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 2005-10-24

00:00:00preglowi'm just wondering if the flash can withstand developing
00:00:07preglowi tend to do a lot of writing when coding rockbox
00:00:10preglowbut i guess it can
00:02:13 Quit pabs (Remote closed the connection)
00:02:25amiconnThe SanDisk TriFlash has an endurance specification for each sector of 100,000 writes (reading a logical sector is
00:02:26amiconnunlimited) under typical conditions.
00:02:46amiconnquote from the datasheet...
00:04:28linuxstbThe Nano's Flash chip:
00:06:40linuxstb"Endurance : 100K Program/Erase Cycles" - so the same as the TriFlash
00:08:29 Join pabs [0] (
00:11:21amiconnlinuxstb: 100000 cycles seems to be common for modern flash devices. The flash ROM chips in the archos and the iriver have the same specification. So we don't need to worry about wearing out the flash when developing rockbox-in-ROM for iriver :)
00:13:35 Quit pabs (Remote closed the connection)
00:15:36amiconnTiMiD: I just verified the x11 simulator remote & main displays both on cygwin and on debian linux. I can't find a missing line...
00:17:04***Saving seen data "./dancer.seen"
00:22:47 Quit _DangerousDan ("Miranda IM! Smaller, Faster, Easier.")
00:32:51 Join XShocK [0] (
00:33:26 Join pabs [0] (
00:37:44 Quit tvelocity ("Leaving")
00:38:30 Quit pabs (Remote closed the connection)
00:39:43 Quit cYmen ("zZz")
00:44:31 Join XavierGr_ [0] (
00:44:35 Nick XavierGr_ is now known as XavierGr (
00:44:46 Join pabs [0] (
00:49:50 Quit pabs (Remote closed the connection)
00:50:28 Join kwaku [0] (i=kwaku@
00:50:49kwakui have a problem..
00:51:07 Join pabs [0] (
00:58:26 Quit kwaku ()
01:17:14 Quit webguest23 ("CGI:IRC (EOF)")
01:20:07 Join pabs_ [0] (
01:24:19 Quit Moos ("Glory to Rockbox")
01:31:29 Quit pabs (Read error: 110 (Connection timed out))
01:43:00 Quit matsl (Remote closed the connection)
01:56:18 Quit einhirn ("Miranda IM! Smaller, Faster, Easier.")
02:01:11 Join gromit` [0] (n=gromit`
02:03:34 Nick pabs_ is now known as pabs (
02:04:28 Join {Mike} [0] (
02:04:33{Mike}MY iriver is screwed :(
02:07:37{Mike}dunno how
02:08:00{Mike}on startup the bootloader says "checksum" and "sum", which match, result 0, and hangs
02:08:20{Mike}the iriver firmware won't load past the iriver logo
02:10:31{Mike}is it just bricked?
02:11:34preglowdoes it attempt to start the disk?
02:11:38preglowdoes the red light ever blink or something?
02:12:33{Mike}the red light is static on when the iriver firmware is on
02:12:48{Mike}trying to load
02:12:59preglowpossible the disk is dead
02:13:01linuxstbDoes the bootloader enter USB mode?
02:13:20{Mike}windows doesn't like detecting it
02:13:23{Mike}I'll try my linux box
02:13:39{Mike}I'll tell you the full message it say
02:14:08{Mike}Rockbox boot loader
02:14:11{Mike}Version 5
02:14:27preglowwhat happens if you start it with the usb cable inserted?
02:14:28{Mike}Batt 3.95V
02:14:35preglowversion 5 supports usb in the bootloader
02:14:46{Mike}then a pause, the right light blinks, then this stuff:
02:14:51{Mike}Loading firmware
02:15:01{Mike}Length: 427F4
02:15:08{Mike}Checksum: 12EE460
02:15:14{Mike}Model name :h120
02:15:18{Mike}Sum: 12EE460
02:15:23{Mike}Result: 0
02:15:26{Mike}then it stops
02:15:42{Mike}can't turn it off, can't boot to iriver firmware without resetting
02:16:07{Mike}if I reset and goto into usb mode, I can mount it
02:16:53linuxstb"Result: 0" means that it's loaded the firmware, checked the calculated checksum against the checksum stored in the rockbox.iriver and they match.
02:17:05***Saving seen data "./dancer.seen"
02:17:37linuxstbThe next thing the bootloader does is run the loaded firmware.
02:17:48{Mike}or not
02:17:51{Mike}as the case may be :(
02:17:55linuxstbIndeed. :(.
02:17:59{Mike}install a new firmware?
02:18:01preglowit actually manages to load the firmware
02:18:14linuxstbIt seems to - see bootloader/main.c
02:18:15preglowhave you used rockbox before?
02:18:57{Mike}it was working fine yesterday :s
02:19:02preglowwell, strange
02:19:08{Mike}I know, tell me about it
02:19:13{Mike}just stopped working :(
02:19:13pregloweverything points to something wrong with the disk
02:19:16linuxstbI wonder what the first thing Rockbox itself does.
02:19:19preglowbut the bootloader actually loads the firmware fine
02:20:26preglow{Mike}: but did you try switching the unit on with usb plugged in?
02:20:51{Mike}woops wrong window
02:20:57linuxstbIt does sound like a hardware problem - if both Rockbox and the iriver firmware are freezing.
02:20:59{Mike}preglow: I can't turn it off
02:21:08{Mike}I can only reset it, or wait for it to run out of batteries
02:21:16linuxstbReset it with USB plugged in.
02:21:18preglowwell, plug in usb then press reset
02:21:58{Mike}think I need to reset my linux box
02:22:09{Mike}it seems to have screwed itself somehow
02:22:13 Quit {Mike} (Remote closed the connection)
02:25:08 Join {Mike} [0] (
02:25:13{Mike}right, got it mounted
02:25:23{Mike}reinstall firmware?
02:25:27preglowso, it works fine?
02:25:35{Mike}got it mounted ok
02:25:37preglowyou can access the disk just nicely?
02:25:44preglowdo an fsck or something
02:25:57preglowand check logs for any warnings from mount
02:26:04{Mike}don't call me a queer again ;)
02:31:03{Mike}while I'm here
02:31:08{Mike}tiny, tiny feature request
02:31:17{Mike}adjustible volume for the menu voices
02:31:24{Mike}they deafen me somewhat :(
02:31:40preglowthat'll come
02:31:49preglowa fix for the volume was commited today, actually
02:32:18preglowit now lowers the volume of playback while playing voices
02:32:34preglowbut like i said, a proper fix will come later
02:32:42{Mike}no fsck errors
02:32:59{Mike}new .rockbox and rockbox.whatever installed
02:33:00preglowtry stuffing in a new daily build
02:33:04{Mike}just did
02:33:05{Mike}no luck
02:33:07{Mike}same as before
02:33:14preglowvery strange
02:33:16{Mike}is it safe to try and reformat the hard drive?
02:33:21preglowwell, sure
02:33:30{Mike}will that do any damage?
02:33:43preglowi've done it myself
02:34:21{Mike}from outside the firmware?
02:41:12{Mike}repartitioned it, reformatted, installed latest build
02:41:16{Mike}no luck
02:42:45preglowhow have you partitioned it?
02:42:54{Mike}exactly as it was
02:42:58{Mike}1 partition
02:43:03{Mike}same class
02:43:08{Mike}filling whole disk
02:43:14preglowthen i've got absolutely no clue
02:43:35preglowif you can, you could return tomorrow
02:43:45{Mike}if I return it to iriver, its voided my warranty hasn't it?
02:43:48preglowmore developers are on around mid day tomorrow
02:43:55preglowyeah, i guess you have
02:43:57{Mike}midday what countries time
02:44:09{Mike}whats that GMT?
02:44:15preglowgmt + 1, i think
02:44:17preglow- 1
02:44:29linuxstbOnly iriver can say that though. e.g. I've sent an iBook with Linux (and only Linux) installed back to Apple for repair without problems.
02:44:30preglowaround 12.00 gmt should do
02:47:34XavierGrMike: You repartioned to a FAT32 right>
02:48:00XavierGryou use Linux so... I don't think you made it NTFS
02:48:11{Mike}XavierGr: yeh
02:48:17{Mike}same paritioning as before
02:48:18{Mike}I checked
02:48:32{Mike}checking for bad clusters, currently
02:48:57XavierGrstrangely though you say that your OS can see the HD....
02:49:12XavierGrthen it shouldn't be an HD problem
02:50:17XavierGrcould you open your files in your PC?
02:52:02{Mike}copied a new build, no problems
02:52:09{Mike}just won't boot :S
03:18:04{Mike}how long should an deep scan of the iriver take?
03:19:15pillapproximately 19 days
03:20:37{Mike}you being sarcastic, or are you just indicating it'll take a verrry long time
03:22:26preglowit does take a long time
03:22:28preglowbut i've gotta go
03:22:42{Mike}thanks for your help
03:33:23{Mike}I'm getting loads of clusters unreadable
03:33:59{Mike}Cluster 72269 is unreadable.
03:34:03{Mike}for example
03:39:00 Quit Febs ("CGI:IRC (Ping timeout)")
03:42:09XavierGrThen it may be a HD fault
03:42:18{Mike}good news
03:42:30{Mike}was doing a fsck of the hdd
03:42:33{Mike}it stopped
03:42:39{Mike}screen said something about a panic
03:42:43{Mike}then turned off
03:42:49{Mike}won't turn on any more, at all
03:42:57XavierGrif the HD is broken then I think it IS good news
03:43:00{Mike}tried pressing resets
03:43:09XavierGrmay be the battery is out.
03:43:14{Mike}plugged into ac
03:44:05XavierGrwait some minutes before starting it.
03:44:48XavierGras I was saying if the HD is broken then probably you could find a replacement. If something else is broken then it should be very difficult to find something to replace it.
03:44:54TiMiD{Mike}: did you tried the badblock command on HD ?
03:45:04{Mike}I don't know what that is
03:45:41XavierGroh TiMiD you are here?
03:46:01XavierGrremember about the little patch with the left button showing the wps when in root?
03:46:28TiMiDyes i'm here :)
03:46:41TiMiDyes I remember :p
03:46:47TiMiDwhat aboiut it ?
03:46:55TiMiD{Mike}: do you run linux ?
03:47:11XavierGrWell I made this to work with click instead of left, I can say that with joystick click it is way more convinient and safe. (not to rewind the track)
03:47:15{Mike}TiMiD: yeh
03:47:52TiMiD{Mike}: badblocks is a command
03:48:00TiMiDto check for badblocks on a disk
03:48:09TiMiDyou should give it a try
03:48:16TiMiDit maps all badblocks
03:48:22{Mike}I will do if I get it booted again :(
03:48:45TiMiDyou shulod also redirect the entry to a file since there can be a lot
03:49:04TiMiDusb bootloader
03:49:12{Mike}won't even turn on
03:49:13TiMiDit gives you disk access
03:49:16TiMiDok :/
03:49:25{Mike}pretty fucked, methinks
03:49:30XavierGrAlso I think that rockbox would need a quick quit button for the main menu (settings) currently all exit buttons of main menu revert to the previous screen
03:49:46XavierGr{Mike}: You press ON and nothing happens?
03:49:48TiMiDXavierGr: you mean you must clikc on select to go to the higher dir ?
03:49:57{Mike}yep, press on, nothing happens
03:50:02{Mike}no flashes, nothing on the screen
03:50:06{Mike}press reset, same thing
03:50:17{Mike}last thing I saw was a panic of some sort
03:50:19XavierGrThat is very akward
03:50:29{Mike}return to iRiver methinks
03:50:42XavierGrdoes the green LED turns on while in main?
03:51:25XavierGrgive it a full charge and try again in the morning
03:51:27TiMiDis it under warranty ?
03:51:32{Mike}thats the only sign of life
03:51:33{Mike}I dunno
03:51:41XavierGrsearch for badblocks and then you could see if it is the disk to blame
03:51:56{Mike}if I get it to turn on, I will
03:52:11TiMiDit should power on even if the battery is empty
03:52:21XavierGreven if it is under warranty I doubt that they will give you an H100 even if they accept the unit you wil probably get something else.
03:52:27TiMiDwhen connected to sector
03:52:44{Mike}should I contact iRiver or the vendor?
03:53:18XavierGralso when you are doing disk checking (if the unit powers on) be sure to have it on mains so that it will never turn off if the battery dies again.
03:53:27{Mike}it was on mains
03:53:48XavierGrthen :(
03:54:12XavierGranyway I think these are bad news but you could always try the next day, just give it a rest for now.
03:54:25XavierGrContact iriver the next day if nothing happens
03:54:27{Mike}I'll leave it charging tonight
03:54:31{Mike}will do
03:54:33{Mike}I dunno
03:54:55{Mike}to be honest, I'd rather have it completely screwed, and iRiver send me a replacement, than them finding out about me flashing another firmware :s
03:54:56XavierGrcome tomorrow midtime and we will see.
03:55:17XavierGryou will be the first one!
03:55:25XavierGrto send a unit with flashed firmware
03:56:07 Quit markun ("leaving")
03:56:41{Mike}if it won't turn on, short of crazy voodoo magic, I don't see what you guys could do
03:57:10XavierGrTiMiD: As I was saying currently in the patch I changed buttons again so, every time you press click the wps will toggle on or off (except if no music is playing, stop)
03:57:57XavierGryeah if it will not turn on I just dont think that anything can be done, but just let Linus know about it.
03:58:12XavierGrHe knows everything about the iriver schematics
03:58:35{Mike}whats his nick?
03:59:15XavierGrHe will probably now it by tomorrow he reads the logs
03:59:51{Mike}does that replace now with know in vim?
04:00:19XavierGrWho knows maybe the iriver technicians will see rockobx on the unit and realize that this firmware is better.
04:00:37TiMiDi don'tthink so
04:01:02TiMiDbut afaik some people returned a flashed iriver without problem
04:01:40{Mike}hows your remote patch going timid, btw
04:02:02XavierGrTiMiD: And yes pressing click while in menu would be good to exit the menu instead of going back (2 keys right now go back, left, A-B)
04:03:13XavierGrBut I guess this changes in button handling will be overuled because of GUI inconcistency
04:03:25XavierGrWhat do you think?
04:04:14TiMiD{Mike}: waiting for more feedback from devs, I modified it like linusN said
04:04:18XavierGr{Mike}: Waits developers for further guidance, developers didn't had the time yet to check the patch and aprove it for comittment
04:04:35{Mike}is it working currently, only tried fireflys
04:04:53TiMiDXavierGr: pressing click is used to play a file oO
04:05:01TiMiDnot to return oO
04:05:11XavierGryes but right is doing that too.
04:05:19TiMiDit's weird
04:06:05TiMiD{Mike}: it's working very well for the parts I implemented, IMHO even better than the current code
04:06:24XavierGrI like this button handling because I can do whatever I want with just one hand (with joystick)
04:06:25 Join Vlad0man [0] (
04:07:07TiMiDXavierGr: well why not :)
04:07:18TiMiDit may be ergonimic after all :p
04:07:28TiMiD(but I only use the remote ;) )
04:07:57TiMiD{Mike}: cool ... I'm quite bored to update my patch each time cvs change :(
04:08:00XavierGrWell as I said I think that devs will not like it, but I will surely keep this for my sake.
04:08:13TiMiDXavierGr: I had an idea
04:08:19TiMiDlong time ago
04:08:30TiMiDa config file for key assignment
04:08:35TiMiDand a new api
04:08:38XavierGryeah I had that too
04:08:49TiMiDinstead of catching buttons you would catch actions
04:09:32TiMiDbut that still an idea and it may probably be not accepted
04:09:55XavierGrThough an config key assignment api would have to be precise and very carefull not to introduce bugs in new assignments.
04:10:26XavierGrthe program would have to check for buttons consistency
04:11:17{Mike}you could, if someone would, write an in-gui program to set them
04:11:38XavierGryes it is doable
04:11:51XavierGrbut I think that this is a waste of the developers reject it
04:11:53TiMiDnot a big deal :p
04:12:27TiMiDthe more difficult part would be to convince them to add a layer over the button api
04:12:55XavierGrThis think is valuable only if it makes to be commited. This is only for users
04:13:05XavierGrso if it isn't commited no one will see it
04:13:07TiMiDthey say : "it's very flexible right now : yu can configure the key assignment as you want, youhave the source"
04:13:41XavierGrI think that there is no need for an API if this goes on button.c
04:14:22XavierGrthough this could easily be done via a script on the source
04:14:29{Mike}aye, I guess
04:14:33XavierGrno runtime process
04:14:37TiMiDI don't think so :p
04:15:11XavierGrimagine a script that will take the keymaps and change button.c accordingly
04:15:11TiMiDI think the source would be more readable if there wasn't all those #define stuff for buttons
04:15:35TiMiDXavierGr: the problem is that it must be changed for each program ..
04:16:08XavierGrWell if you make changes in button.c I dont think so, you just replace a nubmer on the define
04:16:28XavierGrthen the button will change with one of your preference
04:16:43XavierGrexcept if you want different mappings for each function of the player
04:16:45TiMiDyes, but wit this the key mapping would be changed for each program !
04:17:09***Saving seen data "./dancer.seen"
04:17:16TiMiDI want it configurable action by action ;)
04:17:32XavierGrthis will be a beast!!!
04:17:46TiMiDthis will never be accepted
04:17:58XavierGrso many buttons and so many programs
04:18:02TiMiDand this is far far away
04:18:20TiMiDlots of programs
04:18:34TiMiDyou would load a key config like a language file
04:18:46XavierGrthough the script way is easy, doesn't have to change code and needless for commitment.
04:19:36XavierGrthough the user would need the source even if he doesn't counter act with it.
04:19:59TiMiDand it won't be what I need :p
04:20:07TiMiDI need total flexibility :)
04:20:20XavierGryou are very greedy!!!
04:20:25TiMiDI'm an extremist :)
04:20:59TiMiDwell that will be after full remote support
04:21:15TiMiDand maybe a little code cleaning :)
04:21:28XavierGrI am impatient on your remote work.
04:21:32TiMiD(there are many place where code needs to be cleaned and commented :)
04:21:40TiMiDme too :)
04:22:18TiMiDif you want you can use the patch ;p
04:22:25XavierGrI don't know about cleaning (I am not to so advanced yet) but extra comments would help me a lot.
04:22:28TiMiDtell me I'm not crazy
04:23:24TiMiDthe remote display is one line shorter on X11 simulator isn't it
04:23:31XavierGris that the X11 simulator on linux?
04:23:45TiMiD(both with my remote patch and plugins !!!)
04:23:51TiMiDyes this is on linux
04:23:57XavierGrI have never run the X11 I use the windows cygwin.
04:24:01TiMiDamiconn !!!!
04:24:03XavierGrLet me check
04:24:14 Quit _Vladoman (Read error: 110 (Connection timed out))
04:24:35TiMiDcheck if the snow plugin displays snow outsidethe screen on win3é
04:24:56XavierGryes you are right
04:25:02XavierGrand that can be seen on the scrollbar
04:25:30TiMiDamiconn: so I wasn't dreaming after all :)
04:26:11TiMiDI spent lot of time to try to solve this in my code
04:26:19TiMiDthen I noticed it was a simulator bug :)
04:26:35TiMiDsince on the target the last line is displayed correctly :)
04:26:55XavierGryou must never trust the simulator
04:27:11TiMiDit's convenient :p
04:27:19XavierGrThe simulator always turns me down.
04:27:21XavierGryes it is
04:27:35XavierGrBut many times it crashes while main unti will not.
04:27:37TiMiDthe most would be an emulator
04:27:47TiMiDit crashes ?
04:28:12TiMiDunder linux at least it doesn't crash :)
04:28:14XavierGrWin32 simulator always crashes when I try to add a preset (in default code)
04:28:15TiMiDbut you know
04:28:23TiMiDcrashes are always a bad thing
04:28:24XavierGrwhile it works naturally on target as intended
04:28:50TiMiDa crash under simulator not happening on target doesn't means the bug is gone :)
04:29:04XavierGraslo the simulator did this on various occasions with char pointers
04:29:13TiMiDuh ?
04:30:17XavierGrNo the win32 simulator appears to be right with snowflakes
04:30:47TiMiDso you are witness tht there is a bug :)
04:31:02XavierGrand try to build an emulator yourself :p this is a major beast to handle
04:31:28TiMiDI agree :)
04:31:43TiMiDbut it wouldbe very convenient for low level devs
04:31:55TiMiDoops :)
04:32:09TiMiDmy sentence could be interpreted the wrong way :p
04:32:13XavierGrThe worst developing time I had with rockbox was when I was programming the JPEG file scroller (which stills awaits pass for commitment)
04:32:25TiMiDah yeah
04:32:33TiMiDobliged to upload on unit each time
04:32:36XavierGrThe simulator doesn't support this plugin. And every test I did was on my main unit
04:32:45TiMiDbecause no greyscale support
04:32:59XavierGrThat was very tiring and stressfull for the unit.
04:33:21TiMiDsame for me when I was doing fire and plasma evolution
04:33:37XavierGrI feel you.
04:33:59XavierGrAt least your work was commited ;p
04:34:04TiMiDsame also when I started to code my first pligin 'demistify' since simulator was crashing with a "floating point exception"
04:34:20 Quit dpassen1 ()
04:34:26TiMiDI wonder if it's sefull
04:34:32TiMiD3 useless plugins :p
04:35:15TiMiDyou never got your work commited ?
04:35:33XavierGr2 things got commited by me
04:35:58XavierGrsnake mod for iriver and virtual keyboard screen adjustments for iriver too
04:36:07XavierGr2 are pending
04:36:22XavierGrJPEG file scroller and FM advanced preset handling
04:36:54XavierGrlast 2 are very handy IMHO
04:36:58TiMiDI forgot this one :
04:37:05TiMiDgive it a try ;)
04:37:25TiMiDanother demo, anice one I did :p
04:37:31TiMiDI almost forgot it :)
04:37:54TiMiDbtw, what is JPEG file scroller ?
04:37:58XavierGrwhats next? Starcraft demo maybe? LOL
04:38:12TiMiDyou can by now scroll your jpeg file
04:38:17TiMiDlol ^^
04:38:36TiMiDthis starfield demo on my site is very basic
04:38:44XavierGrWell this patch (you can found it on the patch tracker) can scroll between JPEG files located on the same dir
04:39:05XavierGrthus when you choose a jpeg (and has more on the same folder) you can press up/down to change picture
04:39:09TiMiDI have one on my main computer in paris which is a monster : cameras, 3D rotations and so on, but it's not finished :)
04:39:24XavierGrI will give it a try
04:39:41TiMiDStarcraft ... I got a lot this night :)
04:39:58TiMiDI'm coming from a LAN party I orginized with some friends and arggh
04:39:58XavierGryou are playing starcraft?
04:40:09XavierGrLucky you!
04:40:14TiMiDthe best RTC :)
04:40:37XavierGrI used to play... too
04:40:42TiMiDI too
04:40:50TiMiDbut I started again this year
04:41:02TiMiDsince the game is in fact very good :)
04:41:13TiMiDnot like war3
04:41:16XavierGrwhich clan do you handle best?
04:41:18TiMiDnot slow
04:41:24TiMiDI do'ntknow
04:41:39TiMiDI have different strategies with the 3
04:41:41XavierGrI prefer protoss or zergling
04:41:44TiMiDI orefer the zergs
04:41:46XavierGrI mean Zerg
04:42:18TiMiDI like hydralisk + lurker + guardian (my little pleasure :p)
04:42:38XavierGrThere was a korean guy last month who died from excessive playing of starcraft!!!!
04:42:53XavierGrThere multiplayer is like olympic games
04:43:02TiMiDthis night it was more a drinking party than a real lan
04:43:48XavierGrAhhh I had never had the chance of a lan party with friends, only net caffees
04:43:49TiMiD(a friend brought some smirnoff ... parties were . funny :) I should have saved the replays ^^)
04:44:06TiMiDI played both
04:44:12TiMiDbut Iprefer lan :)
04:44:38TiMiDand you can cheat by at least hearing the sounds of the others ;)
04:44:47XavierGrdoes starfield runs on the sim
04:44:51TiMiDif you can't cheat then it's useless to play ^^
04:45:03TiMiDstarfield : yes I thing
04:45:08TiMiDI will try
04:45:11XavierGrnaughty starcraft palayer
04:45:37TiMiDjust remove the #ifndef SIMULATOR
04:45:55TiMiDI'm not very good at that game
04:46:12TiMiDI can't manage attacks
04:46:16TiMiDvery well
04:46:34TiMiDIonly win because I have numeric superiority
04:47:13TiMiD(and because the others sucks more than me ^^)
04:48:12XavierGrwell starfield compiled cleanly now I will test it
04:48:40TiMiDhuhu :)
04:48:48TiMiDyou are the first one to test it :p
04:49:01TiMiDit's one of my prefered
04:49:11XavierGrI am honored!
04:49:21XavierGrwell it crashed right away!!!!
04:49:33TiMiDow ^^^
04:49:38XavierGranyway I will test it on target
04:49:39TiMiDowned :p
04:49:49 Quit XShocK (Remote closed the connection)
04:49:49TiMiDit wored just fine on target :)
04:50:51TiMiDworks fine
04:51:01TiMiDand it's beautifull ;)
04:51:58TiMiDI willmaybe continue to code it with rotations
04:52:09TiMiDZ rotation was working well
04:52:38TiMiDthe only problem was with Y and X rotation since it modify the source of the stars
04:52:47XavierGrwow cool!!
04:52:51TiMiDit works ?
04:54:02TiMiDenjoy it ?
04:54:07XavierGrthe letters need better aligning
04:54:15TiMiDI should post it on the forum maybe
04:54:27XavierGrand be sure to include some BUTTON_LEFT | BUTTON_REPEAT
04:54:48XavierGrbecause it is a little tiring to press one by one
04:55:08TiMiDwhen I satrted codin it I wasn't used to rb
04:55:38XavierGrwhich are the limits?
04:56:17TiMiDspeed : no limit (works also on negatives)
04:56:21TiMiDstars : 3000
04:56:29TiMiDbut you can put more if you want
04:56:38TiMiDit's only a screen limit ...
04:56:47XavierGrwell at 3000 it is almost full of stars
04:57:22TiMiDwhat was wrong with letters ?
04:57:23XavierGrthough the movement is very bad due to the bad LCD refreshes when at high speed....
04:58:12XavierGrthey appear ~5 pixels more down that it should be.
04:58:42XavierGrso the screen can't show the bottom of the letters
04:58:59TiMiDI don't see that on sim :(
04:59:15XavierGrbut it crashes on sim no?
04:59:16TiMiDalso as far as I remember it was fine on target
04:59:22TiMiDno it doesn't crash
04:59:24XavierGrat least on my sim
04:59:39TiMiDtry to put a debugf somewhere to see where it's crashing
05:00:17XavierGrsee I told you that the win 32 sim isnt to be trusted
05:00:36TiMiDtry, mabye it's a real bug :/
05:01:48TiMiDi'l just curious to know at which point it stops
05:02:18TiMiDthen I will post it :)
05:02:27XavierGrwhat's the inline attribute of a function?
05:03:12XavierGrstatic inline void starfield_move_and_draw
05:03:13TiMiDit tells the compiler to put the code inside the calling code
05:03:22TiMiDso there is no overload due to fn call
05:03:37TiMiDbecause I like ptrs and generic fn
05:03:43XavierGrI missed that argument when I was reading C
05:03:56TiMiDnot very important :)
05:06:02TiMiDnot a very complicated code isn't it !
05:06:53XavierGrno it isn't complicated at all.
05:07:24XavierGrahh I know why I saw chopped letters
05:07:52XavierGryou use the user font instead of the fixed font
05:07:59TiMiDok :^
05:08:02XavierGrso I have big letters thus I get choppy ones
05:08:13TiMiDIwill fix that :p
05:08:20XavierGryou should use the get_font_height or something like that
05:08:26XavierGrI dont remember the function well
05:08:30TiMiDchotta matte kudasai ne
05:08:42TiMiDI remeber it don't worry
05:08:53XavierGrbut there is a function that returns the height of a font
05:09:11TiMiDwith my remote patch the font size is stored
05:09:17XavierGrtranslation be me: hey wait ... further I dont know
05:09:19TiMiDso no need to recalc it
05:09:32TiMiDnot chotta
05:09:43XavierGrwhats the difference?
05:09:58TiMiDchotta doesn't exists :p
05:10:22XavierGrI dont know japanese I just hear these on anime series and movies
05:10:48TiMiDyou must have heard this sentence a lot then !
05:10:58TiMiD"wait a little"
05:11:10XavierGrwhat kudasai ne means?
05:11:12TiMiDin polite form
05:11:36XavierGrand "onegai"?
05:11:41TiMiDplease :)
05:11:51TiMiDthe uses are different
05:12:02XavierGrsee no need to learn Japanese watch anime ;)
05:12:12TiMiDhehe :)
05:12:32TiMiDjust go to japan and live there 5 years and you 'll becime bilingual
05:12:59XavierGrI know some words epecially those that are being repeated
05:13:19XavierGrgomme n' asai
05:13:57TiMiDgomen nasai :)
05:14:07XavierGrwhatever lol
05:14:32XavierGrattachi means "I"?
05:14:49TiMiDit's I for woomens
05:15:01TiMiDmens use boku
05:15:04XavierGrthen I dont want to say that
05:15:09TiMiDsure :)
05:15:19TiMiDwell you can use it if you are a men
05:15:33TiMiDIdon't know really the implication of this
05:15:38XavierGrI ment watashi
05:16:14TiMiDI know that usually mens say "boku" to refer as themselves
05:16:23TiMiDbut smetimes they use watashi
05:16:31TiMiDI didn't tried to figure out why
05:17:07XavierGrI like the -kun -chan -san -sensei suffixes!
05:17:45XavierGris there a special meaning for shin
05:17:54XavierGrbecause I have watched a cartoon named shinchan
05:19:17TiMiDIdon't think so
05:19:34TiMiDbut I far far away from an ecyclopedia in this domain :)
05:20:12TiMiDI thought you was asking if there was a shin suffix
05:20:27TiMiDshin means "truth"
05:20:29XavierGrno shin word
05:20:51XavierGrwell it is competely irrelevant with the hero....
05:20:52TiMiDor new in some cases
05:20:59XavierGrwhich is a trouble some kid
05:21:20XavierGrmaybe it is just a name
05:21:29TiMiDit can also mean "heart"
05:22:11XavierGrwhats the relevance between new, heart and truth?
05:22:14XavierGrstrange language
05:22:18TiMiDnothing :)
05:22:37TiMiDbut they are not written wth the same kanjis
05:22:42TiMiDat least
05:22:54TiMiDit's like in french
05:23:00TiMiDvert => green
05:23:07TiMiDverre => glass
05:23:13XavierGrwell your program crashes the simulator when it reaches the ret=main();
05:23:36TiMiDit doesn't even enters into main ?
05:23:54XavierGryes it will not even run the first line of code into main
05:24:15TiMiDI can't do anything about it then :)
05:24:21XavierGrbackstabbing simulator!
05:24:54TiMiDthere must be a reason !!
05:25:18TiMiDsome devs must know the archane of the win32 sim :p
05:25:51XavierGrhehe it is weird though it will crash just entering the main
05:26:14XavierGrI have set an rb->splash just before and just right inot main
05:26:27XavierGronly the first splash triggers then BOOM!
05:27:55TiMiDyes boom
05:28:05TiMiDwindows boom
05:30:38XavierGrI should make an icon ( 6*8 ) for the fmr files in the tree.
05:30:51XavierGrbut I am I haven't found something interesting
05:32:15TiMiDjust draw one yourself :)">
05:32:56XavierGrI am not good at drawing
05:33:08XavierGrand there is no drawing at all in such low resolutions
05:33:22XavierGrsee the page how to make one
05:35:07TiMiDit's "abstract" :)
05:35:42XavierGrI will probably use the R symbol which stands for Radio
05:35:45TiMiDI like the one at (1,6)
05:36:00TiMiDlooks like an antenna
05:36:23TiMiDI think R stands for Rb
05:36:36XavierGrah yes
05:36:50XavierGrwhich 1,6?
05:37:19XavierGrI don;t see an antenna there
05:37:46TiMiDx=0, y=5
05:37:54TiMiDwhen you count from top
05:38:15TiMiDthe 6th from the top
05:38:19TiMiDfirst column
05:38:30XavierGryes got it
05:39:05TiMiDyou can actually also see it as an alien symbol ;p
05:39:09XavierGrits like a crossed X with a | in it
05:41:19TiMiDmysterious one :)
05:41:27TiMiDI wonder what it trulymeans
05:42:01XavierGrthe "X" stands for X-files
05:42:44TiMiDlol ^^
05:43:11DBUGEnqueued KICK XavierGr
05:43:11***Alert Mode level 1
05:43:12TiMiDalien code ^^
05:43:14XavierGrahh sorry for that
05:43:49XavierGrthere an R with lines
05:44:15XavierGror then it is just like RI
05:44:52TiMiDR with a 4 bars E:)
05:45:04TiMiDalien spotted
05:46:16XavierGrwhat about this then
05:46:28XavierGrit looks like a cfg file with an R instead of C
05:47:23TiMiDlooks weird :/
05:47:31TiMiDI prefer my antenna :)
05:48:03XavierGrI like the "node" remark instead "note" for notes that Rasher quoted?
05:48:47XavierGrSpeaking of it where is Rasher? I haven't seen him some days now.
05:49:03TiMiDI don'tknow
05:49:13TiMiDsince I wans't here either :)
05:49:38XavierGrwhen you teamed up?
05:50:33TiMiDteamed up
05:50:41TiMiDwhat does that means ?
05:51:15XavierGrwhen you joined the cause?
05:51:41XavierGrwhen you started fiddling with rockbox and join the chat room?
05:53:12***Alert Mode OFF
05:53:30TiMiDI've been on the irc channel for 1 month">
05:54:54XavierGrhaha maybe I am a blind man after all
05:55:03XavierGrlook who's first
05:57:12TiMiDtoday oO
05:57:15XavierGrMaybe I am these stats maybe a little bit old... who knows
05:57:29TiMiD(wellI wasn't here much today too)
05:57:48TiMiDWednesday 28 September 2005
05:59:18XavierGrhow can I make those icon ".#" to rockbox hex format?
06:00:12XavierGrbut the hex number has something to do with x,y
06:00:15TiMiD8 chr
06:00:23TiMiD=> binary
06:00:29TiMiD=> 1hex
06:01:11TiMiDthat's how I would od it at least
06:01:37XavierGrhmm yes it is 8*6 so 6 hex numbers for 8 binary sets
06:01:41TiMiDsince it's in 6x8
06:01:52TiMiDjust turn it
06:02:31TiMiDright side goes on top
06:03:25XavierGrnot left?
06:04:09TiMiDI think it's right side
06:04:22TiMiDturn it not like a clock
06:04:35TiMiDif you see what I mean
06:04:39XavierGrit goes from bottom to top
06:04:46XavierGrand left to right
06:05:42TiMiDyou are right :)
06:05:50TiMiDI must be sleepy
06:06:09XavierGrMe too but I will not sleep tonight
06:06:13TiMiDI submit my wonderful starfield and I go to bed
06:06:18TiMiDwhy ?
06:06:38XavierGrI woke up on 5.00pm so I will stay awake until 9.00 am to catch my college lesson
06:07:03TiMiDskipping night
06:07:07TiMiDnot good :)
06:07:12XavierGrdont you have anything to do tomorrow?
06:07:26XavierGrit will be 6.00 am there right?
06:08:28TiMiDI have nothing
06:08:32TiMiDno school :)
06:08:50XavierGrlucky you, I will do that on teusday!
06:08:50TiMiDand I wanted to sleep so bad when I left that lan
06:08:54TiMiDbut now
06:09:13TiMiDtuesday I have only 1 hour
06:09:31TiMiDso it's ok :)
06:10:28XavierGrjapanese, in what sort of school are you in?
06:10:43TiMiDEngineer school
06:10:46XavierGror do you do this on free time?
06:10:49TiMiDcomputer science
06:11:01TiMiDsome kind of free time
06:11:04XavierGrAnd what's with japanese?
06:11:16TiMiDbut it counts like a school lesson
06:11:27TiMiDlike a language
06:11:52TiMiDI choosed this because in english I wasn't learning anything
06:12:07TiMiDnot that I have a good level
06:12:19TiMiDbut the lesson really sucked
06:12:28TiMiDa waist of time
06:12:49XavierGrme too
06:12:56XavierGrI did exactly the same
06:13:05XavierGrthough I didn't get another language
06:13:12XavierGrnot that we had one to choose
06:13:22XavierGrRadioIcon = { 0x5f, 0x55, 0x5b, 0x50, 0x55, 0x,55 }
06:13:32XavierGrwow a lot of 5s
06:14:09TiMiDstrange :)
06:14:30TiMiDperhaps a bogus :)
06:16:54XavierGrnow lets see....
06:17:11***Saving seen data "./dancer.seen"
06:18:31XavierGrhmm not bad though....
06:21:38TiMiDet voila !
06:22:00XavierGrwait to see mine
06:23:32XavierGrzoom it and see
06:24:28XavierGrmaybe the line before the last has to be blanked
06:25:32TiMiDit'stoo small for my eyes
06:25:58XavierGrbut every icon is small!
06:25:58TiMiDI zoom with an image tool
06:26:42TiMiDwhich one is the radio ?
06:26:50TiMiDselected ?
06:26:53XavierGroh god....
06:27:00XavierGrthat means that I failed
06:28:27TiMiDthe one I told you was good :)
06:28:35TiMiDmaybe some waves
06:28:41TiMiDyou can draw that
06:29:09XavierGrgrrrr I will take more responses before I add the X-files icon :p
06:29:17 Quit RotAtoR ()
06:31:07TiMiDand waht about configurable icos ? :D
06:33:24XavierGryou are a heretic!
06:33:46TiMiDI just want to recode rockbox ^^
06:33:52XavierGrsome will haunt you down for your beliefs
06:34:15TiMiDanyway some ideas like that are not bad :)
06:34:59TiMiDthat remembers a guy who wanted java on rb ^^
06:35:03XavierGrnot bad, but too much luxuries will cripple the whole project in the end
06:35:20XavierGrRockbox has too many fronts right now.
06:35:34XavierGrAnd come to think of it that it has only discontinued targets
06:35:45TiMiDrb needs a lot of code cleaning
06:36:18TiMiDthere is a lot of duplicated code
06:36:57XavierGrmaybe I am not the expert for that, my knowledge is limited in these aspects.
06:37:09TiMiDI'm mrProper :)
06:37:46XavierGrThen you should seriously talk that with the devs and jump in for a cleanup.
06:38:10TiMiDuseless since I'm doing it while coding the remote
06:38:52TiMiD(not 100% true since some codes like tree would need a lot more than a simple cleaning)
06:39:17XavierGrwell I meant, ask the devs for it then if you get cvs access start the cleanup and not only for the remote.
06:40:12TiMiDThe remote is a very broad part
06:40:41XavierGrwell all is easy to say, but rewriting tree.c is a huge task.
06:40:46TiMiDAlmost all the application code is concerned
06:41:44TiMiDI don't intend to rewrite everything
06:42:04TiMiDjust some parts
06:42:17TiMiDin tree.c I rewrote almost nothing
06:42:26TiMiDmaybe 200 lines
06:42:44TiMiDeverything else was deleted because unneeded
06:43:38XavierGrwait a sec you say that your tree.c is 200 lines?
06:43:58TiMiDno !
06:44:03XavierGror that you just rewrote 200 lines?
06:44:22TiMiD200 lines are my own work
06:44:40XavierGrand the rest is untouched?
06:44:43XavierGror with slight modifications
06:44:54TiMiDit's modofied but it's not really my work
06:45:11TiMiDI have a 1362 lines tree.c
06:45:36XavierGr1.684 original
06:46:05TiMiDthe rest is only comments added, lines removed because unneeded or modified
06:57:12TiMiDwell I'm gonna go to sleep
06:57:16XavierGrtime flies
06:57:16XavierGrit is 8.00 am here
06:57:18TiMiDcu !
06:57:18XavierGrin half an hour I will go to college better to get ready
06:58:04TiMiDthen have a nice day
06:58:33XavierGrand you sweet dreams!!
06:59:55 Join amiconn_ [0] (
07:00:24XavierGroh good morning amiconn
07:04:08 Join webguest58 [0] (
07:04:32 Quit webguest58 (Client Quit)
07:17:33 Quit amiconn (Read error: 110 (Connection timed out))
07:17:33 Nick amiconn_ is now known as amiconn (
07:23:40 Join matsl [0] (
07:28:10 Quit matsl (Remote closed the connection)
07:29:02 Join matsl [0] (
07:31:11 Quit Bger ("[BX] Does your mommy know you use BitchX?")
08:17:12***Saving seen data "./dancer.seen"
08:21:21 Join tvelocity [0] (n=tony@
08:24:07 Join einhirn [0] (
08:35:25 Quit matsl (Remote closed the connection)
08:36:57amiconnTiMiD: These exact display problems shown in your screenshots were fixed with yesterday's 13:09 changes.
08:37:28amiconnIf you still get them you don't have these changes in your tree for some reason
08:39:15amiconnThere were errors both with the remote 'display frame' and the window size (even worse when using the zoom parameter)
08:51:39 Nick Vlad0man is now known as Vladoman (
08:59:11 Join Zagor [0] (
09:02:18 Quit Lynx_ (Read error: 104 (Connection reset by peer))
09:11:04 Join ashridah [0] (
09:15:55 Join cYmen [0] (
09:25:24 Quit solexx (Read error: 110 (Connection timed out))
09:32:00 Join ender` [0] (i=ychat@
09:52:44 Join LinusN [0] (
09:53:27LinusNwas just reading the logs, if this "Mike" comes back, have him email me, linus at
09:55:39 Join Lynx_ [0] (
09:56:11 Join solexx [0] (
09:59:55XavierGrok if i see him i will redirect him to you.
10:00:14XavierGrany idea what is wrong?
10:03:25LinusNi guess a power regulator is busted
10:03:47LinusNi have a similar case here, on the operating table :-)
10:04:00LinusNfixed it by replacing the regulator
10:04:50LinusNgotta go
10:04:53 Part LinusN
10:09:41SlasheriTiMiD: who not to use cvs update and diff to do it? That way you don't need manually do anything to update your patch
10:15:40XavierGrTiMiD is sleeping right now.... :p
10:17:14***Saving seen data "./dancer.seen"
10:20:45Slasherihehe, but he will read the logs :)
10:26:51ashridahSlasheri: that doesn't cover files he's written himself since it'll want to connect to the cvs root if he uses cvs add tho ...
10:31:03 Join markun [0] (
10:51:25 Join phaedrus961 [0] (
10:59:43Slasheriashridah: No, he can use cvs-do add to do that
11:06:24 Join webguest68 [0] (
11:11:14Slasheriashridah: in fact, cvsdo (not cvs-do)
11:12:18XavierGrSlasheri: How things are going on the recording front?
11:14:52SlasheriXavierGr: very little because i have been quite busy on work.. but i try to look it a bit to make it suitable for commit
11:15:13webguest68what ever happened to the idea of having one big WPS graphic and "cutting" from it?
11:15:15webguest68Did that get dropped?
11:16:03XavierGrso whats the progress, until now have you ported all of archos targets recording features or are you in the middle of it. ( I don't mean to be pushy just asking)
11:17:29SlasheriXavierGr: recording to wav, file splitting and peak meter should work
11:17:43Slasheribut there are bugs that cause it to crash
11:18:31Slasherianyway, those should be easy to solve
11:19:08XavierGrwhen you implement it I think I am gonna get on of those fancy stereo microphones
11:19:31Slasherihmm =)
11:19:59Slasherii have never used a stereo microphone :)
11:20:09XavierGrWithout the infamous iriver "click" (in recordings) the recording quality level will be magnificent for a device like this
11:20:22Slasherihehe, yes
11:22:27XavierGrand add to that prerecording and peak meters wow.
11:22:27XavierGrme neither but I have heard that they are super...
11:22:32amiconnSlasheri: Did you add s/pdif recording as well?
11:23:15Slasheriamiconn: not yet (or at least i haven't got it working yet)
11:24:01Slasheriand with s/pdif we might have the sample rate problem because so many devices uses different sample rates
11:24:29XavierGrand how iriver works with that?
11:24:32Slasherifor example 44.1, 48 and 96 kHz
11:24:40Slasherii don't know yet :)
11:24:42Slasheribut we must handle it
11:25:01 Quit webguest68 ("CGI:IRC")
11:29:15amiconnSlasheri: For recording, the coldfire can sync to all allowed s/pdif bitrates (32, 44.1 and 48 kHz)
11:29:43Slasheriamiconn: hmm, that sounds good. Can it automatically detect correct bitrate?
11:31:55zemy soundcard can sync to s/pdif rate... so i assume s/pdif stuff in general can
11:32:19zethough i'd swear my soundcard could take 96khz s/pdif
11:35:01 Join Moos [0] (
11:36:54 Join muesli- [0] (
11:37:21 Join itsthemojo [0] (n=manuelka@
11:37:26itsthemojohi there
11:39:41 Quit itsthemojo (Client Quit)
11:40:11zeah yeah i see a screenshot of its control panel in the manual (the prog for it in linux is a bit different), and it has an s/pdif sample rate section including 32, 44.1, 48, 88.2, and 96 khz
11:46:00 Join whatboutbob [0] (
11:49:21 Join _FireFly_ [0] (
11:50:02muesli-g'moin _FireFly_
11:52:31whatboutbobslasheri: heya. just jumped on to say thanks for taking up the recording implementation challenge. :)
11:52:52 Join XavierGr_ [0] (
11:54:32 Quit XavierGr_ (Client Quit)
12:17:17***Saving seen data "./dancer.seen"
12:20:28 Quit XavierGr (Read error: 110 (Connection timed out))
12:23:46 Quit muesli- (Read error: 110 (Connection timed out))
12:34:56CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
12:34:56*einhirn unlurks
12:35:01einhirnHello all...
12:36:23einhirnJust had the Idea to research for an Infrared Remote Control over my Archos JBR - finding hints that it is already implemented
12:38:21einhirnBut how do I activate support for it? Build switches? I tried to take a look at rbconf but failed to build the "tools" according to because the compiler doesn't know where to find stdio.h et al...
12:40:24einhirnseems to be a Cygwin problem...
12:52:07einhirnOk, cygwin problem gone. Helps to run Setup.bat as Admin - Registry access needed...
12:54:47einhirnSo - rbconf doesn't help with "enable RC5 driver for Remote Pin"...
12:56:00einhirnWell, I'll be around later - maybe someone can help then...
13:04:05TiMiDSlasheri: that the way I'm updating it ... but when files I modified are modified on cvs too, I must update tem locally by hand :/
13:05:11Slasherihmm, there are conflicts cvs update can't automatically solve?
13:05:25amiconnTiMiD: Found my remark regarding the X11 sim?
13:05:59Slasheriwhatboutbob: hehe, thanks :) currently i don't have much time but i try working on it
13:06:30TiMiDamiconn: I'm using current CVS sources (at least that what i believe since I downloaded the sources from the site : "latest" :)
13:07:02TiMiDI will re-download and re-test without my patch
13:28:43 Join RiverFish [0] (
13:30:27RiverFishI have a question about replaygain which I don't think is in the FAQ
13:30:57RiverFishCan you use the replaygain settings to boost the volume output to the max possible level without clipping?
13:31:42RiverFishI find that my iRiver H140 hasn't got enough power for the quiet tracks when listening on a noisy train
13:32:57TiMiDamiconn: you are right :)
13:33:56RiverFishSo if I enable replaygain, set pre-amp to maximum, and prevent clipping to on...
13:34:12RiverFishthen that will boost quiet tracks to their max level without clipping?
13:34:23RiverFishbut only if the tracks contain replaygain tags.
13:34:44RiverFishOr do I need to code this feature myself?
13:35:44RiverFishI'm assuming that replaygain tags contain max amplitude information for each track as well as perceptual loudness information.
13:38:49preglowRiverFish: they need to contain replaygain tags for that to work
13:38:55preglowRiverFish: apart from that, i think it'll work
13:40:53RiverFishWill 12db be enough for the quietest tracks? Happy to make a patch myself if it isn't.
13:42:31preglowmight not be
13:43:30crwlcan't you just add replaygain tags to files, enable clipping prevention and crank preamp somewhere really high?
13:44:17preglowthat's what he's saying he will
13:45:17crwloops, i didn't read *that* line :)
14:17:21***Saving seen data "./dancer.seen"
14:28:58RiverFishthanks for the replies. I'll give it a try and see if 12db is enough.
14:29:42RiverFishIf not then maybe I might patch RockBox to make the limits larger
14:34:58RiverFishHow about allowing volume to go over 100% and making the excess a temporary boost to the pre-amp
14:35:25RiverFishSlightly wierd but it would mean it's easily possible to get a track as loud as possible
14:35:58RiverFishThis is for driving high impedance phones - not annoying my fellow commuters :)
14:36:49RiverFishwould anyone other than me find that useful?
14:39:56 Join Febs [0] (
14:43:20preglowRiverFish: i've thought about doing that, but not via preamp, rather a gain factor that would always be applied if volume is above 100%
14:44:21RiverFishSounds good. I guess replaygain tags would help prevent clipping
14:44:46RiverFishOtherwise there'd be no way of knowing whether the track is going to clip at some point
14:46:19preglowi'm going to need a software gain factor for the eq anyway
15:00:16amiconnWhen digital volume goes above 100% (with a normalised track) there will be ugly distortion. No way to avoid this
15:06:23markunMaybe we can make a sound compressor to screw up the sound :)
15:06:57preglowa limiter would b nice
15:08:32markunRiverFish: Maybe you could use somthing like this:
15:09:58RiverFishI've got a couple already.
15:10:07RiverFish which I'm trying to get repaired boohoo
15:10:13*amiconn wonders why so many people have problems with the H1x0 volume....
15:10:16markunIs doesn't help driving your headphones?
15:10:48markuneven with optical in.. that's nice.
15:12:05RiverFishSounded fantastic when it was working.
15:12:15RiverFishGreedy on the power though. 12v
15:13:18RiverFishEven that didn't always kick out enough to drive er4s
15:13:32preglowi don't have problems with it
15:13:35RiverFishThe problem is only with quiet tracks
15:13:41preglowsome tracks just are mixed low, though
15:13:49RiverFishon loud trains
15:14:07preglowbeside them as well :)
15:14:35RiverFishthat too 4 sure
15:17:26preglowdamn, i wonder how fast this eq is going to be
15:17:33preglowmarkuns crossfeed didn't exactly leave me optimistic
15:18:28markunpreglow: What is a typical number of samples that the crossfeed is processing at a time?
15:18:39preglowyou mean the dsp block size?
15:19:04preglowmarkun: and btw, i had a look at the page you linked me too, and i've got a couple of questions
15:19:30preglowmarkun: first, it mentions 300ms as delay size all the time, and second, it looks like the filters they use are shelving filters
15:19:42preglowi can't imagine where the 300ms figure comes from
15:20:21preglowit'd be a nice figure for a frigging reverb, but not exactly what i'd expect from a crossfeed
15:20:22markunOn my version of the site it says 300 mu-s
15:20:48markunAre we both talking about ?
15:20:53preglowahh, it seems i've found a bug in opera :)
15:21:02preglowit displays
15:21:07preglowit displays mu as a barred m
15:21:14RiverFish300ms crossfeed nice
15:21:29markunI don't know much about filters..
15:21:40markunIs a shelving filter easy to implement?
15:21:53preglowi can design you a couple of shelving filters pretty fast
15:22:02preglowdepends how accurate you need the cutoff to be
15:22:28markunMaybe you can simulate them and listen to the result?
15:22:30preglowbut they'll need more history
15:22:34preglowso it'll be slower :/
15:23:13preglowmarkun: ah, i forgot about that, i've got some time now
15:23:31markunTo me the crossfeed sounds ok as it is, but if you can improve it that would be great.
15:23:45preglowthere's always the volume loss
15:24:28markunYes, imagine all the rockbox volume complaints :)
15:24:48preglowcould you link me to a diff?
15:24:51preglowi've lost the source
15:25:24markunI'll make you a brand new one :)
15:25:44preglowbut we'll seriously need some plugin system for these kind of things
15:26:13preglowallowing for some small use of iram, and unfortunately, some way of mallocing delay buffers :/
15:27:20 Quit ashridah (Read error: 110 (Connection timed out))
15:28:02markunI think there should be a lot less move instructions because I made a local copy of the some of the values from crossfeed_data
15:28:31preglowthe compiler seriously should have optimised away those anyway
15:28:48 Join ashridah [0] (
15:31:26RiverFishsounds like a cool patch. is it work in progress
15:34:05markunbtw I think the unicode patch is quite ready..
15:35:12markunI don't think the summary of the patch reflects the current state:
15:36:40markunI have a japanese and a korean language file sitting here. Shall I wait with committing them till after rockbox has gone unicode?
15:37:01preglowhave you tested how the patch affects the build size for the various platforms?
15:37:12markunNo, I haven't.
15:37:35markunI don't have a compiler for the archos platforms.
15:37:52 Quit _FireFly_ ("Few women admit their age. Few men act theirs.")
15:38:01markunphaedrus961 has tested them I think. I will ask him when he gets back.
15:41:54RiverFishnewbie code question alert: Where do the firmware and codecs call each other?
15:42:16RiverFishI can see codec_api but can't see where the APIs that the codecs call are implemented
15:48:08preglowin codecs.h/c
15:48:24preglowthe codecs are spawned in a separate thread by playback.c/h, afaik
15:48:37preglowand codecs feed the playback engine with data via callbacks
15:52:27 Part whatboutbob
15:57:22RiverFishthx preglow
16:02:26markunpreglow: I'm having trouble compiling gcc for sh1 on FreeBSD so I can't test for the build sizes.
16:17:23***Saving seen data "./dancer.seen"
16:18:52 Join muesli- [0] (
16:21:04 Join Mindship-02 [0] (
16:21:34Mindship-02Can I use a carkit microphone as an ordinary mic?
16:24:27 Quit ashridah ("Leaving")
16:35:55Mindship-02It works wonderfully! (signal is about 20dB weaker, but so is the default of iRivers mic...
16:37:39 Quit Febs ("CGI:IRC (EOF)")
16:44:42 Quit einhirn ("Miranda IM! Smaller, Faster, Easier.")
16:52:44preglowmarkun: i've got the crossfeed setup going in reaktor here, and i'm most certainly not struggling with too low volume
16:52:48preglowrather the opposite
16:53:16markunWow, then what are we doning wrong?
16:53:29markunHow's the sound with headphones?
16:53:48preglowit's ok
16:54:02preglowcan't exactly remember what yours sounded like, but it's better than no crossfeed here
16:55:07preglowi can pop up a sample of what i've got right now
16:55:26markunok, can you do it with one of the songs I sent you?
16:55:40markunThat would make it easyer to compare.
16:57:21markundid you use the shelve filters?
16:57:49preglownot yet
16:58:07preglowi just concated a segment of lucy in the sky with diamonds and this song
16:58:32preglowsince they were the two that was the most jarring without crossfeed, heh
16:59:33markunyes, very annoying :)
17:00:45preglowi needed to adjust the gain down a bit, actually, otherwise it clipped
17:01:57markunThe crossfeed I made shouldn't introduce any additional clipping.
17:02:21 Join dpassen1 [0] (
17:02:38 Join muesli---- [0] (
17:03:18preglowno, because the highpass filter you used had severe gain errors
17:03:31preglowit adjusted the volume down with increasing cutoff frequencuy
17:04:25markunI tried it with substracting a lowpassed signal from the original. It worked quite well as a high pass filter, but the resulting crossfeed didn't sound very good., first segment without crossfeed, second with
17:05:37preglowsettings are 80hz highpass cutoff, 2000hz lowpass cutoff, 0.9 direct highpassed signal and 0.1 crossfeed
17:05:52preglowdoesn't sound as good as yours, i think
17:06:08preglowbut both filters i use are slightly different than yours as well
17:09:17markun'this song' still makes me dizzy
17:09:54 Join Maxime [0] (
17:10:07markunHow can I calculate the cuoff freqs of my filters?
17:10:10 Quit Maxime (Client Quit)
17:10:21 Join Maxime [0] (
17:12:27preglowwith your filters i have no idea
17:13:52markunMaybe I could email the maker of the foobar2000 crossfeed plugin to ask him what he uses
17:14:06preglowdoesn't sound like a bad idea
17:14:13preglowif you know that plugin is good, that is
17:14:21 Part Mindship-02
17:16:47markunI used it at the university before I have my iriver:
17:17:57 Quit muesli- (Read error: 113 (No route to host))
17:19:43 Join XavierGr [0] (
17:21:37 Quit RiverFish ("CGI:IRC (EOF)")
17:26:29 Quit muesli---- (Read error: 110 (Connection timed out))
17:37:19markunI used it at the university before I have my iriver:
17:39:40 Join _FireFly_ [0] (
18:00:41 Join ep_esial [0] (
18:12:28 Quit ep_esial ("Leaving")
18:17:26***Saving seen data "./dancer.seen"
18:19:00 Join DangerousDan [0] (
18:38:06 Join matsl [0] (
19:11:12 Nick jborn_ is now known as JoeBorn (
19:12:11 Quit JoeBorn ("")
19:24:47 Nick Lynx_ is now known as Lynx_awy (
19:26:03 Quit Lynx_awy (Read error: 104 (Connection reset by peer))
19:40:58 Quit dpassen1 (Read error: 104 (Connection reset by peer))
19:56:19 Part pike
20:12:19 Join Paul_The_Nerd [0] (
20:15:42Paul_The_NerdWell, I've encountered an interesting thing. I'm currently listening to track -42. I know it's because I have 1507 tracks but even the fact that it's signed seems odd.
20:16:19XavierGrmaybe your wps has a ''-" symbol?
20:17:29***Saving seen data "./dancer.seen"
20:19:16Paul_The_NerdWhen the track number hits 128 it rolls over to -127
20:22:46linuxstbPaul_The_Nerd: I'm guessing you're using the "%in" tag - to get the ID3 track number. I think that should only contain the track number within the album.
20:23:35linuxstbIt sounds like something odd in your ID3 tag.
20:23:47Paul_The_NerdThe ID3 tags are/should be fine.
20:23:58Paul_The_NerdIt's a very long audiobook split into about 4 minute files.
20:24:34linuxstbBut the ID3 tag for tracknum is only a single byte. I'm guessing the intention was to store values from 1..99 - the limits of CD.
20:25:15linuxstbI don't know if it should be signed or unsigned - but Rockbox is obviously treating it as signed. Which could be a bug.
20:25:18Paul_The_NerdWell, these are Ape tagged.
20:25:44Paul_The_NerdHrm. Ape and ID3v1... That's not right. :(
20:26:44Paul_The_NerdBut yeah, the main thing I was bringing up was that it was signed. I figured I was going well beyond the expected number of tracks (and it's mainly so that the metadata allows me to automatically rename them and have them in a proper sorted order for listening)
20:28:48amiconnI think id3v1 was made with the CD track limit in mind
20:29:09amiconnWe could make it unsigned and support up to 255 tracks this way
20:29:11Paul_The_NerdDidn't know that they'd been ID3v1 tagged along with the Apes. =/
20:29:45markunNow that we support ape and vorbis comments, maybe we should rename the id3 struct in metadata.c to metadata or something.
20:29:49amiconnRockbox itself supports really huge playlists. 1507 isn't really much....
20:30:35Paul_The_NerdYeah, but usually people will have playlists composed of no more than maybe 25-30 songs per disc/album
20:30:59linuxstbDon't forget the %pp WPS tag - to give you the position in the playlist.
20:31:16Paul_The_NerdI use that too.
20:31:51linuxstbI can't think of any reason why the id3v1.1 track number should be treated as signed though.
20:32:47Paul_The_NerdWell, I think returning negative track numbers, ever, is probably not a good thing.
20:34:16amiconnmarkun: id3 is a 'struct mp3entry' which is defined in firmware/mp3data.h. mp3data.[ch] still handles id3 exclusively, everything else is handled in metadata.c (which in turn uses mp3data.c for id3)
20:34:49 Join pengo [0] (
20:35:58linuxstbBut metadata.c also used the mp3entry struct.
20:36:15Paul_The_NerdWhich means Ape Track numbers are still loaded into the ID3->tracknum?
20:37:59linuxstbYes - but id3->tracknum is an "int" - so there's no problem there. Unless Ape track numbers are not integers...
20:39:10linuxstbI don't even know if Rockbox supports Ape tags in MP3 files yet though.
20:39:20Paul_The_NerdMaybe that's it.
20:39:59Paul_The_NerdI'd split it long ago, back when I'd used it on the iRiver firmware, which didn't have bookmarks.
20:40:21Paul_The_NerdIf I can get ahold of my friend's CDs again, I'll just use bookmarks and a better format.
20:40:38amiconnlinuxstb: Afaik it doesn't (and doesn't make sense to me)
20:41:27amiconnI think we don't need to support tagging systems for formats where they aren't 'native'
20:41:53amiconnmpeg audio has id3v1 and id3v2, ogg/* has vorbiscomments etc
20:42:43amiconnOf course rockbox shouldn't crash when it encounters 'alien' tags in a file...
20:43:07 Join lear [0] (
20:43:53markunamiconn: but I still think it's strange to call it id3 in metadata.c and store ape tags and vorbis comments in it..
20:48:34 Join Philip_0729 [0] (
20:49:16 Quit tvelocity (Remote closed the connection)
20:51:54 Join tvelocity [0] (n=tony@
20:52:49 Join SeeSchloss [0] (
20:54:44 Quit Philip_0729 ("bye everybody")
20:55:24 Quit XavierGr (Read error: 110 (Connection timed out))
20:57:49 Quit lear ("Leaving")
21:00:28 Quit Paul_The_Nerd ("Chatzilla 0.9.68a [Firefox 1.0.7/20050915]")
21:02:16 Quit DangerousDan ("Miranda IM! Smaller, Faster, Easier.")
21:17:03phaedrus961anyone know if shorts need to be aligned to a 16bit boundary on the sh1?
21:18:07phaedrus961ok thanks, that helps a lot
21:19:56amiconnshorts and longs should also be aligned on coldfire for performance reasons, but on sh1 they have to be
21:20:01amiconnelse -> address error
21:21:03phaedrus961that's exactly the error I was getting, and this seems to fix it, but I wanted to be sure
21:34:42amiconnBagder/ Zagor: Are you around?
22:05:28 Join Mxm`Pas`Bien [0] (
22:05:28 Quit Maxime (Read error: 104 (Connection reset by peer))
22:10:20markunamiconn: The unicode patch could be committed as it is phaedrus961 and I think. Would you like to take a look at it?
22:11:39markunAnd we've got an email from a japanese rockbox user that all the japanese users are starting to use the patch, so I guess it works :)
22:11:50amiconnDoes it implement at least basic unicode support for all platforms?
22:12:55amiconnI'm thinking about the player here. Of course it can't display non-latin characters (yet), but it should be able to handle unicode .lng files
22:12:57markunBut maybe you could test it on a few units that you suspect might give problems.
22:13:02amiconnThey are utf8, correct?
22:13:24markunYes, they are all converted to utf-8
22:14:03markunphaedrus961: You tested it on the player sim, right?
22:14:03amiconnBtw, the 'newplayer' lcd has some japanese characters...
22:14:30markunFor very very basic japanese support :)
22:15:10amiconnNot that I think it's necessary to use them in the first version...
22:15:35amiconnI have no idea how japanese works, is it like chinese, i.e. one symbol = one meaning
22:16:20amiconnIf so, there must be thousands of symbols... but then I don't understand what a few dozen chars in an lcd controller could help...
22:16:35phaedrus961the player should handle all utf8 strings (read: works for me in the sim)
22:16:37markunI don't understand that either.
22:17:08amiconn "
22:17:08amiconnHW layout of new LCD:
22:17:26 Quit Mxm`Pas`Bien ()
22:17:27phaedrus961afaik the player lcd supports the halfwidth katakana syllabary, not very useful for japanese
22:17:28amiconnOops, inserted the linefeed by accident...
22:17:31***Saving seen data "./dancer.seen"
22:18:57markunamiconn: The japanese char are kata-kana. They represent a syllable each.
22:19:31markunphaedrus961: Do you kow if converting kanji to kata-kana would be readable for japanese people?
22:19:53amiconnIs katakana part of the unicode table?
22:20:21amiconnCurrently reading that :)
22:20:25phaedrus961it would be complicated to convert it, and difficult to read, not worth it imho
22:20:42markunamiconn: and yes:
22:20:42phaedrus961amiconn: yes it is
22:21:26amiconnphaedrus961: Do you mean an automatic conversion? What about a special hand-made 'translation' (it's not exactly a translation, different writing??)
22:22:10 Quit _FireFly_ ("Leaving")
22:22:21phaedrus961that would be possible, but it'd still be a bit difficult to read
22:22:32{Mike}incidentally to anyone who was about last night: contacting Rockbox bout my bricked iRivier ihp140
22:22:46phaedrus961most japanese are not used to reading without kanjis
22:23:22amiconnAh ok. Still could be an interesting thing to do.
22:24:04amiconn"Katakana is still used today to display text on cash registers and other low-resolution displays"
22:26:19markunMaybe some japanese coders can do it if they really want that feature.
22:26:23amiconnHmm, how are fonts handled in the unicode patch? Are there some fonts to test with?
22:27:13phaedrus961fonts have to use the iso10646 encoding
22:27:34phaedrus961I've written a util which can convert any font to this encoding
22:29:16amiconnHow large does a typical font get? Is there lru caching (suppose there is)? Does the built-in font support unicode (guess not)?
22:30:20amiconnWow, an 1.2MB .diff! That beats TiMiD's multi-display patch by far!
22:30:26phaedrus961latin1 is the same in unicode so yes built-in font supports it
22:30:39phaedrus961and there is lru caching
22:31:09phaedrus961most of the diff is the cjk conversion tables, without them it's about 100k
22:31:38amiconnI wasn't precise about the built-in font. So the built-in supports the latin1 range only?
22:33:01phaedrus961built-in will support unicode yes, but the current one doesn't
22:33:07amiconnIf so, we need to think about a dual-storage default font.
22:33:44phaedrus961what is a dual-storage font?
22:33:46amiconnlatin1 part still built into the binary (e.g. for displaying a "font missing" message), the rest loaded from disk as all others
22:34:07phaedrus961ah, ok
22:34:15amiconnOtherwise a vkeyboard with non-latin chars won't work...
22:35:04phaedrus961yes, this is why I haven't added non-latin chars to the vkeyboard
22:35:21 Quit goa ("Client suicide")
22:35:27 Join webguest63 [0] (
22:35:32webguest63hi all
22:36:25amiconnThis is the next logical step of evolution, imho: support for multiple loadable fonts, with special handling of the default font. This would make non-latin translations perfect...
22:37:06 Quit {Mike} ()
22:37:16 Part pengo
22:38:06*amiconn is building unicode binaries
22:39:25amiconnI think I will have to adjust my voice building script for utf-8
22:39:45amiconnHope that is manageable with vbscript....
22:41:02amiconnAarrgh! Binary size is dangerously close to the limit on recorder :(
22:42:08preglowmarkun: you can't just convert kanjii to katakana, there are multiple meanings for every symbol
22:42:09phaedrus961I've noticed that too, maybe I can do some optimizations
22:42:34markunpreglow: I was afraid of that..
22:43:11amiconnphaedrus961: Player target doesn't build :(
22:43:31phaedrus961I've also just found a bug in convbdf which causes the offset table to be written for fonts that don't need it
22:43:41amiconn. /home/jens/rb-unicode/build/player/librockbox.a(id3.o): In function `_setid3v1title':
22:43:49amiconnid3.c:(.text+0x5a8): undefined reference to `_lcd_getstringsize'
22:45:11amiconnOh, and btw: why does id3.c need lcd code???
22:45:20phaedrus961ahh, I need to put the call to that function inside a #ifdef HAVE_LCD_BITMAP
22:45:33phaedrus961forgot, sorry
22:45:33 Quit webguest63 ("CGI:IRC (EOF)")
22:46:20phaedrus961it's only there to make sure the disk doesn't spin up on track change to load new glyphs
22:47:26amiconnHmm. id3.c is for reading and decoding id3 tags. Display code has nothing to do with that. Imho, this is dirty...
22:48:45phaedrus961where do you suggest it should be done?
22:50:00amiconnImho it would be best not to call display code from anywhere outside the gui thread. I'm afraid we have to make an exception in this case
22:50:26amiconnThe playback engine does the tag loading, so perhaps put it there?
22:51:28phaedrus961which file is that in?
22:52:02amiconnUnfortunately this means changing 2 places: apps/playback.c for iriver, and firmware/mpeg.c for archos
22:53:10phaedrus961ok, I'll look at it now
22:54:36amiconnYou used the lru cache from the chinese patch?
22:55:32amiconnI don't think #ifndef ARCHOS_PLAYER is the correct way
22:56:14amiconnChecking for a target plaform is the absolute exception, check for features instead
22:56:34amiconnIn this case I'd think #ifdef HAVE_LCD_BITMAP would be appropriate...
22:57:25phaedrus961that's an easy fix, I'll change that too
22:58:20amiconnI think we should bump the lng version byte when this gets committed
22:58:31amiconnNon-utf8 .lng file might cause confusion...
23:01:06preglowshall we keep the lang files in single byte format, or convert them to utf8 and require people to use unicode-aware editors?
23:01:16phaedrus961i agree, we'd have lots of people wondering why their player shows funny or missing chars
23:02:38amiconnpreglow: I think all .lng files should be converted to utf-8
23:02:48amiconn.lang I mean
23:03:54amiconnThe player unicode support is a bit half-hearted atm, keeping the internal intermediate 'rocklatin' encoding. Something to fix afterwards though
23:04:38amiconnIt could use proper utf-8 encoding, and assign the icons to a special range (iirc there is a range reserved in unicode for private use)
23:06:04 Join webguest12 [0] (
23:06:20phaedrus961the scroll code would also need to be changed to support utf8
23:06:45phaedrus961currently it assumes a single byte charset
23:06:50webguest12Hi guys, wiki spammer struck again,
23:07:01 Part webguest12
23:08:04amiconnphaedrus961: Oops. If that's true for bitmap displays too, that's a show-stopper imho
23:08:32phaedrus961no, only for player
23:08:56phaedrus961but it currently converts to latin1 first, so that scrolling works
23:09:22amiconnAh ok
23:09:30phaedrus961i meant that it would need to be changed if we support more than latin1 on player
23:19:52amiconnQuote from unicodec: "assume big endian" Huh?!
23:20:54amiconnSims are usually little endian, and probably we'll have your first little endian target in the not-too-far future (linuxstb?)
23:21:06phaedrus961where is that?
23:21:10linuxstbDepends how you define "not-too-far".
23:21:30amiconnphaedrus961: line 161
23:22:30preglowlinuxstb: any news on that front?
23:24:02phaedrus961that's because there's no bom in cjk strings and I haven't encountered any with little endian byte order
23:24:21linuxstbpreglow: I synced the latest IPL changes into my code - so at least now the LCD is working properly. But I'm still in the process of getting the ATA driver working.
23:24:34phaedrus961it has nothing to do with the cpu we're running on
23:24:45amiconnphaedrus961: Ah ok. Thought it had something to do with the target endianess...
23:25:38phaedrus961maybe I should change the comment?
23:26:48amiconnyes, probably
23:27:14linuxstbpreglow: The first success report installing IPL on the Nano has just been reported in #ipodlinux :)
23:27:27preglowis that so?
23:27:28amiconnI think unnecessary code should be removed rather than commented out (e.g. in bidi.c)
23:27:39preglowwhat parts of it are functioning well?
23:28:04linuxstbI think it should all be working. It seems almost identical to the latest 4G color ipods.
23:28:24linuxstbBut "podzilla" (the userland app) hasn't been ported to the new LCD yet - it doesn't use the kernel's LCD driver.
23:28:38phaedrus961amiconn: you mean the if(isspace(... line?
23:28:55linuxstbEvery app seems to have its own LCD driver (podzilla, Doom, iBoy etc)
23:29:34amiconnlines 163..165
23:29:50amiconnlinuxstb: Huh? Strange...
23:30:59linuxstbWell, the kernel just has a character based framebuffer driver. Podzilla uses microwindows (so a microwindows driver is needed), and the other apps must do their own thing.
23:31:16linuxstbNot a minimalist approach...
23:31:22preglowwhat, newer ipods support video now?
23:31:44preglowh.264, of all things, i'd hate to code that decoder
23:31:48linuxstbYep, using a nice proprietory Broadcom chip.
23:31:54phaedrus961amiconn: that's commented because it causes problems with the vkeyboard (even without unicode patch) and I don't know if it's necessary for bidi or not
23:32:00preglowlinuxstb: riiiight
23:32:09preglowlinuxstb: small wonder they manage to decode it, then
23:36:34amiconnpreglow: I'm not a codec expert, but I think mpeg-1 or mpeg-2 video could be possible on coldfire. Do you think that would be possible?
23:37:51linuxstbmpeg-1 has been around for a very long time - so it wouldn't surprise me if it was possible on the coldfire.
23:38:23linuxstbBut we need better audio decoding than we have now.
23:39:12amiconnphaedrus961: firmware/export/id3.h, line 95: the array size is changed from 32 to 92. Shouldn't that be 96?
23:39:41 Quit ender` (Read error: 104 (Connection reset by peer))
23:41:39phaedrus961it should be 90 for the string (30*3) plus one for the trailing null = 91, but I made it 92 because it originally had 2 extra bytes
23:42:09phaedrus961I don't know why it was originally 32
23:42:20linuxstbHow fast is the jpeg decoder? Any idea how many 160x128 images that could decode a second?
23:42:44preglowamiconn: yes, i do
23:42:48amiconnNot enough for video, atm
23:43:58amiconnphaedrus961: in font.c there are some variables which don't conform to rockbox style
23:44:39amiconn(e.g. bLongOffset) I know someone who might get very annoyed by such things
23:44:46preglowme, among others
23:45:50phaedrus961how should it be to conform to rockbox style?
23:46:04preglow1. skip the hungarian notation
23:46:07preglow2. no mixed case
23:47:25phaedrus961what is hungarian notation? the b?
23:48:12amiconnThis is used in a number of places in patched font.c, also n etc
23:48:17phaedrus961would long_offset be ok?
23:48:28preglowperfectly ok
23:49:39phaedrus961ok, i'll change that
23:49:50amiconnWhat's the glyph cache file used for?
23:50:51phaedrus961it stores which glyphs are in the lru at shutdown/font change and reloads them at font load
23:51:18linuxstbDoes this patch mean that all strings in Rockbox should now be utf-8 encoded?
23:51:26phaedrus961it was a feature a few people asked for
23:51:39phaedrus961linuxstb: yes
23:52:15linuxstbHave you changed the tag reading routines in metadata.c?
23:52:52amiconnSounds like a useful feature. Hopefully it can handle missing glyphs...
23:52:53preglowsome tags are already utf8
23:53:07linuxstbpreglow: Almost all are I think.
23:53:22amiconnid3v1 definitely not
23:53:23linuxstbThe code that decodes them to latin1 needs to be removed from metadata.c
23:53:35phaedrus961linuxstb: the utf8decode function has been removed for ape and ogg tags
23:53:42preglowso, are we close to utf8 in rockbox?
23:54:18phaedrus961if the patch is deemed acceptable, then yes
23:54:39preglowexcellent, excellent
23:54:51preglowgives me a fuzzy feeling insi
23:55:00linuxstbWhat about strlen - what is that meant to return for a utf8 string?
23:56:08*amiconn spots a place where memmove would really come in handy :/
23:56:12preglowstrlen and co should work on byte level
23:56:20phaedrus961linuxstb: it returns the length in bytes, we have a utf8length function to find the length in chars
23:58:10amiconnphaedrus961: The comment in id3.c lines 361ff needs fixing
23:58:11markunI once did a grep for strlen and started replacing some with utf8length. Hope I didn't change the wrong ones.
23:58:56linuxstbDoes FAT32 support utf8 filenames?

Previous day | Next day