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 2004-09-10

00:04:17Zagorhmm, Tremor makes liberal use of malloc()...
00:11:06 Join Swampylee [0] (
00:14:59 Quit edx ()
00:18:02elinenbehey, do you guys have a cross-compiler for the iriver target?
00:18:42elinenbeso, at least that is not a holdup...
00:18:52elinenbeit sounds like this will be very feasable
00:19:18Zagoryes, i think so too
00:19:46elinenbewhen Zagor says something is feasable −− then it MEANS it's going to happen!
00:20:08Zagorwell, no promises...
00:21:04zeekoebtw, when is someone going to remove the Shutdown entry in the v1 recorder menu?
00:21:10zeekoeit's only in the way
00:21:18lImbushö ?
00:21:21lImbuswhy that ?
00:21:23zeekoethey told me LinusN was going to do that
00:21:41Zagor"they"? :)
00:21:47zeekoeoh well
00:22:20*zeekoe just complains too much i think
00:22:22elinenbeI just ordered a h140 online... −−> Eric is getting ready for iriver:rockbox −− the clash of titans!
00:22:27pike|u got more hw info from the .de guys yet?
00:22:40pike|all chips accounted for?
00:23:30 Quit JoeBorn ("CGI:IRC (EOF)")
00:23:30Zagorpike|: yup, Dave Hooper nailed the last one i think
00:23:44pike|sweet, gulligt
00:24:09pike|is there a dedicated ML for iriver (or perhaps to early?)
00:24:11elinenbe$361 for the h140 −− has anyone seen a better price?
00:24:26LinusNthere is still a mystery chip
00:24:34pike|i can only hunt for low prices in .se
00:24:37elinenbepike|: I think it should just be the rockbox mailing list... I mean it is rockbox −− not ArchosBox
00:24:38Zagorwow, that's cheap. where did you find that?
00:24:45pike|got a h120 myself
00:24:48elinenbeComp-U-Plus Direct
00:24:59elinenbeI've ordered from them before −− pretty good place.
00:25:08Zagori paid more than that for our 120 :)
00:25:09elinenbeand Harmony Computer right here in New York has them for $369
00:25:24pike|elinenbe, thats certainly true, but will the 2 versions be compatible at any level ?
00:25:52Zagorpike|: i agree with elinenbe. we won't benefit anything by splitting up the mailing list.
00:26:18elinenbethere are not THAT many mpe hardware hackers out there −− there really is no benefit ot have more than one list.
00:26:36pike|let me just introduce myself, swede, 30ish, male naturally. Project Manager for XBMC and proud ihp120 owner
00:26:51elinenbeit would be great if we could combine the neuros efforts, the ipod on linux efforts, our efforts etc.
00:27:00elinenbepike|: xbmc is a nice piece of work.
00:27:14LinusN"naturally male", no gender change surgery then. :-)
00:27:15elinenbepike|: quite amazing actually −− I LOVE it.
00:27:23pike|elinenbe, yeah, we're "getting there" slowly but steady
00:28:09elinenbepike|: what is missing? really? other then built in shoutcast support :) (without using the python script)
00:28:26pike|we support shoutcast, just make a strm file
00:28:27elinenbethe iriver has REALLY nice build quality.
00:28:45elinenbepike|: I want to browse shoutcast streams like I do on
00:28:50pike|xbmc dvd handling is kinda weak, mmenues etc
00:29:05pike|elinenbe, hmm, I think that is in the works
00:29:12pike|we got nullsofts blessing
00:29:53elinenbepike|: really? awesome!
00:30:14elinenbepike|: yeah that is a HEAVILY requested feature (running DVD .iso images, and menus)
00:30:31elinenbe</xbmc discussion>
00:32:05pike|yep.. efnet #xbmc is the place for those discussions :p
00:34:55 Join bagawk [0] (
00:37:55 Join SmoothOperator [0] (
00:38:03SmoothOperatorhey all
00:38:21lImbushi Smoo
00:39:04ZagorLinusN: which is a mystery chip?
00:39:43LinusNlast in the hardwarecomponents wiki page
00:39:51bagawkMystery chip on what? The iriver?
00:40:51Zagorlw052a? what's mysterious about it?
00:41:50LinusNreload the page, zagor
00:42:10lImbuslooks like bababa, barzvls. vrml. gugugugu
00:42:29LinusNhad to use some serious photoshop tricks to make out the markings
00:43:13LinusNhas anyone looked on the back of the lcd?
00:43:41Zagorunless you mean "on the board behind the lcd"
00:43:44lImbusLinusN, don't you have a binocular microscope for such things ?
00:43:54 Join tfinn [0] (
00:44:03LinusNlImbus: hehe, no
00:44:24LinusNthere are some markings on the side of it
00:44:32Zagorthe lcd?
00:44:39LinusN0301W and 3ZA
00:44:44LinusNyes, the lcd
00:45:29***Saving seen data "./dancer.seen"
00:45:33pike| is not there, but the back counterpart is
00:45:48lImbusLinusN, too bad. makes soldering smd and such stuff much easier. no coffe for 2 days, and ready you are.
00:45:57Zagorpike|: oops, mistyped link. me fix.
00:46:12pike|large not big, hehe
00:48:41pike|I assume you know the disk is fat(32?) formatted..
00:49:01LinusNlImbus: yeah, we had a microscope like that on my previous assignment, really handy
00:49:01pike|it's like one huge usb memstick when you connect it to the pc, lol
00:49:28Zagoryes, it's the same with the archos.
00:49:46pike|ok. I never had one of those
00:51:13pike|there's no pda o/s running on this cpu already, is there?
00:51:24pike|cpu model I mean
00:51:39Zagorthere's uclinux available
00:51:48tfinnwhat's the processor in the archos?
00:51:58Zagortfinn: hitachi sh-7034
00:52:08pike|just for clarification, I meant the coldfire cpu
00:52:23Zagorpike|: yes, so did I
00:52:31tfinnthe original palms are predecessor to coldfire
00:52:42tfinncoldfire is new-generation 68k
00:53:09Zagortfinn: yup
00:53:11tfinnbut really I would think for an audio player an RTOS would do a better job
00:53:27tfinnsomething like RTEMS or eCOS
00:53:29dwihnoAnother day, another dollar.
00:53:31Zagortfinn: absolutely. we are using our own simple kernel.
00:53:38dwihnoGood morning, fellow boxers.
00:53:53zeekoegood night, dwihno :)
00:54:03dwihnoNah, I just woke up
00:54:11tfinnzag: how much ram/flash is available?
00:54:30Zagordwihno: "i woke up" is not immediately equivalent to "it is morning" ;)
00:54:41Zagortfinn: on iriver 2 MB flash and 32 MB ram
00:54:48dwihnoZagor: Well of course it is :)
00:55:03dwihnoZagor: It's like it's eftermiddag after you've eaten middag :)
00:55:39pike|it can be a Bad Morning also
00:55:49dwihnoIt's friday! :)
00:55:54dwihnoThere's no such thing as a bad friday
00:56:13dwihnoAnd I already bought breakfast - curry/chicken baguette, featuring milky shake deluxe! :)
00:59:05pike|btw, if you want optimized software routines, dont forget to check the foobar SDK
00:59:11pike|he optimized the hell out of some things
00:59:36dwihnohas there been any development already?
00:59:49dwihnoThat would explain why Zagor's still around and not in bed :)
00:59:51Zagorpike|: yes, but it is ia32 only isn't it?
01:00:38pike|I know he fixed some crucual bug in the ISO mpeg specs :p
01:01:15Zagoryes, the specs are buggy. the libmad docs also mentioned that.
01:01:37zeekoe[OT] how is the rockbox windows port progressing? anyone having some info?
01:02:01Zagorafaik nobody has started with that
01:02:30 Nick pike| is now known as pike (
01:06:01LinusNhow many gray shades does the 120 have?
01:06:51pike16/4bpp I *think*
01:06:55 Quit SmoothOperator ("CGI:IRC")
01:07:11pikecan look it up for certain if u wish
01:07:29Zagori thought it was 4 (2bpp)
01:07:30 Quit scott666 (Read error: 104 (Connection reset by peer))
01:07:49hardeephmmm, what did people have in mind regarding the windows port that isn't already done by the simulator?
01:08:09Zagorhardeep: just as a test bench for the reorganisation of the rockbox build system
01:08:18 Join scott666 [0] (
01:09:10pike4 it was... my bad
01:09:58pikehello dithering :)
01:10:34Zagorwell, we play movies and show jpegs with 1bpp on the archos :)
01:10:36hardeepZagor: can't you still use the simulator for this?
01:11:20Zagorhardeep: most of it, yes
01:13:58amiconnZagor: Imho porting of rockbox to Win should be a complete port, including playback (e.g. with libmad), in order to properly modularize the code
01:14:02pikeone major difference between h1xx and h3xx is that the h3xx got a clock, but h1xx doesnt
01:14:53Zagoramiconn: yes, it's mostly a matter of how much work one wants to put into the "test bench"
01:14:56dwihnoHow about iriver batteries?
01:15:09dwihnoAre they replaceable?
01:15:16Zagordwihno: custom li-polymer
01:15:24dwihnoli-poly? neato!
01:15:40dwihnoAlthough I hate the idea of custom batts.
01:15:46dwihnoWell, as long as they are replaceable
01:15:46amiconnZagor: If done properly, this should make further porting much more easy.
01:15:56pikecosts 500sek to replace it @ irivernordic
01:15:57dwihnoamiconn: agreed
01:16:06Zagoramiconn: I agree
01:16:09dwihnopike: quite cheap.
01:16:27pikeyeah, ok price
01:17:54dwihnoas compared to the ipod (ipay?) fes
01:17:59Zagor"FLAC is asymmetric in favor of decode speed. Decoding requires only integer arithmetic, and is much less compute-intensive than for most perceptual codecs."
01:18:06Zagorsounds promising
01:18:07lImbuslol. ok, I now see sek means swedish kroner. I tought about: custom battery swap takes 500 seconds, geeeeez
01:19:21lImbussee, I'm ready for bed. I guess we'll continue laughing tomorrow
01:19:31dwihnoit's already tomorrow ;)
01:21:33lImbusyeah, as you already slept, we went trough that a few minutes ago. But when I come home from the office, it's always today, always time to go to bed, and always late.
01:21:58lImbuseven harder if I ass around in a strange irc all evening long ;-)
01:22:50 Quit Swampylee ()
01:23:38lImbuscya geeks
01:23:58 Quit lImbus (" HydraIRC -> <- *I* use it, so it must be good!")
01:25:21*zeekoe -> bed too
01:25:38zeekoehave fun with being euphoric
01:26:44Zagori'm off to bed too. see you in a few hours.
01:27:01 Quit Zagor ("Client exiting")
01:27:14 Quit zeekoe ("bed time")
01:34:05bagawkit is neat that my silly experiment with the v2 firmware started a new porting of rockbox :)
01:34:56 Join [IDC]Dragon [0] (
01:35:04bagawkhey [IDC]Dragon
01:35:14amiconn[01:34:07] <bagawk> it is neat that my silly experiment with the v2 firmware started a new porting of rockbox :)
01:35:14amiconn[01:34:58] *** [IDC]Dragon ( joined
01:35:19 Join mrrrl [0] (
01:36:17[IDC]Dragonhave been busy beeping
01:37:14bagawkbeeping what?
01:37:29mrrrlHaving a problem. I installed Rockbox 2.2 on Studio 20 and when I select a song and press play, I get Error on the display. Any ideas? TIA
01:37:50amiconn[IDC]Dragon: [OT] Any idea where to get an el cheapo spdif optical->electrical converter?
01:38:20bagawkmrrrl: sounds like you need a .rockbox directory
01:38:54[IDC]Dragonlet me check one more think, I have a question to Linus MAS Nielson then
01:39:11[IDC]Dragonamiconn: Conrad, Reichelt, DIY
01:39:13dwihnoHi Jörg! :)
01:39:28mrrrlI did install the dir .
01:39:35amiconnGrr, this .rockbox dir issue gets annoying. Rockbox could simply create it in case it is missing
01:39:54[IDC]Dragonyeah, I also don't like it
01:40:06amiconn[IDC]Dragon: Conrad != el cheapo
01:40:10[IDC]Dragonsince the Ondio has exchangeable media
01:40:13bagawkamiconn: :) good idea
01:40:54[IDC]Dragonamiconn: get a TORX173 from Reichelt, hook it up to 5V, a resistor and a cinch jack
01:41:03 Join John_Walker_Cefe [0] (iS2k4@
01:41:25John_Walker_Cefenick John_Walker
01:41:39 Nick John_Walker_Cefe is now known as John_Walker (iS2k4@
01:41:43mrrrlIf the .rockbox dir is the common issue with getting the Error on the screen I will check it again...
01:46:07LinusNthe iriver lcd has some markings on the cable strip
01:46:14LinusNS4E1 5002
01:46:14amiconn[IDC]Dragon: I can get it from Segor then, which means no online order (& associated fees & delays) necessary
01:46:47LinusNamiconn: rockbox shouldn't create it
01:46:58LinusNthe user should follow the instructions instead
01:47:16LinusNthe rockbox dir contains other useful stuff
01:47:28LinusNmaybe not that vital for the studio though
01:47:34bagawkLinusN: any ideas if the hardware on the iriver flash based players is almost the same?
01:47:45LinusNi don't know
01:48:55amiconnLinusN: Agreed 100% that rockbox should at least warn if .rockbox isn't there, but either create it afterwards, or run without it (perhaps with somewhat reduced functionality)
01:49:35bagawkif i can find a 128 mb player or so really cheap, i might get one
01:51:33LinusNamiconn: you are aware that it is a bug that it hangs?
01:51:44LinusNand i fixed it some weeks ago
01:52:05LinusNit should continue, but the splash() function crashes
01:52:09amiconnAh, he sad he runs 2.2...
01:52:38[IDC]DragonLinusN: the ondia was a GPIO line to MAS pin "SIBI", any idea why?
01:52:50mrrrlOK, the dir was there, I deleted it, mkdir again, still same problem, did it again, this time it worked. I don't understand why it didn't work the other 2 times. Thx for the help...
01:52:51[IDC]Dragon*ondio has*
01:53:06LinusNmrrrl: mkdir isn't the solution
01:53:28LinusNunzipping the entire zip file is the solution
01:54:17 Part tfinn
01:55:11[IDC]Dragonpin 51, serial input word indication
01:57:19LinusNthe markings on the LCD COG: TF3X23
01:57:30[IDC]Dragondatasheet says we should hold it low
01:57:44LinusN[IDC]Dragon: i have no idea
01:58:37[IDC]Dragonwhat undocumented features does this allow?
01:58:49LinusNwav playback?
01:59:34[IDC]Dragonogg playback?
01:59:56amiconnIirc the (unmodified) wav_io module needs pio in
02:01:02[IDC]Dragonwe have only out, the (bidirectional) buffer is hard-wired for reading
02:01:25amiconnSame as in the jukeboxes
02:02:12amiconnFrom the datasheet: (1) For the MPEG bitstreams the word strobe pin SIBI must always be connected to VSS, bits must be sent MSB first as created by the encoder.
02:02:58[IDC]Dragonand if not, we can send LSB first? ;-)
02:02:59amiconn(2) For the MPEG bitstreams the word During enabling the DSP and its interfaces, it is strongly recommended to hold the SIBC Pin low.
02:03:15amiconnArgh, messed up
02:03:27amiconn(2) During enabling the DSP and its interfaces, it is strongly recommended to hold the SIBC Pin low.
02:03:45[IDC]DragonSIBI, you mean
02:04:47amiconnNo, SIBC (copy&paste). Ah oops, I didn't catch that. This has nothing to do with the "feature" you discovered
02:05:34amiconnThere's more that Micronas doesn't want to tell...
02:09:13 Quit John_Walker ("Itaperuna Script 2004 - iS2k4")
02:13:52amiconnAnyway, I'm off now
02:14:02bagawkbye amiconn :)
02:14:05 Part amiconn
02:14:36bagawkmy computer programming teacher at school is really lame :P
02:24:50bagawktime to go bye
02:25:06 Quit bagawk ("umount /dev/brain")
02:32:39LinusNtime to sleep
02:32:42LinusNnite all
02:32:46 Part LinusN
02:33:43pikeI'm trying to findout what cpu dwells in the recent flash based iriver players
02:37:50pikeearly conclusion, they are ARM cpu's
02:40:32 Join gromit`` [0] (
02:40:39 Quit gromit`` (Read error: 54 (Connection reset by peer))
02:44:28 Join nullobject [0] (
02:45:24nullobjecthi all, i just read zagor's post on for the invite to participate :)
02:45:33***Saving seen data "./dancer.seen"
02:46:59nullobjectso what's the latest on the progress?
02:47:05dwihnodunno really
02:47:14dwihnoZagor and LinusN bought a IHP-something
02:47:41 Quit [IDC]Dragon ()
02:47:51dwihnoThere's been some talk about porting rockbox to win32 as an exercise to prepare for upcoming platforms
02:47:52pikedo they have the data on the LCD yet ?
02:48:01nullobjectoh....I managed to get a copy of dogberts ihp.board file for the coldfire emulator
02:48:07dwihnocheck the wiki, it might be up to date
02:48:24nullobjectanyone had a play around with that yet?
02:52:34nullobjecthas anyone actually gotten the emulator to work...i've been tinkering with that myself ;)
02:52:55nullobjectjust wondering if anyone can shed some light?
02:54:49pikereading the thread @ english iriverlounge forum, subforum firmware, wonder how it will end (hopefully everyone kiss and be friends again)
03:01:08pikesad it almost ended there. very glad it didn't!
03:03:20dwihno*fix contact lenses*
03:06:38dwihnoyeah! 300% visibility increase!
03:08:33nullobjectpretty dumb thing to get upset about if you ask me
03:08:38kaboofaInstant improved vision, how could people not like glasses
03:09:23nullobjecti mean, a forum is supposed to be a space for open collaboration...
03:09:30pikenullobject, I agree 100%. "praise the lord" Zagor is bigger than that
03:09:54nullobjectis Zagor the person in charge of the rockbox project?
03:10:01nullobject(sorry for my ignorance)
03:10:18dwihnoHe started it along with LinusN and Badger
03:10:28dwihnoI think he's the official coach of the team.
03:10:43*dwihno just sits at the bench ;)
03:10:46nullobjectah cool, yeh just added my details to the wiki...i'm keen to get hacking my ihp140
03:11:08pikeI'm just here for the ride. Stuff like this fascinates me
03:11:55dwihnoThis project actually convinced me into buying an archos in the first place
03:12:04nullobjecti'll be much more excited when i can actually boot up my ihp into an alt. firmware...even if it doesn't do anything fancy
03:12:40nullobjecti think getting to that stage will be the hardest part. What was your experience with the other hardware?
03:12:40dwihnoAlthough, when I bought it, there was no support for recorder models (ATA errors, no sound, fishy fat driver to mention a few things)
03:13:18pikeI wonder if the SRS stuff is done purely in hardware, because implementing SRS support will probably not happen... heh
03:14:19nullobjecthmm...not sure, i reckon it's done in software tho because it chews up the battery which would indicate that the cpu is under constant load. If there were a chip for SRS then i don't think it would drain the battery as fast :)
03:14:57nullobjectAlso, it's probably a proprietary algorithm...anyone know about this?
03:15:45pikeproprietary as in you gotta licensce it sounds about right
03:16:07pikenothign I would miss. a parametric eq would be better anyway :)
03:16:26pikebass & trebele adjust != EQ
03:16:41nullobjecti agree totally
03:23:39dwihnoInteresting forum
03:24:57 Quit midkay (Read error: 104 (Connection reset by peer))
03:25:01pikenot much in the english forum, try .de instead
03:25:09 Join midknight2k3 [0] (
03:25:49dwihnogot an url?
03:25:55pikesame just .de
03:26:55dwihnoah, okay
03:28:11pikethe ihp140/h140 harddrive is a little thicker than the h120 harddrive. wonder if the companies that makes these hd's will make bigger hd's the same depth as the one in h120
03:30:46pikeh120=5mm, h140=8mm
03:31:10 Quit mecraw ("Trillian (")
03:39:54dwihnoI'm only accustomed to 2.5" regular laptop disks... is this some new kind of standard?
03:51:40 Join bc|feeding [0] (
03:51:59 Nick bc|feeding is now known as BC (
04:02:02 Quit mrrrl ("CGI:IRC (EOF)")
04:22:49pikeit's 1.8" disks, made for the sole purpose of portable mp3 players :)
04:23:05pikewhen I'm not here I play Burnout3 on the xbox..
04:26:53dwihno1.8"? Coolness!
04:27:18dwihnoI guess you can't buy them yourself?
04:29:23BCooh - no sorry, that looks a 1.8" disk in a pcmcia card
04:30:41dwihnoFishy :)
04:35:05 Quit nullobject ("Leaving")
04:38:03dwihnoLiving in these ages is quite fun
04:38:11dwihnoElectronics becoming smaller and smaller
04:38:16dwihnoFast and faster
04:38:22dwihnoExpensiver and expensiver ;)
04:38:27BC40TB on a holographic credit card recently
04:38:53dwihnoThe new ipods use 1.8" as well?
04:39:10BCas do many of the old ones
04:39:48dwihnoI wonder about the voltage
04:39:54dwihnoand power consumption as well
04:40:15BClower power
04:40:18dwihnoI'd like to see a picture of an archos next to an iHP...
04:40:23 Quit hardeep ("[BX] Mike Tyson says BitchX BITES! Do you HEAR what I'm saying?!")
04:40:25dwihnoI'll have to ask Zagor about that
04:45:35***Saving seen data "./dancer.seen"
04:46:04 Part taaz
04:59:25 Part BC
05:24:15 Join maikeul [0] (
05:32:47 Quit gromit` (Read error: 60 (Operation timed out))
06:13:40dwihnoyeah, found out about that
06:14:11pikethere's prob 60 &/ 80 available also
06:14:47pikeor... there was this new hd that was supposed to be in an upcoming ipod
06:15:00pikespecs about the hd leaked, so steve jobs got mad as hell
06:15:43dwihnoI wonder why ;)
06:16:09dwihnoWindows is a interesting creature
06:16:25dwihnoI've had problems when I work with larger files residing on our server
06:16:36dwihnoThe network driver totally fucked up
06:16:43dwihnoSo now I've installed a newer driver
06:16:50dwihno... it seems to work better now.
06:17:05 Join edx [0] (
06:17:07 Quit edx (Remote closed the connection)
06:19:53pikeyou jumping on the iriver train also, dwihno ? or you'll wait awhile till there's something usable ?
06:20:36dwihnopike: I jumped on the archos train before there was anything usable
06:21:18dwihnobut to answer your question... right now, my unit works flawlessly, and as long as it does, I do not think I will use my hard-earned money on an iriver :)
06:21:28pikeso you'll stick with it till it works no longer? ^_-
06:21:59dwihnoat least until I don't know what to spend my money on ;)
06:23:10dwihnoWhat do you think? Will there be a rockbox for the iriver platform?
06:24:08pikewell, like Zagor says.. they promise nothing. But it sure looks like they're gonna give it a try
06:26:01dwihnoIf the Linus/Bag/Zag team says they will try building a space rocket, I would buy stocks in their company :)
06:26:20pikexchat sucks on so many levels.. oh man, why did I ever switch from trusty old mirc
06:26:46pikexchat windows v2.4.0 = the suckage
06:29:08dwihnoyou can switch back
06:29:19pikeand I will. soon
06:45:37***Saving seen data "./dancer.seen"
06:59:28 Quit scott666 ("i'll be back...eventually...")
07:00:08 Quit midknight2k3 (
07:00:44NJoinmidknight2k3 [0] (
07:53:02 Join ashridah [0] (
07:54:59 Join lImbus [0] (
08:00:47 Join plok [0] (
08:02:49plokIs anyone awake?
08:03:10lImbusI'm trying to get awake *more coffee*
08:03:58plok:) My archos is dying, I wonder if anyone else has seen these symptoms? :
08:04:17lImbuswhich symtoms ?
08:04:25plokIt usually powers off within about 2-30 minutes of use, no matter how charged or new the batteries are
08:04:34ploktried both rockbox and original firmware
08:04:52plokIt's an original 6gig player that I got a long time ago, so I think it's just dying a normal death?
08:05:41plokIt just shutsdown when it needs to access the disk after a while.. strange
08:05:53lImbusI know of bad battery connectors, so you may check that.
08:06:26plokThe things that the battery touches at each end of their housings?
08:07:08lImbusFurthermore, the 4 batteries are 2 in series and two in parallel. so if one side (left or right) is not correctly working, it drains only the other side, leading to such probs.
08:07:45lImbusDo you feel handy opening and taking apart your device ?
08:08:07plokNever done it before, but I'm almost certain that I'll buy a new one anyway, so why not?
08:09:05plokCan I do it with a screwdriver?
08:09:14lImbusit's pretty easy. You'll find a fine explanation on the rockbox-page somewhere. you should get rid of the bumpers and check if everything beneath that is ok.
08:09:30lImbuslike broken soldering points and so on.
08:09:37 Join LinusN [0] (
08:09:59plokHmm..thanks! I'll give it a go. I wouldn't be suprised as I've had it I guess over 4 years now and it's been banged around a lot
08:10:10lImbusscrewdriver ? you'll need a tiny screwdriver (for the 4 screws at the top and the lower boundary) as well as torx
08:10:30lImbusif you want, I can give you some pictures how it should like.
08:10:38lImbusiirc, i made some pics before
08:10:55plokI'd really appreciate that, should I messgae you my email, or do you have a link?
08:11:11lImbusmhmm. lemme first check my pictures
08:11:39plokOf course, with the problems I've been very tempted to just go out and buy a new iRiver H340.
08:11:47plokDo you know if they are good?
08:12:02plokI'd heard some rumours that rockbox might one day make it to the iRiver units?
08:12:36ashridahplok: one of my mates has the h340. they look decent, but the LCD remote that comes with the 1xx series is better (although you can buy it separately and use it with the 3xx)
08:12:48 Quit ashridah ("Client exiting")
08:12:50lImbussee yesterdays irc-logfile. somebody was saying they had a better keypad that the h1xx, but something else was bad. must 've been something like less battery power but more weigth and bigger
08:13:07 Join ashridah [0] (
08:13:48lImbusplok: There has just been bougt the first hp120 and been dissected yesterday. there is no code yet running, and still a lot of work to do. a LOT
08:14:15lImbusI dunno if it's even clear it will work one day.
08:15:21LinusNwell, i can tell you that it will be harder to reverse engineer than the archos
08:15:55lImbusa Linus, didn't see you coming in. good morning. what timeframe do you suppose ?
08:15:57LinusNthe cpu is a BGA package, making it impossible to trace the signals with a multimeter
08:16:23ashridahLinusN: i would have assumed the high frequency more or less ruled out using a multimeter anyway
08:16:27plokoh well, fingers crossed. If I do get a new player, is there any call for an early JB6 player from the developers?
08:16:41LinusNso we will have to rely a heluva lot more on disassembling the original firmware to find out how the hardware is designed
08:17:02LinusNplok: studio or recorder?
08:17:14 Join [IDC]Dragon [0] (
08:17:20LinusNmorning [IDC]Dragon
08:17:33[IDC]Dragongood morning
08:17:35plokit's just a player
08:17:41plokoriginal thing, 6 gig one
08:18:04LinusNplok: sounds like a good paperweight :-)
08:18:28dwihnoWelcome back, Linus
08:18:39[IDC]DragonLinusN: I need to ondio-adjust a few drivers, am not sure about the best way
08:18:54dwihnoPerhaps just IFDEF?
08:19:00[IDC]DragonI2C is on different pins, keys are different
08:19:03dwihnoah, okay
08:19:56[IDC]Dragonthe #ifdef strategy was to not name the hardware, but the feature flavour
08:20:24 Join midkay [0] (
08:20:32[IDC]Dragonso should I do like #ifdef ONDIO_I2C_PINOUT ?
08:20:38plokLinus, do you know if the H340/320 are much the same as the hp120/140's, in terms of porting rockbox to either?
08:20:45LinusN[IDC]Dragon: yes
08:20:46lImbusplok, I'm just uploading 2 pictures to my webspace
08:20:53plokAwesome :)
08:21:07LinusNplok: they look a lot the same on the pics i've seen on the web
08:21:29LinusNi see two major differences, the color lcd and usb2go
08:21:44LinusNsame cpu
08:21:51[IDC]Dragonbut I hope we'll leave that road for "real" porting
08:22:22lImbusplok, did some pics when I swapped the big hdd from the device I got first (a used device) to the brandnew device that was flasheable
08:22:26LinusN[IDC]Dragon: the "real" port will probably have different source files for different drivers
08:22:33LinusNlike i2c_ondio.c
08:23:10LinusNand the config script will select the drivers
08:23:10[IDC]Dragonthe makefile is not really suggesting that, because we name directories, not single C files
08:23:17lImbusplok, I suppose the device I bought used must have had such problems like you have. that's why I has been mended
08:23:42LinusN[IDC]Dragon: today, yes, but we'll change that
08:24:06LinusNthat's the approach we are thinking about now anyway
08:24:07plokIt's only started happening in the past 6 months or so I guess. Also, the battery life has become atrocious, which would seem to fit in with your theory
08:24:35plokawesome :)(
08:24:38 Quit midknight2k3 (Read error: 104 (Connection reset by peer))
08:24:56lImbussorry for bad quallity, it#s just an old cheap digi-cam
08:25:37LinusNa flatbed scanner is the best tool for PCB images
08:26:14lImbusI know, but I didn't want to make pictures of the pcb. It has some 3rd Dimension ;-)
08:26:15[IDC]Dragonnot my LED one
08:26:30LinusNlImbus: saw that now, my apologies
08:26:41[IDC]Dragonit has a focal depth of maybe 0.5 mm
08:26:41plokthey're fine. What was done to fix it? In the wrong one the coil seems to going off into nowhere?
08:27:07lImbuscorrected typo:
08:28:07lImbusplok, I suppose it started detatching, so 'they' (whoever repaired) just stick it to the contact
08:28:27lImbusLinusN: don't bother.
08:29:44plok(lower case was working for me anyway) Hmm.. you mean on the back where we can't see in the second photo? It seems in the first photo, the bit of wire that drags the coil down to one side is gone completely in the corrected version? Am I all wrong? :)
08:30:54lImbuserr. it's the other way around. "HowItShouldBe" is how it should look like on a new device
08:31:02lImbusit WAS a new device
08:31:20lImbusHowItWas looks like a solution. It was the used device I bought like that
08:32:09lImbusand as you saw correctly, I forgot to make pictures of the back, and the wire straight through it is to hold it down solely
08:34:57plokOk. I think I understand. rb_HowItWas.jpg is a correctly functioning configuration also?
08:36:12lImbusyes. it was repairing, but it looks handmade, at least on the backside. so I suppose it's repaired
08:39:14[IDC]DragonLinusN: your port pin assignment table is my valuable resource right now!
08:40:16*plok is away - Automatically set away. - messages will be saved.
08:40:49LinusN[IDC]Dragon: nice to have, isn't it :-)
08:44:42[IDC]Dragonbut I keep theOndio map separate for now
08:45:38***Saving seen data "./dancer.seen"
08:46:55plokIs there a link that will tell me what Ondio means?
08:47:06[IDC]Dragonyes, see twiki
08:47:20lImbusOndio is a smaller Archos device, no hdd, all MMC
08:47:49[IDC]DragonLinusN: the I2C code sets PB5 high, regardless of model
08:48:42 Join amiconn [0] (
08:48:55amiconnhi all
08:49:05lImbushi jens
08:49:05[IDC]Dragonwhich is MAS enable on player, charging off for recorder, power hold for FM/V2
08:49:30amiconn[IDC]Dragon: (sorry for log peeking) I discovered this bug some time ago, when investigating the recording prob with rombox
08:50:08amiconnI did not do something about it yet. Imho PB5->high belongs to something like mas_init()
08:50:25[IDC]Dragonyes, but only for players
08:50:40[IDC]Dragonelse it belogs to charging or power
08:50:48amiconnyes. On recorders it could do the mas poweron reset instead
08:51:28amiconnThen we could switch to using the normal reset in all other code
08:53:38 Part LinusN
08:53:48 Join LinusN [0] (
08:53:52lImbuswhy does everybody excuse for log peeking. what's wrong with that ?
08:56:07 Quit ashridah (Read error: 60 (Operation timed out))
08:56:27ploklImbus, I just went downstairs and bought a H340!
08:57:16amiconn[IDC]Dragon: I already tried that (mas_init() and changing the reset) back then, but reverted it because it didn't help the recording problem
08:58:15plokHmm.. it's much smaller than my archos.
08:58:55lImbusplok: ouch. don't like the feeling of loosing so much precious money in so less time.
08:59:54ploklImbus. I was looking for an excuse anyway.. I think you scared me when you told me my old archos might be fixable :)
08:59:55lImbusI hope you'll get rockbox running on it once
09:00:42plokThere are two usb ports, on says 1.1 Host. They are different form factors
09:01:50plokI have to wait for it to charge before I can play with it :(
09:02:47lImbushost means you can connect another usb-device to it. for example your digital camera to empty the memory of it.
09:03:03plokhow about my archos?
09:03:28lImbuserr. yeah, why not.
09:03:41lImbusif it stands the battery time ;-)
09:04:16plokYeah.. at least now I won't be scared when I'm pulling it apart and having a look :)
09:04:26plokI wonder if you could plug this thing into itself?
09:04:55lImbusyou could. but don't switch it on then ;-) can't be good
09:05:44lImbusevery connected to pc's with the archos usb-a to usb-a cable ? never seen a machine crashing so fast ;-)
09:07:06plokHeh.. of course, my IBM's laptop's USB ports are broken. If you plug anything into it, it instantly shuts down.. sometimes you can hear a spark
09:07:23plokJust happened when it came out of warranty
09:07:36amiconnlImbus: When doing that you may even fry the mobo/ usb card
09:07:55lImbusI know, it was not mine ;-)
09:08:07plokWait, was it an IBM laptop?
09:09:13lImbusno, two standart pc
09:10:48 Join Zagor [242] (
09:12:56pikegooood morning, or.. goood night (for me, lol)
09:13:57amiconnNow gcc 3.3.x is used for the daily builds... Nice :)
09:14:05pikeP0NG 9:13am
09:14:12 Join ashridah [0] (
09:16:40LinusNZagor: morning
09:16:52plokI can't find any LEDs on the H340
09:17:31ashridahplok: i think there's two on either side of the nav pad under the LCD, just like the 1xx series
09:19:02plok....maybe they'll light up when it's playing?
09:19:18ashridahone'll be the hard drive light, the other is the 'charging' light
09:19:28lImbusone of them is directly to the hard drive bus.
09:19:48ashridahlImbus: i had a feeling that'd be the case
09:20:26pikewell hidden...
09:20:59plokthere is as a little reset recessed button on the side of the unit though, I wonder if that resets the firmware?
09:21:16pikereset = just if the unit hangs
09:21:24pikehappened me once on my h120
09:21:29pikeit's a reboot
09:21:32plokahh, thanks
09:22:03plokpike, are you saying there are leds ther just very hard to see?
09:22:08ashridahplok: the device is mostly screwed if you break the ahrdware, from what i've seen in the forums so far
09:22:21pikeI'm saying I dont know where they are, if there are any
09:23:14plokpike, well I've looked over this one about 5 times and I can't see any, but I just bought it and isn't charged yet so I haven't been able to turn it on yet
09:23:35pikesomething should light up when you charge, atleast it does on h1xx
09:23:47ashridahodd. my h140 came charged enough to use the device for an hour or so, and longer if i used the radio
09:24:21plokI haven't tried turning it on yet.. the dude in the stote told me to charge it before using it.. is that baloney?
09:24:38ashridahit's probably a good idea
09:24:41pikeif it's empty, it's always good to give it a full charge
09:24:43ashridahbut almost no battery is sold completely flat
09:24:56ashridahsince most will sulfate and screw themselves
09:25:08ashridahthe li-poly in the iriver series isn't supposed to do that tho
09:25:25LinusNmy iriver was completely flat when i bought it
09:25:52plokhmm.. I'll just charge it when I get home. I can wait a few hours to answer the pressing LED question
09:26:21ashridahit can charge while on, so you can poke at it completely then
09:26:38[IDC]Dragonsorry to bug you again with the ol' Ondio
09:26:53[IDC]Dragonthe buttonsare difficult to translate
09:27:12[IDC]Dragonit has only one common on/off button
09:27:32[IDC]Dragonthe direction keys, and one "option" button
09:27:50Zagorcan you short-press the on/off for stop?
09:27:59 Join pillo [0] (
09:28:09[IDC]Dragonyou mean in the original firmware?
09:28:11Zagoror is it impractically located? (can't see it on your photo)
09:28:21[IDC]Dragonit's on the top side
09:28:21 Quit pillo (Remote closed the connection)
09:28:29Zagorno, i mean is it technically feasible or does it turn off even on short press
09:28:39LinusNbut the browser<->wps toggle will collide with the stop functionality
09:28:51[IDC]Dragonit turns off after a very long press
09:28:53LinusNif the on button is the same as the off button
09:29:17 Join pillo [0] (
09:29:17ZagorLinusN: yes, but we need to rearrange a bit since we have fewer buttons
09:29:40Zagormaybe short press on option to switch between wps and browser and long press to enter menu
09:30:31[IDC]Dragonsee the user manual on how Archos did it:
09:32:28[IDC]Dragonthe UI is quite nice, actually
09:38:09[IDC]Dragonthey use double clicks and long clicks as modifiers (not that I like it, but better than nothing)
09:38:56[IDC]Dragoninstant recording is double clicking the on/off button
09:40:06 Join Lynx_ [0] (lynx@
09:40:38lImbusmental note: I am looking for instant rec on the recorder for a while, and double-on is not yet taken on the recV1, isn't ?
09:41:10[IDC]Dragonwe have no double clicks yet
09:41:43lImbusI mean, there is no way to confound anything.
09:42:19lImbusuhh, there is one: If I am in the browser, want the unit to resume and go to the browser back again, I double click ON
09:50:16 Join Bagder [0] (
10:00:04LinusNlImbus: that's not a separate double-click event
10:00:26LinusNyou perform the same function by clicking it slowly
10:06:41lImbusok. but I DO click very quickly for doing that at the moment. and it works. so implementing doubleclicks could mind some users.
10:07:06lImbusgotta go now. will peek the log, when I#m at work.
10:07:41 Quit lImbus (" HydraIRC -> <- The dawn of a new age")
10:18:08 Quit plok ("I'm outta here!")
10:39:25 Join webguest81 [0] (
10:41:28 Quit Bagder ("Leaving")
10:45:42***Saving seen data "./dancer.seen"
10:48:59 Quit webguest81 ("CGI:IRC (EOF)")
11:09:31[IDC]DragonI'm pretending the Ondio to have a Player keypad now
11:10:12[IDC]Dragonlots of places to fix, we have the graphical LCD, but the limited keys
11:10:49Zagoryeah. the button issue is a rather difficult problem when supporting more platforms
11:10:53dwihno[IDC]Dragon: How far have you gotten btw?
11:10:58 Join lImbus [0] (
11:11:23[IDC]Dragondwihno: see twiki (my standard answer)
11:11:41dwihnowiki wiki wang!
11:11:54dwihnoWiki - "How are YOUR wang today?"
11:23:00[IDC]Dragonany wps.c expert here?
11:23:41Zagori wrote much of it, iirc
11:23:42LinusNi'm here
11:24:03*LinusN is not an expert, but he's here
11:24:24[IDC]DragonI have a problemfor this recorder without a recorder keypad
11:24:26LinusNZagor: it's good that you remember that you wrote it :-)
11:25:00Zagorcvs annotate agrees with me :)
11:25:21[IDC]Dragonthere are 2 status_set_xxx functions to update the charcell display status
11:26:23[IDC]DragonI fall intothere, but have no player
11:26:48[IDC]Dragonjust looking for the clean way, no extra #if
11:27:40Zagor#ifdef _is_ the clean way...
11:28:01[IDC]Dragonnot if it's too many of them
11:28:22Zagorotoh I don't think the player button set is the best starting point
11:28:46Zagorthe ondio has true 4-way navigation, the player doesnt
11:28:51[IDC]Dragonelse it'll be muchworse, I thought
11:29:12 Quit pillo (Read error: 104 (Connection reset by peer))
11:29:13[IDC]Dragonbut the 4 way doen't leave much else
11:29:22Zagorwell it will be rather messy either way, but I think treating it like a player makes a worse UI
11:30:02[IDC]Dragonto get it working and functionaal, I prefered that
11:30:19Zagorup/down = volume, left/right = next/prev & ffwd/rew, option = browse/menu, power=stop
11:30:25[IDC]Dragonthe missing recorder keys would make big parts inaccessible
11:30:34Zagoronly the quickscreens
11:31:12[IDC]Dragonhmm, can I assign it that freely?
11:31:15Zagoractually i think power should be pause. stop is pretty useless on an mp3 player
11:31:56Zagorup/down/left/right and power is pretty simple, but the option button needs some hacking
11:32:07Zagorlunch time, back in a while
11:32:08[IDC]DragonI'd need the functionality of F1(Menu), Play, On, Off
11:32:19Zagorhold option to get menu
11:32:21[IDC]Dragonandhave only Ontion, OnOff
11:32:36Zagoryou don't need play in wps, only pause on/off
11:32:39[IDC]Dragonwe have no long press yet
11:32:45Zagorwe have repeat
11:32:58Zagoranyway, back soon
11:33:18LinusNjust use |BUTTON_REPEAT to sense long key presses
11:34:22[IDC]DragonOK, let me summarize:
11:34:42[IDC]DragonMenu is option+repeat
11:34:56[IDC]DragonPlay is option
11:35:07[IDC]DragonOn is OnOff
11:35:19[IDC]DragonOff is OnOff+repeat
11:36:04[IDC]Dragona repeated button causes no normal (unrepeated) event first?
11:36:47LinusNyes it does
11:37:05LinusNuse the release events
11:37:22LinusNrelease without previous repeat == short press
11:37:46[IDC]Dragonthat would require extra code in the handlers
11:38:19[IDC]DragonI'm trying to port that part eith the button driver, if possible
11:39:10LinusNi don't think rockbox would benefit from a quick'n'dirty port to the ondio
11:39:37LinusNthis is a good opportunity to deal with the problem with different keypads
11:39:57[IDC]Dragonthe Ondio woudn't benefit from the quick'n'dirty port of Rockbox
11:40:30[IDC]DragonI try not to break(touch) the rest
11:40:57[IDC]Dragonbut as the very first step I want to use it somehow with the buttons
11:41:00LinusNwe will have to change that anyway, sooner or later
11:41:12[IDC]Dragonwithout doing a UI overhaul first
11:41:49[IDC]Dragonyes, but when this thing moves, not before
11:42:03LinusNi understand
11:42:24[IDC]Dragonright now I can't write interactive test code
11:42:43[IDC]Dragonbecause I need buttons first
11:44:24 Join vadin2003 [0] (
11:45:26 Quit vadin2003 (Client Quit)
11:47:18amiconnLinusN: A completely different topic: There are a lot of patches in the tracker, dealing with language things, that are (auto- ?) assigned to quelsaruk. However, I haven't seen him around for a long time. Perhaps someone else should deal with these patches?
11:50:35LinusNlunch time
12:19:21 Join [av]bani [0] (
12:19:35[av]baniMCF5249 PLL Configuration Wizard
12:20:14Zagor[IDC]Dragon: about repeat: you have two events: button-repeat and button-release. no need to make a complex state machine.
12:20:34[IDC]Dragonbut a tiny one
12:21:10[IDC]Dragonbecause I get a regular buttonpress first, than repeated ones, then release
12:21:34Zagoryes. so if you get repeat, call repeat(), if you get release, call release(). nothing else required.
12:22:08Zagorgetting release after repeat is not a problem, since you are no longer in the save switch
12:22:25[IDC]Dragonbut I still need to react on the normal (=short) press
12:22:33Zagoryes, that is release
12:23:08Zagorwe are handling this already (albeit more complex) with the ON button on recorders for browser and pitch screen
12:23:09[IDC]Dragonok, short press = release, long press = repeat
12:23:44[IDC]Dragonbut after a long press (repeat) I will also get a relaese (shot press)
12:24:02Zagoryes, but then you are no longer executing the same loop
12:24:17Zagoryou have moved to the menu, for example
12:24:25Zagor...which ignores release events
12:24:32[IDC]Dragonand somebody else may eat that realease, taking it as a short press
12:25:05Zagoryes, that can happen. but we see to it that it doesn't.
12:25:27[IDC]Dragon"see to it"?
12:25:51Zagornot by general rule, but by making sure the code we call after repeat can handle a stray release
12:26:19Zagor[av]bani: nice
12:27:44[IDC]Dragonhmm, but there is no way to tell a stray release from a short press
12:28:06[IDC]Dragonand there may be more repeat events in the queue
12:28:33[IDC]Dragonthis is not clean yet,imho
12:28:45[IDC]Dragonbut I got other issues first
12:28:49Zagorthe queue can be emptied, and releases are only used when we combine long and short press
12:28:58Zagor...which is pretty rare
12:29:16Zagormost places uses only single press
12:29:41Zagorthis is UI code, it can never be clean ;)
12:30:40[IDC]Dragonanyway, Ihave it compiling now with the Player buttons
12:31:18[IDC]Dragonsortes out several confusions about display and keymap not being independent
12:31:54[IDC]Dragonno extra code, just better #if's
12:32:09Zagoryes, i expect we have some bad ones
12:32:37[IDC]Dragondoes anybody mind a commit of that?
12:33:23*[IDC]Dragon goes build-painting
12:45:46***Saving seen data "./dancer.seen"
13:02:36LinusNi can't find a gdb sim target for coldfire :-(
13:02:52Zagordoes m68k not work?
13:03:02LinusNi can't find that either
13:06:12LinusNso we'll have to write one.. :-)
13:06:47LinusNi really thought there would be a 68k sim target...
13:10:10dwihnoLinusN: Luckily your area of expertise covers GDB stubs! :)
13:10:18dwihnoAnd virtually everything else :)
13:10:57LinusNyeah, i did patch the SH sim target to simulate the archos LCD
13:11:31*LinusN is thinking about designing a BDM wiggler
13:12:24[IDC]Dragonthe Ondio is bad for gdb
13:12:26Zagorwe'll need that sooner or later, so why not sooner
13:12:36[IDC]Dragonbecause the serial is taken for the MMC
13:12:37Zagor[IDC]Dragon: why?
13:13:13[IDC]Dragonlet's hope I won't need gdb
13:19:47[av]banilinux kernel has some mcf5249-platform-specific code
13:20:03[av]banian init routine at least
13:22:55[av]banisomoene to talk to about 5249 :)
13:23:42[av]baniThis particular board is based around a Motorola ColdFire 5307 (@90MHz) design, fitted with 16MB of SDRAM, 2MB FLASH, 2 ethernet ports and 2 serial ports.
13:23:51[av]banithey are running uclinux in those constraints
13:23:58[av]baniso uclinux might not be such a bad starting point really
13:24:04Zagoryes, that's no problem.
13:24:15amiconn[IDC]Dragon: You were quite active with that Ondio challenge :)
13:24:42[av]baniyou get a pre made multitasking kernel with full driver support already :)
13:26:02[IDC]Dragonamiconn: I still am
13:28:44amiconn[IDC]Dragon: Two questions (1) The Ondio doesn't have an rtc? (2) Couldn't disassembling the archos fw help us?
13:29:11amiconnA related note: my Ondio is now on its way to me, via dpd...
13:29:35[IDC]Dragon(1) no RTC, (2) I don't think so
13:30:32Zagor[av]bani: yeah, but it's way too heavy to use for a simple mp3 player.
13:30:59Zagorbut some of the drivers can be useful to look at
13:32:14 Join methangas [0] (
13:33:14Mode"#rockbox +o LinusN " by ChanServ (ChanServ@services.)
13:33:48Topic"Sponsor the project: donate your broken iRiver iHP-1xx" by LinusN (
13:37:29*ashridah notes that finding broken iriver players is likely to be a bit difficult until the average person's warranty is up.
13:37:40ashridahonly early adopters are likely to have a broken unrepairable one.
13:38:57 Part [av]bani
13:40:20ashridahdarn. there's some broken flash iriver's on ebay, but that's it
13:42:00LinusNthe warranty doesn't cover everything
13:43:44Zagorthey are selling very cheap new in the US. we could buy one there and have someone ship it.
13:44:29ashridahyeah. the 1xx series is supposed to be discontinued, since they're primarily trying to sell the 3xx series. i imagine the 1xx's will get cheap, then rare, then expensive again :(
13:45:35Zagorcheapest i find is $275, which is about 60% of what we paid here
13:45:50ashridahif mine broke, i'd sell it to you guys for next to nothing plus shipping, provided you didn't want the hard drive (since if it was the hard drive that was broken, i'd just replace that, and keep it)
13:46:22Zagor? where?
13:46:23ashridahZagor: are there any major differences between the 140 and 120 that you know of, minus the hard drive?
13:46:47Zagorsome pictures show different case color, but other than that I don't think so
13:47:19ZagorLinusN: yeah but that's just a bid, with 4 days to go. it will rise, nobody best this early.
13:47:27ashridahyeah, the casing is black paint, and some more recent kits come with a black leather cover, instead of brown.
13:47:44Zagorour 120 came with black leather
13:48:16ashridahZagor: did it? midly interesting. still, i imagine you guys got your own recently.
13:48:26Zagoryeah, yesterday :)
13:49:11ashridahyeah, the guy i know who got one with a black case got his a few months ago. mine is almost 6 months old now (bought it just after easter)
13:49:11 Quit maikeul ("Client exiting")
13:49:25 Join gromit`` [0] (
14:00:16 Part LinusN
14:00:27 Join LinusN [0] (
14:13:31LinusN[IDC]Dragon: #ifndef HAVE_NO_LED
14:13:38LinusNa bit backwards, isn't it?
14:13:51[IDC]Dragonahem yes
14:13:55LinusNwhy now #ifdef HAVE_LED?
14:14:09[IDC]Dragonshould I change theother config files instead?
14:14:35[IDC]Dragonno backlight is next ;-)
14:15:00[IDC]Dragonthe LED was small, I put dummies in
14:15:15[IDC]Dragonbut for the backlight, I don't like that too much
14:15:47amiconn[IDC]Dragon: Iirc we can get rid of the backlight thread?
14:16:26[IDC]Dragonbut #ifdef bracketing every call to the backlight module also won't look good
14:16:59[IDC]Dragonany idea on how to carve this out the best way?
14:18:57Zagoryou could put the #ifdefs in the backlight implementation, creating stubs instead
14:19:48[IDC]Dragonreal stubs or empty macros (avoiding a call)?
14:20:15Zagormacros are fine with me
14:27:54amiconn[IDC]Dragon: I think emtpy macros are in fact a better way than #ifdefing everything. This could be used in a number of other places too (lcd_set_margins() comes to mind)
14:28:36[IDC]Dragonan empty inline function could give the same result, too
14:29:59amiconnOne question is yet to solve: What if such an empty macro gets included into the plugin api struct? Does gcc generate a dummy function?
14:30:26[IDC]Dragonoops, sounds bad
14:30:41Zagoryou will simply get a compile error
14:31:41amiconnPerhaps that works with the empty inline function approach. Iirc gcc does generate a non-inline variant for those cases where a pointer to that function is used
14:32:42amiconnA true empty function does only take 4 bytes (2 instrs) on sh1 (with gcc >= 3.3)
14:36:09amiconnWe could in fact force the compiler to use the same empty function for all those cases by simply defining e.g. dummy_funct() { } and then defining macros for all other functions that are empty referring to dummy_funct()
14:36:40[IDC]Dragonbummer: my Ondio stuff was only compiling until I made a clean make :-(
14:37:09[IDC]Dragonnow keyboard.c heavily fails for missing keys
14:41:46 Nick midkay is now known as midk (
14:42:38[IDC]Dragondefining them as the same like another will fail if both values are in a switch-case
14:45:49***Saving seen data "./dancer.seen"
14:52:44Zagorhmm, libmad uses memmove. not easy to workaround that...
14:53:27midkdoes the ondio use the same size screen, 112x64?
14:53:47[IDC]Dragonmidk: yes
14:53:58 Join Bagder [0] (
14:54:22Zagorhey bagder
14:54:27amiconn[IDC]Dragon: .rvf on the Ondio ;-)
14:54:36Bagderhi ho
14:56:29LinusNZagor: memmove()...badness
14:56:54Zagori'll just add it, it's a small function
14:56:59[IDC]Dragonoverlapping, or why?
14:57:30Zagordoesn't look like a coincidence, it uses memcpy in other places and memmove in just two
14:57:31[IDC]Dragonask amiconn, master of moving memory
14:57:45[IDC]Dragonmaybe an oversight
14:57:58Zagori don't think it is
15:00:20Zagorshould we really fail scrambling when size >200kb? rolo can still load it.
15:00:45Zagorrockbox with libmad is 250kb
15:01:27Zagormethinks a warning should be enough
15:01:34[IDC]DragonZagor: are you testing if it compiles in our platform, or what's that for?
15:02:08[IDC]Dragonwon't run very fast ;-)
15:02:13Zagori was also planning to let it decode a file to wav for testing
15:02:17Bagderhow's the ondio work going?
15:02:21Zagorprobably not, no :)
15:02:49[IDC]DragonBagder: fighting the special cases of having a recorder with limited keys
15:03:07Bagderso most of the drivers work then?
15:03:28[IDC]Dragonwork? compile...
15:04:10[IDC]DragonI'm gradually bringing in the differences for the port mapping
15:04:34[IDC]Dragonthe parts I know of yet
15:04:56Bagderyou're moving on fast, I like it
15:05:26[IDC]Dragonthe bigger task is to write an MMC driver
15:05:41[IDC]DragonAPI-compatible to our ATA
15:05:52Bagderah, right
15:06:18[IDC]Dragonhaven't started on that yet, but found nice code
15:06:26 Join kurzhaarrocker [0] (
15:08:31kurzhaarrockerJust to restart an old flame war: When [IDC]Dragon experiences problems with the keys stuff - why not enhance the key api so that it supports _real_ hold_down_events and double_press events?
15:09:02[IDC]Dragonit may go that way
15:09:06Zagorhow do you separate between hold_down and repeat?
15:09:57Bagderand even repeat and double-press
15:10:00LinusNand how do you separate single-click and double-click?
15:10:14kurzhaarrockergosh I really did start it :)
15:10:26LinusNwe've had this discussion before
15:10:38LinusNbut it's been a while
15:11:00amiconnZagor: If it would be useful on SH1 too, I could extend memcpy.S to also include memmove. The memcpy routine could be reused for down-moving then, and the up-moving routine would use the same optimization techniques
15:11:08kurzhaarrockerI remember that we had it. But my brain doesn't remember enything else about it, LinusN :)
15:13:21kurzhaarrockerConcerning repeat and hold_down: It would be the same thing, just the driver sending multiple events. Like pressed, hold_down, repeat, repeat, repeat..., released
15:14:26kurzhaarrockereg that handlers that are interested in hold_down only don't have to ignore consecutive repeat events
15:14:50Zagorthe events will still come, even if we call them something else
15:15:30Zagoryou're trying to solve a problem we don't have
15:15:46Zagorwe have lots of real problems, i suggest you focus on them instead ;)
15:15:48kurzhaarrockeryes, but then it would be the driver that filteres the events in one spont and not many event handlers that do the same filtering in many places.
15:16:05Zagorno the driver has no idea which events you want. it always sends all events.
15:16:56[IDC]Dragonyou'll reach model-view-controller soon... ;-)
15:17:32kurzhaarrocker<- admits that he is an oo programmer...
15:19:20kurzhaarrockerBut now that we want to support more platforms we might in fact want to introduce a layer of abstract events like ENTER_MENU that is generated by different hardware in different ways.
15:19:33*Bagder added the pretty pictures on the dl.cgi pages too
15:20:08Zagoryes, that is one option. we tried that in tree.c originally. it didn't work out 100% but it might be worth revisiting.
15:20:40[IDC]Dragoncu later
15:20:44 Quit [IDC]Dragon ()
15:21:40Zagorthe UI/button handling is actually one of the hairier issues, architecture wise, when adding new platforms
15:22:05ZagorLinusN: do you have an A-A usb cable?
15:22:15*kurzhaarrocker passes Zagor an electric shaver
15:22:31Zagor:) i could use a good one
15:25:34 Join R3nTiL [0] (
15:32:44 Quit Bagder ("Leaving")
15:35:04amiconnZagor: Got my remark/ question concerning memmove?
15:35:44 Quit kurzhaarrocker ("Trillian (")
15:41:21 Join karacho [0] (
15:42:06karachohi there
15:42:22karachoeveryone feelin comfortable ;) ?
15:43:11karachoi've a question regarding spoken directories
15:43:25karachosomeone aware of that stuff?
15:45:33karachohm, myabe its a better bet to go with the mailinglist.... thx anyway
15:50:26 Quit karacho ("CGI:IRC")
15:53:49Zagoramiconn: yeah, but i don't think we'll need it for sh. if you're itching to optimise, look into the coldfire emac
15:55:15Zagorit is the key to good mp3, ogg and flac performance
15:57:27 Part LinusN
15:58:03Zagormy libmad image crashed horribly :)
15:58:22Zagorbetter fix it in the simulator first
15:59:46dwihnoFix fix! :D
15:59:49dwihnoHooray for Zag!
15:59:53dwihnoZagor makes the day!
16:00:08Zagor:) nice cheering
16:07:48dwihnoHeja Zagor friskt humör, fixa iHP innan vi dör? :)
16:08:01dwihnoNärå. Det behövs' it'
16:11:36 Join edx [0] (
16:22:44 Quit R3nTiL ()
16:26:57Zagorworks nicely in the sim
16:32:21Zagorgotta go
16:32:23 Part Zagor
16:40:25 Join [IDC]Dragon [0] (
16:41:35[IDC]DragonHey, I can browse the Rockbox menuonthe Ondio!
16:41:56[IDC] on the..
16:42:45dwihnoGreat work!
16:43:19dwihnoYou're making the greased lightning appear like a bicycle when you're coding :)
16:43:39[IDC]Dragonno disk yet
16:43:46[IDC]DragonUSB mode to come
16:44:02[IDC]Dragonmost workis left to do
16:44:13dwihnoI'm thrilled you got it working this far :)
16:44:19[IDC]Dragon(sticky space bar)
16:45:53***Saving seen data "./dancer.seen"
16:56:23amiconn[IDC]Dragon: I guess PB14 and PB15 on Ondio are the same as with the jbrs.
16:57:28[IDC]Dragonhave you got you boy, or what leads you?
16:58:37amiconnNo box yet. Too bad this is sent via DPD, a bad solution to send something to someone's home
16:59:48amiconnI was lead to this by comparing the port pin tables for jbr and Ondio. All mas connections you discovered so far use the same port pins
17:02:52 Join webguest21 [0] (
17:03:53 Quit webguest21 (Client Quit)
17:06:43ashridahah, the joys of dodgy programming assignments ( :) )
17:08:05*ashridah notes he probably should have said that in #java, where he originally intended
17:12:45[IDC]DragonROM CRC is the well known one!
17:12:56[IDC]Dragonso I could uart-boot this
17:13:55 Join mecraw [0] (~lmarlow@
17:14:04amiconn[IDC]Dragon: You said uart boot is not possible because of the serial port already taken?
17:14:23[IDC]Dragonwell, if you solder it away
17:14:48amiconnBut then you'll have no mmc access, right?
17:14:55[IDC]Dragonmaybe it can even be tapped at the MMC slot
17:15:10 Quit Lynx_ (" WOW! This IRC Client ownz! HydraIRC -> <-")
17:15:14[IDC]Dragonuart_boot needs no MMC, gdb does
17:18:20 Join webguest30 [0] (
17:19:38 Quit webguest30 (Client Quit)
17:26:48[IDC]DragonI've uploaded an Ondio ajz now
17:29:04elinenbe[IDC]Dragon: everything
17:29:09elinenbe[IDC]Dragon: everything's working now?
17:31:06elinenbewhy the haha? is _anything_ working yet?
17:31:37[IDC]Dragonsee the twiki page, at the bottom
17:31:47[IDC]Dragonthe main menu works
17:31:54elinenbethat's a start
17:32:06[IDC]Dragonunless you do something which does disk I/O
17:32:21amiconn[IDC]Dragon: What flash id does the "hw info" menu item tell you?
17:33:06[IDC]DragonBF D6
17:33:18[IDC]Dragonbut I knew that, saw the chip
17:34:33[IDC]Dragonthey are probably all flashable, since Archos provides a .ajz which does a flash upgrade
17:34:53[IDC]Dragona flashable flash player ;-)
17:37:42elinenbea 1gb mmc card is only about $100 −− that is a nice player for that size.
17:38:28[IDC]Dragona 40 GB disk is also ~$100 :-/
17:42:22 Quit [IDC]Dragon ()
17:49:27 Join Elemeno [0] (
17:49:38ElemenoHi folks
17:50:39ElemenoI'm looking for a replacement mic for my FM recorder. Any suggestions where to look?
17:54:34 Join Zagor [0] (
18:03:56Zagorwhee, the produced pcm file actually sounds great
18:04:07Zagornow why does it crash on the sh...
18:07:34Zagori wonder if the iriver firmware uses all 24 bits of the dac
18:11:45amiconnZagor: Perhaps too small a stack?
18:12:21Zagorwell the thing is my .ajz doesn't even boot, so it's not a problem with the libmad code per se. rather with the build.
18:12:43Zagorwell it boots and shows the logo, but crashed before the browser
18:16:31Zagorahh, worked a little better now. decoding started, then crashed on completion.
18:17:12Zagorthat could be a stack issue. but the display is totally cleared, which is a bit odd
18:17:51 Join R3nTiL [0] (
18:22:56amiconnZagor: Stack issue usually cause all sorts of weird effects: When I implemented my improved screen_dump() I first tried to allocate the 1-row buffer on the stack as well. This lead to a stack overflow, and I got "IllInstr". Now the buffer is static
18:24:24Zagori'm a bit worried about the absence of an error message though. the led is blinking but the screen is clear
18:30:46Zagorhahaaa, success. let's hear it...
18:31:05amiconnPerhaps it crashes in a way the breaks the lcd routine(s). If it crashes by an unhandled interrupt, the screen usually allows rolo reboot. However, this doesn't work most of the time...
18:31:19*ashridah hands zagor the cookie of success
18:31:19amiconnZagor: I'm listening...
18:32:36 Quit ashridah ("sleep")
18:33:07Zagorgahh, typical. now my computer won't mount the usb disk :-{
18:35:14Zagorwheee, works!
18:35:57Zagorit's horribly slow though...
18:36:41Zagorbut that was expected
18:45:54***Saving seen data "./dancer.seen"
18:48:12 Quit Elemeno ()
18:48:35 Quit R3nTiL (Read error: 104 (Connection reset by peer))
19:53:42 Join LinusN [0] (
20:37:49 Part LinusN
20:45:55***Saving seen data "./dancer.seen"
20:48:24Zagorhmmm, how about a mod player...
20:50:00 Join lImbus_ [0] (
20:50:36amiconnZagor: That would be really cool, too bad that we don't have access to that pcm pass-through module for the mas :(
20:50:47Zagoryeah :(
20:51:10 Join benschi [0] (
20:51:16amiconnBut on iriver this sounds like it would be easy once the firmware runs
20:52:02amiconnEspecially as the Iriver is coldfire, and the .mod format emerged from Amiga, which is 68k...
20:53:17Zagorwell the format isn't particularly 68k-specific other than endianness
20:54:23Zagormama mia. libmodplug is *huge*!
20:54:31*Zagor goes looking for something smaller
20:54:40benschiwhats up?
20:54:44 Quit lImbus (Nick collision from services.)
20:54:48benschiis everything running like it should?
20:54:55 Nick lImbus_ is now known as lImbus (
20:55:20Zagorbenschi: yeah, pretty much. i just got the mp3 decoder running in the simulator and on the archos
20:56:22Zagorwe still don't know which display controller it is (although Dave Hooper has made a good guess that might work out)
20:56:54Zagor...and we really would like to get our hands on a broken ihp to remove the cpu from, so we can make proper schematics
20:56:55amiconnZagor: No, .mod isn't 68k specific, but there are many implementations for 68k, even space-saving ones
20:57:56amiconnRemember that it was possible to play (4 channel mods) even on a stock Amiga 500 with only 512K of ram
20:58:16benschizagor, kewl..
20:58:17Zagoryeah, mods require very little procesing
20:58:22benschisorry, i dont own a broken ihp
20:58:30benschiand also nobody of my firends..
20:58:35Zagor(i've actually written music with an amiga tracker myself)
20:58:55benschibut, how long do you think, will it take to get the firmware running on the ihp? only calculations..??
20:58:59amiconnZagor: The only thing that needs additional processing on recent equipment is the stereo downmix
21:00:17pikeonce again, it might be worth to take a look at foobar's code :) Peter Pawloski is a huge fan of tracker tunes himself
21:01:17Zagorbenschi: it will take time. months.
21:01:25benschiok :-)
21:01:33lImbusfoobar, is that the winamp-clone by the nullsoft-developer ?
21:01:34benschihope you get the thing done!
21:01:53pikeby ex nullsoft employee, PP yep
21:02:10pikeand by that definition, anything that plays mp3 is a "winamp clone"
21:02:14Zagorpike: yeah, i have the fb2k code here too
21:02:41amiconnZagor: The similarity between coldfire and 68k may even enable to play module formats which contain some replay code
21:02:48Zagorbenschi: so do I :)
21:03:12Zagorand amiga demos ;)
21:03:29pikehaving been an amiga "addict" the 68k "stamp" doesnt exactly signal "speed"
21:03:44amiconnZagor ;)
21:03:54Zagorthe emac helps to speed up calculations a bit
21:04:23pikeI remember the good ol' days.. mp3, downsampled to mono and 22.05 kHz and played back on my 68030 accelerated amiga
21:04:55amiconnI play mp3s with full resolution on my amiga :)
21:06:09pikehave fun. bbl
21:10:04lImbusZagor [OT]: Has dancer already reported been running on Windows2003 ?
21:10:30lImbusit quits before writing one line of logfile. but touches, and modifies dancer.users
21:10:46Zagori don't know. i've only ran it on linux.
21:17:00 Join HenrikB [0] (
21:22:34HenrikBI'm playing around with the check of the .rockbox dir on startup
21:22:55Zagorhuh? "ColdFire FPU Instruction Set"??? −− CF Programmers Reference Manual
21:23:56HenrikBI've goten the playlist code to play a playlist without a control.file
21:23:57ZagorHenrikB: anything in particular?
21:25:06HenrikBand it shows the missing .rockbox dir message every time you play a playlist if .rockbox is missing
21:25:36HenrikBYou of course also get the message on startup
21:26:24Zagorah, the 5400 series have fpu
21:28:37HenrikBIt is complaing on startup and every time you play a playlist but basics like navigating the disk and playing a single playlist/dirctory works
21:30:29ZagorHenrikB: sorry i'm not following you. what are you trying to do/fix/find?
21:32:30HenrikBI'm trying to give better feedback to users who have forgotten to copy the .rockbox directory on installation
21:32:40Zagorah, ok
21:34:45amiconnHenrikB: What's about our old patch, 848904 (modified boot file detection)?
21:37:20HenrikBIt saves the size and cluster of the bootfile in settings so bootfile detection also works when you power on with the usb cable connected
21:40:09amiconnI know what it does from the description, but it is still open...
21:42:10HenrikBI think there was a feature when I created it and since I've not gotten much feedback I've not commited it
21:42:32HenrikBI know, I should be onth IRC more to get feedback
21:42:53HenrikBs/feature/feature freeze/
21:48:25Zagori don't think i ever really understood the purpose of patch #848904
21:50:54HenrikBIt's only a small improvment on the changed bootfile detection.
21:51:43HenrikBToday if you poeron if the usb cable coneccted and copies a ajbrec file it is not detected as a new bootfile
21:52:04HenrikBs/poeron/power on/
21:53:18Zagori know. i'm just not convinced we need to spend configuration data and code to fix that. the boot check was just a quick convenience hack during development.
21:54:04amiconnIf you aren't going to commit that, you could close it instead. That's what I did with my asm-optimized memcmp(), as memcmp is only used in a few places. If somebody later decides that it would be useful, the patch will still be there
21:54:48HenrikBYeah I know, thats why I have'nt commited it.
21:54:59HenrikBamiconn, I'll do that
22:01:55 Join scott666 [0] (
22:07:03 Quit benschi ("CGI:IRC (Ping timeout)")
22:11:03 Quit silencer_ (Read error: 104 (Connection reset by peer))
22:13:18 Join silencer [0] (
22:31:15HenrikBZagor, amiconn Should I add the the .rockbox check to playlist so we get a "No .rockbox directory, Installation incomplete" message instead of "error accessing playlist control file"
22:31:37Zagorsounds like a good idea
22:34:03amiconnI agree, although the necessity of a .rockbox directory and continued complaints if it is missing could get very annoying on the Ondio
22:35:34 Join zeekoe [0] (
22:36:23Zagorgot the mod player down to 108 kb
22:36:41Zagor(by removing support for lots of obscure formats)
22:37:07 Join Bagder [0] (
22:37:10HenrikBamiconn, Of course, but the hard part is not removing the check, it is "downsizing" the playlist code for Ondio
22:37:21Zagoryo bro
22:41:58Zagorthe bad thing about libmodplug is that it's written in c++...
22:44:52Zagoras usual it's mainly C in classes though, so it won't be much work to sanitize
22:45:57***Saving seen data "./dancer.seen"
22:54:19 Join amiconn_ [0] (
22:58:46 Quit amiconn (Nick collision from services.)
22:58:46 Nick amiconn_ is now known as amiconn (
23:11:00HenrikBBedtime, goodnight.
23:14:03 Quit HenrikB ("Lämnar")
23:16:53*zeekoe is away:
23:32:36*zeekoe is back (gone 00:15:44)
23:45:20 Join _Lucretia_ [0] (
23:45:41_Lucretia_does anyone know if the Xclef MT-500 will be supported at all?
23:45:53_Lucretia_the current firmware is pretty sucky
23:45:58Zagorit's impossible to say
23:46:05Bagderif someone digs up enough info, it might
23:46:09_Lucretia_well, i've emailed them
23:46:13_Lucretia_no response yet
23:46:13Zagorthe hardware is quite similar to the iriver
23:46:18Zagoremailed who?
23:46:33_Lucretia_for specs
23:46:44Zagordon't hold your breath :)
23:47:18_Lucretia_i'd do the work if I could 1) get the specs 2) be guaranteed that I can't fuck up the machine (I use it daily)
23:47:25_Lucretia_I'm not holding my breth
23:47:43_Lucretia_I doubt I'll get it, but would love to see better firmware...oh well
23:47:58_Lucretia_I take it nobody has opened one up and checked the hw?
23:48:00Zagor2) is pretty difficult to guarantee
23:48:15Bagderwe would all cheer and enjoy supporting xclef
23:48:25Zagoryes, it's pretty similar to the iriver. coldfire-based.
23:48:40_Lucretia_if the device is designed such that there is a basic firmware that cannot be overwritten and handles the writing on flash
23:49:00_Lucretia_on = of
23:49:11_Lucretia_then it it *should* be possible to not shag it
23:50:16 Quit edx ("go4ioi")
23:50:16_Lucretia_these are only hdd based devices, the mt-500 is solid state
23:51:00Zagoroh, ok
23:51:23Zagorwe are mostly into hd-based units
23:51:23 Join Dma-Sc [0] (
23:52:20Dma-Scnot much to say, expecting that you guys rule damn big :)
23:52:45Zagor:) thanks
23:53:12Dma-Scyou made my Archos JB Player work like a breeze
23:53:56Dma-Scand now that it died and that i bought a H120, i read that you're working on it, so neat :)
23:54:47Zagorwell, cross your fingers and hope for the best :)
23:55:22_Lucretia_Zagor, no reason not to support the solid state stuff tho ;-)
23:55:39_Lucretia_...but anyways
23:55:46Zagorsidplay looks like it can be difficult to support. in "fast mode" it requires a pentium 166.
23:56:07Dma-Sczagor : eh yep, let's hope it won't use a chip you couldn't identify ;)
23:56:31Dma-Sczagor : is iriver like archos? giving no hw infos?
23:56:32ZagorDma-Sc: we've found all of them except the display controller
23:56:48ZagorDma-Sc: yes, like nearly all companies (except neuros)
23:57:22Zagor_Lucretia_: well our days only have 24 hours ;)
23:57:29Bagderkind of ironic the one single company who publicize their info uses a dead-expensive dev-environment
23:57:54ZagorBagder: yeah. they are not happy about it either.
23:58:03amiconnZagor: SID shouldn't be a problem on coldfire, since an 68060 @50 MHz can do it without much processor load
23:58:20Dma-Scthose guys are so dumb, it save them some work if a great firmware is developped but no they don't give a shit about it
23:58:28Zagoramiconn: yeah, using 68k assembler

Previous day | Next day