Previous day | Jump to hour: 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 | Next day

Seconds: Show Hide | Joins: Show Hide | View raw
Font: Serif Sans-Serif Monospace | Size: Small Medium Large

Click in the nick column to highlight everything a person has said.
The Logo icon identifies that the person is a core developer (has commit access).

Notice: Only Gecko based browsers prior to FF4 support the multipart/mixed "server push" method used by this log reader to auto-update. Since you do not appear to use such a browser, this page will simply show the current log, and not automatically update.

#rockbox log for 2002-08-13

00:05:57adi|worki say the hell with the user
00:06:04adi|workif they are too lazy to read the fucking docs
00:06:07adi|workthe hell with them
00:06:29*adi|work quotes "Ignorance is not an excuse"
00:07:01Zagorwell even if I read the docs and know the limit, it's nice to get a notice when I actually bump into it
00:08:07adi|workbah, humbug
00:08:31adi|workin the same vein, should we have a notice about hitting the 10k file limit?
00:08:36matslBagder: Not sure I understand this code but setting with top 16*11 fixed the problem.
00:08:51Zagoryup, i think so
00:09:04adi|workor how about the 'you are a dumb fuck, unmount the device before pulling the usb cable' notice
00:09:09Bagdermatsl: you mean you modified LCD_WIDTH ?
00:09:10adi|workare we going to code that up too?
00:09:31*adi|work is just in the general fuck the consumer kinda mood
00:09:36Zagornope. we'll just say. "HD Error. Please reformat" :)
00:09:38matslBagder: jupp!
00:09:54Bagdermatsl: well that's the change I propose to commit
00:10:05Bagderwith suitable #ifdef
00:10:53matslBagder: I have the ifdef in my code as well. Time for my first checkin.
00:11:13Bagdergo go go ;-)
00:12:07matslBagder: Another things. Changeing lcd.h didn't trigger a recompile. had to clean and compile.
00:12:10adi|workanyone know where the 200 limit is set offhand?
00:12:31Bagderthe dependencies aren't working properly
00:12:36***Saving seen data "./dancer.seen"
00:12:50Bagderadi|work: its a #define in tree somewhere I think
00:12:54adi|workk
00:13:16matslBagder: Know why?
00:13:35Bagdermatsl: no, I haven't investigated it, only seen it happen myself
00:14:59adi|workwhen we open a directory, is there a way to get a count of the number of subdirectories in it?
00:15:05Zagorno
00:15:08adi|workk
00:15:40Zagori'm adding a notice now
00:15:49*adi|work was just working on that :)_
00:15:52adi|worklol
00:16:28Zagoryeah, i guessed that :)
00:16:41matslBagder: commited!
00:16:47adi|workwell.. if you hit that... ill look at a notice for the 10000 limit
00:16:51Bagderwooo hooo
00:17:03Zagordone that too :) testing now.
00:17:07adi|worklol
00:17:10adi|workyou pain in my ass
00:17:12adi|workdamn you
00:17:13Bagdermatsl: your next commit could be adding yourself to firmware/CREDITS and apps/credits.c! ;-)
00:17:29Zagoryou go take care of your lovely customers :)
00:17:35adi|workbah
00:17:38adi|worki dn't have to deal with customers
00:17:42adi|worki barely have a company ;)
00:18:00matslBagder: Is that the custom? Adding yourself ;-)
00:18:11adi|workmatsl: whats youre real name?
00:18:35Bagdermatsl: *beep* the width should only be non-112 for simulator without HAVE_LCD_BITMAP
00:18:46matslMats Lidell
00:18:59adi|workand you put in code right?>
00:19:37matslBagder: Gosh! I have a bug to fix! ;-)
00:19:45Bagderhehehe
00:19:54Bagderadi|work: he did
00:20:07adi|workk
00:20:24adi|worki was just wondering if he had done other stuff :)
00:20:50matsladi|work: Just filling the mailist with obscure patches!
00:20:58adi|workhehehe
00:21:06adi|workwell.. i just added you to credits
00:21:08adi|workso welcome ;)
00:21:18matsladi|work: Thanks!
00:21:42adi|worknp
00:21:55adi|workgave me something to do since Zagor kept pushing me outta the way ;)
00:23:06Zagoradi|work: if you want to do something, look at adding folder/playlist/song icons on the player. they should use a char prior to the filename to show what type of entry it is
00:23:26Zagorplayer-old has it in the font, player-new needs to use one of the define'able chars
00:24:00adi|worknods
00:24:02adi|worknp
00:24:13adi|workie: F P S
00:24:14adi|work?
00:24:39Zagoryeah, but a symbol like on the recorder
00:24:54adi|worknods
00:25:10adi|workwhere do we have the diagrams on the web?
00:25:20Zagorresearch notes
00:25:33adi|worknm got it..
00:25:38adi|workoh.. hehe got it
00:26:16adi|workZagor when you say old has it in the font, what do you mean?
00:27:07ZagorI mean i'm lying :) ok, both need to define chars
00:27:41Zagorold has a note, but nothing else. i remembered wrong.
00:29:02adi|workummm... okay... so basically pick a char out of the char set on the research page, and reference that, correct?
00:29:15adi|workbecaus with the recorder we had to create the icons ourselves
00:29:35Zagorwe do on the player too. we need to reprogram three chars with icons.
00:29:57adi|workhow do you reprogram chars?
00:30:06adi|worki thought they were fixed?
00:30:11adi|workas in couldn't be changed...
00:30:13Zagorlcd_define_pattern()
00:30:29Zagorthere are a couple of definable chars, I think the first four
00:30:36Zagoror eight
00:30:49Zagoror it differs btwn the versions, can't remember :)
00:30:50adi|workokay
00:30:58adi|workahh. .the blank ones.. got ya
00:31:12Zagorwhen you get this working, you can make them animated! ;)
00:31:21adi|workdont make me hit you ;)
00:31:38Zagoryou'd be a hero! :)
00:31:57adi|workwtf would you want animated icons on the desktop ? :)
00:32:36Zagorbecause we can. is there every another reason?
00:32:42adi|workhehehe
00:34:59adi|workyou said we are already making use of this once somewhere right?
00:35:26Zagorno, but i think we've at least tested it
00:35:33adi|workgrrr...
00:35:50Zagoror you mean the icons on the recorder? they are used in tree.c
00:35:54Bagder... and the simulator doesn't support it... yet
00:36:11Zagori give you the best jobs, adi! :)
00:36:21adi|workno.. what i mean is, i dn't see any refernce to a use of lcd_define_pattern anywhere but where its defined
00:36:28adi|worknot sure what the 'pattern' is
00:36:35adi|workis it one array of the bits
00:36:40adi|workor the full icons itself
00:36:42Zagorah
00:37:02adi|workbecause we pass into lcd_write pattern[i]
00:37:21Zagorit's the full icon, which is an array of bytes
00:37:39adi|workokay.. and the 'length' is really the height of the icon right?
00:37:50Zagori think so, yes
00:37:54adi|workokay..
00:38:01Zagorlength of the array, at least
00:38:06adi|worknod
00:38:09adi|worksame thing
00:38:15adi|workheight*width
00:38:18Zagorshould be, yes :-)
00:39:12adi|workhmmm.. so in this case, lcd_write is drawing pixel by pixel
00:39:20adi|workthat seems rather innefficent
00:40:15adi|workso in theory, we should be able to use the same icons that the recorder uses, no?
00:40:24Zagorwell lcd_write is only used to define the font. after that, you print the char like any other char
00:40:43Zagorif they are the proper size, yes
00:41:15adi|workwhat you just said about lcd_write then doens't make sense on how its used in lcd_define_pattern
00:41:38adi|worknm
00:42:05adi|workhmmmm
00:42:23adi|workicons are located within the recorder subdir, which means the player won't see them...
00:42:24Zagoryou only call lcd_define_pattern once, to define the icon. to draw the icon on-screen, you use lcd_puts()
00:42:34*adi|work nods
00:45:30adi|workso my first call on lcd_define would be like: lcd_define_pattern(LCD_IRAM, "whatever my buffer for the icon", size)
00:45:35adi|workand then to print
00:46:05adi|worklcd_puts(x, y, "whatever my buffer for the icon")
00:46:06Zagornot LCD_IRAM. 'which' is which char to redefine. 0-8
00:46:13adi|workthat don't make sense.
00:46:15adi|workahh.. okay..
00:47:12adi|workso then with the puts, how do i know which 'string' to print?
00:47:28matslLinus didn't like my use-audio-and-param-icons-extensively-patch. Is that something that has been discussed before? How to use these icons on the player?
00:47:45adi|workmatsl where is this patch?
00:47:50*adi|work needs ideas
00:48:02matsladi|work: mailed to the list
00:48:18Zagoradi|work: insert the correct byte, 0 to 8, in a string you send to puts
00:48:48adi|workie: lcd_ascii[0]
00:49:02adi|workcorrect?
00:49:17Bagderas in lcd_puts("\x01moo");
00:49:25Zagorright
00:49:27adi|workk.
00:49:41ZagorBagders model
00:49:50Zagordon't use 0 :)
00:50:08adi|workwhy not?
00:50:18Bagderit won't work ;-)
00:50:29BagderC strings terminate on zero
00:50:35*adi|work cocks his head to the side like a confused dog
00:50:38adi|workright
00:50:43adi|workbut im not printing'0'
00:50:50adi|workim printing the lcd_ascii[0]
00:51:15adi|workfirst location in the stored table no?
00:51:21Bagderno
00:51:36adi|workin other words, i call lcd_define_pattern. Where does the defined pattern get stored?
00:51:38Bagderyou should print the ascii code for the char you redefine
00:51:45Zagorin the lcd controller
00:51:51adi|workahh ... okay
00:51:51Bagderadi|work: you redefine the ASCII code you specify
00:51:57adi|workright...
00:52:21adi|workso then in your example i can just call lcd_ascii[1] then ;)
00:52:43adi|workor, to make life easier: #define PLAYER_FILE_ICON \x01 if you want.
00:53:42BagderI don't know if it is lcd_ascii[letter] or just letter you should use
00:54:04Bagderfor the first four, there's no diff anyway
00:54:16adi|workno, lcd_ascii[0] = \x00 so you were right
00:54:20Zagorlcd_ascii is local to lcd.c so you can't use it anyway.
00:54:29Bagderah
00:54:36Bagderthere you go ;-)
00:54:40adi|workhehe
00:54:41adi|workfine
00:54:44*adi|work pouts
00:55:05matslBagder: Zagor: Comment on the audio-param icons?
00:55:31Zagorcan't remember it. checking the archive...
00:55:54adi|workdo we know the 'size' of the player chars?
00:56:12Zagorthey are exactly as in the picture in "research notes"
00:56:19Zagorbit for bit
00:56:22adi|workk
00:56:28*adi|work goes to count bits
00:56:31Zagorhehe
00:57:07adi|work7 high
00:57:09adi|work5 wide
00:57:13Zagorsounds right
00:57:28adi|workso the recorder icons wont work
00:58:26Zagormatsl: hmm, right you sent me icons for use in the simulator. now where did i put those...
00:59:40adi|workwe have icons already?
00:59:45adi|worklike, ones i don't have to recreate?
00:59:50Bagderno
00:59:59adi|workk
01:00
01:00:02matslZagor: OOPS! I wasn't refering to that at all.
01:00:12Bagdermatsl made some icons we could use in the player sim, that's all
01:00:12Zagormatsl: ehm, ok :)
01:00:33Zagoradi|work: for the USB symbol, play/pause, battery etc
01:00:38adi|workgot ay
01:01:02Bagdermatsl: now *you* can add those icons! ;-)
01:01:09Zagormatsl: can you find it in the archive and point me to it?
01:01:10*Bagder loves handing over work!
01:01:56 Join mistax [0] (lamamba_@213.66.147.135)
01:02:04matslWe have two subjects going here I guess.
01:02:51Zagoryes. ignore me and just show me your posting :)
01:03:11matslZagor: I did som icons for the player. Maybe even easier to mail them again to you!
01:03:21Zagoryup
01:03:33matslZagor: OK. I'll do that.
01:04:02matslZagor: About the audio and params icons is where and whether they shall be shown.
01:04:30matslZagor: I wanted to use them more but Linus disagreed.
01:06:07Zagorcan't find the mail :-(
01:06:34matslWe discussed it over IRC
01:06:40Zagoroh
01:07:07Zagor<Linus> so you think we should light the param icon when in the settings menu?
01:07:42matslZagor: yes and the audio in wps.
01:07:53Zagorok. reading log...
01:08:11matslZagor: You'll find my patch in the list a few days ago.
01:12:07Zagorah, there
01:12:38matslZagor: <icons posted to you!>
01:12:57Zagorso did you and linus discuss this patch after it was posted?
01:13:31matslZagor: Before or after. Don't remember.
01:13:37Zagorok
01:14:22adi|workcan someone check my math on the icons for a record
01:14:26adi|workplayer rather
01:14:31matslZagor: He didn't like the use of the audio icon. Needless he thought.
01:14:36adi|workfigure, dementions are 5x7
01:15:06adi|worksomeone want to tell me what "0x55, 0x00, 0x55, 0x55, 0x55" would look like to them?
01:15:12adi|workand if that makes sense.
01:15:23mistaxhow's it going with the sound quality bug on the players?!
01:15:43adi|workbtw.. on the player.. is pixel distance between chars handle on board or do i have to do that myself?
01:15:48Zagormistax: linus is desperately trying to find someone who can reproduce the problem. do you have it?
01:16:24Zagoradi|work: the character cells on the LCD have a pixel empty between them
01:16:29mistaxyeah, I can play higher with the stock firmware without distortion with more bass!
01:16:43adi|workokay cool
01:17:03mistaxthan with the rockbox...
01:17:48Zagormatsl: i'm not sure I agree with Linus. With just 2x11 chars, i guess sometimes it's not obvious to spot the difference btwn WPS and dir browser
01:18:01Zagorat a glance, i mean
01:18:27Zagormistax: that's not what the bug is about, although I guess he might be interested in that too
01:19:03mistaxok so how shall I notify linus?!
01:19:17mistaxbug report?!
01:19:35Zagorwhere are you located?
01:19:47matslZagor: Jupp! Just what I think. How do we settle these types of things?
01:20:06mistaxnyköping, sweden
01:20:16Zagormatsl: let's discuss it some more tomorrow and then decide
01:20:37Zagormistax: ok, linus will probably be in tomorrow morning. can you be here then?
01:20:51mistaxmorning, what time?!
01:21:04mistaxI just love sleeping..
01:21:10mistax=)
01:21:20matslZagor: ok. I'll sit on my hands then ...
01:21:27Zagorwell he's probably in most of the day
01:21:39adi|workzagor, if i want the icons for the player to be loaded, should they be put in lcd.c or should i put them in a seperate file like they are in recorder?
01:21:46mistaxok
01:22:15Zagoradi|work: they are so small. put them in tree.c for now
01:23:31adi|workk
01:24:24Zagori'm off to bed
01:24:31 Quit Zagor ("Client Exiting")
01:24:32matslcu
01:35:21adi|workumm.. okay.. wtf should i first call lcd_define_pattern?
01:36:31adi|workBagder: you around?
01:36:44Bagderyes, but I have no good suggestion atm
01:37:35BagderI'm off to bed now, night
01:37:38 Part Bagder
01:47:55 Quit matsl ("Client Exiting")
01:48:34 Join matsl [0] (~matsl@as13-4-5.mal.s.bonet.se)
01:48:52 Join matsl_ [0] (~matsl@as13-4-5.mal.s.bonet.se)
01:49:13 Part matsl_ ("Client Exiting")
01:53:06 Part matsl ("Client Exiting")
01:53:49 Join matsl [0] (~matsl@as13-4-5.mal.s.bonet.se)
01:54:17 Quit matsl (Client Quit)
02:00
02:01:06adi|workanyone around that has a player?\
02:01:09adi|worknot a recoder
02:01:53mistaxyeah
02:02:40mistaxwhy?
02:03:33adi|worki may need you to test something for me in a bit...
02:03:36adi|workicons for the player...
02:03:39*adi|work nearly there
02:04:51datazoneme to
02:04:55mistaxI'm right here..
02:05:08adi|workk...
02:05:13adi|workmistax: which do you have
02:05:16adi|workthe old one or new one?
02:05:34datazonei got the new
02:05:37adi|workk
02:06:23adi|workdata you have dcc?
02:06:28datazonesure
02:06:36mistaxold
02:07:05datazonesend
02:07:20adi|workokay.. ill try that too mistax.. one min
02:07:42datazonehmm.. adi, no juice
02:07:53adi|workyou can't grab it?
02:08:20datazonenope
02:08:26adi|workhmm.. okay...
02:08:31adi|workmistax.. let me try with you.. one sec
02:08:35mistaxok
02:09:28adi|workcan you grab that?
02:09:43mistaxit says filetype ignored
02:09:48adi|workhuh?
02:09:49mistaxzip it or something
02:09:49datazonetry again adi
02:09:57adi|workk... ill try ziping.. one sec
02:10:20adi|workoh screw it.. you have email?
02:10:29adi|workactually.. here.. ill web it.
02:10:32datazonecool
02:10:38mistaxno, now I've made it so it doesn't ignore mod files
02:10:42mistaxtry again
02:11:21mistaxWEHEEJ!
02:11:26adi|worktrying
02:11:32adi|workyou don't seem connected
02:11:34mistaxwhat's happening?!
02:11:52mistaxIt says accepting send but nothings happening
02:11:55adi|workpull that down
02:12:29datazoneadi|work: ftp it to ftp.airmail.net in the incoming dir
02:12:38***Saving seen data "./dancer.seen"
02:12:54adi|worktry one more time data
02:13:29datazonenope, no juice, you seem to be firewalled or something
02:13:34datazoneftp it
02:14:12adi|worksent
02:14:23adi|worklet me know if those 1. load
02:14:28adi|work2. see icons in the directory listing
02:14:41datazonedownloading
02:16:38datazoneweird
02:16:55datazonewhar i see is at the right side of the dir names is "AAB->"
02:17:22datazonebut only if it doesnt have to scroll
02:17:26mistaxI can't update my firmware!
02:17:35mistaxI think I'm too low on batteries
02:18:01adi|workhmmm ... i dunno if thats the code or the icon...
02:18:04adi|workhmmmm...
02:18:26adi|worki can't use the sim because it doesn't recognize the lcd_define_pattern from lcd.h
02:18:33datazoneneed to going into the debug menu?
02:18:39datazonedont know if it will help
02:20:22adi|worknope..
02:20:25datazoneadi|work: what exactly are you trying to do?
02:20:33adi|workadd icons to the front of the files
02:20:36adi|workdirectories
02:20:37adi|workand playlists
02:20:46datazoneis the player capable of doing that?
02:22:20adi|workwho has the old?
02:22:39datazonemistax: said he did
02:22:55adi|workmistax grab a new copy if you could
02:23:01adi|workdata, im sending you a new copy too
02:23:17datazoneokay, name it archos-02.mod or something
02:23:25datazoneor do you just want to over write the previous
02:24:01adi|workjust overwritign
02:24:03adi|workno big
02:24:05adi|worksent
02:24:18adi|worki think i was putting the wrong length for the icons
02:25:38mistaxthe address one more time pleez
02:25:53datazoneftp.airmail.net/incoming
02:26:38datazoneadi|work: same thing
02:26:49adi|workokay.. hold on that mistax
02:26:54adi|worki think i have another bug :)
02:26:56datazone"AAB->" at the end of any name that dont need scrolling
02:27:01mistaxdoesn't work(the address I mean)
02:27:17datazonemistax: ftp://ftp.airmail.net/incoming/archos.mod
02:27:39datazoneyou cant view the contents of the dir, you just have to retrieve the file
02:27:44datazoneits a secure drop box
02:30:34datazoneweird, why does it make short dir names uppercase, and long dir names that scroll lowercase?
02:30:41datazonethe 1.2 firmware anyway
02:32:48adi|workdata
02:32:51adi|worktry archos.mod.2
02:32:55adi|workand mistax can't use that
02:32:59adi|workif he's on an old player
02:33:17mistaxI guess I'll have to wait till the tomorrow cuz I got to little batteries left to boot up withthe new firmware!
02:33:35adi|worknp..
02:33:42adi|worki wont have this done tonight
02:34:10mistaxwait a minute, the ftp address is that for the old or the new!?
02:34:15mistaxftp.airmail.net/incoming
02:34:23adi|worknew
02:34:25adi|workyou can't use that
02:34:27adi|workdon't worry..
02:34:33adi|workill work on the old one tomorrow
02:34:55mistaxbut the http address u gave me is that for old or new!?
02:35:00adi|workold
02:35:06datazoneokay, the "AAB->" crap is gone, but its back to normal
02:35:14datazoneits not http
02:35:17mistaxok I'll try that one last time
02:35:25adi|workmistax don't
02:35:29mistaxhttp://phobos.ramapo.edu/~rhak/misc/archos.mod
02:35:30adi|workits not a correct version
02:35:32mistaxwhy?!
02:35:37mistaxok
02:35:38adi|workif its buggy on data's its buggy on yours
02:35:40adi|workits the code
02:36:09datazoneadi|work: do you know if this will even work?
02:36:48datazoneheh, someone need to tell google that "rockbox" is a valid word, and that it shouldnt suggest "rock box" instead
02:38:39adi|worklol
02:38:47adi|workfeel like tryong one last version?
02:38:56datazonesure, i'm all yours tonight
02:39:07adi|workhmm wait one sec
02:44:36adi|workokay.. try archos.mod i just put up data
02:47:06datazoneone sec, in the middle of a ca game
02:47:50adi|workk
02:54:45datazoneokay, let me get it
02:55:11datazonecrack-attack is one addictive game
02:56:27datazonethe "AAB->" stuff is back again
02:57:29datazonehold on
02:57:35datazonei just noticed something
02:57:40datazonegod damn
02:59:24datazoneokay, there is just a blank space in front of the name
02:59:39datazonewget was not over writing the old file
03:00
03:00:24datazoneyou should put a dev number in the mod, so i can make sure i am using the right version
03:09:34adi|workk..
03:09:34adi|workty
03:09:41adi|workgonna work on it more tomorrow..
03:09:46adi|workthanks for being my test whore ;)
03:09:49datazoneokay... sorry about that
03:16:49adi|worknp
03:51:08 Part mistax
04:00
04:12:40***Saving seen data "./dancer.seen"
04:18:51 Join adiamas [0] (~adiamas@216.194.23.29)
04:19:57 Quit adiamas (Client Quit)
04:20:00 Join adiamas [0] (~adiamas@as5300-9.216-194-23-29.nyc.ny.metconnect.net)
04:39:39 Join MeRWiN [0] (~Stalker@12.242.185.10)
04:42:03 Nick adiamas is now known as adi|home (~adiamas@as5300-9.216-194-23-29.nyc.ny.metconnect.net)
04:43:37MeRWiNI wonder when we're letting in the SAVE_TO_DISK feature for good...
05:00
05:18:30 Join elinenbe [0] (elinenbe@pcp02254422pcs.wanarb01.mi.comcast.net)
05:30:38MeRWiNhey elinenbe
05:33:31elinenbehi there.
05:46:18 Join gopher2 [0] (~DAVE@pcp01188022pcs.strl401.mi.comcast.net)
05:46:31 Part gopher2
05:51:44 Join PsycoXuul [0] (psyco@63.205.43.200)
05:51:44 Quit PsycoXul (Read error: 104 (Connection reset by peer))
05:51:59 Nick PsycoXuul is now known as PsycoXul (psyco@63.205.43.200)
06:00
06:03:12 Part elinenbe
06:06:02MeRWiNgrr...
06:12:43***Saving seen data "./dancer.seen"
06:17:56MeRWiNcan't get lcd_puts_scroll() to work right
06:28:37 Join PsycoXuul [0] (psyco@adsl-63-205-44-6.dsl.lsan03.pacbell.net)
06:37:54 Quit MeRWiN (Read error: 104 (Connection reset by peer))
06:45:19 Join elinenbe [0] (trilluser@pcp02254422pcs.wanarb01.mi.comcast.net)
06:45:56HesYou have any idea where do I think they should stick that &%$#@ fundraiser?
06:46:15HesGood morning.
06:48:57 Quit PsycoXul (Read error: 110 (Connection timed out))
06:49:04 Nick PsycoXuul is now known as PsycoXul (psyco@adsl-63-205-44-6.dsl.lsan03.pacbell.net)
06:50:25 Join bruce_ofSherwood [0] (~bruces@12-229-49-19.client.attbi.com)
06:50:34 Join MeRWiN [0] (~Stalker@12.242.185.10)
07:00
07:05:12 Quit MeRWiN (Read error: 104 (Connection reset by peer))
07:10:05adi|homeHes: wanna say that in english again?
07:11:04 Quit bruce_ofSherwood ()
07:11:08 Join MeRWiN [0] (~Stalker@12.242.185.10)
07:11:20 Join bruce_ofSherwood [0] (~bruces@12-229-49-19.client.attbi.com)
07:19:57 Join LinusN [0] (~linus@labb.contactor.se)
07:20:02MeRWiNheya linus
07:20:03LinusNyo
07:20:22LinusNgawd, am i tired of that 200-file limit discussion...
07:20:39LinusNbruce_ofSherwood: are you awake?
07:20:42MeRWiNI've got a player mpeg_info function that displays the file informatino in a seperate menu ... basically all of the id3 info that we've got
07:20:46bruce_ofSherwoodya..
07:21:03LinusNdo you still have the distortion?
07:21:14MeRWiNLinusN: ran into a problem where my computer would blue screen when i plugged in USB... i forgot to include a case for the USB detection :-)
07:21:33bruce_ofSherwooddunno.. is there a new release I should play with (outside of official release version)?
07:21:55LinusNnot that tries to fix any distortion problems...
07:22:08elinenbeI liked the 200 file limit discussion :)
07:22:22LinusNbruce_ofSherwood: 1 question: does the distortion disappear when you raise the treble?
07:22:24MeRWiNI hate that discussion
07:22:32bruce_ofSherwoodumm.. Then I guess I stil have same issue.. I'll try to re-dl current release and see what occurs.
07:22:32MeRWiNI like the response of quit whining and write some code
07:22:33HesMorning
07:22:39MeRWiNmorning hes
07:22:43LinusNMeRWiN: yeah
07:22:46elinenbeThat was a response of frusteration
07:22:52LinusNmorning Hes
07:23:23elinenbeI know it is time to head to sleep if LinusN is up!
07:23:37LinusN:-)
07:25:10LinusNit's frustrating, because we need every byte for MP3 buffering, and then someone wants to waste it on large dir buffers
07:25:41MeRWiNYeah... I wonder how archos did it
07:25:55LinusNwe can, actually, triple the amount of cached names by just changing the storage algorithm for the file names
07:26:23LinusNMeRWiN: i think Archos has a smaller MP3 buffer than we do
07:26:31HesBtw, I think the archos charger code is pretty simple
07:26:48LinusNi think so too, they probably stop when a certain voltage is reached
07:26:53MeRWiNLinusN: Is it alright for me to bind my player_mpeg_info() function to MENU+ON? It works fine... seems to be the only remaining key combo left. It's a nifty screen to have.
07:27:10HesI charged with rockbox yesterday, reached 5.52v (some 5.4v after it had cooled down)
07:27:34Hesthen charged some more with archos - it continued charging for hours
07:27:36LinusNMeRWiN: so MENU+ON switches to a file_info screen?
07:27:45Hesand after that the end voltage wasn't any higher (the batteries were full)
07:27:55HesI guess they go to trickle charge mode and end it with a timer
07:28:18LinusNHes: negative delta detection will probably fail if the battery is nearly full
07:28:18HesI can clearly hear the device make the sound of some xx kHz
07:28:37LinusNwhen you start charging
07:28:41HesLinusN: true... the low delta detection works at that point
07:28:48adi|homewhat i don't get is.. WTF would anyone have more then 200 files in a directory?
07:28:53adi|homeHTF do you find anything
07:29:08HesMaybe he doesn't want to find anything, and is lazy to sort stuff
07:29:09LinusNmaybe it's his "downloaded from the net" folder?
07:29:15adi|homelol
07:29:18Hesjust downloads loads of stuff from the net and plays it with dir shuffle
07:29:29MeRWiNLinusN: yup.. displays all of the ID3 info for the player... left and right scroll through (the scrolling wraps around also). any other key exits it. The only thing it doesn't have is the lcd_puts_scroll() of the info... I couldn't figure out how to get the damn thing to actually scroll. Since it's part of the WPS, I have to always yield() so the buffer can fill itself up, etc. That also causes the normal wps scrolling to start again i think.
07:29:30adi|homewell.. zag added a patch to inform ppl when they hit the limit
07:29:33LinusNPeople are people, never the same
07:29:41adi|homelinus...
07:29:42HesAt least two of my friends have a folder like that on their home computers
07:29:55adi|homei started working on icons for the player, zag asked me to
07:30:07adi|homei compile for the actual player w/no prob
07:30:10LinusNadi|home: great
07:30:14adi|homebut the sim compile don't work
07:30:19adi|homeit bitches about the lcd.h
07:30:25LinusNadi|home: surprised?
07:30:32adi|homebasically.. the sim still runs off the lcd-x no?
07:30:40LinusNno
07:30:54LinusNonly for the text functions
07:30:55adi|homeumm... then i dont getit
07:30:59LinusNand the redrawing
07:31:04LinusNwhat is the error?
07:31:10adi|homeno sec.. ill get it
07:34:25bruce_ofSherwoodquestion about 1.2 player release..
07:34:39adi|homeshoot
07:35:07bruce_ofSherwoodis there a differnt screen on boot for it then shown on website?
07:35:20LinusNon the player, yes
07:35:23bruce_ofSherwoodha..
07:35:25adi|homenods
07:35:30adi|homethe screen on the web is the recorder
07:35:37LinusNthe players with newer LCD has a double-height "ROCKBOX"
07:35:46bruce_ofSherwoodokay. Well. At full charge there is no popping.
07:35:49*adi|home shuts his mouth as he missreads
07:36:14MeRWiNLinusN: Why is it that we don't include the SAVE_TO_DISK by default? It seems as if it's a stable function to include.
07:36:14bruce_ofSherwoodfirst install was at lower voltage (showed 1 bar on bat meter)
07:36:28LinusNMeRWiN: good question
07:36:57LinusNbruce_ofSherwood: I am not sure that it has anything to do with battery status...
07:37:49bruce_ofSherwoodgood point. so why would one copy of the binary allow poooing while another (same version doesnt)? (and other amizing secrets of the universe)
07:37:50adi|homei don't know what the deal is, but the manual page won't display in my netscape
07:38:02LinusNoh?
07:38:12adi|homenods
07:38:23LinusNbruce_ofSherwood: that is the problem
07:38:24adi|homeim using 4.76
07:38:28adi|homeeverything else loads fine
07:38:32LinusNthe noise comes and goes
07:39:05LinusNadi|home: ask Bagder when he wakes up
07:39:18adi|homeLinusN: how much of this error do you want me to spam?
07:39:42adi|homelet me try one more rebuild.. hold on
07:39:53bruce_ofSherwoodHmm.. I'll keep an eye on it.. I am using same reference music. If it turns back up I'll try to report what little I can find out.
07:40:13elinenbegood night....
07:40:20 Nick elinenbe is now known as elinenbe|sleep (trilluser@pcp02254422pcs.wanarb01.mi.comcast.net)
07:40:38LinusNbruce_ofSherwood: one user has noticed that the error comes back if you start the unit with no earplugs inserted
07:40:44LinusNelinenbe|sleep: night
07:40:57bruce_ofSherwoodhmm.. I'l try that now.
07:41:41 Join Bagder [0] (~daniel@as3-3-2.ras.s.bonet.se)
07:41:47MeRWiNyo bag
07:41:47bruce_ofSherwoodbingo
07:41:47Bagdermoo
07:41:47adi|homeokay.. LinusN come to #adi-spam
07:42:04adi|homedon't want to spam the room
07:42:16adi|homebagder.. any ideas why my version of netscape don't load the manual page?
07:42:47Bagderthe front manual page?
07:43:27Bagderyes I see why
07:44:03Bagderfix committed
07:45:02adi|homety
07:45:04adi|homewhat was it?
07:45:26Bagderno foot.t included in the bottom and thus the <table> ended up without a </table>
07:46:22HesI wish I had an oscilloscope around
07:46:39HesAnd a voltage meter connected to a computer so I could graph voltage variations over 12 hours or so
07:46:52adi|homeits going to take a while for that change to take.. grrrr
07:47:22Bagderadi|home: another 15 mins I think
07:47:33bruce_ofSherwoodLinusN: Turn on with no speakers/hdphones - big messy noise. Turn off - add speakers/hdphones, no change. Wonder if I have to let it sit for a few hours.
07:47:38 Join MeRWiN- [0] (~merwin@12.242.185.10)
07:47:45 Quit MeRWiN ()
07:47:47 Nick MeRWiN- is now known as MeRWiN (~merwin@12.242.185.10)
07:48:06LinusNbruce_ofSherwood: maybe, but now you have it, right?
07:48:15adi|homenods
07:48:19bruce_ofSherwoodbig old scratchy time...
07:48:57LinusNok, does it disappear when you raise the treble
07:49:34bruce_ofSherwoodHmm.. starts out a +7 - I'll try both directions
07:50:46bruce_ofSherwoodyes.. at +13 almost gone at +14 is gone
07:51:01LinusNok, you have *the* distortion
07:51:20LinusNnow, what is the ROM firmware version?
07:51:33bruce_ofSherwoodoh.. *the* and not 'another'? :)
07:51:56bruce_ofSherwoodfirmware: 6.05
07:52:00LinusNhehe, some people complain about distortion at max volume
07:52:10LinusNbruce_ofSherwood: you have a pretty new firmware
07:52:19bruce_ofSherwoodnah.. thats just their ear drums bleeding
07:52:44bruce_ofSherwoodunit is about 8 months old
07:52:53LinusNmine has 3.18
07:53:19bruce_ofSherwoodDinosaur skin pouch instead of the neoprene?
07:53:39LinusNyeah, and a wooden stick instead of a remote control
07:54:01bruce_ofSherwoodI'm waiting for the fiber optic headphones.
07:55:03bruce_ofSherwoodany idea on fixes? If not, when will 'save state' work?
07:56:10LinusNbruce_ofSherwood: try this: http://linus.haxx.se/archos.mod
07:56:51*Hes starts his day at work by installing a brand new build off the CVS
07:56:53LinusNi'll have to go to the dentist in a few minutes, and will be back in an hour or two
07:57:16HesI have a new routine added at the start of the day, just after getting a cup of coffee 8-)
07:57:27Bagder:-)
07:57:36*Bagder does that routine before leaving home
07:57:46bruce_ofSherwoodi'll try it. I'll mail you a result. Have fun.
07:57:47Hescheckout, apply pending patches, build and install
07:59:07LinusNoh yeah, i'll have fun for sure!
07:59:09adi|homegod damn lcd.c is a mess
07:59:20LinusNadi|home: yes indeed
07:59:56adi|homewhat was the purpose of the SIMULATOR define in the lcd.c?
08:00
08:00:05LinusNto have special code for the simulator
08:00:12adi|homecan i assume, that which is to be defined in lcd-x11.c is to be defined around in lcd.c?
08:00:16Bagderadi|home: we use the recorder lcd code for the player sim
08:00:34LinusNsilly enough
08:00:35Bagderwhich makes everything a bit messy
08:00:44BagderI don't think it is silly
08:00:46adi|homeright, but the ONLY function defined in lcd-x11.c is lcd_update
08:00:53LinusNgotta go to the dentist. cu!
08:00:53adi|homeso WTF are we doing?
08:01:07 Nick LinusN is now known as LinusN|inpain (~linus@labb.contactor.se)
08:01:08adi|homecause i cna't run a compile on a sim the way it is now
08:01:12Bagderadi|home: we use the "real" functions for the simulator too
08:01:16adi|homenods
08:01:18adi|homeright
08:01:41adi|homewhat im saying is that the only function that should be '#ifndef SIM' in lcd.c should be lcd_update() no?
08:01:50adi|homesince that is the only function defined in lcd-x11.c?
08:01:59Bagderit isn't that simple
08:02:06*adi|home screams
08:02:18Bagderfirst, you can build the sim for recorder OR player
08:02:29adi|homewell, then we have a mighty nice cluster fuck in there :)
08:02:29Bagderfor recorder, the update is basicly the one that should be replaced
08:02:45adi|homebecause to test the player icons i need to use lcd_define_patter
08:02:50adi|homeand thats not recognised
08:02:59adi|homewhich means i also need lcd_write
08:03:00adi|homeguess what :)
08:03:13Bagderright, lcd_define_pattern() must be added to the sim
08:03:26Bagderthe player sim
08:03:46Bagderit would store the pattern in a bitmap
08:03:53Bagdersomehow
08:04:29adi|homeum.. normally i would say sure
08:04:38adi|homebut the ONLY thing defined in the sim lcd is lcd_update
08:04:51Bagderso?
08:04:52adi|homeyour telling me EVERY other lcd function defined in lcd.c accounts for the sim?
08:05:01Bagdernow
08:05:04Bagderno
08:05:21*adi|home doenst' understand how the sim works at all then.
08:05:22Bagderall other functions are in ldc.c because in one way or another they are meant for target
08:05:29adi|homeright i know that
08:05:43Bagderthen what is the problem?
08:05:52*adi|home screams
08:06:01Bagderhehe
08:06:01adi|homedo we have a sepearte lcd_puts for the sim?
08:06:04adi|homehow about lcd_write
08:06:13 Quit bruce_ofSherwood ()
08:06:14Bagderlcd_write() is not used on the sim
08:06:18adi|homehow does the sim work with ONLY the lcd_update defiend for itself
08:06:25Bagderlcd_write() talks to hardware
08:06:38Bagderyou need to understand how the lcd functions work
08:06:45Bagderthe recorder ones
08:06:53Bagderthey all only modify an array
08:07:04Bagderwhatever you do, they update a "frame buffer"
08:07:16*adi|home nods
08:07:17Bagderlcd_update() sends that buffer to the hardware lcd
08:07:24adi|homeright.. that i know
08:07:28Bagderso for the sim, all we need to do is replace that function
08:07:58adi|homeokay..
08:07:59adi|homeso.. let me stop you
08:08:05Bagdercertainly
08:08:06adi|homelcd_write talks to hardware right?
08:08:13adi|homeand the sim doesn't call lcd_write
08:08:21adi|homebut the sim does call lcd_puts right?
08:09:18Bagderthe sim calls?
08:09:32Bagderthe code in apps uses lcd_puts(), yes
08:09:37adi|homeright
08:09:39adi|homeokay...
08:09:40adi|homeSO
08:09:45adi|homelets assume we compile the sim
08:09:52adi|homeand we need to print a playlist to the screen
08:10:03adi|homeat some point 'play_list' gets called
08:10:07adi|homewhich calls lcd_puts
08:10:10adi|homewhich calls lcd_write
08:10:13adi|homeso what am i missing?
08:10:23Bagderlcd_puts() does not call lcd_write
08:10:33Bagderon recorders it never does
08:10:35adi|homevoid lcd_puts(int x, int y, unsigned char *string)
08:10:37adi|home{
08:10:37adi|home lcd_write(true,LCD_CURSOR(x,y));
08:10:37adi|home while (*string && x++<11)
08:10:37DBUGEnqueued KICK adi|home
08:10:37adi|home lcd_write(false,lcd_ascii[*(unsigned char*)string++]);
08:10:37adi|home}
08:10:41Bagderon players, it doesn't while sim
08:11:11Bagderyes, that's because when built as a sim, the player version uses the recorder's lcd_puts()
08:11:22adi|homek
08:11:27adi|homethat makes more sense
08:11:30Bagder... which only modifies the frame buffer
08:12:19adi|homeso then lcd_define_pattern needs to act in the same manner as lcd_puts?
08:12:22adi|homewith the sim
08:12:42Bagderyes, more or less
08:12:44***Saving seen data "./dancer.seen"
08:13:08adi|homei hope you can see why this is so confusing :)
08:13:30BagderI'm not saying it is crystal clear ;-)
08:13:53adi|homeokay, so then whats the point of the check for the define of defined(HAVE_LCD_CHARCELLS)
08:13:55adi|home?
08:14:12Bagderthat's a player-version LCD
08:14:27Bagderwe check for features rather than models
08:14:33adi|homek
08:15:40adi|homeso in the sim.. htf is the define_pattern supposed to work?
08:16:03Bagderit's a bit tricky
08:16:13Bagderwe would need to store the pattern as a bitmap
08:16:25Bagderand then somehow output that bitmap instead of the given ascii letter
08:16:29adi|homeim going to stab someone :)
08:16:43MeRWiNwhat's the syntax in snprintf for using a string value (ie: %d is for decimal)
08:16:50adi|home%s
08:16:55MeRWiNokey
08:16:57MeRWiNthanx
08:17:01adi|homenp
08:17:05Bagderadi|home: for starters, just add the function to the sim and make it do nothing
08:17:19adi|homewell.. we _could_ cheat bagder
08:17:28Bagderlike how?
08:17:47adi|homeyou figure, take the char buf that we get as part of the 'player'
08:17:56adi|homeand force it into the size of 'recorder' icon
08:18:06adi|homethen display it the same way we do a recorder icon
08:18:16adi|homeseeing as a player one is 5x7
08:18:21Bagderyes, but there's only one *little* problem ;-)
08:18:21adi|homeand recorder is 6x8
08:18:25adi|homeif i remember correct
08:18:31adi|homeoh god what?
08:18:33Bagderhehe
08:18:44BagderI enlarged the font used in the player sim the other day
08:19:00adi|homeoh, and like i care?
08:19:02CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
08:19:02*adi|home smirks
08:19:11adi|homei asked earlier.. how large should the icons be
08:19:25adi|homeand i got 'the same size as on the website'
08:19:28adi|homenow your telling me i need to increase them?
08:19:30Bagderadi|home: yes, for target
08:19:36BagderI know how to solve this
08:19:42adi|homeyeah
08:19:55adi|homei bend you over a barrel and shove a cattle prod up your bum
08:20:01adi|homebut i don't see how that fixes the code ;0
08:20:45BagderI think we should just convert the pattern to a bitmap and ignore the size differences
08:21:13Bagderthen we add simulator code in the lcd_puts() that switch()es on the defined_patterns() and uses the set bitmaps if in use
08:21:56BagderI could volounteer to do that part
08:22:28adi|homeumm.. okay.. let me translate from 'intelligent geek' to 'tired geek in pain'
08:22:36adi|homemy icons => bitmap
08:22:55Bagderthat's what lcd_define_pattern() would do in the sim code, yes
08:23:04adi|homebut if you have lcd_puts call the defined_patterns, then defined_patter gets called everytime
08:23:18Bagderno no
08:23:34Bagderlcd_define_pattern() gets a separate implementation in the sim
08:23:45Bagderit converts from a pattern to a bitmap, and remembers that
08:23:46adi|homeokay
08:23:58adi|homeokay.. ala: buffer in global space?
08:24:05Bagderyes
08:24:08adi|homek
08:24:46adi|homeso the new lcd_define
08:25:00adi|homewould that get put in lcd.c with the defines around it correcT?
08:25:25Bagderwell
08:25:34Bagderto make it nicer, it would be in the uisimulator area
08:25:38Bagderbut...
08:25:47Bagderother code already is inserted like that in lcd.c
08:27:59adi|homewell.. wouldn't it make sense to pull out some of that code while were in here?
08:28:54Bagderit would, but that code is used for both x11 and win32 versions so it should probably go into a common place used by either sim-version
08:29:06BagderI mean, preferably
08:31:25adi|homenods
08:31:31adi|homewell.. we do have a common dir
08:31:43Bagderyes, it would be neat to add a file there for it
08:31:59adi|homewell.. let me see if i can do it with the base sim
08:32:01adi|homeuntouched
08:32:02adi|homeand we can expand
08:34:45adi|homedo we have #defines for the diff arch sims?
08:34:50adi|homeie: x11 and win?
08:35:21Bagderwe do, but I've tried to keep them away from code in firmware/ and apps/
08:35:29adi|homeit woulnd't be there
08:35:30Bagderthey still show up though in places
08:35:31adi|homeits for the sim
08:35:39Bagderuse WIN32 for windows
08:35:39adi|homeand i figure having a lcd.*
08:35:41adi|homein common
08:35:43adi|homeit would need it
08:35:44Bagderok
08:36:10adi|homeshould i call it lcd-common.c?
08:36:13adi|homethat makes sense no?
08:36:19adi|homeor just lcd.c?
08:36:19Bagdermakes sense
08:36:33BagderI think lcd-common.c is better
08:36:42*adi|home agrees
08:36:47BagderI like having different names for my emacs buffers ;-)
08:37:22adi|homelol
08:37:38adi|homeso win is WIN#@
08:37:43adi|homeWIN32 even
08:37:48adi|homeand x11 would be #else
08:37:55Bagderyes
08:40:09adi|homehow do i open 3 horizontal windows in emacs?
08:40:18adi|homeall equal size
08:40:31adi|homectrl-x-3 does them verticle
08:40:49Bagderctrl-x-2
08:41:01adi|homeyeah.. thats just 2
08:41:10adi|homewas trying to see if i could do it in one step :)
08:41:13Bagderyes, then again, but they won't be equal size then
08:41:23*adi|home nods
08:41:24adi|homeill deal
08:42:56adi|homethe lcd-x11 should includ the lcd.h as well no?
08:43:20Bagderlcd-x11.h does that
08:43:24adi|homenods
08:50:00adi|homei forgot just how much i hated windows code
08:50:18Bagderoh yes
08:53:52adi|homedo me a favor and make sure im not nuts
08:54:03adi|homecheck and see if the lcd-win32.c and lcd-win32.h are the same file?
08:54:43Bagderno, they're not
08:54:50Bagderyou *are* nuts ;-)
08:54:52*adi|home nods
08:55:04adi|homethats what i thought
08:55:22*adi|home now returns you back to the regularly scheduled voices in your head
08:56:18MeRWiNhow big is the id3->path again?
08:56:56BagderMAX_PATH
08:57:05MeRWiNhow big is MAX_PATH? :-)
08:57:08Bagder260
08:57:10MeRWiNahh
08:57:10MeRWiNk
08:57:34MeRWiNalmost finished with my player wps id3 info screen
09:00
09:00:04*Bagder likes the new charger code
09:03:20HesBagder: does it work better for you since yesterday's changes?
09:03:27Bagderoh yes
09:03:30Bagdermuch better
09:03:37HesExcellent.
09:03:52*Bagder pats Hes on the shoulder, *good work*
09:04:33HesUwe's last suggestion was good, I'm just trying it out (with a small modification, just take two samples at both ends of the range so that we don't get much smaller deltas)
09:07:35MeRWiNTime to submit...
09:10:52MeRWiNanyone got a player and want to test out my new cvs update? :-)
09:15:00Bagdernope
09:17:01MeRWiNdurn
09:18:51MeRWiNFinally, the player can view ID3 :-)
09:19:19Bagderthat is indeed a neat addition
09:22:24 Join matsl [0] (~matsl@as13-4-5.mal.s.bonet.se)
09:28:27 Part LinusN|inpain
09:35:03matslBagder: Is Linus sick?
09:35:11Bagderhe's at the dentist
09:35:37matslwill he come back during the day?
09:35:46Bagderyes, he said he will
09:36:30matslOk. I was hoping for a audio-params-icon discussion with him an Zagor.
09:36:46Bagdersounds like a good ieda
09:39:11MeRWiNis the charset on the player changable?
09:43:50BagderMeRWiN: only N letters
09:43:56Bagderwhere N is 8 at the most
09:46:02MeRWiNahh
09:46:14MeRWiNis it dynamically changeable, or just at load?
09:46:22Bagderdynamicly
09:46:27Bagderlcd_define_pattern()
09:47:00MeRWiNhmm... not fun :)
09:49:13BagderMeRWiN: checked the cvs compile status table... ?
09:49:19Bagderred all over
09:49:50MeRWiNuhh
09:50:12 Join LinusN [0] (~linus@labb.contactor.se)
09:50:21MeRWiNoh, soryr :)
09:50:23MeRWiNeasy fix
09:50:42LinusNbachk flromm the dhentischt...
09:50:46matslLinusN: back again without pain?
09:50:49BagderB)
09:50:59LinusNlucky this isn't a voice chat
09:51:05LinusNpulled a tooth
09:51:15adi|homelol
09:51:20matslnice ...
09:51:26adi|homewould you have to tawk with a lithp?
09:51:43LinusNi'd prefer not to talk at all...
09:52:03*adi|home hands LinusN a role of duct tape
09:52:06LinusNbut if i did, blood would come out of my numb mouth
09:52:24MeRWiNwhere in a function can I declare a variable?
09:52:29matslLinusN: I would like to discuss the audio-param-icon thing when Zagor joins. (I got some support from him yesterday ;-)
09:52:30*adi|home holds up a spiddle cup
09:52:37adi|homeMeRWiN: at the top
09:52:48adi|homeie: int foo(){ int i=0; i++; }
09:53:15MeRWiNhmm.... weird. compiled fine for the player, but not sims
09:53:28MeRWiNor the rec
09:53:46Bagderthey're built with different compilers
09:53:51adi|homedoes it have lcd_define_pattern in it?
09:53:57MeRWiNahh
09:54:06adi|homeim working on fixing the lcd issues
09:54:34adi|homeLinusN: Bagder and i agreed that the sim lcd's are close enough that they should be merged.
09:54:39adi|homeim working on that to begin with.
09:54:41adi|homethe base version
09:54:54adi|homethen add lcd_define_pattern etc. after
09:56:38 Join pyvasene [0] (~pyvasene@ns1.alcove-solutions.com)
09:59:08LinusNmatsl: haven't you given up on that param icon yet? :-)
09:59:17adi|homelol
09:59:35matslLinusN: No and I have checkin access now ;-)
09:59:43adi|homethats okay
09:59:48adi|homewe have deletion access ;)
09:59:55*Bagder looks forward to a commit war ;-)
10:00
10:00:06adi|homelol
10:00:13adi|homei can delete faster then you can commit ;)
10:00:25MeRWiNhmm...
10:00:34*MeRWiN is tracking down the recorder bugs he created :P
10:00:39MeRWiNprobably really easy fixes
10:00:43MeRWiNgot one of them already
10:00:51LinusNi always break the sim builds
10:00:58MeRWiNI broke the recorder :)
10:01:11matsli don't want a war but somehow these YMMW issues must find a solution.
10:01:19Bagderyes
10:01:23adi|homeYMMW?
10:01:44Mode"#rockbox +o LinusN " by ChanServ (ChanServ@services.)
10:02:16Mode"#rockbox +o Bagder " by ChanServ (ChanServ@services.)
10:02:51Mode"#rockbox +o adi|home " by ChanServ (ChanServ@services.)
10:03:01Mode"#rockbox -oo LinusN Bagder " by adi|home (~adiamas@as5300-9.216-194-23-29.nyc.ny.metconnect.net)
10:03:06*adi|home cackles evilly
10:03:07LinusNmatsl: i don't really have a problem with the audio/param icons
10:03:11Mode"#rockbox +oo LinusN Bagder " by adi|home (~adiamas@as5300-9.216-194-23-29.nyc.ny.metconnect.net)
10:03:13*Bagder slaps adi
10:03:19*adi|home asks for more
10:03:27*Bagder slaps adi again and smiles
10:03:32adi|homeyour asking for that barrel son ;)
10:03:32LinusNi just think they are silly in the first place
10:03:42adi|homethat fits us don't it?
10:03:56LinusN:-)
10:04:01matslLinusN: OK. lets w8 for zagor then.
10:04:05MeRWiNwhy is it telling me that BUTTON_STOP is undeclared? I know for a fact that it is :)
10:04:05adi|homei mean, can you name another project where the members list has a responsibility for one of the users being Sarcasm?
10:04:06adi|homei think not
10:06:14LinusNMeRWiN: The recorder has no STOP key
10:06:22MeRWiNoh
10:06:25MeRWiNi need an ifndef :)
10:06:56LinusNyes, #ifdef #HAVE_LCD_CHARCELLS
10:07:13Bagderno its a keypad check
10:07:20 Join bobTHC [0] (~bobTHC@AMarseille-206-2-1-9.abo.wanadoo.fr)
10:07:26LinusNBagder: true
10:07:30MeRWiNyeah
10:07:31MeRWiN#ifdef HAVE_PLAYER_KEYPAD
10:07:33bobTHChi all!
10:07:35Bagdercorrect
10:07:36LinusN#ifdef HAVE_PLAYER_KEYPAD
10:07:37Bagderhey bobTHC
10:10:29bobTHCSorry if I submit too much feature request on sourceforge....
10:10:56adi|homebob don't ever appolgize for that :)
10:10:57BagderbobTHC: no worries, we like feature requests
10:11:21MeRWiNack
10:11:53MeRWiNgreat, now it compiles on the recorder, but not the player :)
10:12:45***Saving seen data "./dancer.seen"
10:13:28bobTHCI want to kwow how and when i can start the translation in french ?
10:13:55bobTHCand why not in spanish...
10:14:00BagderbobTHC: we want to know how too, as we haven't decided on that yet
10:14:20bobTHCok..
10:14:32LinusNwe need a very good system for string handling
10:14:56LinusNmaybe based on a spread sheet with the languages in columns
10:15:15Bagderthere's no good spread sheet standard
10:15:15LinusNand a script that creates .c-files with the string data
10:15:28LinusNBagder: no, give me a better tool
10:15:41BagderI propose a weirdo text-format
10:16:04Bagderand then a reformat tool just as you say
10:16:08LinusNBagder: a weirdo text format is not nice to work with
10:16:17Bagderbetter than a silly spread sheet
10:16:24Bagderand it can be patched
10:16:26LinusNspread sheets are not silly
10:16:29Bagderyes
10:16:33LinusNth100 file formates are
10:16:35Bagderthey're not patachable
10:16:46Bagderth100 ?
10:16:54LinusNit is not the spread sheet, it's the file format
10:17:02MeRWiNwhat's a good way to get rid of an "Unused variable" problem ... the variable is used only in the player, so the recorder complains about it being unused.
10:17:06LinusNspread sheets are wonderful tools
10:17:22LinusNMeRWiN: use it, or #ifdef it
10:17:26Bagderso don't say spread sheet, state a file format ;-)
10:17:26MeRWiNk
10:18:59adi|homecan someone look at lcd-x11.c and tell me why the code in lcd_update that is #if 0'd is still there?
10:19:22LinusNit is there because nobody removed it
10:19:55BagderI left it there when I added the new code just to be able to revert back easier
10:20:00adi|homeum... define 'spread sheet format'?
10:20:03Bagderbut since the new code proved working, it could be removed
10:20:20adi|homeif we are talking cols here, why not comma or colon delineate?
10:20:27adi|homenearly everything ive seen yet imports it
10:20:36adi|homenod
10:20:42Bagderbecause that is crap when you add columns
10:20:50bobTHCSomebody propose the idea of an utility who compil the *.ajz file with only option u want (options,games, language, number of dir entry ).... it's a great idea!!!
10:20:50MeRWiNhow long until the next auto-build?
10:21:02MeRWiN1 or 2 minutes i'm guessing
10:21:03adi|homeso, bring it to something you like, edit, convert back.
10:21:04BagderMeRWiN: every 20 mins
10:21:08adi|homewe are looking for a base version
10:21:34BagderI'll propose a better, in my view, idea
10:21:39adi|homeshoot
10:21:42LinusNbobTHC: we can extend the configure script
10:21:43adi|homebarrel boy
10:21:51LinusNkind of like the Linux kernel build
10:22:06Bagderwe need the english "original" strings in one file
10:22:19Bagdertext, with some kind of identifier on the stings
10:22:35Bagdercould be <id>: <string>
10:23:07MeRWiNBagder: Think I fixed it :) Next auto-build should be perfect (got 1 warning as of this build)
10:23:12Bagdernow, all translators care about their language only, and thus I propose that we deal with one language per file
10:23:23BagderMeRWiN: good work
10:23:32adi|homebobTHC: about your request for a text viewer
10:23:35adi|homewe have a basic shell of one
10:23:38adi|homeit just needs work
10:23:40LinusNBagder: we need two strings per ID
10:23:41adi|homeits on my todo list
10:23:49BagderLinusN: why two?
10:23:55LinusNPlayer and Recorder
10:24:03adi|homenods
10:24:10adi|homethat i would agree with
10:24:10Bagderthey should be different IDs
10:24:21LinusNBagder: why?
10:24:32Bagderbecause there will be strings present in one but not the other
10:24:50Bagderthe ID would only be used to logically connect different language's strings anyway
10:24:51LinusNand?
10:25:05Bagderso why does it matter if they're one or two strings per id?
10:25:37Bagderanyway, that doesn't matter
10:25:38LinusNmaybe you're right
10:25:41MeRWiNTime for bed... g'nite
10:25:47LinusNnite
10:25:47adi|homenite
10:25:58 Quit MeRWiN ()
10:26:46BagderI then suggest we run a tool that outputs a "french.txt" with all the english phrases as comments, with the french ones as <id>:<string>
10:27:03Bagderit should accept the already existing french file as input and fill in those that already exist
10:27:18 Join Zagor [0] (bjst@as9-5-6.k.s.bonet.se)
10:27:30LinusNyo Zagor!
10:27:36bobTHCLinusN : I make some search on the MAS3507D and a lot of MP3 projects use it but i don't find one who use the DSP reprogramming feature of it....
10:27:44Zagoryo
10:27:53bobTHChello!
10:27:56LinusNbobTHC: that is because noone has the documentation
10:28:08LinusNbobTHC: but Vladimir Pantelic has
10:28:18bobTHCYes !!!
10:28:55Zagorvladimir has?
10:29:15Bagdermost likely, yes
10:29:19LinusNi think so, didn't he write a speech codec software for it?
10:29:46LinusNor am i talking about the wrong guy?
10:30:28bobTHCthey already have 2 inside CELP voice Decoder ADPCM encoder...
10:30:52ZagorLinusN: i haven't heard of it. or i've forgot (quite likely) ;)
10:31:22adi|homehow do i remove a file from cvs?
10:31:31Bagdercvs rm [file]
10:31:33 Join edx [0] (~edx@pD9EABA6B.dip.t-dialin.net)
10:31:34adi|homek
10:31:36Bagderremove for real first
10:31:38Zagorremove it locally first
10:31:38bobTHCIf he have it's a really good opportunity......
10:32:02BagderbobTHC: not if he's signed NDA's for the info he has
10:32:05*adi|home nods
10:32:32bobTHCof course but it's not sure...
10:32:56LinusNbobTHC: and we will not be able to release the source code
10:33:08bobTHCand he can help us without broke the NDA...
10:33:12LinusNif we have to sign an nda as well
10:33:56 Quit matp|work (Remote closed the connection)
10:34:07adi|homebah, who ever follows up on those things anyway ;)
10:34:17LinusNZagor: your "limit reached" strings are too wide for the player
10:34:27Zagori know :)
10:35:00Zagori couldn't think of better wording, so I just threw that it
10:35:09LinusNit's ok for now
10:35:29LinusNthere is only one person who is likely to see it anyway :-)
10:35:33adi|homeoh damnit..
10:36:02adi|homecan someone do me a favor and send me the stubs.c file from uisim/common?
10:36:10*adi|home is an idiot
10:38:01adi|homenm
10:38:45matslZagor: have you thought about the use of the params and audio icons since yesterday? (Now that we have LinusN on board.)
10:38:58Zagoradi|home: here you go
10:39:32adi|homety
10:40:04Zagormatsl: ah, right. linus, did you object to the use of "audio" and "param" to flag states of the UI on player?
10:40:39LinusNno, i just can't find a sensible use for them
10:40:44adi|homeumm.. zag.. did you look at that configure patch that got sent in?
10:40:55adi|homei hadn't noticed issues with the upper vs lower case had you?
10:41:09LinusNi don't think those icons can convey any useful information
10:41:29*Zagor is still catching up with his mail
10:41:37adi|homenods
10:41:47adi|homeokay.. then im going to commit it cause it seems to work
10:42:06LinusNbut if we can find a way to use them, go ahead
10:42:14ZagorLinusN: i think you may be forgetting just how cramped the player is. I know i can be confused about wether i'm in wps or dir browser sometimes
10:42:28adi|homeworst case we try the icons, don't like them and burn them in effigy later
10:42:44LinusNok, so (Audio) tells us we're in WPS?
10:42:56BagderI kind of think so too. Throw them in now to have hem tested by the crowds
10:42:57matslLinusN: Jupp!
10:43:00Zagorbut ok, i though you had some strong objections. apprently not. matsl: go ahead with your patch.
10:43:29matslok. ;-)
10:43:51matslLinusN: Your gonna love it ;-)
10:44:47LinusNhehe
10:44:53LinusNi hope so
10:45:58Bagderb-r-e-a-k
10:46:09Bagderred build
10:46:21LinusNwoooweeeewooooweeee
10:46:26Zagorkaboom
10:46:31adi|homeprob my fault
10:46:33adi|homeworking on it
10:47:00adi|homeyup.. my fault
10:47:02adi|homefixing it
10:47:05BagderI like the compile status table, we would be in worse shape without it
10:47:18*adi|home nods
10:47:18LinusNdefinitely
10:47:18HesIt's really good
10:47:27HesMaybe an IRC bot could report changes in the state too 8-)
10:48:10Bagderthat would be pretty cool ;-) have the bot say "red build"
10:48:14Bagderhehe
10:48:47LinusNc00lness indeed
10:48:55LinusNBagder: start dancing!
10:49:04Hesand report new CVS log messages
10:50:15Bagderit would be pretty simple to have it just output the contents of a local file if it exists
10:50:26Bagderand then run a local script that builds it
10:51:19Bagderbut I'm not up to writing that ;-)
10:52:32HesI might
10:55:16adi|homeokay.. im confused
10:55:30adi|homehmmm...
10:55:51BagderI've thought about putting lots of the web scripts in cvs too, to allow others to check them out/use them/fix bugs
10:57:11adi|homeis it a bad idea to have #ifdef SIMULATOR code in backlight.c?
10:57:11adi|homeits either that or we need a file for 2 dummy functions in uisim/common
10:57:11adi|homenot sure which is better.
10:57:11Zagorput it in stubs.c, that's what it's there for
10:57:11adi|homebah.. thats why i deleted it :)
10:57:11adi|homelol
10:57:39LinusNadi|home: some files already have two sets of functions
10:57:44LinusNlike mpeg.c
10:57:49adi|homeright
10:58:01adi|homei was wondering if that was a prob for backlight
10:58:09LinusNplacing the stub functions in backlight.c is ok, imho
10:58:21adi|homeZagor, what do you think?
10:58:22Zagornot a prob, but unless you actually do something it will just clutter the code.
10:58:37BagderI prefer having sim-stuff in the sim dir
10:58:43Bagderas far as possible
10:58:43Zagorif it's purely stubs, they should be in the simulator code and not the firmware code
10:58:43adi|homewhat do you mean 'unless you actually do something'?
10:58:56adi|homeahh.. okay.. their not
10:59:02adi|homefor the sim they are
10:59:06adi|homebut on the target they aren't
10:59:23LinusNadi|home: that's zagors point
11:00
11:00:03adi|homeconfirm: if the sim version is _only_ a stub, leave it out of the main tree.
11:00:03adi|homeyes?
11:00:08Zagoryes
11:00:12adi|homek. done
11:00:13adi|homety
11:00:16LinusNif the stubs are only there to satisfy the compiler, they shouldn't be in the main tree
11:01:12adi|homehow does one revert to a previous branch in cvs?
11:01:39Bagder-r tag
11:01:42adi|homek
11:01:45Bagdercvs up -r tag
11:01:56Bagderor 'co -r tag' in a new place
11:04:02 Nick mbr|gone is now known as mbr (~tiw4mabr@rhlx01.fht-esslingen.de)
11:04:11LinusNhi marcus
11:04:13mbrHi
11:04:21Bagdermorning mbr
11:05:20mbrI ave written a cron thread ...
11:05:22 Join matp [0] (~mpritch@gateway.riversoft.com)
11:05:32LinusNmbr: nice
11:05:43Bagdergee, mbr sure rocks today :-)
11:05:48adi|homeokay.. the build should be fixed now
11:05:50mbrNow i have some questions
11:05:58mbrwe should it go to? firmware?
11:06:13LinusNnot sure
11:06:22Zagorwhat do you use it for?
11:06:22matphi all
11:06:22BagderI think apps
11:06:32adi|homewould it belong with thread.c?
11:06:41LinusNZagor: it is used to schedule tasks
11:06:43adi|homeor is it just a derivative of thread.h?
11:06:46mbrI want to use it to schedule settings save to disk
11:06:47matpwhat's new?
11:06:49Zagor...such as?
11:06:54LinusNi think it should be in firmware
11:06:57adi|homeah.. id say apps then.
11:06:59Zagormbr: ok
11:07:08LinusNthe battery charger will use it
11:07:10 Join RipnetUK [0] (~george@ripnet.fsnet.co.uk)
11:07:21mbrmabe status bar too
11:07:22adi|homewhere are our other thread files?
11:07:34LinusNadi|home: everywhere
11:07:43adi|homewell.. let me put it this way
11:07:47Zagoradi|home: it's not a thread thing, it's more a maintenance thread doing things periodically
11:07:50adi|homembr: it makes use of threads.h right/
11:07:53adi|home?
11:08:02BagderI still vote apps
11:08:08adi|homei vote apps
11:08:11mbradi|home: yes
11:08:11adi|homeand my reason is this
11:08:14*Zagor feels The Great Mixup coming...
11:08:17adi|homenothing in firmware uses threads.h
11:08:22LinusNif we place it in apps, the charger can't use it
11:08:23adi|homeeverything that uses it is in apps/
11:08:29Bagderthe chargre code should be apps too
11:08:43Bagderonly the charger/hw api should be firmware
11:08:43ZagorLinusN: the charger shouldn't use it. the cronthread uses charger
11:08:53LinusNadi|home: backlight uses threads.h, usb uses threads.h ...
11:09:10LinusNZagor: WHAT?
11:09:24BagderI think the diff should be that if I want to build a separate image, what do I need in the kernel
11:09:29BagderI don't need a cron thread
11:09:34Zagorcrontthread is the upper level, charger is the lower. cron calls charger, not the other way around
11:09:52Bagdertrue
11:09:55LinusNcharger registers itself to the cron thread, doesn't it?
11:09:59Zagorno
11:10:02Bagderno
11:10:06Zagori'll have none of that
11:10:20Bagderhehe
11:10:39Topic"The Home of The Code Police" by Bagder (~daniel@as3-3-2.ras.s.bonet.se)
11:10:50Zagorpure, straight-to-the-point code. no registering and cloak-and-dagger stuff
11:11:13Bagderit also makes the line between apps and firmware easier
11:11:17Zagoryes
11:11:22LinusNBagder: really?
11:11:26Bagderyes I think so
11:11:47mbrYou don't want registering functions?
11:12:04LinusNhaving the application bother with charging the batteries is not good, imho
11:12:10mbrThen I don't have a cron :)
11:12:21Zagorno, i don't. we get c++-like problems following the code flow
11:12:43adi|homelol
11:12:49LinusNZagor: so you dont like tick_task either?
11:13:11Bagderhey, let's not shake everything now
11:13:15Bagderlet's stick to the cron ;-)
11:13:37Zagortick_task is different, imho. it's waaaaay more low level
11:13:39LinusNbasically, the firmware will not be able to use the cron thread
11:13:51BagderI think the cron should use the firmware
11:13:59Zagorabsolutely
11:14:10Bagderas the real cron ;-)
11:14:22LinusNyou mean the one that you register to?
11:14:30LinusNwith a crontab
11:14:34Bagderit doesn't matter that much how you use it
11:14:37*adi|home wonders who's on first
11:14:42Bagderto me
11:14:54Zagoryou don't register in cron. you set up a script in one place that executes everything. in the application layer.
11:14:56LinusNmy point is that the firmware will not be able to use the cron thread
11:15:12Bagderthe firmware is *used* from the cron thread, not usee
11:15:16Zagorthe firmware *should* not use the cron thread. why would it?
11:15:28LinusNto charge batteries for instance
11:15:31Zagorwhy are you so fixed on the firmware calling everything else?
11:15:47Zagorthe firmware is the lowest level. it should be called, not call
11:16:11LinusNhave it your way, then
11:16:20Bagderthen of course we can discuss if we should not run the charger in the cron
11:16:28Zagorcharging batteries requires, what? a call to battery_charge_control() every 30 seconds?
11:17:30Zagormaybe i have misunderstood your idea with a cron thread, but then tell me
11:18:39Bagderif I had it my way, the scroll thread would be apps too... ;*)
11:18:55LinusNBagder: i agree
11:19:13LinusNbut i am not sure that i like it as a thread :-)
11:19:26BagderI think we must have it that way
11:19:34Bagdereven if I agree with your concerns
11:19:45Zagorwhy not a thread?
11:19:58Zagorbecause things can draw at the same time?
11:20:05Zagorwe have pause/resume now
11:20:10Bagderyes, that would be the main reason
11:20:18LinusNbecause it updates the LCD behind your back, with a cloak-and-dagger
11:21:24Zagorare you just being grumpy or do you not see the difference?
11:21:45Bagderwe need it as a thread to have it run fine all over
11:21:46LinusNdifference to what?
11:21:55Zagorbtwn cron and scroll
11:21:55mbrSorry have to ga, back in an hour ..
11:22:01LinusNBagder: yes, it is a nice feature
11:22:04adi|homeBagder: if you get a chance, pull down uisim/common/lcd-common.[ch] and see if that make sense along the lines we discussed
11:22:45matslwe have a "undefined reference to `backlight_time'" in the simulator builds. Anyone working on it?
11:23:40adi|homei already got it
11:24:02adi|homeif you refresh your cvs files it should go away
11:24:05Bagderadi|home: I'm not sure why you want to merge the two lcd files when they have nothing in common?
11:24:40*adi|home had a good reason before, he just can't think of it now
11:24:50Bagderwouldn't it be better to just put the common functions in a common file
11:24:57adi|homenone of the makefiles ref. it, its just sitting there now
11:25:16adi|homeand have the include lcd-win32 and lcd-x11 as needed?
11:25:22Bagderso if you add lcd_define_pattern, it would probably look the same on both win32 and x11 and it could be in that file
11:25:51bobTHCfor technical presentation (not the complete doc) of the MAS3587F (recorder DSP) : http://www.electronicstalk.com/news/fli/fli115.html
11:26:16HesAs the charger has started to run more often (it now sleeps 10s, 10s, 10s, 10s, 20s) and the logic is getting a bit more complicated, I don't know if it makes sense to run it from cron
11:26:31Heswould require more global variables
11:27:04adi|homewell. i could just tear that stuff out for now and turn it into a stub :)
11:27:38matsladi|home: I'm still getting it
11:27:44BagderI'm not telling you how to do things, I just pour out a little of my opinions. I have a lot of those ;-)
11:28:05*adi|home slaps matsl
11:28:06adi|homehehehe
11:28:11adi|homegive me a min
11:28:44matsladi|home: mod-build builds. simulator builds don't. That's what I'm seeing.
11:29:04adi|homei know the error.. give me a min :)
11:29:20Bagderwe all see those red blobs hurting our eyes ;-)
11:36:08adi|homeodd...
11:36:13adi|homeits not failing on my sys
11:36:19adi|homeand ive cleaned my cvs
11:37:31adi|homematsl: you wanna try again
11:37:48matslok, I'll try it now.
11:38:39Bagderadi|home: it uses backlight_time() which doesn't seem to exist
11:38:48adi|homeit does
11:38:50adi|homeits in stubs
11:38:52adi|homei replaced it
11:38:57adi|homeand put it back in the Makefile
11:39:03adi|homeunless the revert didn't take
11:39:22Bagderno the stubs wasn't added again to the makefile
11:39:43adi|home" unless the revert didn't take"
11:39:48Bagderright
11:40:11matsladi|home: Need I say that it still fails? It does.
11:40:58adi|homehave you pulled down a fresh makefile in the uisim/x11?
11:41:13adi|home grep "stubs" Makefile
11:41:13adi|homeSRCS = screenhack.c uibasic.c resources.c visual.c lcd-x11.c stubs.c \
11:41:13adi|home$(OBJDIR)/stubs.o: ../common/stubs.c
11:41:15adi|homei have it
11:41:19adi|homeand i just pulled down the fresh one
11:41:34Bagderthat's because you used -r [tag] before
11:41:37Bagderand it has stuck
11:41:49Bagder(sticky tag is actually the cvs term)
11:41:50adi|homeokay.. so where is the problem?
11:42:12adi|homethen your wrong that " no the stubs wasn't added again to the makefile"?
11:42:30adi|homeor do i need to commit that Makefile or something?
11:42:43Bagderdo "cvs up -A"
11:43:01Bagderand you'll get the latest without sticky tags
11:43:13adi|homewtf are these sticky tags
11:43:15Bagderand then you'll get the latest makefile, which doesn't have stubs
11:43:30adi|homei don't get it.. i wanted to prune back a branch, and did what you told me to :)
11:43:32Bagdersticky tags are tags that are remembered on the file name
11:43:48Bagderadi|home: you didn't really explain it that way to me so I misunderstood
11:44:01Bagderyou reverted your local copy of that file to that version
11:44:10Bagderyou can't revert the repository
11:44:20Bagderyou need to redo the change and commit
11:44:31adi|homeheheh thats what i asked :)
11:44:36adi|homeleast i thought i did..
11:44:40BagderI realize that now, sorry
11:44:40adi|homemoot point
11:44:43adi|homenp
11:45:01adi|homeim not as exp in cvs, so i didnt phrase it the right way :)
11:45:08adi|homesccs is sooo much more comfy for me
11:45:16Bagder=)
11:47:52Hadakasvn powa
11:48:09adi|homefixed
11:48:09adi|homenow bugger all of you
11:48:11adi|homeim going to bed
11:48:20adi|home:)
11:48:21adi|homenite
11:48:25Bagdernight adi
11:53:13bobTHCSomebody interested by little digging on the MAS3587F variable pitch possibility?
11:53:41Zagorsure. you found something?
11:56:23bobTHCyes it's possible to define the Oscillator frequency and as is a serial feed i thing it's possible to change the speed of the decoded feed...
11:56:44bobTHCi'm not very clear, sorry for my english...
11:57:37LinusNit's actually about changing the information to the DSP about what frequency the external oscillator is
11:58:20bobTHCtake a look a sourcecode from Dragon Mp3 http://users.pandora.be/DragonMP3/sourcecode/MAS3587F.h
11:58:49LinusNif we fool the DSP by lying about the frequency we might make it playback at another pitch
11:59:57Zagorsounds worth trying
12:00
12:06:32matslUI question: I would like to change the WPS for the player. Are you really satisfied with how it looks now?
12:06:36Bagderthe build went green at last
12:06:48Bagderit was updated this morning
12:06:53LinusNmatsl: we will never be satisfied! :-)
12:07:23Zagormatsl: how would you like it?
12:07:23matslLinusN: Well that's anotherway of saying the same thing isn't it?
12:07:45Bagdermerwin made it show id3 tags
12:07:48*Bagder runs for lunch
12:07:59LinusNmatsl: what is your suggestion?
12:08:19matslHuh... hard to discribe in IRC but I give it a chat.
12:09:07matsldisplay time played, scroll artist, song, album in one line, show playslist.index ...
12:09:26LinusNi like the scroll idea
12:09:27matsl... be smart about id3 tags present or not ...
12:09:47LinusNmatsl: you mean display the file name if there is no id3?
12:09:49matsl.. and fall back to file if they can't be found.
12:09:54LinusNok
12:09:58matslLinusN: Jupp!
12:10:05LinusNi like that
12:10:18LinusNwould you display the playlist index in the scroll?
12:10:58matslLinusN: No I don't thinks so. My code displays like this now: "01/12 04:44"
12:11:19LinusNok, so the total time is not displayed?
12:11:53matslLinusN: No. Its displayed when the song is paused but I'm not sure about that idea.
12:12:46***Saving seen data "./dancer.seen"
12:12:57LinusNhow about displaying things in different time slots?
12:13:13LinusNlike time for 3 secs, playlist index for 3 secs and so on...
12:13:28matslLinusN: That has been proposed yes but I thought it was confusing.
12:13:40LinusNok, i haven't tried it myself
12:13:46matslLinusN: did you try that patch and looked for yourself?
12:14:10LinusNno i haven't (working on other stuff atm)
12:15:59matslLinusN: OK. i could try to join forces with the guys that have supplied patches and add a fourth WPS screen
12:16:17LinusNdo so
12:16:49LinusNit can't hurt
12:17:31matslLinusN: Right I'll do that.
12:17:36bobTHCfor more Mas3587F digging take look to http://users.pandora.be/DragonMP3/sourcecode/MAS3587F.c too (it's basic commands)
12:18:09ZagorbobTHC: i see nothing about changing pitch
12:20:16bobTHCof course i just talk about the Oscillator frequency variable and that can be divert to make it...
12:23:09ZagorbobTHC: can you test this?
12:24:20bobTHCtest my idea?
12:24:24bobTHCwith a patch?
12:24:30Zagoryes
12:24:58bobTHCnope
12:25:08Zagorok
12:25:39bobTHCi'm not C coder, perhaps next year !
12:25:56Zagorok, no problem
12:26:51 Quit edx ("be back in two hours")
12:30:15*bobTHC go to lunch
12:33:23*mbr is back
12:34:03mbrZagor: you want a staic list of cron jobs, defined at compile time?
12:34:11notchokay this is killing me... how do you do those prettly coloured broadcast meaasages!
12:34:31mbrAnd each job determines if something should be done
12:35:03bobTHCnotch use the "/me" command
12:35:18*notch testing
12:35:23bobTHC:)
12:35:27notchwhoho!
12:36:01mbrzagor: and every run all defined cron functions will be called?
12:36:29Zagormbr: i'm not 100% sure. i just want to keep the complexity down
12:36:54Zagori haven't been part of the cron discussion, so i'm sorry if i sounded a bit harsh. could you fill me in a bit?
12:37:03matpmatsl I did the patch for the recorder to allow you to define the wps screen. i'm sure something sim could be accomplished for the player
12:37:57matslmatp: is your patch in CVS?
12:38:22mbrZagor: The problem was to schedule a write to disk when settings are changed. For example by pressing F3 or changing the volume in wps
12:38:33matpmatsl: I haven't checked for a few days. I don't think so.
12:38:58mbrAnd so Linus mentioned, that we should have a cron thread that handles that events.
12:39:18Zagorbut then the disk will still spin up just to write settings
12:39:19matpmatsl: I posted it to the list. you have a player I take it ...
12:39:30LinusNZagor: that is inevitable
12:39:45matslmatp: jupp!
12:39:45LinusNyou'll have to save the settings sooner or later
12:39:52mbrBut you also should define a timeout, because the disk possibly never spinns up.
12:39:57Zagorwe do spin the disk sooner or later too
12:40:13LinusNno, we just might turn off the unit before that
12:40:15Zagorwe have always discussed writing settings when we spin the disk for another reason
12:40:15mbrZagor: or never ..
12:40:37matslmatp: I have been thinking of your customization patch and yes it could be used for the player ...
12:40:37Zagorthat's a special case, and rather unusual i would think
12:40:39LinusNI was thinking of a "Save Settings" menu option
12:40:50ZagorLinusN: I can buy that
12:40:51matpmatsl: there's 2 lines, right. i was thinking of allowing you to define which id3 tags get scolled on which lines
12:41:15LinusNthe "save when spinning up" thing is very good for Resume, but not for settings
12:41:22matslmatp: right.
12:41:33matpmatsl: say: line 1: artist, track. line2: bitrate
12:41:38matslmatp: But there is more to it than id3 tags
12:41:41RipnetUKi asked before, but cant we have a non-blocking save to disc?
12:41:42Zagorit's good for settings too
12:41:51matpmatsl: yes, definitely
12:42:04Zagori bet 95% of people want to try their new settings before powering off. and then they'll spin the disk.
12:42:16matpI want to be able to toggle the timer bar as well as the icon bar
12:42:22LinusNZagor: what if the settings don't apply to MP3 playing?
12:42:26matslmatp: now we only have one line scroll too which makes it hard just to caoncatenate things
12:42:28LinusNlike LCD contrast
12:42:49Zagorstill, why change settings and then immediately power off? it doesn't seem like normal user behaviour.
12:42:56dwihnopixel based area scrolling would be neato :)
12:43:00matpmatsl: there is a patch for multi line scrolling though?
12:43:12LinusNstill, that normal user will wonder why the settings aren't saved
12:43:22RipnetUKdid neone try my multiple line scroll hack?
12:43:30mbrThat is my tought, too
12:43:33Zagorno, because he won't do that. you're inventing a problem without knowing if it's likely to happen.
12:43:38matslmatp: I know
12:43:39LinusN"to save the settings, play a song"
12:43:50matpnot yet. I would like to try it with my wps patch. i will try and do so today
12:43:50mbr:)
12:44:14Zagorremember dir browing spins the disk too
12:44:15matpmatsl: what else would you add other than id3 tags?
12:45:05matslmatp: I would like to display time played, playlist info (index, amount) etc.
12:45:12LinusNZagor: not when you leave the menu
12:45:22matslmatp: time played doesn't scroll nicely.
12:45:34dwihnoGood afternoon! :D
12:45:40ZagorLinusN: think back. how many times have you changed a setting and then immediately powered off the unit?
12:45:44*dwihno sings some Caramell beats for Zagor - I know you like it ;D
12:45:58mbrnevertheless, the problem of not written settings is still there, even if we have a scheduled function
12:46:36LinusNZagor: i did it yesterday
12:46:37Zagormbr: exactly
12:46:40LinusNwhen i set the clock
12:47:03Zagorhaha. you chose the one single setting that's not stored on disk :-)
12:47:06LinusN:-)
12:48:03LinusNZagor: you mean that it's acceptble that the settings are not saved if the user shuts off the unit without having used the disk?
12:48:30Zagoryes. at least until we get a dozen reports about how annoying that is.
12:48:37matpmatsl: why not?
12:48:43Zagori don't think we'll get even 2
12:48:50LinusNi don't think the user cares *that* much about battery usage
12:49:12RipnetUKi think a explicit save to disc option is better, because then you can try settings, and abandon them if u dont like them. Also easier to implement
12:49:40ZagorRipnetUK: we have to flush settings anyway. volume is changed outside the menu, for instance.
12:49:44matslmatp: I tried it and thought it hurt my eyes. Both scroll and seconds changeing at the same time. To much motion for my likeing.
12:49:49mbrRipnetUK: you mean a write immidiately
12:49:53LinusNand then you can inform the user when the settings aren't save because of low battery
12:50:20matslmatp: Time ticking should be a none scroll display I think.
12:50:22matpmatsl: it would be nice to have partial line scolling
12:50:51ZagorLinusN: i think you have forgotten just how annoying that is on the player, saving every volume change to disk
12:50:54matslmatp: Culd fix it but we don't have many chars left to scroll then.
12:51:21LinusNZagor: you have a point
12:51:22matpmatsl: how many chars on the player?
12:51:27matslmatp: 11
12:52:03matslmatp: My current code has first line scrolling major id3 info ...
12:52:34matslmatpf: second line stabel while playing (time) and second oine scrolling other info while song is paused.
12:52:42matpmatsl: why not allow the user to define the format
12:53:14matslmatp: Yes, why not!
12:53:59matpmatsl: :) It's tricky to think of how to express it in a menu
12:54:54matpmatsl: But possible. I guess this may be useful for the recorder as well
12:55:38matslmatp: I'll think about it.
12:56:07matslmatp: Any feeling about how the wps-code looks like after such a change?
12:57:07matpmatsl: lots of #ifdefs, and messiness :)
12:58:31matslmatp: That could be a problem then. It is an mp3 player after all. Its the music we want.
13:00
13:00:23matpmatsl: I'm not aware of any cd/minidisc/mp3 players that allow you to define the output in this way. but it's a very nice feature, I think.
13:02:31matslmatp: agreed. I could very well imaging choosing some id3 tags this way could be feasable.
13:03:21matpmatsl: my code is pretty clean. iterate through the lines, and switch on the user-defined value. I haven't really thought about how to attack multiple tags per line.
13:04:37matslmatp: (Just at this moment) I got a completely different idea.
13:04:53matpmatsl: just need a flag for whether it's multi, and an array of tags per line
13:05:00matpmatsl: shoot
13:05:55matslmatp: Each line could be a "printf-style" string that's interpreted
13:06:03*Bagder returns
13:06:09Bagdergee you talk a lot
13:06:32matslplease join
13:06:51matpmatsl: yeah, I thought that before, but I'm not keen on parsing the line. also, it's not that intuitive for the user.
13:07:08BagderI'm all for a custom wps screen
13:07:17Bagdercustomisable
13:07:32matpmatsl: we should think of the wps screen operating in different modes.
13:07:35Zagori think matp's array solution sounds a lot easier for the user. just select which info tags you want, in which order.
13:07:38matslmatp: each id3-tag could be represented by a letter, time a letter etc. If line is longer than 11 chars it scrolls.
13:08:25matpmy code currently displays the title of the tag, e.g Artist, Album. It's not hard to do this.
13:08:45matpBut there are real-estate issues on the player
13:09:07matslmatp: ??
13:09:12matp(which I've never seen in action :) )
13:09:42matpwell, I'm assuming the menu is limited in the player ?
13:09:49matpthe menu size, I mean
13:10:21matslmatp: display has two lines??
13:11:45matpI mean, if you want to actually define the line in the menu, you have 11 chars to do this ?
13:12:29matpIf so, I can see why you might want to keep this to printf style tokens, rather than the full names
13:12:50matslmatp: ok
13:13:29matslmatp: Its a matter if input as well. easier to input on char than a word. keybord on the player isn't that friendly.
13:13:39Zagorone char per tag, i'd say. 'A' for Artist, 'a' for the actual string "Artist:"
13:13:45matpBagder, have you tried the definable wps patch ?
13:13:56Bagdersorry, no
13:13:59Zagormatsl: upper line is tag list, lower line explains which tag you have just selected. like:
13:14:03matpYeah, but on the recorder, I like Artist, not %A
13:14:26ZagoraA lL tT sS <−−- cursor on A
13:14:30ZagorArtist
13:14:31BagderI agree with matp, we should probably do differently on them
13:14:41Zagoryup
13:14:45matpZagor, yeah, that would work
13:15:06BagderZagor: that's a neat idea
13:15:18LinusNwhen we implement this, we REALLY want persistent settings! :-)
13:15:25Bagderhehehe
13:15:28Bagderamen
13:15:50LinusNwhy not enable it by default, anyway?
13:15:52matpyeah, I'm getting pissed off having to redfine my wps screen each time i boot up
13:16:07matslmatp: change the code
13:16:20matpI could do with someone helping out with the persistent settings
13:16:43LinusNmatp: what do you mean?
13:16:51matpI had a look, and it wasn;t immediately clear.
13:17:07LinusNput this first in the file:
13:17:16LinusN#define SAVE_TO_DISK
13:17:21LinusNsettings.c
13:17:41matpI have recorder, which saves to rtc anyway
13:17:45LinusNoh
13:17:58LinusNwhat was your problem then?
13:18:31matpI wasn't sure how to go about extending the config block.
13:18:47matslZagor: With one char of info an one for actual value we get 26 items that can be displayed in the wps.
13:19:22matpI have a line->tag mapping. 7 lines, do I define 7 config slots?
13:19:33LinusNmatp: look at how the new lcd contrast setting does it
13:20:05Zagormatsl: cool, isn't it? :-)
13:20:09matpDoes it squeeze it into 1 bye, or soimething?
13:20:11matpbyte
13:20:24LinusNmatp: which one?
13:21:29matpIt's OK, I'll take a look at the code before I comment. It's not to hand. I may get a chance this afternoon.
13:22:05matpDo you think the line by line approach in the menu is a good way to go? It feels a bit clunky to me...
13:22:21LinusNthe lcd setting is easy, it just uses an unused byte, it doesn't try to use a few bits in an already allocated byte
13:22:55matslZagor: We would need some way to handle fields. Like padding with zeros but also displaying * or something if value is to big. (Avoiding scrolling)
13:23:22matpLinusN: OK, but I have 7 fields I want to save. Should I use 7 bytes, or cram them into 1?
13:23:45Bagderwe need disk saving on the recorder too...
13:23:53LinusNthat all depends on how many bits each value requires
13:23:56Bagdersince otherwise we'll soon spend all those 44 bytes
13:24:24LinusNif each field takes 2 bits you only need 2 bytes
13:24:46matpThat's why I didn't extend it. Problem is, I don;t know how may values those fields will have.
13:25:14LinusNi urge all recorder users to test the lcd contrast setting in all possible ways
13:25:32Zagormatsl just said each tag can have one of 26 values
13:25:47matpcurrently i have 5 or 6 values. matsl has pointed out more. it guess I should work out maximum possible size
13:25:58*bobTHC is back
13:26:03LinusNok, so 4 bits may be enough
13:26:04matp:)
13:27:15LinusNif the new format setting becomes reality, those settings will probably not fit in the RTC RAM
13:27:30*matp nods
13:27:42LinusNZagor: is your OldPlayer alive?
13:27:51Zagoryup
13:28:07LinusNtry the lcd contrast and tell me if i should use a higher default
13:28:14matpwhen is write-to-disk settings planned?
13:28:15RipnetUKthe LCD contrast patch -I get ablank screen on my recorder20 - you should ahve a higher default :)
13:28:25RipnetUKi assume its on 0
13:28:35LinusNRipnetUK: pull out your batteries
13:28:44RipnetUK? what to blank the CMOS?
13:28:48LinusNyes
13:29:02Bagderheheh
13:29:04RipnetUKshouldnt it check for a zero value, and substitute something visible :)
13:29:10Bagderis that a step in the upgrade to rockbox 1.3 process? B)
13:29:22RipnetUKhehe
13:29:24matpgot to lunch. brb
13:29:34LinusNRipnetUK: 0 might just be the value you have chosen
13:29:52RipnetUKi havent been able to choose a value yet, because I cant see the screen :-O
13:30:05LinusNi meant that 0 is not an invalid setting
13:30:08RipnetUKdoesnt cmos momory start with random values?
13:30:21LinusNwe have a checksum
13:30:27 Quit matp (Remote closed the connection)
13:30:51RipnetUKi see - smart
13:31:02RipnetUKyeah, that fixed it
13:31:26 Join matp [0] (~mpritch@gateway.riversoft.com)
13:31:41matpouch, x died
13:32:01RipnetUKanything below 18 seems invisible
13:32:39RipnetUKand above 35
13:32:41LinusNi can go as far as 12
13:33:13LinusNand 42
13:33:16 Nick elinenbe|sleep is now known as elinenbe (trilluser@pcp02254422pcs.wanarb01.mi.comcast.net)
13:33:40matslZagor: In addition to add tags to a line we could have multiple lines displayed on one screen line. Each with its own time slot.
13:34:13elinenbehelp me please. how do you run the loadable fonts patch? do I need to "compile" the fonts? and where do I put them?
13:35:17RipnetUKive also found that the proportinal fonts break the time/date setting screen - it expects fixed width.. sorry
13:36:15LinusNhey Zagor!
13:36:44Bagderelinenbe: you need to build the bdf2ajf in tools/
13:37:05Bagderelinenbe: then convert a .bdf font to a system.ajf and put that in your archos root
13:37:41bobTHCread this (about Recorder DSP) : Micronas is unable to ship its MAS3587F to the United States because the chip, an encoder/decoder designed to support the ripping of tracks from a CD, violates the Audio Home Recording Act of 1992, von Janecek said. But a change in the chip's firmware has enabled Micronas to sell it as the MAS35x9F MP3 decoder with AAC and G.729 voice codec features.
13:38:11bobTHCAAC and G.729 ?
13:39:24elinenbeBadger: thanks
13:40:38LinusNbobTHC: it can honor the Copyright bit in the S/PDIF stream
13:40:58LinusNand in the MP3 stream
13:41:19LinusNso the software can refuse to record copyrighted data via S/PDIF
13:41:35bobTHCyes it's about SCMS..
13:43:36 Nick LinusN is now known as Linus|lunch (~linus@labb.contactor.se)
13:49:36Zagorelinenbe, bagder: the makefile should do all that for you
13:50:56Zagormatsl: imagine several scrolling lines, with a timeslot each. we can create true mayhem on the player! :-)
13:53:52matslZagor: I know. Maybe we could use the WPS state as well for customization!
13:54:10BagderRipnetUK: is your multiline scroll working good?
13:55:04Bagderyour file (i.e not patch) is really hard to apply now
13:59:51elinenbethe atari-small font on this page: http://hea-www.harvard.edu/~fine/Tech/x11fonts.html is a great font!
14:00
14:00:33Zagorooh, cool
14:00:40elinenbeit is GREAT!
14:00:47*Zagor hurries getting loadable+proportional working
14:00:57Zagoroh, it's not proportional :-(
14:01:08Zagorstill looks good
14:01:08elinenbeno, but it works great.
14:01:09Bagdershouldn't be that hard to make it
14:01:25elinenbeso much information on the screen :)
14:02:16Zagortoo bad it's ascii only :-(
14:02:35Bagderwe'll hire dwihno to extend it! ;-)
14:02:35bobTHC:(
14:02:44elinenbeyeah, only 128 chars... how many chars are normal bdf fontsets
14:02:46elinenbe?
14:03:15Zagorthere is no "normal". but we want at least full iso-8859, i.e. 256 chars
14:03:15BagderI guess they can basicly be enormous if they provide a huge set
14:03:34Zagoryeah, some provide big sections of unicode and are, well, big... :-)
14:04:04Bagderwe want chinese rockbox!! ;-P
14:04:15bobTHCand russian!!
14:04:31Bagderrussian already works
14:04:40elinenbeSince I am fluent in both, I will get around to tranlating everything.
14:04:51Bagderheheh
14:06:10elinenbethe only problem with the loadable fonts is their height. most of the screens are quite messed up...
14:07:14Bagderyes, we have some work to do there
14:07:21elinenbelike when using a larger font.
14:07:50elinenbehere are all the unicode fonts : http://www.cl.cam.ac.uk/~mgk25/ucs-fonts.html
14:09:39 Join data2one [0] (~datazone@chaos.iadfw.net)
14:12:17 Quit data2one (Client Quit)
14:12:47***Saving seen data "./dancer.seen"
14:13:56 Join datazone-work [0] (~datazone@chaos.iadfw.net)
14:16:46dwihnoBagder: I'm not your font slave! :)
14:16:52Bagderdarn!
14:16:59dwihnoBagder: I did it for myself since I thought the default font was no-good
14:17:04dwihno:)
14:17:07Bagderbut then... what are you? B)
14:17:26dwihnoI am the bishop of Caramell
14:17:29Bagderaaaah
14:18:56 Nick Linus|lunch is now known as LinusN (~linus@labb.contactor.se)
14:26:00 Quit elinenbe (Read error: 104 (Connection reset by peer))
14:26:19 Nick matsl is now known as matsl|working (~matsl@as13-4-5.mal.s.bonet.se)
14:27:27RipnetUKbadger - the multi-line scroll did NOT look good, i think its better to scroll the entire screen (except the status bar) left and right, and not the lines 1 by one
14:28:04RipnetUKsorry didnt do a patch, but was low-quality hacking, and jsut a proof of concept really, rather than a serious submission
14:28:11RipnetUKit didnt take into account proportinal fonts etc.
14:28:45Bagderok
14:28:46RipnetUKI am thinking that we should be having a wider (3x screen width perhaps) screen buffer, and scroll that when we update the LCD instead of scrolling line by line? i think that would look better than each line scrolling seperate
14:28:58RipnetUKalso easier to do I would imagine?
14:29:10BagderI've thought about that too
14:29:13 Nick matsl|working is now known as matsl (~matsl@as13-4-5.mal.s.bonet.se)
14:29:19Bagdermaking a bigger "virtual screen"
14:29:32matslRipnetUK: I tried with different scroll speeds and that looked ok.
14:29:59matslRipnetUK: But only on the player.
14:30:17RipnetUKit worked on the player :) im suprised i didnt check that
14:30:42matslRipnetUK: No. my own proof of consept experiment.
14:31:00 Join elinenbe [0] (trilluser@pcp02254422pcs.wanarb01.mi.comcast.net)
14:31:10matslRipnetUK: I did an experiment with scrolling two lines.
14:31:37 Quit elinenbe (Read error: 104 (Connection reset by peer))
14:31:38RipnetUKi definately think the way forward is to have a larger virtual screen, with the actual screen a window on it. You could have the scroll thread maintain 2 rectangles one for the area which is scrollable, and the other for the screen region to scroll in
14:31:40matslRipnetUK: Each line could scroll eitherfast or slow.
14:32:05RipnetUKi see, it looked naff on the recorder with 8 lines scrolling and wrapping / repeating at different points.
14:32:48BagderRipnetUK: it would only be a matter of enlarging the screen buffer actually, and setting the X.Y for the visible screen window
14:33:04RipnetUKyeah, thats another reason to like the idea
14:33:23RipnetUKwe could also have an option to bounce back and forward like the 'real' firmware
14:33:29Bagderbut we would need the scrolling and possiblt scrolling multiple lines anyway
14:33:44RipnetUKwould we? what for?
14:33:52Bagdersince the strings can be very long
14:34:00 Join elinenbe [0] (trilluser@pcp02254422pcs.wanarb01.mi.comcast.net)
14:34:02Bagderlonger than we can possibly set the virtual screen to
14:34:37 Join Bino_monk [0] (Bugger@vpooley.fsnet.co.uk)
14:34:43RipnetUKok, i see. But if we had a screen buffer say 4x or 8x the screen width, and only scrolled up to where was needed. Do we use a bitplane to store the graphic, or a byte per pixel? if its only a bit per pixel, surely we can afford the storage space?
14:35:00Bagdera bit per pixel
14:35:03RipnetUKmaybe we could get away with one scrolling line then
14:35:38Bagderyou're right, it might suffice with one line then
14:35:48Bagderit could be worth trying out anyway
14:36:05Bagderit would also make it very easy to make it scroll smoothly
14:36:18Bagderor on a sinus ;-)
14:36:39dwihnosine bouncing!
14:36:42RipnetUKhehe scrolling copper bars
14:38:21RipnetUKyou could even implement 'scroll bars' to allow the user to do the scrolling (i would hate that tho, but some may like it)
14:38:24dwihnoare there any free ia32 emulators other than bochs?
14:40:22RipnetUKi think you would have to copy from frame buffer to LCD in two steps, to handle the fixed region, and the scrolling rectangle
14:40:54Bagderit would depend on if the scrolling rect would be fixed on the LCD or on the "virtual screen"
14:41:23Bagderbut I guess it would be insanely weird to have it scroll on the virtual one
14:41:54 Join mistax [0] (lamamba_@213.66.147.135)
14:43:04RipnetUKalso the status bar wants to be fixed, as does the progress bar and the time/bitrate/samp freq stuff, in fact, only 4 lines need to scroll on the normal WPS
14:43:51Bagderall that taken into account, I think it is easier to just extend the current scroll funtionality
14:44:58RipnetUKhmmm... not sure I agree, maybe we should have 2 frame buffers, one normal size, and a scrollable one as well, and when it displays it plots the normal one, then overlays with a portion of the scrollable one?
14:45:07RipnetUKbut that would make the player harder
14:46:27Bino_monkall sounds very technical & sorry to interupt chaps...but is there any plans to add the function to scan through tracks before u get to 1.3? That with persistant settings would be the bees knees for my JB6000
14:46:27Bagderwell, if I make the scrolling smooth, and we enable four lines scrolling, then what would the visual difference be?
14:47:07ZagorBino_monk: yes, that will probably be added soon
14:47:20Bino_monkI see u added the remote support on the latest CVS... smart!
14:47:33RipnetUKnone, but i reckon implementing my way would be a lot easier, plus if we ever want graphics and stuff, we could just plot it on the virtual screen, and would all be taken care of, plus no worries with proportional/ different sized fonts
14:47:47Bagdertrue
14:48:09Bagderbasicly we want a huge bitmap to scroll
14:48:13RipnetUKguesss its swings & roundabouts
14:48:49Zagorthe problem I see is the same Bagder mentioned, with one line being much longer than the others. like the filepath on today's wps
14:48:59LinusNZagor: i think we should enable disk writes by default
14:49:28RipnetUKbut i tried it with independant scrolling for different lines, and IMHO it looks crap
14:49:42Bagdermaybe we should consider not showing the full path
14:50:02Bagderthe stock firmware doesn't
14:50:10Bino_monkthat'd look poo
14:50:17Bino_monksorry..
14:50:19ZagorLinusN: we need to somehow check that sector 61 isn't used
14:50:46LinusNis it possible to run partitionless on the player?
14:50:54Zagoryes
14:51:05Bagderwe need the storage on the recorder too
14:51:15ZagorBagder: yes
14:51:34LinusNSo where do we store the config when sector 61 is occupied?
14:52:01LinusNat the end of the FAT?
14:53:17LinusNAs long as the MBR contains a partition table we should be able to use sector 61
14:54:01RipnetUKcant we do a 'partial' file write, where you must have a file called .ArchosConfig in the root, of size (say) 4k, and it works out the position of that, and writes to that?
14:54:19RipnetUKthat would save creating new dir entries and so on, and allow it to be backed up
14:54:21mistaxLinus I was told to talk o u about the sond quality bug on the players
14:54:31LinusNok
14:54:43mistaxso let's do it..
14:54:54BagderRipnetUK: we don't have any file write code yet, that's why
14:54:57LinusNso you have distortion?
14:55:48mistaxwell yeah, I can play song louder and with more bass without distortion on the stock firmware than on the rockbox!
14:55:57 Quit matp ("Client Exiting")
14:56:07RipnetUKbut if we know what sector the fixed file is at, cant we use sector write code?
14:56:12RipnetUKbrb
14:56:39LinusNmistax: ok, an experiment: does the distortion go away when you raise the treble?
14:57:52mistaxone more brob! I can't get my player to recognize an upgrade so I'm sitting here with the original 4.09 mod!
14:58:14LinusNyou need to do an fsck/scandisk
14:58:21mistaxfsck?
14:58:34LinusNunix file system check
14:58:44mistaxok, scandisk it is
14:58:48LinusNyup
14:59:12LinusNand scandisk on USB1.1 is a real party! :-)
14:59:40Zagorwe can also simply refuse to support partitionless disks
14:59:53mistaxyuck! how long time? 6 gb on my 450 mhz with usb 1.1?
15:00
15:00:10LinusNZagor: yes we can, Archos probably does
15:00:31LinusNmistax: i don't know. i have never done it
15:00:54LinusNZagor: or just don't save settings on them
15:01:27Zagorthat was my thought, yes
15:01:58LinusNZagor: how to tell if the disk is partitionless?
15:01:58Bino_monkmistax: about an hour+, quicker if u have no files on the disk (obviously)
15:02:19Bino_monkfor a full disk surface scan
15:02:27datazone-workif its partitionless the firmware wont load, now will it
15:03:09LinusNi don't know
15:03:16ZagorLinusN: we can remember which sector we mounted the fs on
15:03:38LinusNZagor: can you implement such a function?
15:03:41Zagorif (rootsector != 0) save_config_buffer()
15:03:41Zagorsure
15:04:24LinusNi really want to have the save-to-disk option on by default
15:04:27datazone-workokay, call me crazy, but if there is no partition, then there is no filesystem, then how does the player load the rockbox firmware?
15:04:30 Nick mistax is now known as grummel|scandisk (lamamba_@213.66.147.135)
15:04:49Zagordatazone-work: there is a partition, just no partition *table*
15:04:55LinusNdatazone-work: the file system is placed directly on the disk
15:04:57Zagorthe partition starts at sector 0
15:05:11 Nick grummel|scandisk is now known as mistax|scandisk (lamamba_@213.66.147.135)
15:05:11LinusNas if it was only one partition
15:05:17Zagornormally, there's a partition table on partition 0 and the first partition starts at sector 64
15:05:25Zagoron sector 0, i mean
15:05:25LinusN63
15:05:32Zagoruh, yes
15:05:37datazone-workhiw would that occur?
15:05:40datazone-workhow
15:05:51ZagorMicrosoft...
15:06:04LinusNtry to format a blank removeable media with Win2K
15:06:16datazone-worki see
15:06:30LinusNit creates the filesystem on sector 0 with no partition table
15:06:33datazone-workforgive my ignorance then
15:06:34LinusNmakes sens to me
15:07:09LinusNi would really like to have a choice
15:07:18LinusNbut Winl00se doesn't give me that
15:07:40Hadakahmm
15:07:45datazone-worktoms root+boot+disk with usb support
15:08:10Zagordwihno: yo, font-slave. how's it going? ;*)
15:08:39HadakaI've done whole-disk filesystems on linux too, at times
15:09:00ZagorHadaka: yeah, but you had a choice :-)
15:09:01LinusNme too
15:09:13datazone-workso instead of sdb# you just put it on sdb?
15:09:15datazone-workcreepy
15:09:22LinusNindeed
15:09:29Hescreepy but works 8-)
15:09:33Hadakabut um, doesn't win2k have that volumemanagement program that can do partition tables?
15:09:42LinusNbut why waste space on a partition table when you only want one partition?
15:09:54datazone-workcause you got 6 gigs!
15:10:07LinusNHadaka: yes, but it doesn't manage removable volumes
15:10:10datazone-worki dont think you are going to miss a few ks
15:10:53LinusNi can only speak for the PC-card drivers
15:10:54 Join matp [0] (~mpritch@gateway.riversoft.com)
15:11:03ZagorLinusN: umm, because the spec says so? oh, sorry. forgot it was MS we're talking about.
15:11:19Hadakadatazone-work: yeah, except these are devfsd days ;)
15:11:28Hadakait doesn't manage removables? oh poop
15:11:39LinusNat least not pccard drives
15:12:35LinusNi had a hell of a problem with ATA pccards and win2k
15:13:12LinusNi had a hardware unit which only accepted partitioned ata-cards
15:13:22*Hes has charged his 1800 mAh battery set for 3 hours now, going at 5.57v and expecting the zero delta detection to hit any minute now
15:13:28LinusNand when a card was f'cked up i had to reformat it in w2k
15:13:47LinusNand w2k formatted it partitionless, unconditionally
15:13:58LinusNHes: take cover!
15:14:15Hadakahm
15:14:28Hadakapartitionmagic helps hopefully ;)
15:14:36LinusNit didn't
15:14:39Hadakaoh?
15:14:46Hadakathat is _so_ fucked up
15:14:54LinusNwhen it comes to mind, i think it's the pccard drivers that is ste problem
15:14:56Hadakaeverybody should just use linux
15:15:14*Bagder wears his Linux tshirt today
15:15:33HesDuring last LCD_WIDTH (112) minutes voltage has gone from 5.41 to 5.57v
15:15:35Bagderkernel source code looks good ;-)
15:15:46*LinusN is happy that Bagder wears a t-shirt
15:16:14HadakaI have my trusty Totoro T-shirt ;)
15:16:49*Zagor wears no shirt
15:18:04Bagderice cream time!
15:19:46 Join lohap [0] (lohap@213-48-172-125.bra.cvx.blueyonder.co.uk)
15:20:14lohapi know this is the wrong place but what is the rating of the charger for the jukebox recorder 9v or 12v cause ive lost mine lol
15:20:38LinusN9v 600mA
15:20:45LinusNbut you can use 12 too
15:20:59LinusNHadaka uses 12V 1A
15:21:13lohapthanks who manufactors them cause i found one that is that but i aint sure if its the correct one
15:21:39LinusN"kings electric co, ltd"
15:21:49lohapyep i got the right one lol
15:22:25Zagorif you're buying a new one, get a better model
15:22:38lohapno i found it lol
15:22:45Zagorok
15:22:49lohapunder my 2600
15:23:11lohapo btw i was wondering im i wanted to create icons for the jukebox recorder how would i go about it lol (newbie_
15:24:05Zagorcheck in apps/recorder/icons.c
15:24:31lohapthanks lol ill have to download source lol
15:24:38LinusNZagor: now enable save-to-disk!
15:25:11ZagorLinusN: there's still a lot of #ifdef HAVE_RTC #else save_to_disk() #endif
15:25:24LinusNand?
15:25:29Zagorrecorder needs to save to disk too
15:25:31Bagderalso, we should only save when we need to
15:25:44LinusNcome on, how long do you want to wait?
15:25:57LinusNi can see that you are recorder users
15:26:10Bagderso why don't the player users fix the code?
15:26:26Zagorwell sure, we can activate it first and fix i later
15:26:44Bagderis that really smart?
15:26:49LinusNBagder: the code doesn't need fixing for the player users
15:26:55Bagderit doesn't?
15:27:02LinusNno
15:27:08LinusNit works
15:27:23Bagderwhat about saving when spinning and only if it is really needed?
15:27:27LinusNat least the last time i tested it
15:27:37LinusNBagder: that is a feature, not a bug
15:27:38ZagorBagder: it only saves when exiting menus
15:27:53Bagderyes, but the settings may not be changed, right?
15:28:00Bagdermight not
15:28:14LinusN?
15:28:24Bagdergo to settings, leave
15:28:27Bagderit saves
15:28:30Bagderthe same
15:28:31LinusNyes
15:28:43Bagderthat's all I'm saying
15:28:55BagderI don't like that
15:29:08LinusNyou have a point, but nobody will notice it unless we enable the disk saves
15:29:15Bagderof course
15:29:21Bagderbut you want to enable it like this
15:29:23BagderI don't
15:29:27LinusNand the "player users" won't fix it
15:29:35Zagori'll add a checksum comparison and only save if it differs
15:29:47bobTHCi Find some .bdf fonts (i'm under Windows) but i cannot view it .... the package start from 4x6 to 10x20 somebody interested ???
15:30:11ZagorbobTHC: 4x6 sounds interesting
15:30:44LinusNi wonder how 'm' looks like in that font...
15:30:51Zagorhehe
15:31:46lohapbtw what is the anme of the format that the icons are stored in "binary images"
15:31:53Hes1800 mAh batteries 196 minutes from empty condition (hd fails), ended at 5.58v for long-term zero-delta detection
15:32:57ZagorHes: nice
15:33:03Bagderlohap: all bitmaps are stored like this: byte 0 is 8 pixels high, the lowest bit being at the top, byte 1 is the next column etc
15:33:30*matsl is away: I'm busy
15:33:43bobTHCi propos it because they include all caracterset : Latin, Greek, Cyrillic, Armenian, Georgian), all mathematical and phonetic alphabet symbols, all typographic punctuation, all box-drawing characters, control code pictures, graphical shapes and some more that you would expect
15:33:44*Zagor writes "AC/DC converter, 12V, 4A" on his shopping list
15:34:13ZagorbobTHC: nice, but it gets awfully big too...
15:34:20*matsl is back (gone 00:00:50)
15:34:30*Hes makes the charger disable charging at > 10v ext voltage
15:34:46Zagordidn't you use a 12V charger?
15:34:47bobTHCyes but we can make a special version of it...
15:34:57ZagorbobTHC: yes
15:35:21HesI'm using the stock charger... rockbox measures a bit over 8.7v at the external input
15:35:21LinusNHes: Hadaka will kill you for that
15:35:24matslIRC Q: How do you get this funny '*' comments
15:35:30ZagorHes: oh, ok
15:35:31HesLinusN: Oh?
15:35:43LinusNHes: the scale factor might not be correct
15:35:45HadakaHes: yes, I will
15:35:52HesLinusN: Right, I was suspecting that
15:36:02LinusNuse a multimeter
15:36:04Hadakamatsl: /me foo
15:36:06Zagorisn't the official Archos Travel Charger 12V?
15:36:11HadakaZagor: yes.
15:36:20Hadakathat's what I'm using.
15:36:24*matsl like this?
15:36:26HesI'm wondering if we're using too high charge currents
15:36:28Zagorthen I wasn't dreaming
15:36:45Hessince we don't trickle which the archos charger seems to be doing
15:37:05matslHadaka: Thanks. So now I can (try to) be funny too. Thanks!
15:37:16 Quit ironi__ (Read error: 110 (Connection timed out))
15:37:25ZagorHes: well we shouldn't trickle until they are full, right?
15:37:37HesHey, I want your comments on this one: should the battery meter on the WPS be either disabled, made to scroll, or the displayed value be made much lower during charging?
15:37:45*matsl Silly battery discussion. Of to drill som holes in a wall ;-)
15:38:02bobTHCfor download font : http://www.cl.cam.ac.uk/~mgk25/download/ucs-fonts.tar.gz
15:38:15HesSince it will show 'full' very early before the batteries are actually full
15:38:18LinusNHes: i think it should be disabled, or animate
15:38:22ZagorHes: during charging it can blink or whatever. archos animates it during charging.
15:38:32lohapis ther any apps out ther to convert a .bmp to a .c
15:38:36Hespeople might take a habit of disconnecting the charger too early
15:38:48Zagorwe could simply replace the battery icon with a wallplug icon
15:38:51HesI think so too.
15:39:01Bagderlohap: nope, not that I know of
15:39:48matpcan we have uk style plug?
15:39:49HesZagor: I am thinking that we might want to actually limit the effective charge current by trickling with a 50-80% duty cycle when charging with a high-capacity charger
15:40:05Hesquick charging is fast, but reduces capacity & battery life
15:40:14bobTHCloahp : http://www.ticalc.org/archives/files/fileinfo/163/16392.html
15:40:20Zagordoes it? not according to the battery makers.
15:40:48BagderbobTHC: non-rockbox program will most likely not generate compatible image C source
15:41:01HesI am wondering why archos trickles before the battery being full, and takes so long to charge
15:41:06LinusNZagor: since we don't have a temperature monitor we do risk wearing them out
15:41:14lohapthanls bobTHC
15:41:19lohapthanks*
15:41:24bobTHCok...
15:41:29HesWe take a larger risk by charging at full current
15:41:45LinusNHes: some manufacturers recommend "topping off" with ~2h trickle after a fast charge
15:42:26HesDo we have anyone with a computer-controlled voltage meter & an oscilloscope? I want to know what archos is doing.
15:42:40HesWhat their voltage curve looks like and what duty cycle are they using.
15:43:00lohapLinusN if a create a nokia stile menu with icons and a ui map can someone who knows c do the backend and then people can seee what they think
15:43:12HesWouldn't want to sit by the device for a day with a multimeter in hand 8-)
15:43:45LinusNlohap: i'm sure there are quite a few people on this list that can help you with that
15:44:31lohapgood good see im a graphic designer so i can come up with the icons just not the backend
15:44:47LinusNlohap: good
15:45:29lohapright of to workj to create icons btw in pixels how large is the recorder screen
15:45:52LinusN112x64
15:46:01lohapthansk
15:46:04lohapthanks
15:46:10lohapdamn my lappykeyboard
15:46:56LinusNZagor: do we have a nice, small proportional font yet?
15:49:03Zagorno
15:49:10Zagoronly dwihnos hard coded
15:51:15Hesusw-pr-cvs.sourceforge.net: Connection refused
15:51:37BagderHes: wait a while, it sometimes does that
15:53:06Bagder"stay a while, staaaaaay forever!"
15:53:21Bagder(c64 nostalgia)
15:53:36RipnetUKbadger - sounds like Slayer 'Dead Skin Mask' to me
15:53:51Bagder"mission impossible"
15:54:02Bagderone of the absolutely first speach syntesizers
15:54:09Bagdersounded awful
15:54:57*Zagor played it just a few days ago
15:58:15Bino_monkdoes firmware/rockbox have any control over battery charging with the players or is it hardware?
15:58:29HesIt's hardware
15:58:38Bino_monkk
15:59:16lohaphi i have create an icon can someone check that its ok to go on the recorder
15:59:21lohapie have a look at the bmp
16:00
16:00:12LinusNlohap: as long as it's monochrome and small, it should be no problem
16:00:44lohapwell this is a menu icon so its quite lage
16:00:46lohaplarge
16:01:19elinenbewhere is it?
16:01:23elinenbeI'll check it out.
16:01:49LinusNtime to go home, cu guys!
16:01:53Zagorbye
16:02:01 Part LinusN
16:02:02elinenbebye
16:02:05lohapright ill wack it on http
16:03:01lohaphttp://213.48.172.125/jb.bmp
16:03:40mbrHes, are you there?
16:04:34HesYup!
16:04:54mbrJust read your mail
16:05:04mbrAbout the animated battery
16:05:25mbrCurrently the power plug is blinking during charging
16:05:43HesIt wouldn't need to be animated, it just shouldn't show the capacity bar while charging
16:05:55Hesso that the user wouldn't think the battery is full an hour before it actually is
16:06:11Hes(even though the plug is blinking)
16:06:38mbrBut don't showing the battery is not so good in my opinion
16:06:48Zagormbr: why not?
16:07:00mbrAnimation like the stock firmware would be better
16:07:10Zagorand skipping the plig?
16:07:32Zagori.e. we don't need space for both the plug and the battery, since they are mutually useful
16:07:43HesLocalized power plugs would be c00l and useless. 8-]
16:07:54*Zagor smacks hes :)
16:08:20mbrZagor: not fully mutual exclusive at the moment
16:08:22HesProvide the plug icon in the language file (encoded into a character string 8-)
16:08:29Zagormbr: explain
16:08:41mbrplug appears when charger is pressent, blinks when charging
16:09:15mbrA visible icon for charger present was a feature request
16:09:57mbrOr do you think we shall remove the battery even if we are not charging?
16:10:08Zagorgood point
16:10:56HesYou have a point there.
16:10:56mbrexchange the battery with a solid plug when charger inserted and blinking plug when charging?
16:11:17 Part Bino_monk
16:11:23Zagorno, we still want to know battery status if not charging
16:11:51Hesright, especially when in 'discharge' mode
16:11:55Zagoryes
16:12:10mbrAnd why not animate the battery when charging?
16:12:23Hesanimating the battery would be fine
16:12:51***Saving seen data "./dancer.seen"
16:13:47HesI'll have to run to the second-hand guitar shop before it closes... seeya
16:14:01mbrbye
16:14:18Zagorbye
16:14:25BagderGUI-tar shop, that's where you buy interfaces for archive tools? ;-)
16:14:39Zagorhaha
16:17:03lohaplol is there any way to view a .c image on the recorder sim
16:17:36Zagornot without coding a viewer :-)
16:18:02lohapbar cause i created an icon and i want to view it on the archos screen
16:19:48 Part elinenbe
16:22:52bobTHCzagor u think it's possible to use a bmp to c converter ?
16:23:09Zagorno. 'c' is not an image format.
16:23:19Zagorevery c program wants a different format
16:23:32bobTHCi don't understand all...
16:23:37Bagderonly if the converter is very flexible
16:23:50bobTHChttp://www.simtel.net/pub/pd/41531.html
16:24:47bobTHCit's just say bmp to c/c++ sourcefile converte.. :|
16:24:55Bagderyes
16:25:00Bagderso then it might not work
16:25:37Bagderthey have to be the "correct" C
16:25:49RipnetUKdid anyone have any more thoughts about how to implement multi line scrolling? basically the discussion was a large bitmap, part of which is overlaid on the real display area, displaying a differnt 'window' of the large bitmap to scroll vs scrolling each line one by one similar the the one line we have now
16:27:29ZagorRipnetUK: i think a combination is probably best. i like the big screen buffer idea, but sometimes I think we'll want to keep one-line scrolling
16:29:17RipnetUKi agreed that we still need one line to be able to scroll a v. long piece of text (full path?) but I think the ID3 info should scroll as a block, not line by line, as i tried it line by line and it gave me headache!
16:29:48Zagori think so to
16:29:53Zagortoo
16:29:58RipnetUKi may have a go tonight to implement the big buffer... i think we would need duel sprintf commands, to write to the other buffer
16:30:18 Quit lohap ()
16:34:12 Nick mistax|scandisk is now known as mistax (lamamba_@213.66.147.135)
16:35:37Zagorhow do you mean the other buffer?
16:37:48bobTHCsomebody test the font?
16:38:18 Join ironi [0] (~ironi@80.88.116.93)
16:39:01RipnetUKthe large scrollable region buffer
16:39:51RipnetUKit will be kinda like a screen in its own right, only larger (wider) than the real screen. A selected region of that will be copied to the real buffer before the LCD update. By sliding that selected region left adn right, the entire block will scroll
16:40:13RipnetUKincluding any graphics etc that may be in that buffer
16:40:42RipnetUKit will be like a windows scrollbox, but automatically scrolling left and right
16:40:52Zagorwhy is that another screen? i thought that was *the* screen?
16:41:18RipnetUKbecause the main screen will contain some fixed lines, which dont scroll, like status bar, and progrwess bar
16:41:30Zagorhmm, right
16:41:39RipnetUKso we will say that (for example) lines 2-6 will be overwritten with the 'scrollbox'
16:42:37RipnetUKor a region thereof
16:42:45Zagorright
16:44:05RipnetUKdoes that sound OK to you?
16:44:32Zagori think so :-)
16:44:43Zagorit just needs to mature in my mind a bit
16:44:54 Join elinenbe [0] (trilluser@pcp02254422pcs.wanarb01.mi.comcast.net)
16:44:59Bagderif the scrollbox is like 2x or 3x the width, I sine wave might be neat for the back and forth movement
16:45:07elinenbewoohoo!!!! ROCKBOX!
16:45:11BagderI mean, it might be
16:45:16ZagorBagder: you and your sine waves... :-)
16:45:23elinenbeBadger: agreed. That would give a nice flow.
16:45:35Zagorhas anyone experimented how fast we can update it anyway?
16:45:43Bagderas the stock firmware's back-and-forth is pretty annoying at thends
16:46:10elinenbeBadger: agreed. I did a demo with a 3D dog and we used sine waves for the head movement and such, and it came out awesome
16:46:10Zagorwe'll get bug reports for motion sickness ;)
16:46:19BagderZagor: linus measured the lcd_update() to 20 ms when moved to iram
16:46:42Bagderat least I think he said so
16:46:57Zagorwow, that's faster than I thought
16:47:05Zagor50 fps! ;)
16:47:14Bagderyes, it moved from 36 => 20
16:47:26Bagderor something like that
16:47:48elinenbeat the finish of the sokoban levels you can see a quick lcd_inverse lcd_update of the whole screen and it almost gives a nice greyscale
16:48:02elinenbenow if you were to just invert certain parts of the screen.
16:49:56ZagorBagder: make bounce smoother!
16:50:23RipnetUKbug reports for motion sickness <−−- LOL
16:50:39BagderZagor: it uses a rather tiny sine table, and it moves too fast in it
16:51:07RipnetUKwe could make it a user definable function - x^2 for REAL motion sickness potential :)
16:51:13Zagoralso, you need to get rid of the font dependency. the fixed wonts go out the window soon.
16:51:40elinenbeso, soon there will be no more hard coded fonts?
16:51:43elinenbenice :-)
16:51:44RipnetUKi believe that scrolling window would make font independance work easier, as you could just blindly print, rather than working out the width
16:51:57elinenbeI like my handwritten bdf font.
16:52:20elinenbewell, you still need to work out the height for the treeview, menus, and such
16:52:34RipnetUKyeah, and the one long scrolling line
16:52:59elinenbecurrently they are a little messed. I think we should target a few bdf fonts and see how they look
16:53:03RipnetUKalthough maybe we could do away with that, and have 2 scrollbox regions, one which is v. long and 1 line high, and 1 for the ID3 info which is shorter
16:53:39RipnetUKaaarrrghhhh i miss object orientation... (but of course i appriciate why we do not use c++ here :) )
16:55:33 Join Bino_monk [0] (Bugger@vpooley.fsnet.co.uk)
16:56:45Zagorelinenbe: the current bdf code doesn't draw them 100% correct
16:56:56 Quit Bino_monk (Client Quit)
17:00
17:02:00 Join Bino_monk [0] (Bugger@vpooley.fsnet.co.uk)
17:07:01RipnetUKanother advantage of doing the scrolling with a window is that is makes the marker easier to keep in one place at the left of the treeview
17:09:12BagderZagor: new bounce is smoother ;-)
17:09:28Bagdercan't really remove the font depency as the code is now
17:09:34Zagoroooo
17:09:41Bagderbecause the font output functions don't do plain OR
17:10:41Bagderso I need to use lcd_bitmap() for each letter
17:10:48Zagorok
17:11:57ZagorBagder: i received top-row player-sim icons from matsl. feel like adding them?
17:12:17Zagori'm pretty busy myself (settings on-disk & font code)
17:12:29Bagdermatsl has commit access himself now, I'll help him add them himself
17:13:11Zagorok. i didn't know matsl was versed in X11 programming :)
17:13:25Bagderit isn't x11 programming
17:13:34Bagderremember that the player sim uses the recorder lcd functions
17:13:36Zagorah, lcd_bitmap
17:13:38Zagorright
17:13:42Bagderyeps
17:14:42BagderI'll go home, see ya in a while
17:14:45 Part Bagder
17:17:38matperr, i just built the current cvs and the screen blanks out!!!
17:18:02mbrmatp: Recorder?
17:18:06matpyes
17:18:12matpwhat gives?
17:18:32mbrLinus added contrast setting. the contrast is too low
17:19:03matpso I up the default in settings.c ?
17:19:13mbrF1, down, right, 4*down, right, lots of ups
17:19:59matpmagic
17:20:19Zagoror pull the batteries to clear the RTC RAM
17:20:33mbrand loos all other settings :)
17:20:41matpdrastic
17:23:08RipnetUKsurely there should be a min/max on the contrast?
17:23:28Zagoryeah, and a higher default value
17:26:18RipnetUKdammit i had to set the clock again - not easy with prop. fonts enabled, as the setup screen screws up
17:26:57mbrI'll look into that
17:28:08Zagorgood
17:28:14matpcan I change font yet then? how ?
17:29:50Zagor../tools/configure
17:29:58Zagoronly on recorders, of course
17:30:26Zagorif you choose "no" to loadable fonts, you can select "proportional". that's the easiest option for now.
17:32:11matpi seem to be in a loop during make
17:32:38matpill try make clean
17:32:57matphmm, that's looping too
17:34:15RipnetUKmatp?is your clock correct?
17:34:28RipnetUKi found that having my RTC set to yesterday did that
17:34:58RipnetUKits VMWare (udner which my Linux install runs) - it seems to screw up the RTC time, so now I reset it from a time server as part of my build script
17:35:38matpno, i'm nntp'd
17:38:32RipnetUKare u in the same time zone as the contributors?
17:39:16RipnetUKi am in GMT and it seems to work OK (actually about 10 miles from Greenwich itself)
17:40:09matpi'm in gmt too
17:40:17matprichmond
17:40:18notchanyone tried the latest buld on the recorder?
17:40:26notchbuild
17:40:37RipnetUKhmmm... recorder (with prop fonts) builds fine on mine - havent tested the object file yet tho
17:40:45notchScreen blanks on startup and doesnt recover
17:40:50matphe
17:40:55RipnetUKnotch - thats the contrast
17:41:06Zagori'll fix the default contrast
17:41:14notchokay I cannot see to ajust it !
17:41:23mbrnotch
17:41:31HesWhy was the config revision increased?
17:41:32mbrnotch:F1, down, right, 4*down, right, lots of ups
17:41:32matpF1, down, right, 4*down, right, lots of ups
17:41:46matp:-)
17:42:28ZagorHes: "to force default values for the LCD contrast setting /Linus"
17:42:31notchwhoa! you have to get to 40 or something on my recorder!
17:42:45Zagormine too. i think 40 should be default
17:43:01notchyep.
17:43:06RipnetUK40 is a bit dark, but readable on mine
17:43:12RipnetUK36 is about right...
17:43:13matpmore like 26 on mine
17:43:24HesI wonder why that was required, the same logic should have worked as for the other stuff
17:43:25Zagorok so 30 is a reasonable compromise
17:43:26RipnetUK26 is a bit sun-bleached on mine
17:43:34Hesall of the unused RTC space is set to 0xff by default
17:43:37*notch notch thinks the animated battery is cool
17:43:40ZagorHes: i agree
17:43:57Hesso the code could check if the location was 0xff and keep the default value if it was
17:44:09Hesand use a default value which can be seen
17:44:19RipnetUKor just limit it from 25-50 or some sensible range
17:44:42matpmy battery icon moves too much when the disk spins
17:44:47Hesincrementing the config revision will trash all the other config too
17:44:55Zagorumm, he added it in another place. now we store it at 0x12 and 0x0a...
17:45:05Hesuh?
17:45:19notchOcasionally I get 'end of list press on to continue' when using playlists, but the unit recovers... known bug?
17:45:34mbrhave to go, bye
17:45:34*matsl back again from drilling holes but still finding battery discussion ;-)
17:45:46Zagorbye mbr
17:45:53Hesmbr: thanks for the animation
17:45:54bobTHCbye
17:46:02 Nick mbr is now known as mbr|gone (~tiw4mabr@rhlx01.fht-esslingen.de)
17:46:05RipnetUKnotch - i get that too
17:46:06mbr|goneHes: np
17:46:23matpnotch: i used to get this a lot when I was walking with the unit.
17:46:39RipnetUKhavent been able to work out a definate pattern, but the unit is sitting still on a desk
17:46:54matpbad mp3?
17:47:11RipnetUKcould be... will look out for that next time it happens
17:47:19RipnetUKbut it happens while the music is still playing
17:47:44Zagornotch: i think it's read errors. we don't retry yet.
17:48:02RipnetUKbut surely that would kill the music playing?
17:48:24notchHDD read errors? My HDD may be on the way out?
17:48:47Zagornah, read errors are normal. only if retry fails is it a sector error.
17:49:02RipnetUKnotch - good excuse to chuck a 60gig in there :)
17:49:20notchI dont have that many mp3's!
17:49:36RipnetUKnor do I but I have easily outdone a 20
17:50:03RipnetUKbut im worried about other drives using too much power
17:50:25notchI need to buy some more CD's (or a cable modem)!
17:50:37RipnetUKi went for the broadband option ;-)
17:50:53Zagor20/40 is not much difference. the 60-gigers i've seen, though, use much more power.
17:50:55notchIt should state the max power consumption on the case, but average is anyones guess...
17:51:04Hesusw-pr-cvs.sourceforge.net: Connection refused
17:51:05Hesgrr.
17:51:11RipnetUKi thought the normal fujitusu drives were special low power?
17:51:13*Hes wants to commit.
17:51:15bobTHCmy toshiba 30 gig use less that the original
17:51:20Zagornotch: don't read the case, read spec sheets on their web sites
17:51:40RipnetUKi would have thought that one with a larger cache would be an advantage?
17:51:56*notch relies on linus for data
17:51:58bobTHCand he had 2 Mb of cache !
17:52:00ZagorRipnetUK: no, they are standard laptop drives
17:52:08RipnetUKhow easy is it to swap the drives? i heard you had to unsolder/cut metal tabs
17:52:29bobTHCthe original only have 512kb !!!
17:52:30RipnetUKi thought Archos selected them because they were lower power than normal? maybe i was wrong
17:52:32ZagorRipnetUK: no, it's easy. no soldering.
17:52:59ZagorRipnetUK: http://rockbox.haxx.se/mods/disassemble.html
17:53:05RipnetUKcool... i knew it was no soldering to actually swap the drive, but I thought you had to cut/unsolder the metal underneeth the battaries
17:53:24RipnetUKso bob - do you still get the same battary life?
17:53:29notchdoes a bigger buffer help us if we are constantly powering up/down?
17:53:54notchHDD buffer that is....
17:53:55Hesnotch: we don't power the drive off, just tell it to sleep
17:54:17 Quit mistax ()
17:54:25notchHes: I thought linus decided to turn it off....
17:54:27bobTHCsorry i say toshiba but i have a IBM travelstar ! (end of working day very hard!)
17:54:51 Join Bagder [0] (~daniel@as3-3-2.ras.s.bonet.se)
17:55:36bobTHCthe travelstar need less power, have more cache and is cheap !
17:55:57bobTHCthe 3 gig for 150 Euro
17:56:01bobTHC30gig
17:56:26notchanyone have any opinions on fugitsu?
17:56:55bobTHCthe original is fujitsu
17:57:34notchmmn should have guessed - they are the cheapest here.
17:58:40RipnetUKi wouldnt bother with 30 i would go straight to 40
17:58:46RipnetUKanyway, gotta travel home now
17:58:50RipnetUKcatch ya later
17:59:19Bagdersee ya RipnetUK
17:59:20bobTHCwait a little and toshiba come out a 60gig 9.5mm
17:59:39RipnetUKbob - but would be V expensive
17:59:59RipnetUKbut it should drop the price of the 40 a bit
18:00
18:00:01bobTHCof course when it's new it's expensive!
18:00:15RipnetUKseeya
18:01:41bobTHCtoshiba said more than 100gig before next spring !!
18:02:42BagderZagor: freshmeat posting please
18:03:11*bobTHC go home
18:03:15bobTHCbye all
18:03:22Bagderbye bobTHC
18:03:49 Part bobTHC
18:04:03ZagorBagder: aj
18:04:04Zagorah
18:04:04matpis scrolling broken?
18:04:57Zagoris it?
18:05:32matpProportional font. Manic_Street_Preachers does not scroll in the dir browser
18:06:17matpBut "And You Will Know Us By The Trail Of Dead" does scroll
18:06:33Bagderbad width check probably
18:06:41matpsounds like it
18:07:26Zagoryes, the width check in scroller bugs slightly with propfonts
18:07:38Zagori've meant to fix it for weeks :-)
18:07:43 Quit pyvasene ("using sirc version 2.211+KSIRC/1.2.1")
18:08:59matpThe fonts look nice though :)
18:09:07matpmore sexy
18:09:16Bagderhttp://forums.dreamincode.net/index.php?act=ST&f=20&t=3710&s=23d42eddec5e44356552774da0fea408
18:09:18Zagori hate that freshmeat want the changes in "Complete sentences in 3rd person english"
18:09:35Zagorhow the heck to you translate a fix list into prose? looks ridiculous
18:09:36BagderZagor: they always rewrite whatever you type in there anyway ;-)
18:09:48Zagoryeah, i'll just paste the list and let them do whatever they want
18:10:07BagderZagor: I always end up with too much text and then the interface complaints ;-/
18:10:26Zagorit's, um, "suboptimal"
18:12:11Zagor"Your changes summary is 1221 characters long. This exceeds the fixed limit of 600 characters."
18:12:20Bagdersee what I mean? ;-)
18:12:32Zagorwell, we did a lot of changes. whaddayagonnadoaboudid?
18:12:55***Saving seen data "./dancer.seen"
18:13:00Bagdercut of all mentioned bug fixes to start with ;-)
18:13:03Bagdercut off
18:14:26*Hes bangs his head at the keyboard after realizing that export CVS_RSH=ssh does not stick
18:14:54Bagderput it in your .profile/.login/.tcshrc/[fill in]
18:15:05HesI know i know i know
18:15:08Hesjust forgot it.
18:15:37matpi create a dir called "A". linux displays it as "a", but rockbox displays it as "A". why???????
18:16:01elinenbeHes: how can you make that stick. I have to type that in every time I open up a shell...
18:16:06HesDifferences in FAT/VFAT/FAT32 implementations
18:16:29Heselinenbe: put the commands in the startup script of your shell (some common samples listed above by Badger)
18:17:05notchanyone know how to do that in win32 (cygwin)
18:17:15 Join Electrocut [0] (Electrocut@rennes-1-a7-62-147-99-223.dial.proxad.net)
18:17:24notchhey tanguy!
18:17:25Hesthe VFAT/FAT32 filesystems contain two copies of the file names, a 8+3 one (filename.ext) for compatibility with older systems not supporting long filenames
18:17:28Zagormatp: names shorter than 8 chars are stored as "shortnames", always uppercase. We don't change the case of names, so it's displayed as 'A'. Linux is the one modifying the information, not us. (And windows is a lot worse)
18:17:37Bagdercygwin runs bash, doesn't it? .bashrc
18:17:52Hes.bashrc in the home directory (whatever directory your shell starts in)
18:18:20matpsilly linux
18:19:00Zagori'm surprised you get it to show 'a'. I've been struggling to get my "A-HA" dir to be called "A-ha" without success.
18:19:31 Quit Electrocut (Client Quit)
18:19:58notchI cant find a file called .bashrc - should I create one?
18:20:00matpREM is bad as well :)
18:20:05Bagdernotch: yes
18:21:13*notch hides in shame from his lack of linux knowhow...
18:21:29Bagderbut you learn
18:21:40matslZagor: remember the '#### ' (progress bar!?)-thing I experimented with the parameter settings?
18:22:12Zagoryup
18:22:40matslZagor: I kind of liked it ;-) You had some similar idea?
18:22:56Zagordid I? :-)
18:23:07Zagori think it's fine, I just want numbers too
18:23:34elinenbeHes: can't get it to work with cygwin here. whatever... no big deal.
18:23:57matslIn my code the numbers are on the same line as the parameter name.
18:24:18matslThe progress bar alone on line 2.
18:24:42*Zagor added persistent settings to the simulator :-)
18:24:52Bagderzagor rocks!
18:25:00Zagormatsl: is there enough room on the first line?
18:25:08HesDoes cygwin have truss/strace? 8-)
18:25:28Bagdermatsl: btw, I happily pass the "put icons on the player simulator display" job over to you ;-)
18:25:40matslZagor: Almost. Problem with negative dbs and treble.
18:26:01Hesif it has, you can do truss or strace bash 2>&1 | grep open (or something along those lines) to see what files it opesn
18:26:02Hesopens
18:26:08matslBagder: I have happily forgotten about that job ;-)
18:26:14Bagder*g*
18:27:36matslZagor: With treble you get "Treble-15DB"
18:27:39 Join mistax [0] (lamamba_@h135n1fls34o883.telia.com)
18:27:43Bagderwhen I get this script of mine working, I'll cheer like crazy
18:28:55matslBagder: The icons job looks like a low priority thing I could look into. yes.
18:29:28Bagdermatsl: it is basicly a matter of converting the image into bitmaps and then use lcd_bitmap() to put them on screen
18:29:54Zagormatsl: I think I prefer numbers to a progress bar, personally
18:30:15Zagormaybe centering it on the line for prettiness :)
18:30:17matslBagder: Seems feasable. maybe not much time for rockbox the next few days though.
18:30:45Bagderwell, this is rather low prio, and I don't have much time myself
18:31:47matslZagor: Numbers are fine, thats why I weant to keep them. Volume is OK. %-is known to be 0-100. With DBs though you don't know where you are on the scale.
18:34:19matslZagor: Think about it anyway. Must do some other thing now.
18:34:33*matsl getting into must-cook-food-or-die-mode
18:34:51Zagor:)
18:35:21Zagor"cook or die", wasn't that an old game? ;)
18:35:42Bagderskate or die!
18:35:54Zagoryeah, i know
18:40:42elinenbecheck out the icons and UI of the RioVolt/iriver 250 mp3CD player. That is a VERY nice UI
18:44:05*notch gone to feed diet coke habbit
18:44:10 Quit notch ()
18:46:07Zagor250? I only find 100, 150 and 350 on Iriver's web site.
18:47:46Zagorah, found it on the "international" site.
18:47:57*Zagor hates it when companies hide information from their customers
18:49:07Zagortoo small pictures to see anything, though
18:50:46BagderWelcome bagder34@netscape.net
18:50:51BagderYou are currently using 0.00 MB (of your 5.00 MB Limit)
18:50:56Bagder*FINALLY*
18:51:00 Join zaknafein [0] (zaknafein@faerun.ugr.es)
18:51:33elinenbeZagor: you need to download the manual pdg
18:51:35elinenbe.pdf
18:51:39elinenbeI give you the link
18:51:43Zagorok
18:51:59*Bagder is *so* happy
18:52:50elinenbehttp://down.iriver.com/iriveramerica/support/manuals/imp350e-v200_manual.pdf
18:52:53elinenbethere you go.
18:54:34ZagorBagder: ?
18:54:45Zagorah, ok
18:54:53BagderZagor: I'm logged in to netscape webmail
18:55:01Bagderit takes 17 curl invokes
18:55:12Zagorhaha
18:55:21Zagorthat's insane
18:55:22Bagderhorrible
18:55:55elinenbecheck out the modes on page 14 of the manual −− that is the supported playmodes.
18:56:32elinenbeand then pages 19 onwards show the UI
18:56:35elinenbeit is quite nice.
18:57:37Zagorthat thing has a very nice lcd!
18:57:48elinenbeand, that is a nice manual. 40+ pages is not too shabby
18:58:06Zagortoo bad they waste it all on icons and leave two (2) whole lines for showing files :-)
18:58:46Zagoroh, ok. that's their wps
18:58:58*Bagder likes the term WPS
18:59:05Bagdersounds very weird ;-)
18:59:09HesDo you know what Linus's logic was with the contrast changes?
18:59:14 Nick dwihno is now known as dwihno|party (dwihno@Bald067.Baldakinen.Umea.SE)
18:59:20ZagorHes: no idea. i'm fixing it now.
18:59:22HesOk.
18:59:27dwihno|party<−− is gooone fishing
18:59:59HesI was just about to fix it... i guess the added 0x12 location is unnecessary
19:00
19:00:08Zagoryes
19:00:20Zagori'm reverting back to version 0 too. works just fine.
19:00:34HesYeah
19:00:37Hes#define MAX_CONTRAST_SETTING31
19:00:37Hes#define DEFAULT_CONTRAST_SETTING 30
19:00:56HesMaybe have a #define MIN_CONTRAST_SETTING 5 or so
19:01:19Hesand make the settings loading code use the default if the configured value is < MIN_CONTRAST_SETTING
19:01:34BagderZagor: "cp -r manual manual-1.2" ? ;-)
19:01:34Hesin case the byte is set to 0 by older code versions
19:01:44ZagorBagder: ok
19:01:59Zagordone
19:02:26BagderZagor: we should somehow allow people to view either one
19:02:31Zagoryes
19:02:46Zagorwe link to the new from the daily page
19:02:56Bagdergood idea
19:03:19 Quit zaknafein ()
19:04:11Zagorfixed
19:04:38Bagdergoodie, so now all UI-fiddlers can update the manual too
19:04:42Zagoryup
19:08:18BagderRyan Koski is right, we need a playlist browser
19:09:04Zagoryes. but it's not the most needed feature...
19:09:12Bagderno
19:19:00Zagorfor some strange reason, the LED doesn't turn on during writes
19:19:16Zagorah, now I know
19:22:35Bagdernice sim fix for the sector stuff
19:22:48Zagorthanks
19:23:01Zagorseems win32sim doesn't include stub.c though
19:23:12Bagderneed to fix the win32/Makefile I bet
19:23:28Zagoryup. but it's supposed to include it, that's why it's in common/...
19:23:44Bagderyes
19:23:55Zagorhow often is the last5 script ran?
19:24:53Bagderevery 20 mins
19:24:57Bagder10,30,50
19:25:26Zagorah. might be nice to do it 01,21,41 to be in sync with the compile status table
19:25:40Bagdertrue
19:25:42BagderI fix
19:26:05Bagderdone
19:26:19Zagorgood
19:30:45HesZagor:
19:30:46Hes+ if ( global_settings.contrast < MIN_CONTRAST_SETTING )
19:30:46Hes+ global_settings.contrast < DEFAULT_CONTRAST_SETTING;
19:30:46DBUGEnqueued KICK Hes
19:30:46Hes+ }
19:30:53Zagoryeah, fixed that
19:33:00Hes+ DEBUGF( "save_config_buffer: rtc_write failed at addr 0x%02x:
19:33:00Hes%d\n", 14, r );
19:33:08Zagoroops :)
19:33:11Hes14 ?
19:33:30Hesprobably 0x14 + i or so 8-)
19:33:40Zagoryes
19:34:16Zagori broke the rtc save...
19:34:24HesI noticed. 8-)
19:34:42HesHey, I think we should write to disk on recorders too.
19:34:48HesAnd load from disk if rtc read fails.
19:35:17Zagoryes. but I wanted to to it gradually
19:35:18HesWrite to RTC immediately after changes and write to disk when it's comfortable
19:35:25HesOk
19:36:55HesAnd I'm not sure if cksum is good for checking if the config changed
19:37:09Hesit might be easy to change the config in a way that the cksum does not change
19:37:32Zagorhmm, right
19:37:40RipnetUKwhy does the UI sim not include the debug menu? it is just because its no use?
19:37:47HesIt's a very simple xor anyway, not cryptographically strong hash 8-)
19:37:57 Quit Bino_monk ()
19:38:23ZagorRipnetUK: precisely. the debug menu shows mostly hardware things that the sim doesn't have
19:38:46HesIt was intended to detect if the stock firmware (or something) had corrupted the rtc
19:39:46RipnetUKk
19:41:25 Part elinenbe
19:41:32RipnetUKsorry to be a C lamer, but does __attribute__ ((section (".idata"))); mean to store this variable or whatever in the data segment?
19:41:46BagderRipnetUK: it means the data goes to the idata memory
19:42:02Bagderthe idata area is defined in the link control file
19:42:08Bagder... and happens to be internal ram
19:42:57RipnetUKi see, thanks, i do Delphi normally, and we just have a simple project file... how are we doing for internal RAM? im trying to implement the scrollbox stuff - can I use a bit more Internal ram for the buffer?
19:43:26BagderRipnetUK: we only have 4K internal ram, it can only be used for very special things
19:43:55Zagor*doh* I added stubs.c in two places :-(
19:44:56RipnetUKok, i will try and be economical with it :)
19:45:25 Nick Zagor is now known as Zagor|out (bjst@as9-5-6.k.s.bonet.se)
20:00
20:07:12 Join lithermon [0] (gxhb@modem-792.clefairy.dialup.pol.co.uk)
20:12:57***Saving seen data "./dancer.seen"
20:35:56 Quit lithermon ()
21:00
21:28:17RipnetUKdammit - my limited c is getting to me again... can anyone help me out with a bit of beginners c please?
21:43:00Bagdersure
21:44:37RipnetUKthanks - im trying to call a function i defined in lcd.h from lcd-x11.c but I keep getting undefined reference to `draw_all_scrollboxes'
21:44:49RipnetUKeven though i declared the function prototype in lcd.h
21:45:18RipnetUKi have downloaded a c tutorial to my palmtop for reading tomorow, so I am RTFMing but im dying to try my ideas tonight :)
21:45:21Bagderundefined reference means that the function does exist when it links
21:45:26Bagderdoesn't
21:45:44RipnetUKbut surely lcd.c is used in the uisim?
21:45:52Bagderyes it is
21:46:12Bagderand your function draw_all_scrollboxes is in it?
21:46:22RipnetUKi prototype it at top of lcd.c, so im not sure why it can't be 'seen' from lcd-x11.c
21:46:36Bagderthe prototype doesn't matter to the linker anyway
21:46:42Bagderthe prototype is for the compiler
21:47:08RipnetUKyes
21:48:39RipnetUKit compiled ok when I remove the call in lcd-x11.c,
21:50:06Bagderweird
21:53:20RipnetUKi can call it OK from within lcd.c - i have proved this by changing the call to a deliberate error, and re-compileing, to proove it is in fact getting compiled, but if i call it from the other c file, it cant link
21:54:56Bagderyou don't happen to declare the funtion "static" ?
21:55:20RipnetUKyes, is that naughty :)
21:55:36Bagderno, not if you only use it within the same file
21:55:38Bagder;-)
21:55:56Bagderbut if you want to reach it from other files, you can't have it static
21:55:59RipnetUKah ha, thanks i think i am there now... that would be it... id better get reading my c tutorial tonight...
21:56:05RipnetUKmany thanks
21:56:17Bagderno problemos
21:56:31RipnetUKmany thanks
21:56:43RipnetUKsorry, that up arrow/enter was meant for my xterm :)
21:56:57RipnetUKand as if by magic... thanks again it works now
21:56:57Bagder"too many xterms error" ;-)
21:57:23CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
21:57:23*Bagder walks off and takes a shower
21:57:37RipnetUKwoohoo and it even did what I expected
22:00
22:03:51Bagderneato
22:12:58***Saving seen data "./dancer.seen"
22:18:53 Nick Zagor|out is now known as Zagor (bjst@as9-5-6.k.s.bonet.se)
22:23:38 Join datazone_ [0] ([RCMXTY3UA@207.136.36.203)
22:25:50BagderZagor: what do you think about seeking in the mpeg thread ?
22:26:07Bagderbeyond the id3v2 tag
22:26:59Zagori think we have no choice: we must do it
22:27:18Bagderyeps
22:27:38Zagorthe id3 reader should mark where StartOfData is, and then mpeg.c seeks there before starting to read
22:27:47Bagderit pretty much do that
22:27:55Bagderas it sets the size of the id3v2 header
22:27:57Zagorin the mp3info struct?
22:28:07Bagderyes
22:28:10Zagorok
22:30:11Zagori'll fix
22:30:16Bagdercoolio
22:30:36Bagderwe should use one of those giant-bloated-bmp-tag mp3s to test with
22:30:49Bagderor possibly generate even worse examples
22:30:49Zagoryup
22:31:42 Part Bagder
22:33:15Zagori'm downloading the one posted on the list
22:33:19 Quit datazone (Read error: 110 (Connection timed out))
22:43:40 Join elinenbe [0] (trilluser@pcp02254422pcs.wanarb01.mi.comcast.net)
22:45:19 Nick datazone_ is now known as datazone ([RCMXTY3UA@207.136.36.203)
23:00
23:14:39RipnetUKBagder - i have now read a c faq and am well aware of what static does :) thanks for yer help
23:37:12 Join aku [0] (zohosa@dhcp024-208-181-223.columbus.rr.com)
23:45:58RipnetUKdid someone say we avoided dynamic memory allocation?
23:46:08Zagoryes, very much
23:46:32RipnetUKwhy?
23:46:58Zagori'm writing a lengthy explenation in an email. basically, it's useless in an embedded system
23:47:06RipnetUKok, thanks :) will take your word for it
23:47:28RipnetUKi guess the overhead of a memory management system is too much to deal with?
23:50:34Zagorno, that's not it. the reason is: there's no point. dynamic memory does not give us more memory, since we have no disk paging. it only gives us messier bugs.
23:51:46RipnetUKok, thanks

Previous day | Next day