, but I have a little concern with Rockbox being shipped with a lot of wps's. Imho, Rockbox should be shipped with only a few wps to show what could be done, and perhaps offer a downloadable theme-pack on the server.
20:12:44 Join Mark_ [0] (
20:12:56 Join Mark__ [0] (
20:13:09tucozhshah, this has nothing to do with yours. I like the clean look on yours.
20:13:24hshahno worries mate
20:13:40hshahwhats your reasoning behind that anyway?
20:13:49hshahi mean why not have a few wps's?
20:15:11tucozI do not see the point in having a zillion themes shipped in the standard build
20:15:39hshahwell obviously not too many, but anything below 20 is ok isn't it?
20:15:43Sanare you guys shipping rockbox as the standard firmware for the players?
20:15:51 Join tim66 [0] (n=tim@
20:15:55hshahSan... we can't do that - lol
20:15:58Sani know
20:16:05Sanbut have people asked you?
20:16:13hshahbut i think someone said something about it...
20:16:22hshahLinusN is the person to ask i think
20:17:13hshahtucoz - rockbox should be for everyone, and some don't know how to make wps/compile the code etc etc so why not include as much as possible
20:17:25tucozhshah, maybe. In my opinion, the shipped wps's should be some sort of showcase, or maybe popular ones. Then again, I trust the developers do their choices when they commit stuff.
20:17:27hshahits not like extra wps are going to reduce performace or eat up lots of space
20:18:02markunSan: there were some talks with Neuros
20:18:13Sanoh, yeah, i remember that
20:18:38Sanwhats wrong with the button config in rockboy?
20:18:41amiconnhshah: They do eat up lots of space, at least in their current implementation
20:18:55hshahhow much space?
20:18:56tucozhshah: well, not much. Maybe you are right. I just think that it might be a better idea to offer a theme-pack for people wanting to have _lots_ of themes on their player
20:19:09hshahnow that wud be a better idea tucoz
20:19:13hshahofficial theme packs
20:19:34amiconnSeveral megabytes depending on the HD size
20:19:57hshahi wouldn't call anything less than 5mb a lot...
20:20:11amiconnThe problem is that a number of wps'es uses dozens of bmp files. Each file needs at least one cluster
20:20:24amiconn...and cluster size is 32KB on H140...
20:20:25TiMiDanyway since hshah's wps is 'clean style', it should be bundled too
20:20:28tucozI do think I could change opinion if the themes only used one .wps and one .bmp file. But, as it is now.
20:20:34TiMiDwe have only picture wps
20:20:45tucozTiMiD: I agree
20:21:15hshahthanks TiMiD :)
20:21:20TiMiDtucoz: commit it then (I have some connexion problems)
20:21:22 Quit trinidad (Remote closed the connection)
20:21:27amiconnThe wps module contains >200 files.
20:21:30tucozTiMiD: I can't
20:21:35tucozI am no commiter
20:21:53TiMiDwell what is a Mb when you have GB ?
20:22:09tucozIt is not the space I am concearned about
20:22:20hshahtrue it adds to the download size of rockbox
20:22:22TiMiDdownload size ?
20:22:24amiconnThey also increase RAM usage of the dircache
20:22:25hshahand some people r still on 56k
20:22:33amiconnYes, that as well
20:22:39tucozamiconn: that is a good point
20:22:46TiMiDthe bundled one are very nice
20:22:52amiconnThat's the main reason why we also don't bundle voice files
20:23:00TiMiDmaybe we should filter the next incoming
20:23:09linuxstbamiconn: I still like your idea of having a single large bitmap that you "copy and paste" from.
20:23:20amiconn...although I consider voice UI a much more useful feature than graphical wps
20:23:34TiMiDmake a page with a picture or 2 of the wps
20:23:46TiMiDand a zip file to extract to the root of the player
20:24:06TiMiDI never look at the wps
20:24:07tucozYes, I like those. Still, I do think that the bundled ones should show something special, have a high standard etc. Just not all of them.
20:24:12TiMiDbut they are cute :)
20:24:26hshahbut the idea about theme packs is good
20:24:36hshahespecially if official rockbox ones were made by the devs
20:24:54tucozhshah: yes, especially now, as we have the theme-concept
20:24:55TiMiDyou don't want a wps made by me ;)
20:25:03amiconnWhen the windows installer will be extended to support iriver, we could also offer 2 versions.
20:25:03linuxstbYou don't want devs designing the WPS...
20:25:13TiMiDhehe :)
20:25:20hshahwhat i meant was, they should be collected by devs
20:25:31TiMiDor by users
20:25:35TiMiDwith a vote system
20:25:38hshahso the devs see all the available ones... choose ones they like to be for rockbox... and then make packs
20:25:55amiconnOne minimal install, and a full version for those users with a fast connection which want an all-in-one package
20:26:33amiconnThis all-in-one package might easily become 10..20MB, perhaps even more
20:27:08tucozamiconn: yes, that could work. Or just offer the extras in their own zips. (, etc) ?
20:27:38amiconnThat's what we do right now with the voice files
20:27:46hshahyeah - i wud prefer the extras
20:27:58tucozamiconn: yes, but more easy accessible. Like on the download-page
20:28:01hshahif the users don't know how to install the themes then they should be shot :p
20:28:16tucozhshah: that would be as simple as installing rockbox
20:28:27tucozjust unzip the .zip onto the player
20:28:28hshahyeah i know - but some even have problems with that
20:28:47amiconntucoz: I agree 100% that the voice files should be made accessible from the download page, at least those matching the (current) release
20:29:01hshahsome people have emailed me after completing my guide and asking what they do next...
20:29:04 Quit ghode|afk (Read error: 104 (Connection reset by peer))
20:29:12tucozhshah: hehe, that is true. Still, if they ask for themes, then I guess they have rockbox installed already
20:29:51amiconnWe already had requests where the user didn't understand what was meant by 'root directory' ...
20:29:55hshahyeah - ur always gonna get some idiot asking... what program do i unzip it with, or where do i put the files... and oh this theme is not working...
20:30:53amiconnHmm, there might be an additional problem with theme packs. Not all themes are suitable for all targets
20:31:06tucozah, that is true
20:31:50tucozBut, the buildscript only bundles the suitable themes as it is now?
20:32:00 Join ghode|afk [0] (
20:32:37_FireFly_TiMiD: i got it :) now the format_buffer is only with 387 bytes filled instead of 759 bytes(with my wps)
20:33:12tucozI see.
20:35:48tucozThen the thempacks will have to be for a specific platform(s), which in turn will yield a number of .zips, which is again not a good idea.
20:37:04amiconnThere will be at least 4 theme packs, depending on lcd resolution and depth. 6 when all ipods will be supported some day
20:37:35tucoz..or just an extra link under the platform on the download page.
20:37:47tucozlike an extras link or something like that
20:38:12 Quit Acksaw ()
20:39:46 Part tucoz
20:44:11preglowipod data bus _is_ 16 bits wide
20:44:18preglowso perhaps we should do some tests with thumb code
20:45:21 Join Rhino [0] (
20:48:19hshahbk to themes - i think rockbox should have a max of 10 built in themes and the rest should be addons
20:49:02 Quit Mxm`Pas`Bien (Read error: 104 (Connection reset by peer))
20:49:03 Join Maxime [0] (
20:54:51 Join _arkascha [0] (
20:57:11 Join frederic [0] (
20:57:19 Quit _arkascha (Remote closed the connection)
20:59:12 Quit arkascha (Read error: 110 (Connection timed out))
21:00:47 Quit Rhino ("CGI:IRC (Ping timeout)")
21:01:18 Join webguest33 [0] (
21:03:01***Saving seen data "./dancer.seen"
21:03:17 Join Rhino [0] (
21:03:40Rhinohi all
21:04:16 Join trinidad [0] (
21:05:06trinidadi have installed the latest daily firmware for my archos, fixed the problem i had with mounting and write access, but it seems that whenever i past folders into the drive it doesn't transfer all the data
21:05:19 Quit trinidad (Client Quit)
21:05:31 Join Philip [0] (
21:05:44RhinoI have a Roxkbox programming question ... can anyone help?
21:06:20_FireFly_simply ask Rhino
21:07:03tim66hi all, anyone else get an internal compiler error when building 'dumb' ???
21:07:13Rhinothanks dude, I was wondering if some sort of simplified directory structuring is in the pipeline ... if not I was going to have a bash at it.
21:07:23RhinoI have 200 artists
21:07:43Rhinomy directory structure is in \my music\artist\album\track# - track
21:07:58Rhinoand it's a pain having to scroll thru the artist list to get to the one I want
21:08:15Philipdo you have lots of differnent genres??
21:08:36RhinoI was wondering if someone is planning to implement A, B, C virtual directories that automagically filter the artists
21:08:53Rhinoif not is it doable?
21:08:56 Join muesli_- [0] (
21:09:11RhinoI don't want to go to the bother of starting it if there is a fundamental reason why it won't work
21:09:41Philipi can't see any reason it shouldn't.... but i'm not a coder :(
21:10:42RhinoI've been coding c/c++ for 18 years so I can't forsee any programming issue, I was just worried that there may be a fundamental issue with the Rockbox API preventing something like this being implemented
21:10:56 Quit Wett (Read error: 110 (Connection timed out))
21:11:23tim66Rhino: virtual directories? what do you mean? would an artists folder with playlists for each artist, and the same for genre, etc. give the sort of functionality you're looking for?
21:11:29amiconntim66: m68k-elf-gcc 3.3.x ICEs on dumb. Use 3.4.x
21:12:10amiconnRhino: You can already scroll page-wise
21:12:19 Join linuxstb_ [0] (
21:13:13 Quit Mark__ ("Leaving")
21:13:24Rhinodidn't know that ... I thought you could only do it line-by-line ... 2 ticks while I try it
21:13:49Philipdo you want the ability to essential search your files via ID3 tag data??
21:13:49tim66amiconn: cheers, I'll give it a shot, I misread the wiki :-(
21:13:58Rhinook ... give me a clue ... how do I jump by page?
21:14:28amiconnOn iriver, hold Play while using the joystick
21:15:54RhinoPhilip: no, I want to go into a directory ... if it has more than, say, 20, folders present a single character list of matching folders, e.g. A, B, C, G, K
21:16:13Rhinowhen you click on the folder it would show you the real folders
21:16:30amiconnI wouldn't like rockbox to obfuscate the real dir structure
21:17:20Rhinomakes browsing so much simpler, I implemented it in my AlbumView application ...
21:18:11Philipme neither i have my music in Genre\Artist\Album\##-title but my genres are vague at best, not true genres
21:18:36amiconnPhilip: same here
21:20:03fredericI like filtering as it is done on the Rio Karma - list of letters on the left you can always see the "contents" of each letter on the right and switch between
21:21:28RhinoBTW I'm not the only one who would like this approach to browsing ...
21:23:13TiMiDRhino: of course everything can be done, but the code that handles the filetree is in my opinion ugly enough ...
21:23:29Philipi think i would use it once i get more music...
21:23:47linuxstb_Rhino: Why don't you just create your own subdirectories A, B, C, D etc and put your artists in those? Or am I missing something.
21:25:02 Quit Coldtoast ("Peace and Protection 4.22")
21:26:56RhinoMainly because for sync purposes I mirror what I have on my main fileserver where I don't have the problem.
21:27:48RhinoI use my AV app to browse/play
21:28:14RhinoAlso it's something that computers were designed for in the first place :D
21:28:28 Quit Moos ("Glory to Rockbox")
21:29:03dwihnoRhino: You're kidding me ;)
21:29:27dwihnoHowever, a icon-based UI is needed.
21:29:32RhinoI haven't looked at the code, is it mostly C or is it C++/STL?
21:29:39TiMiDRhino: hahaha
21:29:42dwihnoOnly C (and some ASM)
21:29:57Rhinooooo ... down'n'dirty
21:30:17dwihnoand faster than a rabbit on spee^H^H^H^Hrollerskates
21:30:32TiMiDRhino: no ooooo :)
21:30:44PhilipI was wondering, are the H300 builds updated with every change as they are announsed on the RB website??? for example will the files have changed since sunday???
21:31:24dwihnoDoes the Win32 sim compile without cygwin btw?
21:32:04TiMiDRhino: try to read it first, you will see that lot of parts can't do anything but what they habe been (foot)coded for
21:32:40linuxstb_Philip: Yes, the builds get updated automatically. The bleeding edge builds after each cvs commit, and the daily builds once each day.
21:32:45 Quit webguest33 ("CGI:IRC")
21:32:53Sanso any changes in the H300?
21:33:23Rhinoyeah, I'll pull the code and see what I can do. if I get anything working I'll pop back and show you guys to see what you think
21:33:35Rhinothanks for your help dudes.
21:33:41linuxstb_Rhino: One complication is the dircache.
21:33:57Sananyone working on the H300 besides LinusN?
21:34:11linuxstb_But if you do this at the application level, it shouldn't matter.
21:34:30Sanphillip, I have 6617 .gbc roms
21:34:35Sanbeat that
21:35:05RhinoIs the dircache an API layer? If so thenm that'd be the best place to implement it
21:35:20Philipi only asked because there is not an obvious link (from the daily builds page) to the H300 build so did not know wheaver or not it was updated in the same manne, thankyou
21:35:22TiMiDlinuxstb_: in my optinion, if something like this is implemented, it won't interract with dircache sinceit will be at the application level
21:35:45Philipsan: can you play any of them???
21:35:55 Quit frederic ("Chatzilla [Firefox 1.5/undefined]")
21:36:01Sani got metal gear solid playing
21:36:04ender`what's the H120 screen resolution?
21:36:09Rhinois the dircache really a directory cache or is it a full FS API?
21:36:14Sanbut couldn't controll the person
21:36:16TiMiDa dir api
21:36:39Rhinothat should be fine then ... just need to put in virtual folders
21:36:45Philipoh cool just noticed daily build link to H300 :D
21:37:28RhinoI assume that for each folder there is a struct pointing to the parent and children?
21:37:37TiMiDRhino: you mean behind the dircache api ?
21:37:49Rhinono, the dircache API itself
21:37:51Sanhow do I download the dailybuild? re download the folder?
21:38:34RhinoI presume it's a tree structure of structures
21:38:44FebsSan, it's a .zip file. Just download it and extract its contents to the root of your player.
21:39:04TiMiDbut if it's implemented at the fs level, apps won't be able to see the normal tree structure which I presume could be bad in some cases
21:39:11Sanbut where, the place i downloaded it first?
21:39:21TiMiDfor me it's something that has to be implemented inside the viewer
21:39:35linuxstb_I agree.
21:39:46RhinoTiMiD: not really as all they would see if an empty read-only folder list
21:41:00TiMiDimagine you remove a file in the dir
21:41:04Santy febs
21:41:18Santhe wiki is so confuzzeling
21:41:25TiMiDand you set the file limit to 20 for starting the lettrer virtuals folders
21:41:51ender`actually, where can i find the logo that rockbox displays when it starts up?
21:41:53Philiphas anyone tested tag db in RB yet??
21:41:58TiMiDall apps which had the path /a/aaa will be wrong
21:42:08TiMiDbecause now it would be /aaa
21:42:10Sani clicked on the link and got this
21:43:20Rhinotrue, but the virualness could be controlled by the app and when it wants a "real" filename it can ask the dircache to convert it
21:43:31FebsYou need to click on either the daily build (top of page) or a bleeding edge build (bottom of page).
21:43:54FebsDon't click in the table−−those aren't the builds, they are the compiler logs.
21:44:04Philipsan: if you go to the daily builds page there is now a direct link the H300 build
21:44:27Sansee it now
21:44:33Sanwasn't tere yesterday
21:44:39Santhats why I was confused
21:45:35TiMiDmaybe a layer over the fs then
21:45:42TiMiDbut not the default choice
21:46:11Rhinodefinately, need to keep backwards compatibility
21:46:16TiMiDthe only app that could benefit fromm that would be filetree ....
21:46:47TiMiDanyway take a look at the sources and if you are not afraid, go :)
21:46:49Rhinonow you're into realms I don't know as I haven't pulled the code yet
21:47:00Rhinohehehehe ... defo!
21:47:16Rhinoagain, thanks for all your help dudes ... if I get anywhere I'll let you know
21:47:29TiMiDcu !
21:48:05 Part Rhino
21:49:51Philipjust crashed as i was seaching ID3 tag database...
21:49:59Philipon H300
21:51:17TiMiDPhilip: H300 is at a very early stage ...
21:51:34TiMiDand it seems that ID3 too :)
21:51:35Philipi know, just letting people know
21:51:59Philipthe data base loads, i can seach using text input
21:52:19Philipthen if i select any of the artists it says loading and hangs
21:57:48ender`can anybody point me to the place where the rockbox logo that's displayed while the player is starting is stored?
21:58:24Philipi thought it was contained in the rockbox.iriver file...
21:58:31 Join bbad [0] (n=bbad@
21:59:27ender`what about the sources?
22:00:40Philip?? sorry I don't know enough about it to know
22:01:11 Quit Kohlrabi (Nick collision from services.)
22:01:17 Join Kohlriba [0] (
22:02:13linuxstb_apps/recorder/logo-h300.c (for example)
22:03:50ender`what about H120?
22:04:07TiMiDguess :)
22:04:35ender`i'm trying, but i only see h300, nano and ipod in there...
22:05:13ender`apps/plugins/logo.c ?
22:05:28TiMiDbecause ifrb wwas logically organized, it wouldn't be fun ^^
22:05:42*TiMiD whispers apps/recorder/icons.c
22:05:51^BeN^where i can get the rockbox simulatur?
22:06:04preglowyou don't, you build it yourself
22:06:22^BeN^i think that i can download it
22:06:27*preglow stares at the build table in disbelief
22:06:32ender`ok... and the picture that was used to create that doesn't exist? :)
22:06:37preglowyou can, but then you need to compile it
22:06:48TiMiDender`: yes it's in tools/
22:06:54TiMiDbmp2rb I believe
22:06:58^BeN^compile its not a problem
22:07:08^BeN^wher i can get the source
22:07:16TiMiDbut it seems to be buggy since someone added support for ipod or H300
22:07:17preglowcvs, daily builds
22:07:54ender`that's just the converter as far as i can see
22:08:46linuxstb_You can create it again - e.g take a screendump.
22:09:41linuxstb_But the source image is on the Rockbox website somewhere - a very large tiff file which was scaled down.
22:10:50 Quit San ()
22:13:29 Join tvelocity [0] (
22:13:42 Quit ender` (Read error: 113 (No route to host))
22:13:43 Join ender1 [0] (i=ychat@
22:14:04ender1great, trying to view the screendump bsoded my windows :p
22:14:25linuxstb_Ah, my virus works...
22:14:49ender1you successfully brought down M$'s fastfat.sys
22:15:06ender1let's see if it happens again...
22:15:52linuxstb_preglow: How did you discover the ipod has a 16-bit data bus?
22:16:03amiconnpreglow: What about teh build table?
22:16:19preglowamiconn: it's... green
22:16:26preglowlinuxstb_: portalplayer datasheets
22:16:35preglowlinuxstb_: s/datasheets/advertising propaganda
22:16:36 Quit ender1 (Read error: 113 (No route to host))
22:16:44linuxstb_You mean "product briefs" ?
22:17:05amiconnpreglow: And? Apart from that there were better times. It's not completely green
22:17:25preglowamiconn: ah, no, but the current one is
22:17:49 Join ender` [0] (i=ychat@
22:18:11ender`this seems pretty reproducible...
22:18:28ender`it happened both times when i tried to view the 2nd image...
22:23:00 Join Jozi [0] (
22:23:39preglowlinuxstb_: seems they're bolting fat32 support onto their own bootloader now
22:23:50Jozicould someone help me with the bootloader
22:24:04Jozii was told theres a new version out for iriver h120
22:24:22Joziall i can find on the site are the daily builds
22:24:35Jozino version 6 bootloader
22:25:09linuxstb_preglow: Yes, and hfs and ext2/3 I believe
22:25:40Jozipreglow: thankk you
22:25:54preglowJozi: np
22:27:14Joziare you one of the development guys for the iriver fw?
22:27:33Jozicool, love the fw
22:27:45Jozikeep liking this player more and more
22:30:30preglowme too, heh
22:30:38preglowthink i'll reboot and do some eq work
22:33:08amiconnpreglow: fed up by arm exception vectors? ;)
22:37:22 Quit XavierGr (Read error: 113 (No route to host))
22:37:27 Quit _FireFly_ ("Leaving")
22:40:13Jozibootloader updated, thank you
22:40:28*ender` now has scans of his iHP-120 showing the RockBox logo
22:41:00Jozijust wondering the other day is rockbox more efficient with bat life than the original iriver fw?
22:41:24 Join Kaggen [0] (
22:44:27TiMiDJozi: it's difficult to test since the settings differs a lot, but some benchmarks gave iriver more battery time
22:44:47TiMiDwell almost all :)
22:44:53TiMiDsearch on the forum
22:46:01wubblathen have a look at:
22:46:11wubblaBattery lifetime:
22:46:17wubblairiver h1x0 firmware: long
22:46:22wubblaRockbox: longer
22:46:27TiMiDyes I know
22:46:49wubblanow why's that? :P
22:46:51TiMiDbut that's not what the users who tested concluded at least
22:46:58muesli_-irivers fw lasts longer at the moment
22:47:42muesli_-tested it myself
22:48:23TiMiDhehe I don't know why since both use the same codecs
22:48:31TiMiDat least for ogg
22:49:07TiMiDit would be interresting to do an ogg benchmark to see if it's the mp2 decoder in rb that doesn't performs as fast as in iriver fw
22:49:17TiMiDor if it's something else
22:49:32TiMiDsince ogg codec is the same on the 2 players
22:49:53 Join hardeep [0] (i=hardeeps@norge.freeshell.ORG)
22:49:57 Nick StrathAFK is now known as Strath (
22:53:00 Join amar [0] (
22:55:19ender`see my iRiver in all of it's dusty glory:
22:55:42muesli_-well done m8
22:55:56muesli_-wellcome on the good site
22:56:43*ender` cheated with the JPEG plugin to get that image
22:58:34 Quit DreamTactix291 (Read error: 110 (Connection timed out))
23:00:44ender`Windows chkdsk doesn't seem to like the dump files rockbox creates
23:01:27ender`this is what it says: More than one DUMP_000.BMP entry in folder \.
23:02:06ender`(and dir /x shows that all BMPs have the same short filename, which seems to be what's throwing the fat driver off)
23:02:06 Quit Febs ("CGI:IRC (EOF)")
23:02:28linuxstb_I was about to ask you that. Sounds like a bug in the Rockbox fat driver.
23:03:02***Saving seen data "./dancer.seen"
23:03:57wubblais there any way to find empty directories with a simple linux shell command/bash-one-liner?
23:04:44Maximeniow the max files per dir is at max 900 ?
23:06:33ender`linuxstb: please note that i'm currently using 1 month old snapshot
23:08:09 Quit linuxstb (Read error: 104 (Connection reset by peer))
23:08:13 Join linuxstb [0] (
23:08:15 Join frederic [0] (
23:14:49 Part linuxstb_
23:15:00 Quit frederic ("Chatzilla [Firefox 1.5/undefined]")
23:15:09 Join linuxstb_ [0] (
23:15:33 Quit linuxstb (Nick collision from services.)
23:15:38 Nick linuxstb_ is now known as linuxstb (
23:19:40ender`hm, it's kinda confusing when you navigate to an empty directory...
23:21:23muesli_-add some files ;)
23:21:33ender`what are the new crossfade settings - delay and duration?
23:22:09muesli_-hold on
23:22:25muesli_-just posted by me today ;)
23:24:55ender`ok, then another thing: what's the shuffle repeat mode?
23:25:55Joziwas playing a game there. Pity bat isnt as good then.
23:26:48 Quit markun (Remote closed the connection)
23:26:53 Join markun [0] (
23:31:31 Quit hshah ("Leaving")
23:32:32 Part Kaggen
23:35:51TiMiDhas H300 a gold button ?
23:36:53TiMiDno hold :)
23:37:07Philipoh typo
23:37:23TiMiDif it had a gold button, lot of people couldn't afford to buy one ;)
23:37:55TiMiDit seems there is one
23:38:19Philipyes hold button on side of player
23:38:21TiMiDbut that it's not implemented in simulator
23:38:33TiMiDthat's why I get those errors
23:38:44TiMiDand red build grrrr
23:39:39 Quit tvelocity ("Leaving")
23:43:27amiconngold build on red button? jk
23:43:52amiconnSeriously, there seems to be a stub missing for h300 sim that is present for h120 sim
23:44:20Philipyeh the RB logo is red and the Background colour is gold :)
23:45:16 Part Philip
23:45:18amiconnTiMiD: The fix is trivial... Me fix?
23:46:15amiconnWithin uisimulator/x11/button-x11.c and uisimulator/win32/button.c a section needs to be #ifdefed differently
23:46:26amiconn(at the bottom of these files)
23:47:08TiMiDI will try then :)
23:52:28TiMiDyou can call that a genious patch ^^
23:54:05 Quit _DangerousDan ("Miranda IM! Smaller, Faster, Easier.")
23:54:40amiconnBagder: You could also add an iPod color win32 sim if you feel like it. It should build ok
23:55:03amiconn...and even work to some extent unlike the colour x11 sims ;)
23:57:22xmixahlxsome dude has dozens of refurbed ihp-120's
23:57:58xmixahlx...if anyone is interested
23:58:54preglowanyone have the url of that guy making his own h1x0 firmware?

Previous day | Next day