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

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

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

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

#rockbox log for 2008-11-11

00:03:13linuxstbamiconn: OK, if that's the case for the as3514, then we should remove that #if in as3514.h. But I guess what I don't like about as3514.h is that it contains a mixture of both private definitions (like the I2C address - which IMO should just be declared somewhere more private), and the public interface for the audio codec part of it. Although the wm*.h files do the same thing...
00:04:09linuxstbIt also doesn't feel right to define ADC_* there.
00:04:45XavierGrI just got a very strange Windows error with my gigabeast
00:05:43XavierGrI inserted the USB cable, Rockbox still navigates (it shows that it charges) and windows pop up an error window saying: "There is no disk in the drive. Please insert a disk into drive /Device/Harddisk4/DR18"
00:06:16 Join Rob2222 [0] (
00:10:22gevaertsXavierGr: can you try going to the debug menu and choose "reconnect usb storage" ?
00:11:01BigBambiXavierGr: I would imagine the more the better
00:11:11XavierGrgevaerts: ah sorry, I rebooted and now it works as it should
00:11:23JdGordonXavierGr: is that the stock battery?
00:11:28BigBambiJdGordon: No
00:11:31XavierGrgevaerts: I will have this in mind if I can reproduce it
00:11:37BigBambiJdGordon: Stock is 700 mAh
00:12:23XavierGrI tried to shove an ipod 4g 1200mAh battery in the beast (like in the gigabeat F), but the S is too tight for that battery
00:13:07BigBambiI just went for the biggest one I could find that said gigabeat S, which was the 1000
00:14:45 Quit MethoS_ (Read error: 110 (Connection timed out))
00:15:59jhMikeSXavierGr: might as well add it to the pool. does anyone have the stock batt but /me? :)
00:16:20*JdGordon does
00:16:29*amiconn too
00:16:36JdGordoni did a bench a while ago, and gonna do a new one today
00:16:41BigBambiI've now found my stock battery, I'll do one with that tomorrow when this one has finished and I have swapped them
00:16:42JdGordonI think my beast is fucked though
00:16:52linuxstbDoes anyone have any idea about the status of the Philips SA9200 port, and why it isn't in the build table?
00:17:05BigBambiOn the other hand, if JdGordon is going to do one with the standard, I won't bother :)
00:17:19 Quit tyfoo2 ("Carpe diem")
00:17:20JdGordonlinuxstb: isnt that one of the ports you started?
00:17:23BigBambiOn the other hand, if JdGordon's beast is fooked, I will do one!
00:17:25jhMikeSJdGordon: Not sure how you'd get it beyond useable.
00:17:44linuxstbJdGordon: No ;) I can at least remember the status of my ports...
00:17:49*BigBambi has run out of hands
00:17:51jhMikeSunless the hardware went
00:17:59XavierGrI changed mine because I had serious issues with the original. It wouldn't last more than 5 hours
00:18:18BigBambiXavierGr: That is normal
00:18:28JdGordonjhMikeS: I dont know... it seems to die after 2 hours or so with the disk not spining up
00:18:29*jhMikeS 's always ran a long time
00:18:33*amiconn doubts that his beast would even last 4 hours with its stock battery
00:18:40XavierGrBigBambi: oh I thought It should last around 10 hours
00:18:43BigBambiMine does about 5 hrs on stock
00:18:51BigBambiXavierGr: Those claims were arse
00:19:04*JdGordon got about 6 last bench though
00:19:06jhMikeStoffee82 must have given me a more beasty than beasty beast
00:19:18BigBambiXavierGr: Have a look at some reviews - the vast majority I have found say very nice, but battery life is crap and nothing like claimed
00:19:34JdGordonand worse withh rockbox
00:19:47amiconnjhMikeS: If we had proper scaling, rewriting some asm blobs for armv6 would be interesting
00:19:47BigBambiThe OF keeps the disk spun up in menus ffs
00:19:53LambdaCalculus37linuxstb: The SA9200 port was started by lowlight, and from the last mention he made, he just needed to get the touchpad working; audio, LCD and buttons are all working.
00:20:15amiconn(e.g. mpegplayer's idct)
00:20:17XavierGrJdGordon: worse, really?
00:20:21jhMikeSamiconn: yeah, that voltage scaling probably helps alot.
00:20:27JdGordonisnt that usuallly how it works?
00:20:46jhMikeSthe ARM DSP instructions are supposed to decode mp3 at 10mHz.
00:21:01 Join synergist [0] (
00:21:12amiconnjhMikeS: That'd be neon, not armv6 dsp
00:21:33amiconnThe arm1136 doesn't have neon
00:22:05jhMikeSThe SIMD != neon?
00:22:23XavierGrshould I charge it with rockbox or OF?
00:22:30XavierGror it shouldn't matter?
00:22:32amiconnNeon is similar to VFP, but for integer. ARMv6 dsp only uses the arm registers
00:23:11jhMikeSamiconn: so neon is really its version of emac
00:23:12amiconnBut it's still quite useful - compare the (effective) clock needed to filter ape -c5000:
00:23:27jhMikeSyou did a conversion for APE?
00:23:29kugelJdGordon: How far did you come with booting into rockbox? and how did you get the offset value? I'm having problems with my fuze
00:23:55JdGordoni got as far as "file not found", that number was taken from the forum thread
00:24:14JdGordonim going to play this arvo maybe... 1 more exam to go
00:24:20jhMikeSXavier: I doubt it's too important at this point. OF seems to kick a few more mV into it though but then being conservate helps battery lifetime.
00:24:25kugelOh, seems I didn't read carefully enough, I thought you found it out
00:24:29amiconnARM7TDM: ~2000MHz (v4), ARM920T (v4) 1720MHz, ARMv6: 476MHz
00:24:42kugelthe number seems to fit for the fuze too, but I can't boot a build because of checksum errors
00:24:56JdGordonthen the number is probably wrong
00:25:02amiconnThanks to smlad(x) & friends...
00:25:07JdGordonwe need a proper method of getting it
00:25:12jhMikeSamiconn: that's why you used the ffs function for division?
00:25:14JdGordoni.e reading it from the of section probably
00:25:32jhMikeSor just for arm4 only?
00:25:33amiconnjhMikeS: Hmm? That's something completely different
00:25:33XavierGrBigBambi: heh the GigaeatRuntime page has a bench of mine with 6 hours. So it seems that my battery wasn't bad after all.
00:25:46BigBambiXavierGr: hehe, good stuff :)
00:25:52 Join SomeIdiot [0] (n=farf@unaffiliated/someidiot)
00:25:56jhMikeSamiconn: ah, ok.
00:26:22amiconnI was talking about the ARMv6 simd stuff. The division optimisation is for ARMv4 only (but uses ideas from the ARMv5+ implementation in libgcc)
00:26:25BigBambiXavierGr: I bought a 1000 mAh as it was an easy way to get an improvement in playing time without any coding :)
00:27:04XavierGrBigBambi: indeed
00:27:18jhMikeSamiconn: I wasn't sure if a mod to division played a role in the numbers. that's quite an improvement in clock efficiency though.
00:27:33SomeIdiotis a version of rockbox for the creative zen far off?
00:27:37JdGordonhmm. I got 5h from r18574
00:27:51LloreanBigBambi: Where'd you get it? eBay or some site?
00:27:59amiconnThe filters don't do any division. The division is needed for the entropy decoder (range decoder), and helps all comptression levels
00:28:04BigBambiLlorean: ebay I think
00:28:23*LambdaCalculus37 heads to eBay to look for some Cameron Sino batteries for his beast
00:28:49LloreanBigBambi: But it was Cameron Sino, and labeled as for Gigabeast?
00:29:00amiconnPerhaps I should put the armv5+ division implementaion into rockbox as well, for 2 reasons? (1) It would make sure that we use the fast implementation, not the one used for OPTIMIZE_SIZE. (2) It would allow saving a single cycle...
00:29:21*amiconn needs to look for a mini g2 battery :/
00:29:36JdGordonyou want mine?
00:29:43amiconnCurrently I can't time ape -c5000 on the mini - it dies around 1/4 through the track...
00:29:45BigBambiLlorean: It was sold as a replacement for the Gigabeat S, and the picture was of a Cameron Sino battery. I wasn't quite sure what would actually turn up, but what has arrrived seems legit, but it could of course be fake
00:29:49*JdGordon has a very dead mini2g
00:30:04LambdaCalculus37JdGordon: What's wrong with the mini?
00:30:05jhMikeSamiconn: ARM5 code is in SVN now?
00:30:16*LambdaCalculus37 just got one fixed up and fully working
00:30:20amiconnjhMikeS: The armv5 code is in libgcc...
00:30:26JdGordonLambdaCalculus37: broke at least one of the bits on the mainboard
00:30:44LloreanBigBambi: Well, the key factor for me is "Cameron Sino seems to make a GigabeatS battery"
00:30:53LloreanSince I like them, now I know to look for one by them. Thanks
00:30:55amiconnI am undecided whether it's worth putting it into rockbpx as well
00:30:56BigBambiLlorean: Yes :)
00:31:21jhMikeSamiconn: something you patched to libgcc?
00:32:59LambdaCalculus37JdGordon: Was it a connector, or a chip?
00:33:10*amiconn is probably too tired to follow
00:33:54 Quit Rob2223 (Connection timed out)
00:34:00JdGordonLambdaCalculus37: i tihnk one of the caps or something.. a chip... I also have broken the touchpad connector thing twice! :'(
00:34:02jhMikeSamiconn: ok, I'm confused. you mean just compiling C-code for ARM6 not any optimized assembly code.
00:34:12JdGordonso i have a working lcd, battery and case
00:34:14XavierGrLlorean: I bought 4 batteries from seller grade_a_battery. For: H300, 2 * 4g ipod, gigabeat s all camero sino
00:34:23JdGordonsold off the microdrive already
00:36:18LambdaCalculus37JdGordon: Can you take some pictures of the board and email them to me? Maybe it may not be so bad to fix.
00:36:23JdGordonalso a leather case if anyone wants
00:36:45amiconnjhMikeS: #if defined(CPU_ARM) && (ARM_ARCH == 4) / #define DIV(a, b) udiv32_arm(a, b) / #else / #define DIV(a, b) (a / b) / #endif
00:37:02JdGordonLambdaCalculus37: yeah sure, pm me a email addy
00:37:16JdGordonif it can be fixed it would be great, my soldering skills are crap
00:37:28jhMikeSamiconn: ah, I see. I thought you might have committed ARM6 SIMD optimization to SVN already as well.
00:37:47JdGordonI think the broken chip is a cap or voltage thingy so should be replaceable
00:38:04 Quit krazykit ("Connection reset by beer")
00:38:05amiconnjhMikeS: Again, that's something completely different (filters, not the range decoder), and *is* in SVN
00:38:20amiconnFor quite some more time than the armv4 division optimisation
00:38:48 Quit Vorador ("Ex-Chat")
00:39:29 Quit faemir (Remote closed the connection)
00:40:06*jhMikeS has had gaps in his attentance of late
00:40:40amiconnBut you hopefully don't have gaps in your source tree ;)
00:40:48 Part SomeIdiot
00:40:54jhMikeSno, no...they grew back in :)
00:41:26pixelmaZagor: did you follow the weird failed builds? They also happen on different servers now and n1s made the remark for the "5" rated one that it failed while trying to build rombox. Even though the others show different errors but it could be an explanation why OndioSP and Player builds fail - those two still have rombox. Could there be any relation to your changes to "make" lately?
00:41:31amiconnSee apps/codecs/demac/libdemac/vector_math16_armv6.h for the simd stuff
00:42:49ZagorI won't say no, but since the builds have worked for days without these problems I'm not ready to blame the Makefiles quite yet
00:43:26amiconnZagor: Perhaps rombox + make -j (on > 1 core)?
00:43:27gevaertsMaybe someone should look at the -j settings on servers that fail and servers that don't
00:43:48Zagoris ondio sp the only rombox target left?
00:44:01pixelmaPlayer and OndioSP
00:44:28jhMikeSthe "...arm7.h" is a bit strange. It looks more like arm4 arch. to me :)
00:44:53Zagoracbuild should print the make command line used
00:45:40amiconnjhMikeS: _arm7 is correct though. It is armv4 of course (note the missing 'v'). Using that code on a v4 arm9 is slower than using C...
00:46:26jhMikeSah, oops...missed the "v" there
00:47:09amiconnI know that arm versioning vs. arm architecture versioning is a mess...
00:50:55Zagoramiconn: you're right, there is a -j issue with rombox
00:52:12ZagorI didn't apply the fix "rambox" got
00:55:59jhMikeSrambox or is that "diskbox"?
00:56:00JdGordonok, beast bench started
00:57:21 Quit jgarvey ("Leaving")
00:57:27 Quit ender` (" Sometimes I lie awake at night, and I ask, "Where have I gone wrong?" Then a voice says to me, "This is going to take more")
00:58:40BigBambijhMikeS: Is there a specific place you would prefer for beast benches? (I'm off to bed now, I'm certain it will be finished before I wake up :P)
00:59:06ZagorjhMikeS: ?
01:00:35pixelmaat least something those disks are still useful for ;)
01:01:36kugelJdGordon: well, i do think that it is the correct number
01:01:44jhMikeSBigBambi: how about wiki attachments? I really don't have any other.
01:01:57BigBambiOK, no probs
01:02:08jhMikeSZagor: hahahaha...I guess that is it.
01:02:38kugelJdGordon: it seems to be dependant on the firmware size (e200v2 and fuze have the same firmware size). Also, the filesize as well as the header of my rockbox binary is read correctly (e,g, it says "model fuze")
01:03:05kugelit was also the only number which enabled me to go beyond disk_init failed
01:03:05XavierGrBigBambi: you can upload your results in the wiki I think
01:03:31*jhMikeS gets a new logo idea :)
01:03:47BigBambiXavierGr: Sure, I was just thinking that jhMikeS might have a different preferred area
01:04:36 Join MethoS- [0] (
01:04:46jhMikeSBigBambi: at least there, they are together. I have no public upload space.
01:04:56 Join saratoga [0] (n=9803c264@gateway/web/cgi-irc/
01:05:01BigBambiRoger, will do
01:05:05saratogaany reason not to commit the ipod accessory patch?
01:06:30XavierGrmine is charging so hopefully in an hour or two I will start a bench too.
01:08:28kugelJdGordon: last but not least, I cannot load from my microsd either, which doesn't have OF offset at all, for the same reason
01:08:49kugelis there a way to check the the rockbox binary on the pc side?
01:09:00kugeli.e. the checksum
01:13:58JdGordonthat probably means the checksum isnt being written correctly
01:14:08JdGordonsaratoga: no, its going in soon...
01:18:09 Quit BigBambi (Read error: 104 (Connection reset by peer))
01:18:23 Quit MethoS (Read error: 113 (No route to host))
01:18:26 Join BigBambi [0] (n=Alex@rockbox/staff/BigBambi)
01:22:26kugelJdGordon: did you have a go at building a full rockbox for your e200v2?
01:22:44kugelshould be very possible without dozens of stubs with the recent commits
01:23:20JdGordonnot yet
01:32:14JdGordonkugel: if you want to investigate... have a look at the builds header and see what the bootloader value gets for the checksum
01:35:20gevaertsJdGordon: if you add a partition table, you have to shift everything by one sector, so the driver will have an offset anyway
01:36:21gevaerts(unless you virtually overwrite the OF sector 0, but IMHO that's even worse
01:36:48 Quit MethoS- (Remote closed the connection)
01:37:32 Quit stripwax (Read error: 104 (Connection reset by peer))
01:37:34JdGordonnot add a real one... just make the rockbox disk driver think there is one
01:37:58gevaertsWell yes, but where will you put it?
01:38:09JdGordonin ram
01:38:38gevaertsI mean, will you mask sector 0?
01:38:49JdGordoni mean... just setup "static struct partinfo part[8]; /* space for 4 partitions on 2 drives */" with the correct numbers
01:40:36JdGordonhmm... dont even need to do that
01:40:37gevaertsSo you'll fill in the partition info bypassing MBR parsing? That could work
01:41:11JdGordonfat_mount takes a start sector, thats enough to do whats needed
01:41:17JdGordonI guess...?
01:41:54gevaertsExcept for USB. There you need to present something that any OS understands, so there will have to be some offsetting there
01:42:16linuxstbWe probably don't want to expose the firmware partition via USB though.
01:42:25JdGordonprobably not
01:42:40gevaertsBut I agree that the low-level driver needs to be as straightforward as possible
01:42:49linuxstbSo just superfloppy (i.e. no partition table)
01:43:16***Saving seen data "./dancer.seen"
01:43:24*gevaerts has been thinking of a storage filter layer that could do things like working with offsets or faking MBRs
01:44:57gevaertsOf course you need to be careful when going down that road, as it also leads to loopback devices and RAID
01:44:58JdGordonmaybe the easiest way is to add a #define SUPERFLOPPY_OFFSET in the target config and use that if no partition table is found?
01:45:02JdGordondefaulting to 0 of course?
01:45:44*JdGordon wishes we wern't dealing with magic numbers here thoughg
01:46:23gevaertsI'd vote for something a bit more general, like a target-specific function
01:46:57 Quit tvelocity (Connection timed out)
01:49:55 Quit lasser (Read error: 110 (Connection timed out))
01:49:59 Part homielowe
01:53:15 Quit culture (Read error: 110 (Connection timed out))
02:03:25 Join fdinel [0] (
02:09:12 Join nuonguy [0] (
02:14:03 Quit Zagor ("Client exiting")
02:15:05 Quit saratoga ("CGI:IRC")
02:16:01UnhelpfuljhMikeS: i have a battery bench run for the orig battery... do we have a tool to process that to the curves?
02:17:20jhMikeSUnhelpful: no tool in particular. after several the results can be combined in the right way.
02:19:22 Join homielowe [0] (
02:21:13kugelJdGordon: well, the bootloader printf's the checksum in the header and calculates another and compares them
02:21:28kugelJdGordon: so I think it does automatically that what you suggested?
02:22:06Unhelpfultool == a calculator, ok. there's a different table for charge vs discharge, isn't there? do i just run battery_bench while charging for that, or is there something else to do?
02:22:14kugelparticulary load_firmware does that
02:22:22JdGordonkugel: and you said ignoring the checksum caused a data abort?
02:23:35kugelat least when I tried the last time, I compiled some newer builds in the mean time
02:23:55kugelnot data abort, but undefined instruction (look at my forum post)
02:25:06JdGordonok, rebuild a main binary but replace main() with a while loop which just flashes the backlight or something simple (wth as little init as possible) and see if that loads without the checksum
02:25:16JdGordoncould be reading the wrong sector...
02:26:46 Join Darksair [0] (n=user@
02:27:15kugelwill do later
02:33:12 Quit PaulJam (".")
02:33:13 Join hsitz [0] (n=4c68889d@gateway/web/cgi-irc/
02:34:27hsitzAny help out there for problem with Sansa e260?
02:35:14JdGordonnot if you dont ask a question
02:35:57 Part pixelma
02:36:11 Join pixelma2 [0] (n=marianne@rockbox/staff/pixelma)
02:36:24hsitzok, thanks. Had initial install working fine. Just installed 8gb sdhc card and in building database I get error
02:36:45hsitzError is 'undefined instruction'. Screen is stuck, machine is stuck. I can't even turn it off.
02:37:14linuxstbkugel: Have you tested the RAM, as funman suggested?
02:37:30LambdaCalculus37hsitz: Have you tried holding the Power button for up to 30 seconds?
02:37:56hsitzlambda: I've only tried up to 15 or 20. I'll try longer. . .
02:39:22hsitzlambda: That did it. I think I can get things going from here. Thanks!
02:41:11 Quit tchan ("WeeChat 0.2.7-dev")
02:41:22 Join tchan [0] (
02:45:20 Quit hsitz ("CGI:IRC (EOF)")
02:48:10 Nick fxb is now known as fxb__ (
02:51:57kugellinuxstb: no, I haven't done anything the past few hours
02:52:17kugellinuxstb: the of would lie though if it's not 8MB
02:53:10linuxstbThat's not the only issue though - even if it does have 8MB RAM, the init may not be correct. So it's useful to test.
02:53:42JdGordonkugel: and I wouldnt put it past the of to lie about its ram size
02:54:02JdGordonespecially on a flash target, ram isnt *that* important
02:54:25kugelwhere is the ram initialized?
02:55:06kugelI think I modified accordingly (if that's the correct place)
02:56:18kugelI think i see something I forgot (in system-as3525.c)
02:56:27kugelline 170
02:59:08kugelone of those ancient assumtions by funman
02:59:10 Quit Makuseru (Read error: 104 (Connection reset by peer))
03:02:45 Join BHSPitMonkey [0] (n=stephen@unaffiliated/bhspitmonkey)
03:03:02linuxstbkugel: What did you change in
03:05:13kugelmemorysize with 8, since I got linker errors. but that's probably related then
03:06:03linuxstbYou should change the memory definition in tools/configure - everything else should follow from there (apart from system-as3525.c)
03:08:25kugeli did that but that didn't seem to do it, so i explicitly changed (the linker error disappeared)
03:10:36*linuxstb tries
03:11:00 Quit Darksair ("Use the Force, Luke!")
03:12:05*linuxstb sees there are lots of other things to fix in the main fuze build...
03:14:14 Quit Kopfgeldjaeger ("Serverwechsel")
03:14:59 Quit Zarggg ()
03:16:08linuxstbkugel: Have you tried the "lcd_enable" function?
03:19:16 Quit Chronon ("ChatZilla 0.9.83 [Firefox 3.0.3/2008092417]")
03:24:21kugelno, not yet
03:28:14 Quit cg (Read error: 110 (Connection timed out))
03:28:44kugellinuxstb, JdGordon: that single line messed my whole day. the checksum is correct now
03:29:37kugelloading from sd doesn't work quite well though
03:29:43kugelmicrosd that is
03:32:49 Join AndyI [0] (i=AndyI@
03:34:08*kugel sees the bootlogo of the main binary for a split second
03:35:19 Quit LambdaCalculus37 ("Ka-chunka")
03:43:05 Quit nuonguy ("This computer has gone to sleep")
03:43:17***Saving seen data "./dancer.seen"
03:44:58 Quit AndyIL (Read error: 110 (Connection timed out))
03:48:30 Join Zarggg [0] (
03:51:07JdGordonarg, came back to my beast and it is off... except powering it back on the bootloader says the batt is at 3.672 and the main build rekons there is plenty of batt left
03:51:12 Quit Zarggg (Client Quit)
03:52:21JdGordondid ~3h according to the batt bench .txt
03:53:47 Quit reacocard (Read error: 110 (Connection timed out))
03:58:29kugelnot much :(
03:58:57 Join reacocard [0] (n=reacocar@
03:59:39soapIt was doing much better when you got it?
03:59:47soapI feel guilty if not.
04:00:05kugelJdGordon: noticed? it was really that one line in system-as3525.c. I'm so angry...
04:00:40JdGordonkugel: i have an exam in 30min.. i have more important things to be ignoring....
04:01:04 Quit fdinel ("Miranda IM! Smaller, Faster, Easier.")
04:01:06JdGordonsoap: yeah, was "fine" when i got it... i dunno what the problem is
04:01:30JdGordonsoap: also, you menitoned redoing the ipod accessory wiki page? should we wait for that before commiting?
04:02:55kugelJdGordon: heh, ok, good luck then
04:03:06soapI don't see the point. Worst case situation I see is duplicate feedback from users.
04:03:19JdGordonok, cool
04:03:29kugellinuxstb: yep, it's lcd_enable
04:04:11kugelgood night all
04:04:16 Quit kugel ("ChatZilla 0.9.83 [Firefox 3.0.3/2008101315]")
04:06:00 Join Zarggg [0] (
04:21:31 Join Bagder_ [0] (
04:27:49 Join _Bagder_ [0] (
04:31:39 Join miepchen^schlaf_ [0] (
04:37:42 Quit Bagder (Read error: 110 (Connection timed out))
04:38:33 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
04:42:33 Quit Bagder_ (Read error: 110 (Connection timed out))
04:42:34soapHow is one to contact wiki contributers if there is a question on the content they offer upon the wiki? Very few people outside the "regulars" leave an email address in public view, fill out IrcNicks, or mention their forum/IRC nick on their personal wiki page. Reading through the IRC logs on the last couple of them, they didn't even mention their "alter ego" nick in here. Must I go through the wiki admins (Sweden?) to find out contact info?
04:47:29soap2: Can we declare it policy that a copyright notice and license notice MUST be inside all contributed themes/WPSs? This does not appear to be current policy (least I can not find the requirement posted anywhere), but if we are going to insist on code-like licensing rules for themes/WPSs should we not also require explicit notices in the code of the WPS/theme?
04:49:18LloreanYes, if we're going to insist on licensing, the license used should be included in the file somewhere too. Preferably (in my opinion) as comments within the .wps, but I don't know if that's strictly necessary. It'd at least avoid clutter of additional files
04:49:21 Join blkhawk- [0] (
04:49:33soapThis comes back to point / question #3. It's once again getting pretty bloody hard to figure out if people are carrying out the requirements of the CC-BY-SA license. This most recent set of uploads my Michael Ross (who is that? how can I contact him?) clearly is borrowing graphical elements from other themes. No mention of where those elements came from, much less attribution, is given anywhere.
04:52:13LloreanI don't know if there's a good solution to this so long as we're using the wiki.
04:52:33 Quit Bensawsome ("The awsome is gone :(")
04:55:13soapShould this be discussed on the dev mailing list? Perhaps an agreed policy? Something broad enough in scope to cover future issues? I feel as if we're constantly playing catch-up on WPS poster behavior. That isn't the fault of the posters so much as it is my failure to come up with comprehensive and proactive guidelines.
04:55:56LloreanWell, I don't know how we can consistently enforce any sort of guidelines without a site designed to keep records.
04:56:19LloreanMake the email address available to a larger selection of us, when a theme is reported as problematic.
04:57:06soapIF we had a formal policy regarding the posting of creative works to the wiki (even if it were a policy I don't like ;) ) I would feel emboldened to simply waste-bin such themes. But as it stands now I'm not sure anything outside "spirit" was broken.
04:57:33 Join jhulst_ [0] (n=jhulst@unaffiliated/jhulst)
04:57:56LloreanWell, CC-BY-SA requires attribution
04:58:02LloreanIf he's not attributing, it's license violation.
04:58:08 Quit jhulst (Read error: 113 (No route to host))
04:59:56soapwe allow banned forum users to post to the wiki?
05:01:43LloreanWell, they aren't connected.
05:05:01 Quit smacnay ("Ex-Chat")
05:05:31 Quit blkhawk (Read error: 110 (Connection timed out))
05:06:20 Nick blkhawk- is now known as blkhawk (
05:09:41 Join webguest23 [0] (n=4a4d2b16@gateway/web/cgi-irc/
05:10:36webguest23is anyone here?
05:11:40 Quit webguest23 (Client Quit)
05:33:26 Join Bensawsome [0] (n=Bensawso@unaffiliated/bensawsome)
05:38:45 Quit esthar (Read error: 104 (Connection reset by peer))
05:43:20***Saving seen data "./dancer.seen"
05:43:21 Join aarcane [0] (
05:50:14 Quit Horscht ("I got raided by the FBI and all i got is this lousy quit message")
05:59:27 Join esthar [0] (
06:02:25 Quit XavierGr ()
06:02:41 Join fyrestorm [0] (
06:03:44 Join Makuseru [0] (
06:07:43JdGordon3 hr 30min beast nech... that really really sucks :'(
06:08:47 Quit fyre^OS (Read error: 60 (Operation timed out))
06:15:03 Quit Bensawsome ("The awsome is gone :(")
06:15:16 Join Photoguy [0] (n=Adams@
06:15:31PhotoguyHi all.
06:15:51 Join gartral [0] (n=Gareth@
06:16:01gartralgood morning everyone
06:16:03PhotoguyCan someone see if they can answer this?
06:17:02gartralis it the connector, or the cable?
06:17:14PhotoguyThe connector..
06:17:25PhotoguyWell, kind of both.
06:17:33gartralif its the connector, your kinda stuck... id its your cable, buy a new one >.>
06:17:48gartralbeyond that, this is hardware, not software
06:17:58PhotoguySo, you don't think there is any way to boot form msd?
06:18:08advcomp2019Photoguy, i see you posted that on ABi too.. i think there is a way but i think you need to build your own build but i could be wrong
06:18:21gartralim not a dev, but i remember something about them trying it
06:18:40PhotoguyMaybe I'll wait, and see if anyone asnwers my post.
06:19:13gartralbeyond that, your only option is to swap mem units with another sansa e2xx and install with that, then swap back
06:19:16advcomp2019Photoguy, o wait.. i misread it.. rockbox does have a file moving option
06:19:27PhotoguyThe stupid thing is, I was copying somthing in rockbox, and mistakely deletd the folder (stupid!!)
06:19:51PhotoguyNo the problem is, there is not rockbox currently on the player.
06:20:19PhotoguyThen, I STUPIDLY shut down the player..
06:20:26gartraltake your mem unit out of that box, put it in one that works, intall, and swap back...
06:20:54PhotoguyI could have just ccopied a new build onto it while it was still on. :P
06:21:10Photoguy*turned on
06:21:29advcomp2019hmm.. that is another issue then
06:21:49PhotoguyHow could I be so dumb?
06:22:20PhotoguyI guess I need to ask an exoert about booting from MSD
06:22:40PhotoguyNot that you guys are dumd..
06:22:54PhotoguyBoy, I can't spell.
06:23:20gartralno, youve proven you can spell, what your haveing trouble with is typeing out what your spelling
06:24:00gartralgetting back OT i have found a DAP that has a similar name to our project...
06:24:08gartralcalled roq box
06:24:25Photoguy"roq box" ?
06:24:57PhotoguyWow, that's expsensive!'
06:25:35gartralyep >.> ive noticed... but the point is, its not in our similar names section of the wiki, and ide like to put it there if i may
06:26:48PhotoguyWell, thanks for the inspiration..
06:26:56PhotoguyTo buy a new one :P
06:27:31 Quit Photoguy ("Leaving")
06:29:57 Join massiveH [0] (
06:33:25 Quit BigBambi (Read error: 113 (No route to host))
06:36:43gartralso who else thinks this is a good candidate for the similar names project?
06:37:12gartralthe roq box, i mean
06:38:45scorchegartral: lots of items are named "rockbox" or "rock box"
06:42:59LloreanIf you think it belongs on the page, add it...
06:43:10gartralyea, and isnt the point of that wiki page to help people keep track of what we are, and arnt talking about?
06:50:35 Join J-23 [0] (
06:52:20JdGordonhmm... my monoeared e200 has already been "fixed" once
06:52:41JdGordonwith a worse soldering job than i could do :p
06:57:25 Quit jhMikeS (Nick collision from services.)
06:57:31 Join jhMikeS [50] (n=jethead7@rockbox/developer/jhMikeS)
07:06:44 Nick jhulst_ is now known as jhulst (n=jhulst@unaffiliated/jhulst)
07:11:38 Join massive_H [0] (
07:12:13 Quit massiveH (Nick collision from services.)
07:12:15 Nick massive_H is now known as massiveH (
07:14:36gartrali know this is ot, but why do some users put an _ at the end of there names?
07:15:00 Quit knnylftn (
07:15:00 Quit Seed (
07:15:00 Quit trisiak (
07:15:00 Quit Unhelpful (
07:16:32NJointrisiak [0] (
07:17:03scorchegartral: if you know it is offtopic, then why say it?
07:17:18NJoinSeed [0] (
07:19:38 Quit reacocard (".")
07:20:22 Join reacocard [0] (n=reacocar@WL-112.CINE.HMC.Edu)
07:20:48NJoinknnylftn [0] (
07:31:07 Join MU{lappy} [0] (n=Militant@
07:43:24***Saving seen data "./dancer.seen"
07:47:32 Nick fxb__ is now known as fxb (
07:50:31 Quit reacocard (".")
07:52:16 Join reacocard [0] (n=reacocar@WL-112.CINE.HMC.Edu)
07:52:34 Quit amiconn (Nick collision from services.)
07:52:40 Join amiconn [50] (n=jens@rockbox/developer/amiconn)
07:53:01 Nick fxb is now known as fxb__ (
07:53:03 Quit ch4os (Read error: 60 (Operation timed out))
07:56:56 Join ch4os [0] (n=ch4os@gentoo/user/ch4os)
08:05:06 Part gartral
08:09:13 Part LinusN
08:11:51 Join LinusN [0] (n=linus@rockbox/developer/LinusN)
08:15:31 Join Darksair [0] (n=user@
08:15:40 Join bertrik [0] (
08:38:35 Join Zagor [0] (n=bjorn@rockbox/developer/Zagor)
08:41:38 Quit pixelma2 ("-")
08:42:04 Join pixelma [50] (i=pixelma@rockbox/staff/pixelma)
08:42:07 Part pixelma ("Cycling Channel")
08:42:11 Join pixelma [50] (i=pixelma@rockbox/staff/pixelma)
08:42:15 Part pixelma ("Cycling Channel")
08:42:27 Join pixelma [50] (i=pixelma@rockbox/staff/pixelma)
08:42:31 Quit pixelma (Remote closed the connection)
08:43:37 Join Bagderr [0] (n=daniel@gateway/web/cgi-irc/
08:43:55 Join pixelma [50] (i=pixelma@rockbox/staff/pixelma)
08:45:02 Nick Bagderr is now known as B4gder (n=daniel@gateway/web/cgi-irc/
08:46:34 Quit bertrik (Remote closed the connection)
08:48:28 Quit GodEater (" ajax IRC Client")
08:49:33 Quit miepchen^schlaf_ ()
08:50:25 Join ender` [0] (
08:50:38 Join GodEater [0] (i=c2cbc962@gateway/web/ajax/
08:56:14 Quit massiveH ("Leaving")
08:57:46 Join spiorf [0] (
09:00:14 Join Rob2223 [0] (
09:01:39 Join nima [0] (n=nima@unaffiliated/nima)
09:04:20nimaHello. My iPod has entered disk mode, and I can't find any way to get it out. Is there any alternative to waiting for the batteries to die?
09:04:38nimaIt's a 1GB iPod nano
09:06:10GodEaterreset it
09:06:17GodEaterHold Menu+Select together for about 10 seconds
09:06:22 Join tyfoo [0] (
09:06:46nimaThanks, GodEater. That worked
09:07:12GodEaterno problem
09:07:13 Quit amiconn (Nick collision from services.)
09:07:15 Quit tyfoo (Client Quit)
09:07:19 Join amiconn [50] (n=jens@rockbox/developer/amiconn)
09:08:38 Join kugel [0] (n=chatzill@unaffiliated/kugel)
09:10:07 Quit BHSPitMonkey ("Ex-Chat")
09:14:39 Quit MrDuck (Read error: 113 (No route to host))
09:15:07 Quit kugel ("ChatZilla 0.9.83 [Firefox 3.0.3/2008101315]")
09:16:20 Quit amiconn (Nick collision from services.)
09:16:26 Join amiconn [50] (n=jens@rockbox/developer/amiconn)
09:16:55 Join pixelma_ [50] (i=pixelma@rockbox/staff/pixelma)
09:18:35 Quit Rob2222 (Read error: 110 (Connection timed out))
09:20:44 Join skipper_ [0] (n=skipper@
09:22:59 Quit jhulst (Read error: 113 (No route to host))
09:23:27 Join CaptainSquid [0] (
09:24:33 Quit pixelma (Nick collision from services.)
09:24:34 Nick pixelma_ is now known as pixelma (i=pixelma@rockbox/staff/pixelma)
09:26:57 Join BHSPitMonkey [0] (n=stephen@unaffiliated/bhspitmonkey)
09:29:59 Quit JdGordon ("Konversation terminated!")
09:31:53 Join JdGordon [0] (
09:33:44 Join l403 [0] (n=l@
09:39:54 Join einhirn [0] (
09:43:26***Saving seen data "./dancer.seen"
09:52:00 Join kugel [0] (n=chatzill@unaffiliated/kugel)
09:58:33 Quit BHSPitMonkey (Read error: 104 (Connection reset by peer))
10:07:08 Quit MU{lappy} (Read error: 113 (No route to host))
10:14:21 Quit Seed (Read error: 110 (Connection timed out))
10:22:37 Join JonathanRoitgrun [0] (n=52e09642@gateway/web/cgi-irc/
10:23:01JonathanRoitgrunhi. i've just signed up for the wiki, i'd like write permission to upload a theme plz?
10:23:18 Quit kugel (Remote closed the connection)
10:23:55Zagorouch. alac.c:(.text+0x588): relocation truncated to fit: R_ARM_PC24 against `.icode'
10:24:07 Join MU{lappy} [0] (
10:24:16JonathanRoitgruna .wps that is, sry
10:26:20 Join kubzior [0] (n=kubz@unaffiliated/kubz)
10:27:21Zagorit's a little messy when many codecs want different -O flags
10:27:34 Quit l403 (Read error: 104 (Connection reset by peer))
10:28:04 Quit JdGordon (Read error: 104 (Connection reset by peer))
10:28:47 Join culture [0] (
10:29:36 Quit JonathanRoitgrun ("CGI:IRC (Ping timeout)")
10:30:27 Join JdGordon [0] (
10:31:47 Join JonathanRoitgrun [0] (n=52e09642@gateway/web/cgi-irc/
10:32:45 Join jonaroit [0] (i=52e09642@gateway/web/ajax/
10:32:46 Quit JonathanRoitgrun (Client Quit)
10:32:49ZagorIt would be good if some devs could take a look at my single make effort and voice opinions, so I don't spend too much effort going in the wrong direction. Latest patch is in FS #9534.
10:33:20jonaroithello all. can I get write permission to the wiki here? (i'd like to upload a WPS mod)
10:34:48 Join l403 [0] (n=l@
10:38:35 Join lasser [0] (
10:39:20 Join Mastrojani [0] (n=4dfda952@gateway/web/cgi-irc/
10:40:38jonaroitum, anybody?
10:40:50Zagorjonaroit: sure, what is your wiki name?
10:41:44*pixelma guesses "JonathanRoitgrund"
10:41:56 Join YOJI [0] (n=4dfda952@gateway/web/cgi-irc/
10:42:38jonaroityes, JonathanRoitgrund
10:42:47jonaroitsry I was afk for a while
10:44:19 Quit Mastrojani ("CGI:IRC (Ping timeout)")
10:44:57 Quit YOJI (Client Quit)
10:46:08Zagorjonaroit: added
10:46:45jonaroitthanks, great. just making sure, it is perfectly fine for me to submit a mod of an existing theme? (cabbiev2) that is, if I mention that it's based on it?
10:46:46 Join bodymind [0] (n=bodymind@
10:47:20jonaroitgreat, thanks
10:47:26 Quit MU{lappy} (
10:48:44kubziorso is there port for sony walkman nw-e010 series? :f
10:49:12NJoinMU{lappy} [0] (
10:49:19Zagorkubzior: no, you'll be first! :-)
10:49:40kubziorZagor: fine, but I'm not good programmer ;p
10:51:22 Join Thundercloud [0] (
10:52:26 Join DerDome [0] (n=DerDome@
10:52:59 Quit DerDome (Client Quit)
10:53:41 Quit Nibbler (Read error: 113 (No route to host))
10:58:34 Quit Thundercloud (Remote closed the connection)
11:01:08 Part kubzior
11:01:09bodymindanyone on sansav2?
11:01:53 Join Photoguy [0] (n=Adams@
11:03:32 Quit Darksair ("(define zero (lambda (f) (lambda (x) x)))")
11:04:53 Quit jonaroit (" ajax IRC Client")
11:06:08PhotoguyAnyone here that likes to help?
11:07:37ZagorPhotoguy: simply ask the question. if someone can answer, someone will.
11:08:05PhotoguyHere's the problem...
11:08:39PhotoguyDo you think it's possible?
11:09:02ZagorPhotoguy: no. your best bet is trying to clean/fix the connector.
11:09:24PhotoguyNot possible at all, to boot from MSD?
11:09:51B4gderbut you can copy files from the microsd
11:10:01B4gder(as I presume MSD means)
11:10:07Zagorcan the OF do that?
11:10:16PhotoguyRight, but rockbox is not installed..
11:10:23PhotoguyNot that I know of.
11:10:26 Join kugel [0] (n=chatzill@unaffiliated/kugel)
11:10:34B4gderand you can't use usb so you're toast
11:11:33PhotoguyI'm in Israel, so I can't get another one easily/
11:11:56PhotoguyThis looks like a v1, what do you think?
11:12:09JdGordonaaaaaaaerrrrrrr.. farking sansa DB refresh
11:13:06ZagorPhotoguy: how is the connector broken? visible destroyed, or simply not working?
11:13:07JdGordonPhotoguy: shalom... yeah probably a good bet thats a v1
11:13:39Photoguy@ Zagor, a couple of pins are bent..
11:13:59PhotoguyBut it's still works sometimes.
11:14:16PhotoguyThe other problem is, my dock is broken :P
11:14:37PhotoguyI know, it's a hopeless case.
11:15:01Photoguy@jdGordon, good bet, huh?
11:15:11PhotoguyI don't want it if it's v22
11:15:17JdGordonwoo!!!! ata error -1 booting a main build on a e200v2
11:15:37JdGordonPhotoguy: the v2's will have rockbox pretty soon (probably) but if its refurb its probaly a v1
11:15:59PhotoguyOk, how soon?
11:16:15PhotoguyI suppose I can wait..
11:16:56JdGordoncould be months.. could be days....
11:17:05JdGordonwe are activly working on it
11:17:39PhotoguyThanks for what you do!
11:19:36 Quit culture (Read error: 110 (Connection timed out))
11:19:48bodymindJdGordon: how is it going? have you booted?
11:19:59JdGordongot as far as ATA error
11:20:08JdGordonwhich means not very far :p
11:20:40bodymindi get:
11:20:42bodymindLoading firmware:
11:20:45bodymindfile not found.. :s
11:21:25JdGordonyes, because you dont have rockbox.sansa....
11:21:28bodymindi suppose it's because i'm making a Bootloader and not a normal build
11:21:36 Quit linuxstb (Read error: 60 (Operation timed out))
11:21:36bodymindok.. how do i build it?
11:21:47bodymindi hit make
11:21:51JdGordonyou cant easily yet
11:22:19bodymindand i get lots of undeclared: CPUFREQ_DEFAULT, CPUFREQ_NORMAL..etc..
11:22:37bodymindi have some C knowledge but not so low level... i would like to help... 8)
11:23:38JdGordonthere really is not much you can do that isnt being done already...
11:23:38 Join MrDuck [0] (
11:23:41JdGordonnot yet anyway
11:25:16kugelI came to the rockbox logo
11:25:18bodymindok :)
11:25:19JdGordonunless you want to build, test, comment out a line, build test, comment another line... rinse repeat
11:25:21kugelbut not to the main menu
11:26:03JdGordonkugel: you definatyl got the main buil logo?
11:27:00JdGordonkugel: can I see your diff?
11:27:47 Join yoji [0] (n=4dfda952@gateway/web/cgi-irc/
11:28:14JdGordon"if( !send_cmd(drive, SD_APP_CMD, 0, MMC_RESP|MMC_ARG, &response) || !(response & (1<<5)) )" is failing for me
11:30:01PhotoguyGo, you can do it!!
11:30:15JdGordonarg, wtf? the main build reflips and mirrors the display!?
11:32:47 Quit yoji ("CGI:IRC (Ping timeout)")
11:33:26JdGordonok, I have the logo showing now
11:33:37JdGordonkugel: thats as far as you got? just the main build logo?
11:35:19 Join yoji [0] (n=4dfda952@gateway/web/cgi-irc/
11:35:39 Quit yoji (Client Quit)
11:35:47Photoguy@JdGordon, should I order a v1? or wait?
11:36:05kugelJdGordon: yes
11:36:15JdGordonget whatever there is
11:36:49JdGordongetting better.... white screen
11:37:09PhotoguyWait, you think I should buy a v2?
11:37:27JdGordonarg, bah, going backwards
11:38:05PhotoguyI wouldn't see it for a month and a half anyway, because i'm going back to the u.s. then.
11:38:44kugelJdGordon: much fuze simulator code though (which I still wait for to be committed9
11:40:14 Join BigBambi [0] (n=Alex@rockbox/staff/BigBambi)
11:40:59PhotoguyHi Bambi.
11:41:07 Join cg [0] (
11:41:33PhotoguyKugel and JdGordon are discussing the v2 firware, they're getting somewhere!
11:41:47 Quit spiorf (Remote closed the connection)
11:43:28***Saving seen data "./dancer.seen"
11:44:05 Quit Photoguy ("Leaving")
11:44:36kugelJdGordon: I also have some weirdness at the build logo. It's shifted left a bit
11:45:13JdGordonnoticably? how many pixels?
11:45:43 Join YOJI [0] (n=4dfda952@gateway/web/cgi-irc/
11:48:16 Join what [0] (n=79b45a2d@gateway/web/cgi-irc/
11:48:23*JdGordon loves that commenting out EMPTY functions breaks things :'(
11:48:49kugelJdGordon: 20pixel maybe
11:48:58markunwhat: hi
11:49:36whati'm korean
11:50:36markunwelcome. Do you have a question about rockbox?
11:51:16whati want to have a samsung t9 rockbox
11:52:35whatis it developing?
11:53:00markunI don't think anyone is actively working on it. I only see this:
11:54:00markunah, I see it's the same CPU as the Meizu M series
11:54:29whatI know C languages.
11:54:45markunwhat: do you know if it's possible to enter the DFU rescue mode
11:55:17whatmarkun: no
11:55:50markunwhat: here is a checklist to help you if you want to work on the port:
11:55:52whatmarkun: sorry yes.i know
11:56:05whathmm yes thankyou
11:56:18kugelJdGordon: would you like to look at committing ?
11:57:18JdGordonyeah ok, ill have a look
11:58:41*JdGordon has a patch for the as35.. bootloader but isnt sure if he should commit it
11:58:47JdGordonthat delay is driving me mad
11:59:16whati can't english well haha;;
11:59:41markunwhat: do you know if there is some kind of rescue mode if we break the firmware?
12:00:36whatno hmm]
12:01:24whatmarkun: no hmm]
12:02:52bodymindwhere do i define CPUFREQ_DEFAULT? do you have any patch already?
12:03:33markunwhat: I have no time to look into it right now, maybe tomorrow
12:04:09whatmarkin: hmm what time is it ?
12:04:16whatmarkun: what time is it?
12:04:17JdGordonkugel: you have a few changes in that diff which shouldnt effect the sim... and you really should fix every lang for the wheel (seen as its a no brainer)
12:05:43whatmarkun: what time is it now? hmm
12:06:07J-23what: 12:05 in my timezone
12:06:29whatJ-23: hmm my timezone is 8:06PM
12:06:34kugelJdGordon: the other to changes affect the memory, which are essential anyway
12:06:52JdGordonyes, but shold be seperate
12:06:58BigBambibeast battery bench 7:02 with 1000 mAh
12:09:34JdGordonbugger you! :p
12:10:57JdGordonanyone totally against moving all the config-<target>.h files into export/configs/<target>.h ?
12:12:48 Join moos [0] (
12:14:45markunwhat: it's 12:14 here
12:15:02markunand try to make longer sentences..
12:15:17whatmarkun: yes sorry
12:15:37bodymindhere is 11:14 8) portugal \m/
12:17:08BigBambiwhat: Please don't just type rubbish into the channel. If you have a question, just ask it
12:17:45whatBigBamBi: OK!
12:18:49markunwhat: if you just want to chat you should go to #rockbox-community
12:19:54kugelJdGordon: i updated
12:21:36JdGordonyou havnt fixed plugins at all yet?
12:21:49kugelplugins are not compiled
12:21:49JdGordonplugins="" works?
12:21:53kugelcan be done later imho
12:22:21 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
12:22:22pixelmawhat's the Fuze display's resolution?
12:22:27kugelis it supposed to be "no"?
12:22:42pixelmaok, boring ;)
12:22:42JdGordonI would assume "no" is correct?
12:23:12kugelpixelma: heh, looked for adapting some default themes? too bad h300 ones can be used :p
12:24:12pixelmawell, not only themes but also other graphics. Already got a plan for cabbiev2 on the Clip :)
12:25:01kugelJdGordon: judging from other configure targets "" works as well as "no" for not compiling plugins, but yea, I'd assume "no" is slightly more correct
12:25:16JdGordonmeh, it works so its fine
12:26:26J-23what should be CPUFREQ_* values for e200v2 and where are they set?
12:26:52 Quit MrDuck (Read error: 110 (Connection timed out))
12:27:43kugelJdGordon: Do you have an idea why the BOOTDIR macro isn't used in load_firmware?
12:27:55JdGordonin which?
12:28:22kugelso if I change the bootdir in my target config, the bootloaeder will still look in /.rockbox
12:28:45kugelI stumbled upon that yesterday while trying to boot from the microsd
12:28:57JdGordonyeah, looks like its unused
12:29:21JdGordonbootloader/common.c load_firmware() is where you want to change it
12:30:06kugeluhm yea
12:30:08pixelmais there/shouldn't there be a keymap file for the Fuze? I didn't see one...
12:30:23JdGordonarg, doh, forgot to svn add it
12:31:08 Join PaulJam [0] (
12:32:05whatmarkun: Can you make a samsung t9 rockbox?
12:33:26markunno, because I don't have such a player and no time, but if anyone wants to work on the port we'll help him as much as we can
12:33:41JdGordonIs there any reason why the bootloaders dont follow BOOTDIR? I want to change it but dunno if its an oversight or there was some discussion on this?
12:34:27kugelJdGordon: load_firmware definitely found rockbox.sansa on my microsd when I changed the path
12:34:48JdGordonyeah of course.. but i wonder if there is a reason it wasnt changed
12:35:05whatmarkun: i want make a rockbox but rockbox homepage is all of the english , i can't english well..
12:35:41bodymindwhat: use google translator ^^
12:35:46kugelI think it just that nobody noticed it yet, since it just always has been /.rockbox
12:36:16markunwhat: that will be a bit of a problem yes. Maybe you can find some other korean devs who want to work on it and who speak english?
12:36:50whatbodymind: but it is very strange
12:37:15JdGordonkugel: no, it was origionally in / only, it was moved into .rockbox so there might be a reason it wasnt changed to BOOTDIR
12:39:03J-23hmm, what's default timer frequency on e200v2's?
12:39:42pixelmaJdGordon: not sure if it has anything to do with it but the Archos still expect the firmware file in the root (because the original firmware expects it there and this way you can just "run" Rockbox)
12:39:56B4gderJ-23: we run the timer tick at 100Hz on all targets
12:40:01whatmarkun: it is samsungT9 port
12:40:14B4gderwhat: it's the beginning of one, all the info we have on it
12:41:37whatB4gder: OK Thankyou!
12:45:45kugelpixelma: load_firmware searches in /.rockbox, and if it fails in /
12:48:08 Join Llorean1 [0] (
12:48:32 Quit scorche (Read error: 54 (Connection reset by peer))
12:48:59kugelJdGordon: thanks for committing
12:50:25 Join scorche [0] (i=Blah@rockbox/administrator/scorche)
12:51:29kugelB4gder: can you comment on load_firmware not using the BOOTDIR symbol?
12:51:59B4gderI think it should
12:52:09B4gderat least the bootloader should
12:52:20B4gderis load_firmware used elsewhere?
12:52:30 Quit Llorean (Read error: 60 (Operation timed out))
12:54:57kugelB4gder: a quick grep says no (not even in rolo.c)
12:55:26kugelbut there's a load_firmware_file in the iriver flash plugin
12:55:44 Join ap0 [0] (
12:56:55 Join kachna [0] (
12:59:16kugelJdGordon: ok then, I'll put a new patch up enabling rockbox.sansa to be executed
13:01:02 Quit kugel ("ChatZilla 0.9.83 [Firefox 3.0.3/2008101315]")
13:02:41 Quit BigBambi (Read error: 104 (Connection reset by peer))
13:02:57 Join BigBambi [0] (n=Alex@rockbox/staff/BigBambi)
13:06:12JdGordonBOOTDIR is used by misc.c but not the bootloader
13:07:35J-23hmm, why oldmenuapi.c requires pcm_record.h?
13:08:13 Nick fxb__ is now known as fxb (
13:09:22JdGordonof course!
13:10:59J-23Is it possible to record to memory card instead of internal memory?
13:11:12pixelmathose weird failing builds still happen
13:12:00J-23Where can I change it? I don't see such option in Recording Settings menu
13:12:30pixelmaJ-23: look for "Set as recording directory"
13:13:02pixelmain the context menu on a directory in the file browser
13:13:46pixelmaor in the manual
13:14:35 Join Darksair [0] (n=user@
13:16:38 Join tyfoo [0] (
13:16:47 Join zorgblaubaer [0] (
13:17:30zorgblaubaerhey can anyone here help me out with testing the meizu m6 stuff?
13:17:54 Quit BigBambi (Read error: 104 (Connection reset by peer))
13:18:09 Join BigBambi [0] (n=Alex@rockbox/staff/BigBambi)
13:19:21what0x30070000 Is this hardware adress?
13:19:46B4gderare there other addresses?
13:19:53 Quit Makuseru (Read error: 104 (Connection reset by peer))
13:21:53markunwhat: somewhere on the website you should be able to find a link to the S6L8700 datasheet, there you can find all the addresses
13:29:30 Join fredddy [0] (
13:29:59zorgblaubaerhi i've got a problem with the svn. i always get the error: repository moved permanently to /'viewvc.cgi/';please relocate... can anyone help? i want to try the m6sl bootloader
13:30:12J-23What source file defines NB_SCREENS?
13:30:33B4gderzorgblaubaer: what exactly are you trying then?
13:30:33JdGordonor grep
13:31:07zorgblaubaerfirst of all i want to see any folders at all ;)
13:31:17zorgblaubaeri try to connect via rapidsvn
13:31:47B4gdereh, why not just follow the instructions that we have?
13:31:55B4gderif you wanna go fancy, you're on your own
13:32:50zorgblaubaerokay thanks ill try ;)
13:32:57zorgblaubaeryes just found it thank you
13:34:30zorgblaubaerworked like a charm, thanks again... me so dump xD
13:36:16 Quit Zarggg ()
13:42:31zorgblaubaererr yes so can anyone help me with compiling the bootloader? or is it better to ask on the forums?
13:43:32***Saving seen data "./dancer.seen"
13:44:12zorgblaubaerthanks, ill go through this
13:44:49J-23did anybody run Rockbox normal build on e200v2?
13:49:14B4gderJ-23: I think the forum thread is pretty clear on that progress
13:49:23J-23ahh, sorry
13:59:11 Quit skipper_ ("Leaving")
14:04:24fredddyhi I got my Fuze today and want to help with getting rockbox started
14:05:08fredddyI`ve already compiled an bootloader with svn and patched it with mkamsboot
14:05:53fredddyam I right if something goes wrong I can boot in the OF ?
14:05:59 Quit what ("CGI:IRC (EOF)")
14:06:12 Join what [0] (n=79b45a2d@gateway/web/cgi-irc/
14:08:48 Quit YOJI ("CGI:IRC (EOF)")
14:15:23zorgblaubaeri am getting error messages while compiling ;( i used the bootloader option in autoconfig... i receive 2 "undeclared"s while compiling
14:15:58zorgblaubaerGPIO5_SET and GIO6_READ
14:16:30zorgblaubaerah and GPIO5_CLR and struct storage_info has no member names serial
14:16:48stu8ballIs it legal to have a wildcard in a playlist file, and will Rockbox use it right if it is and I do?
14:17:27B4gderrockbox doesn't support wildcards like that, no
14:24:16JdGordongevaerts: what do you tihnk of;topic=14064.0;attach=3302 ?
14:24:35JdGordonI tihnk untill a target comes along which needs something more complicated that could work
14:24:50 Quit __lifeless (Remote closed the connection)
14:28:40gevaertsJdGordon: I think there are already targets where that isn't sufficient : the AMS sansas ;)
14:28:56JdGordonthat is the fix for the ams sansas
14:28:59gevaertsYou don't want that for the external SD card
14:29:24JdGordoncrap, yeah, they are going to show up as superfloppies also wont they? :(
14:29:38gevaertsThey're bound to
14:30:02JdGordonok, back to the drawing board :p
14:33:30gevaertsI still think a target specific function is better, i.e. storage_data_offset(IF_MV_NONVOID(int drive)), possibly guarded by HAVE_STORAGE_OFFSET
14:34:12Zagorarghh. why does ld care which order libraries are placed on the command line?
14:34:42Zagor-lcodec -ltremor produces "undefined reference to `mdct_backward'"
14:34:59JdGordongevaerts: ok, ill do that then
14:35:00Zagor-ltremor -lcodec gives no error
14:38:04 Join wookieneck [0] (
14:40:10gevaertszorgblaubaer: sorry, meizu not compiling is my fault. I'll fix it today
14:40:15 Quit kachna (Read error: 110 (Connection timed out))
14:41:57wookieneckhey, just found out about rockbox this morning. Im freaking excited at the prospect of neutering my Ipod, and I promise to read the manual before I ask stupid questions.
14:42:06 Join Photoguy [0] (n=Adams@
14:42:28zorgblaubaergevaerts: ok thanks ;) ill check back later than
14:43:27 Part B4gder
14:45:32 Join Seed [0] (
14:45:41 Part LinusN
14:46:09 Nick JdGordon is now known as JdGordon|zzz (n=jonno@rockbox/developer/JdGordon)
14:47:38PhotoguyIs there a way to boot form MicroSD (e200)
14:47:53PhotoguyMy Usb connetor is shot.
14:48:18 Join tyfoo2 [0] (
14:48:51BigBambiPhotoguy: Not at present
14:49:08PhotoguyWhat about an alternative firmware?
14:49:08 Quit what ("CGI:IRC (EOF)")
14:49:14 Quit tyfoo2 (Client Quit)
14:49:21BigBambiYou would need to do some fairly big bootloader modifications (I'm guesing there is no SD driver in the bootloader)
14:49:24PhotoguyJust so I can transfer the needed files.
14:49:34BigBambiIf you got a new bootloader going, you would need USB to update it
14:49:44 Join tyfoo2 [0] (
14:50:06PhotoguyAll I need to do is transfer the .rockbox folder to the internal..
14:50:11PhotoguyThen I'd be good.
14:50:19BigBambiCan you already boot Rockbox?
14:50:26BigBambiAre you looking to update or install?
14:51:13PhotoguyInstall, I had it installed before, but I had a cop/paste error, which overwrote the .rockbox folder.
14:51:41BigBambiI don't think the OF lets you do copy/pasting, so without USB I think you are stuck
14:51:59PhotoguyOk, I guess I'm going to have to accept it.
14:52:16 Join cmacis [0] (
14:52:38 Join LambdaCalculus37 [0] (i=44a04303@gateway/web/ajax/
14:52:54cmacisIf there's developers of rockbox here, I'd like to start by saying that it is brilliant!
14:53:29BigBambiOn the behalf on those that did the work, thanks very much :)
14:54:13cmacisI'm still getting used to it, but I like it.
14:55:24PhotoguyIt sure is capable isn't it?
14:56:45cmacisyes, very slick. I do, of course, have a question. Is the only video type supported MPEG?
14:56:49 Nick jfc^3 is now known as jfc (
14:57:00gevaertsMPEG1 and MPEG2, yes
14:57:30cmacisah, I'll make sure I get an encoder that deals with that then.
14:58:06BigBambior in fact
15:02:38 Quit tyfoo (Connection timed out)
15:06:54 Part wookieneck
15:13:06 Join Schmogel [0] (
15:20:45gevaertszorgblaubaer: if you update now, it should compile
15:23:41 Join robin0800 [0] (
15:30:49 Join nplus [0] (n=nplus@141.25.Globcom.Net)
15:39:11PhotoguyI did it!!!
15:40:40PhotoguyI fixed it!!
15:41:08 Quit Seed (Nick collision from services.)
15:41:13 Join Seed [0] (
15:41:21bodymindPhotoguy: how?
15:41:52PhotoguyBy holding together my usb connecter (in three parts)
15:42:22bodymindaw right :)
15:42:27PhotoguyThen I quickly copied rockbox folder..
15:42:38Photoguy(Still holdong it together.
15:42:46PhotoguyAh, yes!
15:43:21*Photoguy 's arm is sore
15:43:34***Saving seen data "./dancer.seen"
15:46:00PhotoguyActually it fell apart, before it totally copied, fortunatly it had already copied the bootloader/main files needed.
15:46:23PhotoguyThen, I could copy the rest via MicroSD.
15:49:30 Join dany_21a [0] (
15:50:18 Join n1s [0] (
15:50:58zorgblaubaeri am trying to launch the meizu m6sl bootloader with the linux dfu tool. i start it with "m6sl rockboot.ebn rockboot.ebn" but when it tries to send the second file it prints an GET CPU error and after a few seconds the battery charging screen of my meizu pops up. i tries commenting some stuff out in the source but nothing worked so far. maybe the usb_dev_close() function is bugged?
15:51:15zorgblaubaercan anyone help? ;)
15:51:39gevaertszorgblaubaer: the second one won't work, that's normal.
15:51:58*gevaerts goes to get his m6sl
15:52:26zorgblaubaergevaerts: ok but i dont think its normal that the battery charging pops up
15:55:44*gevaerts investigates
15:55:59 Quit Seed (Read error: 104 (Connection reset by peer))
15:56:18 Join {phoenix} [0] (
15:58:32gevaertsdenes_: ping
16:01:34markungevaerts: do you get the same?
16:02:02gevaertszorgblaubaer: I seem to remember vaguely that m6sl is broken atm. The latest crt0.S doesn't work properly on it yet
16:02:11 Join kronflux [0] (
16:03:34gevaertsmarkun: yes. The problem is that the current crt0.S uses the m3 memory layout
16:04:05zorgblaubaergevaerts: thats a pity.. i thought i could see some code in action so i would get a better access to the development ;)
16:05:18gevaertsYou could go back to an older revision, like r18717 (I think that one should work on m6sl)
16:05:50gevaertsYou won't see much though, just backlight toggling in response to the touchpad
16:06:46zorgblaubaergevaerts: ok thank you ill try that. maybe i will look into what crt0.S exactly means tonight and try to fix it.. i have a feeling ill feel crazy tonight
16:07:00 Quit denes_ (Read error: 110 (Connection timed out))
16:07:03 Join denes_ [0] (
16:07:29markunzorgblaubaer: if you feel bored, you could try to fix the SPI communication with the LCD on the M6SL :)
16:07:41gevaertsmarkun: crt0.S needs to be either duplicated or fixed to work on all meizus. I just thought at the time that it wouldn't be right to hold back one port (m3) because the others weren't keeping up
16:08:33zorgblaubaermarkun: ill keep that in mind
16:08:57markunzorgblaubaer: right now we're seding and receiving something, but we always get the chip ID instead of the actual value of the register we are trying to read from..
16:09:38markunmaybe we only need to fix this to get the LCD init working and then the rest "should be easy" :)
16:10:01zorgblaubaeri see
16:10:05gevaertsThe rest of the LCD, yes :)
16:11:00J-23Where (on the player) should rockbox.sansa file be put?
16:11:03J-23to main directory?
16:11:47linuxstbIn /.rockbox/
16:13:13linuxstbBut the correct way to install is to do "make zip" and then unzip the created to the root of your device.
16:14:13J-23metronome.c says that TIMER_FREQ is undefined when building for e200v2
16:14:59*Zagor now has complete ipod video build with a single 'make'
16:15:12gevaertsZagor: congratulations!
16:15:25n1sZagor: nice
16:16:15LambdaCalculus37Zagor: Nice!
16:16:22linuxstbZagor: Have you measured the speed difference?
16:16:27gevaertsOnly 74 files touched :)
16:16:57Zagorlinuxstb: no. how do I clear ccache?
16:17:01Zagorgevaerts: :-)
16:17:13linuxstbNo idea...
16:17:32gevaertsZagor: unless you touched that bit, you could run configure −−no-ccache
16:17:51gevaertsOr else ccache -C for the hard way
16:19:06zorgblaubaergevearts: i tried the rev you mentioned and a few other older ones but the behavious is persistend. since i got to get ready for university now ill try again later, but thanks for your help ;)
16:21:37Zagorthe high number of touched files is due to me changing the way some include files are handled
16:21:43J-23hmm, for some man e200v2 bootloader displays "Loading firmware" forever and never loads it
16:21:56Zagornamely generated (bitmaps) and library files (codecs and plugins)
16:22:05J-23on an e250v2
16:22:53kronfluxI have the sansa fuze if anyone needs any testing done. preferrably without blowing it up :p but its a risk I'd be willing to take
16:24:42kronfluxdont know whether thats the same type of player you're talking about, but hey! just trying to contribute ;)
16:25:29 Join Thundercloud [0] (
16:26:21LambdaCalculus37kronflux: It (sort of) is. :)
16:26:44kronfluxw00t, I was (sort of) right! :D
16:26:52gevaertsJ-23: have you tried multiple times? I understand that it's still a bit of hit-and-miss
16:27:16Zagorspeed is 2:05 for my build vs 3:08 for the old version. that is with ccache -C and single core (-j1)
16:27:29 Join XavierGr [0] (n=xavier@rockbox/staff/XavierGr)
16:28:01Zagorsurprisingly big difference, I must say
16:30:27gevaertsNow you need a cygwin tester
16:30:37ZagorI do
16:31:06*gevaerts wonders if cygwin would run in wine
16:31:19Zagorand then I have to fix all other targets, and sims, and...
16:37:15 Join krooven [0] (n=598b3478@gateway/web/cgi-irc/
16:38:37 Join Kopfgeldjaeger [0] (
16:39:19kroovenhi, is any1 know about useing rockbox on fake ipods?
16:39:23Zagorapps/codecs/wma.c:467: warning: 'packetlength' may be used uninitialized in this function
16:39:49Zagorkrooven: every player needs a port to be supported.
16:39:59Zagorwho knows the wma codec?
16:40:33rasherZagor: iirc saratoga
16:41:06kroovenbut can i know this port? i cant get info about it
16:41:58Zagorkrooven: all working ports are listed on the front page of
16:42:19Zagorports in various stages of progress are discussed in our forum
16:42:22 Quit kronflux ("Leaving")
16:43:42 Join miepchen^schlaf [0] (
16:43:44XavierGrhmpf the beast bench kept going for only 6.35 hours. My previous bench was 6 hours (though with 160 CBR mp3 instead of 192 VBR now)
16:43:58XavierGrvery poor battery performance, I expected something more.
16:44:19kroovenzagor: well my is not list there so there no change it will work or should i try it like it was a real ipod?
16:44:44XavierGr(me previous benchmark was with the stock battery 700mAh, the newer one is with a cameron sino 1000mAh, that's why I am surprised)
16:45:22LambdaCalculus37XavierGr: Better than mine right now. I'm getting just under 6 hours on my beast with Q5 VBR Vorbis.
16:45:29gevaertskrooven: can you install new firmware with itunes?
16:45:38Zagorkrooven: no, it won't work. the "fake ipods" only look like ipods on the outside
16:46:21 Join jgarvey [0] (
16:48:18kroovengevaerts: no
16:48:30BigBambiZagor, gevaerts I haven't been following, but saw cygwin testing mentioned. What would you like testing?
16:48:48XavierGrLambdaCalculus37: stock or upgraded battery?
16:48:51kroovenbut mybe anyone know where i may find some other firmware?
16:48:54ZagorBigBambi: FS #9534
16:49:04n1s<Zagor>arghh. why does ld care which order libraries are placed on the command line? <- isn't that because the codec lib is compiled with function sections and linked with gc-sections so if nothing depends on a function it is gc'ed (tremor relies on the mdct in the lib)
16:49:06ZagorBigBambi: hang on, new patch coming
16:49:16BigBambiWill do :)
16:49:17gevaertsZagor: apps/codecs/codecs.make seems to be missing
16:49:27Zagorgevaerts: oh, thanks
16:50:32Zagornew patch posted
16:50:32LambdaCalculus37XavierGr: Stock.
16:50:52Zagorshould work with ipod video or sansa c200
16:50:57LambdaCalculus37I'm waiting for a couple of CameronSino batteries that I ordered to arrive.
16:52:46Zagorn1s: but should that really depend on the order on the command line?
16:52:47 Quit l403 (Read error: 113 (No route to host))
16:52:54 Quit krooven ("CGI:IRC")
16:52:55BigBambiZagor: So make times with and without the patch?
16:53:29ZagorBigBambi: well start with seeing how far you get. I would be surprised if it works in cygwin...
16:53:35gevaertsZagor: LANG seems to be a bad name for a variable. perl complains about locale things
16:53:37BigBambiah, OK :)
16:53:46 Join Nico_P [50] (n=nicolas@rockbox/developer/NicoP)
16:54:19Zagorgevaerts: perl?
16:54:29gevaertsgenlang things
16:54:40n1sZagor: i don't know if it should but i would guess it does, btw about your request for comments, I think the largest problem with our buildsystem was/is that few people understands make well enough to maintain it so you might not get many insightful comments :)
16:54:56gevaertsApparently there's a variable named LANG set with all .lang files
16:55:01Zagorn1s: :-)
16:55:12Zagorgevaerts: yes, in apps/lang/lang.make
16:56:11Zagorrename it LANGS or something. it's only used in the top two lines
16:56:30gevaertsIt doesn't stop on it, so I won't worry too much
16:56:43gevaertsIt falls back to "C", so that's fine
16:57:06XavierGrLambdaCalculus37: I have an 1000mAh gigabeast battery, I expected around 30-40% more juice
16:58:22gevaerts8m27.702s to 5m22.573s here (ipod video, building on eee900)
16:59:00Zagora definite difference
17:00:38gevaertsIndeed. Well worth doing I think. What exactly needs to be done for each target?
17:01:22rasherIs it also simpler? Because that'd be more important than any speed difference, if you ask me
17:01:52Zagorgevaerts: the only "known" thing is removing the STARTUP line from each line. I expect there are also a number of other small fixes for some targets.
17:02:35Zagorrasher: I would say so, yes. especially in terms of dependencies. and the fact that you can specify any build target you want and it builds only it.
17:02:49gevaertsrasher: IIUC the goal was correctness, i.e. to fix dependency bugs
17:02:53LambdaCalculus37XavierGr: There's still no proper power management on the beast AFAIK. jhMikeS is working on that.
17:04:01ZagorI have to go. I'll pop in later tonight.
17:04:03 Quit Zagor ("Client exiting")
17:07:04BigBambiWell, it does indeed fail on cygwin
17:08:15*bodymind was so excited building e200v2.. and suicides with timer_freq undefined....
17:08:33 Nick Llorean1 is now known as Llorean (
17:09:09 Quit CaptainSquid ("Miranda IM!")
17:09:17 Quit zorgblaubaer (Remote closed the connection)
17:09:24BigBambiI selected ipod video - why is iaudio_bl_flash giving errors?
17:09:33*gevaerts thinks that BigBambi's paste counts as "spectacular"
17:09:48gevaertsscramble is a shared tool
17:09:52BigBambiah, OK
17:10:27BigBambiI'll add the link to the flyspray task
17:10:39fredddydoes someone with a fuse get more then just disk_init failed ?
17:12:35 Quit Zambezi (Remote closed the connection)
17:13:30 Join kachna|lappy [0] (
17:14:10gevaertsBigBambi: maybe try to run make in tools/ first
17:15:22BigBambijust "make" whilst in tools? As that just says "Run make in your build directory!"
17:15:23 Join l403 [0] (n=l@
17:15:47rashermaybe make scramble
17:15:58BigBambigood thinking!
17:16:05gevaertsah, yes. make scramble works :)
17:16:08BigBambiOK, that works
17:18:29 Quit Photoguy (Read error: 104 (Connection reset by peer))
17:25:58rasher_Bagder_: Why must I remove my download dir before running?
17:26:05BigBambigevaerts: seems to work now
17:26:19rasher_Bagder_: ignore me
17:27:10rasherBigBambi: did you notice that your paste is incomplete?
17:27:17BigBambiit is?
17:27:38rasherThe lines got cut off
17:27:52BigBambiah yes, woops
17:27:56BigBambiI'll fix that
17:30:12 Join MrDuck [0] (
17:31:28 Quit kachna|lappy (Read error: 104 (Connection reset by peer))
17:40:35 Join kugel [0] (n=chatzill@unaffiliated/kugel)
17:41:58 Join bmbl [0] (n=Miranda@unaffiliated/bmbl)
17:43:37***Saving seen data "./dancer.seen"
17:43:45kugellinuxstb: reading your commit, I guess my findings yester where a bit helpful. but now you beat me to have patch making fuze (and e200v2) build compile :)
17:48:34J-23what does "ATA error -1" message mean?
17:49:20 Join webguest99 [0] (n=57a7505c@gateway/web/cgi-irc/
17:49:57webguest99is some here?
17:50:44webguest99i have a big problem.. after installing rockbox i cannot restore my ipod
17:54:35 Nick _Bagder_ is now known as Bagder (
17:55:54 Join massiveH [0] (
17:57:45advcomp2019webguest99, did you read the manual and wikis
17:59:28LambdaCalculus37Also, what iPod model?
18:00:33 Join kugel_ [0] (
18:01:06kugel_J-23: ata_sd_as3525.c should tell you
18:02:38 Join kugel__ [0] (
18:02:42 Join faemir [0] (
18:02:44 Quit massiveH ("Leaving")
18:02:49 Quit kugel (Nick collision from services.)
18:02:50 Nick kugel__ is now known as kugel (
18:03:38webguest99its a 5.5 80gig
18:04:04markunwebguest99: I know very little about the ipods, sorry
18:04:21webguest99and when i start it allways" connect to your computer.use itunes to restore." is shown...
18:04:22markunhopefully someone else can help you. I think that restoring should always work.
18:04:51webguest99yes but restore stops because it says ..files used form another system
18:06:32 Quit voltagex (Remote closed the connection)
18:06:43 Join voltagex [0] (
18:08:04LambdaCalculus37Are you using your iPod on a Mac, or a Windows PC?
18:08:35webguest99i cannot format ist by force because windows did not recognize it..
18:08:48LambdaCalculus37Does iTunes recognize it?
18:09:04webguest99not allways
18:09:24webguest99i 'm trying anapod reinitializer..but it did not seem to work..
18:10:38LambdaCalculus37I'm not 100% certain about this, but you may be having file system problems. Can you copy any files to the iPod?
18:10:58 Quit kugel_ (Read error: 145 (Connection timed out))
18:11:06webguest99no, because it is not recognized as known File system on windows
18:11:57LambdaCalculus37Can you google for a utility called Swissknife?
18:12:08LambdaCalculus37Download and install that, then format the iPod drive with it.
18:12:21LambdaCalculus37Then try restoring the iPod again afterwards.
18:12:42LambdaCalculus37If it still doesn't work, then we can figure out another solution.
18:12:43gevaertswebguest99: first of all try booting it into emergency disk mode if you're not already doing that
18:13:10LambdaCalculus37webguest99: gevaerts is right. Boot into disk mode (MENU+SELECT, then SELECT+PLAY immediately after).
18:13:29LambdaCalculus37Hold MENU+SELECT until the iPod reboots.
18:13:31 Quit perrikwp (" ajax IRC Client")
18:13:52webguest99i know is the only way to start ipod
18:13:58webguest99otherways porblem screen is shown
18:14:46webguest99with hp usb disk storage format tool it says "device is write protectet"
18:17:17gevaertsDo you have administrator rights?
18:21:19 Join MethoS [0] (
18:21:28gevaertsIs itunes running? Maybe it interferes
18:23:06webguest99no i closed it
18:23:23webguest99i have retried with hold switch on/off...
18:23:35rasherYou may need to kill some background services as well
18:24:27gevaertsIf you can't access the disk properly, maybe there's something wrong with the disk itself. Disk mode should never fail
18:24:29webguest99i close all tasks with *itunes or *ipod
18:25:33 Join perrikwp [0] (i=982137c3@gateway/web/ajax/
18:25:35 Quit {phoenix} (Remote closed the connection)
18:25:52kugelgevaerts: it's not gevärts???
18:26:53gevaertskugel: indeed not :)
18:27:17kugelI gotta change my mind :S
18:27:23LambdaCalculus37webguest99: gevaerts is Belgian, and that's enough off-topic chatter for now. Back to business.
18:27:31LambdaCalculus37webguest99: You may have a dying hard drive.
18:28:29kugelLambdaCalculus37: yea, back to business. I think the backlight fading patch is matured enough now, I'd like to get it committed.
18:28:31webguest99it has worked fine... i tried it on another mainboard.. from a 5g ipod.. it did not work.and now it didn't work on 5.5G ether
18:28:40kugelI just fear another epic binsize discussion
18:29:07gevaertskugel: which targets is this for? sansa and beast?
18:29:13 Join {phoenix} [0] (
18:29:26kugelgevaerts: yep, and from what I've seen easily expandable to e200v2 and fuze too
18:29:47kugelgevaerts: basically all targets who have backlight brightness setting can use it
18:30:07gevaertsSo only targets with either flash, lots of RAM, or both. Just shout loudly at anyone who brings up binsize
18:31:14kugelwhat does it have to do with flash? you mean because reduced ram results in more disk spin ups?
18:31:38 Join bertrik [0] (
18:32:00gevaertsyes. That's the main binsize argument (which is true IMHO). Flash doesn't have that issue, and the beast has 64MB of RAM.
18:32:06LambdaCalculus37kugel: I have no problems committing it, but it's not entirely up to me as well.
18:32:08kugelbut as most of the time, the patch would - if any - cause a disk spin only a few frames earlier
18:32:44 Join domonoky [0] (n=Domonoky@rockbox/developer/domonoky)
18:33:24 Part domonoky
18:33:28kugelLambdaCalculus37: I understand and I don't ask you (directly). I'm fine with the work out the patch-discuss-reject for binsize reasons schema
18:33:31 Join obo [0] (n=obo@rockbox/developer/obo)
18:33:38 Join domonoky [0] (n=Domonoky@rockbox/developer/domonoky)
18:33:39kugelerr s/reject for binsize/evaluate the worthness
18:33:42 Quit bodymind ("Ex-Chat")
18:33:42 Quit webguest99 ("CGI:IRC")
18:34:10gevaertskugel: sure, but it's not the only feature that's not used by everyone :) If you have 500 features taking 1K each, you've halved the buffer on the Archoses
18:34:39kugelmost features I care about wouldn't hit archos anyway
18:34:48LambdaCalculus37And not to mention nearly all of it on limited mem targets (m200, iFP...)
18:34:51kugelbut I'll do a binsize test, just to be prepared
18:35:15gevaertsBut as I said, on sansa classic or beast, it shouldn't be an issue
18:35:16LambdaCalculus37Do binsize tests on the e200, beast, and any other targets this affects.
18:35:22n1sgevaerts: and killed the voice UI while at it (it needs to keep the voicefile in ram ~1MB on archos)
18:35:46LambdaCalculus37No wonder I have almost no buffer on my Archos. ;)
18:35:56*LambdaCalculus37 sees only 439KB buffer with Rockbox 3.0
18:36:17kugelyou have an archos?
18:36:30LambdaCalculus37Jukebox 5000 and JBRv1.
18:36:50kugeli imagine the disk is spinning all the time
18:37:13LambdaCalculus37On the JBR, yes. And the battery's getting sapped very quickly as a result.
18:37:39kugelgoing from that point of view, the rest of the buffer can be used for features too, since the disk is spinning all the time anyway :)
18:37:58n1skugel: IIUC you are doing this by adjusting the backlight brightness, are you sure the (incredibly feature packed) pmic in the beast has no other (nicer) way to do this?
18:38:25 Part cmacis ("Leaving")
18:39:04kugeln1s: I talked about that with jhMikeS. We agreed on deactivating the hardware fading since it's hard to handle and does only feature fading down as of now
18:39:16kugelor was it only fading up? one of those
18:40:26LambdaCalculus37kugel: Let's have another talk with jhMikeS about your patch, and if he agrees with it as well, I'll commit it.
18:41:00kugelLambdaCalculus37: iirc JdGordon was also in favor of this patch if it uses the backlight thread
18:41:04kugeland it does that
18:41:43LambdaCalculus37Well, seeing as it's using the backlight thread, then I don't think there should be very much an issue.
18:42:48kugel JdGordon quite messed with it
18:44:43kugelone thing I'm not sure about is though, I focused on having the old code still working (seeing that this is a feature which should be imho able to be deactivated) the lcd-*.c files have quite a bit #ifdefs now
18:45:21 Join tessarakt [0] (
18:45:58kugelmaybe I can factor it out a bit more
18:46:29LambdaCalculus37kugel: See what you can do with it. If you can knock a few of the #ifdefs down a bit, it might make it a little better.
18:48:46 Join nuonguy [0] (
18:51:08*BigBambi is confused
18:51:29 Quit einhirn ("Miranda IM! Smaller, Faster, Easier.")
18:51:36BigBambigevaerts: I just checked out a new source (same as last time) and tried Zagor's patch again) and it worked
18:53:05gevaertsBigBambi: weird. Did you measure timing?
18:53:16BigBambino, I will do now
18:54:52 Quit moos ("Rockbox rules the DAP world")
18:56:37pixelmaLambdaCalculus37: 439 kB buffer? How did you manage that?
18:57:49 Join freddy__ [0] (
18:58:15 Quit fredddy (Read error: 110 (Connection timed out))
18:58:59pixelmaI have 1.29MB of free buffer on my Ondio (ok, there are hardware and feature differences to the RecorderV1 but shouldn't be that much)
18:59:38 Quit _Tristan_508 (Client Quit)
19:00:03kugeldircache and load to ram maybe?
19:00:21n1skugel, none of those are included on archos
19:00:23pixelmano, they don't exist on lowmem targets
19:00:26 Join Tristan [0] (
19:00:29kugeloh didn't know
19:00:36n1spixelma: he mentioned a voice file i think
19:00:54 Join meven [0] (
19:01:12pixelman1s: I have voice files too, they are not counted there it seems
19:01:13 Quit BigBambi (Read error: 104 (Connection reset by peer))
19:01:28 Join BigBambi [0] (n=Alex@rockbox/staff/BigBambi)
19:01:53n1spixelma: a voice file should be counted, as having it loaded makes the buffer smaller
19:02:09domonokythe voice file is loaded in the remaining buffer.. (of course only when no music is playing)... so he would need a 439kb voicefile to get it to load :-)
19:02:55domonokyn1s: no extra buffer for the voicefile, it uses the audiobuffer...
19:03:04pixelmait isn't currently, and on the hardware codec targets it won't make the buffer for music playback smaller, it's unloaded during playback and you couldn't have music and voice at the same time anyways
19:04:01n1sah, i thought it worked in a similar way as on swcodec except for the no voice and music at the same time bit.
19:05:37pixelmathen I would probably have next to no free buffer for playback... the voice files I use are usually +1MB
19:06:34domonokyits very hard to make a understandable mp3-voicefile under 1MB...
19:09:00 Quit Darksair ("(define (add-1 n) (lambda (f) (lambda (x) (f ((n f) x)))))")
19:09:05 Join Bensawsome [0] (n=Bensawso@unaffiliated/bensawsome)
19:09:35 Join jhulst [0] (n=jhulst@unaffiliated/jhulst)
19:12:02pixelmakugel: one argument against your patch was also that it will look worse than the current backlight fading and maybe having two methods is not an ideal solution
19:12:53kugelpixelma: I don't remember that argument
19:13:01pixelmajust remembering what I read so far, didn't try myself and couldn't care less
19:13:07kugelif it looks worse on the beast, LambdaCalculus37 didn't tell me
19:13:24 Join petur [50] (n=petur@rockbox/developer/petur)
19:13:26kugelon e/c200 it can't look worse as there's no fading
19:13:44pixelmaI meant compared to what is there on the Irivers for example
19:14:45kugelI can't control how it looks if the lcd doesn't offer enough/linear brightness levels
19:15:07kugelLambdaCalculus37: you might comment on the quality of the fade on the beast now
19:15:55linuxstbkugel: Why do you fear an epic binsize discussion? Is your patch causing an epic binsize increase?
19:16:20kugelno, it's just my experience
19:16:30pixelmakugel: if I understand and remember correctly you could and that's what current backlight fading does with pwm
19:17:29kugelI really don't remember it's looking bad. I can imagine though that it cannot compete with hardware solutions
19:23:39bertrikI quite liked how the creative zen of my brother fades down to a low level after a short timeout instead of switching off immediately
19:24:10bertrikbut that adds more complications than just fading ...
19:24:23*linuxstb senses even more backlight settings ;)
19:24:54gevaertsI think that if a low level is good enough for the screen to be readable, you should just use that level
19:25:01pixelmaas far as I know, current backlight fading is done in software too and you'll probably need something to compare directly (don't know if the sim could be helpfull there). It's just theory on my side but I thought I remember amiconn speaking with you about this not that long ago, currently searching the logs for it
19:26:20kugelbertrik: more complications as in looping through the lowest brightness level a few times while fading
19:27:30bertrikkugel, maybe it's not that complicated w.r.t. fading, but indeed it adds an extra backlight setting as linuxstb remarked
19:27:37LambdaCalculus37kugel: Backlight fading on the beast looks fine to me.
19:27:39 Join tvelocity [0] (n=tony@
19:31:30dany_21ahello all... quick question: which file do i need for mkamsboot as "<boot file>" for the fuze - bec of:
19:32:02kugelbertrik: to make the pause configurable? I wouldnt do so
19:32:52kugeldany_21a: the bootloader file you build
19:33:18kugeldany_21a: tools/configure −−target=58 −−type=b
19:36:17dany_21akugel oh... of course - thx, worked
19:37:44kugeldany_21a: no problem
19:37:53 Quit nuonguy ("This computer has gone to sleep")
19:38:17BigBambigevaerts, zagor: 24m37 without the patch, 16m25 with the patch
19:38:45 Join Horscht [0] (n=Horscht@xbmc/user/horscht)
19:39:37gevaertsSo it seems to be about one third faster everywhere
19:39:42 Quit Schmogel ("Miranda IM! Smaller, Faster, Easier.")
19:40:02*gevaerts thinks that this is a nice side effect from fixing dependencies :)
19:40:20BigBambiIt is a nice speed up to be sure
19:40:41 Join ktulu64 [0] (n=ktulu@
19:41:41 Quit perrikwp (" ajax IRC Client")
19:42:57 Part ktulu64
19:43:38***Saving seen data "./dancer.seen"
19:45:28 Join Seed [0] (
19:46:09 Join allele [0] (
19:47:11alleleI'm trying to write directly to the frame buffer, but currently rb->lcd_framebuffer[x][y] = value; doesn't seem to be the right route
19:47:13Nico_Pgevaerts: is it also one third faster on linux?
19:48:32n1sallele: you need to lcd_update after that i think
19:48:58gevaertsNico_P: 8m27.702s to 5m22.573s for me (on eee900), and Zagor gor 3:08 to 3:05
19:49:07alleleno, I have that, but that code won't even compile
19:49:11gevaerts3:08 to 2:05 that is
19:49:18Nico_Pbetter :)
19:49:29alleleI'm trying a single dimensional array
19:49:30Nico_Pthose are nice figures indeed
19:49:33n1sallele: then you need to share a bit more details :)
19:50:02gevaertsIndeed. We need to get that finished soon
19:50:07 Join MethoS- [0] (
19:50:21*amiconn is surprised by his beast's runtime test result
19:50:31Nico_Pall that by simply unifying the makefiles?
19:50:38Nico_Por are there other improvements?
19:51:19gevaertsIt doesn't build intermediate libraries anymore, which may also save a bit
19:52:11LambdaCalculus37amiconn: How long did it run for?
19:52:16allelewell prettymuch I'm working on a plugin that I would like to display normal for other players (particularly the ipod nano 1g) and for the sansa e200 the image will be rotated 90 degrees. I'm thinking since I already have a h file for each player I may as well add a custom display method that would be applied to each method and just return values to be written to the frame buffer.
19:52:16amiconnallele: This won't work in plugins. rb->lcd_framebuffer is simple pointer to fb_data, not a pointer to an array
19:53:19n1samiconn: you got more runtime than expected?
19:53:23amiconnallele: Also, writing directly to the framebuffer will get you into serious trouble on the greyscale and monochrome targets (pixels are packed on these)
19:53:33amiconnn1s: Yes, almost 5:30
19:53:43amiconnI expected <4 hours
19:53:50alleleamiconn: alright, well could the method return an array that I can just use lcd_butmap_transparent in then?
19:54:00n1sthat's about what i get on mine with q4 vorbis
19:54:23amiconnThe bitmap functions work on the framebuffer
19:54:31 Quit Seed ("cu, Andre")
19:54:55amiconnIf you want to display rotated on portrait displays, you need to draw rotated
19:55:05LambdaCalculus37amiconn: Is your beast on the stock battery or a different battery?
19:55:20amiconnSInce the bitmaps can be display specific, there's no need to go through hoops
19:55:23amiconnLambdaCalculus37: stock
19:55:31amiconn(I think)
19:56:01LambdaCalculus37Not bad... that's roughly what I'm getting (5:40 on my S with stock battery and Q5 Vorbis)
19:56:41amiconnThis was my standard test folder (mp3 lame vbr, ~200kbps)
19:57:24amiconnI'd expect the codec to have almost no effect on runtime for a beast SVN build
19:57:31 Quit MethoS (Read error: 145 (Connection timed out))
19:57:47amiconn(CPU running at fixed clock, and the CPU is essentially bored with mp3, vorbis etc)
19:58:25 Join Seed [0] (
19:58:29 Quit linuxstb (Read error: 60 (Operation timed out))
20:00:20 Join bluebrother [0] (
20:01:01 Join _Andrew_ [0] (n=on@
20:01:11 Part dany_21a
20:03:17allelealright follow up question: native bitmaps for each player are all arrays of shorts, right?
20:04:13 Join bodymind [0] (
20:07:57 Quit Thundercloud (Remote closed the connection)
20:08:05allelealright, thanks for the help:-D
20:08:21 Part allele
20:09:57 Quit bodymind ("Ex-Chat")
20:12:57*n1s stops messing with the beast for today with no progress :/
20:13:44Nico_Pn1s: what were you working on?
20:14:10n1sNico_P: I'm trying to get fm working
20:14:57n1sI can read/write to the fm chip, it seems like seeking is sort of working but i can't get any sound
20:15:30bertrikyou probably need to enable an analog input to the codec
20:16:15 Join dany_21a [0] (
20:16:33n1sbertrik: that's what i have been trying, i found some settings for enabling analogue inputs, setting volume etc but i seem to have missed something
20:19:33LambdaCalculus37n1s: Do you have a patch for the FM radio on the beast? I'm willing to help test.
20:19:36 Join perrikwp [0] (i=d1a8d351@gateway/web/ajax/
20:21:00n1sLambdaCalculus37: i get no sound so it is not very useful :) also everything is a major mess and hacked into one of the debug screens... I will try to clean up what i have and post a patch in the tracker
20:21:23n1sbut not today
20:21:56 Quit ap0 ("Ba")
20:26:47 Quit robin0800 (Read error: 104 (Connection reset by peer))
20:28:23LambdaCalculus37n1s: No problem. Even if there's no sound, it's useful to get information out of the debug screen to at least see that it's being detected.
20:28:50n1sthe only thing i can think of except me missing something obvious is if they connected the fm to the mic input...
20:29:05n1sLambdaCalculus37: ok, i'll pastebin the whole mess :)
20:31:25bertrikhave you been able to define a reasonable low-level API to the fm registers?
20:31:53bertrikdo you just read/write ALL registers when reading/writing something?
20:32:04n1sLambdaCalculus37: here it is,
20:33:40 Quit l403 (Read error: 104 (Connection reset by peer))
20:33:52n1sbertrik: i have not made anything nice yet, but i plan to, I only write a couple of regs now. The config settings i've touched so far are in the 4 first regs and i always write all of those
20:34:20 Join webguest21 [0] (n=584a203b@
20:35:14n1sI'm thinking of an api like the one for the wm8978 where you have register clear/set/write functions that then call target specific functions and do any necessary magic
20:35:51amiconnThere is an api for tuner control iirc
20:36:18amiconnIt became necessary when Ondio support was added, so it's not even new
20:36:33 Quit webguest21 (Client Quit)
20:36:35n1samiconn: yes but i was thinkign about how the driver communicates with hardware
20:37:16amiconnHmm? Why would you need an api for that? It's target specific anyway...
20:37:20n1sLambdaCalculus37: look in the first item of the debug menualso, it hangs some times for me when trying to seek so keep your paperclip handy
20:38:10*LambdaCalculus37 holds up his paper clip ;)
20:38:29n1samiconn: this fm chip is used in several targets so they could share the driver by implementing separate functions for writing/reading etc and sharing the rest
20:39:03 Part dany_21a
20:39:10amiconnThe lower layer is i2c, and there you have your api
20:39:13bertrikamiconn, when using the i2c interface, you cannot read arbitrary registers but only a block of registers
20:39:33bertrikthis is different from the GPIO way of interfacing with the si470x
20:39:37amiconnHave a look at the tuner drivers. They do call fmradio_i2c_*
20:39:41 Join fdinel [0] (
20:40:08n1samiconn: that's the kind of thing i mean...
20:41:25 Quit SUSaiyan ()
20:42:58bertrikwe don't know yet how any of the ams sansas communicate with the si470x tuner chip, maybe they also use i2c
20:43:38 Join massiveH [0] (
20:48:43 Join Zagor [242] (
20:49:57 Quit PaulJam (".")
20:51:13 Quit bertrik (Remote closed the connection)
20:51:45 Join l403 [0] (n=l@
20:51:58n1sjhMikeS: assuming the fm chip is connected to the aux in of the wm, should enabling AUXL2LMIX and raising AUXLMIXVOL in WMC_LEFT_MIXER_CTRL and the same for right be enough or am i missing something?
21:02:57 Join funman [0] (
21:03:00 Join robin0800 [0] (
21:03:06ZagorI'm stunned my patch worked in cygwin
21:04:46 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
21:05:47funmanJdGordon|zzz: doesn't r19085 need a button_init_device() ?
21:06:06 Quit J-23 (Remote closed the connection)
21:10:42 Quit _Andrew_ (Read error: 110 (Connection timed out))
21:10:56Zagorif we can come up with a way to avoid having to run mkdir for each compile, I bet cygwin would speed up even more
21:11:12gevaertsZagor: try again. Maybe you can find a way to break it :)
21:15:15jhMikeSn1s: Are you sure it's on AUX?
21:16:49n1sjhMikeS: no, that's one of my possible problems... do you think they used another input?
21:17:33jhMikeSPossibly. I'd print out that block diagram and plan carefully by drawing the paths out. It makes it much easier for me anyway to plan.
21:18:05jhMikeSI'm sure some analog bypass is also needed
21:18:33 Part lfaraone
21:20:08jhMikeSThe only other possible path looks like L2/R2
21:21:06 Join culture [0] (
21:22:20n1sjhMikeS: i will look into that then, is is possible to have several sources inputting to the mixer at the same time or will i need to disable the dac first?
21:22:21funmanI wonder if the writes of PL180 datasheet were funny when naming the 'tx fifo half empty' and 'rx fifo half full' flags :P
21:23:16funmanas far as i could see, the sansa ams pl180 controller tells us that there is a rx fifo overrun, maybe it should just try again in case of an error
21:23:21jhMikeSn1s: shouldn't have to stop the DAC. you either need input to the ADC or to bypass the analog signal to the outpus.
21:24:03 Join Nibbler [0] (
21:25:36jhMikeSwhen this is done, FM recording might as well be added soon after
21:32:30Nico_PZagor: your makefile work is looking good :)
21:33:04Nico_PI gave it a try on an ipod video build and got a nice speedup
21:33:14BigBambiIt is a massive boon on cygwin
21:34:06LambdaCalculus37Zagor: Your makefile has sped up building on OS X quite a bit.
21:34:20Nico_PI wish my understanding of make was a bit more than just basic
21:34:28ZagorLambdaCalculus37: more than on linux/cygwin?
21:35:01LambdaCalculus37I haven't compared with Linux/Cygwin yet, but I'll at least compare with Cygwin later.
21:35:22BigBambiLambdaCalculus37: The speed up on both linux and cygwin is about a third
21:35:26BigBambiWhat was it on OSX?
21:35:30 Quit freddy__ (Read error: 60 (Operation timed out))
21:36:19 Join Strife89 [0] (n=michael@
21:36:27*gevaerts expects the same one third win as everywhere else
21:36:35LambdaCalculus37BigBambi: Approximately 25%; built with a beast build.
21:36:47 Quit Strife89 (Client Quit)
21:36:50BigBambiI thought that the patch was for ipod video and e200 only?
21:37:16Nico_Pso I guess recursive make is really harmful then :)
21:37:19*LambdaCalculus37 smacks himself
21:37:41 Join Strife89 [0] (n=michael@
21:37:51*LambdaCalculus37 shouldn't have so many players at once and should watch his builds more closely
21:37:52BigBambior even just ipod video (not sure where I got e200 from)
21:38:16LambdaCalculus37BigBambi: I'll try a video build with Zagor's makefile.
21:38:30Strife89I'm curious about the Clip code. I was trying to build a sim and there are some undefined values in "sound.c".
21:38:52LambdaCalculus37I got crossed also from finally being able to update GNU make to 3.81+ and seeing a big boost in build speed.
21:41:28funmanI got much more reliable SD reads on the sansa ams with this patch : . can you give it a try and confirm it's not the hazard ?
21:41:55*Strife89 will stay away from file I/O testing for now. ;)
21:42:19 Join MethoS-- [0] (
21:42:30Strife89Most of you know why. :)
21:43:41***Saving seen data "./dancer.seen"
21:47:14 Join freddy__ [0] (
21:49:15preglowgevaerts: anything in the way of using a struct of function pointers instead of a big switch with ifdefs?
21:49:30preglowthen an indexed array of those structs, of course, which will be ifdeffed
21:51:20funmanthe ata-sd-pp.c driver also retries when it the receive fifo was full before it could read it
21:51:30 Quit Strife89 ("Bye, guys!")
21:51:50gevaertsFor read and write that would be faster I guess, while not being extremely wasteful
21:51:55 Quit jhulst (Read error: 110 (Connection timed out))
21:52:00funmanI wonder if using DMA would solve this problem, I don't know if a DMA request has to be made each time the 16*32bits FIFO is full, or one time per sector
21:53:30gevaertspreglow: I'm mainly stuck on the currently non-drive-specific functions, like sleep() or spin(). My current thinking is to make the code pass a file descriptor with those. Any ideas?
21:54:51pregloweh? i don't even see how file descriptors fit in with those calls at all :)
21:55:01*amiconn would just "broadcast" these
21:55:09gevaertsThey're mostly called from apps/ code
21:56:49gevaertsSleep is indeed doable as broadcase, but not everything is. What happens is e.g. the mpeg player decides that it will need data again soon, so it tells the disk not to spin down.
21:56:50preglowwell, it depends if broadcasting is the right thing
21:56:57preglowi don't really know how and why they are called
21:57:00gevaertsYou don't want all disks spinning in that case
21:57:06amiconngevaerts: Where's the problem?
21:57:33amiconnata_spin() doesn't spin up the diskif it's sleeping, it merely prevents it from spinning down if it's spinning
21:57:34preglowbut no, broadcasting them might not always be the right thing to do
21:57:49amiconnSo broadcasting a _spin() call doesn't do any harm
21:58:04gevaertsamiconn: suppose you're on a HD+SD system, and you're watching an mpeg file from the SD card. You don't want to keep the HD spinning
21:58:19amiconnIt won't
21:59:05amiconnYou're loading the movie from SD, hence the hdd won't be spinning, and a _spin broadcast won't cause it to
21:59:26preglowit might be spinning
21:59:31preglowyou don't know that
21:59:36 Quit MethoS- (Read error: 110 (Connection timed out))
21:59:42preglowif i was browsing for a movie to watch, and had just accessed the hd, it might be spinning
21:59:49 Quit LambdaCalculus37 (" ajax IRC Client")
21:59:59amiconnbesides, mpegplayer doesn't call _spin
22:00:06Lloreanpreglow: Especially if the video is high-bitrate, and the disk spun to load mpegplayer.
22:00:19*gevaerts should look at the code so he can give accurate examples
22:00:20*Llorean would really like mpegplayer to default to "just play the video" rather than "stop and give you a menu"
22:00:34amiconnata_spin() is used during browsing, in order to prevent spindown and hence repeated spinup waits
22:01:03preglowanyway, i don't believe broadcasting them will always be the right thing
22:01:07amiconnTo the contrary, both standard playback and mpegplayer call ata_sleep() after buffering finishes
22:01:19preglowit might work now, but if there's another way that's easy to put in, that should be done
22:01:28amiconnNo, but it will be *almost* always the right thing, and it's KISS
22:01:31gevaertsamiconn: it does, for seeking
22:01:31preglowbut will callers of those functions always have a file descriptor to pass?
22:01:45amiconnpreglow: no
22:02:04 Join saratoga3 [0] (n=9803c264@gateway/web/cgi-irc/
22:02:35saratoga3Zagor: reagarding the wma.c warning, does adding "=0" to the variable define fix the warning?
22:02:42 Quit cg (Read error: 110 (Connection timed out))
22:03:02gevaertsIt's a bit academic right now, as there is only one target-in-progress that has HD+something else, and that's safely hidden in a drawer
22:03:05saratoga3i'm not sure why changing a makefile would introduce a warning though
22:03:34amiconnZagor: I really want to know what's wrong with the build system :\
22:04:12preglowsomeone give me an example why someone would keep the disk spinning
22:05:17linuxstbjpeg slideshow...
22:05:58 Quit freddy__ (Remote closed the connection)
22:06:04preglowis dircache default on these days, btw?
22:07:58amiconnBut even with dircache there's a reason to keep the disk spinning while browsing: .talk clips
22:08:08 Quit meven (Remote closed the connection)
22:08:09 Quit nplus (Remote closed the connection)
22:08:20preglowsure, i wasn't thinking of having anyone remove the possibility anyway :)
22:08:32preglowbut i do think we should enable that by default on 16meg++ targets
22:08:42preglowyes i do
22:08:52amiconnI mean I don't
22:09:33*amiconn has it enabled on his hdd-based 32MB+ targets
22:09:33preglowif i wasn't so completely convinced it would drive me mad, i would try browsing without it now
22:09:51preglowyeah, 16 meg might be marginal
22:09:57amiconnI don't use it on 16MB or lower, and it doesn't feel bad
22:10:20preglowstill only the old iaudios that are 16 meg ram?
22:10:45amiconnBut the main reason why I'm against enabling it by default is that it misbehaves if there are too many files
22:10:58preglowmisbehaves in what way?
22:11:00amiconnYes, M3, M5 and X5 (of the hdd targets)
22:11:26amiconnFails to initialise, and tries a foreground scan on each boot over and over
22:11:45amiconnAlso, if it's near its limit, but not above, the scan takes ages
22:12:46*amiconn thinks dircache should auto-disable itself if it fails because of hitting the buffer limit
22:13:31preglowwhile showing a "press key to remove" splash message
22:14:22gevaertsSo the storage_*() consensus is to just broadcast?
22:14:38preglowwell, i'd say that is ok for now
22:15:19gevaertsok. That will allow me to get on with it :)
22:15:26preglowexactly :)
22:17:35funmanwhat should I do if I have a patch which makes a bad driver become a bit less bad: commit it, or post it on flyspray for more testing ? (specifically for the sansa ams)
22:17:39*n1s listens to the friggin radio \o/
22:17:46gevaertsHow's the D2 SD driver coming along?
22:17:48funmann1s: :)
22:17:49BigBambin1s: horray!
22:18:00preglowgevaerts: hacking on it now, i'm gonna try really hard to have something working soon
22:18:06BigBambin1s: or even hurray. Either way, congrats :)
22:18:07preglowthere is something very wrong going on, tho
22:18:18preglowcard switch states as it should
22:18:34preglown1s: what target? :V
22:18:49n1spreglow: the beast
22:18:53preglown1s: sweet
22:19:30*preglow is looking forward to hacking on dab
22:19:37funmann1s: the beast has a si4700, right ? do you know the differences with the 4701 / 4702 (inside sansa ams) ? (I suppose you have read some docs about it)
22:19:45preglowi think i need to start on that after sd, if only to have some work ahead that interests me...
22:20:13gevaertsfunman: how much is a bit, and how confident are you that it won't break other ports?
22:21:11funmangevaerts: a bit is a lot, but not perfect though; and the diff is in target-specific code
22:21:15n1sfunman: not sure, if i read the chip id correctly it's actually a 4701 in the beast (difference between 4700 and 4701 is RDS)
22:21:31n1sso i think this one can do RDS
22:21:39gevaertsfunman: commit it
22:21:40BigBambin1s: really?
22:21:41preglowfunman: i don't see any reason not to commit it if it makes something bad a little better
22:21:46BigBambin1s: That is awesome
22:22:09preglowfunman: unless it also does something apart from just fix that badness, of course
22:22:12funmann1s: very nice, good work ! (I hope you could at least listen to a good emission)
22:22:16*gevaerts wants RDS
22:22:21*preglow wants dab
22:22:28n1sBigBambi: no promises, but i _think_ so
22:22:31*gevaerts hands preglow a DAX ;)
22:22:42funmanok thanks for the hints, there is still a major problem: I added more mess to the error codes reporting; I think I should fix that before committing
22:22:57gevaertsThats different of course :)
22:23:14funmantoo bad I had already used -666 and -42 :(
22:23:16preglowgevaerts: oh, d2 has dab, i just need to make more basic stuff work first :&/
22:23:30gevaertspreglow: I know, but mine doesn't
22:23:39preglowbad choice then :)
22:24:49Zagorsaratoga3: yes, =0 silences the warning
22:25:13preglowsaratoga3: did you ever hear any more from that guy who wanted to port back wma to ffmpeg?
22:25:15 Quit {phoenix} (Remote closed the connection)
22:39:09funmanI constantly hit the panic in drivers/fat.c:1530 - I have absolutely no knowledge of FAT, does this condition always mean that the filesystem is invalid/corrupted ?
22:39:24 Join kronflux [0] (
22:39:53funmannote: I hit it when trying to 'open' a music file in rockbox filebrowser, note that I have no decoder built in my rockbox firmware (maybe only libspeex for the voices)
22:40:11Zagorfunman: that is definitely a bad panic
22:40:32funmanah no I just hit it when opening a directory - something is definitely wrong
22:41:03funmanI suspected that not being friendly with other threads would help (i.e. the FIFO would have less chance of becoming full if we keep the CPU in the SD thread)
22:42:24funmanperhaps the solution is to use interrupts to read from the FIFO as soon as possible, but i'm not friendly (yet?) with the threads code; for example to use inactive sleep while waiting for the interrupt to happen, and resume after it happened
22:44:51preglowso you actually yield when inbetween issuing the read command and reading back the data you get?
22:46:09funmannot exactly, we yield() in the middle of reading the data from the FIFO, if it wasn't filled within the requested timeout
22:46:33funmanor more exactly, half-filled (we read only half by half to ensure the fifo doesn't overrun)
22:46:35preglowshould that ever take so long that that is necessary?
22:46:42preglowi don't think the portalplayer sd card driver does this
22:46:59funmanit does - i copied the logic bit per bit before thinking of one
22:47:54funmanit timeouts between each FIFO read after 0.5 second
22:48:28preglowwell, do you have sd dma available?
22:48:35preglowthat would probably be the nicest solution
22:49:01funmani 'just' need to understand what is dma and read the datasheets :)
22:49:13funmannote that the OF use dma for transfers
22:49:21pregloweverything should
22:49:24preglowit is the best way
22:49:28preglowbut not as easy as the trivial, of course
22:50:19funmani think i'll poste here a pastebin of my current diff, and not commit it until i read more about dma. i don't think anyone rely on a stabler SD to write code for sansa ams
22:51:33 Join bertrik [0] (
22:51:53 Join Thundercloud [0] (
22:52:16funmanif you need more stable SD reads (for quite reliable rockboox boot, but not reading music files) apply this :
22:54:29 Quit aarcane ("Leaving")
22:55:08*domonoky is still investigating why he only gets to the logo in the rockbox app on m200v4..
22:55:48funmandomonoky: have a try at this patch, several reads on the SD are made by rockbox when booting
22:56:42 Join bodymind [0] (
22:58:08*domonoky tries...
22:59:09 Quit n1s ()
22:59:11 Nick bodymind is now known as bodymind_ (
22:59:25funmanis there a reason for libmusepack not being built by 'make rocks' ?
23:00:00funmanMEMORYSIZE <= 2
23:00:11 Quit bodymind_ (Client Quit)
23:00:24 Join bodymind_ [0] (
23:00:28 Quit bodymind_ (Read error: 104 (Connection reset by peer))
23:00:43 Join bodymind [0] (
23:01:05funmanbut CODECLIBS in apps/codecs/Makefile is not conditional on MEMORYSIZE (unlike apps/codecs/SOURCES)
23:01:20 Join _lifeless [0] (n=lifeless@
23:02:07funmanand apps/plugins/Makefile
23:03:07domonokyit takes a loong time to boot, but it works.. :-)
23:03:39funmanso this is definitely not a fix, at best a temporary workaround
23:04:32bodymindhow can i calculate TIMER_FREQ for sansav2?
23:04:52 Quit domonoky (Read error: 104 (Connection reset by peer))
23:05:15 Quit Nibbler (Read error: 113 (No route to host))
23:05:42funmanbodymind: I guess by first understand what its meaning is
23:07:51bodymindyes u are right... maybe google :$
23:08:23 Quit petur ("reading time too")
23:09:40bodymindit's used for plugins.. is it for calculate time?
23:09:52bodymindhow can i disable the compilation of plugins?
23:10:33funmanuse only 'make bin' , or replace plugins="yes" by plugins="" in tools/configure, then re-run configure
23:11:45kugelfunman: the v1 sd driver is probably not the best to copy from, as it's known to have a serious bug (which is apparently only noticeable when using usb or doing big transfers to an external card)
23:12:14kugelfunman: you don't build codecs on purpose?
23:12:42funmankugel: it's also my only code reference for rockbox, and as far as i can see conforms to the SD specification
23:13:47funmanyes, I tried to build and use them as well; but rockbox crashes (due to the SD driver, but not only, also data aborts in apparently thread-related code). And I think we should not put our attention on the codecs yet :)
23:15:08 Join _Andrew_ [0] (
23:21:13 Join Schmogel [0] (
23:22:06soapLlorean, How should I go about removing copyright-infringing themes from the wiki? Make a list and submit it to the Swedes?
23:23:13LloreanAs far as I know, that's the only way to get the actual files completely removed.
23:23:20 Join moos [0] (
23:23:22soapOr is the idea of removing WPSs using Microsoft/Apple/CompanyX graphics still a subject which needs debated?
23:23:32kugelfunman: building them doesn't hurt though
23:23:42BagderI don't think it's much of a debate, mostly a case of identifying
23:24:24soapI'll work on a list of filenames and paths tonight for you. Does that work well? One lump sum message?
23:24:29kugelfunman: so you can to normal make and make zip etc
23:24:36Bagdersoap: sounds great
23:24:43funmankugel: the problem is more the numerous plugins not building
23:25:15kugelyou disable plugins in configure
23:25:22funmanadd_to_list_l() just crashed in thread.c:1197 :/
23:25:38kugelthey're lacking keymap. and I think plugins have even a lower priority than codecs
23:26:08kugeldeactivating plugins != deactivating codecs
23:27:39funmankugel: oh .. ok, I think I left plugins activated in configure, thinking that it was somehow required for some mystical reason
23:28:24kugelthey don't build for me
23:29:23funmankugel: but then I would have to do real work instead of shutting on/off my Clip 20 times in a row to see the mistify demo at least once again ? :)
23:36:14 Join Chronon [0] (
23:38:54 Join Zarggg [0] (
23:39:02 Quit bertrik (Remote closed the connection)
23:41:19 Join RoC_MasterMind [0] (
23:43:41 Join bertrik [0] (
23:43:45***Saving seen data "./dancer.seen"
23:43:49funmanI was chatting a bit with the famous sansa ams hacker fdinel, he told me he is experienced with DMA; now let's just motivate him ! :)
23:44:24 Quit RoC_MasterMind (Remote closed the connection)
23:45:10preglowthat's the way!
23:45:35 Quit tvelocity (Read error: 145 (Connection timed out))
23:46:21 Join ap0 [0] (
23:49:50 Quit funman ("leaving")
23:49:57preglownow, why the hell won't the sd card go into ready state
23:50:13 Quit Schmogel (Read error: 104 (Connection reset by peer))
23:56:07 Join bodymind_ [0] (
23:56:14 Quit bodymind (Read error: 110 (Connection timed out))
23:57:24preglowbecause i'm an asshat
23:58:16Bagdernow that's a word we don't use often enough here! ;-)
23:58:50 Quit bmbl ("Woah!")

Previous day | Next day