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

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

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

#rockbox log for 2007-10-28

00:02:36krii have followed the guide/faq on the site, (manuall install)
00:02:40 Quit jhMikeS (Read error: 104 (Connection reset by peer))
00:02:56krinow, im doing it again replacing all files and doing it again, but it wan't start.
00:03:07n1skri: what is it that doesn't work?
00:03:18linuxstb_kri: Obvious question, but is it definitely an ipod video, not an ipod classic?
00:03:45kriyeah i defently orderd video because i didnt want classic
00:03:58n1skri: it doesn't start at all or rockbox doesn't strt?
00:04:06krirockbox dosent start
00:04:18linuxstb_What do you see on your ipod's screen?
00:05:03 Join jhMikeS [0] (n=jethead7@rockbox/developer/jhMikeS)
00:05:05 Quit bertrik ("bye")
00:05:08krino splash screen as i suppose to...
00:05:25krijust the usual apple ...
00:05:37linuxstb_So it freezes at the apple logo?
00:05:58krinow i have redone the manuall install and will see.
00:06:29 Quit linuxstb_ ("ChatZilla [Firefox]")
00:07:09 Join zerdik [0] (
00:09:37 Quit Robin0800 (Read error: 110 (Connection timed out))
00:10:06Zagorbloody tax forms. next year we'll accept no US gsoc participants...
00:10:13n1skri: does ipodpatcher report success?
00:10:37 Quit davina (Remote closed the connection)
00:10:58jhMikeShaha...the IRS has a long arm huh?
00:11:31jhMikeSor is it tax on your end because of US participants?
00:12:12 Join saratoga [0] (i=9803c6dd@gateway/web/cgi-irc/
00:12:33Zagoryeah, google (well, I guess the IRS) requires that we fill out a W8BEN tax form. and to do that, we need to fill out a W7 form and apply for a TIN number... ain't bureaucracy great!
00:13:18*linuxstb blames saratoga ;)
00:13:19n1sat least it makes sure a lot of people have jobs :-)
00:14:22linuxstbZagor: So you'll receive the money in about 2010?
00:15:07krik. more stupid question here. the text is very small. can i rezise it?
00:15:09Zagoryeah that's another thing. they want an address to mail the check. a CHECK??? what is this, 1974?
00:15:47linuxstbkri: Did you install the fonts package? (i.e. unzip to your ipod)
00:16:10peturZagor: in the US, it is 1984 ;)
00:17:09saratogafor some reason the checks are very popular in the US
00:17:18saratogai have no idea why
00:17:53saratogaanyway, i suspect you'd have to deal with the IRS anyway, since goolge is in the US
00:18:32Zagoractually if you have no us activity (mentors or students) you can just sign a single google form and be done with it.
00:19:41Zagorbut I guess banning US people would be considered ... unfriendly :)
00:20:11linuxstbAnyone know how/if I can do a recursive diff between two different directories, ignoring .svn folders?
00:21:24 Quit donutman25 ("ChatZilla [Firefox]")
00:21:51 Join fm2 [0] (n=chatzill@
00:21:58*linuxstb discovers "-x .svn" and thinks it works
00:22:17*petur only knows a windows tool - winmerge
00:22:50fm2Is there any system in error codes in buffering API? I think it'snot good that just numbers (e.g. -2) are used and not constants.
00:23:21fm2I mean named constants. Which would make reading and maintaining code easier.
00:24:03peturfm2: then do not look at disk.c/file.c/fat.c/... ;)
00:25:08fm2petur: that was made in a prehistorical epoche when we could not write programs the right way! ;-) But now...
00:25:38fm2It's 21-th century after all!
00:25:44 Quit Toxicity999 (Remote closed the connection)
00:26:47n1sfm2: I'm sure a patch with logical error names would be accepted :-)
00:27:59fm2n1s: but for that to be possible, a system must exist in the codes. Does it? Does anybody know?
00:28:06amiconnThe numbers are constructed in a logical way. I would seriously dislike named errors for stuff that's not normally user visible
00:28:11jhMikeShow bout assigning ranges to subsystems?
00:28:21 Join fakeplasticpete [0] (
00:28:22amiconnIt would just be bloat. This isn't Windows...
00:28:27jhMikeSamiconn: bloat?
00:28:30n1sfm2: Nico_P is the person to ask
00:28:52jhMikeSintegers are bloat?
00:28:56fm2amiconn: but if I see "if (rc == -2)" I ask what -2 means
00:29:16amiconnfm2: The error numbers are (or are meant to be; not sure whether the playback stuff adheres to these rules) providing a backtrace
00:29:25peturjhMikeS: he means the 0x80408020 stuff the MS uses
00:29:26jhMikeSfm2: I agree. Only certain select things need strings but the source could be nicer with #define constants
00:30:27fm2Constants would make code more readable and would also enforce (and ensure) some discipline and system in error codes
00:30:38amiconnEach function that returns an error due to a downlevel function failing returns both its own error and the downlevel function's error code, multiplied by 10
00:30:38ZagorI think we should simply assign a US person to recieve the money and put them in the paypal account. this tax form business is silly.
00:31:01amiconnAll error codes are negative, and this system allows for 9 different error codes per function
00:31:05jhMikeSpetur: sure, and that would also help code be more robust sine you'd have some idea what's going on.
00:31:47linuxstbZagor: Wouldn't that require the US person to do tax things?
00:32:38jhMikeSI think the system now is awful and I'd never use that myself. At least common errors would have defined constants too. ERROR_DISK_FULL :)
00:32:41Zagorlinuxstb: not as I understand it:
00:32:52kriso what happend with the music that was on? have i owerwrite it? i think not, there is used space...
00:33:08fm2 amiconn: I don't quite like this approach but if it's the way the things are done in RB then Ok. The approach mixes up levels. Besides, it's only usable if you know the 'deepness' of the function call.
00:34:12Zagorfm2: mixes levels?
00:34:13jhMikeSnot to mention FILE_NOT_FOUND or something is just that no matter what system returns it
00:34:58fm2Zagor: propagating error code from a lower level to a higher level where it can't (and shouldn't) be interpreted anymore
00:35:47Zagorfm2: the fact that users can sometimes see these codes has helped us a lot in the past
00:35:50jhMikeSthe e200 ata driver was a real treat for the current approach
00:35:51amiconnI didn't invent this system, but I this it makes very much sense
00:36:04amiconn*I think
00:36:45ZagorjhMikeS: file not found isn't exactly what the purpose of these error codes is for
00:36:53amiconn(just that I would have chosen base 16 instead of base 10, and then displayed the error code in hex. << 4 is cheaper than * 10
00:37:04fm2jhMikeS: but I think each subsystem should have its own "error code domain", with different subsystems possibly having same values. I.e. the values don't necessarilyhave to be globally unique
00:37:26jhMikeSit could be...and we could pack it as hex so more codes can be used...if displayed, a small decoding function would be nothing
00:38:18amiconnfm2: The advantage of the current system is that it provides a backtrace. With unique error codes per system, you wouldn't have that anymore
00:38:54fm2jhMikeS: but if you have e.g. -52 how do you know which func added 5 and which 2? In order to interpret the value you have to know the call stack.
00:39:08jhMikeSI guess each has it's own advantage and I like aspect of both to have the cake and eat it too?
00:39:29fm2amiconn: 52 does not provide the call stack IMHO
00:40:00krii moved the 'music' folder to the root for rockbox. and now i loaded the database in the ipod, is that right?
00:40:00amiconnfm2: You check the function which produced the error code, and then check which place in it generated -2. Then you check what call it did before that. This function generated -5
00:40:32kriso i just move the music now to the folder? or?
00:40:46n1samiconn: if the errors are just #defined constants we could still print the number to get the backtrace but the code could be like if(ret == ERROR_FOO)
00:41:04Zagorn1s: but that wouldn't help us find bugs
00:41:31n1sZagor: no but it makes the code easier to read which I believe was the argument for
00:41:34fm2amiconn: correct. What I meant is the return value alone does not provide the call stack, it has to be found out manually (as you described)
00:41:35jhMikeSBugs? There's bugs? :)
00:41:58Zagorn1s: the error code is there for finding bugs, not presenting pretty strings to users.
00:42:11fm2Zagor: I think easy to read code is 50% to bug free code since bugs can be spotted easier
00:42:25jhMikeSI've had use for handling some things the other way...and being able to return errors to show splashes in higher level code rather than lower.
00:42:51Zagorfm2: what does easy to read code has to do with this?
00:43:17 Quit darkless (Connection timed out)
00:43:31Zagorthe point is you know what went wrong in every step of the call chain. you will never get that with "pretty" error codes.
00:43:41n1sZagor: I don' t mean that we should display them just #define the numbers we use to identifiers that say something so I don't have to read hundreds of lines of code to find out what an error means
00:43:55fm2Zagor: you can easily write -1 instead of -2 in "if". And that's hard to notice. But if you have named (per #define) consts, you'll see it
00:43:57jhMikeSbut a thread could save an error code in TLS
00:44:16Zagorfm2: these values are not checked. only (if <0)
00:44:31fm2n1s: exactly my thoughts. I'm not proposing displaying errors as strings
00:44:48Zagorn1s: sure it could be nice. but rather a heavy load to maintain.
00:45:07n1sZagor: fair enough :)
00:45:08fm2Zagor: I've seen some places with "if (xxx == -2)"
00:46:18fm2Zagor: just "xxx < 0" would be ok (yes/no logic), but if concrete values are checked I refer named consts
00:46:29Zagorfm2: I bet there are a few places, but that's not the base of the concept
00:46:34jhMikeSfm2: yeah, I hate that. some errno thing would be nice is many situation as a supplement to if (retval < 0)
00:46:43 Join darkless [0] (
00:47:17fm2jhMikeS: that I did't want to say. I disagree in this point.
00:47:20Zagorchecking specific values is really flirting with bugs
00:47:46jhMikeSfm2: in my point?
00:48:06fm2jhMikeS: yes, "some errno as a supplement"
00:48:27fm2jhMikeS: it's not thread safe
00:48:34fm2jhMikeS: just for one
00:49:53fm2Sorry, I must leave
00:49:58 Part fm2 ("Bye")
00:51:06linuxstbamiconn: Do you know if there are any examples in Rockbox of two different devices (with different CPUs) using the same LCD controller, and sharing a driver?
00:51:54amiconnNo, and most often it wouldn't make sense
00:52:18amiconnE.g. the H300 and the (type 1) Color and the Nano have the same lcd controller
00:52:27 Join jhulst_ [0] (n=jhulst@unaffiliated/jhulst)
00:52:33amiconnSame goes for X5 and big H10
00:52:43 Quit jhMikeS (Nick collision from services.)
00:52:49 Join jhMikeS [0] (n=jethead7@rockbox/developer/jhMikeS)
00:53:10amiconnAnd the same was also true for the (dropped) gmini port - it had SSD1815 like the recorder
00:53:27amiconnBack then, it actually *was* a single driver, but that was before target tree
00:53:39jhMikeSargh...disconnect...and "not thread safe"...coming from your's truly. lol.
00:54:36kri'undefinde instruction at 00125A90 (0='
00:54:37amiconnIf there's something controller specific that can be factored out, and it's more than just a few lines, we could introduce firmware/drivers/lcd
00:55:12amiconnBut most often the hookup is so different that there isn't much that could be shared
00:55:22krimy new ipod is broke. :(
00:55:45amiconnEven H1x0 and M5 do *not* share a driver - and they *do* have the same cpu, and the same lcd controller
00:56:05amiconn(well, almost same cpu)
00:56:20linuxstbamiconn: For my MP3/DAB player, the only differences with the Archos SSD1815 driver are lcd_device_init(), lcd_write_command() and lcd_write_data(). But that doesn't leave very much anyway, so maybe just copy and paste is easier.
00:57:28amiconnI bet lcd_default_contrast() is also different...
00:57:37linuxstbYes, forgot that...
00:57:52amiconnOh, and the flip stuff too - unless your target is also 112x64
00:58:01linuxstbMy current implementation is a copy of the driver in my player's target-tree.
00:58:04amiconnOtherwise the xoffset will be different
00:58:16linuxstbThe xoffset is just defined as 132-LCD_WIDTH
00:58:18amiconn(the controller is 132x64)
00:58:28linuxstbMy width is 128
00:58:42amiconnAh, yes
01:00:42amiconnIf you compare lcd-h100.c and lcd-m5.c, you'll notice that they're also mostly c&p
01:00:43 Quit saratoga ("CGI:IRC (EOF)")
01:00:50jhMikeSamiconn: somehow we manage to pull it off for tuners. I can see a similar way for lcd.
01:01:12linuxstbI think I'll just keep it separate as well, at least for now.
01:01:20amiconnThe init sequence is different, and there are some more due to different gpio pins and different hookup of the lcd controller's A0 bit
01:01:27keanurasher, around?
01:01:37rasherkeanu: yeah
01:02:01rasherAnd I'd appreciate it if you'd leave me a message other than "around".. Just ask your question next time.
01:02:05amiconn(on H1x0 it's on gpio while on M5 it's on the coldfire's A1 bit)
01:02:32keanurasher, is there anything in the Makefile that needs to be changed, other than taking out "-g" to reduce the sim file size?
01:03:20rasherkeanu: I left you a message in the logs..
01:03:44keanurasher, oh, ok. I'll look back through them
01:03:47 Quit darkless (Read error: 113 (No route to host))
01:05:51keanurasher, thanks
01:06:05 Quit jhulst (Connection timed out)
01:06:30rasherRemoving -g helps a bit, but stripping removes even more
01:06:45linuxstbamiconn: On a slightly different subject, I've been helping TMM try to figure out the LCD driver in his iaudio 7 (a 160x128 colour LCD), and the code in the original firmware we think is the lcd driver does something which strikes me as strange - it enables/disables a GPIO pin before/after writing to a register in the lcd controller. Have you seen that in other LCD drivers?
01:07:05amiconnI just mentioned one... H1x0
01:09:02 Quit kri ("quit")
01:09:04amiconnThe Epson controller's A0 bit selects control or data. On H1x0 it's hooked to gpio.
01:09:49linuxstbThanks. grep missed it as it's in the asm file...
01:10:10 Quit spiorf (Remote closed the connection)
01:10:10TMMthat makes sense
01:11:00linuxstbTMM: So I'm now willing to bet money we've found the lcd driver - you just need to implement it now ;)
01:11:30amiconnIt's a bit nasty - and one of the spots where I ran into problems wrt buffered writes. Cowon's choise to use an actuall address bit on M5 is definitely the better one
01:12:04 Quit Zagor ("Client exiting")
01:12:12amiconnBut there are numerous weird design choices in our targets....
01:12:33*amiconn can't type :/
01:15:26 Quit HellDragon (Client Quit)
01:15:45*n1s finds it interesting/annoying that vorbis performance can vary quite a lot with changes in the core that don't affect codecs at all...
01:16:38n1sright after tomal's latest optimization i got 341% realtime on my testfile, 3 days ago 319, today 326...
01:17:09n1swhich also means that my asm nets 0 speed gain/loss :-/
01:17:10jhMikeSn1s: one big one on cf was when the codec buffer lost alignment
01:17:16amiconnCache aliasing...
01:17:35***Saving seen data "./dancer.seen"
01:17:59n1samiconn: when I tested different builds with tremor i also tested mp3 and that was unaffected...
01:18:32 Quit moos (Read error: 110 (Connection timed out))
01:18:40n1sjhMikeS: is it aligned now?
01:18:45amiconnNot all codecs would suffer from this. Even those which potentially do wouldn't change their performance in a consistent way
01:19:02jhMikeSn1s: last I looked into it I made sure it was line aligned
01:19:15 Join HellDragon [0] (n=Nocebo@unaffiliated/helldragon)
01:19:41jhMikeSthe performance difference was about 28% boost to %12 boost or so (no test_codec then)
01:19:42amiconnIt's just that one time critical function which resides in dram calls another one, which is also in dram. If they (partially) map to the same cache line(s), one throws the other out of the cache
01:20:05n1samiconn: is there anyway to predict/benefit from it without having to run everything through an emulator?
01:20:13amiconnThe coldfire instruction cache is direct mapped, so this effect is often visible, unlike on PP, where the cache is set associative
01:20:25 Join kri [0] (
01:20:42jhMikeSn1s: yes, align code is still there
01:21:09amiconnThe only way to strictly avoid this is to put all time critical functions in iram - but iram is more important for data
01:21:32n1samiconn, jhMikeS: any idea which function is the culprit or how to find out which one it is?
01:21:51kriplease, it worked and i listen to the ipod and it stoped, flashing 'undefined instruction at [nr] (x)' and its there i can't do nothing. can't restart the ipod, turn it off, or load it on the comp, what to to? is it broke?
01:22:20n1skri: press and hold menu+select
01:22:30krinothing happen
01:22:33amiconnn1s: If you have any idea which functions are time critical, and know all the functions it calls, or is called from, you can check the map file
01:22:52linuxstbkri: Hold it for a few seconds - it NEVER fails.
01:22:58amiconnBut doing this is rather difficult...
01:23:06krii can hold it for a min
01:23:29n1samiconn: I guess I will have to dig a bit in the code to see what it calls often and go from there then...
01:23:45linuxstbkri: Then you're doing it wrong... Press MENU and SELECT (the centre button) together and hold them down for about 5 seconds.
01:23:59amiconnYou might find out by using profiling - if you can get that to work
01:24:09*amiconn *never* tried it
01:24:42kriwhat went wrong i almost start crying. :/
01:25:00amiconnkri: Make sure Hold is disabled, otherwise the hard reset won't work
01:25:02n1samiconn: I actually tried it and the code runnign on target seemed to work but the script couldn't map the adresses to symbols...
01:25:26 Join atsea-32 [0] (i=atsea-@gateway/tor/x-d272899186ee6ed9)
01:25:47kriits okey now. i need to now why it went wrong... was it because i moved music folder from root to .rockbox ?
01:26:11n1skri: undefined instruction is a bug
01:26:21 Join Toxicity999 [0] (n=bryan@unaffiliated/Toxicity999)
01:26:56 Join Robin0800 [0] (
01:32:11kkurbjunamiconn, JdGordon was talking to you before about the lists that were not showing properly on the M:Robe, I thought I found the problem, when I move the character buffer out of the list drawing function (make it a global) it works, the list drawing shows properly. I thought this might be a stack overflow, but as I look more into it it apears that that is not the problem either, there is more than enough stack left when I look at a
01:32:19 Join darkless [0] (n=darkless@
01:32:49 Join JdGordon [0] (
01:34:21kkurbjunJdGordon, I found the problem with the lists, but I don't know /why/ it's happening yet
01:34:35JdGordon:) well its a start.. whats the problem?
01:34:49kkurbjunand I have a workaround, but I'm afraid that it might just be masking a larger problem
01:35:12kkurbjunin list.c on line 315 there's a buffer defined
01:35:22kkurbjunif you move it out of the function as a global the lists will work
01:35:59kkurbjunif you look at the logs right before you came in there's an explanation of my debugging process so far
01:36:25JdGordonmaybe it is a stack problem then?
01:36:39kkurbjunI don't think it is, there appears to be more than enough stack
01:36:56kkurbjunand I've also tried increasing the stack space/decreasing the buffer size
01:37:58 Quit darkless (Read error: 104 (Connection reset by peer))
01:38:49JdGordonhaha i logged in 2 min too late :p
01:39:41JdGordonthat buffer doesnt make sense though... why would some lists work and some not?
01:40:28kkurbjunI don't think any lists work
01:40:35kkurbjunregardless of the type
01:40:38 Join psycho_maniac [0] (
01:40:39JdGordonnone work for you?
01:40:50kkurbjunwell, the menu's work, but those aren't lists
01:40:55JdGordonyes they are
01:40:55Robin0800any one know why the ipod video is stuck on 100% boost with latest svn?
01:41:13kkurbjunthey use the settings_list code?
01:41:15 Quit hcs ("Leaving.")
01:41:38kkurbjunI mean the option_select code
01:41:44kkurbjunnone of that works for me
01:41:59JdGordonthey all end up getting drawn by the same code in list.c
01:42:40kkurbjunI think it's a problem in the option_get_valuestring flow, but I don't know why it would only be showing in hardware
01:43:13kkurbjunthe callback_get_item_name for settings that is
01:43:31linuxstbamiconn: contains what we think is the lcd init sequence for the iaudio 7, and I was wondering if you recognised it. The LCD is 160x128 and advertised as 18-bit colour.
01:45:21jhMikeSlinuxstb: I know you didn't ask me but that's renesas stuff
01:45:40TMMjhMikeS: renesas?
01:46:10linuxstbjhMikeS: Sorry for not asking you ;)
01:46:12jhMikeSThe driver manufacturer. HD66789, HD66773
01:46:25TMMwow, cool
01:46:36TMMso, it was pointless to RE this then? :)
01:47:18linuxstbYou had to RE it to discover the controller...
01:47:30TMMI suppose
01:48:00 Join darkless [0] (n=darkless@
01:51:04amiconnLooks like the 789R
01:51:47JdGordonkkurbjun: im comiling with your suggestion...
01:51:50 Quit TotallyInfected ()
01:52:12TMM789R? what brand is that?
01:52:43 Join TotallyInfected [0] (
01:52:50jhMikeSalready said :)
01:52:52amiconnRenesas HD66789R, the same thing that's e.g. used in the iriver H300
01:53:19TMMO, awesome
01:53:20linuxstbIs there a datasheet in the wiki?
01:53:25JdGordonkkurbjun: no change...
01:53:33TMMso, we 'just' have to port it over? :)
01:53:36jhMikeSthe e200 uses that too
01:53:38amiconnThe 773 doesn't have R10
01:53:56amiconn(0x10 I mean)
01:54:56JdGordonmoving that entry_buffer out of the function, the same lists dont show
01:55:57 Quit japc (Remote closed the connection)
01:56:07jhMikeSooh, that reminds me, theres a delay in the e200 that's probably not required now
01:57:09kkurbjunhmm, works for me - I have to go now though, I'll look into it more later
01:57:57amiconnTMM: Looks like there is a typo in your init...
01:58:13TMMamiconn: only one?
01:58:23amiconn0x34 near the end of lcd init 2 should probably be 0x43
01:58:43TMMI thought I had fixed tat
01:59:02TMMI have fixed that
01:59:05TMMthat's an old paste :)
01:59:47linuxstbTMM: But this probably means you don't have to RE the lcd_update functions - just copy them from elsewhere in Rockbox.
02:00:21TMMhurrah! :)
02:01:03amiconnPerhaps not
02:01:10amiconnThe e200 does some dma trickery
02:01:44amiconnI'm not sure how that works, and it probably depends on how the controller is hooked up
02:02:12jhMikeStrickery? running DMA while trying to send commands would scramble them.
02:02:34jhMikeSIt's in ram write mode normally and uses HWM
02:02:41linuxstbThe H300 is also using DMA...
02:02:56amiconnHmm, but in a different way.
02:03:05TMMwell, datasheet is nice :)
02:03:09jhMikeSbut it's not a memmapped framebuffer
02:03:42amiconnjhMikeS: Does the e200 use the PP's dma engine for the lcd updates?
02:03:51jhMikeSnot really memmapped, but in-memory. the display is constantly refreshed from RAM
02:04:12jhMikeSamiconn: A special one similar to the controller on gigabeat
02:04:34amiconnHmm, and that pays off? Constant ram access costs bandwidth, and power...
02:05:23jhMikeSnot for these designs apparently
02:05:34amiconnTMM: The lcd_write_reg() suggests that the hookup uses the 9 bit interface mode
02:06:11 Quit psycho_maniac (" HydraIRC -> <- Chicks dig it")
02:06:21 Quit TotallyInfected ()
02:06:29amiconnBtw, the HD66789R is just 16 bit −− advertising people.... :\
02:06:43 Join TotallyInfected [0] (n=ebola@
02:06:45 Quit ender` (" Two things are infinite: the universe and human stupidity, even though I'm not yet sure about the universe. -- A. Einstein")
02:06:46n1show often should a codec yield? Tremor seems to do it about 100 times per second...
02:07:11TMMamiconn: let's sue them for false advertisement! I want my 2 bits of extra color information damnit!
02:07:11jhMikeSamiconn: that is rather annoying re: the color depth :)
02:07:16linuxstbn1s: Good question - we should probably have some guidance for that...
02:08:03amiconnI think 100 times per second is about right - once per tick
02:08:06jhMikeSn1s: that sounds about good to me
02:08:20 Quit cooz ()
02:08:47 Quit Robin0800 (" HydraIRC -> <- Like it? Visit #hydrairc on EFNet")
02:09:00jhMikeSI'd do more speed thread response elsewhere. mostly it will just return to the codec immediately.
02:09:29n1sok, just checking :-) I remember the midiplayer yields when the buffer is full which could take about 0.4 seconds to fill :-) I guess that will change when it's turned into a codec
02:09:56amiconnlinuxstb: The ipl (and rockbox) bcm driver is definitely based on the rom code, not full OF code
02:09:58jhMikeSfor the buffer to fill?
02:10:18amiconnThat magic constant (0xfffa0005) appears several times in the rom, but nowhere in the OF
02:11:19 Quit petur ("Zzzzz")
02:11:20n1sjhMikeS: the buffer is that long so whenever it throws a buffer miss it took longer than 0.4 doesn't happen very often now but it can't yield much more than 4-5 times a second
02:12:07linuxstbamiconn: My guess is that the Broadcom chip is running different firmware when the OF is running, so it wouldn't surprise me if the LCD API is completely different - does your research agree with that?
02:12:09jhMikeSthat's massive starvation for the system. the yields are very fast.
02:12:24amiconnlinuxstb: So far it looks like that, yes
02:14:29*jhMikeS thinks it's telling that his favorite targets to listen to use the SW tone controls
02:14:37linuxstbAttempting to load the "OF" broadcom firmware could be interesting...
02:15:55*jhMikeS wants to be able to use SW tone controls whether or not HW ones are present
02:16:55amiconnlinuxstb: That would first require to find out how the bcm on-disk firmware is stored, how it's supposed to be loaded into the chip, and run
02:17:13linuxstbamiconn: DId you miss my reply earlier? It's simply a FAT16 filesystem.
02:17:36amiconnYes, but that doesn't tell anything about how the image is loaded
02:17:41linuxstbBut that doesn't answer the second question - how to transfer it..
02:18:11linuxstbI'm not saying it's simple ;)
02:18:16amiconnIs it encrypted or scrambled? Is there some header that splits the image into blocks which need to be loaded at different addresses?
02:25:18 Join darkless_ [0] (
02:33:33 Quit darkless (Read error: 110 (Connection timed out))
02:33:35 Quit TotallyInfected (Remote closed the connection)
02:34:15 Join TotallyInfected [0] (
02:35:07amiconnlinuxstb: I found the ipod HW revision code on the G5.5
02:35:41 Quit lee-qid ("aufwiederbyebientotsayonara")
02:35:41amiconnIt's at 0x405c in the ROM, instead of 0x2084 like on all other ipods
02:37:36amiconn8 bytes before that is some kind of signature for that value
02:38:06amiconn'rVwH' - read as little endian 32 bit it becomes 'HwVr'
02:41:12amiconnAnyone with a plain G5 (not G5.5) around?
02:41:28krizzz. switch theme and it went black apperently missing some files
02:43:10 Join webguest52 [0] (i=5932d76c@gateway/web/cgi-irc/
02:47:24linuxstbamiconn: "0004050: 0000 0000 7256 7748 0000 0000 0500 0b00 ....rVwH........" (from my 5g's flash dump)
02:48:10amiconnSo it's also at 0x405c: HW version 0xb0005
02:48:19amiconnThis G5.5 is 0xb0011
02:48:39amiconnThat means we can fix our HW version check for G5/G5.5
02:48:59linuxstbIt's in firmware/export/hwcompat.h
02:53:27amiconnHmm. Nasty ifdefing...
02:59:49 Quit n1s ()
02:02:59 Join Strath [0] (
02:06:57 Join magix [0] (
02:09:43magixhey can someone help me with watching movies on ipod nano?
02:10:40karashatadepends, what's your problem?
02:12:10 Join kri_ [0] (
02:17:37***Saving seen data "./dancer.seen"
02:18:02 Nick fxb__ is now known as fxb (
02:19:27 Join keen- [0] (
02:20:50keen-has anyone managed to get rockbox working on sansa's c100 series as well as c200?
02:22:38linuxstbkeen-: The c100 is completely different hardware to the c200
02:23:02linuxstbBut there was someone around here earlier today who expressed an interest in porting Rockbox to the c100.
02:24:09keen-hmm ok
02:30:37 Quit kri (Read error: 110 (Connection timed out))
02:33:24 Quit Toxicity999 (Read error: 104 (Connection reset by peer))
02:33:57 Join Toxicity999 [0] (n=bryan@unaffiliated/Toxicity999)
02:41:34 Join hcs [0] (n=agashlin@rockbox/contributor/hcs)
02:41:53 Join Craig [0] (
02:43:43CraigI've got a question about the mpegplayer on the sansa e280: With recent releases, the video plays much faster than the sound, anybody else experiencing this?
02:44:00linuxstbHave you turned off the "limit fps" option?
02:44:20CraigOh, uh .... dunno, I should check that?
02:44:34magixHow do i convert my ipod into winpid?
02:44:51magixHow do i convert my ipod into winpod?
02:45:24linuxstbmagix: With the instructions here -
02:46:32magixk thnx
02:51:09 Quit animeloe ("Leaving")
02:52:30 Quit webguest52 ("CGI:IRC (EOF)")
02:53:04magixcan i put videos on a nano with xp?
02:53:18 Join TotallyInfected_ [0] (
02:54:05karashatamagix: of course, as long as they're encoded properly for Rockbox to be able to play them
02:54:23magix<karashata> can i pm u?
02:55:14 Join animeloe [0] (n=animeloe@unaffiliated/animeloe)
02:55:50linuxstbmagix: Instructions for playing videos in Rockbox are here -
02:56:40magixye but i dont have any of the settings i need it from step 1
02:56:47 Quit animeloe (Client Quit)
02:58:40 Part Craig
03:02:49 Quit kugel ("Benutzer ist abwesend.")
03:03:21 Join kugel [0] (i=kugel@unaffiliated/kugel)
03:03:21 Quit TotallyInfected (Read error: 110 (Connection timed out))
03:03:37 Join kripso [0] (
03:08:57 Join Craig [0] (
03:09:33 Join animeloe [0] (n=animeloe@unaffiliated/animeloe)
03:09:51magixcan someone help me with installing linux on ipod nano?
03:10:01CraigFPS limit fixed the sync issue, thanks.
03:10:38 Part Craig
03:14:57 Join beetlejuice [0] (
03:16:22 Nick fxb is now known as fxb__ (
03:17:59magixdoes this work on 2nd generation ipod nanos?
03:24:18 Quit fakeplasticpete (
03:24:18 Quit [omni] (
03:24:18 Quit Strath (
03:24:18 Quit zicho (
03:24:18 Quit thegeek (
03:24:18 Quit XavierGr (
03:24:18 Quit TheDarkOne[lappy (
03:24:18 Quit GodEater_ (
03:24:18 Quit chandlerc (
03:24:18 Quit billytwowilly (
03:24:18 Quit ATravelingGeek (
03:24:18 Quit GodEater (
03:24:18 Quit _pill (
03:24:18 Quit homielowe (
03:24:18 Quit karashata (
03:24:18 Quit mirak (
03:24:18 Quit Thundercloud (
03:24:18 Quit freqmod_nx (
03:24:18 Quit ze (
03:24:18 Quit sd (
03:24:18 Quit Cronoh (
03:24:18 Quit Soap (
03:24:18 Quit Seed (
03:24:18 Quit sounddude (
03:24:18 Quit tchan (
03:24:18 Quit Zom (
03:24:18 Quit marcosou1ce (
03:24:18 Quit annulus_ (
03:24:18 Quit Galois (
03:24:18 Quit w0rd54 (
03:24:18 Quit Presence (
03:24:18 Quit TotallyInfected_ (
03:24:18 Quit DogBoy (
03:24:18 Quit JRoT (
03:24:18 Quit krazykit (
03:24:18 Quit FOAD (
03:24:18 Quit courtc (
03:24:18 Quit gtkspert (
03:24:18 Quit Nimdae (
03:24:18 Quit lostlogic (
03:24:18 Quit jepler (
03:24:18 Quit puzzles (
03:24:18 Quit micols (
03:24:18 Quit Toxicity999 (
03:24:18 Quit kri_ (
03:24:18 Quit darkless_ (
03:24:18 Quit JdGordon (
03:24:18 Quit jhulst_ (
03:24:18 Quit zerdik (
03:24:18 Quit SirFunk (
03:24:18 Quit Bagder (
03:24:18 Quit linuxstb (
03:24:18 Quit Isolinear (
03:24:18 Quit J (
03:24:18 Quit nomel (
03:24:18 Quit RaRe (
03:24:18 Quit daurnimator (
03:24:18 Quit crashd (
03:24:24 Quit shodanX (
03:24:24 Quit ivan` (
03:24:24 Quit pabs (
03:24:24 Quit blithe (
03:24:24 Quit jurrie (
03:24:24 Quit Weiss (
03:24:24 Quit wookey_ (
03:24:24 Quit tuplanolla (
03:24:24 Quit DiDjCodt (
03:24:24 Quit kugel (
03:24:24 Quit atsea-32 (
03:24:24 Quit andrewg867 (
03:24:24 Quit inakimhnh (
03:24:24 Quit TMM (
03:24:24 Quit bb (
03:24:24 Quit Rick (
03:24:24 Quit newbyx86 (
03:24:24 Quit dionoea (
03:24:24 Quit joshin (
03:24:24 Quit Echelon (
03:24:24 Quit crashmatrix (
03:24:24 Quit bagawk (
03:24:24 Quit n17ikh|Lappy (
03:24:24 Quit Ave (
03:24:24 Quit magix (
03:24:24 Quit HellDragon (
03:24:24 Quit miepchen^schlaf (
03:24:24 Quit kkurbjun (
03:24:24 Quit idnar (
03:24:24 Quit alienbiker99 (
03:24:24 Quit delYsid (
03:24:24 Quit Dave2 (
03:24:24 Quit rvvs89 (
03:24:24 Quit in-jane (
03:24:24 Quit maxkelley (
03:24:24 Quit beetlejuice (
03:24:24 Quit amiconn (
03:24:24 Quit scorche (
03:24:24 Quit illissius- (
03:24:24 Quit rotator (
03:24:24 Quit midgey (
03:24:24 Quit sneakums (
03:24:24 Quit solatis (
03:24:24 Quit BHSPitMonkey (
03:24:24 Quit sslashes (
03:24:24 Quit EnterUserName (
03:24:24 Quit Bjoern-Erik (
03:24:24 Quit feisar (
03:24:24 Quit guyzmo (
03:28:14 Join bb [0] (
03:28:15***Server message 505: 'logbot :Private messages from unregistered users are currently blocked due to spam problems, but you can always message a staffer. Please register! ( )'
03:29:53 Quit hcs ("Leaving.")
03:38:59 Quit J3TC- (".UPP.")
03:41:52 Join webguest94 [0] (i=47fece89@gateway/web/cgi-irc/
03:42:50 Quit webguest94 (Client Quit)
03:52:32NJoinalienbiker99 [0] (
03:52:32 Join bagawk [0] (
03:52:32 Join dionoea_ [0] (
03:52:32 Join hcs [0] (n=agashlin@nat01-southtowers-ext.Rutgers.EDU)
03:52:32NJoinatsea-32 [0] (i=atsea-@gateway/tor/x-d272899186ee6ed9)
03:52:32NJoinkkurbjun [0] (
03:52:32 Join Rick [0] (i=rick@unaffiliated/rick)
03:52:32NJoinnewbyx86 [0] (
03:52:32 Join delYsid [0] (n=user@debian/developer/mlang)
03:52:32NJoinEchelon [0] (
03:52:32 Join rvvs89 [0] (n=rvvs89@pdpc/supporter/active/rvvs89)
03:52:32NJoinn17ikh|Lappy [0] (
03:52:32NJoinAve [0] (
03:52:33***Server message 505: 'logbot :Private messages from unregistered users are currently blocked due to spam problems, but you can always message a staffer. Please register! ( )'
03:52:33***Server message 505: 'logbot :Private messages from unregistered users are currently blocked due to spam problems, but you can always message a staffer. Please register! ( )'
03:52:33***Server message 505: 'logbot :Private messages from unregistered users are currently blocked due to spam problems, but you can always message a staffer. Please register! ( )'
03:52:36***Server message 505: 'logbot :Private messages from unregistered users are currently blocked due to spam problems, but you can always message a staffer. Please register! ( )'
03:53:01 Join Echel0n [0] (
03:53:26 Quit n17ikh|Lappy (Killed by (Nick collision))
03:53:38 Join n17ikh|Lappy [0] (
03:53:38 Join TMM_ [0] (
03:53:38NJoin[omni] [0] (
03:53:38 Join rvvs89_ [0] (
03:53:46NJoinamiconn [0] (n=jens@rockbox/developer/amiconn)
03:53:46NJoinscorche [0] (i=Blah@rockbox/administrator/scorche)
03:53:46NJoinillissius- [0] (
03:53:46NJoinrotator [0] (n=e@rockbox/developer/rotator)
03:53:46NJoinmidgey [0] (
03:53:46NJoinsneakums [0] (
03:53:46NJoinsolatis [0] (
03:53:46NJoinBHSPitMonkey [0] (n=stephen@unaffiliated/bhspitmonkey)
03:53:46NJoinsslashes [0] (n=sslashes@
03:53:46 Join EnterUserName [0] (n=dave@pdpc/supporter/student/GeekZoid)
03:53:46NJoinBjoern-Erik [0] (
03:53:46NJoinfeisar [0] (
03:53:46NJoinguyzmo [0] (
03:53:47***Server message 505: 'logbot :Private messages from unregistered users are currently blocked due to spam problems, but you can always message a staffer. Please register! ( )'
03:53:51NJoindionoea [0] (
03:53:51 Join idnar [0] (
03:53:51NJoinStrath [0] (
03:53:51NJoinzicho [0] (
03:53:51NJointhegeek [0] (
03:53:51NJoinTheDarkOne[lappy [0] (n=marc@
03:53:51NJoinGodEater_ [0] (n=bryan@rockbox/staff/GodEater)
03:53:51NJoinchandlerc [0] (n=chandler@
03:53:51NJoinbillytwowilly [0] (
03:53:51NJoinATravelingGeek [0] (n=ATG@pdpc/supporter/student/ATravelingGeek)
03:53:51NJoinGodEater [0] (n=bryan@rockbox/staff/GodEater)
03:53:51***Server message 505: 'logbot :Private messages from unregistered users are currently blocked due to spam problems, but you can always message a staffer. Please register! ( )'
03:53:52 Join AlexC [0] (
03:53:52 Join andrewg877 [0] (
03:53:52 Join maxkelley_ [0] (
03:53:52NJoinToxicity999 [0] (n=bryan@unaffiliated/Toxicity999)
03:53:52NJoindarkless_ [0] (
03:53:52 Join JdGordon [0] (n=jonno@rockbox/developer/JdGordon)
03:53:52NJoinzerdik [0] (
03:53:52NJoinSirFunk [0] (
03:53:52 Join Bagder [0] (n=daniel@rockbox/developer/bagder)
03:53:52NJoinlinuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
03:53:52NJoinIsolinear [0] (
03:53:52NJoinJ [0] (
03:53:52NJointuplanolla [0] (
03:53:52NJoinnomel [0] (
03:53:52NJoinRaRe [0] (
03:53:52NJoindaurnimator [0] (i=daurn@unaffiliated/daurnimator)
03:53:52NJoincrashd [0] (
03:53:52NJoinshodanX [0] (
03:53:52 Join ivan` [0] (n=ivan`@unaffiliated/ivan/x-000001)
03:53:52NJoinpabs [0] (
03:53:52NJoinblithe [0] (
03:53:52NJoinwookey_ [0] (
03:53:52NJoinjurrie [0] (
03:53:52NJoinWeiss [0] (
03:53:52NJoinDiDjCodt [0] (
03:53:53***Server message 505: 'logbot :Private messages from unregistered users are currently blocked due to spam problems, but you can always message a staffer. Please register! ( )'
03:53:54 Quit Echelon (Dead socket)
03:53:59 Join HellDragon [0] (
03:53:59NJoinjoshin [0] (n=joshin@VDSL-130-13-11-27.PHNX.QWEST.NET)
03:53:59NJoinTotallyInfected_ [0] (
03:53:59NJoinpuzzles [0] (n=dan@xmms2/developer/puzzles)
03:53:59NJoinDogBoy [0] (
03:53:59NJoinJRoT [0] (
03:53:59NJoinkrazykit [0] (
03:53:59NJoinFOAD [0] (
03:53:59NJoinmicols [0] (
03:53:59 Join courtc [0] (n=court@unaffiliated/courtc)
03:53:59NJoingtkspert [0] (
03:53:59NJoinNimdae [0] (
03:53:59NJoinjepler [0] (n=jepler@emc/developer/jepler)
03:53:59NJoinlostlogic [0] (n=lostlogi@rockbox/developer/lostlogic)
03:54:00***Server message 505: 'logbot :Private messages from unregistered users are currently blocked due to spam problems, but you can always message a staffer. Please register! ( )'
03:54:27NJoinXavierGr [0] (
03:54:27 Join psycho_maniac [0] (
03:54:27NJoinmirak [0] (
03:54:27 Join crashmat1ix [0] (
03:54:27NJoin_pill [0] (
03:54:27NJoinhomielowe [0] (
03:54:27NJoinkarashata [0] (
03:54:27NJoinsd [0] (n=sd@
03:54:27NJoinfreqmod_nx [0] (
03:54:27NJoinze [0] (
03:54:27NJoinCronoh [0] (
03:54:27NJoinSoap [0] (n=Soap@rockbox/staff/soap)
03:54:27NJoinSeed [0] (
03:54:27NJoinsounddude [0] (
03:54:27NJointchan [0] (n=tchan@lunar-linux/developer/tchan)
03:54:27NJoinZom [0] (n=zom@
03:54:27NJoinmarcosou1ce [0] (
03:54:27NJoinannulus_ [0] (
03:54:27NJoinGalois [0] (
03:54:27NJoinw0rd54 [0] (
03:54:27NJoinPresence [0] (
03:54:31 Quit kkurbjun (Killed by (Nick collision))
03:54:32 Quit Rick (No route to host)
03:54:34 Join Gibbed [0] (
03:54:34 Join J3TC- [0] (
03:54:35 Join _jhMikeS_ [0] (n=jethead7@
03:54:35 Quit dionoea_ (Connection reset by peer)
03:54:37 Nick Gibbed is now known as Rick (
03:54:44 Quit zerdik ("Disconnecting")
03:54:52 Quit jhMikeS (Nick collision from services.)
03:55:35NJoinbeetlejuice [0] (
03:56:46 Join miepchen^schlaf [0] (
03:56:47***Server message 505: 'logbot :Private messages from unregistered users are currently blocked due to spam problems, but you can always message a staffer. Please register! ( )'
03:56:52NJoinin-jane [0] (
03:58:33 Join bagawk_ [0] (n=lee@unaffiliated/bagawk)
04:00:42 Join Dave2 [0] (i=dave@freenode/staff/dave2)
04:01:39beetlejuicedoes anybody know an alternate way to commit the database on boot up? my 20gig iriver h10 won't commit the database, and as a result is continually freezing and not allowing me to access the database at all. I have tried re-installing rockbox, twice, but to no avail.
04:02:54CtcpVersion from freenode-connect!freenode@freenode/bot/connect
04:03:03karashataI'm not aware of any, myself... do you have any free space on the drive for the database to be commited to?
04:03:19beetlejuiceoh, I don't know. good iead
04:03:41beetlejuice6.55gig free
04:03:51karashatathat's way more than enough...
04:03:59beetlejuicemy thoughs exactly
04:04:03karashataI couldn't tell you why it's not, then
04:04:33beetlejuiceit was working before, but after I finished adding more music, it stopped
04:04:53karashataI don't use the database, myself
04:04:57 Quit rvvs89 (Connection timed out)
04:05:06beetlejuicejust the file browser?
04:05:10karashataI have folders structured how I want, and playlists that I can use
04:05:55beetlejuiceyeah, okay. the only problem is, because the database won't commit, there are a whole heap of problems actually playing songs
04:06:12karashataeven out of the file browser?
04:06:46beetlejuiceyeah, it'll play that dir, but when it tries to skip to the next one, it throws an error
04:07:22karashatahmm, that sounds unusual...
04:07:36karashatadoesn't sound like something the database should affect, either
04:08:12karashataare you skipping to the next directory on the file tree, or are you using the random directory skip?
04:08:52beetlejuiceand another related thing, whenm I do try to enter the database, it tells me that it's not initialised, when I ask it to, to tells me its initialising, but does nothing
04:08:58beetlejuiceumm... I think its on random
04:08:59 Join billenium [0] (
04:09:50beetlejuiceyeah, it was on random
04:09:55beetlejuiceI've turned it off
04:10:08karashatahmm, the random skip should, at least from my experience, skip to the next directory in the tree unless you've run the random-folder-advance plugin...
04:10:24karashataotherwise, I don't know
04:10:46beetlejuiceno random folderadvance plugin on thos thing
04:10:53beetlejuicebummer, thanks for the help
04:11:06karashatayour welcome
04:11:19beetlejuicemabye: are there specific characters that throw the database out?
04:11:42beetlejuiceI copied the files from an ipod, so mabye that's It...
04:12:06karashataI wouldn't know, none of my files have any characters in the tags that aren't standard english characters
04:16:02 Quit bb (Nick collision from services.)
04:16:07 Join bb_ [0] (n=bb@unaffiliated/bb)
04:16:28 Quit bagawk (Connection timed out)
04:17:24 Join nnakihfdn [0] (i=0@
04:17:41***Saving seen data "./dancer.seen"
04:25:53 Quit darkless_ ("Leaving")
04:26:00 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
04:26:03 Join donutman25 [0] (n=chatzill@
04:26:18 Join miepchen^schlaf [0] (
04:29:50 Join webguest02 [0] (i=4cefbbf6@gateway/web/cgi-irc/
04:30:16webguest02has rockbox for the ipod video been fixed to better handle the battery?
04:30:25karashatanot yet
04:30:56karashataleast, last I hear the battery life hasn't significantly improved yet
04:31:09webguest02do you know when that might happen? Last time I used it my battery would die after about three hours of use maybe less.
04:32:11 Quit webguest02 (Client Quit)
04:33:24karashata...if he'd wait for an answer next time...
04:33:28*karashata shrugs
04:35:44 Quit mirak (Read error: 110 (Connection timed out))
04:37:44psycho_maniac3 hours? it lasts longer then that i thought.
04:38:11donutman25hey devs would this long copy and paste be of any help to you th get the sansa running better?
04:38:24karashatapsycho_maniac: I thought it lasted longer, myself
04:38:51karashatathough, I haven't looked closely at the runtime page in the wiki, since I don't care much about the iPods...
04:39:54billeniumCan anyone give me Ubuntu Cygwin help?
04:41:25 Quit midgey ()
04:41:28psycho_maniacthe wiki says that my ipod (the 80gig) lasts about 10 hours.
04:42:23karashata...maybe the wiki needs updating..?
04:43:02psycho_maniachmmm that could be.
04:43:20karashataor maybe the people getting that kind of battery life are being very conservative, listening at really low volumes, using a minimal WPS, not turning the backlight on very often at all...
04:43:50karashataI tend to get about 11, maybe 11.5 hours with my H10 20GB
04:44:41psycho_maniacactually in the 2nd square (dont know the name) its under 90-10% runtime its 8h 50min
04:46:46psycho_maniacbut 3 hours sounds pretty low.
04:47:20karashata3 hours seems lower than normal, the last time I looked, the lowest average tended to be 6, maybe 5, hours...
04:48:00billenium :this is for Window Users only... What if we dont use Windows?
04:49:10ATravelingGeekbillenium: try
04:49:30ATravelingGeekYou'll need all the tools listed there on your host, but otherwise it should work
04:49:33*billenium huggles AtravelingGeek
04:49:50ATravelingGeekWhat distro/os are you using?
04:50:55billeniumFeisty FAwn
04:51:04ATravelingGeekyou'll need to install pretty much everything
04:52:36billeniumHow long do you think it would take a semi-noob that catches on fast to setup?
04:54:14ATravelingGeeknot very long
04:54:22psycho_maniacnot too long. i did it . and now i can compile my own build with patches.
04:54:31ATravelingGeekinstall everything by apt-getting, grab all the source, compile, ???, profit!
04:54:46billeniumwhats this profit i hear off?
04:55:15ATravelingGeekYour own personal build that is worth an infinite amount to you but little to everyone else :)
05:01:03psycho_maniaci only use one patch but its one of the best imo
05:01:57billeniumi use the beast patch
05:05:51 Quit kripso (Read error: 104 (Connection reset by peer))
05:07:01 Quit scorche (Connection timed out)
05:16:06 Join bb [0] (n=bb@unaffiliated/bb)
05:16:07 Quit Strath (Read error: 104 (Connection reset by peer))
05:18:27 Quit rotator ()
05:19:03 Join Ishakabibble__ [0] (
05:22:25 Quit Ishakabibble__ (Connection reset by peer)
05:24:42 Join BHSPitLappy [0] (n=Me@unaffiliated/bhspitmonkey)
05:26:31 Join AndrewJ [0] (n=KeyLime@
05:26:39AndrewJanyway to get rockbox louder at all?
05:27:00psycho_maniacturn up the volume?
05:27:17Shaidbuy a headphone amp?
05:27:40 Quit bb_ (Read error: 110 (Connection timed out))
05:27:42BHSPitLappyrockbox gets pretty dang loud
05:27:45psycho_maniacor better headphones. since i got new headphones i never have mine over 6bs
05:28:16AndrewJhow do you get it OVER 6?
05:28:23AndrewJmine just gets to 6 and stays
05:28:46psycho_maniaci ment 0 dbs
05:29:10BHSPitLappyyou mean meant
05:29:29AndrewJim using JVC marshmallows
05:29:55ShaidI never go higher than -14 or so
05:30:23AndrewJhow would I go about making a patch to make it louder?
05:30:44AndrewJ(how do i even apply patches? :D)
05:30:48psycho_maniacandrewJ: what player do you have?
05:30:49BHSPitLappyAndrewJ, just use audacity to amplify your songs?
05:31:24AndrewJsansa e200
05:34:59psycho_maniacAndrewJ you partially deaf?
05:35:13AndrewJmaybe.... i listen to very load music :D
05:35:21AndrewJno im not
05:35:22BHSPitLappypsycho_maniac, he's going for it
05:35:31AndrewJgetting there
05:35:50Soappsycho_maniac: I made the 90-10% runtime column because it is really the only half-way reliable part of a LiIon's discharge.
05:36:33BHSPitLappyhe's trying to become one of the idiots who play their headphones so loud the whole class/bus/room can hear them clearly
05:36:43SoapThe first 10% of the discharge is very non-linear and varies greatly depending on charging conditions, and the last 10% is not reliably recorded (esp on 64MB models) because of the way battery_bench writes data.
05:37:18SoapSo, if you want to compare runtimes betweeen different testers you should use the 90-10 number, not the total.
05:37:43psycho_maniacthank you
05:37:59SoapThat being said - none of the numbers are worth jack squat if the condition of the battery is not know, preferably by doing a test with the original firmware.
05:38:29psycho_maniacSoap: could you maybe explain what this means: "The iPod Video has the HW version at a different address in ROM". and what advantages this does? thats the newest commit in svn.
05:40:33Soapthey found where Apple stores the hardware revision string for the 5th and 5.5th generation iPods. It was not at the same address as (all?) the previous generations. This will allow Rockbox to more reliably tell if it is running on a 5 or on a 5.5. What that info is being used for I can say - I can only ASSume the whole large sector issue with the 5.5 80B.
05:41:35 Quit beetlejuice (Read error: 110 (Connection timed out))
05:48:58Soapjust pray I understood the discussion correctly. ;)
05:49:19Soapfor looking at the diff, it is not what I expected.
05:53:28 Quit AlexC (Connection timed out)
05:58:23 Join advcomp2019 [0] (n=advcomp2@unaffiliated/advcomp2019)
06:17:42***Saving seen data "./dancer.seen"
06:22:03 Join midkay [0] (n=midkay@rockbox/developer/midkay)
06:23:22 Nick maxkelley_ is now known as maxkelley (
06:23:36 Join kri [0] (
06:28:41nomelhey, anyone know who calculated the inductor constant for the sansa c200 fm?
06:39:07 Quit gromit` ("Coyote finally caught me")
06:42:55 Quit kri (Read error: 104 (Connection reset by peer))
06:43:14 Join kri [0] (
06:59:14 Join AlexC [0] (
07:01:43 Part toffe82
07:02:45 Quit nomel ()
07:09:42 Quit Toxicity999 (Remote closed the connection)
07:10:18 Join Toxicity999 [0] (n=bryan@unaffiliated/Toxicity999)
07:16:15 Join jhulst_ [0] (n=jhulst@unaffiliated/jhulst)
07:19:02 Quit AndrewJ ()
07:31:07 Join midgey [0] (
07:49:54 Quit kri (Read error: 104 (Connection reset by peer))
07:50:15 Join kri [0] (
07:50:35 Quit psycho_maniac (" HydraIRC -> <- Would you like to know more?")
07:52:10 Join psycho_maniac [0] (
08:01:02 Quit TotallyInfected_ (Read error: 104 (Connection reset by peer))
08:01:31 Join TotallyInfected [0] (
08:09:47 Quit midgey ()
08:11:38 Join kri_ [0] (
08:12:39 Quit psycho_maniac (" HydraIRC -> <-")
08:12:54 Join scorche [0] (i=Blah@rockbox/administrator/scorche)
08:17:45***Saving seen data "./dancer.seen"
08:19:38 Quit jhulst_ ("Konversation terminated!")
08:24:35 Quit AlexC (Read error: 110 (Connection timed out))
08:28:21 Join Llorean [0] (
08:28:22 Join Rob222241 [0] (
08:29:08 Join psycho_maniac [0] (
08:30:04 Quit kri (Read error: 110 (Connection timed out))
08:31:18 Nick rvvs89_ is now known as rvvs89 (n=rvvs89@pdpc/supporter/active/rvvs89)
08:34:43 Quit psycho_maniac (" HydraIRC -> <- s0 d4Mn l33t |t'z 5c4rY!")
08:35:09 Join advcomp2019_ [0] (n=advcomp2@unaffiliated/advcomp2019)
08:35:36 Part karashata
08:35:50 Quit _jhMikeS_ (Nick collision from services.)
08:35:54 Join jhMikeS [0] (n=jethead7@rockbox/developer/jhMikeS)
08:36:49 Join ompaul [0] (n=ompaul@gnewsense/friend/ompaul)
08:37:12 Quit advcomp2019 (Nick collision from services.)
08:37:15 Nick advcomp2019_ is now known as advcomp2019 (n=advcomp2@unaffiliated/advcomp2019)
08:43:56 Join kubiix [0] (
08:46:16 Quit Rob2222 (Read error: 110 (Connection timed out))
09:08:14 Part Llorean
09:30:54 Quit BHSPitLappy (Read error: 110 (Connection timed out))
09:35:18 Join kripso [0] (
09:36:39 Join bertrik [0] (
09:38:22 Join Redbreva [0] (
09:54:23 Quit kri_ (Read error: 110 (Connection timed out))
10:02:52 Join TotallyInfected_ [0] (n=ebola@
10:03:51 Join Thundercloud [0] (
10:06:06 Quit BigBambi (Read error: 113 (No route to host))
10:11:22 Join desowin [0] (
10:11:31 Join kri [0] (
10:13:13 Quit TotallyInfected (Read error: 104 (Connection reset by peer))
10:13:23 Join AlexC [0] (
10:14:25 Quit kubiix ("Miranda IM! Smaller, Faster, Easier.")
10:17:13 Nick marcosou1ce is now known as |marcosource (
10:17:17 Nick |marcosource is now known as marcosource (
10:17:48***Saving seen data "./dancer.seen"
10:21:20 Join norbusan [0] (
10:21:24 Part norbusan
10:27:42 Join BigBambi [0] (n=alex@rockbox/staff/BigBambi)
10:29:20 Quit kripso (Read error: 110 (Connection timed out))
10:29:45 Quit markun (Read error: 104 (Connection reset by peer))
10:29:52 Join markun [0] (n=markun@rockbox/developer/markun)
10:30:03 Join Genre9mp3 [0] (
10:30:35 Join ender` [0] (
10:33:05 Join pixelma [0] (n=Marianne@rockbox/staff/pixelma)
10:33:21 Join kri_ [0] (
10:36:11 Join markun_ [0] (n=markun@rockbox/developer/markun)
10:43:12 Quit desowin ("use linux")
10:43:52 Join davina [0] (
10:46:05 Quit markun (Read error: 113 (No route to host))
10:46:24 Join cooz [0] (
10:47:02 Join ddalton [0] (
10:47:08 Part ddalton
10:50:59 Quit TotallyInfected_ ()
10:52:21 Quit kri_ (Read error: 104 (Connection reset by peer))
10:52:31 Quit kri (Read error: 110 (Connection timed out))
10:53:47 Join kri [0] (
10:54:35 Join hannesd [0] (
10:54:41 Join [1]seb [0] (
10:55:19[1]sebanybody online?
10:56:14BigBambiLots of people, just ask your question
10:57:36[1]seboh i see, sorry
10:59:05 Join pepie34 [0] (
10:59:28[1]sebwhen im clicking through the database a little text flashes, it says searching, 0 found ()
10:59:42BigBambihave you initialised the database?
10:59:47[1]sebi have sansa e250
11:00:20 Quit kri (Read error: 104 (Connection reset by peer))
11:00:20BigBambiAnd what music files do you have on there?
11:00:22bertrikI have a sansa e260 and see that too, but it hasn't bothered me so far
11:00:29[1]sebit didn't happened before my rockbox freezed completely
11:00:39 Join kri [0] (
11:00:44BigBambiOh, so after the splash you still see your music?
11:01:31amiconnThe splash is just the progress display of the database
11:01:51BigBambiI thought you meant it found 0 files when they were there
11:01:53amiconnSeems that it disappears really fast for you
11:01:58*BigBambi doesn't use the database
11:02:22pixelmaand that confused me on my Ondio too when seeing it the first time
11:02:23amiconnSo you'll just see the '0 files found' flashing by, then it disappears
11:02:28BigBambiIn that case don't worry, it just finishes the search before the splash is updated
11:02:37BigBambiOr something :)
11:02:40[1]sebi didnt notice it before i removed battery
11:03:18bertrikremoved the battery, what for?
11:03:29[1]sebrockbox freezed entirely
11:03:41bertrikif it hangs, you can just press off button for 15 seconds and it will turn off too
11:04:13bertrikthat is a hardware poweroff, so it should work even if software is completely frozen, IIRC
11:04:14BigBambiOr sometimes even longer I've heard
11:04:27[1]seb15? ok, didnt know
11:04:43BigBambiPeople have reported up to 30s being necessary
11:05:17[1]sebso, i guess i have to get used to the splash...
11:05:29BigBambiOr fix it
11:05:44bertrikmy e260 hangs when playing an ogg just after playing an mp3
11:05:52[1]sebhow ^^?
11:05:57 Quit pepie34 ("Ex-Chat")
11:06:28BigBambi[1]seb: get coding
11:06:46bertrikheh, that's open-source for you
11:06:54[1]sebi started C two weeks ago
11:07:07BigBambiYou are on your way
11:07:32BigBambibertrik: Do you do any skipping, or does it hang on just normal playback
11:08:03bertriki've seen it hang on normal playback too, but not easily reproducible
11:08:30bertriki can get my sansa really confused by switching between mp3 and ogg, very reproducibly
11:09:08BigBambiIs this bang up to date? Nico_P and lostlogic (and jhMikeS I think) were committing various fixes last night
11:09:18 Join PaulJam [0] (
11:09:48bertriklast tested version 15338 had the bug
11:10:22bertrikBTW, it doesn't hang as I said earlier, but playback just doesn't start, it hangs on shutdown after that
11:11:22BigBambiwell, 15338 is the latest, so I guess check flyspray or grap Nico_P when he is here
11:14:18bertriksvn 15305 still works, 15307 doesn't, 15306 was the mob commit ....
11:15:06BigBambiI'm surprised and impressed such a major thing as MoB has gone off so relatively smoothly
11:25:23[1]sebis there a way to switch off the system messages?
11:25:44BigBambiNot without editing the code
11:27:52coozi had same playback hanging problems on sansa c200 - stayed with r15305
11:30:38 Join kri_ [0] (
11:30:46 Quit amiconn (" bbl")
11:31:05[1]sebok, thanks for your support
11:31:25[1]sebim off, have a nice sunday!
11:31:56 Quit kri_ (Read error: 104 (Connection reset by peer))
11:31:59 Part [1]seb
11:32:17 Join kri_ [0] (
11:33:13 Join hannesd_ [0] (
11:36:07 Join [1]seb [0] (
11:37:10 Join Xerion [0] (
11:38:22[1]sebi just read in the manual that i can set a directory cache, but i dont have that option in my rockbox...
11:38:36BigBambi[1]seb: It doesn't exist on the sansa
11:38:44BigBambiAs it makes no real sense on flash targets
11:38:56BigBambiBut if it is in the manual it shouldn't be
11:39:13BigBambiWhat page is it on?
11:39:48 Join spiorf [0] (
11:40:02 Join kubrick [0] (
11:40:23coozthere's few more things which are in sansa manual and not in code
11:40:31coozand in other way too :)
11:40:35BigBambiSuch as?
11:40:42[1]seband no battery option either
11:41:09coozgive me moment
11:41:12BigBambiCould you make a list if you find stuff and pop it on flyspray?
11:42:01BigBambieither :)
11:42:22pixelmaI think some things aren't taken out because they are expected to be available soon(ish). I mean that especially for the c200 and some plugins for example
11:42:46BigBambiBut I don't think dircache and flash is one is it?
11:43:02coozin c200 manual virtual keybord section if empty
11:43:35BigBambic200 is quite a new port, so it is more expected that the manual needs a little more work
11:43:38cooztrick with rec button to delete took me a while to firure out
11:44:24pixelmaafaik dircache on flash only makes sense for auto-updating the database. The problems with Sansa is that the current implementation of dircashe doesn't work well with multi-volume that's why it is disabled
11:44:52 Join Lear [0] (i=chatzill@rockbox/developer/lear)
11:45:21pixelmacooz: I'm aware of this. I wanted to create a better button mapping for the virtual keyboard and then update the manual with it
11:45:22Slasheripixelma: auto-updating on flash targets should work now without dircache
11:45:23BigBambiAh OK - is the preferred solution though to allow auto-update (isn't it just removal of items?) without dircache, no matter how slow this makes it
11:45:32BigBambiSlasheri: I knew it!
11:46:22pixelmaSlasheri: thought it was slow then
11:46:34BigBambiSo I guess that dircache is unlikely soon to extend to the E200?
11:46:39Slasheripixelma: on flash it shouldn't be slow, however i have no way to test that
11:46:42coozpixelma: it's not bad atm, just volume buttons could have some different functions
11:46:53BigBambiIn which case it might be better to remove the section on it from the manual
11:47:05*BigBambi takes a first look at the manual source
11:47:24Slasherii have had in mind to commit that multi-volume patch for dircache
11:47:34pixelmacooz: problem is that I had two ideas, both have pros and cons... and I'm undecided :\
11:47:36 Nick fxb__ is now known as fxb (
11:47:40*BigBambi stops looking :)
11:48:04BigBambiSlasheri: So shall I not bother altering the manual?
11:48:07 Quit kri (Read error: 110 (Connection timed out))
11:48:22coozpixelma: any place/link about those ideas, or only your head? :)
11:48:28SlasheriBigBambi: probably it's best to wait some more :)
11:48:33BigBambiOK, will do
11:48:55pixelmaSlasheri: and I don't trust it, if even a manual "update now" doesn't work correctly on my Ondio with plugged and unplugged MMC...
11:49:27 Quit hannesd (Read error: 110 (Connection timed out))
11:49:28 Nick hannesd_ is now known as hannesd (
11:50:08 Quit AlexC (Success)
11:50:14pixelmacooz: only in my head at the moment (and 2 patches at a different box). I think I'm going to pop them up on flyspray today
11:50:42Slasheripixelma: hmm.. removable media certainly can break the db..
11:50:45pixelmaso they could also be disscussed there and maybe people come up with even better ideas
11:51:04*BigBambi whispers file tree
11:51:45pixelmaBigBambi: I use file tree myself, but if people demand the database...
11:51:57BigBambiI know, when will they learn
11:52:21BigBambiAlthough I don't have one, being glued to itunes with an iPod would do my head in
11:52:31BigBambiCopy and paste makes me happy :)
11:52:54bertrikI have my music files organised per album already, I hate it when 'media managers' want to flatten the file/directory structure again
11:53:13 Quit atsea-32 (Read error: 104 (Connection reset by peer))
11:53:35PaulJamBigBambi: the database is not only for itunes userd. i use it for example to make playlists by genre.
11:53:50pixelmait would have one advantage when used with multivolume - music will be completely in one list, with file tree you have the internal memory and what is on the other volume is inside a virtual subfolder
11:54:12BigBambiPaulJam: Sure, but that is optional. I would hate to be forced to use it just to listen to albums for example
11:54:19BigBambipixelma: true
11:56:18pixelmaeven though I don't use it myself this way, just try the database from time to time
11:58:38 Quit pixelma (" lunch")
11:59:17 Join Frazz [0] (
11:59:22 Join atsea-32 [0] (i=atsea-@gateway/tor/x-010b1eac913f038e)
12:02:51[1]sebah, cool, in the filetree there's no splah flashing
12:04:20 Join moos [0] (
12:04:31Slasheri[1]seb: not in db either if it's loaded to ram
12:05:19Slasheriif one is using db more than occassionally, it really should be loaded to ram to improve performance
12:05:47BigBambithere is an option in the database settings
12:05:54BigBambiThe manual should have this in it :)
12:06:14[1]sebauto update?
12:06:25BigBambiload to ram
12:07:26BigBambiShould be the option before auto update
12:07:30 Join mikola [0] (
12:07:33[1]sebno, sorry
12:07:45PaulJamis the load to ram option even available on flash based targets?
12:07:52BigBambiSlasheri: Does that also not exist for flash/E200?
12:08:11BigBambiPaulJam: I'm guessing not if it isn't there :)
12:08:22 Join Bagder_ [0] (
12:08:42SlasheriBigBambi: hmm, need to check that
12:08:43 Part mikola
12:09:40Slasheriah, currently it's enabled only with dircache, but that is not a requirement
12:10:01BigBambiChangy changy? :)
12:10:11Slasherii will :)
12:10:17BigBambiMerci :)
12:14:56LearStrange, got no audio for the end of a track. Went to the buffering debug screen, played a little with up/down and the next track started, seemingly when it should...
12:17:51***Saving seen data "./dancer.seen"
12:20:04 Quit atsea-32 (Read error: 104 (Connection reset by peer))
12:21:45 Join lee-qid [0] (
12:25:09 Quit spiorf (Remote closed the connection)
12:25:09 Quit Bagder (Connection timed out)
12:25:11 Join mirak [0] (
12:26:46 Quit Genre9mp3 ("I don't suffer from Rockbox psychosis. I enjoy every minute of it.")
12:30:30 Join Nico_P [0] (n=nicolas@rockbox/developer/NicoP)
12:39:08 Join amiconn [0] (n=jens@rockbox/developer/amiconn)
12:39:16 Nick markun_ is now known as markun (n=markun@rockbox/developer/markun)
12:42:26 Quit bertrik ("bye")
12:43:36linuxstbNico_P: Around?
12:46:09*linuxstb looks at a WMA file that seems to have an id3v2 tag, complete with JPEG in the middle of it...
12:46:36 Quit ender` (" Top reason why compilers are like women: Miss a period and they go crazy")
12:46:37 Nick fxb is now known as fxb__ (
12:46:42 Join Maha` [0] (i=tabac@gateway/gpg-tor/key-0xB9002659)
12:49:11 Join pixelma [0] (
12:50:04 Join styleism [0] (
12:51:00 Quit Shaid (Read error: 104 (Connection reset by peer))
12:52:27 Join ender` [0] (
12:53:49 Quit Maha` ("Leaving")
12:54:24 Join jumpatrain [0] (i=tabac@gateway/gpg-tor/key-0xB9002659)
12:55:18jumpatrainmusepack files use apeV2 for tagging, i tried playing them but rb cant read the frakkin tags nor get the replaygain straight
12:57:10moosjuppatrain: don't you have mixed id3 tags+APE?
12:57:33jumpatrainmoos foobar doesnt like id3 in ape tags at all
12:57:38jumpatrainape or nothin it seems
12:57:47moosyup rockbox too
12:57:52linuxstbDo you have an example file? Rockbox should support APE tags in Musepack files.
12:58:06mooslinuxstb: it does :)
12:58:21markunmoos: you also use musepack with ape tags, right?
12:58:44moosyeah, and with replaygain, and all is fine
12:58:48linuxstbjumpatrain: Also, what exactly happens? Does the file play? If so, what is displayed in the WPS and "show track info" screens when you play the track?
12:59:18 Join darkless [0] (
12:59:21jumpatrainthe file plays sure.
12:59:37jumpatrainit doesnt show any tags where other files do show tags
12:59:40 Quit TMM_ ("Ex-Chat")
13:00:03jumpatrain"show tag info" yields nothin
13:00:20moosjumpatrain: which soft do you use for tagging?
13:00:26linuxstbThen I repeat my request for an example file...
13:00:45jumpatrainhmm. ok gimme a moment
13:00:54jumpatrainmoos foobar 2000 of course :)'
13:01:29moosjumpatrain: then that's strange, but like linuxstb said, better to see the fle :)
13:01:29LearCould be it has a trailing ID3V1 tag.
13:01:40jumpatrainthe tagging is done in .cue files, where i ripped the album into one file
13:01:43moosLear: that what I thought too
13:01:47jumpatrainalthough, mp3 does this just fine
13:03:54 Join atsea-32 [0] (i=atsea-@gateway/tor/x-5567191f4be2a4f5)
13:10:20preglowbuffering really is in pain here
13:11:16linuxstbjumpatrain: That file works fine...
13:11:39preglowbuffering just doesn't seem to stop after having done its first full buffer fill
13:11:46preglowhard disk light will just blink until i press stop
13:11:51linuxstbjumpatrain: i.e. all the tags seem to being read correctly.
13:12:18 Join kripso [0] (
13:13:03jumpatrainlinuxstb guess it must be on my end
13:13:18 Nick parafin|away is now known as parafin (
13:13:30linuxstbpreglow: I've looked at that WMA file, and the ASF parser seems to be working as it should. But there are two issues - 1) Because the comment tag is so big, the metadata buffer for that track is full before it reads the later tags in the file, so they're skipped. 2) The "show track info" screen seems to dislike long comments.
13:14:40preglowlinuxstb: sounds like a similar issue we had with wav
13:14:58preglowlinuxstb: we shouldn't rely on what metadata we want being located at the very start of the file
13:15:01 Quit homielowe (Remote closed the connection)
13:16:39linuxstbit's not metadata we don't want (we want comments), it's just that the comment dominates...
13:17:07preglowwell, is that a problem?
13:17:14preglowespecially with mob, we should be fine, right?
13:17:34linuxstbIt's not a problem for Rockbox (it's just skipped), but this WMA file also contains an id3v2 tag embedded as an ASF tag... The id3v2 tag even includes an embedded jpeg ;)
13:17:43preglowsweet lord
13:20:19linuxstbI guess a sensible thing to do would be to limit the size of comment tags that we read - to ensure that some space is left for later tags. I don't think a 629 character comment is reasonable...
13:20:38jumpatrainlinuxstb these is what my cue files looked like and rb couldnt figure it out
13:21:00linuxstbIn fact, this file has comments longer than 629 chars, Rockbox truncates them to 629.
13:21:45preglowlinuxstb: space for later tags? don't we have around 32 megs of tag space for most targets now? :D
13:22:05amiconnBagder_: Looks like we have a hung build...
13:22:18linuxstbjumpatrain: You're trying to use that cuefile in Rockbox?
13:22:45jumpatrainlinuxstb right.
13:23:36linuxstbpreglow: I don't think that's exploited yet - we're still using the same metadata struct, but it's now in the audio buffer, rather than a static array. I don't think it has a variable size.
13:23:50linuxstbjumpatrain: Then why are you asking about APE tags?
13:24:44jumpatrainthat was a mistake, i apologize.
13:25:20linuxstbI don't think Rockbox reads replaygain tags from cuefiles - just the standard TITLE/PERFORMER tags.
13:25:54preglowlinuxstb: i know, but we could do that
13:26:14linuxstbpreglow: Sure.
13:26:15preglowbut anyway, anyone else having the "disk not turning off, ever" buffering thing?
13:26:26preglowit's pissing me off
13:26:28linuxstbI haven't updated to MoB yet...
13:26:28 Join webguest26 [0] (i=57b0fa6f@gateway/web/cgi-irc/
13:27:07preglowi didn't bring my charger :/
13:27:15linuxstbpreglow: Does musepack have "native" support for gapless?
13:27:30amiconnpreglow: Never happened here so far with MoB
13:27:37amiconn(mostly tested on H300)
13:28:53 Part Redbreva ("Out for the day...!")
13:29:13preglowlinuxstb: aye
13:29:25preglowlinuxstb: skip amount is in the header
13:29:50preglowlinuxstb: why?
13:29:57linuxstbpreglow: Just curious.
13:30:30 Quit kri_ (Read error: 110 (Connection timed out))
13:35:42 Join Fraser [0] (
13:38:10Nico_Plinuxstb: I'm back
13:38:18 Quit webguest26 ("CGI:IRC (EOF)")
13:39:20linuxstbNico_P: DId you read the logs regarding the issue with files with long comment tags?
13:40:09Nico_Plinuxstb: from where you asked if I was there and after?
13:40:27linuxstbAlthough I'm just looking at your MoB patch (I'm late...), and I see you have three static buffers for track metadata?
13:40:29linuxstbNico_P: Yes
13:41:12[1]sebwhat's mob?
13:41:44Nico_Plinuxstb: yes. I didn't expect to have to use one for the previous track, but otherwise it wasn't possible to make the progressbar continue after track change
13:41:54Nico_Plinuxstb: what's the problem with long tags?
13:42:30linuxstbIIUC, the problem is that Rockbox is still using a fixed-sized buffer for metadata - the mp3entry struct.
13:42:59linuxstbSo if this buffer becomes full (e.g. with a long comment tag), later tags in the file are skipped.
13:43:12PaulJamhmm, was tere a recent cahnge that could cause that the genre tag for m4a files doesn't work anymore? with a recent build it shows only garbage in the WPS and tag info.
13:43:39Nico_Plinuxstb: ah, then maybe we could remove the buffer from the struct mp3entry and put it in a handle
13:43:42pixelmaPaulJam: ii
13:44:00pixelmaoops, IiRC bluebrother reported issues with genre too
13:44:44pixelmabut he didn't see what it's related too
13:45:01*pixelma gives up correcting typos
13:45:44linuxstbNico_P: This isn't straightforward - we don't know in advance how big the buffer will need to be. We would need to do an initial parse of the tags to calculate the space needed, and then do the real parsing, which seems overkill...
13:45:57 Join bertrik [0] (
13:46:35LearActually, tags should be limited, to what the line scrolling code can handle (something like 260 chars last time I checked, but that was a while ago).
13:46:45Nico_Plinuxstb: what kind of parsing is involved? isn't there a header indicating the total length?
13:47:47linuxstbMaybe, but I don't think that's universal. Also, that will be the length of the tags in the file, when parsed, they will take a different amount of space (e.g. conversion to utf8)
13:48:52 Join lazka [0] (
13:49:22Nico_Plinuxstb: what solutions do we have then?
13:49:43 Join kri_ [0] (
13:50:34 Join n1s [0] (
13:50:35linuxstbI'm not sure... What facilities does MoB offer us now?
13:51:40linuxstbDo I understand correctly that there is "core" metadata stored in the mp3entry struct, and that's copied from the audio buffer to the static prev/current/next track structs. Plus we have the ability to store arbitrary metadata in the audio buffer, which isn't copied?
13:52:11Nico_Plinuxstb: yes, that's it
13:52:34Nico_Pbut currently this arbitray metadata possibility isn't used
13:52:41linuxstbAnd the arbitrary metadata will be kept on the buffer for the lifetime of the track?
13:52:42 Quit Frazz (Read error: 110 (Connection timed out))
13:52:58 Join Frazz [0] (
13:53:16 Join advcomp2019_ [0] (n=advcomp2@unaffiliated/advcomp2019)
13:53:16Nico_Pthen we have the problem of pointers which could become outdated byt that can be handled nicely
13:54:01linuxstbSo the WPS is just accessing the static structs at the moment?
13:54:10Nico_Pif the metadata parsers could allow it, it would seem a good solution to me to have a string buffer as a handle separate from the mp3entry handle
13:54:17LearHm, it might be nice for the codec to be able to say that some track metadata is no longer needed. Not sure if it is needed yet though.
13:54:30Nico_Pnot always, but most of the time (see audio_current_track and audio_next_track)
13:54:59linuxstbWhat are the problems if we try and get rid of those 3 static structs?
13:55:19 Quit Fraser (Success)
13:55:35Nico_Plinuxstb: the main problem is for the codec with the current id3
13:55:37linuxstbOr put another way, is there some metadata which needs to be in those static structs, and couldn't be kept to the audio buffer?
13:55:49Nico_Pprevious id3 is to keep the progressbar updated during the track trnasition
13:56:08linuxstbBut you just need "previous length" for that?
13:56:26Nico_Plinuxstb: "elapsed" and "length" are the most important I think. probably other things the codec needs, but the string I'm pretty sure are safe to move
13:56:41Nico_Plinuxstb: not only... It's a bit twisted
13:57:40 Join Rondom [0] (
13:57:58linuxstbSo you need things like samplerate, and #channels?
13:58:29Nico_PI think
13:58:35linuxstbHow do you envisage album-art working?
13:59:31 Join kri [0] (
13:59:34linuxstbIn my mind, the image is the same kind of data as the strings needing to be displayed in a WPS.
14:00:25Nico_PI agree. I was thinking of having the bitmap in a handle and users of the image access it via bufgetdata
14:00:48Nico_Pit would be looked for by audio_load_track
14:00:51Nico_Pcuesheet would work the same too
14:02:19 Quit [1]seb (" HydraIRC -> <- *I* use it, so it must be good!")
14:03:22Nico_Plinuxstb: the codec API has a pointer to the current struct mp3entry. it would be unsafe to set it to a location in the main buffer, where it could move, that's the reason for the static struct
14:07:04 Join Strath [0] (
14:07:52 Quit kripso (Read error: 110 (Connection timed out))
14:07:56 Join kripso [0] (
14:09:12n1slinuxstb: FS #8039 is a hw limitation, right?
14:09:36 Quit styleism ()
14:09:48 Quit advcomp2019 (Read error: 110 (Connection timed out))
14:10:13linuxstbn1s: Recording gain control hasn't been implemented for the ipods yet.
14:10:44linuxstbAnd microphone mode shouldn't even be enabled for the 5g - it just has line-in via the dock.
14:10:51n1sok, i'll close it as it isn't a bug then
14:11:36linuxstbThere must be other tasks about recording on ipods, so it's more likely a duplicate.
14:15:15Nico_Ppreglow: about the disk never spinning down issue, do you have the latest svn?
14:16:57 Join TMM [0] (
14:16:59linuxstbn1s: I can't find any duplicate reports... I would leave it open.
14:17:03 Quit kri_ (Read error: 110 (Connection timed out))
14:17:13n1shmm, that particular thing has not been reported before afaics either
14:17:20 Join kri_ [0] (
14:17:44n1slinuxstb: is the data abort after recording bug still alive?
14:17:55***Saving seen data "./dancer.seen"
14:18:51linuxstbI don't get data aborts, but playback is still broken after recording on the ipods - the PCM buffer fills, but never empties.
14:19:21PaulJamlittle question, are the warnings " Clock skew detected. Your build may be incomplete." and "File `...' has modification time 0.14 s in the future" in cygwin something i should worry about?
14:19:31n1sah, doesn't seem to be a lot of devs interested in recording on those targets unfortunately
14:20:04n1sPaulJam: yep, someone in the future is editing your files :P
14:21:28 Quit kri (Read error: 110 (Connection timed out))
14:23:09 Join austriancoder [0] (n=austrian@rockbox/developer/austriancoder)
14:25:04linuxstbaustriancoder: Hi.
14:25:30austriancoderlinuxstb: hi
14:25:37n1sjhMikeS: about your comment in FS #7402 has that fix been committed?
14:26:21linuxstbaustriancoder: Have you read about Zagor's USB work?
14:26:32austriancoderlinuxstb: nope
14:26:51austriancoderlinuxstb: i have only seen #7962
14:27:26 Quit kripso (Read error: 110 (Connection timed out))
14:27:37jhMikeSn1s: good timing. just got up. lemme look.
14:27:44linuxstbOK, there's not much more to add, apart from he's still working on it. Have you done any more work on your implementation?
14:28:59amiconnhi Nico_P
14:29:12Nico_Pamiconn: hi
14:29:30amiconnThe bug I reported for next track metadata before MoB obviously extended to current track metadata with MoB :/
14:29:30austriancoderlinuxstb: I am hacking at the moment on a failing ch9 test... I will have a loot at patch #7962, maybe it will help me
14:29:40 Join kripso [0] (
14:29:56jhMikeSn1s: Are data aborts still happening on other players or it's just a failure to restart audio playback now?
14:30:12Nico_Pamiconn: could you ellaborate?
14:30:33n1sjhMikeS: according to linuxstb the data abort is gone but as you say playback doesn't work after recording
14:30:44amiconnWithin the last few seconds of a track (when the codec already switched to the next track internally), 'current track' info in wps is from the next track (and 'next track' info is from 2 tracks ahead, as reported before MoB)
14:31:13Nico_Phuh? weird, I thought I had checked that and it was ok
14:31:28jhMikeSn1s: hmmm...them probabaly something that needs addressing in the setting the chip back to playback mode or some little IIS tweak is needed for certain players.
14:31:33amiconnElapsed time is correct
14:32:06 Join kri [0] (
14:32:22linuxstbjhMikeS: Do you have a PP502x ipod?
14:32:50jhMikeSn1s: after fixing the data abort on H10 I found I needed the IISTX to always be enabled and all worked nicely
14:33:03jhMikeSlinuxstb: no
14:33:28linuxstbaustriancoder: You may want to talk to Zagor - IIUC his intention is that his USB code will be committed to replace your code.
14:33:54jhMikeSMy PP targets are: iPod 3G, H10-20GB, e260
14:33:56 Join spiorf [0] (
14:34:04Nico_Pamiconn: it's perfectly OK for me... :/
14:34:13linuxstbjhMikeS: Do we have recording on the 3G?
14:34:15Nico_Peven next track metadata is correct
14:34:17austriancoderlinuxstb: hmm.. i think it would be better to integrate his work in to mine
14:34:28linuxstbaustriancoder: Hence the need to talk ;)
14:34:32amiconnlinuxstb: 3G can't record. It just has a DAC
14:34:43jhMikeSlinuxstb: not at all. Are they supposed to? (haven't look at retailos too much)
14:34:50linuxstbReally? I could have sworn that IPL could record on the 3G...
14:34:50Nico_Pamiconn: has there been skipping involved?
14:35:06amiconnNico_P: No, but resume
14:35:45amiconnlinuxstb: Hmm, seems I confused sth. WM8731 can record
14:36:16Nico_Pamiconn: resume to the last few secs? also it's expected that the playlist index will increment a little too early, but metadata should be fine
14:36:27Nico_Pamiconn: I would be very interested to hear a repro recipe
14:36:32jhMikeSIf the wm8731 driver was similar to other targets. I found out the chip was left tuned off after recording was stopped. Probably the problem.
14:36:40amiconnNot sure, have to test again
14:36:50 Join kugel [0] (i=kugel@unaffiliated/kugel)
14:36:57amiconnI observed it on ipod video, while testing something totally unrelated
14:37:35jhMikeSI added shadow regs and defines to wm8731 which make it much easier to track what's happening
14:37:58amiconnjhMikeS: The driver is the same for all targets with WM8731
14:38:33jhMikeSI redid the driver for wm8731. Is there something with wm8731 that is having the problem?
14:38:56preglowNico_P: lemme check exactly what rev it is
14:39:30Nico_Ppreglow: my last two commit were attempts to fix the problem so if you still have it the attempts obviously failed
14:39:36linuxstbI don't think so - the problem devices are ipod 4G, Color and Nano (WM8975) and ipod Video (WM8758)
14:40:20preglowNico_P: 15334
14:40:41jhMikeSwm powerdowns were being set on wm8731 when recording was disabled
14:40:47jhMikeS*all wm
14:40:49Nico_Ppreglow: then update and tell me if the problem is still there ;)
14:41:30preglowif i can find a cable
14:42:03Nico_Pr15335 and r15336 are (or should be) fixes to that problem
14:42:28*jhMikeS sees the problem with wm8975. the whole chip is turned off there too.
14:42:53linuxstbjhMikeS: Is it an easy fix?
14:43:14billeniumokay i finished install now, do i have to restart for it to work?
14:43:18 Join webguest57 [0] (i=53ed3c9e@gateway/web/cgi-irc/
14:43:38billeniumi finished installing JRE*
14:44:09*linuxstb wonders if billenium is in the right channel
14:44:22billeniumi dont like xchat...
14:44:26webguest57could you give me a permission to write a wiki page?
14:44:36linuxstbWhat is your wiki name?
14:44:48jhMikeSlinuxstb: I doubt it's too difficult. I imagine setting it up with shadow regs would make it easier to configure for rec/play.
14:44:49webguest57Iriver Clix2
14:45:00BigBambiwebguest57: What name did you sign uo with
14:45:07linuxstbwebguest57: You're required to use your real name in the wiki...
14:45:49BigBambiHeh, IriverClix was the username....
14:46:26webguest57i typed Iriver i need to change my wiki name?
14:46:32*jhMikeS notices all the gain ranges in wm codecs seem to be pasted from the uda
14:46:33BigBambiYes, to your real name
14:47:13linuxstbjhMikeS: So basically make the driver more like the wm8731?
14:47:28BigBambiwebguest57: From the registration page, "Your WikiName must be your real first and last name concatenated, like JohnSmith or SteveAnderson. It's not JohnS, SteveA or NinjaBoy"
14:48:18jhMikeSwm8731 and wm8721 have difference layouts and reg names than those chips. Probably more like the gigabeat driver but with shadow regs and recording support.
14:49:00webguest57i rgistered as john black
14:49:11 Quit kripso (Read error: 110 (Connection timed out))
14:49:14BigBambiSo your wiki name is JohnBlack?
14:49:59BigBambiAccording to the wiki registration agent it is IrverClix
14:50:13BigBambiWhat username do you use to log in
14:50:33BigBambiAh, I see mow
14:50:37BigBambiOK, one mo
14:51:02 Quit kri_ (Connection timed out)
14:51:31BigBambiwebguest57: OK, you should be good
14:51:56webguest57so now i can write a wiki page?
14:52:09BigBambiEr, one second
14:53:00BigBambiI got my alphabet wrong the first time :/
14:53:26webguest57thanks alot!
14:53:50BigBambino problem
14:53:52linuxstbwebguest57: I'm guessing you're interested in a port to the Clix2?
14:55:55webguest57yes i am!but i am an absolout beginner!so i find people who could help me.moreover my english isnt very well because i am from russia!
14:56:01PaulJami tested a little bit with the uisimulator and m4a files and it seems that the problems with the genre tags started with r15306 (initial MoB commit).
14:56:27Nico_PPaulJam: not a big surprise :)
14:56:58Nico_PPaulJam: you're getting garbage tags?
14:57:43PaulJamNico_P: on the target i get garbage tags, and in the uisim it (more or less) crashes.
14:57:48*jhMikeS want to take the "based on IPL" remark out of wm8731.c since it's not even one bit based on it. :)
14:58:04webguest57<linuxstb> and i am not quite understanding how to write a wiki page!could you help me?
14:58:18Nico_PPaulJam: what's "more or less"? no segfault? also, does it happen with all m4a files?
14:58:52linuxstbjhMikeS: Then do it...
15:00:11preglowjhMikeS: then go ahead
15:00:33SoapYou are from Russia and your name is John Black? webguest57?
15:00:43 Join Maddin94 [0] (i=51adac56@gateway/web/cgi-irc/
15:00:47PaulJamNico_P: the window of the simulator closes and immediately opens again and a splash complaining about a missing .rockbox dir shows. the simulator was compiled with cygwin under windows. when i remove the genretags from the files they work fine.
15:01:17 Join rep|icant_ [0] (n=rep|
15:01:24*preglow disappears
15:01:25linuxstbwebguest57: I can create a wiki page for you, and then you will just need to edit it. I assume you want a page to add information about the Iriver Clix2?
15:01:38linuxstbpreglow: Clever trick.
15:01:45jhMikeSpreglow|disappeared: what did you change in there?
15:02:05Maddin94well, do you know, when the usb-support will work?
15:02:53n1sMaddin94: impossible to say
15:03:28webguest57<linuxstb>yes i need to ad information!its from iriver clix2 has Same CPU as the Cowon D2 (Telechips TCC7801)
15:03:54Maddin94n1s: I thougt so
15:05:03Nico_PPaulJam: I'd be interested in one of your tracks
15:05:16linuxstbwebguest57: - click on the "Edit" link at the bottom and add your information.
15:06:01 Quit Maddin94 ("CGI:IRC")
15:06:43 Join Robin0800 [0] (
15:07:05webguest57where is edit?on the bottom of what?
15:07:44Nico_PPaulJam: I got a segfault when viewing the track info :) (this is good)
15:08:29PaulJamyou got the file? (my client shows that it was aborted)
15:08:35BigBambiwebguest57: Is John Black your real name?
15:08:39Nico_Pyep I got it
15:09:00Nico_PPaulJam: is it only 2.6 MB?
15:10:08Nico_Pwell the genre string definitely isn't correct
15:10:33BigBambiJust doesn't sound very russian
15:11:01 Quit rep|icant (Read error: 110 (Connection timed out))
15:11:46webguest57<BigBambi>i know!my grandparents where from england!my russian name is ваня which is the same as john!
15:11:57 Join rzr\ [0] (
15:12:02BigBambiwebguest57: Not a problem, just wondering :)
15:13:31 Quit davina (Remote closed the connection)
15:14:03rzr\Hello all. I'm having an issue with my brand new Sansa e270 which appeared after installing rockbox. Windows will not detect the device now via USB, even though the player lights up and shows me the USB icon when I plug it in.
15:14:16BigBambirzr\: You need to reboot to the OF
15:14:30BigBambiRockbox doesn't (yet) have a USB mode
15:14:36rzr\The original firmware I take it
15:14:56BigBambiAlthough it is being worked on
15:14:59bertrikEasy way to go into OF, is to turn off the sansa while in rockbox, then plug in the charger cable
15:15:18 Join davina [0] (
15:15:20BigBambi*Clarification - rockbox doesn't yet have a USB mode on portalplayer targets
15:15:20rzr\I will try that.
15:16:04Soapdoes rockbox have a USB mode on non-portalplayer targets? Or does the hardware have a USB mode? ;)
15:16:19rzr\So it's not possible for me to transfer files using when booted to Rockbox then
15:16:30Soapcorrect at this time rzr\
15:16:32BigBambiSoap: Damn you, you pedantic ****
15:16:56rzr\Alright, thank you. Now I know. :)
15:17:45Nico_PPaulJam: I know where the problem comes from... now I need to solve it
15:17:57Nico_Pn1s: were you getting bogus genre strings too?
15:18:51PaulJamNico_P: pixema said that bluebroter reported this aswell
15:19:00Nico_PPaulJam: only for genre?
15:19:17bertrikanyone familiar with building and using the rockbox simulator from a VirtualBox under windows xp?
15:19:29PaulJami saw it only with genre.
15:19:50Nico_PPaulJam: FYI, the problem wasn't introduced by MoB, but it was triggered by it.
15:19:59linuxstbbertrik: You want to build a win32 sim under Linux?
15:20:12Nico_Pit's a problem with metadata moving, but metadata wasn't moved before MoB
15:20:18n1sNico_P: I haven't looked for that at all so dunno
15:20:38Nico_Pn1s: didn't you report metadata problems? maybe I confused with bluebrother
15:20:43bertriklinuxstb: no, i run windows xp as my main OS and now i want to build and use the rockbox simulator in a virtual box
15:20:58linuxstbbertrik: And Linux is running in the virtual box?
15:21:07 Quit hcs ("Leaving.")
15:21:30linuxstbAre you having a specific problem? It should just work like any other Linux install.
15:22:30kugelbertrik: you can build the sim under linux
15:22:58n1sNico_P: i don't think I did at least :)
15:23:07Nico_Phehe :) sorry then
15:23:09bertrikyeah, the ui sim builds runs but it complains about missing .rockbox dir
15:23:28bertrikI cannot figure out how to share files between windows and linux running under virtual box
15:23:42linuxstbYou need to run "make install" after compiling the sim.
15:23:53kugelI didn't figure that out yet too
15:24:17linuxstbAs for sharing files, that's not specific to Rockbox - I'm sure the virtualbox documentation will cover that.
15:24:17 Quit webguest57 ("CGI:IRC")
15:24:48JdGordonBagder_: did a distro get credit for a new curl feature or something ?
15:25:16PaulJambertrik: maybe this helps you:
15:26:59 Quit n1s ()
15:27:00billeniumThe VMware quickstart guide is confusing me. I did that first step, downloading, then i opened it... Then what do i do? download the 7zip? (running Ubuntu Feisty Fawn
15:28:17linuxstbbillenium: So you've installed vmware player?
15:28:37billeniumi just downloaded the .tar
15:28:59linuxstbSo you're running Ubuntu on your computer already?
15:29:12billeniumUbuntu = my OS
15:29:14linuxstbThen you don't need vmware - that's just for Windows users.
15:29:34linuxstbAll you need to do is download the Rockbox source code, and run the tools/ script
15:30:05billeniumso its more simple if you just have ubuntu?
15:30:05linuxstbSee here -
15:32:49 Join kri_ [0] (
15:33:42 Quit Lear ("ChatZilla [Firefox]")
15:35:11amiconnBagder_: ping...
15:35:49 Join SirFunk_ [0] (
15:38:13Nico_PPaulJam: I have a fix
15:39:26 Join kripso [0] (
15:41:50 Join dongs [0] (
15:41:56dongshello my fellow israelis
15:42:00dongshow much ram is in a typical ipod?
15:42:09dongs(im too lazy to open browser and check)
15:42:37rasherIn the future, don't be.
15:42:50dongsyes, but thats waht you're here for
15:42:56PaulJamNico_P: cool, that was fast.
15:43:22 Quit SirFunk (Read error: 110 (Connection timed out))
15:43:57rasherdongs: Stop spamming unless you have something Rockbox related to share.
15:45:12 Join kugel_ [0] (
15:45:26 Quit kugel (Nick collision from services.)
15:45:28 Nick kugel_ is now known as kugel (
15:46:01 Join Arathis [0] (
15:46:10dongsrasher: what?
15:46:18dongsrasher: do you know if that 32megs is external sdram or what.
15:47:05keanudongs, in what way would it matter?
15:47:17keanukeanu, it's DDR3 SODIMMS
15:47:23dongsnice try.
15:47:35Nico_PPaulJam: committed. You might want to update ;)
15:47:37dongswhen you troll, address the proper nick.
15:48:17keanudongs, sorry - i've been rushed the past few isn't the first time i've addressed myself
15:48:30 Quit rzr\ ("Generic quit message...")
15:48:45linuxstbdongs: Yes, SDRAM is external to the PP.
15:48:46coozdongs: did you see that: ?
15:48:47Nico_Pargh the build server is stalled... Bagder_?
15:49:10dongscooz: no, that deosnt seem to be relevant to my current line of questions
15:49:14keanudongs, I apologize for that 'joke'
15:49:14dongslinuxstb: great
15:49:31 Quit kugel (Client Quit)
15:51:24 Quit kri (Read error: 110 (Connection timed out))
15:53:12 Quit billenium (Read error: 110 (Connection timed out))
15:55:11 Part pixelma
15:55:29 Quit jhMikeS (Nick collision from services.)
15:55:35 Join jhMikeS [0] (n=jethead7@rockbox/developer/jhMikeS)
15:56:13 Join kugel_ [0] (
15:56:19 Nick kugel_ is now known as kugel (
15:56:46 Join toffe82 [0] (
15:57:34 Quit kri_ (Read error: 110 (Connection timed out))
15:58:40 Join J-23 [0] (
15:59:03J-23Can I find UTF-8 fonts for Rockbox?
15:59:55linuxstbJ-23: Have you seen this page?
16:00:15 Join Calcifer [0] (
16:00:24 Quit kugel (Client Quit)
16:03:02*J-23 looks for Polish small fonts.
16:03:39J-23How I can edit FNT files?
16:03:48JdGordonyou dont
16:03:53JdGordonyou edit the bdf
16:04:36PaulJamNico_P: thank you. everything looks fine here with the latest build.
16:04:49J-23these files in /.rockbox/archos/fonts?
16:05:00JdGordonno, in the source files
16:05:35JdGordonNico_P: do you tihnk it would be difficult to stop the spinup if you press play when a playlist is loaded but stopped? (so the tracks had already been buffered?)
16:06:15Nico_PJdGordon: I don't understand the situation you're describing
16:06:24JdGordonsorry, im half asleep :)
16:06:27 Join Sedgewick [0] (
16:06:36JdGordonif you press stop while music is playing.. then press play again
16:06:47JdGordoncurrently it would need to rebuffer the playlist again
16:07:08Nico_PJdGordon: that would be possibe if we don't close the handles on stop
16:07:17Nico_Plet me see
16:07:48 Quit spiorf (Connection timed out)
16:08:42Nico_Phmm I think things would quickly become hard to manage... how do you then distinguish between resuming and a new playlist?
16:09:15JdGordonI guess, if the playlist is changed it is a new one, otherwise its the same?
16:09:17Nico_Pand what if another "user" needs the buffer. It would need to know the playback system hasn't freed the handles and ask it to
16:09:37Nico_Pyeah but playback.c isn't very aware of the playlist
16:09:43 Quit Xerion (Read error: 104 (Connection reset by peer))
16:10:00JdGordonok, well something to think about anyway :) im goign to bed
16:10:06linuxstbBut isn't the idea of "stop" to clear everything?
16:10:06 Quit JdGordon ("Konversation terminated!")
16:10:12Nico_Plinuxstb: exactly
16:10:20 Join Rincewind [0] (
16:10:47Nico_Pthere's a task on the tracker added by pondlife with a similar issue
16:11:14Nico_PFS #8029
16:14:01Nico_Plinuxstb: seen my commit to id3.c?
16:14:52PaulJamcould it be that "gather runtime data" doesn't work anymore?
16:16:07 Quit Calcipher (Read error: 110 (Connection timed out))
16:17:17linuxstbNico_P: Just looked at it now.
16:17:34markundongs: you're back!
16:17:42 Quit kubrick (Read error: 104 (Connection reset by peer))
16:17:57***Saving seen data "./dancer.seen"
16:19:04markunare you still a member of the GNAA?
16:19:38 Join spiorf [0] (
16:19:43 Join DanielSiebrecht [0] (i=57b0f925@gateway/web/cgi-irc/
16:20:59 Quit DanielSiebrecht (Client Quit)
16:21:34J-23How I can check size of font?
16:24:16 Join kubrick [0] (n=repulse@unaffiliated/funky)
16:24:29markunJ-23: are you looking at the .bdf files?
16:24:35Nico_PjhMikeS: is there a way to know when and how a mutex is blocking a thread?
16:24:59J-23markun: FNT. I'm trying to make Polish font for Rockbox
16:25:23markunJ-23: it's not possible to see the size of the font, but somethimes it's specified in the name
16:25:34J-23convert Polish Times New Roman for Rockbox*
16:26:30J-23Somebody uses Font Convertor?
16:27:51 Join kugel [0] (i=kugel@unaffiliated/kugel)
16:30:19J-23it returns errors when I try to make 6x10 font
16:31:01 Quit FOAD ("I'll be back")
16:33:43 Join mf0102 [0] (n=michi@
16:34:36J-23What size's rockbox-default?
16:35:43J-23How I can set such size in Font Convertor?
16:36:54 Join FOAD [0] (
16:36:54 Quit FOAD (Client Quit)
16:37:08 Part dongs ("emoquit")
16:37:31 Join FOAD [0] (
16:40:55 Join midgey [0] (
16:41:54 Nick fxb__ is now known as fxb (
16:42:06 Nick fxb is now known as fxb__ (
16:42:13markunJ-23: I don't think the font convertor recognizes the rockbox .fnt format. If you use the original .bdf file the size should be found
16:44:23J-23But how I can give 6x10 size to Font Convertor?
16:47:23 Join [1]seb [0] (
16:47:56[1]sebhey BigBambi, are you there?
16:48:59[1]sebcool, remember me with the "search...0 found"-problem?
16:49:19BigBambiindeed so
16:50:25 Quit J-23 ()
16:50:28[1]sebi got a custom build from rockbox-lounge, with the load to ram option on. works properly
16:50:36BigBambiOK, good
16:51:19 Quit kripso (Read error: 104 (Connection reset by peer))
16:51:28BigBambiBe aware that custom builds aren't supported here, but Slasheri has said he is going to remove the dircache requirement for flash targets in SVN, so when he does that the official builds will also work as supposed
16:51:39 Join kripso [0] (
16:51:54BigBambiAlthough I don't think it is a problem, at the moment, just slightly annoying
16:52:24 Join ilgufo [0] (
16:54:01[1]sebcool, so i check back later, so i can switch to original rockbox
16:54:29 Quit FOAD ("I'll be back")
16:55:50[1]sebthanks again...have a nice evening
16:55:58 Quit [1]seb (" HydraIRC -> <- Po-ta-to, boil em, mash em, stick em in a stew.")
16:57:23 Quit moos ("Glory to Rockbox")
16:58:15 Join zardos [0] (i=53e279b4@gateway/web/cgi-irc/
16:58:47amiconnBagder_: ping... ...
16:59:06PaulJamseems like r15306 (MoB) broke the "gather runtime data" feature. will someone look at this or should i open a bugreport?
16:59:22BigBambiPaulJam: Always best to open a bug report
16:59:29BigBambiThen more people can see it
17:00:00zardosDose some one here know how to turn down the sencitivity of the compiler in vwware Rootbox? I mean manny of my own aplications compile fine in cygwin but not in linux, some one know why?
17:01:09krazykitzardos, are you talking about rockbox specifically, or just your programs in general?
17:02:57zardos rockbox i guess
17:03:50Nico_PPaulJam: yeah, bug report is always better
17:04:00Nico_PI'll look into it though
17:04:25 Quit Toxicity999 (Read error: 104 (Connection reset by peer))
17:04:40krazykitzardos, do you have any specific error messages?
17:04:47 Join Toxicity999 [0] (n=bryan@unaffiliated/Toxicity999)
17:05:06Nico_PPaulJam: any details on how it's broken? I don't even know what the feature does
17:05:39zardoshold on a sec
17:07:19krazykitzardos, but please don't paste the errors here. paste them at and put the link here
17:07:32 Join Buschel [0] (
17:08:45PaulJamNico_P: the gather runtime data feature collects playcount, playtime and when it was played last time in the database (used for autoscore, recently plaed etc.). now the old stats are still shown, but don't get updated anymore.
17:09:43zardoskrazykit i need to do a recompile so it will take a while, is the no setting to tund down the error threshold to allow uglyer code?
17:10:31krazykitnot as far as i know, anyway. you could check gcc's documentation if you really want to
17:10:40Nico_PPaulJam: OK, I think I know why it might be broken
17:13:04Nico_PPaulJam: you're saying the stats are shown correctly in the track info while playing?
17:13:36 Nick advcomp2019_ is now known as advcomp2019 (n=advcomp2@unaffiliated/advcomp2019)
17:13:56 Join gromit` [0] (
17:15:43PaulJamthe stats that were collected before the MoB show correctly in the wps (my WPS only shows autoscore) and in the Database, but they don't update, so in the last played view i see the tracks that were played last before i updated the build.
17:16:39Nico_Pok. the fact that the info is shown in the WPS at least tells me something
17:19:26 Join kri_ [0] (
17:31:56 Join matsl [0] (
17:33:33lostlogicNico_P: what would have broken that... it just uses the end_of_track callbacks in playback.c right?!
17:34:04Nico_Plostlogic: yeah but in the main buffer, "elapsed" is always 0
17:35:58lostlogicooh :)
17:36:18 Join tpickers66 [0] (
17:37:04lostlogicNico_P: also, wrt FS #8037, most of my music is mp3 and ogg so my player switches back and forth constantly and I've not seen that bug, so it's definitely not an everybody or an always bug.
17:37:21 Quit kripso (Read error: 110 (Connection timed out))
17:38:05 Quit ilgufo ("So Long, and Thanks For All the Fish -")
17:38:26 Quit atsea-32 (Remote closed the connection)
17:41:32Nico_Plostlogic: that means we need to find out what's so special with bertrik's setup
17:42:46tpickers66anyone here got a copy of the ihpfirm tool? the link on the IriverToolsGuide wiki page isn't working...
17:43:20 Quit donutman25 ("ChatZilla [Firefox]")
17:44:27amiconnBagder_: ping...
17:44:54linuxstbtpickers66: That is obsolete - the code is incorporated into scramble/descramble - two tools that come in the Rockbox source code.
17:45:03bertrikplease let me know if I can do anything, Nico_P, perhaps send you one of my oggs?
17:45:15Nico_Pbertrik: yeah that might help
17:45:24Nico_Pbertrik: also have you tried in the sim?
17:45:34bertrikyes, I tried the sim, it works fine there
17:45:44tpickers66linuxstb: thanks. maybe I'll update the wiki page, once I've worked out how the new tools work
17:46:04lostlogicis the e200 faster or slower than the ipod G5 for decoding?
17:46:08 Join Crash91 [0] (n=evil91@
17:46:10lostlogicand is it HDD or flash based?
17:46:18linuxstbtpickers66: They are documented a little here -
17:46:34linuxstbIt's flash-based. But the CPU should be the same.
17:47:12lostlogicand what's its ram size?
17:47:21linuxstb32MB I think.
17:47:21tpickers66linuxstb: okay, thanks.
17:47:30markunlinuxstb: nice work on the new port
17:47:45lostlogicbertrik: does it happen when you press skip, or when you let it transition or both?
17:47:49lostlogicbertrik: also, do you use crossfade?
17:47:51bertrikI'll try once more with just two songs on my sansa
17:48:05bertriki tried with a completely fresh, default, install
17:48:12bertrik(wiped the .rockbox dir)
17:48:21lostlogicok, so not crossfade
17:48:26 Join ilgufo [0] (
17:48:37*amiconn points lostlogic to
17:49:06bertrikI hope I'm not overlooking something obvious
17:49:11linuxstbmarkun: The Telechips CPUs seem quite nice - not over-powered like the Gigabeat F, but fast enough to be comfortable in Rockbox... And power consumption seems very good (the OF for my device quotes 50 hours MP3 playback on a AA battery.
17:49:25lostlogicamiconn: but I'm sooo lazy :-D yes, I should RTFM just as much as anyone else.
17:49:28*linuxstb is trying to convince others to buy Telechips players...
17:49:50 Join atsea- [0] (i=atsea-@gateway/tor/x-c0bf80ce1fa2b5d6)
17:49:58amiconnAre there any monochrome or greyscale telechips targets?
17:50:11amiconn(or transflective colour)
17:50:18amiconnAnd please no touch crap
17:50:23linuxstbThe only mono/greyscale one I know is the one I'm working on.
17:50:35lostlogiclinuxstb: your commit says that iAudio 7 is telechips based? I have been kinda thinking about one of those for a secondary player.
17:50:36amiconnThat's flash based, correct?
17:50:48linuxstbYes, but only 1GB, and no card slot.
17:50:55amiconnhmm :\
17:51:01amiconnDAB would be nice...
17:51:09linuxstbIt's not a great target, but I find it technically interesting.
17:51:13*amiconn <−− dab fan
17:51:22 Join illissius` [0] (
17:51:34amiconn(unlike dvb and that hdtv crap)
17:51:47linuxstbThe DAB tuner is only Band III, so won't work on every German DAB multiplex IIUC.
17:52:24linuxstbThe Cowon D2 with DAB seems nice, but has a touch screen, and I share your dislike of them.
17:52:27amiconnIn germany the majority of multiplexes uses band II as well. L band is too expensive because of its short range
17:52:36amiconn*band III
17:52:40Soaplinuxstb: If/when telechips becomes the next supported arch., how many current-gen targets will that open up?
17:52:53linuxstbI've listed some on the TelechipsInfo wiki page.
17:53:08*amiconn has both a car stereo and a hifi tuner with DAB support
17:53:10linuxstbI'm not sure how many of them are considered "current gen" though.
17:54:36 Quit zardos ("CGI:IRC")
17:54:55SoapJust curious, not trying to make you do the legwork if you don't know off the top of your head.
17:54:56lostlogiciAudio7 is in production afaik
17:55:34linuxstblostlogic: Yes. TMM has just bought a 16GB version, and is working on the Rockbox port. He also bought an iAudio 6, which he says is the same hardware, but with a hdd.
17:55:41linuxstb(0.8" hdd iirc)
18:00:26 Quit Crash91 ("Bye Bye!")
18:02:34 Quit lazka ("I'm off now")
18:04:29lostlogicoh yeah, it look the same even.
18:04:39jhMikeSso we're stuck with a stuck build for awhile? yay
18:04:39 Quit Frazz (Read error: 104 (Connection reset by peer))
18:05:19 Quit illissius- (Connection timed out)
18:05:59linuxstblostlogic: It looks nice, but lacks expandable storage...
18:07:05 Join scharkalvin [0] (
18:07:18 Join geek0091 [0] (i=4c06a06c@gateway/web/cgi-irc/
18:07:55geek0091anyone here?
18:07:56lostlogiclinuxstb: yeah, so I see −− 16G is a kinda inconvenient size fo rme −− not enough to hold all of my music but more than a days worth *shrug* still it looks like a nice player. I know a coworker is getting one (regardless of rockbox support).
18:08:07 Join webguest80 [0] (i=4ae94966@gateway/web/cgi-irc/
18:08:40 Join geek0091_ [0] (
18:08:50geek0091_anyone here, I need help
18:09:36jhMikeShelp with rockbox or your taxes or something?
18:10:03linuxstbgeek0091: You need to ask a question
18:10:05jhMikeSjust ask the question
18:10:06*geek0091 please
18:10:40markunlinuxstb: did you find any nice HDD based telechips DAPs?
18:10:53scharkalvintype everything on one line, don't keep pressing <return> till your thought is complete.
18:11:13geek0091_when I try to install rockbox in the sandisk e260r, I get an error msg after updating the drivers, and runing it, and it says it runs successfully on the computer, but on my sandissk, it sats Unknow bootloader, can install
18:11:34 Quit geek0091 (Client Quit)
18:11:40 Quit kugel (Read error: 104 (Connection reset by peer))
18:11:48 Nick geek0091_ is now known as geek0091 (
18:12:14scharkalvinSounds like sansapatcher ran ok, but you did not unzip the rockbox software correctly to the root directory of your sansa.
18:12:36linuxstbmarkun: No, I haven't seen any.
18:12:43 Join kugel [0] (i=kugel@unaffiliated/kugel)
18:12:56geek0091nonono, when I try to update the patcher to get the e260r to e200, after updating the service driver, it give me an unkown bootloader error
18:13:03Soap0.8" HDD? Is that a microdrive?
18:13:12linuxstbscharkalvin: sansapatcher isn't used for the e200r install
18:13:30geek0091it uses a patcher
18:13:55scharkalvinAh, I missed the "r" part. But I thought the newest bootloader (version 0.5) fixed that.
18:13:57geek0091and you have to update the driver, to the e200 without hte extra security feature
18:14:23geek0091note the "r"
18:14:48linuxstbI don't know what you mean by "update the driver". The procedure is: 1) Run e200rpatcher to patch the Sansa bootloader. ; 2) Install the Rockbox bootloader and Rockbox itself.
18:15:17geek0091when I do that, it gives me the error though
18:15:33geek0091it says download the e200 manufac driver
18:15:36linuxstbDo what, I listed two things...
18:15:56geek0091in the tutorial online, it says to do that
18:16:17linuxstbPlease don't use "it" and "that" when it isn't obvious what you're referring to.
18:16:38linuxstbI assume you're following these instructions?
18:17:08linuxstbAnd you're working through "Step 1 - Original Sansa bootloader patching" ?
18:17:58***Saving seen data "./dancer.seen"
18:19:38geek0091the install instructions are to broad, so I don't know if I missed something
18:19:52linuxstbWhen you said "I get an error msg after updating the drivers", what was that error message?
18:20:15geek0091unkown bootloader, when I run the patcher
18:20:24 Quit scharkalvin ("using sirc version 2.211+KSIRC/1.3.12")
18:20:50 Join donutman25 [0] (n=dagni@
18:20:59linuxstbSo e200rpatcher worked - i.e. it succeeded in uploading the patching program onto your Sansa, but when that program ran on your Sansa, it displayed an "unknown bootloader" message on your Sansa's LCD?
18:21:31geek0091when I just ran the patcher, on a fresh version of the e260r firmware, it says that the upload of the application failed, so im puzzled.
18:22:18linuxstbAre you sure you have an E260R, and not a normal E260? You could try connecting your Sansa to your computer normally, and then run sansapatcher to see if it detects it.
18:22:53geek0091how do you run it
18:23:07linuxstbIt's described in the Rockbox manual for the e200
18:23:22geek0091and in the menu, it has the rhapsody, and on the back, non the less
18:26:26 Join Xerion [0] (
18:26:28 Join petur [0] (
18:27:36geek0091trying the normal install real quick
18:28:07Nico_Pbertrik: skipping to the OGG track you sent me works fine on my gigabeat
18:28:12Nico_PI'll try on the H300
18:28:39geek0091it booted into rockbox, how do I boot into the origanal one now? just wondering
18:30:01linuxstbYou will need to hold a key whilst the Sansa is booting. I don't know which one - it should be in the manual.
18:30:08bertrikNico_P: it works for me when I select the tracks from the files browser, it hangs when I use the database ... :?
18:30:25Nico_Pbertrik: ah, that's a good indication
18:31:10geek0091how do I run my music now
18:33:08geek0091my sandisk wont turn off now, and the screen is black
18:33:32 Quit Sedgewick (Read error: 110 (Connection timed out))
18:34:44JRoThold the power button for about 15 seconds geek0091
18:35:16geek0091how do you boot into the origanal firmware now though
18:35:28rashergeek0091: You hold left/prev while booting
18:35:36rashergeek0091: This is covered in the manual
18:35:58geek0091were is it
18:36:27rashergeek0091: It would suite you to make a minimum of effort to find the question to your answers.
18:36:32JRoTbtw on my sansa e260 i installed the new bootloader, but still i builds a DB in OF on usb connect
18:36:52JRoTanyone know how to fix that?
18:37:05geek0091kk, well mines working now, I think i installed a regualr e200 when the select and record button, so im good
18:37:48geek0091tahnks all
18:37:50bertrikJRoT: that may have something to do with multiple copies of the NVPARAMS data IIRC
18:38:04bertrikthat's where the database refresh bit is stored
18:38:09JRoTah oke
18:38:43bertrikthe latest sansapatcher wipes the NVPARAMS to make sure that the bit is at the expected address
18:39:02bertrikbut I don't know any other details about it
18:39:29JRoToke will take a look at the site for the latest bootloader
18:39:40 Join kripso [0] (
18:42:18amiconnBagder_: ping... ...
18:43:00 Quit kripso (Read error: 104 (Connection reset by peer))
18:43:19 Join kripso [0] (
18:44:41 Join DanielSiebrecht [0] (i=57b0f4d5@gateway/web/cgi-irc/
18:49:43 Join TotallyInfected [0] (
18:50:37 Quit idnar (Nick collision from services.)
18:50:37 Quit DanielSiebrecht ("CGI:IRC (EOF)")
18:50:39 Join idnar_ [0] (i=mithrand@unaffiliated/idnar)
18:51:39 Join Daniel_S [0] (i=57b0f4d5@gateway/web/cgi-irc/
18:51:46Nico_Plostlogic: here?
18:54:21 Quit ilgufo (Nick collision from services.)
18:54:24 Join gufo [0] (
18:57:25 Quit kri_ (Read error: 110 (Connection timed out))
18:59:40bertrikNico_P: any luck reproducing FS #8037 yet?
18:59:42 Join Bam1 [0] (
18:59:58Nico_Pbertrik: no, not yet
19:00:21 Join Llorean [0] (n=llorean@
19:00:29Bam1compiling rockbox is soo much easier on UBuntu
19:01:40bertrikyeah, most importantly: much faster!
19:03:53geek0091were can you find rockbox addons, like other applications
19:04:37 Join BHSPitLappy [0] (n=Me@unaffiliated/bhspitmonkey)
19:05:56geek0091Ubuntu is great version of linux, I prefer knoppix though
19:05:58PaulJamgeek0091: maybe you can find something here (but you need to compile them yourself):
19:05:59rashergeek0091: there aren't many - pretty much every useful one is already included in the install
19:06:16 Join kkurbjun [0] (
19:06:24geek0091ok, thanks, but I also heard about doom on rockbox, whats that about
19:07:03PaulJamdoom is included, see in the manual how to use it.
19:07:50kkurbjunlinuxstb: arm armv4 binaries not completely compatible with an armv5 processor?
19:10:01geek0091thank you all for getting my rockbox to work, I hope that my c and C++ skills may contribute to the project when I can, anyways, thanks for your help :)
19:10:16 Quit Rondom (Remote closed the connection)
19:11:02*bertrik is glad that rockbox isn't written in C++
19:11:28 Join Rondom [0] (
19:11:49*Bam1 laughs
19:11:55 Quit geek0091 ("ChatZilla [Firefox]")
19:12:43Bam1i love the gAIM irc feature
19:12:51 Join pixelma [0] (n=Marianne@rockbox/staff/pixelma)
19:12:57Bam1possibly wrong place xD
19:13:50linuxstbkkurbjun: AFAIK, they should work, but you want gcc to know exactly which processor you're using, so it can optimise correctly, and also make use of the extra instructions available on armv5
19:15:03linuxstbkkurbjun: But I don't think my gcc patch is working correctly for ARM926EJ-S - I tried an m:robe build for that CPU, and it failed.
19:15:19kkurbjungotcha, maybe I should set the m:robe up to compile for the arv5TEJ.. I wonder if there's any benefit from letting it know it's a J version
19:15:34kkurbjunwhere is it failing?
19:15:39linuxstbDid you read my mailing list post about this issue?
19:16:01Bam1is the mailing list email or IRL mail?
19:16:09kkurbjunI saw it, I don't recall the details, I'll go back and reread
19:17:52linuxstbIt failed with the same message I got originally when compiling with the old Rockbox arm-elf-gcc
19:18:44linuxstbI don't completely understand how the multilibs are supposed to work, but I think we'll need to add an alias so that arm926ej-s uses the armv5 libs.
19:19:12kkurbjuncan't we specify the core, ie, armv5tej?
19:19:32kkurbjunor the arch
19:20:20markunonly says "telechips"
19:20:41markun2.5" HDD I think
19:21:28linuxstbAh yes, I remember that now - the person who runs that website posted to the forums offering assistance...
19:22:33markunweren't you looking for a 2.5" HDD player anyway?
19:23:23linuxstbYes. That may be interesting, but I don't want any more players at the moment...
19:23:38 Nick idnar_ is now known as idnar (i=mithrand@unaffiliated/idnar)
19:24:51 Join webguest74 [0] (i=4ca81f01@gateway/web/cgi-irc/
19:25:13webguest74uhh hellloz
19:25:40webguest74duudez i need he;p
19:26:14linuxstbJust ask.
19:26:23 Quit webguest74 (Client Quit)
19:26:32 Join tim [0] (i=4ca81f01@gateway/web/cgi-irc/
19:26:48 Quit donutman25 (Remote closed the connection)
19:27:15 Nick fxb__ is now known as fxb (
19:27:40timneed help heere
19:27:45linuxstbtim: Just ask...
19:27:48Bam1PMFG my computer only has 64mb of ram
19:28:10*Bam1 kills self
19:28:31LloreanBam1: This is an on-topic channel. Please remember.
19:29:00Bam1oh yeah
19:30:00*Nico_P has a mutex leak somewhere
19:30:48timoh my bad. im new here. well i just got rockbox yesterday. installed it and it works fine. i have a 5.5g 80 gig ipod vid. i download the coverflow theme for it. but it wont go into coverflow mode. the dude who made it says"This is my first WPS for the iPod. Inspired by the iPhone and its cover flow feature, I came up with this. Enjoy! Tested with EvilG's Fusion build, but should work on any build with AlbumArt, Bitm
19:32:44timwhat are fusion builds? what patches?
19:32:51markuntim: you are out of luck here if it only works with unoffial builds
19:33:04markunbut they have a separate forum section
19:33:21timno.. i downloaded it from this website!
19:33:58kkurbjuntim, you need a custom build, we only test/support the official build without any patches
19:33:58alienbiker99tim. it needs a special build. and it doesnt work like you think. its not the actualy coverflow
19:34:31bertriktim, you downloaded it from
19:35:15timthats where i got it from
19:35:18bertrikI think that's not part of the official rockbox website, although it looks very similar
19:36:16timohh. where can i get the official themes from the official rockbox website?
19:36:22alienbiker99coverflow would be a cool addition if album art is ever integrated
19:37:21timapple sucks big nuts for not sending a update for oldr ipods to get coverflow
19:38:50timayo it says i need a EvilG build. Whats that?
19:38:58Lloreantim: That's a somewhat invalid assumption to make: that they can. Older iPods are a lot slower than what the newest ones are expected to be.
19:39:17krazykittim, it means it needs an unsupported build, which you can find in the forums.
19:40:23timoh ok llorean. thx. i didnt know. but what build am i supposed to download? i clicked the link and it showed a lot of thinggs
19:41:16 Join FOAD [0] (
19:41:55kkurbjunlinuxstb, is your patch applied by default when building with, or is there a flag?
19:42:33timo one more thing. what is ipod linux? and can it go on my ipod?
19:42:34kkurbjunI'm not seeing the changes you talked about in the mailing list message
19:43:08linuxstbkkurbjun: It should be included by default now.
19:43:19linuxstbAssuming it's been fixed...
19:43:32kkurbjunand it should be patching t-arm-elf?
19:44:23linuxstbYes, the patch seems to be on the server now -
19:44:44timlinux. do u know if ipod linux can go on my ipod. and if it can, how can i download it and install it? i have a 5.5g ipod vid 80 gig
19:45:05linuxstbYes, I think it can, but I've no idea how. Check at
19:45:36timthx man. oh and whats that link for that yu posted?
19:45:39kkurbjunlinuxstb, yeah, that doesn't seem to be included on my run, I'll have to go back and patch
19:46:22linuxstbkkurbjun: I've tried finding documentation on how to use the MULTILIBS options, but can't...
19:46:53linuxstbtim: A patch to gcc. Nothing that affects Rockbox on ipods.
19:46:58timwhat are patches?? im confused
19:47:30timuh whats a ggc? yea ima newb
19:47:32kkurbjundo the newer versions of gcc include the multilib features for the other cores we are targeting.. I'm not really sure what this whole multilib business is about.
19:47:35linuxstbtim: Just ignore us...
19:48:08 Quit tim ("CGI:IRC (EOF)")
19:48:26linuxstbkkurbjun: From what I understand, "multilib" is a way to build different versions of libgcc (and the other libs gcc uses) for multiple architectures.
19:48:35 Join Deeno [0] (i=4ca81f01@gateway/web/cgi-irc/
19:48:45linuxstbBy default, only one libgcc is build.
19:48:49 Part Bam1
19:49:08kkurbjunahh, I see, so it wouldn't be included by default on a later version of gcc anyway
19:49:46linuxstbI don't think so, no. The number of combinations is endless...
19:49:59 Quit JETC- (".UPP.")
19:50:34linuxstbDeeno: Hello tim.
19:50:40 Join JETC- [0] (
19:50:43DeenoHey linux!
19:51:21Deenodont know why it signed me out. just did. and i couldnt go back as tim
19:51:55linuxstbkkurbjun: I think my way of building an armv5 lib is wrong - I'm specifying cpu=arm9e, so when gcc looks for an arm926ej-s lib, it doesn't find one, so defaults back to the armv4 one.
19:52:48keen-hmm, for rockbox it says there's .mod support for sansa c200 but they don't show up in my player's listing... am i doing something wrong
19:52:58kkurbjuncan you use -march=armv5 instead of specifying the cpu?
19:53:14 Part pixelma
19:53:17linuxstbMaybe, yes.
19:53:40kkurbjunis there any benefit for specifying the cpu instead of the arch?
19:53:50crashdwhat's the current run of mp3 players (ipod barred) looking like ?
19:54:06crashdi need to replace my 5g 30gb, and i'd prefer a rockbox enabled product, or at least, one that doesn't totally suck
19:54:06 Nick gufo is now known as ilgufo (
19:54:13linuxstbkeen-: What is "it" ? There isn't .mod support in Rockbox, but there's a patch for it.
19:54:36keen-oh, well that would explain it then
19:54:58linuxstbcrashd: In terms of hard-disk based DAPs, there aren't any currently being made afaik.
19:55:11Deenohow do u install patches from this place : what are these builds here? how do u install the patches here and the builds?
19:55:20linuxstbkkurbjun: I'm not sure.
19:55:40keen-do i have to have some kind of special dev kit to patch and compile using the mp3 player? or how does that work, i'm unfamiliar with compiling on non-x86 platforms
19:56:06crashdlinuxstb: that's a shame ;\
19:56:10kkurbjunhmm, I'll have to play around with it more later when I get time, I have to run for now.
19:56:12crashdi dont really wanna downgrade to ~16gb or whatever
19:56:14linuxstbDeeno: "patches" are changes to the Rockbox source code. To use them, you need to apply those changes to a copy of the source code, and compile your own version of Rockbox. Or use a pre-compiled version someone else has made which includes those patches.
19:56:22linuxstbcrashd: We all agree with that...
19:56:38crashdthe battery on my 5g has gone to the dogs and it's time to get a new one tho
19:56:43 Quit amiconn (Nick collision from services.)
19:56:50 Join amiconn [0] (n=jens@rockbox/developer/amiconn)
19:56:59linuxstbkeen-: The Rockbox wiki has instructions on setting up a dev environment - see the "index" page, and then the "For Developers" section.
19:57:36Deenooh mk thx. how do u install those things
19:57:47keen-ok thanks :o
19:58:06linuxstbcrashd: If I didn't own any Rockbox devices, I would probably buy a Gigabeat F - they're relatively cheap, sound good, have a fast CPU and work well with Rockbox.
19:58:55linuxstbe.g. a used F40 (40GB) can probably be found for around $100 USD on ebay.
19:59:12 Join jhulst_ [0] (n=jhulst@unaffiliated/jhulst)
20:01:01BigBambicrashd: If you don't want radio or recording, gigabeat F is a very good choice
20:10:18 Join rep|icant [0] (n=rep|
20:11:18lostlogicNico_P: just popped back in, wasup?
20:13:41Nico_Plostlogic: in your changes to add_handle, the size param switched from size_t * to size_t, but the comment still states that on return, size will contain the real allocated data
20:14:56 Join Falen [0] (i=51edfa08@gateway/web/cgi-irc/
20:15:04 Quit Deeno ("CGI:IRC (EOF)")
20:15:13lostlogicNico_P: my bad −− forgot to change the comment. the function now only returns success if either the caller doestn't reuiqre the full size, or the full size is allocated.
20:15:38lostlogic*doesn't require
20:15:51Nico_Plostlogic: yeah, so the caller doesn't need to know what size was allocated anymore?
20:15:52 Quit midgey ()
20:16:11 Quit Falen (Client Quit)
20:17:02 Join Gurkan [0] (
20:17:23 Quit Gurkan (Client Quit)
20:17:40 Join Falen [0] (
20:17:44 Quit rep|icant_ (Read error: 110 (Connection timed out))
20:18:02***Saving seen data "./dancer.seen"
20:18:49lostlogiccorrect −− it's a pass-fail operation now. Committing detailed commentage now.
20:19:37 Quit Daniel_S ("CGI:IRC")
20:19:56Nico_Plostlogic: also I've implemented jhMikeS' suggestion of having one mutex per handle
20:20:21 Join K4rP4D [0] (n=KrPD@unaffiliated/krpd)
20:20:35Nico_PI lock it on almost every operation except buffering
20:20:38lostlogicsounds... strange, I was ignoring that discussion because I didn't get the cost benefit equation.
20:21:02Nico_PI'm not sure it's of any use either...
20:21:22lostlogichave a patch posted somewhere so I can maybe understand the tradeoffs in code better?
20:21:37Nico_PI'll post it now
20:22:19scorcheBagder_: around?
20:23:02 Join Dark_Apostrophe [0] (n=darkapos@supporter/monthlybyte/DarkApostrophe)
20:23:09Nico_Plostlogic: I'm also trying to do more things in the main buffer as opposed to the static mp3entry structs
20:23:16Dark_ApostropheIs there a problem with mpegplayer in the current Rockbox release?
20:23:27Dark_ApostropheI just upgraded to today's version
20:23:38Dark_ApostropheAnd the video is very fast, while the audio is normal
20:23:43lostlogicNico_P: that makes sense −− less memcpy in typical use.
20:24:00crashdwhen i initially bought my ipod i considered a gigabeat
20:24:12Nico_Pyes, and also it should help with the runtimedb bug
20:24:29lostlogicNico_P: definitely.
20:24:51lostlogicNico_P: I'd be concerned about the effect of a move_handle on the status tracking :-
20:25:37Nico_Pyeah I'm aware of that. the WPS uses audio_current_track to get the mp3entry*, and the codec uses callbacks to set the elapsed time
20:26:00Nico_Pbut I don't think I can completely get rid of the curtrack_id3 static struct
20:26:01 Join BiptoN [0] (i=4cc869a6@gateway/web/cgi-irc/
20:26:29lostlogicyeah, definitely not −− but there does need to at least be some (possibly fully mutex protected) way to copy-back onto the main buffer
20:26:34BiptoNhello everyone
20:27:50amiconnBagder_: ping ...
20:27:53BiptoNhas anyone attempted to port rb to a microtrack 2496 yet?
20:27:56Nico_Plostlogic: for the mutex business... nothing very interesting, just a bunch of mutex_(un)lock calls
20:30:12 Join Guile`` [0] (n=Guile@
20:31:31jhMikeSNico_P: why? I wasn't sure of the usefulness either. Just odd schizophrenic thinking it was. :P
20:31:52PaulJamDark_Apostrophe: do you have "Limit FPS" enabled in mpegplayer?
20:32:01Nico_PjhMikeS: why what?
20:32:12Nico_Pwhy did I write it?
20:32:16lostlogicNico_P: So I can see some real benefit if there was some way to hold the handle's mutex locked across the move_handle call in shrink_handle
20:32:20jhMikeSIf no usefulness was found why bother?
20:32:43lostlogicNico_P: then we could safely do keep some accounting data on the buffer and alter it (with handle mutex held)
20:32:55Dark_ApostrophePaulJam: No
20:32:59lostlogic(of course we still wouldn't want to alter it on every packet decode because that would be SLOOOWWWW)
20:33:04lostlogic(I think)
20:33:05jhMikeSMutexes can't be moved while threads are block on it anyway since they contain pointers and the thread references it.
20:33:14Nico_PjhMikeS: good question :p I think I wanted to see if I was going to hit an ugliy blocking situation proving the code wrong
20:33:18lostlogicjhMikeS: :(
20:33:36PaulJamDark_Apostrophe: then enabe it (you should enable "Skip Frames" too)
20:33:45Dark_ApostropheOh, okay
20:33:49jhMikeSthey'll need to contain more pointers when priority inheritance is implemented
20:35:20 Join webguest09 [0] (i=4a469ecd@gateway/web/cgi-irc/
20:36:09 Join Frazz [0] (
20:36:09lostlogicNico_P: the only way that I can think of holding soem state safely over a move_handle would be to directly copy the data from old-to-new locations of the struct on buffer while holding the lock of _both_ the old and new copies.
20:36:24webguest09is there a way to have rockbox, e200 firmware, and rhapsody firmware on my e200r?
20:37:00lostlogicmove_handle would hafta initialize the new mutex location
20:37:18lostlogic(and that should be safe, since my changes to move handle require that the move is at least the length of a memory_handle)
20:37:23jhMikeSwhy need to lock the new copy? who would know about it?
20:38:06 Quit BiptoN ("CGI:IRC (EOF)")
20:39:58 Quit Falen ("ChatZilla [Firefox]")
20:41:06webguest09maybe something like holding down the << key to go into rhapsody firmware, the center button to go into e200 firmware, and then rockbox on normal boot.
20:41:22lostlogiconce we unlock the llist_mutex anyone could have found out about it.
20:41:38 Quit TMM ("Ex-Chat")
20:41:49lostlogicis it possible to initialize a mutex to a locked state?
20:41:52scorchewebguest09: you would have to edit the source of the bootloader, but yes, that would be possible
20:41:53jhMikeSfinish the move then refer to the new location
20:41:55Nico_Plostlogic: doesn't the llist_mutex protect the moving?
20:42:21webguest09how would i edit that?
20:42:22jhMikeSlostlogic: nope. mutexes are owned. semaphores can be.
20:42:27Nico_Plostlogic: as long as users don't keep the pointer to the old location it's safe
20:42:49jhMikeSIf you want it locked. init it. call lock_mutex
20:42:57Dark_ApostrophePaulJam: Works just fine now :D
20:43:51jhMikeSNico_P: that should be illegal use without a reference count where they say "I don't need this data"
20:43:55Dark_ApostrophePaulJam: Thanks
20:44:50Nico_PjhMikeS: agreed. btw, how do you think writing to a handle could be done?
20:46:30jhMikeSso you need to change some members after it's passed to other code, right?
20:46:49 Quit webguest09 ("CGI:IRC (EOF)")
20:47:25Nico_Pno, actually simply letting users write to the handle's buffer...
20:47:41Nico_Pe.g. modifiy a struct mp3entry
20:49:13 Join mako [0] (
20:49:20 Part mako
20:49:52lostlogicNico_P: yes, it protects the actual moving, but it's unlocked by the time shrink_handle has an opportunity to lock the per-handle mutex
20:50:13jhMikeSone thing I'm not sure of here is which functions are actually called on the order of thousands of times per second?
20:50:20lostlogicnot that that's a possible problem since it doesn't yield in between :-P
20:50:40lostlogicjhMikeS: nothing in buffering.c should be called that often (right Nico_P)
20:50:55Nico_PjhMikeS: find_handle is called the most. but I don't think any would be called thousands of times per seconds... maybe tens of times per second
20:51:29jhMikeSso nothing is really a big deal there. Frankly, I think users of data should protect the actually data themselves.
20:53:08Nico_Phmm anyway with cooperative if I do "bufgetid3(CUR_TI->id3_hid)->elapsed = value - latency;" there's no risk of the handle having moved while I write
20:53:09jhMikeSWhatever method best suited I suppose. You could also use a mutex pool but then care must be used
20:53:49 Quit webguest80 ("CGI:IRC (EOF)")
20:53:52jhMikeSthat's true...and I'm going to rush ahead on preemtive just for that :P
20:54:37Nico_Phehe... with preemptive I could implement something like lock_handle and unlock_handle internally using a mutex for each handle
20:54:42 Join MournBlade [0] (
20:54:54Nico_Plock_handle; write; unlock_handle
20:55:10jhMikeSwhat's that got to do with it? you could do that now.
20:55:23Nico_PI could but do I need to?
20:55:32lostlogicI hope rockbox never becomes preemptive.
20:56:07jhMikeSlostlogic: I hope it does and I'm not the only one. Too many advantages and hardly any disadvantage.
20:56:42*Nico_P has nothing against it
20:56:54 Quit kubrick (Read error: 113 (No route to host))
20:57:13Nico_Pwith dual core we need to think preemptive anyway
20:57:20jhMikeSI hate slow code controlling time-critical code and low-latency audio is very desireable. Cooperative can't really provide that very well.
20:57:49 Quit amiconn (Nick collision from services.)
20:57:56 Join amiconn [0] (n=jens@rockbox/developer/amiconn)
20:58:37jhMikeSNico_P: with COP it's even better since it's actually parallel. :)
20:59:24Nico_PjhMikeS: If I were to implement the (un)lock_handle mecanism I described above, I'd need to actually use the mutex everywhere the handle is accessed, right?
20:59:41jhMikeSI'd never want GUI widgets to be MT like that or anything.
21:00:51jhMikeSNico_P: I think intimate knowledge about the details should say where and when. If by accessed you mean only the handle itself perhaps.
21:01:22 Join kri_ [0] (
21:01:58 Join MethoS- [0] (
21:02:02Nico_Phmm I need to think about it... Ideally I could get completely rid of the static mp3entries
21:02:11Nico_Pthat would require changes to the codecs though
21:02:35 Join desowin [0] (
21:02:35jhMikeSIt's like mutexing the file system or something. Who cares what threads do to the file if the shares it? We only care about the FAT and the descriptors themselves. The app must worry about it's own data.
21:02:41 Quit TotallyInfected ()
21:04:03*Nico_P needs to be off
21:04:47 Quit Nico_P (Remote closed the connection)
21:08:38jhMikeShmm...we have a handle cache. could be the ticket if it has many slots.
21:09:39 Join Thundercloud_ [0] (
21:09:55 Quit jhulst_ (Read error: 110 (Connection timed out))
21:12:18 Quit amiconn (" HydraIRC -> <- Like it? Visit #hydrairc on EFNet")
21:13:32 Join webguest35 [0] (i=4a469ecd@gateway/web/cgi-irc/
21:14:16 Quit webguest35 (Client Quit)
21:15:00lostlogicwe definitely need a way to populate the next_track id3 even if its not on the buffer... that's been broken since MoB went in I think
21:16:01jhMikeSit occurs to me next track id3 could be allocated on the buffer and not need a static buffer. it's only need when using the playback engine right?
21:16:35jhMikeSI just mean something similar to the codec malloc buffer
21:18:06lostlogicah yeah, it could be 'leased' during playback in the same way, when the buffer is reset
21:18:41 Quit kripso (Read error: 110 (Connection timed out))
21:20:18 Join webguest27 [0] (i=4a469ecd@gateway/web/cgi-irc/
21:20:40 Quit Thundercloud (Read error: 110 (Connection timed out))
21:24:53jhMikeSwe could lease all kinds of stuff there and not waste space permanently. actually, stopping playback, leasing space and then restarting it is really simple.
21:25:03webguest27i have a sansa e200r and when i put it into recovery mode and connect it, the 16mb frive doesn't show. I'm using windows me. I've put it into recovery mode on windows xp before and the drive showed. My quetion is, is there someway i can make this 16mb drive appear on windows me. Is there something i could download? Any info would help. Thanks
21:25:15lostlogicoh, builds are still stuck.
21:25:18 Quit spiorf (Connection timed out)
21:25:33jhMikeSstill? no swedes around today?
21:25:42scorchewebguest27: well, this is more of a windows issue =/
21:25:44lostlogichaha, apparently not :)
21:26:22 Join TMM [0] (
21:26:27 Join TotallyInfected [0] (
21:27:22webguest27yeah, i was just wondering if there something i could download to make it show but i guess not.
21:29:21webguest27i was trying to convert my e200r into a e200 but the directions to do this say i need to drop certain files in recovery mode
21:29:22BigBambiwebguest27: No, it is not displayed over USB
21:29:51BigBambiYou cannot see the firmware partition of an R over USB, that is partly why installation is an arse
21:31:43 Join advlaptop2019 [0] (n=advcomp2@unaffiliated/advcomp2019)
21:31:45 Quit kri_ (Read error: 104 (Connection reset by peer))
21:32:12 Join kri_ [0] (
21:32:55 Quit ilgufo (Remote closed the connection)
21:33:37 Join ilgufo [0] (
21:34:47*jhMikeS should forget about a small size increase from inlining software corelocks since they reduce heavy scheduler use performance by about 25%. It would like like help alot.
21:35:15jhMikeSs/like like/likely/
21:35:39 Join kkurbju1 [0] (
21:35:58webguest27one last question, once the bootloader is patched, will it remain patched for good or is there a way to unpatch it?
21:36:01 Quit kkurbjun (Nick collision from services.)
21:36:08 Nick kkurbju1 is now known as kkurbun (
21:36:26 Nick kkurbun is now known as kkurbjun (
21:37:14BigBambiwebguest27: You can undo it. I believe it involves using an unpatched bootloader in either recovery or manufacturing mode, but I haven't got a Sansa so I'm not too sure
21:37:34 Quit desowin ("use linux")
21:38:25 Quit quaal ("Leaving")
21:40:42webguest27okay, so installing new firmware won't unpatch it?
21:40:48jhMikeSlostlogic: why couldn't we cache like 16 handles in a direct-mapped cache?
21:41:27 Join spiorf [0] (
21:41:57 Join Daniel_S [0] (i=57b0f5da@gateway/web/cgi-irc/
21:42:51BigBambiwebguest27: Coming a new rockbox build won't
21:42:58BigBambiWhat do you mean by new firmware
21:44:19BigBambi-coming + copying
21:46:18rasherwebguest27: Removing the bootloader is done by simply running sansapatcher and select uninstall rather than install. (or doing the same in RButil). Assuming it's not a e200R
21:46:42BigBambirasher: It is
21:47:01BigBambiindeed so
21:47:36 Join karashata [0] (
21:48:25webguest27thanks for the info.
21:48:26*jhMikeS wonders where all these "R" models come from. Not even BestBuy had them on the shelves at last look.
21:48:52 Join crashmatrix [0] (
21:50:19webguest27if i already have the bootloader patched and i put e200 firmware on the root and reboot, will it be e200 firmware or mess up my sansa
21:50:22jhMikeSI really have no idea where you get one...sorry to say :)
21:50:38webguest27i got mine at bestbuy
21:50:40 Quit K4rP4D ("Leaving")
21:52:24krazykitwebguest27, it probably won't update or anything, really. nothing that can't be recovered from, anyway
21:54:12webguest27so if i put the e200 firmware on the root and reboot, i'll still have the rhapsody firmware? I am wondering if doing this would give me e200 firmware because i don't want rhapsody anymore
21:55:07BigBambiwebguest27: I think there are instructions in the wiki to turn an E200r into an E200
21:59:08 Quit kugel ("ChatZilla [Firefox]")
21:59:57 Quit crashmat1ix (Read error: 110 (Connection timed out))
22:00:49 Part Dark_Apostrophe ("Tu ne cede malis, sed contra audentior ito")
22:13:12preglowjhMikeS: what did i change where?
22:18:04***Saving seen data "./dancer.seen"
22:19:01 Quit ilgufo ("So Long, and Thanks For All the Fish -")
22:20:11jhMikeSpreglow: saw your moniker in wm8731.c as the last one to commit to that file
22:21:56jhMikeSjust too lazy to look, I am
22:22:39jhMikeSI also want SW tone controls available on everything. All the best sounding targets IMO are also the ones that use them. The HW ones are just plain inferior (too steep). :)
22:23:11preglowjhMikeS: i probably just fixed up the stereo width step size
22:23:31preglowhah, nice
22:23:40preglowbut having free tone controls are also nice
22:23:44preglowall ipods have sucky ones, though :/
22:23:53jhMikeSnot when they sound the shiite
22:24:21preglowi wish there was a way of having an option for it that i wouldn't also be against introducing :)
22:25:09jhMikeSOn gigabeat they are free :) and the HW ones sound horrible. The dynamic bass helps but it's still not so great.
22:25:22 Quit advlaptop2019 ("Leaving")
22:25:45preglowi'm still really annoyed by the overflowing that sometimes happens with eq/tone controls
22:25:52preglowit'll only happen when stuff is clipping very badly, but anyway
22:26:05 Quit mf0102 ("Verlassend")
22:26:46jhMikeSI can actually hear the boost go up suddenly as a bass goes down the scale on GB. No such effect exists with the SW ones with the gentler rolloff.
22:26:47 Quit webguest27 ("CGI:IRC (EOF)")
22:27:26preglowjhMikeS: also, the sw tone controls and hardware scaler functionality aren't in sync, that also annoys me badly
22:28:04 Join webguest67 [0] (i=4ca81f01@gateway/web/cgi-irc/
22:28:06 Quit davina (Remote closed the connection)
22:28:39webguest67helloz. i got a question. im bout to try to hack my cousin's ipod nano. its 2nd gen. can it be hacked?
22:28:50jhMikeSyou mean there's a HW scaler to prevent clipping? AS3514 has an AGC block on the output.
22:30:02 Quit Buschel ()
22:30:03bertrikthe AGC on the AS3514 is currently enabled AFAIK
22:30:09karashatawebguest67: if you mean you're going to try to install Rockbox to it, don't, there isn't a build for the 2G Nano
22:30:16preglowjhMikeS: for gains above 0 db, the volume is adjusted down in filtering, and boosted in the dac
22:30:18jhMikeSyep...not for the mic though. I should enable it there.
22:30:31preglowjhMikeS: so we can do positive gains with no clipping if the analogue volume isn't maxed
22:30:45karashatawebguest67: otherwise, you're in the wrong channel for "hacking" it...
22:30:52webguest67damn. karashata, is there anything i can do to it?
22:31:05jhMikeSon that one I think it should work that way and won't let analog clip.
22:31:05 Join davina [0] (
22:31:41karashatawebguest67: look into iPodLinux and see if they have any information on the 2G Nano
22:31:48karashatathat's all the help I can give you
22:31:57preglowjhMikeS: like, a 1 db analogue gain really is combined with overall gain of -1 dB in filtering with 1 dB boost in dac
22:32:13preglowjhMikeS: problem is, of course, that tone controls has latency, hw does not
22:32:22webguest67Ok Thank You
22:32:54webguest67can u use the ipodwizard on it karashata
22:33:09jhMikeSit's part of the analog block DAC involved
22:33:15karashataI have no idea, I don't own a nano
22:33:46jhMikeSIt comes right after the 5-channel analog mixer
22:33:50bertrikwhat is the 0 dB level anyway? 1 mW or 1V into 16 ohm or something?
22:33:54webguest67oh mk. i hacked my ipod vid and i was wonderin if i can do anything to that. ok thank you very much
22:35:51jhMikeSbertrik: no idea actually for that
22:36:27preglowbertrik: the 0 db level is the maximum gain the hardware can produce with guaranteed no clipping
22:36:51bertrikah ok, makes sense in a way
22:36:52preglowbertrik: it has nothing to do with physical values
22:36:54jhMikeSpreglow: you were pushing for low latency eh? :)
22:37:11 Join amiconn [0] (n=jens@rockbox/developer/amiconn)
22:37:35preglowbertrik: some players can do gains higher than 0 db, but that means the music will clip if it's not recorded with some headroom
22:37:37amiconnBagder_: ping .. .. ..
22:37:42preglowjhMikeS: i don't think i've made a secret of it :)
22:38:05*jhMikeS won't say "preemptive" in certain company .... shhh :p
22:38:10amiconnpreglow: Today I went ahead and reported the msac bug ...
22:38:16preglowamiconn: cool!
22:38:35jhMikeSand they'll just say "well don't use it then" :)
22:39:42preglowamiconn: we don't currently use a patch for binutils, no?
22:39:45jhMikeSI mean just the GCC team really care if it has bugs? It doesn't seem like it.
22:39:50preglowjhMikeS: no, they don't
22:39:56preglowjhMikeS: they suck quite badly at that
22:40:21jhMikeSAnd they probably don't want anyone having commit access there that would actually fix them.
22:40:22preglowyou would think one of the more important pieces of software would care more if they had bugs
22:40:30linuxstbamiconn: It's the law - no Swedes are around when the build server hangs.... ;(
22:40:50preglowand they are still the only people with access to it? :/
22:40:52amiconnIt becomes a major annoyance :(
22:40:55jhMikeSThey just don't want it used without supervision :)
22:41:22 Quit webguest67 ("CGI:IRC (EOF)")
22:41:27amiconnjhMikeS: I prefer analog volume controls on targets which have them any day over the digital ones
22:41:45preglowamiconn: even the ipods -6 -> 9 dB ones? they sound seriously bad
22:41:59amiconnI have no ipod with tone control
22:42:10amiconnThe wm8721 has none
22:42:27jhMikeSamiconn: volume controls? I was talking tone controls.
22:42:28preglowit sounds like crap
22:42:39amiconnMeh, I meant tone controls of course
22:43:11amiconnThe best tone controls so far are the ones in the MAS35xxF though
22:43:17jhMikeSwhich target has analog tone controls?
22:43:32amiconnUDA1380 has them
22:43:57jhMikeSthen why is the cutoff affected by samplerate?
22:43:59amiconnOh, and the G5, altjough they're still mapped as "hw eq"
22:44:26 Join pixelma [0] (i=pixelma@rockbox/staff/pixelma)
22:45:06amiconnjhMikeS: Are they? Rockbox only uses 44.1kHz for main playback so how'd you notice that?
22:45:22 Join webguest55 [0] (i=4ca81f01@gateway/web/cgi-irc/
22:46:24*linuxstb tests FS #7440 before preglow asks again
22:46:53preglowuda1380 doesn't have _analogue_ tone controls, does it?
22:47:02preglowthey're in the codec, but i think they're digital
22:47:09 Quit pixelma (Client Quit)
22:47:22amiconnYes, they're actually digital, but the codec handles them
22:47:38amiconnSame for the MAS, the tone controls aren't really analogue
22:47:58webguest55yeah im bk again. um i saw this list of emulators for linux. somethin called idarcnes, ipod mame,ines, and igamegear. can these be installed on 80 gig 5.5g? and if they can, how cam you install them?
22:48:36preglowamiconn: in that case, the cutoff will probably be different for different sample rates
22:48:47preglowi don't think they actually use separate coefs for separate rates
22:48:52karashatawebguest55: this isn't the channel to ask about that in, this channel is for Rockbox-specific issues
22:48:52scorchewebguest55: well, those are for linux...
22:48:58karashatasorry to say
22:49:20jhMikeSamiconn: it says it is in the datasheet
22:49:27webguest55sryz ppl
22:49:33amiconnpreglow: Possibly, but I still prefer them.
22:49:58preglowamiconn: for good sound or not being software?
22:49:59jhMikeSthey do sound better on that one though. the WM hardware ones are just awful.
22:50:01amiconnWhy waste cpu cycles for something the codec can do equally well, or maybe even better?
22:50:11preglowamiconn: you should try the wm ones, then
22:50:14preglowthey are horrible
22:50:15jhMikeSFor gigabeat, we need to waste a few
22:50:29preglowand coldfire can't even cut
22:50:32preglowonly boost
22:50:37amiconnAgain, I can't... unless you mean the "hw eq" ones on G5, and those are ok
22:51:02preglowanywho, an option would be cool for it, but i don't want more options
22:51:06amiconnWell, while most tone controls allow for both cut & boost, I never needed cut...
22:51:30jhMikeSjust have them both. let the user use them both at once if they're on the cutting edge. :)
22:51:47 Join pixelma [0] (i=pixelma@
22:52:00amiconnjhMikeS: Haha, and then complain why it sounds even more like BoS
22:52:26preglowit annoys me that ac broke being able to easily compile any target using sw tone controls by adding a define
22:52:34jhMikeSsure, why not? it's not my problem. some might like it just to max out the bass.
22:52:36amiconnThere are already both tone controls (either hardware *or* software) and the sw eq - I think that's more than enough
22:54:04preglowme too
22:54:06jhMikeSbeh...I'd just rather replace them on gigabeat and use SW only. I can't even listen to the HW ones there and I just can't match with eq quite right.
22:54:19preglowjhMikeS: i wouldn't really mind
22:54:29preglowit's more of an issue on ipod, which hasn't god cycles to burn
22:54:41preglowthe latency issue still bothers me...
22:56:19jhMikeSThough I'll say I _can_ listen to the adaptive bass mode. It's not quite so bad. Some is clipping problems some is just too steep a filter.
22:57:41preglowif only arm had something like emac, *sigh*
22:58:41jhMikeSbut with COP, SPC's with echo play quite nicely with SW tone controls so why would iPods be so bad off? They actually don't cause too much boosting on sansa or H10.
22:59:03preglowwell, they're just an eq_filter call
22:59:06*jhMikeS mixed up two lines of thought in that sentence.
22:59:16preglowjust one of the filters can be made faster too
22:59:23preglowbut i've optimized for the case of both being used
22:59:38 Join MajorC [0] (
23:00:00jhMikeSMy MP3s boost around 40% with SW tone controls on.
23:01:10jhMikeSsame files about 35% without
23:01:10 Quit webguest55 ("CGI:IRC (EOF)")
23:01:15preglowa quick comparison would be cool
23:01:29linuxstbpreglow: Do you know why safetydan implemented the bass/treble cutoff options as 0,1,2 and 3, rather than the cut-off frequencies?
23:01:40preglowlinuxstb: because cutoff frequencies are sample rate dependent
23:01:45jhMikeSmake that 34% without. So 5-6% more.
23:01:52preglowlinuxstb: they we start using anything but 44.1khz, they'll be wrong
23:02:28preglows/they/the day/
23:02:37 Quit J3TC- (Read error: 113 (No route to host))
23:02:41linuxstbAre you sure? I can't find any reference to that in the datasheet.
23:02:47 Quit JETC- (".UPP.")
23:02:56preglowlinuxstb: i'm quite sure i've seen it in at least one data sheet, it also makes sense
23:03:01 Join J3TC- [0] (
23:03:14preglowlinuxstb: if the filters are digital, and they usually are, they'd have to start changing coefs by sample rate to make cutoff be constant
23:03:22preglowand we're not exactly talking studio audio here
23:03:42 Quit pixelma (" reboot")
23:06:31keanuJdGordon: when you're reading the logs, does e200rpatcher remove the signature check (so that all that'd need to be done to do an e200r -> e200 conversion is patch the BL and put the e200 firmware in recovery mode) or does it do something else?
23:07:04 Quit kri_ (Read error: 104 (Connection reset by peer))
23:07:21 Join JETC- [0] (
23:07:24 Join kri_ [0] (
23:07:35linuxstbThe WM8983 datasjeet states the cutoffs in terms of frequencies, and doesn't have any reference to it being based on a specific samplerate (afaics). But that datasheet obviously isn't 100% applicable...
23:08:43preglowisn't that codec very flexible wrt. choosing sample rates?
23:09:23 Join pixelma [0] (i=pixelma@rockbox/staff/pixelma)
23:10:28linuxstbI think so, yes.
23:10:40preglowthen it's really unlikely
23:10:48preglowit'd have to store a huge table of coefs
23:10:55preglowor calculate them, even more unlikely
23:11:11preglowthis can be quickly resolved by some measurements
23:11:33preglowand i think it's kind of a moot point anyway, the choices are so few the exact frequency isn't really interestingf
23:11:43preglowand tone controls never seem to have exact cutoff frequency controls
23:12:06linuxstbIt just seems odd to see the cut-off freqs as 0, 1, 2 and 3...
23:12:29 Nick Bagder_ is now known as Bagder (
23:12:47jhMikeSaha...the kicker...hello
23:12:58preglowlinuxstb: i don't like it either, but can you think of anything better
23:13:05preglowlinuxstb: also, i just found a reference in the uda datasheet
23:13:44preglow- Two settings: f_c = 1.5 kHz and f_c = 3 kHz assuming sampling frequency is 44.1 kHz
23:13:46jhMikeSpreglow: we change the menu items based on current samplerate ;)
23:13:48*Bagder does the kick dance
23:14:02 Join ptw419 [0] (
23:14:13preglowBagder: why not make a kicker interface that a couple more people can have access to?
23:14:30 Join kripso [0] (
23:14:44jhMikeSus kids cannot be messing around with the big toys :)
23:14:45Bagderyeah I guess it is about time I make one
23:16:16 Quit austriancoder (Remote closed the connection)
23:16:49*linuxstb notices austriancoder has left, so expects Zagor to appear at any moment...
23:17:36jhMikeShaha...I'm sure he thought Zagor might be coming so booked
23:19:17JRoTdoes the sansa charge if i put in usb under rockbox?
23:19:18alienbiker99heh. austriancoder commented on the usb patch today
23:19:52jhMikeSalienbiker99: where? in FS?
23:21:36JRoTdoes the sansa charge if i put in usb under rockbox? anyone??
23:21:55jhMikeSJRoT: low-current mode
23:22:18 Quit MajorC ()
23:22:19JRoTis that a yes or a no?
23:22:23sdit does not
23:23:14jhMikeSlast I checked it seemed to but slowly
23:23:22JRoTok, cause the OF is keeping the backlight on
23:26:16alienbiker99i guess austriancoder didnt know zagor started to work on usb
23:27:14 Quit matsl ("Riece/3.1.2 XEmacs/21.5-b28 (fuki, linux)")
23:27:23 Quit kri_ (Connection timed out)
23:28:12linuxstbHe said he was aware of the flyspray task.
23:30:55*jhMikeS reads much more into the statement...arf "
23:32:09 Join einhirn [0] (
23:33:01 Quit Frazz (Read error: 110 (Connection timed out))
23:33:15 Quit kripso (Read error: 110 (Connection timed out))
23:34:19 Nick parafin is now known as parafin|away (
23:36:30 Join safetydan [0] (i=dc9d468b@rockbox/developer/safetydan)
23:36:58*petur had not even noticed austriancoder was here - need to install dual core cpu and more memory in head :-/
23:37:25preglowlinuxstb: anywho, does the patch work ok?
23:37:39preglowlinuxstb: how is the cutoff menu inserted into the current menu?
23:38:32jhMikeSpreglow: it's only for HW based? can't remember now if you might have mentioned already.
23:39:52preglowjhMikeS: what is only for hw based? if you're talking about cutoff support, i am planning to extend it to other targets too, and eventually sw tone controls
23:40:13jhMikeSyes, I was asking about cutoff
23:40:18safetydanpreglow, linuxstb added a comment to FS #7440. Sounds like it sort of works, but resets the gain every time the cutoff is changed
23:40:54preglowjhMikeS: for swtc we can of course add full proper cutoff support, but i think i'll limit it to a couple of choices
23:40:58*jhMikeS forgot how his own bloody PTS queue worked :p ... it's not quite a normal queue
23:41:09 Quit Robin0800 (" HydraIRC -> <- Organize your IRC")
23:41:38preglowlinuxstb: would "0 (low), 1, 2, 3 (high)" be any better?
23:42:10preglowlinuxstb: on other targets that only support two choices, i would use "low, high"
23:42:12linuxstbI think so. Maybe 1-4 would be better as well.
23:42:19preglowahh, yes, no 0, definitely
23:42:28jhMikeSpreglow: makes sense. did you use the choices I made when you committed it or switch it? maybe that's why I like it so much :)
23:42:29linuxstbsafetydan: BTW, I've just noticed a typo - base_control
23:42:40preglowjhMikeS: can't remember, really
23:42:43 Quit Daniel_S ("CGI:IRC (EOF)")
23:43:02linuxstbsafetydan: Sorry, I meant "base cutoff"
23:43:10preglowlinuxstb: how is the cutoff menu inserted into the current menu?
23:43:20linuxstbsafetydan: (in settings_list.c)
23:43:38preglowhow as in appearance
23:43:58linuxstbpreglow: "Bass", "Bass cutoff", "Treble", "Treble cutoff". That also looks a bit ugly, but I'm not sure how else it could be done.
23:44:05preglowbass/treble/bass cutoff/treble cutoff sounds like too many entries for that single feature
23:44:10jhMikeSjust being a bugger :)
23:44:47preglowbut i don't want them in a sub menu either, bass and treble should be easily accessible
23:44:59 Quit Arathis ("Bye, bye")
23:45:08jhMikeSBass->Cutoff/Gain, Treble->Cutoff/Gain <== ??
23:45:29preglowsame trouble, it buries an essential setting like bass and treble control too deeply, imho
23:46:03linuxstbI had forgotten just how crowded all the settings menus are...
23:47:23preglowthey are that
23:47:28preglowand i'm starting to really dislike it
23:47:32preglowwe need some kind of cleanup
23:47:38jhMikeSpreglow: access issues are a result of the way things are done
23:47:58preglowalso, "general settings" is indeed starting to be a bit too general for me
23:48:02preglowyou have sound settings in one separate thing
23:48:03preglowthen bang
23:48:06pregloweverything else in general
23:48:25linuxstbI think just some gradual evolution is in order. e.g. "Playback settings" could be moved to the top level, next to "Recording Settings"
23:48:52preglowbasically branch essential stuff like that further down
23:49:05pregloweverything you need to just access and tweak once should be buried deeper
23:49:20jhMikeSrethinking of widgets needs to be done and some more advanced stuff than lists used
23:49:36 Quit tpickers66 (Read error: 113 (No route to host))
23:49:46safetydanlinuxstb, I have real trouble not typing base instead of bass sometimes. Did the cutoff setting actually work though?
23:49:49preglowright now we manage screen space really weirdly
23:50:06linuxstbsafetydan: It's hard to tell, as the gain is reset...
23:50:19jhMikeSfrankly it's contrary to the way settings are displayed on pretty much every device I see
23:50:40preglowi don't like how some settings waste screen space like mad
23:50:51preglowlike "sort case sensitive", yes/no
23:50:59preglowwastes all the screen on even h120
23:51:00preglowlooks stupid
23:51:08preglowsome kind of overlay would look better
23:51:21preglowbut i guess this stems for the fact that our display drivers are line based
23:51:24linuxstbAh, viewports...
23:51:25preglowoverlays would be hard to do
23:51:30preglowlinuxstb: yes, viewports ...
23:51:33jhMikeSI think the displaying of the value with the text line would be just fine and pretty much everything in the universe does that
23:51:42*amiconn actually likes those simple lists
23:51:42 Join tdtooke [0] (
23:52:37safetydanyeah, the consistency of a list for everything is nice sometimes
23:52:38preglowjhMikeS: yes, i would like that as well
23:52:42preglowsimpler _and_ prettier
23:52:46amiconnSure, some screens look a bit empty on higher resolution, but it allows precise control and is straightforward
23:53:19amiconnAnd it adapts to almost all font sizes with no problems
23:53:19preglowamiconn: i don't get "precise control", what solution wouldn't offer that?
23:53:21jhMikeSI frankly can't see how precision would be affected at all.
23:53:41 Quit davina (Remote closed the connection)
23:53:42linuxstbThere's something to be said for consistency - the current use of lists gives that.
23:54:00amiconnSetting + value on one line looks really ugly if either the screen is small, or the font is big
23:54:22LloreanWell, if you're changing I/O expectations (for example the graphical equalizer screen) you lose precision in the sense that if you try to treat it like the lists, it won't work.
23:54:23jhMikeSwhy is it that if we use things that people are already used to seeing, it's an issue?
23:54:23pixelmajhMikeS: I was very confused by the recording setting of the X5 original firmware which follows a similar concept, IIRC
23:54:33amiconnThe best (i.e. worst looking) example for this is the trigger settings screen
23:54:40LloreanWasn't there a patch that showed setting values *in* the list?
23:54:42tdtookeI just noticed something I'd like to clarify before I start compiling. I noiced for the commit for r15348 that some defines were added for BUFFER_FULL, DATA_NOT_READY, etc.. in apps/playback.c I noticed -2 replaced with DATA_NOT_READY while buffering.h has BUFFER_FULL as -2. Is that a typo or am I looking at this wrong?
23:54:55*amiconn *hates* that screen
23:54:56jhMikeSthat's the trigger settings without proper design
23:55:10linuxstbtdtooke: Sounds like a mistake...
23:55:51jhMikeSpixelma: I'd have to look again. Perhaps they made some particular blunders that could be avoided.
23:56:11amiconnjhMikeS: Imho that can't be made good looking while retaining the flexibility wrt font size etc
23:56:13*preglow thinks the trigger screen is ok, but it overrides the button handling really badly
23:56:51preglowit still has bugs too, i see
23:57:12pregloweverything is inited to "db", but when i try to adjust, it switches to "%"
23:57:16amiconnpreglow: Try the trigger screen with a fairly big font, e.g. Helvetica 24 from rasher's site. Then you''ll get an impression how bad it looks at low resolution even with smaller fonts
23:57:16preglowwho maintains that?
23:57:19linuxstbtdtooke: Maybe the error codes were wrong, and the new version is correct...
23:57:20jhMikeSSome things need sorting out with 1) Default font size for hires screen 2) Just two fonts available
23:57:59Lloreanlinuxstb: Aren't error codes unique to the file/function they're in anyway? For example, -2 is used elsewhere for other things too, isn't it (like in the bootloader)?
23:58:14linuxstbLlorean: They're from a common list in buffering.h
23:58:37amiconnjhMikeS: That has nthing to do with default font size
23:58:41pixelmajhMikeS: I don't understand (2)
23:58:50LloreanAh, he just said one's in playback.c and the other is buffering.h, I assumed he meant two different -2s.

Previous day | Next day