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

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

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

#rockbox log for 2007-11-28

00:00:20 Quit DM| (Read error: 110 (Connection timed out))
00:00:21 Join male [0] (
00:01:59 Join frawfraw [0] (
00:02:22frawfrawmy ipod's battery ran flat and now it won't recharge
00:02:26Nico_PjhMikeS: in queue_wait, execution doesn't seem to get past set_irq_level
00:02:30frawfrawi don't know what to do
00:02:38 Join DM| [0] (
00:02:54Nico_Pbut now I get a prefetch abort
00:03:27Nico_P(I added printf calls before and after the set_irq_level call)
00:04:00jhMikeSset_irq_level? ?? odd
00:04:21Nico_Pdoes the printf call that is right after it have a chance of working?
00:04:28PaulJamfrawfraw: have you already tried to insert the charger and then reset the ipod?
00:04:53 Part przemhb
00:05:08frawfrawpauljam: i have the charger inserted but nothing happens
00:05:32jhMikeSNico_P: perhaps find a way to dump an indicator without using printf directly in the routine?
00:05:33Lloreanfrawfraw: Try leaving it attached to the charger for several hours
00:05:55frawfrawboth: wait, do you mean the wall charger or the USB cable?
00:06:17Nico_PjhMikeS: how?
00:06:36Lloreanfrawfraw: A wall charger is best
00:06:43LloreanUSB cable might work.
00:07:15jhMikeStwo before, one after, they should match values. printf the values in the abort handler
00:08:49 Part LinusN
00:09:03frawfrawok, thanks. i guess I will see if some life comes back
00:09:07 Join DaCapn [0] (
00:09:48 Quit jgarvey ("Leaving")
00:12:23*jhMikeS notices crt0.S sets cpsr and not cpsr_c when changing modes
00:12:37 Quit frawfraw ()
00:13:16jhMikeSright at new_start
00:13:25jhMikeSnewstart...whatever :p
00:14:50jhMikeSand why does it need two instructions to set it (mov r0, #0xd3; msr cpsr, r0)??
00:15:04stripwaxanyone with an ipod nano able to test the pacbox patch with PACBOX_ON_COP enabled?
00:15:32Nico_PjhMikeS: yeah with the counters I think set_irq_level(HIGHEST_IRQ_LEVEL); is what causes the prefetch abort
00:17:13*Llorean learns that Rockbox is apparently a conspiracy.
00:17:25Nico_PLlorean: didn't you know? :p
00:17:26jhMikeSthe one after set_irq_level is one less?
00:17:29LloreanI and the other forum mods are in cahoots to keep things secret!
00:18:05Nico_PjhMikeS: yes, well actually I get "-" as display, but I tried incrementing the other counter again after the call and it still ahs the same value
00:19:13 Join hcs [0] (n=agashlin@rockbox/contributor/hcs)
00:19:32 Quit MethoS-- (Remote closed the connection)
00:19:53jhMikeSI find the code in crt0.S a bit odd myself.
00:20:47 Join Soap [0] (n=Soap@rockbox/staff/soap)
00:20:47jhMikeSI mean there must be some specific reason for doing it that way, no?
00:20:48 Nick billeniumZzZ is now known as Bam2550 (
00:21:03 Nick Bam2550 is now known as Billenium (
00:21:39Nico_PjhMikeS: that way as opposed to what other way?
00:22:13jhMikeSsetting all of cpsr instead of cpsr_c
00:22:51 Quit davina (Remote closed the connection)
00:24:08markunLlorean: I hope you are not going to waste too much time replying to this post:
00:26:50nanokmarkun: secret, progress sllow?
00:26:59nanokwell, i must say..
00:27:06Lloreanmarkun: I was going to reply, but I've decided against it
00:27:13markunLlorean: good
00:27:14LloreanThe e200R thread is a really, REALLY bad example
00:27:18 Join psycho_maniac [0] (
00:27:23LloreanHundreds of, frankly, quite stupid suggestions for what we could try
00:27:36LloreanPeople with nearly no technical knowledge repeating the same thing forty or so other people had already said
00:27:41psycho_maniacanybody have a ipod video with wall charger?
00:28:28psycho_maniaci think i found a bug
00:28:56nanoki remember i had to choose a rockbox-enabled dap, and ipods being out of the question (personal reasons, no matter), i decided for the e200 instead of the c200. i am almost sorry now, one of the main reasons was the sd slot support was not here yet for the c200, it only took a few weeks untill that was fixed..
00:29:33markunnanok: is the c200 better?
00:29:46LloreanIt's smaller
00:29:55stripwaxpsycho_maniac - I do, but I hope there's more detail to the bug description than that ;-)
00:29:59LloreanWith a smaller screen, no wheel (a nicer button interface, I feel), and less storage
00:30:06 Join linuxstb_ [0] (
00:30:22markunLlorean: I'm getting a bit pissed from reading his post again, by I guess that was his goal, to get a discussion going.
00:31:11stripwaxWhen a thread is closed because 'the answer can be found in the manual', does it get eliminiated from search results?
00:31:15scorche|wi sounds like a troll, rather than a discussion
00:31:18*Soap will be in Germany first two weeks of December. If anyone wants something from the USA I can bring it.
00:31:43scorche|w(say me)
00:32:08Lloreanstripwax: Not yet. I'm looking for a way to do that.
00:32:17markunSoap: which cities will you visit?
00:32:19stripwaxLlorean - great
00:32:21LloreanI really want a "no-index" flag
00:32:30nanokmarkun: not better or worse as such. it is just different, mainly, as Llorean pointed out, it is smaller (considerably so), and lighter
00:32:33Soapmarkun, will be working in Hamburg, should have the weekends off.
00:32:35 Quit Psy-Krow ("don't taze me bro!")
00:32:38LloreanBut at least telling people to read the manual first helps *reduce* the number of threads they'll start
00:32:50psycho_maniacstripwax: im trying to figure out the bug first.
00:32:57stripwaxLlorean - exactly (and then closing the thread, presumably)
00:33:15psycho_maniaci know what it does but it has to be in some setting. when the backlight turns off it will not come back on.
00:33:32stripwaxpsycho_maniac - is that related to the wall charger at all?
00:33:47Lloreanstripwax: We try not to lock threads, just so the person can come back and say "I read this part of the manual, but don't understand what Blah means"
00:33:57Lloreanstripwax: Believe it or not, we do try to consider people. :)
00:34:06nanoksome people should really read the "ask questions the smart way" guide. i would go as far as saying it should be a requirement for being allowed to use "the internet" :)
00:34:11psycho_maniaci just found out it is not. i thought it was though.
00:34:19stripwaxLlorean - true :) I didn't mean "and then closing the thread immediately after saying to look in the manual" :)
00:34:28Lloreanstripwax: Alright, just making sure.
00:34:37stripwaxactually you're right, it doesn't make sense to lock the thread at all, I suppose
00:34:44LloreanWe really could be a lot more totalitarian than we are, but I think we're keeping the balance about where it has to be.
00:35:05Nico_PjhMikeS: any suggestions?
00:35:07 Join Psy-Krow [0] (
00:35:54 Quit Zagor ("Client exiting")
00:36:03jhMikeSNico_P: none yet except to try making crt0.S more "normal" :)
00:36:25Nico_Phaha I have no idea what "normal" is for that ;)
00:36:38n1smaybe we should raise funds to get jhMikeS an S? ;)
00:36:41jhMikeSuse cpsr_c instead of all of cpsr
00:36:59*stripwax rebuilds sliding_puzzle for h120,rec,ipod5g, just to be sure..
00:37:15 Quit n17ikh|Lappy (Read error: 104 (Connection reset by peer))
00:37:16*jhMikeS has to get at least 3 new kinds of devices to be on the up 'n' up. :) DM320, TCC-based, and iMX-31 based.
00:37:25stripwaxJust noticed that a lot of targets don't actually show the key help in sliding_puzzle so I'll add that at the same time
00:37:53LloreanjhMikeS: Might see if refurbdepot still has any.
00:37:54 Quit iamben_ (Read error: 104 (Connection reset by peer))
00:37:54 Join n17ikh|Lappy [0] (
00:38:32jhMikeSLlorean: that's sells factory refurb stuff?
00:38:35psycho_maniachas there been any changes in the lcd settings lately?
00:38:36 Quit linuxstb (Read error: 113 (No route to host))
00:39:31n1spsycho_maniac: not lcd settings but backlight, yeah
00:39:44LloreanjhMikeS: I don't know who does the refurbishing
00:39:52LloreanThey also had an S60 last time I looked, but that was the first google result
00:39:55psycho_maniacerr thats what i ment to say.
00:40:20 Join homielowe [0] (
00:40:25n1spsycho_maniac: update to the latest build and reset your settings and then check if it still happens
00:41:30psycho_maniacthats what im going to do.
00:42:43toffe82jhMikeS: you want a gigabeat S ?
00:43:06 Join iamben [0] (
00:43:22psycho_maniaci think it has to do with a certin setting because when i load my custom cfg file the backlight will not turn on. well see when i get the newest build.
00:43:30 Join darkapostrophe [0] (
00:43:34jhMikeStoffe82: perhaps. sort of on the cache crunch atm. whacha got?
00:43:40jhMikeScash even
00:43:49 Join barrywardell [0] (
00:44:08psycho_maniacwow thats really cheap for a S and i just bought a F
00:44:18 Join Traveler7 [0] (
00:44:26jhMikeScache != cash but a cash cache is good to have :p
00:44:53n1spsycho_maniac: the meaning of the value 0 for the backlight settings changed it now means "off" it used to mean "on" (always) I guessing that is why your settings cause it to not come on
00:45:22Traveler7hehe, was playing with rockbox today, wouldn't let me run my plugin while playing music
00:45:41 Join J3TC- [0] (
00:45:43Traveler7did let me run chopper though, but it caused my music to play in fast forward
00:45:50 Join mokkurkalve [0] (
00:45:52jhMikeSI've seen S models going for about $159 on ebay when I checks awhile ago
00:46:06toffe82jhMikeS: I got a s30 for 60$ from somebody who tell me that it was not working properly, I test it and it seems to work fine, I have to reload the software on it, if you want it
00:46:07*jhMikeS is typing especially badly today
00:46:43jhMikeStoffe82: hmmm...what's it missing as far as the accessories?
00:46:44 Quit animeloe ("Leaving")
00:47:27toffe82video cable, usb and headphones
00:47:33psycho_maniacis that for every backlight setting?
00:48:01jhMikeStoffe82: it's just a non-proprietary USB cable?
00:48:11toffe82standard usb
00:48:17toffe82the same as the F
00:48:20jhMikeSwhat's the video cable type?
00:48:43toffe82jack 4 point to video and sound
00:49:08toffe82look the diagram on the wiki, it is a standard cable I think
00:50:36n1spsycho_maniac: there are only two timeouts, one while the charger is connected and one while it's not but as i already said a reset of the settings should fix it, didn't it work?
00:50:37jhMikeStoffe82: check PM
00:50:41*nanok is starting to seriously consider these toshiba toys..
00:50:59nanoki hate the non standard usb "dock connector" on my sansa :(
00:51:23 Quit Dark_Apostrophe (Connection reset by peer)
00:51:23 Nick darkapostrophe is now known as Dark_Apostrophe (
00:51:35n1sit also has a seperate charger connector so no need to charge over usb :) (I use my h300 charger for it)
00:53:40 Join animeloe [0] (n=animeloe@unaffiliated/animeloe)
00:53:42psycho_maniaci just resaved my backlight settings and i think that did the truck
00:54:20mokkurkalveanybody in the knowing of e200 here now? I have trouble.
00:54:30psycho_maniacdont the S come in black?
00:55:09krazykitmokkurkalve, ask your question and somebody can usually help. no need to ask to ask a question :)
00:55:39 Quit n1s ()
00:55:54nanokmokkurkalve: just ask
00:57:43advcomp2019should or should not the e200 reboot when you plug in the usb cable even tho there was a commit that says "Enable reboot on USB-insert for e200, as it works when done c200-style"
00:58:31mokkurkalveI've posted a lengthy explanation in the sansa part on the forum under my older nick Mr_Rabid_Teddybear. Anyway. After using sansapatcher under Linux i couldn't boot either firmware. And after using "Recovery Mode" as wiki prescribes I get "load main image fail". And that's how far I get...
01:00:00nanokadvcomp2019: i think not. no reboot. only change boot order if the cable is plugged while the sansa is off (to have the of first)
01:00:46 Quit ender` (" If I had only finished this sentence,")
01:00:50 Quit Febs ("ChatZilla 0.9.79 [Firefox]")
01:01:05nanokadvcomp2019: it wouldn't make much sense to reboot it anyway, as the user might want to only charge it (and now, thanks to zagor, that is allready possible)
01:01:19 Quit animeloe ("Leaving")
01:02:10advcomp2019nanok, o ok.. i think i might have misunderstood that commit then
01:02:15mokkurkalveso basically I can still get into the "Recovery Mode" so I guess there's still hope. I think I might need a differnt .mi4 file. For all practical purposes my new player are bricked right now. Hope there's a way out....
01:02:24psycho_maniaci guess my problem was not really there. i just resaved my backlight settings and they work now.
01:02:39 Quit Traveler7 ("Java user signed off")
01:03:54nanokadvcomp2019: to be honest, i don;t really get what that commit might be about. but maybe i don;t have a current enough build. how old is that commit?
01:04:14nanokdarn ;
01:04:39Lloreanadvcomp2019: No, it should reboot when the cable is inserted, but it doesn't work reliably yet
01:05:05linuxstb_mokkurkalve: Have you tried other firmwares? There is a collection here -
01:05:41nanokLlorean: may i ask what the reason for that is? or is it only intended (and needed) for the upcoming usb storage part?
01:06:10Lloreannanok: Well, Rockbox has no USB. For the moment, it should reboot into the original firmware, and use it's USB. But the reboot doesn't happen reliably
01:06:18advcomp2019Llorean, thanks because it does not work on my e280r that is why i asked
01:06:55barrywardellmokkurkalve: I suggest getting the 01.02.18F firmware from the site linuxstb_ gave a link to and putting the mi4 from there on in recovery mode
01:06:59nanokLlorean: it does have the ability to use usb power for charging while running rb, which is, imho, essential
01:08:38mokkurkalveI will try that, thanks. Report back soon (knocks on wood)
01:08:57pixelmananok: and you still can use this, hold "select" when inserting the USB cable and you'll get it into charging mode and prevent the reboot into original firmware
01:09:25pixelmathis is similar to how it is done on other targets
01:09:39Lloreannanok: And by holding down a button while plugging in the USB cable you can prevent the reboot. Should be "Select"
01:10:14pixelmatwice ;)
01:10:22advcomp2019this is odd.. now it is not charging my sansa in rockbox.. it is slowly draining it
01:10:25 Join donutman25 [0] (n=chatzill@
01:10:40preglowjhMikeS: what, you're going for a gigabeast s too now? :P
01:10:56preglowgigabeast, haha, nice typo, thom
01:11:04preglowi think i'll stick with it
01:11:12LloreanI like it.
01:11:13nanokpixelma: Llorean uhm, okay, that's why i was asking for what it's intended earlyer. it does work okay if you press select, it makes sense in this case, agreed
01:11:30jhMikeSeven gigabeastier than the F :)
01:11:33preglowLlorean: well, if there's one dap that does qualify for the name, i think it's it
01:11:36Lloreannanok: Once the USB stack is working, Select will probably still prevent connection, but instead of rebooting it'll just use Rockbox's USB mode
01:11:45Lloreanpreglow: Absolutely
01:11:57preglowjhMikeS: do you know the timings for that fpu thing?
01:12:19nanokLlorean: i see, i think that makes sense also. i personally like being able to "charge only"
01:12:33jhMikeSnot a clue yet. I suppose the datasheet is available to look at :)
01:12:45*nanok is wondering if that is really enough to save my ass at work in case of a security audit :-P
01:14:11***Saving seen data "./dancer.seen"
01:14:56 Join animeloe [0] (n=animeloe@unaffiliated/animeloe)
01:15:18*jhMikeS just needs to get a clue on the pp5002 cache thing for the moment
01:16:30 Quit psycho_maniac (" HydraIRC -> <- Po-ta-to, boil em, mash em, stick em in a stew.")
01:16:39 Quit animeloe (Client Quit)
01:18:09preglowjhMikeS: any new clues on what's going wrong?
01:18:29 Quit scorche|w ("CGI:IRC (Ping timeout)")
01:20:46jhMikeSI know if run with cache disabled, the battery level is ok and no crash happens. Once things are initialized without incident, it all operates correctly. It's something during the init.
01:21:25mokkurkalveUnfortunatly that failed. Get "Load main image fail. Switch to Recovery mode". So. Are next step "Manufacturing Mode"?
01:21:55barrywardellno,not yet
01:22:10barrywardelldid you copy over the bootloader too? or just the mi4?
01:22:13jhMikeSbut the batt level is still messed up of course if the build happens to run
01:22:45mokkurkalvejust the PP5022.mi4
01:23:53mokkurkalvethere's a version.txt there from before, thats all there is. I only copied the .mi4 over
01:24:28 Quit Mouser_X (Success)
01:24:35barrywardelltry copying over both the bootloader and the mi4
01:24:43barrywardellmake sure it says that it's upgrading
01:24:47*preglow wonders why the ground is shaking
01:25:29mokkurkalvewhere do I find the bootloader? I used sansapatcher before
01:26:02*linuxstb_ wonders why preglow's ground is shaking
01:26:18 Quit Robin0800 (" Want to be different? Try HydraIRC -> <-")
01:26:24preglowwish i knew
01:26:29preglowbut it's irrefutable
01:27:08hcspassing invisible giant?
01:27:18barrywardellmokkurkalve: there should be a .rom file in the zip you downloaded
01:27:46mokkurkalveThe wiki stresses to only copy .mi4 to the recovery partition
01:27:47preglowjhMikeS: we don't get too many of those around here
01:27:54Nico_PjhMikeS: you're getting an S! good news :)
01:28:08Nico_PI've changed cpsr to cpsr_c and it didn't change anything
01:28:11jhMikeSpreglow: In Michigan we don't either but they do happen every 5-10 years
01:28:49preglowsure, we get small ones
01:28:52preglowstopped now
01:29:01nanokjhMikeS: i initially missread (my internal bash completion running wild) 5-10 minutes
01:29:24mokkurkalveYes. There is also BL_SD_boardSupportSD.rom in the sevenzip. How safe is this?
01:29:33jhMikeSnanok: I'd hope not :)
01:29:40barrywardellmokkurkalve: copying the BL_SD_boardsupportedSD.rom over is fine too.
01:30:09jhMikeSpreglow: yeah, here too...I noticed it, others didn't. It was on the news last time so it wasn't a hallucination.
01:30:12 Join Chronon [0] (
01:30:13Nico_Plinuxstb_: I've been fiddling with the main binary and the problems seem to be in IRQ handling
01:30:23mokkurkalveI thought that file was for use when you go to "Manufacturing mode". OK
01:30:34barrywardellit can be used for that too
01:30:58mokkurkalveI'll try anything once....
01:31:26jhMikeSpreglow: cool thing was the house rocking back and forth and my desk lamp swaying at about 1HZ :)
01:31:39preglowhaha, leet
01:31:41preglowthis was more like 10hz
01:31:50 Join linuxstb__ [0] (
01:31:52 Nick linuxstb__ is now known as linuxstb (
01:32:12 Quit donutman25 ("ChatZilla 0.9.79 [Firefox]")
01:33:35mokkurkalveOk. Now my screen shows random noise..... Don't look good
01:33:54mokkurkalveActually it's patterned. not random
01:33:56 Join animeloe [0] (n=animeloe@unaffiliated/animeloe)
01:34:28advcomp2019with r15801 it will charge in rockbox, but now with r15839 it is not charging. i will still use OF for charging since it still buggy.
01:35:00barrywardellmokkurkalve: Random how?
01:35:04 Quit animeloe (Client Quit)
01:35:15barrywardelldid it say it was updating?
01:36:09jhMikeSNico_P: I don't know how that function should ever abort in a sense...looking at a disasm might tell where some memory is accessed around that point
01:36:28mokkurkalvei copied over the 2 files. Detached USB. Now I have patterns of white noise on my screen.
01:36:43Nico_PjhMikeS: I'll probably try that tomorrow
01:36:43jhMikeSusually I just add −−save-temps to GCCOPTS and force a recompile of the particular file
01:36:58barrywardellmokkurkalve: can you get back into recovery mode?
01:37:12mokkurkalveIt updates when detatced. did last time. now only noise
01:37:17 Quit linuxstb_ (Nick collision from services.)
01:37:44mokkurkalvebarrywardell: I must take out batttery to close it down, then well see
01:37:51nanokadvcomp2019: are you using linux or windows when charging (or what)
01:37:56barrywardellmokkurkalve: just hold power for 10 seconds to power off
01:38:19nanokmokkurkalve: or more like 15-20. just be patient ;)
01:38:28advcomp2019nanok, both
01:38:52nanoki also took my battery out for that reason, only after i found out there is a more ellegant way..
01:39:19 Part toffe82
01:39:22nanokadvcomp2019: ahm, okay. i am not sure how it works in windows, it seems it is a bit strange for now
01:39:22 Join animeloe [0] (n=animeloe@unaffiliated/animeloe)
01:39:59Nico_PjhMikeS: I did it, but what kind of file should I look for? I don't see anything but the .o files
01:40:02nanokadvcomp2019: btw, you have an "r" one, right? (i think there should be no difference in this case though)
01:40:09jhMikeSNico_P: things should run even if the core isn't put to sleep. m:robe does
01:40:13mokkurkalvePhew! Recovery mode back... what now?
01:40:26Nico_PjhMikeS: ie if core_sleep isn't implemented?
01:40:27jhMikeSNico_P: an .s file in the firmware directory
01:40:43advcomp2019nanok, yea i do have an e200r
01:40:44jhMikeSNico_P: yeah, it should run...but will just burn cycles in sleep_core
01:40:45Nico_Pah yes, found it
01:41:21barrywardellmokkurkalve: can you post the contents of version.txt file somewhere (eg.
01:42:12 Join jhulst [0] (n=jhulst@unaffiliated/jhulst)
01:43:12 Quit animeloe (Client Quit)
01:44:24 Quit stripwax ("Miranda IM! Smaller, Faster, Easier.")
01:44:35Nico_PjhMikeS: there is a printf call and some incrementations added compared to svn
01:44:53preglowjhMikeS: for some weird-ass reason, i'm thinking of making a float path through dsp as well, for targets like gigabeat s, but i don't even know if it'd be worth it
01:45:35 Quit obo ("KVIrc 3.2.4 Anomalies")
01:46:00 Join animeloe [0] (n=animeloe@unaffiliated/animeloe)
01:46:12preglowjhMikeS: about the only good thing about it i can think of is it'd make some things easier, plus make sure we never have to worry about overflow again
01:47:05jhMikeSpreglow: but that has saturation too! :)
01:47:11barrywardellmokkurkalve: are you still unable to load the of?
01:47:30jhMikeSit's arm5 with a Q bit afaik
01:47:31*preglow suddenly remembers he should finish his eq plotter
01:47:44preglowthat'd definitely be easier with floats...
01:48:31jhMikeSwhy not look up the timings in the reference material?
01:49:45hcsHas anyone else noticed that SPC tag information seems to become corrupted a few seconds before the end of the track?
01:50:08jhMikeShcs: yep, the metadata seems to get released before it's finished using it
01:50:27 Quit animeloe (Client Quit)
01:50:48Nico_Pstrange, it should be in a static struct
01:51:11Nico_Phcs: maybe file a bug report
01:51:18mokkurkalvebarrywardell: Yes. same message "load main image fail".
01:51:29preglowjhMikeS: couldn't find it :/
01:51:50Nico_PjhMikeS: any comments on that asm bit I posted?
01:51:55preglowgod damn, this piece of junk is getting sluggish
01:52:17jhMikeSNico_P: I take it line 17 there is your counter?
01:53:09barrywardellmokkurkalve: weird. maybe try a different firmware version? I use 1.01.11A without any problems. you probably have to update the bootlaoder too
01:53:24Nico_PjhMikeS: proably, yes
01:53:56jhMikeSif not, that's very weird :) but I'm seeing how set_irq_level should be doing anything bad :\
01:53:59mokkurkalvebarrywardell: In recovery mode?
01:54:30barrywardellyeah, since you can't get into the of to use normal usb mode
01:55:39Nico_PjhMikeS: I should've pasted that .L74 part
01:56:03 Join animeloe [0] (n=animeloe@unaffiliated/animeloe)
01:57:10 Quit animeloe (Client Quit)
01:57:24Nico_PI think it's bed time now... I can't think anymore
01:57:31jhMikeSlooks like the printf call to me
01:58:12 Quit linuxstb (Read error: 113 (No route to host))
01:58:13Nico_P.L74 has printf, the counters, the core var
01:59:22jhMikeScore var shouldn't exist on single core
01:59:46hcsNico_P, jhMikeS: FS # 8242 for the SPC metadata thing
01:59:46 Quit ompaul (Client Quit)
02:00:06*preglow wonders if the gigabeast s tv out is any good
02:00:13preglowi did it again!!
02:00:39mokkurkalvebarrywardell: Unfortunately only same message as before with 1.01.11A
02:00:46Nico_PjhMikeS: a bit more complete:
02:00:53jhMikeSI'll certainly attempt to find out if it is somehow
02:00:55preglowwell, at least imx31 is well documented
02:00:57Nico_Phcs: thanks
02:01:10jhMikeSI think the TV out is a separate IC though
02:01:33jhMikeSI'm not sure about the docs on that one
02:01:42Nico_PjhMikeS: is printf .L74 or .L74+4?
02:01:42preglowjhMikeS: if it does decent tv out, i'll just have to get one
02:01:46 Quit Chronon ("Work-->Home")
02:01:48preglowjhMikeS: bastard's perfect for decoding video too
02:02:05advcomp2019mokkurkalve, could you try one of the e200R firmwares to see if you changed it to an e200R by accident
02:02:09preglowif not perfect, then at least thorougly excellent
02:02:18jhMikeSNico_P: should be L74+4 since bx r3 is used
02:02:50barrywardellmokkurkalve: it definitely says updating main image when you copy over the mi4?
02:02:55 Join animeloe [0] (n=animeloe@unaffiliated/animeloe)
02:02:56Nico_Phmm yeah
02:02:59preglow8 state pipeline
02:03:11jhMikeSpreglow: the DM320 would be too and I'd love to implement libmpeg2 as a hardware decoder but with the same API
02:03:31mokkurkalvebarrywardell: it only says updating finished
02:03:46preglowjhMikeS: on the dsp, i take it...
02:03:49mokkurkalvewhen detaching usb
02:04:07jhMikeSI think it's a specialized hardware accelerator for video/jpeg
02:04:30 Quit animeloe (Client Quit)
02:04:36 Join goffa [0] (n=goffa@
02:04:43preglowjhMikeS: hmm, the vfp doesn't have the best latency instructions ever
02:04:52preglowjhMikeS: even an add is five cycles, but throughput is nice
02:04:53barrywardellmokkurkalve: what happens if you copy over a the mi4 and call it firmware.mi4 instead?
02:05:05preglowjhMikeS: as a matter of fact, an add is 9 cycles...
02:05:19mokkurkalvebarrywardell: I'm not sure if it actually does anything. Or if it just says "updating finished" everytime I detach usb
02:05:24jhMikeSbut that's across the entire pipeline?
02:05:27 Join animeloe [0] (n=animeloe@unaffiliated/animeloe)
02:05:49mokkurkalvebarrywardell: just the .mi4 this time or .rom also?
02:05:55Nico_PI'm off to bed now, gnight all
02:06:04barrywardelljust the mi4
02:06:09 Quit Nico_P (Remote closed the connection)
02:06:21preglowjhMikeS: i assume, yes, throughput is 1 cycle, so you should be able to do tons of them in parallel
02:06:42preglowjhMikeS: but 9 cycles is a lot if you're dealing with calculations with a lot of inter-dependency
02:08:42jhMikeSwhatever can keep throughput going for emac should be similarly doable though, no?
02:08:51 Join saratoga [0] (i=98039da7@gateway/web/cgi-irc/
02:09:04 Quit J3TC- (".•«UPP»•.")
02:09:16preglowjhMikeS: depends, but i expect so
02:09:29preglowjhMikeS: armv6 core... i'm going to need one of these bastards
02:09:49preglowjhMikeS: too bad there'll be little incentive to use the fancy stuff for optimizing :)
02:10:27mokkurkalvebarrywardell: It sait it was reading and updating. But the result remains the same...
02:11:09saratogahow is the S port coming along these days? i saw theres finally a good way to run code on it
02:11:26preglowsaratoga: still not running rockbox, but i see a decent amount of devs are getting them now
02:11:33XavierGrdamn you all with the gigabeat S discussion now I am itching to buy one :p
02:11:42XavierGrI can't justify to buy yet another DAP!
02:12:00Lloreansaratoga: Loading Rockbox and absolutely immediately freezing with the Rockbox splash on screen.
02:12:22preglowXavierGr: i can justify bying that one...
02:12:39preglowanyone have any gigabeat s on hand?
02:12:40*scorche wonders if XavierGr will the 6th person to get an S from him =P
02:12:46 Quit hcs ("Leaving.")
02:12:50preglowscorche: you got some?
02:13:06preglowman, i'm starting to sound like i'm buying drugs
02:13:27scorchenay...tons have passed through my hands though (i receive and re-send abroad for a good many rockboxers)
02:13:29 Join J3TC- [0] (
02:13:36barrywardellmokkurkalve: did the of ever work for you?
02:13:59preglowhow much are they in general any way?
02:13:59XavierGrI am currently watching some Ses on ebay
02:14:21jhMikeSpreglow: faster is better anyway...then really fancy things can be implemented since cycles actually do useful work
02:14:26scorchepreglow: all but 1 of them came from refurbdepot
02:14:35XavierGrwhat type is the gigabeat S HD?
02:14:41XavierGrnormal ATA or ZIF?
02:14:42preglowjhMikeS: the vfp looks quite neat, it can even flush denormals to zero
02:14:52mokkurkalvebarrywardell: before I ran sansapatcher I was in the Original Firmware, yes, if that's what you mean...
02:15:25 Quit goffa_ (Read error: 110 (Connection timed out))
02:17:41barrywardellmokkurkalve: the only other thing I can think of is using e200tool recover. then copying over a new bootloader rom and mi4 while in that recovery mode
02:17:48LloreanXavierGr: ZIF iirc
02:18:54mokkurkalvebarrywardell: Is that entering "manufacturing mode" as described in the wiki?
02:19:08XavierGrLlorean: ah damn, I wanted to bid on a faulty one, I could use the disk for my H100 if it was unusable
02:19:21XavierGrthough ZIF can be converted to ATA...
02:19:52XavierGryup looks like a ZIF to me
02:20:26 Quit Jon-Kha (
02:20:26 Quit Slasheri (
02:20:26 Quit Hadaka (
02:21:23NJoinJon-Kha [0] (
02:21:23NJoinHadaka [0] (
02:21:23NJoinSlasheri [0] (i=miipekk@rockbox/developer/Slasheri)
02:22:25saratogaI'll probably pick up an S once rockbox is semifunctional on it
02:22:32saratogait simply looks like too much fun to work on
02:22:53saratogathat or a TMS target, since their DSP sounds interesting as well
02:22:56 Join My_Sic [0] (i=4c449196@gateway/web/cgi-irc/
02:23:15My_SicHi everybody
02:23:20barrywardellmokkurkalve: yeah
02:23:20 Quit animeloe ("Leaving")
02:24:13preglowi'd love to pick up an s and help on the port
02:24:15preglowbut we'll see
02:24:21preglowfirst, i need cash, that's going to help
02:25:14My_Sicit is possible to display lyrics in rockbox ? in id3 tag or in extern txt files ?
02:25:29saratogadoes the S have any DSP ability or would it be a plain ARM target as far as codecs are concerned?
02:25:34mokkurkalvebarrywardell: Think I'll wait till tomorrow doing that. Bit tired now... Thanks for all help so far!
02:25:45 Join animeloe [0] (n=animeloe@unaffiliated/animeloe)
02:26:23barrywardellok, good luck
02:27:17jhMikeSpreglow: what does that help anyway? it can only do calcs on normalized numbers in hardware?
02:27:48 Nick fxb is now known as fxb__ (
02:28:23preglowjhMikeS: that's what i'm saying, just treating them like zero ensures we're never bogged down with them
02:28:52preglowjhMikeS: any kind of filtering will sooner or later produce denormals
02:30:28 Quit My_Sic ("CGI:IRC (Ping timeout)")
02:30:46 Join My_Sic [0] (i=4c449196@gateway/web/cgi-irc/
02:31:00jhMikeSI suppose a simple exponential decay would do that
02:31:21saratogadoes it do any sort of hardware fixed point ops?
02:31:36preglowsaratoga: well, it has all the usual arm stuff, plus more
02:31:45preglowsaratoga: it even has saturating fixed point math and simd integer operations
02:32:11preglowjhMikeS: the x86 fpu can't switch off denormal handling, so you have to resort to all kinds of tricks to avoid them
02:32:24preglowjhMikeS: a pentium 4 will more or less just die if it sees a stream of them
02:32:40preglowaltogether nicer to just not worry about them
02:33:04saratogapreglow: but no way to do a Q15.16 fixed point multiply in one op?
02:33:10saratogai really want a CPU that can do that
02:33:17saratogawouldn't even be that hard
02:33:57mokkurkalvebarrywardell: I solved the problem! I took an old laptop I have here running Debian. It only have USB 1.1. And copied the files over there. And recovery worked. Think the problem is that USB 2.0 controllers on my main PCs MoBo must be going west...
02:34:18preglowsaratoga: what, arm can, if you don't count the shifting back to Q15.16
02:35:04preglowbut then again, so can coldfire...
02:35:46 Quit My_Sic ("CGI:IRC (Ping timeout)")
02:35:46saratogabut thats 3 ops in general
02:36:01mokkurkalveDang! It was USB troubles with my main comp all the way....! Guess sansapatcher will work if I run it on the old laptop then...
02:36:08saratogacould easily be done in one op if they'd put a shifter on the outside of the ALU
02:36:37saratogawould only take a handful of transistors for the shifter, and a few more to handle setting where the decimal point goes
02:37:51jhMikeSpreglow: I thought that that hardware could handle them
02:38:04preglowjhMikeS: handle what?
02:38:15preglowjhMikeS: it handles them fine, but waaaay slower
02:38:35preglowjhMikeS: sse can handle denormals in several ways, including flush to zero, but the fpu has no choice but to process them the compliant way
02:39:03 Join My_Sic [0] (i=4c449196@gateway/web/cgi-irc/
02:42:11 Part pixelma
02:44:02 Join goffa_ [0] (n=goffa@
02:44:31 Quit My_Sic ("CGI:IRC (Ping timeout)")
02:45:29 Quit TMM (Remote closed the connection)
02:52:58 Join webguest86 [0] (i=450e16de@gateway/web/cgi-irc/
02:54:17 Quit goffa (Read error: 110 (Connection timed out))
02:56:02 Quit DerPapst ("So Long And Thanks For All The Fish!")
02:57:09 Join hcs [0] (n=agashlin@rockbox/contributor/hcs)
02:57:25 Join spence_ [0] (
02:59:29webguest86Am a green newbie so excuse me! I installed the "current" build on my Sansa e250, by unzipping the file - got a .rockbox folder and the usual contents. The main menu has a "Files" submenu. From there I see <microSD>, Photo, Record, Tmp, VIDEO, root, and Version. If I select <microSD> I can navigate the folders on the SD drive. I assume that if I select root I can do the same on the main drive - but instea
03:00:23scorchecut off at "but instea"
03:00:46webguest86but instead it just starts playing a selection on the SD drive. What's wrong
03:01:06scorchewell, "Files" is browsing the root of the device
03:01:16webguest86The menu structure doesn't seem to match at all that in the PDF
03:02:07webguest86That's reasonable - how do I browse the files on the main drive?
03:02:32LloreanYou already are.
03:02:48LloreanAs soon as you select "Files" you're seeing the main drive
03:02:51webguest86Are you saying nothing is there?
03:02:51scorche" Photo, Record, Tmp, VIDEO" and such are folders on the device itself
03:03:12scorcheplaced there by the original firmware
03:03:23webguest86I had a Music folder that contained numerous subfolders containing tracks
03:03:39saratogaperhaps it is hidden?
03:03:45LloreanOn the e200 series, the Music folder is automatically hidden by the original firmware
03:03:46preglowwhat flavours does the gigabeat s come in? 30 and 60?
03:03:48preglow30 seems most common
03:03:53scorchepreglow: yes
03:04:06krazykitthat folder is marked as "hidden" by the OF, so you need to either use the database, set your file browsing mode to "all", or use a different folder
03:04:18scorcheactually, only 1 out of the 5 i sent was a 30 too :P
03:04:33webguest86krazykit: thanks I'll try one of those
03:04:45preglowmmmmmm... nice low dollar
03:04:45PaulJamis there a special reason why hidden folders aren't shown in the suppoted view? this seems to confuse many users.
03:05:15saratogai think its mostly to avoid them seeing the .rockbox folder and such
03:05:21scorchePaulJam: showing hidden folders confuse them too...specifically seeing the .rockbox folder
03:05:21LloreanPaulJam: Because honestly, people shouldn't be hiding their "Music" folder
03:05:29saratogapeople only get confused because of the sandisk firmware, which won't be a problem once we get USB
03:06:14PaulJamcan't you treat folders with a dot in front differnt than filders with the "hidden" attribute?
03:06:38mokkurkalveThen I have Rockbox allright on my new e280. Dang! Well, I need a new MoBo on my PC. USB corruption all the way... It's a bit embarrasing... Well, thanks all!
03:06:57LloreanPaulJam: A dot in front is "hidden" in Linux terms, the hidden attribute is hidden in FAT32 terms. We really shouldn't differentiate between hiddens.
03:07:00webguest86krazykit: i tried setting the file view mode to "all", and it WORKED. Thanks.
03:07:12LloreanPaulJam: And again, under what conditions would a normal user "Hide" his music folder?
03:07:23LloreanIt's only the problematic design of the e200 firmware.
03:08:26PaulJamwell, for example the sansa firmware hides the music folder without the user interacting.
03:09:02LloreanPaulJam: Yes, but the user doesn't have to put their songs in the MUSIC folder that the e200 firmware uses
03:09:06LloreanIt'll even find music in other folders
03:09:35LloreanTheir own file browser will show them that the folder's hidden while they're putting music in it...
03:09:56 Join psycho_maniac [0] (
03:10:42 Quit goffa_ (Read error: 110 (Connection timed out))
03:11:00PaulJamok, i thought the sansa uses some itunes-like software where the user doesn't see that the folder is hidden.
03:12:10LloreanThey can use an MTP transfer program too, if they want to.
03:12:22LloreanBut in MSC mode, you'll see the folder is hidden
03:12:29LloreanAnd can unhide it, at least until the next time you boot the OF
03:14:15***Saving seen data "./dancer.seen"
03:17:56 Quit webguest86 ("CGI:IRC (EOF)")
03:19:09 Quit PaulJam (".")
03:29:09barrywardellmokkurkalve: good to hear you have it working now
03:31:03 Join karashata [0] (n=karashat@
03:31:40 Quit Thundercloud (Remote closed the connection)
03:31:48mokkurkalvebarrywardell: But a bit embarrased over raising a stink over such an issue, when it boiled down to PC hardware trouble only... ;D Well, thanks again!
03:32:55 Quit saratoga ("CGI:IRC (EOF)")
03:38:45 Quit Billenium ("Ex-Chat")
03:39:48 Join nekulturniy [0] (
03:44:55 Quit jhulst ("Konversation terminated!")
03:47:37 Join billenium [0] (
03:49:11billeniumDoes rockbox use less battery life? because my friend and I both had ipod nanos 1st gen and they both died. So we plugged them into my computer. They both had the "Please wait very low battery" and it stayed like that for 30 minutes. So we both took ours out and did a reboot. Mine opened up to Rockbox and i quickly put his in. His did nothing.
03:49:50billeniummine opened up to rockbox and i quickly put mine in*
03:52:56hcsso why exactly is there "-Skipped 766 measurements from 00:03:45 to 04:12:26-" in battery bench?
03:55:05 Quit OlivierBorowski (Remote closed the connection)
03:55:28 Join Rob222241 [0] (
04:00:52 Quit nekulturniy (Read error: 110 (Connection timed out))
04:05:16 Quit barrywardell ()
04:12:56 Join Febs [0] (
04:13:08 Join goffa [0] (n=goffa@
04:13:36 Quit Rob2222 (Read error: 110 (Connection timed out))
04:19:57 Join jhulst [0] (n=jhulst@unaffiliated/jhulst)
04:21:51 Join spliifdown [0] (
04:22:31spliifdownjust wondering if i should bother waiting for a 2nd gen nano hack to come out
04:22:41spliifdownor just splash out on the touch
04:23:42billeniumspliifdown: dont hold your breath
04:24:01billeniumspliifdown: I think one of the reasons they released the 2nd generating was because they didnt want you changing the firmware
04:25:23spliifdownoh orite
04:25:29SoapI wouldn't get paranoid like that - but the 2nd gen nano will not be supported for a long time.
04:25:41spliifdownits tottallly eencrypted
04:25:50SoapPromises aren't usually made about future ports, but that is a pretty safe promise on my part.
04:26:28spliifdownwhat programming language does it use the 2nd gen i mean
04:27:17spliifdownanyone have a clue
04:27:42 Quit goffa (Remote closed the connection)
04:27:45 Join goffa [0] (n=goffa@
04:28:32psycho_maniaccan we promise a nano 2nd gen in 50 years? ;)
04:28:49psycho_maniacs/nano 2nd gen/nano 2nd gen port
04:28:54Soapcould be FORTRAN for all anyone knows, or all it matters. Neither Rockbox, nor IpodLinux use any part of the original firmware, so the language it was written in before compiling doesn't much matter.
04:29:05safetydanspliifdown: it's likely C/C++ with asm but there's no way to know for sure. Doesn't matter though. Or what Soap said
04:29:46spliifdownyou do need to since you wil have to decompile it am i not right?
04:30:08spliifdownsince most of the coding in these apps are reverse engineered
04:30:08Soapyou don't decompile it back to source code.
04:30:35LloreanAnd less is reverse engineered than you might expect
04:31:03spliifdownyou need to knw how the software interacts with the hardware
04:31:14spliifdownin order to write the code
04:31:41spliifdownand is that where the problem lies?
04:31:45Soapyou decompile it back to assembly language, and saying "most" is not an accurate representation of the amount of info in Rockbox gleaned from reverse engineering.
04:32:03Soapone of the problems.
04:32:30spliifdownrockbox i understand the effort put into it
04:32:30Soapthe other major issue is getting custom software, even if it had working drivers, to be "allowed" to run on the hardware.
04:32:56spliifdownoh orite
04:33:08spliifdownso its actually not worth hacking it
04:33:31Soapdidn't say that.
04:33:33psycho_maniaccan anybody with a gigabeat reproduce this? i cannot and i have the latest build
04:33:44Soapand perhaps you misunderstand what I mean my "allowed"
04:34:08spliifdown"allowed" did you mean it legal terms?
04:34:23SoapI mean there is a bootloader which launches the firmware on the iPods, and it is this bootloader which will need to be "faked out", as it were, to loading Rockbox instead of Apple.
04:35:29spliifdownalot of work
04:35:35spliifdownif i were to begin this
04:35:47spliifdowni would need to learn c++
04:35:50spliifdownand asm
04:35:57safetydanasm, yes, C++ not so much
04:36:08spliifdownbut i think i should best start off with python?
04:36:15safetydanbut better would be a general understanding of how hardware works
04:36:24safetydanpython isn't going to help you with that
04:36:40spliifdownnot with that as a programming starting point
04:37:00spliifdowni knw how to use vb, but anyone with brain can use that
04:38:56spliifdownis java similar to asm?
04:39:19spliifdownif it is might be able to get my hands dirty asap
04:41:00spliifdownneway Soap, hopefully i can get my mind going on this
04:41:17spliifdownthanks for your time, very appreciated
04:41:29spliifdownits 4am here so i best be off to sleep
04:41:31spliifdowncheers mate
04:42:34spliifdownthanks safetydan
04:42:38 Quit spliifdown ()
04:45:11 Join Mouser_X [0] (n=mouser_x@
04:48:12 Quit FunkyELF ("Leaving")
04:49:49lostlogicwill anyone be upset if I spend 128 bytes of code size to make stripping tags easier?
04:51:29safetydanlostlogic: will it fix the bug with tag stripping and gapless?
04:52:08lostlogicthat's my point
04:52:22lostlogicit could be done without the extra 128 bytes but it would be more difficult
04:52:26 Join webguest26 [0] (i=450e16de@gateway/web/cgi-irc/
04:53:08LloreanDoes more difficult also mean "more complex, and harder for future developers to understand when they see it"?
04:53:53safetydanI'd say use the bytes if it leads to clearer code
04:54:09 Join goffa_ [0] (n=goffa@
04:54:17lostlogicok, thanks for the feedback, I'll see how much complexity it would add before making a final decision
04:54:51LloreanI'm all for saving space, but I prefer to do it by shooting down redundancy rather than saying "space is more important than doing it a good way"
04:55:47 Join webguest26df [0] (i=4c10ea95@gateway/web/cgi-irc/
04:55:53LloreanAlso, anyone know when mpegplayer got dithering, and what I can expect from it?
04:55:54safetydanand there's plenty of redundancy to sort out
04:56:03webguest26I'm trying to find some menu entries like Settings -> General Settings -> Display -> Browse Fonts or Settings -> General Settings -> Display -> Browse .WPS files.
04:56:24webguest26the manual says they're there, but I cannot find them
04:56:34psycho_maniacwhat player?
04:56:40webguest26dfhi i vae a question, is there a way to zoom into photos without having to stop the music from playing?
04:56:47Lloreanwebguest26: They're under theme settings now, I believe
04:56:54karashatawebguest26: those options have been moved to Theme Settings now
04:56:56Lloreanwebguest26: Unfortunately, I don't think whoever changed it updated the manual
04:57:10webguest26I can get as far as the Display menu, but only get LCD Settings, Scrolling, and three other things.
04:57:16webguest26I have the e250 Sansa
04:57:30webguest26and current build of Rockbox
04:57:55webguest26dfplease help i really need to know thank you. :)
04:57:56karashataTheme settings is in the root of the settings menu now, not in the General Settings
04:57:58webguest26I want to create a wps file, and then I assume I would browse to it in the settings menu
04:58:41webguest26dfthank you i got to go bye! :)
04:58:49psycho_maniacgo to Settings>Theme Settings> webguest26
04:58:52webguest26ok - I see "Theme settings" - and it contains .wps and fonts.
04:58:55 Quit webguest26df (Client Quit)
04:58:58webguest26Am I using the wrong manual?
04:59:32safetydanwebguest26: no, I think the manual is just out of date for that part
04:59:34 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
04:59:35webguest26using rockbox-sansae200.pdf, and getting the same information from the online Sansa e200 manual
04:59:51psycho_maniacin that menu browse themes, browse fonts, browse .wps files, show icons, clear backdrop, line selector type, and colours
04:59:52webguest26ok - can't expect manuals to keep up with the times i guess
05:00:00 Join miepchen^schlaf [0] (
05:00:11n17ikh|Lappythat's... an interesting error
05:00:13Lloreanwebguest26: You *should* be able to expect manuals to stay up to date.
05:00:23LloreanUnfortunately, not everyone is consistent in updating it when they change things
05:00:31 Quit goffa (Read error: 110 (Connection timed out))
05:00:32n17ikh|LappyI got an "undefined instruction" error on my H10 20gb with the latest build
05:01:05n17ikh|Lappyshould I try and delete my config file or some such and try again? or is it just a problem with the build itself, most likely?
05:01:57karashatan17ikh|Lappy: depends, what were you trying to do when you got that?
05:02:04n17ikh|Lappypower it on :/
05:02:37karashatatry clearing your settings and see if it happens again
05:02:56n17ikh|Lappygo go gadget reset button
05:03:12karashataturn it on and flip the hold switch into the lock position before the Rockbox screen comes up to clear them the easy way
05:03:41n17ikh|Lappydoes that actually save the cleared settings?
05:03:53n17ikh|Lappyor is it just for that boot
05:04:06n17ikh|Lappybecause, in the past, on the iriver it just did it for the boot with the hold switch
05:04:20n17ikh|Lappyon my sansa e series, it apparently saves the cleared settings automagically
05:04:29n17ikh|Lappybut that's with a much newer build
05:04:30karashataI'm not sure, I think is will save them if it gets a chance to write to disk, but I could be wrong
05:04:53karashataI've had it work either way
05:05:08karashatasame player and most recent builds, and I never had it crash on me, so...
05:05:28karashatawhat screen do you have it set to start up on when it loads up?
05:05:55n17ikh|Lappyprobably... the previous one
05:05:58n17ikh|Lappywhich is most likely the wps
05:06:08n17ikh|Lappyok, yeah
05:06:16n17ikh|Lappybooting with the config cleared fixes it
05:06:24n17ikh|Lappynow to find exactly WHAT causes it
05:06:43karashatait might be a glitch with trying to start on the previous screen it shut down on
05:06:51n17ikh|Lappylet me change that
05:06:55karashataI haven't had it do that myself, so...
05:07:33n17ikh|Lappywhere's that hide
05:07:40n17ikh|Lappynot the setting itself, but where it keeps the previous screen
05:07:57karashatano idea, you'd have to change the setting
05:07:59n17ikh|Lappycourse, I could just delete that one setting
05:08:54karashataI think it would be easier for you to change the setting to start on a specific screen than to try to find the one setting in the config file and delete its value
05:09:18karashataI have mine set to start on previous screen and it booted fine to the WPS when I just tested now
05:09:26karashatadefault is to boot to main menu
05:09:42karashataI'll see if I can reproduce your problem and figure out what it was
05:10:00n17ikh|Lappydon't know what caused the problem, though
05:10:05n17ikh|Lappyother than starting in the WPS
05:10:17karashatait may have just been that the upgrade changed something and it didn't like it from where it had been shut down from your previous build
05:10:29n17ikh|Lappywell, also the dynamic playlist was gone
05:10:44n17ikh|Lappywhen I went to resume play in the new build, it just said nothing to resume
05:11:10n17ikh|Lappyyeah, that's what I said
05:11:15karashatawhat build were you upgrading from?
05:11:16n17ikh|Lappypretty sure I didn't delete any music files
05:11:21n17ikh|Lappyan old one? dunno
05:11:28n17ikh|LappyI haven't used my h10 in a month or so
05:11:37n17ikh|Lappysince I got the e260
05:11:46n17ikh|Lappyflash players are so much nicer
05:12:02karashatayou probably ran into something being changed between your old build and the recent one that your settings didn't get along with very well
05:12:14n17ikh|Lappynot really worth worrying about
05:12:24karashataif it doesn't happen again, no it's not
05:12:29 Join goffa [0] (n=goffa@
05:14:17***Saving seen data "./dancer.seen"
05:17:14 Quit goffa_ (Read error: 110 (Connection timed out))
05:17:48 Join webguest31 [0] (i=185699d8@gateway/web/cgi-irc/
05:18:24 Quit hcs ("Leaving.")
05:18:37 Quit Soap (Read error: 110 (Connection timed out))
05:18:45 Join hcs [0] (n=agashlin@rockbox/contributor/hcs)
05:21:31 Quit webguest26 ("CGI:IRC (EOF)")
05:22:21 Quit hcs (Client Quit)
05:23:14 Quit webguest31 ("CGI:IRC (Ping timeout)")
05:23:29 Quit midkay ("Leaving")
05:24:24 Join atsea- [0] (i=atsea-@gateway/tor/x-51bc3756db61d251)
05:29:56 Join midkay [0] (n=midkay@rockbox/developer/midkay)
05:30:18 Join goffa_ [0] (n=goffa@
05:33:40 Join darkapostrophe [0] (
05:33:51 Join l7 [0] (
05:34:42 Quit goffa (Read error: 110 (Connection timed out))
05:35:29lostlogicwho has mp3s with trailing ape or id3v1 tags and experiences gapless breakage?
05:35:33lostlogicgoing to need a tester shortly.
05:41:52Mouser_Xlostlogic: There was talk about making the MIDI player into a codec. I was going through the logs, so I don't know how far the discussion got. However, if I'm not mistaken, MIDI shouild be flagged as atomic, correct? In the discussion I read, this wasn't mentioned. I just though it ought to be pointed out to someone, though.
05:42:42lostlogicMouser_X: yes, you are correct and I also saw the discussion (I was just masterfully lurking out of it). Nico knows that and was referencing it when he said something like "Buffering supports these things"
05:43:05Mouser_XAh. I probably missed that.
05:43:55lostlogicsafetydan, llorean, anyone else:
05:44:10lostlogicshould fix gapless, if you have any badly tagged files to test, also for general code review.
05:44:11Mouser_XAlso, ADXs aren't looping still. I was hoping that when 8092 (backwords seeking, IIRC) was fixed, that when the codec attempted to loop the file, it would reload the overwritten portion, or something.
05:44:28lostlogicMouser_X: what happens when they try to loop?
05:44:31lostlogicit should.
05:45:01Mouser_XI haven't had a chance to test with the rebuffer stuff yet, but HCS says they're still behaving the same as they did before the fix.
05:45:16Mouser_X(He did test, but he was on his way to bed, so I mentioned it instead.)
05:45:34lostlogicgotcha - well nico's fixage should make them work, not sure why it's not.
05:46:20Mouser_XBefore the fix, they'd loop, as long as they hadn't been overwritten. If the currently playing ADX was near the end of the buffer, it'd be overwritten, and thus wouldn't loop. This is, apparently, still happening.
05:46:37Mouser_XI figured it was worth mentioning, at least.
05:46:53lostlogicyeah, for sure −− it should at least be able to loop... eventually... with nico's changes.
05:46:54 Join FunkyELF [0] (
05:47:08lostlogicit would just require a big buffer dump
05:47:12Mouser_X(Even though I'm aware that you're not too fond of ADX's "unique" features.)
05:47:29safetydanlostlogic: I don't have any broken files to test, but I can't see anything obviously wrong with that patch
05:47:31lostlogicI'd prefer to pretend them out of existance, yes.
05:47:59FunkyELFhey guys, I looked at the online docs for rockbox on a video ipod and it didn't say anything about being able to select rockbox or the default apple firmware. I saw on youtube that you can do that. Is that how following the docs will set it up it just didn't mention it?
05:48:26Mouser_XFunkyELF: Dual-boot is how Rockbox works.
05:48:48Mouser_XIn other words, if you install according to the directions provided, it will dual-boot.
05:48:54Mouser_XIf not, it probably won't boot.
05:49:04scorcherockbox is dual-boot by deafault, but if you want to have an actual menu to select from, then you will want to look towards Loader2 (ipl's BL)
05:49:09Mouser_X(That is, if you don't follow the directions, you might have trouble booting.)
05:49:37Mouser_Xscorche: Ah, I didn't know that Loader2 had a menu thingy.
05:50:00FunkyELFis it common to "brick" you iPod while installing rockbox on it?
05:50:16lostlogicnearly impossible.
05:50:26FunkyELFgood to know
05:50:53l7how likely is bricking with sansas?
05:51:25FunkyELFI have a 5th gen 60Gb video iPod, my brother just got an 80Gb iPod classic which apparently can't run rockbox.....he said he'll trade with someone at work who has an 80Gb Video iPod if I install it and like it
05:51:26Mouser_XFunkyELF: Bricking is only likely if you use the "Sledgehammer method" to install. That is, beating it senseless...
05:51:41FunkyELFl7, my Sandisk bricked itself
05:52:00lostlogicFunkyELF: and you ere unable to restore?
05:52:00*Mouser_X installed Rockbox on his e250 *very* easily.
05:52:04scorchel7: unlikely (although it is simple to get it into a brick-like-state if you dont follow the instructions)
05:52:08lostlogicFunkyELF: I temp-bricked mine but was able to fix with the instructions
05:52:13l7FunkyELF: wow, how did it happen?
05:52:45FunkyELFlostlogic, l7 firmware aparently had a race condition in it....just freezes after a while
05:52:47 Part safetydan
05:53:04lostlogicFunkyELF: that's not bricked, don't scare people
05:53:22FunkyELFI saw tons of posts of people complaining about it and how SanDisk knows about it but doesn't do anything other than send you another one which will freeze after a while too
05:53:25l7FunkyELF: was that OF or rockbox?
05:53:48FunkyELFThis had nothing to do with rockbox....just the crappy Sansa M250
05:56:07FunkyELFwill my iPod be able to work with all those iPod accessories with rockbox?
05:57:04Mouser_XNot Likely. There's a wiki page that can tell you more though.
05:57:10 Nick billenium is now known as BilleniumZzZ (
05:57:18Mouser_X(I don't have an iPod, and thus I don't remember the page.)
05:57:23lostlogicI am awesome. gapless fix works first shot.
05:57:31l7FunkyELF: did sandisk try to fix it?
05:57:37FunkyELFdoesn't seem impossible, just a lot of reverse engineering to see what that little port provides
05:57:40Mouser_Xlostlogic = AWESOME
05:57:43Mouser_XNice job.
05:57:56FunkyELFl7, they sent me a new one, different firmware I think
05:58:03psycho_maniacwhat was wrong with gapless?
05:58:09l7oh okay, nice service i guess
05:58:19Mouser_Xpsycho_maniac: It doesn't work, if your MP3s have trailing tags.
05:58:32psycho_maniacdefine: trailing tags
05:58:40Mouser_XID3 v1, or APE tags.
05:59:07lostlogictags on th eend of files would break codecs with variable frame sizes (like mp3 vbr)
05:59:08psycho_maniacFunkyELF: look at this page:
06:01:44 Quit karashata ("I will *SO* make you regret that... Later... *is a lazy dragon*")
06:01:46lostlogicit was one of the last remaining regressions from MoB
06:02:59lostlogicand the fact that it worked first shot is a testament to how much better the playback system is thanks to Nico_P.
06:09:24lostlogicbagh, red
06:11:11psycho_maniacwow lots of red.
06:11:23lostlogicall hwcodec targets
06:11:37lostlogicoh and code size
06:11:40lostlogicwtf 1/2k!?
06:11:45*lostlogic scowls
06:12:08psycho_maniacwill it still be safe to download the current build?
06:12:42lostlogicno problem on swcodec targets
06:15:16lostlogicI hate binary size increases *cry*
06:24:21 Join fiXXXerMet [0] (
06:25:45fiXXXerMetI was following a post about turning the e260r into a e260 and then putting rockbox on it. It said to hold down the select button and press power, while hold is enabled. I did this, and it worked, but now I can't turn the device off. The blue light stays on, and it won't go off (nor will the screen come on).
06:26:04lostlogicfiXXXerMet: hold the power button for longer
06:26:08lostlogiclike 30-60s
06:26:16fiXXXerMetYeah? ok.
06:27:00fiXXXerMetWas does that button combo do anyway, when it turns the blue light on?
06:27:27scorcheputs teh device into manufacture mode
06:31:04 Part Llorean
06:32:16 Part fiXXXerMet
06:48:31 Join Spanky [0] (i=407e8b3f@gateway/web/cgi-irc/
06:49:31 Quit Spanky (Client Quit)
07:01:35 Join Traveler4 [0] (
07:03:04Traveler4is it faster to add images then update the lcd at the end of a cycle in rockbox compared to ipodlinux programs with openGl?
07:09:56 Join goffa [0] (n=goffa@
07:10:55 Join woodensoul [0] (n=woodenso@
07:12:09woodensoulLooking for a pre-built simulator for the iRiver H320, but rasher's site isn't working right now.
07:14:20***Saving seen data "./dancer.seen"
07:15:51 Join axionix [0] (
07:16:34 Quit Psy-Krow ("don't taze me bro!")
07:17:38 Quit goffa_ (Read error: 110 (Connection timed out))
07:17:56 Quit axionix (Client Quit)
07:18:18 Join axionix [0] (
07:20:01 Quit Traveler4 ("Java user signed off")
07:21:03psycho_maniachow does the %cf tag work? what would you use for a conditional?
07:23:32 Quit l7 ("Lost terminal")
07:24:38 Quit axionix ("leaving")
07:24:57 Join axionix [0] (
07:29:21 Join moos [0] (
07:29:29 Quit moos (Read error: 104 (Connection reset by peer))
07:30:39 Join kubiix [0] (
07:31:18 Join GTroy [0] (
07:32:08 Quit DogBoy ("Leaving")
07:32:14*GTroy can't seem to find where the sansa with rockbox is to mount
07:33:12lostlogicGTroy: check dmesg for the latest USB storage device mapped
07:33:32*GTroy did that but forgot what to look for
07:39:31 Quit woodensoul ()
07:39:49 Join woodensoul [0] (n=noneofya@
07:42:56 Join goffa_ [0] (n=goffa@
07:48:33 Quit goffa (Read error: 110 (Connection timed out))
07:57:11 Join goffa [0] (n=goffa@
08:00:19 Quit psycho_maniac (" HydraIRC -> <- The professional IRC Client :D")
08:02:50 Quit goffa_ (Read error: 110 (Connection timed out))
08:09:50 Join LinusN [0] (i=linus@rockbox/developer/LinusN)
08:13:06 Join DogBoy [0] (
08:13:40 Join CaptainSquid [0] (
08:21:32 Join GodEater_ [0] (n=bryan@rockbox/staff/GodEater)
08:25:55woodensoulHow can i make a backdrop that is supported by Rockbox?
08:26:11woodensoulI see that it says 220X176 for the H320, 1 bit, or 24 bit.
08:27:09GodEater_so what's the actual question ?
08:27:09LinusNyou have to look up what resolution your player has, and create a bitmap with exactly those dimensions
08:27:11 Join goffa_ [0] (n=goffa@
08:27:30woodensoulI just figured it out.
08:28:10woodensoulI used irfanview, but I had to check "change color depth" and set it to 24BPP.
08:29:09 Quit goffa (Read error: 110 (Connection timed out))
08:29:53woodensoulAnother question: why does the sound from the simular skip when performing other functions within Windows?
08:30:43 Quit Mouser_X (Read error: 104 (Connection reset by peer))
08:31:38 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
08:34:17GodEater_haven't noticed that myself - the sim behaves just fine for me =/
08:34:52woodensoulWell I just started using it yesterday and I want to use it as my default media player since I love Rockbox on my DAPs so much.
08:35:07GodEater_I do that too :)
08:35:31woodensoulit seems to skip when copying files, opening folders through explorer
08:35:54 Join Rob2222 [0] (
08:39:02LinusNwoodensoul: the simulator is not optimized in any way, and is just a debugging tool
08:41:24woodensoulLinusN: I see. I've been loving it learning to make WPSs and I'll probably go ahead and start using it for my default audio player and just cut back on other windows operations while listening.
08:46:28 Join aliask [0] (
08:46:36woodensoulIs there any way to use the Simulator to play music that is not in the "Archos" directory?
08:48:51amiconnUse a symlink (or on windows, a junction point) within the archos dir
08:49:37*amiconn uses junction points to have only one copy of his testfile tree for a whole bunch of simulators
08:51:51woodensoulWhere can I learn how to do that?
08:52:10 Quit jhulst (Read error: 113 (No route to host))
08:53:22 Quit Rob222241 (Read error: 110 (Connection timed out))
08:53:56 Quit male (Read error: 110 (Connection timed out))
08:56:03 Quit goffa_ (Remote closed the connection)
08:56:06 Join goffa [0] (n=goffa@
08:57:29 Join male [0] (
08:58:12 Join petur [0] (n=petur@rockbox/developer/petur)
08:59:38 Join spitball_ [0] (
09:02:15 Join ender` [0] (
09:03:32 Quit goffa (Remote closed the connection)
09:03:36 Join goffa [0] (n=goffa@
09:05:08 Join Mouser_X [0] (n=mouser_x@
09:08:26 Join robert10_infoxp [0] (n=robert10@
09:08:53robert10_infoxpalguien habla español?
09:10:34 Quit robert10_infoxp ("Adios amigos de")
09:11:01scorcheingles por favor
09:11:08LycoLocoa little late, scorche
09:14:04 Quit goffa (Remote closed the connection)
09:14:08 Join goffa [0] (n=goffa@
09:14:21***Saving seen data "./dancer.seen"
09:14:30woodensoulGoogle is my friend. ;-)
09:14:45woodensoulWhy is it that the controls listed arent' the same on my keyboard in the sim?
09:16:41 Join Zagor [0] (
09:17:08GodEater_good question
09:17:17GodEater_where is the list you refer to ?
09:31:35 Join goffa_ [0] (n=goffa@
09:32:44 Quit male ("User disconnected")
09:34:20 Join Doomed_ [0] (
09:34:20 Quit alienbiker99 (Read error: 104 (Connection reset by peer))
09:37:47 Join Nico_P [0] (
09:38:09 Quit Zagor (
09:38:09 Quit FunkyELF (
09:38:09 Quit iamben (
09:38:09 Quit DM| (
09:38:09 Quit zicho (
09:38:09 Quit lids_ (
09:38:09 Quit Bagder (
09:38:09 Quit jepler (
09:38:09 Quit TTThomas (
09:38:09 Quit gromit`` (
09:38:09 Quit RaRe (
09:38:09 Quit maraz (
09:38:09 Quit J (
09:38:09 Quit weswh- (
09:38:09 Quit blithe (
09:38:09 Quit Weiss (
09:38:09 Quit SirFunk (
09:38:09 Quit DiDjCodt (
09:38:09 Quit joshin (
09:38:09 Quit [omni] (
09:38:09 Quit fxb__ (
09:38:09 Quit z35 (
09:38:09 Quit ivan (
09:38:09 Quit daurnimator (
09:38:09 Quit ender` (
09:38:09 Quit spitball_ (
09:38:09 Quit DogBoy (
09:38:09 Quit advcomp2019 (
09:38:09 Quit GodEater (
09:38:09 Quit Redbreva (
09:38:09 Quit Toki (
09:38:09 Quit krazykit (
09:38:09 Quit newbyx86 (
09:38:09 Quit LycoLoco (
09:38:09 Quit aliask (
09:38:09 Quit Rob2222 (
09:38:09 Quit CaptainSquid (
09:38:09 Quit J3TC- (
09:38:09 Quit DaCapn (
09:38:09 Quit sslashes (
09:38:09 Quit marcosource (
09:38:09 Quit tuplanolla (
09:38:09 Quit [mbm] (
09:38:10 Quit Shaid (
09:38:10 Quit HellDragon (
09:38:10 Quit rvvs89 (
09:38:10 Quit GodEater_ (
09:38:10 Quit darkapostrophe (
09:38:10 Quit animeloe (
09:38:10 Quit spence_ (
09:38:10 Quit Dark_Apostrophe (
09:38:10 Quit amiconn (
09:38:10 Quit Seedy (
09:38:10 Quit Zom (
09:38:10 Quit crwl (
09:38:10 Quit sneakums (
09:38:12 Quit tedrock (
09:38:12 Quit Lynx_ (
09:38:12 Quit jmworx (
09:38:12 Quit qwm (
09:38:12 Quit courtc (
09:38:12 Quit ATravelingGeek (
09:38:12 Quit ChanServ (
09:38:12 Quit Jon-Kha (
09:38:12 Quit Slasheri (
09:38:12 Quit Hadaka (
09:38:12 Quit rasher (
09:38:12 Quit goffa_ (
09:38:12 Quit axionix (
09:38:12 Quit homielowe (
09:38:12 Quit japc (
09:38:12 Quit nanok (
09:38:12 Quit hannesd (
09:38:12 Quit BHSPitMonkey (
09:38:12 Quit sbeh (
09:38:12 Quit lastebil (
09:38:12 Quit JETC- (
09:38:12 Quit parafin|away (
09:38:12 Quit DataGhost (
09:38:12 Quit crashd_ (
09:38:12 Quit jurrie (
09:38:12 Quit nicktastique (
09:38:12 Quit DrCrow (
09:38:12 Quit lodesi (
09:38:12 Quit thegeek (
09:38:12 Quit gtkspert (
09:38:12 Quit markun (
09:38:12 Quit Langly (
09:38:12 Quit UncleRemus (
09:38:12 Quit guyzmo (
09:38:12 Quit Rick (
09:38:12 Quit Xerion (
09:38:12 Quit maddler (
09:38:12 Quit preglow (
09:38:12 Quit LinusN (
09:38:12 Quit goffa (
09:38:12 Quit Mouser_X (
09:38:12 Quit Soap_ (
09:38:12 Quit scorche (
09:38:12 Quit XavierGr (
09:38:12 Quit Nimdae (
09:38:12 Quit Ave (
09:38:12 Quit Kohlrabi (
09:38:12 Quit BrianHV (
09:38:12 Quit inakinfbi (
09:38:12 Quit feisar (
09:38:12 Quit Doomed_ (
09:38:12 Quit woodensoul (
09:38:12 Quit kubiix (
09:38:12 Quit atsea- (
09:38:12 Quit n17ikh|Lappy (
09:38:12 Quit jhMikeS (
09:38:12 Quit GTroy (
09:38:12 Quit lostlogic (
09:38:12 Quit shodanX (
09:38:12 Quit maxkelley (
09:38:12 Quit bagawk (
09:40:09 Join jhMikeS [0] (
09:40:10***Server message 505: 'logbot :Private messages from unregistered users are currently blocked due to spam problems, but you can always message a staffer. Please register! ( )'
09:44:06NJoinChanServ [0] (ChanServ@services.)
09:44:06 Join bagawk [0] (n=lee@
09:44:06 Join alienbiker99 [0] (
09:44:06NJoinNimdae [0] (
09:44:06NJoinmaxkelley [0] (
09:44:06 Join scorche` [0] (n=scorche@rockbox/administrator/scorche)
09:44:06 Join miepchen^schlaf [0] (n=hihi@
09:44:06 Join davina [0] (n=davina@
09:44:06 Join kubiixx [0] (
09:44:06 Join Soap__ [0] (
09:44:06NJoinDoomed_ [0] (
09:44:06NJoingoffa_ [0] (n=goffa@
09:44:06 Join Zagor [0] (n=bjst@rockbox/developer/Zagor)
09:44:06NJoingoffa [0] (n=goffa@
09:44:06NJoinMouser_X [0] (n=mouser_x@
09:44:06NJoinender` [0] (
09:44:06NJoinspitball_ [0] (
09:44:06NJoinaliask [0] (
09:44:06NJoinRob2222 [0] (
09:44:06NJoinGodEater_ [0] (n=bryan@rockbox/staff/GodEater)
09:44:06NJoinCaptainSquid [0] (
09:44:06NJoinDogBoy [0] (
09:44:06NJoinLinusN [0] (i=linus@rockbox/developer/LinusN)
09:44:06NJoinwoodensoul [0] (n=noneofya@
09:44:06NJoinGTroy [0] (
09:44:06NJoinkubiix [0] (
09:44:06NJoinaxionix [0] (
09:44:06NJoinFunkyELF [0] (
09:44:06NJoindarkapostrophe [0] (
09:44:06NJoinatsea- [0] (i=atsea-@gateway/tor/x-51bc3756db61d251)
09:44:06NJoinspence_ [0] (
09:44:06NJoinanimeloe [0] (n=animeloe@unaffiliated/animeloe)
09:44:06NJoinSlasheri [0] (i=miipekk@rockbox/developer/Slasheri)
09:44:06NJoinHadaka [0] (
09:44:06NJoinJon-Kha [0] (
09:44:06NJoinJ3TC- [0] (
09:44:06 Join Dark_Apostrophe [0] (n=darkapos@supporter/monthlybyte/DarkApostrophe)
09:44:06NJoiniamben [0] (
09:44:06NJoinhomielowe [0] (
09:44:06NJoinn17ikh|Lappy [0] (
09:44:06NJoinDaCapn [0] (
09:44:06NJoinDM| [0] (
09:44:06NJoinadvcomp2019 [0] (n=advcomp2@unaffiliated/advcomp2019)
09:44:06NJoinrasher [0] (n=rasher@rockbox/developer/rasher)
09:44:06Mode"#rockbox +o ChanServ " by
09:44:06NJoinamiconn [0] (n=jens@rockbox/developer/amiconn)
09:44:06NJoinjapc [0] (
09:44:06NJoinzicho [0] (
09:44:06NJoinSeedy [0] (
09:44:06NJoinlids_ [0] (n=lds@
09:44:06 Join Bagder [0] (n=daniel@rockbox/developer/bagder)
09:44:06NJoinGodEater [0] (n=bryan@rockbox/staff/GodEater)
09:44:06NJoinnanok [0] (n=nanok@
09:44:06NJoinhannesd [0] (
09:44:06NJoinZom [0] (
09:44:06NJoincrashd_ [0] (
09:44:06NJoinSoap_ [0] (n=Soap@rockbox/staff/soap)
09:44:06NJoinjepler [0] (n=jepler@emc/developer/jepler)
09:44:06NJoinBHSPitMonkey [0] (n=stephen@unaffiliated/bhspitmonkey)
09:44:06NJoinjurrie [0] (
09:44:06NJoinTTThomas [0] (
09:44:06NJoinLycoLoco [0] (
09:44:06NJoinXavierGr [0] (
09:44:06NJoingromit`` [0] (
09:44:06NJoinnicktastique [0] (n=nick@unaffiliated/nicktastic)
09:44:06NJoinDrCrow [0] (
09:44:06NJoinRedbreva [0] (
09:44:06NJoinRaRe [0] (
09:44:06NJoinmaraz [0] (
09:44:06NJoinDiDjCodt [0] (
09:44:06NJoinblithe [0] (
09:44:06NJoinJ [0] (
09:44:06NJoin[omni] [0] (
09:44:06NJoinfxb__ [0] (
09:44:06NJoinivan [0] (n=ivan`@unaffiliated/ivan/x-000001)
09:44:06NJoinWeiss [0] (
09:44:06NJoindaurnimator [0] (n=daurn@unaffiliated/daurnimator)
09:44:06NJoinz35 [0] (
09:44:06NJoinSirFunk [0] (
09:44:06NJoinjoshin [0] (n=josh@unaffiliated/joshin)
09:44:06NJoinweswh- [0] (
09:44:06NJoin[mbm] [0] (i=mbm@openwrt/developer/mbm)
09:44:06NJoinsslashes [0] (i=sslashes@
09:44:06NJoinmarcosource [0] (
09:44:06NJointuplanolla [0] (
09:44:06 Join rvvs89 [0] (n=rvvs89@pdpc/supporter/active/rvvs89)
09:44:06NJoinShaid [0] (
09:44:06NJoinHellDragon [0] (i=jd@unaffiliated/helldragon)
09:44:06NJoinnewbyx86 [0] (
09:44:06NJoinToki [0] (
09:44:06NJoinkrazykit [0] (
09:44:06NJoinXerion [0] (
09:44:06NJointedrock [0] (
09:44:06NJoinguyzmo [0] (
09:44:06NJoinmarkun [0] (n=markun@rockbox/developer/markun)
09:44:06NJoininakinfbi [0] (i=0@
09:44:06NJoinsneakums [0] (
09:44:06NJoinLangly [0] (
09:44:06NJoinUncleRemus [0] (
09:44:06 Join courtc [0] (n=court@unaffiliated/courtc)
09:44:06NJoingtkspert [0] (
09:44:06NJointhegeek [0] (
09:44:06NJoinparafin|away [0] (
09:44:06NJoincrwl [0] (
09:44:06NJoinshodanX [0] (
09:44:06NJoinATravelingGeek [0] (n=ATG@pdpc/supporter/student/ATravelingGeek)
09:44:06NJoinjmworx [0] (
09:44:06NJoinqwm [0] (
09:44:06NJoinLynx_ [0] (
09:44:06NJoinlodesi [0] (
09:44:06NJoinmaddler [0] (
09:44:06NJoinRick [0] (i=rick@unaffiliated/rick)
09:44:06NJoinJETC- [0] (
09:44:06NJoinlastebil [0] (
09:44:06NJoinsbeh [0] (n=sbeh@
09:44:06NJoinDataGhost [0] (
09:44:06NJoinBrianHV [0] (
09:44:06NJoinfeisar [0] (
09:44:06NJoinlostlogic [0] (n=lostlogi@rockbox/developer/lostlogic)
09:44:06NJoinKohlrabi [0] (
09:44:06NJoinpreglow [0] (n=thomj@rockbox/developer/preglow)
09:44:06NJoinAve [0] (
09:44:07***Server message 505: 'logbot :Private messages from unregistered users are currently blocked due to spam problems, but you can always message a staffer. Please register! ( )'
09:44:07***Server message 505: 'logbot :Private messages from unregistered users are currently blocked due to spam problems, but you can always message a staffer. Please register! ( )'
09:44:13***Server message 505: 'logbot :Private messages from unregistered users are currently blocked due to spam problems, but you can always message a staffer. Please register! ( )'
09:44:22 Quit goffa (Read error: 110 (Connection timed out))
09:44:28 Quit Soap_ (Read error: 104 (Connection reset by peer))
09:45:24 Quit Doomed_ (Read error: 104 (Connection reset by peer))
09:45:38 Join Doomed_ [0] (
09:45:42CtcpVersion from freenode-connect!freenode@freenode/bot/connect
09:45:42 Quit alienbiker99 (Read error: 104 (Connection reset by peer))
09:46:09 Quit kubiix (Read error: 104 (Connection reset by peer))
09:46:26 Join needhelp [0] (
09:46:28needhelpHi! I need your help. Iam collecting points in page listed below. If you be so kind, please click url below.(sorry for the spam, thank you)
09:46:30 Part needhelp
09:46:51markunat least he's saying sorry :)
09:47:07 Join spiorf [0] (
09:48:01 Quit miepchen^schlaf (Read error: 104 (Connection reset by peer))
09:48:10 Join miepchen^schlaf [0] (
09:48:13 Quit goffa_ (Remote closed the connection)
09:48:16 Join goffa [0] (n=goffa@
09:50:13 Join barrywardell [0] (
09:51:20 Quit Kohlrabi (Read error: 104 (Connection reset by peer))
09:51:23 Join Kohlrabi [0] (
09:51:37 Join pixelma [0] (i=pixelma@rockbox/staff/pixelma)
09:52:01 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
09:53:40 Part GTroy
10:17:14 Quit idnar (Nick collision from services.)
10:17:16 Join idnar_ [0] (i=mithrand@unaffiliated/idnar)
10:22:06 Quit goffa (Remote closed the connection)
10:22:11 Join goffa [0] (n=goffa@
10:25:06Nico_Plostlogic: nice commit :)
10:27:20peturshame he didn't use the correct FS# syntax to create the link...
10:28:12linuxstbOr a shame the FS linker doesn't support FSxxxx...
10:28:34Bagderthose lame script writers should be shot
10:30:56 Quit miepchen^schlaf ("Verlassend")
10:39:02linuxstbpreglow: Would it be worth putting your rbspeexdec.c program in SVN - so people can test their encodings?
10:46:32 Join OlivierBorowski [0] (
10:47:01Nico_Paliask: what's the status of interrupts on the S?
10:47:30aliaskNico_P: ptw419 did some work on them, but never got them working
10:47:54aliaskNot entirely sure what he did
10:48:17*linuxstb summons ptw419
10:48:39Nico_Paliask: what kind of experience did he have with ARM?
10:49:31linuxstbAs we're going to need USB in the bootloader anyway (which I'm assuming is going to require interrupts), maybe we should concentrate on getting interrupts working well in the bootloader build, and forget about the main Rockbox build for now?
10:49:47 Quit goffa (Remote closed the connection)
10:49:50 Join goffa [0] (n=goffa@
10:51:08Nico_Plinuxstb: makes sense... what are they used for in the current bootloader? (or what could we use them for as test?)
10:51:18 Join webguest16 [0] (i=dce3b189@gateway/web/cgi-irc/
10:51:20GodEater_linuxstb: summoning ptw419 requires a small sacrifice and some sort of ritual dancing I think...
10:51:30*linuxstb fetches the goat
10:51:46 Quit webguest16 (Client Quit)
10:51:51 Join webguest16 [0] (i=dce3b189@gateway/web/cgi-irc/
10:51:51aliaskNico_P: He was learning as he went I think.
10:52:43linuxstbNico_P: The first thing would probably just be the ticker interrupt - we can test that's working by just displaying the value of current_tick.
10:52:56webguest16I have a ipod shuffle look alike, can I install rockbox on this look alike?
10:53:13linuxstbNo. You can't even install it on a genuine ipod shuffle.
10:53:42aliaskNico_P: Currently the current_tick variable is faked and just points to a hardware timer which increments at 100hz.
10:53:47 Join webguest75 [0] (i=5614d05d@gateway/web/cgi-irc/
10:53:59linuxstbwebguest16: Because all mp3 players are different and Rockbox needs to be made to work on every one individually - which is a lot of work.
10:54:06webguest75hi. is the rickbox support irc?
10:54:15linuxstbNo, Rockbox.
10:55:45 Quit webguest75 (Client Quit)
10:56:22linuxstbaliask: Ah, do you know the resolution of the hardware timer?
10:57:23aliaskLet me check
10:57:53Nico_Plinuxstb: what's the code that enables interrupts? set_irq_level?
10:58:33linuxstbYes, plus you need to ensure the exception vectors (normally located at address 0x0, but possibly they can be remapped) point to interrupt handlers.
10:58:51aliasklinuxstb: 27mhz
10:59:13linuxstbAh, the "mpeg tick" frequency (IIRC).
10:59:15GodEater_is that normal ? Seems a very off figure....
10:59:34*GodEater_ looks suspiciously at his keyboard
10:59:54BagderGodEater: someone switched your d and f keys again? ;-)
11:00:02Nico_Plinuxstb: ah then it's probably a problem with the exception vectors, isn't it? set_irq_level should just be the same as any other ARM, shouldn't it?
11:00:29aliask0x0 is not writable on the IMX31
11:00:43GodEater_Bagder: apparently not, they switched the wiring between my middle and ring fingers instead
11:00:46linuxstbNico_P: Yes, that would be the first place I would look. We need to try and understand how memory is being mapped - that should be the code in crt0.S and maybe the system.c for i.mx31
11:00:56markunaliask: then where does it point to?
11:01:21linuxstbaliask: But that's where ARM exception vectors live... On PortalPlayer for example, we remap DRAM to start at 0x0 so we can modify the exception vectors.
11:01:40aliask0x0 is "Secure ROM" according to the datasheet
11:01:57linuxstbBut I think PP also has registers to remap the exception vectors elsewhere - so maybe the imx31 has the same ability.
11:02:01markunah, like that. On the Gigabeat F we also remap DRAM to 0x0
11:02:42linuxstbDoes the imx have any IRAM?
11:03:21aliaskI think so, 0x1FFFC000 is "RAM" according to the IMX datasheet, 16kb worth.
11:03:21 Quit webguest16 ("CGI:IRC (EOF)")
11:03:47linuxstbBy default the Telechips devices have IRAM at 0x0, and SDRAM at 0x20000000 (IIRC), so the exception vectors can be in IRAM.
11:04:28*GodEater_ wonders what on earth "Secure ROM" is
11:04:43linuxstbROM with a big "Keep Out!" sign on it.
11:05:06GodEater_guard dogs and watch towers too presumably
11:06:37 Quit barrywardell ()
11:06:48Bagderpossibly it is or can be encrypted
11:07:26linuxstbHow is it decrypted?
11:08:25Bagdergood question, it just struck me that I recall imx having some crypto suff in it
11:08:57BagderI'm just speculating
11:08:58Bagderignore me
11:10:39linuxstbFrom one of Freescale's product briefs - "Includes secure ROM designed to provide restricted access and interrupt-free execution for sensitive code"
11:11:35Nico_Pthere isn't much about it in the datasheet
11:13:20linuxstbSo I guess we need to either remap DRAM over the top of it, or move the exception vectors from 0x0.... Does the datasheet describe how to use interrupts?
11:14:24***Saving seen data "./dancer.seen"
11:14:45aliasklinuxstb: There's an "Accelerated Vector Interrupt Controller", which is documented, but this is just designed to help the ARM interrupts as far as I can see
11:18:05GodEater_aliask: can't we use that anyway though ? Assuming I'm understanding it correctly, the ARM interrupts are processed there anyway ?
11:18:58 Join Thundercloud [0] (
11:19:06 Join goffa_ [0] (n=goffa@
11:19:15aliaskWell, I think it's supposed to be used to tell where an interrupt came from. IIUC you usually just get a general "interrupt", which would jump to an interrupt function. That function should check the AVIC to see what caused it, and act accordingly.
11:19:40aliaskBut, I'm no expert
11:19:45 Quit spitball_ ("leaving")
11:23:33*GodEater_ prods at the "experts" for comment
11:27:12 Join maxkelley_ [0] (
11:27:19 Quit maxkelley (Remote closed the connection)
11:27:58Nico_Pto get basic interrupts working I doubt we need to care about the AVIC
11:28:08 Join pondlife [0] (n=Steve@rockbox/developer/pondlife)
11:28:09aliaskHrm, it seems that the MMU code has been moved into a generic ARM MMU file. Good stuff.
11:28:53Nico_Phow to we remap DRAM to 0x0?
11:29:01aliaskI worry though that if we reset the MMU we'll lose the memory mapping to the LCD
11:29:35aliaskNico_P: map_section() in firmware/target/arm/mmu-arm.c
11:30:38 Join idnar [0] (i=mithrand@unaffiliated/idnar)
11:30:42 Quit goffa (Read error: 110 (Connection timed out))
11:31:17GodEater_the reference manual seems to indicate that you have to use the AVIC, regardless of whether you want to use it's advanced features or not =/
11:31:41Nico_Paliask: mmu-imx31.c has map_section too... shouldn't it be removed?
11:31:50GodEater_section 9.4.3 if anyone is interested...
11:32:04Nico_PGodEater_: I don't understand much of it :/
11:32:05aliaskNico_P: Yep
11:33:31markunNico_P: is the mmu from the imx31 different from the general ARM one? (used by the m:robe and gigabeat f)
11:33:37Nico_PGodEater_: doesn't that part say interrupt handlers are looked for in 0x18 + MMU_OFFSET?
11:33:40jhMikeSComment before I have to leave: the first thing I'll do after receiving the S is redo the whole avic system. We don't dynamically install handlers so it can be much simpler I think.
11:33:49Nico_Pmarkun: I have no idea
11:34:06Nico_PjhMikeS: it's ordered?
11:34:30jhMikeSWhy does that matter?
11:34:33aliaskmarkun: I don't think so. The imx31 datasheet points you to the ARM documentation on the CP15 register, which the arm-mmu code seems to use
11:34:46jhMikeSThe F uses a vector table
11:35:03Nico_PjhMikeS: to know if you're going to get it soon :)
11:35:53jhMikeSoh :P
11:36:00markunaliask: could you use the general arm-mmu code then and maybe update it with any improvements from the imx31 code?
11:36:02jhMikeSYeah, you could say it's ordered. :)
11:38:24 Quit woodensoul ()
11:38:41jhMikeShow come avic_enable_int isn't called?
11:39:11 Join PaulJam [0] (
11:39:29 Quit idnar_ (Read error: 110 (Connection timed out))
11:42:01Nico_Pmaybe I'm confused but the interrupt vectors seem to go in the VECTOR0 to VECTOR63 regs
11:44:29linuxstbNico_P: The diagram in section 9.4.3 of the reference manual (page 433) may help (and the whole of chapter 9)
11:45:20GodEater_that's the section I quoted earlier...
11:45:45Nico_Plinuxstb: yes, and fig 9-31 is what made me think what I said
11:46:22Nico_Pdon't we just need to init VECTORx with the address to an ISR and we're good?
11:46:25 Join ewan_ko_08 [0] (i=ewan_ko_@
11:46:49jhMikeSthe address of the handler, not those branch instructions I take it?
11:47:37linuxstbYes, it looks like VECTORx is the address.
11:47:39jhMikeSlooks like those ldr pc, ..blah.. should just be replaced
11:47:46jhMikeSor eliminated even
11:48:08linuxstbYou still need those for the core ARM handlers (at address 0x0, or wherever that vector table has been remapped to)
11:48:16jhMikeSjust copy a hardcoded table with 64 weak aliases and it should work
11:48:31GodEater_is it just me, or does the manual not say where the table is ?
11:48:41ewan_ko_08how do i reformat my sansa e260 if it says that it is write protected?
11:48:51Bagderewan_ko_08: why do you try to reformat it and how?
11:49:05ewan_ko_08cause it doesn't turn on anymore
11:49:09Nico_PGodEater_: 0x6800_0100 (VECTOR0)–0x6800_01FC (VECTOR63)
11:49:16ewan_ko_08it displays load main image failed
11:49:23GodEater_Nico_P: ah thanks ;)
11:49:40Bagderewan_ko_08: and how do you try to format it?
11:49:45linuxstbjhMikeS: You're talking about the vector_init() function in avic-imx31.c ?
11:50:01ewan_ko_08through recovery mode
11:50:30Bagderewan_ko_08: you put the magic file in there?
11:50:44ewan_ko_08but i cannot copy the files because it says that it is write protected
11:51:00Bagderewan_ko_08: but you get to see the 16MB "disk"?
11:51:06ewan_ko_08Badger: yes
11:51:16linuxstbWhat is telling you it is write protected?
11:51:33jhMikeSlinuxstb: yeah, that's just to accomodate dynamic installation without changing the table
11:51:45 Quit Febs (Read error: 110 (Connection timed out))
11:52:06ewan_ko_08is there anyway i could fix this?
11:52:17Bagderewan_ko_08: What is telling you it is write protected?
11:52:25 Nick fxb__ is now known as fxb (
11:52:26ewan_ko_08Badger: yes
11:52:35Bagderyes is not an answer to that question
11:52:58BagderWhat says write protected?
11:53:09Bagderhow do you try to copy the file?
11:53:20Nico_Pyeah, why are avic_init and vector_init never called?
11:53:23ewan_ko_08Badger: when i try to copy the files an pop up message appears saying that the media is write protected
11:53:35linuxstbewan_ko_08: Are you using Windows?
11:53:38ewan_ko_08Badger: i tried drag and drop and also copy paste
11:53:53ewan_ko_08linuxstb: yes
11:54:07 Quit Dark_Apostrophe (Read error: 104 (Connection reset by peer))
11:54:07 Nick darkapostrophe is now known as Dark_Apostrophe (
11:54:30Bagderto me it sounds like the device is broken somehow
11:55:02ewan_ko_08badger: but the computer can still recognize it when it is in recovery mode
11:55:05linuxstbewan_ko_08: Can you try on a different computer, or with a different USB cable?
11:55:28Bagderewan_ko_08: sure, the device is what prevents the writing, not the computer
11:55:52BagderI've never seen anyone have this problem before
11:56:05ewan_ko_08linuxstb: i don't have any other computer
11:56:39 Join Febs [0] (
11:57:49ewan_ko_08linuxstb: do you think the problem is with my computer?
11:59:15pixelmathere was someone yesterday with the "load main image" problem on his e200 after installing rockbox and bootloader - he then tried at another computer (said "old laptop with USB 1.1") and got it working
11:59:48ewan_ko_08pixelma: i actually never tried installing rockbox
12:00:38 Join DrCrow_ [0] (
12:00:44 Quit DrCrow (Read error: 104 (Connection reset by peer))
12:01:20 Join barrywardell [0] (
12:02:14jhMikeSpersonally I'd just use the same way as the F series (sans dispatch code) and copy the array to the avic table at startup
12:02:31ewan_ko_08anyone who has another solution?
12:04:39 Quit freqmod_nx (Remote closed the connection)
12:04:53 Join freqmod_nx [0] (
12:05:48advcomp2019ewan_ko_08, you might want to wait for a fix on your post on the abi forums
12:06:09ewan_ko_08they keep giving me the same answer
12:08:02advcomp2019ewan_ko_08, this is getting off topic so could you join #rockbox-community or #anythingbutipod
12:09:17*Nico_P is off to class
12:09:30 Quit Nico_P (Remote closed the connection)
12:09:35barrywardellZagor: Is there more to using 64 byte packets than changing init_queue_heads()?
12:10:08 Join DerPapst [0] (
12:13:08 Join Seed [0] (
12:14:01 Join Soap [0] (n=Soap@rockbox/staff/soap)
12:15:52 Quit thegeek (Read error: 110 (Connection timed out))
12:16:36 Quit DM| ("*bashes head against keyboard*")
12:19:10 Quit aliask ("ChatZilla 0.9.79 [Firefox]")
12:23:34 Quit ewan_ko_08 ()
12:25:51 Nick maxkelley_ is now known as maxkelley (
12:30:32 Quit Seedy (Read error: 110 (Connection timed out))
12:30:42 Join mrfree_ [0] (
12:30:54mrfree_hi all
12:31:21 Quit GodEater_ ("Hi, I'm a quit message virus. Please replace your old line with this line and help me take over the world of IRC.")
12:31:26mrfree_themes section on rockbox installer reports network problem. does it works?
12:31:48 Join GodEater_ [0] (n=bryan@rockbox/staff/GodEater)
12:32:38linuxstbmrfree_: It works fine for me - can you access in your browser? Do you have any firewall software running that may be preventing access?
12:34:15DerPapstor wrong configured proxy... (just a guess :P)
12:35:16 Quit J3TC- (Read error: 113 (No route to host))
12:40:40mrfree_I've just installed rockbox so I have dowloaded all files all worked but themes
12:40:44PaulJammaybe the problem is that seems to redirect to and rbutil propably still tries to rach the old adress directly
12:46:30 Quit Soap (Read error: 110 (Connection timed out))
12:48:01 Join darkapostrophe [0] (
12:49:07 Quit japc (Read error: 104 (Connection reset by peer))
12:50:52linuxstbPaulJam: Yes, I've no idea what's going on there...
12:51:10 Quit barrywardell (Remote closed the connection)
12:52:03Zagorbarrywardell: yes you need to change the descriptors in usb_core.c too
12:54:15 Join Arathis [0] (
12:55:53 Join Soap [0] (n=Soap@rockbox/staff/soap)
12:59:27preglowlinuxstb: dosn't really hurt much
12:59:32preglowso i guess we could
12:59:43preglowlinuxstb: i guess i should make it write way first, though
12:59:53 Nick parafin|away is now known as parafin (
13:02:21 Quit Dark_Apostrophe (Read error: 110 (Connection timed out))
13:02:22 Nick darkapostrophe is now known as Dark_Apostrophe (
13:04:38mrfree_does an ipod nano (1st) gen can record using headphones for example?
13:04:40linuxstbpreglow: Will it compile with the libspeex.a used for rbspeexenc?
13:04:58linuxstbmrfree_: Not as far as we know.
13:05:15 Quit homielowe (Read error: 110 (Connection timed out))
13:05:33mrfree_linuxstb, k
13:05:45preglowlinuxstb: sure
13:10:38mrfree_if I try to start recording on my ipod nano then playback stop working and I need to restart the ipod
13:10:41mrfree_is it normal?
13:14:25***Saving seen data "./dancer.seen"
13:19:25preglowlinuxstb: i'll just add some quick wav writing here, and i'll commit it
13:20:20linuxstbmrfree_: No, but it's a known bug.
13:21:16mrfree_linuxstb, k
13:24:13 Join J3TC- [0] (
13:24:54 Quit kubiixx (Read error: 104 (Connection reset by peer))
13:25:01 Join kubiix [0] (
13:32:23 Quit J3TC- (Read error: 104 (Connection reset by peer))
13:32:55 Join J3TC- [0] (
13:37:01 Quit mrfree_ ("Leaving")
13:38:04DerPapstamiconn: you're here?
13:40:29 Quit CaptainSquid ("Miranda IM!")
13:41:13 Quit Soap (Read error: 110 (Connection timed out))
13:41:42preglowlinuxstb: should rbspeexdec be compiled by default?
13:46:38GodEater_is there a good reason for not compiling it by default ?
13:48:01preglowwell, yeah, most people should never need it
13:48:09preglowbut if we don't compile it by default, most people might never see it either
13:48:13linuxstbI can't see any harm in compiling it - most of the work is compiling the lib
13:48:48preglowtacking on some cheap wav stuff now
14:01:10 Quit spence_ (Read error: 113 (No route to host))
14:01:25 Quit [omni] (Read error: 104 (Connection reset by peer))
14:02:42 Join RaRe` [0] (
14:04:46 Quit Arathis (Read error: 110 (Connection timed out))
14:07:43 Join [omni] [0] (
14:10:47preglowlinuxstb: aren't tools automatically rebuilt if deps change?
14:11:08preglowlinuxstb: i just did a "make" in rbspeex/, and it just seemed to discover that all the files in codecs/libspeex/ had been touched
14:13:18 Join webguest18 [0] (i=c1713011@gateway/web/cgi-irc/
14:14:46 Quit webguest18 (Client Quit)
14:17:26 Quit RaRe (Read error: 110 (Connection timed out))
14:18:10linuxstbpreglow: Yes, a "make voice" or "make voicetools" should update the voice tools.
14:20:22preglowright, but not a regular make
14:20:23preglowmakes sense
14:20:33preglowmy brain might have intended that pun
14:20:43preglowbut i refuse to take responsibility for it
14:21:21 Join spence_ [0] (
14:23:26 Quit Rob2222 (Read error: 104 (Connection reset by peer))
14:24:08 Join Rob2222 [0] (
14:24:48 Quit Jon-Kha (
14:24:48 Quit Slasheri (
14:24:48 Quit Hadaka (
14:25:53NJoinJon-Kha [0] (
14:25:53NJoinHadaka [0] (
14:25:53NJoinSlasheri [0] (i=miipekk@rockbox/developer/Slasheri)
14:27:03 Join stewball`ghost [0] (n=WTFOMGBB@
14:27:23 Quit amiconn (Nick collision from services.)
14:27:31 Join amiconn [0] (n=jens@rockbox/developer/amiconn)
14:35:40 Nick fxb is now known as fxb__ (
14:41:53 Quit hannesd (No route to host)
14:42:47preglowlinuxstb: there
14:45:33 Join fokussiert [0] (
14:46:24 Quit spence_ (No route to host)
14:51:06 Join japc [0] (n=japc@
14:52:16linuxstbpreglow: Do you want to tell your friend on the mailing list?
14:52:50*linuxstb notices you already did...
14:54:08linuxstbWhat do people think of this patch? sdoyon seems keen on committing it.
14:55:34peturhaving an indication what the default is would be nice, but that wouldn't help people who use the voice interface
14:56:57 Join spence_ [0] (
15:01:42 Quit mokkurkalve ("leaving")
15:01:58 Quit stewball`ghost (Read error: 104 (Connection reset by peer))
15:09:56 Join stewball`ghost [0] (n=WTFOMGBB@
15:10:12*preglow is reminded of rbutil...
15:13:38 Join goffa [0] (n=goffa@
15:14:12PaulJamsome forum guy around?
15:14:28***Saving seen data "./dancer.seen"
15:14:49linuxstbWhat kind of forum guy?
15:15:47PaulJamsmeone who knows how i can correct the link in
15:16:59peturlike this?
15:17:42linuxstbTry surrounding it with [url][/url]
15:17:57peturseems to work without it too
15:17:58PaulJamthanks petur
15:18:15*linuxstb probably got there after petur...
15:18:19peturcan't users edit their own posts?
15:18:24linuxstbWhat did you change?
15:18:41peturremoved some [] that didn't belong there
15:19:09PaulJampetur: i didn't expect that the link works without the []
15:19:33peturI've added them now anyway ;)
15:21:02peturPaulJam: if you're logged in, I would expect you to be able to modify your own posts
15:21:08GodEater_yes - all users can
15:21:22GodEater_otherwise it would be pointless us shouting at them about double posting ;)
15:21:24preglowlinuxstb: can you think of any good reasons why a codec might want to alter the metadata struct?
15:21:41*petur sees PaulJam isn't logged in...
15:22:38PaulJampetur: yes, i can modify my own posts, it's just that i didn't think that the link would still work when removing the [] characters.
15:23:42 Join Llorean [0] (
15:24:25 Quit DerPapst (Read error: 110 (Connection timed out))
15:24:32linuxstbpreglow: Maybe some codecs don't have full metadata parsers in apps/metadata/, so the codec does it instead.
15:24:44 Quit goffa_ (Read error: 110 (Connection timed out))
15:25:31preglowlinuxstb: yeah, but that's not a good reason
15:25:57preglowspc.c seems to duplicate most of it...
15:26:14preglowmetadata/spc.c does most of it, then codecs/spc.c happily comes along and does it again
15:26:16linuxstbpreglow: I agree, codecs should ignore the metadata.
15:26:45linuxstbWhy are you asking?
15:27:09preglowdidn't adam gashlin write the spc codec? i wonder why he's posting bugs for his own codec
15:27:13 Quit goffa (Remote closed the connection)
15:27:17 Join goffa [0] (n=goffa@
15:27:20preglowlinuxstb: just because of the spc metadata bug that was posted today
15:28:42Lloreanpreglow: He maybe can't get to it right now? Having the bug in the tracker prevents someone else from reporting it again (in theory), lets it be known it's known about, and serves as a sticky-note reminder if you're busy at the time
15:29:59preglowhe might also just not know how to fix it, but i put in a pointer
15:30:20preglowsince the metadata is garbled two seconds before the end of the track, i almost assume it's got to be something the codec itself is doing to the metadata
15:30:26preglowsince a new track starts decoding about at that time
15:32:28linuxstbIs this when an SPC track follows another SPC track?
15:34:50linuxstbpreglow: Does it look like the codec is reading more metadata than metadata/spc.c does?
15:37:39preglowlinuxstb: didn't have too close a look
15:37:58preglowlinuxstb: i got that impression from the report, yes, regarding the sequence
15:40:20 Join jac0b-work [0] (n=jac0b-wo@
15:40:26 Part jac0b-work
15:43:03 Join hannesd_ [0] (
15:44:36johnf1911question about "time remaining"
15:44:46johnf1911does it build up some kind of battery history over time?
15:44:55johnf1911it's estimates have been increasing since a fresh install / settings clear
15:45:03johnf1911it started saying four fours
15:45:13johnf1911now it says like, 9 hours when I go to shut down
15:45:59 Join spence__ [0] (
15:46:46 Quit spence_ (Read error: 104 (Connection reset by peer))
15:48:35 Join barrywardell [0] (
15:50:17 Join MethoS- [0] (
15:51:30preglowjohnf1911: no battery history, afaik
15:51:42preglowi think the entire feature is useless, it depends on too many factors, at least on swcodec
15:53:53linuxstbI'm sure it could work nicely if someone put some effort into it...
15:54:03markunjohnf1911: are you a programmer?
15:54:23preglowlinuxstb: that certainly doesn't seem to happen
15:54:29Zagorhmm, my c200 bootloader says "model id: e200" after I reformatted the disk.
15:54:49preglowlinuxstb: and it'll need recalibrating every once in a while
15:54:53 Join goffa_ [0] (n=goffa@
15:55:14 Quit spence__ ("See Ya!")
15:55:18 Part LinusN
15:56:21 Join spence [0] (
15:56:36johnf1911markun: Senior sys / net admin, but I can program too
15:57:43linuxstbZagor: The Sansa or Rockbox bootloader?
15:58:05Zagorand now it takes a very long time to boot the OF
15:58:08 Join ApotheoZ [0] (
15:58:17johnf1911preglow: I like it myself, I know about the many limitations, but I've been using it for years, and know the meaning of the values (power up the estimate is low, power down estimate is fairly accurate based on session consumption) also, most of my files are my own rips, same quality / codec
15:58:30johnf1911markun: were you going to link me to the source for that portion?
15:59:07Zagorsomeone with a c200 around?
15:59:08linuxstbZagor: Hmm, IIUC that's read from from rockbox.mi4 file...
15:59:34 Quit spence (Client Quit)
16:00:02linuxstbOr maybe the original firmware MI4...
16:00:14Zagorlinuxstb: yeah, it's OF mi4: /* Read mi4 header of the OF */
16:00:49 Quit goffa (Read error: 110 (Connection timed out))
16:00:52linuxstbOK, that's a bug in sansapatcher - it's writing e200 there, regardless of whether it's a c200 or e200
16:01:40linuxstbLine 595 of rbutil/sansapatcher/sansapatcher.c
16:02:20 Join Arathis [0] (
16:04:57 Quit spiorf (Remote closed the connection)
16:05:01linuxstbZagor: Are you still trying to get large blocks working, or will you enable UMS with 64-byte transfers?
16:05:28Zagori'm working on 64-byte transfers currently
16:05:48 Join iamben_ [0] (n=ben@
16:05:59linuxstbDo you know what speeds can be achieved?
16:06:52Zagortests show around 1MB/s but those are raw transfers with very little processing.
16:07:23Zagorand no disk i/o. so basically I don't know :)
16:07:44linuxstb;) Whatever it is, it's better than nothing which is what we have now...
16:08:25linuxstbAnd it will be nice to be able to try and port it to the Gigabeat S - that MTP nonsense is getting annoying...
16:09:00ZagorI can imagine that. I gotta go, see you later.
16:09:02 Quit Zagor ("Client exiting")
16:09:39markunjohnf1911: battery_time() in firmware/powermgmt.c
16:09:56preglowot: threads in a linux process are supposed to share pids, yes?
16:13:54 Join H10_007quick [0] (
16:14:49 Join DerPapst [0] (
16:14:54H10_007quickbarrywardell: a recent commit says they did something with the H10 button driver, Could you explain what it changed? or was there no functional change?
16:15:11preglowH10_007quick: just made it a bit safer
16:15:31H10_007quickso nothing functional?
16:15:50H10_007quickdid it stabalize the readings anymore?
16:16:53 Quit iamben (Read error: 110 (Connection timed out))
16:17:00 Join scorche|w [0] (n=42c007b2@rockbox/administrator/scorche)
16:19:00 Quit J3TC- (Read error: 110 (Connection timed out))
16:19:49markunpreglow: maybe the new rockbox voice files are what this guy is looking for:
16:24:44 Nick fxb__ is now known as fxb (
16:25:17 Join karashata [0] (n=karashat@
16:29:37 Join goffa [0] (n=goffa@
16:30:09 Join perrikwp [0] (i=98216687@gateway/web/cgi-irc/
16:35:35 Quit perrikwp ("CGI:IRC (Ping timeout)")
16:40:38 Join jgarvey [0] (
16:40:43 Quit goffa_ (Read error: 110 (Connection timed out))
16:41:49 Quit MethoS- (Remote closed the connection)
16:44:41 Join TMM [0] (
16:47:44 Join RaZorbacK [0] (
16:48:01johnf1911markun: thank you
16:48:43 Quit H10_007quick ("ChatZilla 0.9.79 [Firefox]")
16:48:50RaZorbacKhi all, is there someone who could give me the right syntax to use with genlang to generate a file with marked strings for translation (###)
16:49:45RaZorbacKi think it's in the wiki but not way to find it
16:50:42linuxstbHave you seen this page?
16:51:10RaZorbacKargg how do you do to find what you want so wuickly :(
16:51:27linuxstbGoogle for genlang
16:51:48RaZorbacKlinuxstb: good tip, thx !
16:52:14linuxstbSearching the wiki for "genlang" would also have found it - click on "search" at the top of any wiki page.
16:52:53RaZorbacKi've tried .... ok though :)
16:53:28linuxstbDid you use the Search feature, or type something into the "Go" box - it's confusing...
16:53:44RaZorbacKin the go box ...
16:54:13linuxstbThe Go box is for entering the name of a wiki page - e.g. LangFiles
16:54:44RaZorbacKwe have to know the wikiword
16:55:07RaZorbacKso we have to use the other search box
16:55:14RaZorbacKgood to know
16:55:21RaZorbacKthx for your time linuxstb
16:57:24 Quit ApotheoZ ()
16:58:05 Join toffe82 [0] (
17:03:10 Join Nico_P [0] (n=nicolas@rockbox/developer/NicoP)
17:07:47lostlogicNico_P: thanks.
17:07:51 Quit sup (Read error: 110 (Connection timed out))
17:08:10lostlogicI was sad to see how much binary size it cost, but it works nicely.
17:08:54Nico_Pah yes, I hadn't seen... no problem though
17:09:17Nico_Pfunny how the nano and the mini got much more than the others
17:09:33lostlogicyeah and H10, must have to do with a missing feature of that PP variant?
17:09:39 Join hannesd__ [0] (
17:09:41 Quit hannesd_ (Read error: 104 (Connection reset by peer))
17:09:48 Join obo [0] (n=obo@rockbox/developer/obo)
17:09:54 Join |obo| [0] (
17:10:14 Quit |obo| (Client Quit)
17:10:30 Nick hannesd__ is now known as hannesd (
17:11:29linuxstblostlogic: Am I right in thinking that you load the entire file into RAM, then (after the file has been completely loaded) use the metadata parser to strip the tags?
17:12:12Nico_Plinuxstb: yes
17:12:41Nico_Plinuxstb: only the trailing tags are stripped that way though*
17:12:47lostlogiclinuxstb: yeah, not necessarily the whole file is in RAM at that time though −− some of it may have been shrunk-off if it's longer than the buffer or such
17:13:25lostlogicjust whenever a file finishes buffering it gets it's ape/id3v1 tags cut off the end
17:13:58 Join _jz [0] (
17:14:03_jzHello !
17:14:15 Join kubiixx [0] (
17:14:16 Quit kubiix (Read error: 104 (Connection reset by peer))
17:14:20_jzIs there a known problem in the recent builds, with Sansa E200 and MicroSD ?
17:14:29***Saving seen data "./dancer.seen"
17:14:53_jzSince a few builds, the presence of a micro-SD makes everything very slow, and when I get to read something from the SD it keeps skipping and acting like very hogged
17:15:06_jzthough when i get to read something from internal memory it's almost OK
17:15:11lostlogicNico_P: does it make sense to add: if (h == cur_handle) { buf_widx = h->widx; } to bufcuttail?
17:15:12barrywardelllostlogic, Nico_P: the binsize difference is because of RAM usage - the ones with the big bin change have a 2048 byte RAM increase
17:16:23linuxstb_jz: Have you tried reinstalling an older build? If so, does the problem go away?
17:16:38lostlogicbarrywardell: is RAM size specific to buffers or is it including code? I thought it included code
17:17:02_jzlinuxstb: i was asking if the problem was known before trying to do so
17:17:18 Join preglow_ [0] (
17:17:31Nico_Plostlogic: it does... it's not a problem if it's not there and it probably won't bring a hug benefit, but it probably should be there
17:17:54_jzlinuxstb: i'll try the daily first...
17:18:03barrywardelllostlogic: I'm not sure, but maybe it's to do with something that's aligned to 2k boundary not quite fitting?
17:18:04lostlogicNico_P: that's what I thought
17:18:16linuxstb_jz: I don't think it's known, but I may have missed someone mention it... Have you looked in the forums and/or flyspray for similar reports?
17:18:33_jzi didn't have much time lately
17:19:03lostlogicbarrywardell: The only buffers I added to the code are "TAG" and "APETAGX" so I doubt that −− I think it has more to do with a series of array-dereference-shifts involved in ape tag parsing
17:19:58 Quit preglow (Nick collision from services.)
17:20:05 Nick preglow_ is now known as preglow (
17:20:44barrywardelllostlogic: I mean maybe something unrelated to the commit. some other buffer somewhere. maybe not but just a thought
17:21:14_jzother weird problem :
17:21:20 Join krick [0] (i=c27f0814@gateway/web/cgi-irc/
17:21:49_jzfor a few times, each time I reboot i'm on the same track (from internal memory), with "play" on, 0:00, ??? ?bps, and all seems frozen
17:21:58_jzonly solution for now: reboot using crappy OF :(
17:22:00kricklostlogic: why do you make an element wise comparisom and don't use memcmp? (I mean when stripping the tail tags.)
17:22:23_jzneeds 10sec to shutdown, any other button is unresponsive
17:22:25_jzany idea ?
17:22:26linuxstblostlogic: There should already be functions to safely get 32-bit little-endian integers from the buffer - see metadata_common.[ch]
17:22:28lostlogickrick: dunno −− I just copied the code from the old version
17:22:45lostlogicI'll change both of those and it should reclaim some code size. thanks guys.
17:22:57linuxstblostlogic: Looking at them, they should be using uint32_t, rather than long though...
17:23:20linuxstbI guess that would cause problems on 64-bit?
17:23:34lostlogicdepends what you eventually use them for right?
17:24:08linuxstbAnd possibly if they're meant to be signed or unsigned.
17:24:37 Quit krick (Client Quit)
17:26:43lostlogicwell in my case, I just use them as simple signed numbers that are likely to be in the range representable by a short ;)
17:27:51 Join J3TC- [0] (
17:34:27linuxstblostlogic: Thinking about your commit, I'm not sure that changing id3->filesize at a random point during the track's lifetime is a good idea - codecs often use that value for things. e.g. the NSF codec will fail with an error if it can't read filesize bytes, and other codecs use it for the upper-bound for seeking...
17:34:59preglowthat does sound nasty indeed
17:36:19 Quit atsea- (Read error: 104 (Connection reset by peer))
17:36:47lostlogiclinuxstb: any codec that would depend on the filesize should also be incompatible with trailing tags, if I understand correctly.
17:37:22lostlogic(moreover, we did this pre-MoB to strip these tags too, with the same potential side effect ;))
17:41:15linuxstbYes, but codecs might read ci->id3->filesize during initialisation, and then later find it's wrong. So the fix is not helping them...
17:42:00linuxstbWouldn't it be simpler (even though it's slightly less efficient) to just check for trailing tags before buffering the file?
17:42:53lostlogicamiconn repeatedly vetoed that *shrug* I have no love of doing it this way, but it works for the test case I was provided and nearly precisely duplicates the pre-MoB functionality
17:42:55linuxstbThat would also mean you could actually use the tags - e.g. support APE in MP3, which is very often asked for (even though many devs don't like the idea).
17:43:39linuxstbI'm just wondering exactly how expensive one extra seek/read per mp3 file will be.
17:43:47preglownot very...
17:44:06preglowi think i'd prefer that, it's way less hacky
17:44:23linuxstbIt also removes those two new functions from the buffering API...
17:44:40markunlinuxstb: what was amiconn's reason against it?
17:44:43preglowwhat does the newly commited way do? remove them when reading the last chunks?
17:45:03lostlogicpreglow: when a file finishes buffering it calls a callback which strips the tags from the just-finished handle
17:45:10preglowthen change filesize? that's nasty
17:45:19lostlogicand yes, it then changes the filesize
17:45:31Slasherihmm, sounds very bad..
17:45:32preglowis there any reason codecs need to know filesize, btw?
17:45:45linuxstbmarkun: IIRC, the fact that an extra seek/read is needed... Hopefully he'll correct me if there was more.
17:46:13lostlogicSlasheri: this is exactly what the playback engine pre-MoB did.
17:46:26preglowit is?
17:46:29lostlogic(only it wasn't a callback, it was coded into playback.c)
17:46:43linuxstbpreglow: Some use them for seeking. Some read the whole file into a local buffer.
17:46:47lostlogicthe sequence of events is identical
17:46:50preglowincløuding changing filesize?
17:47:00 Join RonOverdrive [0] (
17:47:05preglowwell, shrug
17:47:17preglowthen no further harm is done, at least
17:47:20linuxstbWe don't want to duplicate everything the old playback engine did though...
17:47:23lostlogic tracks[track_widx].filesize -= len;
17:47:25preglowwe could instate some kind of policy on filesize usage
17:47:27lostlogicwas the LOC
17:47:34preglowlike, only depend on during init
17:47:44preglowand then be aware it might be wrong compared to what you'll get...
17:47:45_jzlinuxstb: yes it works with an old rockbox build !
17:48:38linuxstbI still think it would be cleaner and simpler to just check it before buffering though, with the advantage that you could use the tags.
17:50:05lostlogicwell now that the code for stripping them is over in metadata.c it should be somewhat easier to do that ;)
17:50:13*lostlogic hides (goes to work)
17:50:46linuxstbBut that's relying on the buffering API - the metadata parsers work directly on the files.
17:51:16Nico_Plinuxstb: it would be simpler if the metadata parser did all the work
17:52:41linuxstb_jz: Can you identify the last daily build which worked?
17:54:05linuxstbNico_P: I'm not against it, as it seems the more logical approach. But it depends what others think.
17:54:23Nico_Pof course
17:54:31*lostlogic is strictly neutral, but has no inclination to reimplement it
17:56:40_jzlinuxstb: no I can't. i took an old one stored on my laptop
17:56:56_jzlinuxstb: i'd say roughly between 3 and 6 weeks
17:57:24_jzi'll stick to that old version for some time i think...
18:02:22_jzlinuxstb: a quick search in the forum doesn't bring any result
18:02:34_jzi'll look in flyspray
18:02:43RonOverdrivehi I was wondering if someone can help me with my 30gig 5.5gen ipod video? I just bought it and installed Rockbox on it without issue, but for some reason I have a hard time syncing it up. In Rockbox mode the ipod crashes within a few seconds and in Apple mode after a few hundred songs it locks up
18:02:53 Join Mathiasdm [0] (
18:03:05 Part RaZorbacK
18:03:55_jzlinuxstb: maybe this one is related :
18:04:05 Join webguest72 [0] (i=4f447077@gateway/web/cgi-irc/
18:04:46_jzwell i didin't know if my problem was database related but the wier filebrowser bein stuck is there too
18:04:55RonOverdrivealso does anyone know a program in Linux that will reconize that I have Rockbox installed to synce formats other then MP3 and AAC?
18:06:42webguest72_jz I had very similar problems after updating a fer weeks back. sluggish loading of tracks & skipping audio. A complete removal of Rockbox & Bootloader & reinstall sorted it.
18:07:13_jzso maybe it's just when upgrading rockbox ?
18:07:31_jznow I wiped out rockbox (lost my customized theme and settings, what a jerk ;) and put that old rocbox and it works...
18:07:53_jzmaybe i'll try putting the most recent update fresh on top of it (backuping it first now ;)
18:11:32 Join bertrik [0] (
18:12:02 Join sup [0] (
18:13:29 Join mf0102 [0] (n=michi@
18:15:17 Quit petur ("work->home")
18:19:38 Quit linuxstb (Read error: 113 (No route to host))
18:20:57 Join linuxstb [0] (n=chatzill@rockbox/developer/linuxstb)
18:21:29DerPapstRonOverdrive: iirc there was an attemp to manage rockboxed ipods with amarok
18:22:06DerPapstbut i don't know if the patch ever made it into the main branch of amarok
18:23:21RonOverdriveno it hasn't, because I'm using the main branch
18:23:29RonOverdrivebut what about my other problem?
18:24:56LloreanRonOverdrive: There is no "Rockbox" mode while syncing on iPods
18:25:09LloreanBoth are Apple USB modes, so if you're having any problems in USB mode, there's something wrong with your iPod, or the apple firmware
18:26:20amiconnmarkun, lostlogic: The reason for not checking trailing tags before buffering is the nasty seek involved
18:26:58preglowand changing "ci->filesize" during the lifetime of the track is any less nasty???
18:27:10markunamiconn: nasty in what way?
18:27:12RonOverdriveis there a way I can re-install the apple firmware in case something went wrong with the update?
18:27:16preglowmarkun: nasty in being slow
18:27:19amiconnA seek is somewhat costly
18:27:50preglowamiconn: metadata handlers seek around like mad anyway
18:27:57preglowone more seek isn't going to matter much
18:28:15amiconnThat is, if that seek needs to happen anyway because there is no id3v1 tag, the metadata reader could mark the trailing tag for removal. But if there is an id3v2 tag, it doesn't, and shouldn't, check for trailing tags
18:28:41amiconnpreglow: The mp3 metadata reader does *not* seek to the end if it finds an id3v2 tag
18:29:20pixelmamarkun: do you know any tips and tricks to make espeak+mbrola output less noisy? Lame isn't that good in compressing it and so the voice file ends up too big for my Ondio (testing swedish, the mp3 voice file for Ondio is twice the size of the speex voice file for my c200 using the default settings, can't use it this way...) :\
18:29:39Slasheriamiconn: how costly one seek is, does it take for example more than 0.5s to seek end of a 200MB file?
18:29:51preglowme brb
18:29:54Slasheriif not, i don't think that matters when buffering
18:30:25amiconnDepends on the target, I think
18:31:03 Nick tuplanolla is now known as Tuplanolla (
18:31:14 Join Doca [0] (
18:31:55 Quit Doca (Client Quit)
18:32:10markunpixelma: no, sorry
18:32:12 Quit webguest72 ("CGI:IRC (EOF)")
18:33:12markunmaybe you can filter noise with a tool like sox
18:36:27pixelmahmm, ok. Thanks anyway.
18:38:32RonOverdrivequestion, will updating the firmware remove the ipodbootloader?
18:40:27markunpixelma: do you like the quality otherwise?
18:41:13linuxstbRonOverdrive: Updating the Apple firmware will remove the Rockbox bootloader, if that's what you're asking.
18:42:01 Join nicktastic [0] (n=nick@unaffiliated/nicktastic)
18:44:36scorche|wDerPapst: actually, i think i talked about that a bit with the person from amarok at google...dont quite recall the results though...
18:45:39pixelmamarkun: yes, pronounciation sound quite ok to my "foreign" ears overall, this noise is a bit disturbing. Listening to some samples I got the impression that swedish is even better than some english voices by them.
18:48:01pixelmambrola really makes a difference with regards to sounding "robotic"
18:51:11DerPapstscorche|w: heh...
18:52:33DerPapstbut i don't think i would use some music management tool to sync my music for rockbox
18:52:40scorche|wyou want me to ask him about it?
18:52:46scorche|wah...alright :)
18:53:37 Join Domonoky [0] (
18:53:37DerPapstbecause i don't think amarok would build the database when transferring files
18:54:07Nico_Plinuxstb: I'm having trouble getting avic-imx31 to link... it says "undefined reference to `_avicstart" in avic_init. I've been trying to make it work but haven't managed, and I don't quite understand the intent of the asm line... do you think you could have a quick look?
18:54:10scorche|wfrom what i do remember, they have gotten tons of requests for a plugin that will do that
18:55:22DerPapstimo such a thing would only be usable if it builds the database while transferring files so that one doesn't have to do it on the targed. otherwise it's kinda useless
18:56:59 Join webguest66 [0] (i=18ebd67f@gateway/web/cgi-irc/
18:57:04webguest66hey folks
18:57:45webguest66glad to have found rockbox - but have it all set up and now can only access database - but no files are in my file list, no vids or no music - just blank...................?
18:58:13karashatatry setting your file view to all instead of supported
18:58:22karashatathe folder your music is stored in is probably hidden
18:58:35webguest66i'll give it a shot
18:59:20webguest66the thing is i can see it in the database view
18:59:58karashatahmm, what DAP are you using?
19:00:00 Quit RonOverdrive ("Leaving")
19:00:11LloreanThe database will happily index hidden files..
19:00:58webguest66still nothing
19:01:07webguest66sorry, what's a DAP?
19:01:16karashatadigital audio player
19:01:26karashatamp3 player, basically
19:01:48webguest66ipod vid 5gen
19:01:49nanoksansa, i would guess
19:02:10karashataahh, the music was loaded onto it with iTunes, no doubt
19:02:38webguest66just got it all moved into the database - that works well
19:02:42karashatawhich means your best bet, unless you want to reload your music onto it via drag-and-drop, is to use the database
19:02:54karashatasince iTunes manages files strangely
19:03:04webguest66ya that's why i'm glad to be done with it
19:03:36webguest66so, anything i load on there via drag and drop will likely show up in the file view?\
19:03:47karashatait should, yes
19:03:58webguest66makes sense - thnx a lot
19:07:23 Join webguest55 [0] (i=3b5c974a@gateway/web/cgi-irc/
19:08:07webguest55I get this message when try rockboxui.exe
19:08:08webguest55No filesystem found. Have you forgotten to create it?
19:08:23webguest55how do I create the file system?
19:08:43DerPapstmake install?
19:09:21linuxstbwebguest55: Did you download rockboxui.exe or compile it yourself?
19:10:47webguest55I compiled it myself
19:11:09webguest55and then copied to the players root directory
19:11:15linuxstbThen type what DerPapst said - "make install"
19:11:19 Join hcs [0] (n=agashlin@rockbox/contributor/hcs)
19:11:34linuxstbWhy did you copy it? You're meant to run it from where it's created.
19:11:35rasherwebguest55: it won't work with the player's root directory at all
19:11:56rasherWell, some parts will, but lots of things will break.
19:12:17DerPapstyou build it in ./rockbox/build and you're supposed to run rockboxui.exe from there
19:12:46webguest55you mean .rockbox/build
19:13:11DerPapstbut you have to "make install" first so that all plugins and codecs are copied to the correct location in ./rockbox/build/archos/.rockbox
19:13:38 Join elinenbe [0] (n=elinenbe@
19:13:56webguest55yes I did this and then it created a and the rockboxui.exe
19:14:08DerPapst./ refers in my example to the directory where you downloaded the rockbox source code to
19:14:30***Saving seen data "./dancer.seen"
19:15:01linuxstbwebguest55: Do you have the "unzip" program installed? Just type "unzip" in the terminal to test.
19:15:08webguest55I then extracted the into the players root folder along with the rockboxui.exe
19:15:25DerPapstso in cygwin in ./rockbox/build type "make install" and then "./rockboxui.exe −−background"
19:15:58linuxstbwebguest55: The UI sim doesn't need to go anywhere near your device - it's independent (and will conflict with the real Rockbox install on your device).
19:15:59DerPapstyou don't use the sim on your player at all
19:16:01pixelmaI'm completely confused - what do you want to compile: a simulator or a target build?
19:18:09Nico_Plinuxstb: any chance you could help me out with that small bit of asm?
19:18:23DerPapstsimulator is the thing that runs on your pc and target build is (obviously) the thing that runs on your DAP
19:18:35 Join l7 [0] (
19:18:55webguest55a simulator, my player does not have a screen so I'm trying to use this software to see if I can play the songs through this ui
19:19:29DerPapstwhat player do you have?
19:20:06linuxstbNico_P: I'm not very good with inline assembler... My guess would be that that's code from Linux which hasn't been fully ported to Rockbox.
19:20:15webguest55it looks like a shuffle but its a fake
19:20:49Nico_Phmm... I've looked at gcc asm docs and still can't figure out how to write it properly
19:20:57karashatawebguest55: it's an unsupported player, you won't be able to run rockbox on it
19:21:11karashatathe players you can use rockbox on are listed on the front page of the site
19:21:25 Join Chronon [0] (
19:21:29linuxstbNico_P: Can you find where _avicstart is defined?
19:22:03Nico_Pmy guess is that it's supposed to the the avicstart int that is declared just above
19:22:44preglowi'm decent at inline asm
19:23:16linuxstbpreglow: target/arm/imx31/gigabeat-s/avic-imx31.c - look for avicstart
19:23:20Nico_Ppreglow: then maybe you could help me with avic-imx31.c... the asm in avic_init is wrong
19:23:47linuxstbNico_P: I don't think the asm is wrong, it's just that it's referring to a symbol not defined in Rockbox...
19:23:47Nico_P"undefined reference to `_avicstart"
19:24:34Nico_Pit still fails if I add _avicstart as an int above
19:24:44preglowand _avicstart is what? a global symbol?
19:24:58linuxstbIt doesn't appear to be defined anywhere in Rockbox...
19:25:06preglowthen what exactly are you trying to do?
19:25:15Nico_Pno idea, but isn't an _ in inline asm a possible optional prefix for a C var?
19:25:53 Join Rincewind [0] (
19:26:00preglowdepends on the platform
19:26:07preglowbut what are you trying to do?
19:26:19linuxstbAs I said, I think this is just a work-in-progress port from Linux... We probably don't need that as a symbol - we can just use a #define to the address of the AVIC vectors.
19:26:23Nico_Ppreglow: I'm not sure really, I guess trying to read a reg val into a C var
19:26:54bertrikI guess it tries to read the address of an external asm function into a C var
19:27:15preglowsomeone send me a gigabeat s, please :>
19:28:25Nico_Ppreglow: ask toffe82 :)
19:28:34preglowNico_P: that line, as it stands, loads the symbol _avicstart into the variable avicstart
19:28:44preglowbut if you don't know what _avicstart is...
19:29:31 Quit J3TC- (Read error: 104 (Connection reset by peer))
19:29:45 Join male [0] (
19:29:45Nico_Ppreglow: what kind of symbol is _avicstart then?
19:29:54preglowwhy don't you just do it from scratch instead of using some code you don't know what is?
19:30:01preglowNico_P: i have no idea
19:30:02 Join J3TC- [0] (
19:30:09preglowit's an address to something, but god knows what
19:30:25Nico_Ppreglow: ok... at first I simply wanted to try that code out
19:30:30bertrikI guess it's a label for the first entry in the vic table
19:30:39 Quit linuxstb (Read error: 104 (Connection reset by peer))
19:30:54 Quit Febs (Read error: 110 (Connection timed out))
19:30:55preglowwhy in the world use inline asm for that simple thing? just so they can alias a variable name?
19:31:10Nico_PI was going to ask the same question
19:31:13bertrikpreglow: my thoughts exactly
19:31:30 Quit hcs (Read error: 113 (No route to host))
19:31:44 Join hcs [0] (n=agashlin@rockbox/contributor/hcs)
19:32:38preglowhcs: there?
19:33:15hcspreglow: yeah, and I saw your note, I suck
19:33:34Nico_PI'll try to understand that code better globally and then maybe I'll be able to make it simpler and correct it
19:34:42Nico_Ppreglow: to answer your question, I don't just do it from scratch because I wouldn't know where to start. I'm very inexperienced in low level stuff
19:36:24preglowNico_P: what other people are porting?
19:36:28preglowhcs: you tried to remove it?
19:36:45Nico_Ppreglow: aliask and linuxstb. jhMikeS is getting one
19:36:58preglowok, then you'll have a low-level guy onboard soon :>
19:37:09Nico_Pthere is ptw419 too but he's not around often
19:37:12hcspreglow: not yet, will do in a few minutes
19:37:15 Quit _jz (Read error: 113 (No route to host))
19:37:21preglowhcs: not guranteed to be the reason, but it's bad anyway
19:37:33preglowhcs: the entire "two seconds before track change" deal kind of makes it sound like the reason, though
19:38:37 Join _jz [0] (
19:39:11 Quit webguest55 ("CGI:IRC (EOF)")
19:41:01preglowSlasheri: is there any reason why dircache should just disable itself?
19:41:12 Join miepchen^schlaf [0] (
19:41:46 Join Frazz [0] (
19:41:54preglowSlasheri: suddenly the debug screen says it's not initialized, and the last used size isn't even _near_ the limit
19:42:52preglowanyone know why the status bar is so slow in appearing in newer builds?
19:44:09Lloreanpreglow: I asked about that recently but nobody even replied at the time. It's been happening for a few weeks now, I think.
19:44:39 Join linuxstb [0] (n=chatzill@rockbox/developer/linuxstb)
19:46:38preglowLlorean: it has indeed, and it's pissing me off
19:46:50preglowthe player is unresponsive while it's not there too
19:47:02preglowi can't believe noone has noticed it
19:47:21linuxstbIs this just the Nano, or all targets?
19:47:22LloreanWhat player are you using?
19:47:26Lloreanlinuxstb: Gigabeat F too
19:47:31preglowcan't say i've noticed it on nano
19:47:36amiconnHmm, now that you mention it, I noticed that
19:47:48LloreanIt's especially noticeable if your backdrop contrasts with your set background color, because you get a "background color" line at the top until the status bar fills in
19:47:54amiconnDidn't annoy me so far, hence forgot again
19:48:51amiconnSeems swcodec specific though
19:49:10PaulJami have seen the dircache disabling itself occassionally too, but never found a way to reproduce, and it didn't happen often.
19:49:12preglowjust tried nano, doesn't happen there
19:49:24Slasheripreglow: if dircache founds an internal problem, it disables itself
19:49:37Slasherithat will be logfd
19:49:41Lloreanpreglow: Could it be either some sort of setting, or some how tying into the presence of a disk then?
19:49:48preglowperhaps i should run a logf build, then
19:49:57preglowLlorean: i have no idea, i haven't touched my settings
19:50:19preglowSlasheri: btw, still no idea about why foreground scans happen so often?
19:50:39nanokpreglow: a quick receipe to reproduce? i could try on my sansa (haven;t noticed it so far)
19:50:41Slasherihmm, nope.. i haven't experienced that
19:50:43*amiconn never sees foreground scans without reason
19:50:49preglowamiconn: i see them all the time
19:51:11preglowi think there's a bug filed
19:51:20preglownanok: reprouce what? statusbar? dircache?
19:51:21amiconnI see foreground scans happen sometimes, but there's always a reason. Mostly hard reset before
19:51:31nanokpreglow: status bar, sorry
19:51:37preglownanok: i just switch mine on
19:51:40Slasheripreglow: if dircache fails runtime, then next boot might cause a foreground scan
19:51:45nanokyou allready said dircache you don;t know yet how to reproduce
19:51:55nanokdamn ;
19:51:59preglownanok: but no, i can't remember having done anything to provoke it
19:52:19nanokpreglow: okay, so just turn on.. let's see
19:52:35*amiconn has a suspicion
19:52:41preglowamiconn: backlight bug posted, might that be resulting from your stuff? isn't too specific...
19:53:06amiconnWhat backlight bug?
19:53:26nanoknope. 4 seocnds or so to boot, responsive right away, nothing wrong visible. but i am using a plain text wps, don't know if that matters?..
19:53:31*DerPapst had a backlight bug too
19:53:39Lloreanamiconn: There's some sort of bug relating to the backlight timout
19:53:56preglowamiconn: just posted in the tracker
19:53:59LloreanAfter the first backlight timeout on boot, it "stays off" until the user goes to the menu and sets a new timeout.
19:54:04preglowi just assume you're all on the bugtracker ml
19:54:06LloreanThen everything goes back to normal, and stays normal
19:54:09amiconnProbably the fade-in/fade-out times. Readjust them,
19:54:11DerPapstif you enable fade in and fade out it might happen that it stays off or on
19:54:17Lloreanpreglow: RSS feed here. :) But same diff
19:54:38amiconnpreglow: I'm not on the bugtracker ml. Never were.
19:54:59preglowLlorean: prefer ml, that way i can keep bugs/features i care about hanging around until i resolve them/i have them resolved :)
19:55:01 Join Redbreva_ [0] (
19:55:15preglowamiconn: 8245, then
19:56:04 Quit webguest66 ("CGI:IRC")
19:56:18nanokLlorean: speaking of reboot when usb plugged on sansa: it does try to do it, just checked now, but it only crashes the player. i will update right away and test again
19:56:44nanokcharging also doesn't seem to really work
20:00:22nanokLlorean: reboot works fine it seems with the current build. but i only tested once
20:01:01 Quit Redbreva ("Miranda IM! Smaller, Faster, Easier.")
20:01:04*preglow thinks all developers should be on the flyspray ml
20:01:32 Nick Redbreva_ is now known as Redbreva (
20:01:58amiconnLooks like the delayed status bar happens when 3 conditions match:
20:02:43amiconn(1) Voicing of directories is enabled. (2) Directory talk clips are enabled. (3) File type announcement is enabled
20:03:01preglownone of them are enabled here
20:03:05preglowi've got a voice file, but that's it
20:03:13amiconnThat is, if there is any directory in the list so the cursor starts on a directory in the browser
20:03:24preglowthat it does
20:03:58 Quit zicho (Read error: 104 (Connection reset by peer))
20:04:26amiconnI tested on G5.5. If any of the mentioned conditions isn't met, the status bar is (nearly) instantaneous
20:04:34 Join zicho [0] (
20:06:30preglowwell, like i said, i only have a voice file, that's that
20:06:43 Join moos [0] (
20:07:15preglowabsolutely all voice related options are off, but like i've said before, that hasn't prevented me from having rockbox speak from time to time...
20:08:14 Join petur [0] (n=petur@rockbox/developer/petur)
20:08:36preglowa week ago rockbox played a voice clip even though i had all voice stuff disabled
20:08:40amiconnBtw, I have no idea about the backlight bug. Never seen this here. A recipe to reproduce would be helpful...
20:08:40preglowi checked directly afterwards
20:08:41pondlifepreglow: Did it call you Dave?
20:08:45preglowpondlife: roger
20:09:12*pondlife would really like a HAL-9000 english.voice..
20:09:33preglowamiconn: it's probably an effect of a playback related bug, though, there has been some glitchiness around track changed since mob was introduced
20:09:47preglowi've had codecs crash, dircache disable itself, wrong files get played
20:11:11 Quit japc (Read error: 110 (Connection timed out))
20:12:56preglowamiconn: removed voice file, status bar is still slow
20:13:36amiconnpondlife: Btw, I re-checked what is wrong with the L&H SAPI5 voices here
20:14:05amiconnThey work for admins, but not for restricted users
20:15:00pondlifeBut this doesn't apply to the MS voices?
20:15:10pondlifeThe MS Reader installer must be borked
20:15:11amiconnNo, and not to my AT&T voices either
20:15:24 Quit LycoLoco (Read error: 104 (Connection reset by peer))
20:15:48 Quit J3TC- (Read error: 104 (Connection reset by peer))
20:16:04preglowNico_P: btw, working any more on completing cuesheet conversion to mob?
20:16:22amiconnI get "This voice cannot be played. Please try selecting another voice or selecting a different audio output device" when I try to select an L&H voice as user in the speech panel
20:16:26 Join J3TC- [0] (
20:16:30Nico_Ppreglow: I've started, but not worked on it for a few days
20:16:39amiconnAll other SAPI5 voices work
20:16:42preglowNico_P: ok, good to know you're on it anyway
20:17:22 Quit zicho (Read error: 104 (Connection reset by peer))
20:17:32 Join zicho [0] (
20:19:55preglowbtw, am i the only one that finds the "backlight on hold" setting silly?
20:20:04hcspreglow: crap, I don't have the sansa cable on me
20:20:39hcspreglow: not really, if you want the backlight on all the time except when you're not doing anything
20:22:20preglowwhy would i want backlight always on just because my hold switch is on?
20:22:21preglowi don't get it
20:22:44amiconnThe 'always on' setting is probably superfluous
20:22:49hcsoh, that particular setting
20:23:09amiconnThe other variants are somewhat useful
20:23:37preglowamiconn: what about always off? why do we need that? so you can save battery by disabling the backlight at once when you switch hold on?
20:23:44preglowi think that setting is a waste of space
20:23:58preglowit does nothing very usable
20:24:30 Quit obo (Read error: 110 (Connection timed out))
20:24:33 Quit axionix (Read error: 104 (Connection reset by peer))
20:25:20bertrikwell, what about the upside down function then? It's a nice hack, but totally useless IMO.
20:25:21 Join axionix [0] (
20:25:33 Join obo [0] (n=obo@rockbox/developer/obo)
20:26:43*amiconn finds 'upside down' quite useful from time to time
20:26:46bertrikoh, it also turns the buttons upside down, nice
20:26:50preglowbertrik: not as useless
20:26:57preglowbut i'd rather remove that option as well, of course
20:27:01moospreglow, amiconn, bertrik: this way, sure we could find users hating one seetting, and this probably for quasi all options
20:27:03preglowand that one actually eats quite a bit of space
20:27:03hcsbertrik: what if you live in australia?
20:27:22preglowmoos: i don't intend to ask all users
20:27:41Lloreanamiconn: I find "Always On" quite useful when combined with a dock.
20:27:43nanoki use upsidedown all the time
20:27:45preglowif we asked all users what settings to use, we'd be bloated beyond anything now
20:27:46moospreglow: I mean all features are used, and all users have her resistents ;)
20:27:56preglowlikewise, i am not going to ask all users when removing settings
20:28:02preglowmoos: sure, but i don't care
20:28:05pixelmabertrik: someone in the forum said he used his sansa e200 upside down for the controls
20:28:26preglowwe are the developers, not the users
20:28:31preglowusers are free to patch their builds however they want
20:28:39moospreglow: teah but that's a taste thing
20:28:47amiconn'upside down' is helpful if you have your dap sitting on the table. Depending on the orientation of the earphone connector, you can move around a bit more (and avoid kinks in the cable)
20:28:54pixelmaand I could imagine that it is useful on the c200 for left handed people
20:28:57moosIf a user want an always baclight option???
20:29:50preglowmoos: we have aa "backlight always on" option
20:29:55preglowmoos: i don't want to remove that, i think that's a good option
20:29:59 Join Zagor [0] (
20:30:14Lloreanpreglow: Always Off is pretty good too, if your DAP has a screen that can be seen without the backlight on.
20:30:14nanoki think upside down is also usefull for certain dap's when worn on that string around the neck: for the sansa e200, for example, fi you use the upside down option, it is painless, otherwise it is really hard
20:30:18moospreglow: ...even if w don't use it :)
20:30:21 Join Davide-NYC [0] (
20:30:33amiconnAnd 'upside down' is cheap - it's only implemented on targets where the lcd controller allows reversing the update directions
20:30:35preglowLlorean: sure, but that's a backlight option, no? not a "backlight on hold" ?
20:30:56amiconnSo all it needs is a little setup commands, on some targets maintaining an offset variable, and the button flip filter
20:31:04 Quit hcs ("Leaving.")
20:31:04Lloreanpreglow: Oh, missed the "on hold" bit.
20:31:17moospreglow: missed the always off here :)
20:31:21Lloreanpreglow: Always Off is not as useful on hold I guess (unless it overrules the "on plugged in" option)
20:31:23preglowLlorean: "backligh" is fine by me, i just think "backlight on hold" is useless
20:31:32LloreanBacklight on Hold is used by me
20:31:37preglowLlorean: why?
20:31:41preglowand what for?
20:31:43LloreanI put my DAP in a dock
20:31:43 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
20:31:48LloreanBacklight always on with the speakers
20:31:48 Join miepchen^schlaf [0] (
20:31:56preglowLlorean: what does that have to do with hold?
20:31:57LloreanWhen I go to sleep, I flip the hold switch so the screen is darkened
20:32:08LloreanMusic doesn't stop, but no annoying glow
20:32:11LloreanI might be a special case though
20:32:16preglowok, so you're using the hold switch as a light dimmer
20:32:18bertrikheh, so maybe another option: "backlight on dock" ? :P
20:32:34preglowyou can do that just fine without this setting
20:32:37Lloreanbertrik: That's "backlight while charging"
20:32:45Lloreanpreglow: How?
20:33:01preglowLlorean: by switching backlight setting?
20:33:11PaulJamplease don't remove the backlight on hold option. i really like it.
20:33:23preglowLlorean: it takes more time, but i really think we shouldn't bother with settings that duplicate functionality just so you can press three buttons less during a day
20:33:32 Join jhulst [0] (n=jhulst@unaffiliated/jhulst)
20:34:36Lloreanpreglow: I do agree the setting is of arguable value, but I think it should be put on a "Can be removed if necessary" list rather than a "We should go ahead and remove it" list.
20:34:46amiconnLooks like every option has its users, and usage cases
20:35:02moosamiconn: Amen, taste thing
20:35:10Zagor"if you build it, they will come" :-)
20:35:15preglowi'm not contesting that, i'm just saying that's not a reason to not remove some
20:35:17amiconnIf it was me to decide, I would remove a bunch of options, but others than those mentioned so far
20:35:28bertrikamiconn: which ones?
20:35:43 Quit Davide-NYC ("ChatZilla 0.9.79 [Firefox]")
20:36:07preglowi'd just love to remove these esoteric fringe options that really just duplicate functionality
20:36:09Lloreanpreglow: I do agree with the fact that "just because people like it, it's not a reason to remove it", but at the same time if there's no use for "Backlight on Hold", then there's no use for "Backlight while Charging" either.
20:36:21LloreanOr rather, there's no reason to keep it, not "No use"
20:36:22preglowLlorean: i'd love to remove that...
20:36:26preglowLlorean: but it's marginally more useful
20:36:44LloreanI'd say "On Hold" is actually more useful.
20:36:50PaulJami use the hold switch to turn backlight on or off. if you despertaely want to remove a setting, why not remove the "sort case sensitive" option. I'd be surprised if someone uses it.
20:37:01Llorean"On" on hold means that you can use it to show the WPS without "First Keypress" methods.
20:37:05amiconnbertrik: Fade on stop/pause, crossfade, crossfeed, curesheet support, replaygain...
20:37:15pondlifeBacklight on charge makes a lot of sense
20:37:15nanoki have a hunch every developer would remove other options. this is probably why of's tend to seem lacking to most people: they contain only what a small group of individual thought meaningfull, and i honestly think this is where rb differs so much, being a community effort
20:37:18moospreglow: more useful for *who*? ;)
20:37:27preglowmoos: for me, of course
20:37:30amiconnOh, not to forget the sw eq
20:37:32preglowi can't speak for anyone else
20:37:35pixelmaZagor: to comment on your question earlier, the bootloader I have also says "model id: e200" when starting the original firmware (starting Rockbox is not verbose). It's one that was made with barrywardell's bootloader commit which disabled the OF's database refresh.
20:37:52Lloreannanok: A balance does have to be maintained with Rockbox. It's a semi-constant war against feature creep
20:37:52pixelmaon e200 that was
20:38:15Lloreanpreglow: On a semi-related note, what's the reason for wanting to remove them? Complication in the code, or simply "Keep the menus streamlined"?
20:38:25Zagorpixelma: can you run "fdisk -l" on your c200 and tell me what partition type you have?
20:38:42pondlifeLlorean: That reminds me, any progress on the menu proposal?
20:38:51preglowLlorean: "keep the menus streamlined"
20:38:57preglowsimplicity is good
20:39:04Lloreanpondlife: Not so far. I've got the printout still, but I haven't managed to go at it with a red pen.
20:39:09preglowand right now rockbox isn't very simple at all
20:39:15Zagorpixelma: actually, ignore that. partition type isn't changed by mkfs
20:39:25Lloreanpreglow: I think then that the backlight settings at least are well named, and not particularly confusing to the user.
20:39:33nanokLlorean: i do understand that, but i was trying to think out loud, and figure out some decent criteria which could be used. i guess space(if it si so important, but on current dap's i think it is not so much, they tend to have plenty of space)+overhead of maintainig the feature - how usefull it is (but how do you determine this one "objectively"?)
20:39:38Lloreanpreglow: There's a lot of other places were the confusion slips in. :)
20:39:49preglowsure, just mentioning things i notice
20:40:16LloreanHonestly, I'd like to see an improved menu system of some sort, but I haven't come up with a proposal yet. Separate from reorganization, that is.
20:40:18preglowi use so little of rockbox i usually don't see much
20:40:35moosthe only thing we can be sure, no one dev:user use all the plenty options :)
20:40:41 Join Arathis2 [0] (
20:40:50pixelmaZagor: and it only takes longer to boot into the OF if it detects a microSD - haven't observed changes in boot time yet
20:41:11Lloreannanok: Space isn't a big issue. Code size and RAM usage is, in a very, very minor sense (Most modern DAPs have plenty of RAM, but there's no rule that says a flash player won't have only 1mb, so there's no sense wasting RAM we don't need to), code complexity (hard to maintain), and user confusion (overwhelmed by features and their interplay)
20:41:20preglowLlorean: just coupled with a wish for rockbox to approach settings a bit different than it usually has
20:41:30preglowseems to me developers pretty much just add stuff they want
20:41:38pondlifeThe display and theme options really need sorting out...
20:41:59pondlifePlus the two menus both called System always confuse me for a short while
20:42:02pixelmayes, and they don't help getting the manual up to date :)
20:42:04Lloreanpreglow: How would you do it, then?
20:42:20preglowLlorean: i'd want more discussion, that's for sure
20:42:28preglowLlorean: but of course, you can't overdo that either, people tend to never agree
20:42:30Zagorpixelma: ahh, found the OF internal formatting function. that fixed my problem.
20:42:38nanokLlorean: wouldn't it be possible to have a system for disabling features at runtime or buildtime to save ram, instead of removing them alltogether, and to make it simpler also, in the process?
20:42:44pondlifeWe need a document to show a proposed layout (on the wiki?)
20:42:50Lloreanpondlife: I'd really like "related" settings to be nested. Like, first you choose "Backlight" then you see options for "On Hold" "On Battery" and "On Charger" rather than three long setting names... sorta, more grouping, and less long lists.
20:42:59preglowand that's if you manage to get people to discuss it at all
20:43:03pixelmaZagor: interesting...
20:43:11Lloreanpreglow: More discussion is definitely needed, especially since most of the basics are covered and we're getting into more fringe feature territory all the time
20:43:12nanokalso, there is a saying about not needing to make it too simple: if you assume your users are idiots, you will soon have idiots for users :)
20:43:31pondlifeIt's not over-complex, just badly laid out IMHO
20:43:51amiconnLlorean: The problem with that approach is that you then need a way to know where you are in the tree
20:43:53preglowLlorean: yup
20:43:56pixelmaLlorean: and then there are remotes with backlight settings - part of the "backlight" submenu or an own "remote" submenu?
20:44:15Lloreannanok: Some features are disabled at build time (RTC for H120), but one of the goals is to try to have Rockbox be as similarly functional as the hardware permits. And most cases, a new *feature* won't get rejected, but a new way to do something that can already be done, will be.
20:44:56Lloreanamiconn: That's why I haven't come up with an actual proposal yet. I haven't yet come up with a way that's less problem-filled than the current way.
20:45:01*bertrik thinks the current menu structure is quite OK
20:45:18nanokLlorean: that i understand, at least to some extent, it is obviously not reasonable to have useless complications
20:45:19Lloreanpixelma: Backlight->Main/Remote->On Batter/On Charger/On Hold
20:45:32amiconnI think the current menu system works fairly well, just a few options need reordering
20:45:53Lloreannanok: Well, one thing, keeping it simpler makes the code easier to maintain. Having 40 options that do the same thing might "simplify" it for people who have a preference for one of them, but probably confuses new users. Etc.
20:46:07preglowLlorean: but existing ways to do soemthing that already can be done shouldn't be removed once they're already in?
20:46:09 Quit hannesd (
20:46:12Lloreanamiconn: I didn't say the current one is bad, mind you. I'm just not sure it's the best we can have. :)
20:46:26amiconnThe table-driven settings are macro hell, but once you get used to them, they're not that difficult to handle
20:46:36Lloreanpreglow: See, that's where it gets iffy. Taking things away from users is always problematic. Or even changing them (view the shouts of outrage about the dB volume, or keymap changes).
20:46:56nanokLlorean: that kind of "complication" should be configurable via a text config file imho, it shouldn't be something that has to be "builtin", as much as possible
20:46:57 Join spiorf [0] (
20:47:00amiconnAnd the result is quite compact... although there is some non-const stuff that should be const and I can't figure out why it isn't :/
20:47:05preglowLlorean: i just believe we should go by the same arguments for stuff that's already in and stuff that someone wants in
20:47:08pondlifeThose menu macros are not very readable.. and they seem to get out of patch-sync quicker than most code..
20:47:10Lloreanpreglow: I'd say "Take 'em out if it makes the code easier to maintain, or if they're definitely introducing confusion, leave them if they're simple enough to have it be merely preference whether they're in or not"
20:47:17preglow"i'm used to it" isn't a good reason
20:47:39Lloreannanok: How do you mean? I'm not sure I understand that statement
20:47:58Lloreannanok: We *really* don't like options that can only be accessible via text config. These are portable devices, and should be configurable wholly from within the device
20:48:01pondlifeIMHO all settings should be settable without needing a text editor.
20:48:20pondlifeSome of the icon stuff needs improving their IIRC
20:48:32preglowwe should rather try making backlight handling on hold always doing the right thing, like i wouldn't really mind if the "backlight on hold" off setting was made default
20:48:39pixelmaLlorean: that could be confusing (or needs a special case) for targets without remote. Just what comes to my mind, I'm not saying that the idea is bad or so, in fact it's the opposite.
20:48:49amiconnWell, making wps'es requires (and always required) a text editor
20:48:55Lloreanpreglow: And that's a very valid point of view. And one I'm *almost* in favour of. But you have to admit, "not giving them the option in the first place" is very different from "letting them develop a habit, and taking it away". So I think while they should be *similarly* considered, the weight is a little different.
20:49:00Zagormy opinion is that we have much more important issues than menu or setting complexity
20:49:09amiconnI think that themes don't need to be creatable on-device
20:49:14nanokLlorean: i mean when the feature is not something new rb can do, but a new way for it to do it, having it in a new part of the menu and so on. probably many of these are in that category. also conditionals could be included there "if hold switch on set backlight to off" or something like that, should be easy to configure on the fly, based on events
20:49:22preglowZagor: indeed, but none that are so easily resolved :>
20:49:23Lloreanpreglow: I certainly wouldn't mind the backlight always being off on hold, but devices with remotes might disagree.
20:49:31 Quit barrywardell (Remote closed the connection)
20:49:36amiconnpreglow: What would "the right thing" be?
20:49:38Zagorpreglow: easy in code, perhaps. but not with consensus...
20:49:43 Quit mf0102 ("Verlassend")
20:49:50preglowZagor: never hurts to try :)
20:49:57preglowLlorean: wouldn't be applied to remotes, of course
20:50:04Lloreannanok: I'm still not quite understanding. You still need to add complexity to the code to support the new methodology.
20:50:13amiconnI wouldn't like backlight being off on hold...
20:50:19nanokrb could come with a default config, which is what the developers agree on, and if any user wants to play with it, they can, on theyr own, using events and conditional statements and such. i am not sure how is that would be to do though
20:50:28Lloreanpreglow: I meant though that the with the H120 on hold, the remote not on hold, some people might want the screen to light up despite that. Though I wouldn't mind overruling them.
20:50:32preglowamiconn: "off" sounds right to me, once you switch hold on, it's logical to assume you're not going to be looking at the display much longmer
20:50:46nanokLlorean: yes, agreed. if it is not deisgned like this in the first place, it will surely not be easy
20:50:50Lloreannanok: Sounds like you'd then need to support a scripting language, which is another layer of added complexity...
20:51:02Zagoramiconn: then set "backlight on hold" to "off"
20:51:08amiconn"hold" is for avoiding accidental button events, not for preventing a quick glance on the display...
20:51:16Zagorthe option isn't "turn on backlight on hold"
20:51:17nanokLlorean: yes, but not such a complex one. something similar to the wps i think
20:51:33preglowamiconn: sure, and you do that when you put it somewhere you usually can't see it
20:51:34Lloreanamiconn: The LCD doesn't go off if the backlight's off... and to light the backlight you'd have to press a button, non-hold, anyway, right?
20:51:43amiconnpreglow: Not in the car...
20:51:52preglowaccidental button presses in a car???
20:52:00preglowi usually keep my hands on the steering wheel or gearshift
20:52:01amiconnWell, usually not
20:52:47Lloreannanok: Still, all that does is increase the code size further (strings, code to handle the new options, plus code to handle the scripting) compared to simply statically including the options in the menus. It addresses the "user confusion" problem by hiding it, perhaps, but makes the "code complexity" and "code size" problems worse in the process.
20:52:57amiconnBut imho there is no single "right thing" to do - that's why there is an option
20:53:01Lloreannanok: Not to mention, if you hide an option from users, the vast, vast majority won't ever find it anyway.
20:53:27preglowamiconn: often there isn't, no, but sometimes you can deduce behaviour that is logical and easy to get used to, really eliminating the need for an option in the process
20:54:09amiconnBtw, when that option was first introduced, I didn't like it, because it doesn't really add new functionality, but now that it exists, I can see a few usage cases where it's useful
20:54:59amiconnWell I 'might' agree that it would be logical to switch off backlight on hold, unless it's set to 'always on'
20:55:00NJoinhannesd [0] (
20:55:05amiconnBut then Llorean would complain...
20:55:20LloreanNot really
20:55:27 Quit Seed (Nick collision from services.)
20:55:35 Join Seed [0] (
20:56:03LloreanWhat I'd recommend though is that precedence be "On Battery" > "On Hold" > "On Charger" if "on" conflicts with "off"
20:56:21LloreanThen if you set it Always On for Battery it will always, always be on
20:56:33 Join psycho_maniac [0] (i=psycho_m@
20:56:34LloreanIf you set it to a timer on Battery, Off on hold, On on charger, turning hold on, on charger, will turn it off.
20:56:35preglowbut i don't really see a reason to turn the backlight on when you switch hold on. hold is for keeping accidental key presses from happening, and backlight doesn't enter into that
20:56:51preglowso what the option really becomes is a customizable button thing, and we all know what we feel about that...
20:57:15*amiconn summons JdGordon
20:57:34LloreanOf course, that still means having two setting, "On Battery" and "On Charger" and simply dropping "On hold" but works for me
20:57:35amiconn'hold' is not a button ;)
20:57:47nanokLlorean: well, that is another discussion. a set of configs could be provided with the rb, "default" "full" "minimal", whatever, you get the picture, the full one being intended to showcase what can be done rather than to be used, not to mention the countless theme-makers out there, which would just jump at the occasion to customize them further. so that shouldn;t be a problem
20:58:07 Quit Arathis (Read error: 110 (Connection timed out))
20:58:22scorche|wamiconn: i think he is somewhere above the atlantic
20:58:26Lloreannanok: The problem is that you're still forgetting entirely everything but "the desire to customize"...
20:58:30amiconnAh, hmm :/
20:58:40*pixelma hands nanok some '
20:58:46*amiconn has a question concerning the settings system and voice
20:59:13amiconnI have an idea how it might be possible to handle numeric settings with decimal point
20:59:20Lloreanpreglow: Would you find the order of precedence idea acceptable, with "On Hold" simply being "Always off, unless it's set to always on on battery"?
20:59:20nanokLlorean: about code complexity, ofcourse, it is arguable, and i am the last to be able to asses how much it would affect it, but i think , maybe, on the long run it would be worth it (after it is done, there is less to maintain for every release, and having a more modular system means less gui-related work for the developer who just wants some functional feature added, and so on)
20:59:30 Join MethoS- [0] (
20:59:32lostlogicscorche|w: naw, he's over land now ;), just running up the coast from Cali to here.
20:59:36preglowLlorean: oh, sure
20:59:41 Join przemhb [0] (
20:59:44preglowLlorean: if it's set to always on, then you don't want hold to change that
20:59:44nanokLlorean: yes, i answered in reverse, sorry for the confusion
20:59:51Lloreannanok: How does adding code to allow custom addition and removal of items from menus mean less to maintain?
21:00:14Lloreanpreglow: But at the same time, I'd rather "Always on when charging" be overruled by turning hold on, still.
21:00:23LloreanJust "Always on when on battery" overruling on-hold
21:00:41amiconnThe idea of making options optional was already discussed several times, and turned down
21:01:14preglowLlorean: what, because of your particular usage scenario? problem with that is that i don't think it's very logical
21:01:50nanokLlorean: the point was that you don't need to care about that part everytime you add or change something, the interaction with the gui would be via a "standard" interface, which would be easy to make backwards compatible and have generally more separated from the "functional" part
21:01:58scorche|wlostlogic: ah...he should have met me in cali
21:01:59 Nick Arathis2 is now known as Arathis (
21:02:00preglowLlorean: i can agree to some degree that switching hold on when it's charging might signify something else than you wanting to avoid spurious key presses, though
21:02:02jhMikeSamiconn: do you have any idea about when this p5002 stuff was originally reported?
21:02:03nanokLlorean: uhm, it is an endless discussion, it was just an idea
21:02:26Lloreanpreglow: I know I'm not the only one who uses the hold switch to turn off the backlight when plugged in. But I think an order of precedence with "On battery" always-on always overruling the "Off on hold" or "Whatever you set on-charger to" doesn't seem that illogical
21:02:28nanokand i am certainly not a good one to sustain it, even if there is something good in it :)
21:02:34amiconnjhMikeS: Unfortunately not. In fact I don't know about other reports, but I observed it for quite a while myself
21:03:01Lloreannanok: Sorry. Don't mean to sound critical, but you seem to be suggesting *adding* a bunch of stuff, which means that there's new stuff to maintain, rather than less stuff to maintain.
21:03:09amiconnFor quite some time I thought it might be a hardware problem of either my 2nd Gen or my firewire port
21:03:13preglowLlorean: yeah, i'm easing into the idea
21:03:23Lloreanpreglow: Ot
21:03:42Lloreanpreglow: If I could type... It's not perfect, but I think it would be the combination that would least break the way people normally tend to use it now, while still losing a setting
21:03:49amiconnBut now that I compared behaviour with my 1st Gen, and also "cmp"ed files, I know that it isn't
21:04:12preglowLlorean: yeah,i think that could work
21:04:17amiconnjhMikeS: I think it did not happen back in August
21:04:29preglowLlorean: at least not so many people use "hold" while charging that they'd be too confused by the behaviour
21:04:44 Quit obo (Read error: 110 (Connection timed out))
21:04:45preglowand those that do probably want the backlight off :-)
21:04:46jhMikeSamiconn: Ok. All I know is 1) both the crashing and batt problem disappear with the cache off. 2) never waking up the COP to have it init has no effect (one core running is enough)
21:04:48LloreanI think the only people who use hold while charging are the ones who expect it to overrule the "While Charging" backlight settings currently
21:04:55preglowyeah, heh
21:05:14preglowso, eliminate one option, and still accomplish what people want \o/
21:05:15nanokLlorean: i understand that, but the reason i suggest it in the first place is that i think it can be made in such a way to make things easyer after. meaning have some things which rarely change separated from the functional part, which is worked on all the time by the developers, and have less technical people/non=programmers be able to handle that, and thus help the developers not worry about it, thus saving time and effort
21:05:16jhMikeSamiconn: that far back? hmmm...
21:05:22amiconnjhMikeS: Yes, I know it happens with single core... of course then you don't see the cop crashing
21:05:29nanokLlorean: it would be an added effort at first, ofcourse
21:06:01jhMikeSamiconn: I mean never even waking it to leave crt0-pp.S at all.
21:06:02nanokLlorean: it's just a thought, probably not very well forumlated, and not very clear (it just came to me while watching the discussion)
21:06:15Lloreannanok: I don't really understand what we're talking about any more... what things would non-technical people be working on at all in terms of core code?
21:06:35amiconnjhMikeS: ?? Looks like I'm confused...
21:06:55jhMikeSI thought something could get clobbered when the COP goes into cop_main to sleep on single-core build but no. This is really odd.
21:07:01 Join obo [0] (n=obo@rockbox/developer/obo)
21:07:59 Join stripwax [0] (
21:07:59jhMikeSamiconn: Never waking up the COP so it even goes past the startup code it still behaves the same.
21:08:15amiconnDoes the cop crash then?
21:08:33nanokLlorean: that's the point, not core code. menu layout, features which are used/accessible and those which are not, and so on. everything which means user interaction. i don't know, something like x windows system, on top of which you put a window manager, desktop env., and so on, but it is not the problem of the video card drivers developer, nor the x-windows system developer; ofcourse, not on that scale, it was just an example of such reasoning
21:08:44jhMikeSit never crashes if it's not woken. the CPU still crashes or has bad battery reading
21:08:53amiconnWith the build I have on my PP5002 ipods, I see both cores crashing. First the cop, and later the cpu
21:09:11jhMikeSyeah, if they both perform init they will
21:09:31amiconnIf you disable the cache to make it boot - where are you doing that?
21:09:39Lloreannanok: The thing is, the underlying code still has to support any menu entries they want to add to the layout... So in the end, the core code still has to increase complexity to match any increased featureset...
21:09:41amiconnAnd what happens if you enable the cache later?
21:10:12jhMikeSamiconn: just commenting out ipod_init_cache. (you also need to block calls to the cache routines)
21:10:27 Quit karashata ("I will *SO* make you regret that... Later... *is a lazy dragon*")
21:10:52amiconnjhMikeS: Maybe the cache is flushed too late, so it writes old data from the cache to the ram?
21:11:28jhMikeSit's flushed in the bootloader (it is on by default). disabling it in the bootloader has no effect either. :\
21:11:46amiconnWhy does it flush and invalidate in rockbox, btw? Shouldn't that be done in the bootloader, *before* passing control to the loaded rockbox binary?
21:12:03jhMikeSit is handled in the bootloader
21:12:17amiconnWhat happens if you enable the cache later?
21:12:20nanokLlorean: yes, but it wouldn't have to support combinations of them. for example, if you have an event for "hold on", and a set of possible settings for the backlight, you can combine them however you want, without needing the core code to have support for specifically "backlight off if hold on", built into it
21:12:44nanokLlorean: i mean hold switch "on"
21:12:45amiconnWhat's really irritating is that it didn't happen with every build in the past
21:13:05jhMikeSamiconn: same thing happens no matter what.
21:13:42jhMikeSyeah, it is frustrating since the basic init hasn't changed at all.
21:13:48amiconnWith dualcore, there weren't enough builds so far. The 2 builds I tried both crash spectacularly
21:14:13nanokLlorean: this is i think the bottom line: you can only worry about "atomic" functions which can be afterwards combined in any way, no need to worry about and maintain others. i am not saying it is better, just saying it might be a way to look at it *maybe*
21:14:22jhMikeSyou can probably throw some noops somewhere and have it work mysteriously
21:14:34***Saving seen data "./dancer.seen"
21:14:59Lloreannanok: In that case, all you need is a few simple things. Various "States" for the device (Playback on/off, Hold on/off, etc) and the ability to load a .cfg file in transitions between various states or combinations of states.
21:15:21LloreanWhich then boils down to "The user can click on a .cfg whenever they want" and saves the code of having to monitor these states and find files for them.
21:15:23nanokLlorean: as i said, it just occured to me now, i am not advocating it, just trying to figure out what it is that ran through my mind :), and express it coherently, so maybe someone who understands the big picture can make use of it (or discard it)
21:15:27amiconnNow that would be wasteful... spinning up the disk for every state change
21:16:40amiconnjhMikeS: In fact I already pondered going through all commits since August (in steps of 10 first), and write down which ones crash and which don't
21:17:05amiconnBut that'd be a lot of work, and afterwards I might need a new firewire socket for my laptop :/
21:17:09nanokLlorean: amiconn i ws not thinking about config files for each state change, i was thinking about a config which is "read" and loaded into memory at boot, for example, only once
21:17:32amiconnWasting memory is bad either
21:17:33jhMikeSamiconn: if a build does work, do yours show low battery or is that only 3g?
21:17:52preglowLlorean: i might cook up a patch for this and demand some feedback :)
21:17:53Lloreannanok: It's the same thing though. The .cfg files could be cached just as effectively, the "wasting memory" argument stands. Did you know you can load a .cfg by clicking on it in the file browser, like a song?
21:18:05Lloreanpreglow: Sounds like a good idea.
21:18:08amiconnI sometimes see low battery percentages first, but I always thought these are just due to the disk spinning at boot
21:18:12preglownot today, though
21:18:25 Quit psycho_maniac (" HydraIRC -> <- Organize your IRC")
21:18:39nanokLlorean: yes, i know that, haven't played with it though, and don't know to what extent it can be usefull for the present discussion
21:18:47amiconnLow bat shutdown is disabled on 1st/2nd Gen, as those can't be shut down anyway, and we don't handle suspend yet
21:19:02nanokbut considering the discussion was initiated in the first place, probably not much
21:19:39 Quit Mathiasdm ("Ik ga weg")
21:19:47Lloreannanok: .cfg files can change an arbitrary number of settings, from 0 to "all of them", and can very effectively be used to change "profiles" such as an "in-car.cfg" that changes backlight preferences while plugged, font, theme, volume, enables the Car mode, etc, all in one swoop.
21:19:47jhMikeSamiconn: for me, the start % will be low, when the disk spins down at first, the battery indicator starts flashing (changing the disk spindown timeout changes when that happens). After awhile the filtered battery level will return to normal.
21:20:15amiconnHmm. Didn't observe that
21:20:38amiconnBut 1st/2nd Gen has no pcf, they have a dedicated (and different between 1st and 2nd Gen) battery adc
21:20:42jhMikeSIt's probably a side effect...again, no such phenomenon with cache off.
21:21:13nanokLlorean: yes, i understand that, but what i was trying to get through here was a system where the cfg file is designed such that you don't need to have all options "in-core", but only atomic "options", which can be combined after in any way
21:21:35amiconnThey need bit-banged serial transfer... and the one in the 2nd Gen is so slow that the driver needs to use GPIO interrupts in order to be efficient
21:22:03Lloreannanok: There are very, very, very few options that can be described as a combination of atomic options.
21:22:21 Quit jhulst (Read error: 113 (No route to host))
21:23:02nanokLlorean: hm, okay. in that case i am way off, i suppose. i thought this was mainly what the discussion was caused by,
21:23:09 Part pondlife ("Gone")
21:23:16 Quit stripwax ("Miranda IM! Smaller, Faster, Easier.")
21:23:25nanokLlorean: this is what i understood from your statement about "new ways to do old things" as features
21:23:42nanokbeing rejected as features, that is
21:28:34jhMikeSamiconn: this is by far the strangest pp behavior I've encountered. none of the usual coherency problem seem to apply here. those have always been fairly quick to track down and fix.
21:28:41 Quit stewball`ghost (Read error: 110 (Connection timed out))
21:30:15jhMikeSI havnen't found any "exotic" things the OF is doing either. Basically, it does as we do and makes sure the cache is turned off as its very first operation.
21:34:07preglowdone any binary searches through commits yet?
21:34:38jhMikeSnot yet. this will be the first time ever I have to do something like that. :p bleh...what a pain.
21:35:02preglowespecially considering it sometimes works when it's in fact buggy :)
21:35:34jhMikeS...and having to go back so far...I'll probably just go by month at first.
21:35:45amiconnYes, and because of that binsearch won't really work
21:35:55amiconnIt would need to be an exhaustive search
21:36:49 Quit obo (Read error: 110 (Connection timed out))
21:36:57 Quit spiorf (Connection timed out)
21:37:02jhMikeSwell, if I can nail a month down. then go by week...etc.
21:39:41 Join rlpowell [0] (
21:39:56amiconnHmm, ~1800 revisions from August to now
21:40:27amiconnThe first time I observed this behaviour should be deducable from the irc logs
21:40:39amiconnIt was on some weekend
21:40:56jhMikeShigher-order searches are your friend here :)
21:41:00amiconnAnd I usually update all targets once per week
21:41:19amiconn...but not always
21:41:29 Quit axionix (Read error: 110 (Connection timed out))
21:42:22 Join homielowe [0] (
21:42:24jhMikeSI have no idea when it popped up. I remember you commenting after the initial dualcore commit, then I made a change to remove swp and you said it worked again.
21:43:30jhMikeSperhaps swp had nothing to do with it
21:43:33 Quit amiconn (Nick collision from services.)
21:43:42 Join amiconn [0] (n=jens@rockbox/developer/amiconn)
21:44:23 Join barrywardell [0] (
21:44:34przemhbif someone would be interested in checking workaround patch for FS #8238 then it is on the tracker
21:45:00przemhbthe patch verifies wheter or not H10 radio was tuned in
21:45:20 Quit Gnu47 ("I'm off to stalk my Pah'Mach'kai")
21:45:35jhMikeSis the H10 tuner bitbanged i2c?
21:45:35 Join Gnu47 [0] (
21:45:39przemhbif it wasn't it sets PLL frequency untill the radio gets tuned in
21:46:21przemhbjhMikeS: what do you mean by bitbanged?
21:46:48jhMikeSnot using the hardware i2c but by setting GPIO pins in software
21:47:27przemhbyes, it is
21:50:43jhMikeSI'm gonna slow it down a bit and see if 80MHz will actually tune the radio (it can neither read nor write the chip at that frequency).
21:51:55jhMikeSI know it's udelay based but worth a shot anyway.
21:52:49przemhbI've tested increased udelay up to 50us, but it did not help for the FS #8144
21:53:04 Quit Zagor ("Client exiting")
21:53:33 Join jhulst [0] (n=jhulst@unaffiliated/jhulst)
21:53:41amiconnjhMikeS: This H10 tuner problem is weird. It seems only the 20GB model is affected
21:53:59amiconnI have no problem tuning back & forth at both 30 and 80MHz on my 6GB model
21:54:49jhMikeSalmost sounds a bit like the c/e200 stuff
21:55:48 Join Zagor [0] (
21:57:42 Join goffa_ [0] (n=goffa@
21:59:48amiconnjhMikeS: My first mention of this PP5002 crashing bug in the logs is from October 16. An earlier one is from Octoer 6, but that I am not sure about myself
22:00:34 Join desowin [0] (
22:01:07jhMikeSprzemhb:'s as though there's no bus traffic at all. :\
22:02:11amiconnjhMikeS: Log from 16 Oct 2007 starting 10:06
22:02:35*amiconn likes baregrep :)
22:02:47 Quit goffa (Read error: 110 (Connection timed out))
22:03:07jhMikeSwhich on the 6th?
22:03:25amiconnErrr, that earlier mention is from the 4th
22:03:44amiconnStarting 19:43
22:03:52 Quit male (Read error: 110 (Connection timed out))
22:03:59jhMikeSprobably why I'm not finding it :)
22:05:13 Join donutman25 [0] (n=chatzill@
22:06:58 Join movmov [0] (n=bbnet2@
22:07:45jhMikeSI only see talk about the H10 and remember this discussion (was before getting to the bottom of PP5020 instabilities).
22:08:28przemhbjhMikeS: if you take a look into m patch you'll find an interesting thing - I had to use sleep() in order to make it work as simple delays - udelay() and TIME_AFTER() - were insufficient
22:08:42 Quit movmov (Client Quit)
22:09:18 Join goffa [0] (n=goffa@
22:09:41 Join obo [0] (n=obo@rockbox/developer/obo)
22:10:06jhMikeSprzembh: I noticed that too and I'm curious as to the reason and how it could matter.
22:10:56przemhbjhMikeS: I have no idea...
22:12:30 Quit donutman25 ("ChatZilla 0.9.79 [Firefox]")
22:13:26 Join mawell [0] (
22:14:14 Quit Gnu47 ("Don't follow me")
22:14:29 Join Gnu47 [0] (
22:15:01amiconnjhMikeS: You're right - as I said, I wasn't sure about this 4th Oct mention
22:15:13 Join japc [0] (
22:15:20amiconnThat was H10 - not visible when greping
22:16:35amiconnAnd Oct 16 was the broken swp() fix - but perhaps not only
22:16:56 Join stripwax [0] (
22:20:09przemhbbarrywardell: hi
22:20:34jhMikeSamiconn: well, we know the problem is surviving deep changes anyway
22:21:11 Part mawell ("Kopete 0.12.3 :")
22:21:34amiconnHmm. Judging from the Oct 16 conversation, I'd say it started with the new scheduler code...
22:21:36przemhbbarrywardell: haven't you experienced any problems with your H10 tuner? (like FS #8238 FS #8177 FS #8144) which version of H10 have ou got?
22:21:39 Quit goffa_ (Read error: 110 (Connection timed out))
22:21:52barrywardellI have 20GB
22:22:09 Join linuxstb_ [0] (
22:22:21przemhbbarrywardell: have you tried to set starting screen to FM tuner?
22:22:35barrywardellno I haven't used the tuner much
22:22:43barrywardellI'll try now
22:22:50przemhbbarrywardell: in my case the tuner doesn't tune in
22:23:01jhMikeSamiconn: possible, but why only a serious problem on 5002? though it's certainly not out of my consideration yet.
22:23:13amiconnI don't know :\
22:24:28barrywardellI just set start screen to tuner, then turned off and back on and everything worked fine
22:24:29jhMikeSof course it could affect other targets but in a much more subtle/as yet undetected manner
22:25:02przemhbbarrywardell: lucky you; my tuner doesn't tune in
22:25:09barrywardellI did just reproduce 8177 though
22:25:15jhMikeSI suppose a revision checkout before/after those changes should say
22:25:28*amiconn doesn't understand how the current scheduler works... :(
22:26:40 Join ompaul [0] (n=ompaul@gnewsense/friend/ompaul)
22:27:08przemhbbarrywardell: m friend's H10 is seriously affected by all those tuner issues - it hardly works :/
22:27:46jhMikeSamiconn: simple list manager really...move the thread from one place to another depending on if it's running, blocked, or timeout.
22:28:24 Join linuxstb__ [0] (
22:28:24Gnu47Welcome linuxstb__! You have achieved a new peak of 162 users for #rockbox!
22:29:04barrywardellprzemhb: have you checked if it's getting stuck in those while loops?
22:29:15preglowi don't think the scheduler is that hard to follow
22:29:15pixelmaGnu47: what if you wouldn't count the two other instances of linuxstb? ;)
22:29:51Gnu47oops.... didn't realise i'd let that run in here :/
22:29:54jhMikeSif it's on the running linked list (which is circular) it's just scheduled thread->next basically. if blocked, sleeping, or blocked w/tmo, it's somewhere else and gets no CPU cycles.
22:30:12 Quit Zagor ("Client exiting")
22:31:04przemhbbarrywardell: no, I haven't (which while loops)?
22:31:54 Join axionix [0] (
22:32:32barrywardellwaiting for !SCL
22:33:59 Quit homielowe (Read error: 110 (Connection timed out))
22:35:44 Join linuxstb___ [0] (
22:35:45 Quit linuxstb (Read error: 110 (Connection timed out))
22:35:46 Nick linuxstb___ is now known as linuxstb (
22:35:57 Quit linuxstb_ (Nick collision from services.)
22:35:59 Quit Rincewind ("bye")
22:36:47barrywardellAlso, in scan mode for me it doesn't find any stations
22:36:51 Quit Frazz ("Leaving")
22:36:59 Join end-schranz [0] (n=bb@
22:37:17end-schranz[ERR] Firmware partition doesn't contain Apple copyright, aborting.
22:37:17end-schranz[ERR] Failed to read firmware directory - nimages=0
22:37:25end-schranzhow to fix this?
22:38:02linuxstbend-schranz: What did you type?
22:38:14end-schranz./ipodpatcher /dev/sdc -ab loader.bin
22:38:29przemhbbarrywardell: for me too
22:38:37linuxstbIs /dev/sdc your ipod, and does it contain a working Apple firmware partition?
22:39:00end-schranzyes: i only changed the partition layout to install ipodlinux
22:39:04jhMikeSamiconn: were you going to check out and test any earlier revisions? no sense in duplicating efforts.
22:39:37amiconnNot now
22:40:06linuxstbend-schranz: ipodpatcher disagrees with you... Does your ipod start normally?
22:40:37end-schranzlinuxstb: no
22:40:55linuxstbThat's the problem - ipodpatcher requires a working ipod to modify.
22:41:05end-schranzlinuxstb: so i have to install the bootloader first and then partition it?
22:41:14jhMikeSI guess I'll go ahead then, starting at the 16th. Is there a way to get a flat revision list from the SVN browser?
22:41:42end-schranzlinuxstb: thx i will try
22:41:42preglowsvn log? :>
22:41:44 Part end-schranz
22:43:47jhMikeSpreglow: works too from the browser. :)
22:44:03 Quit desowin ("use linux")
22:44:11*preglow distrusts and avoids his browser
22:45:03 Quit obo ("KVIrc 3.2.4 Anomalies")
22:45:22stripwaxThere's a comment posted to my pacbox patch saying that the long-term plan is to move all codecs to COP so that the main CPU is left alone for foreground tasks and that running pacbox on COP isn't compatible with that plan - does that mean the patch is unlikely to ever be accepted?
22:45:41preglowjhMikeS: would you know what the hell 8245 is talking about with mpegplayer?
22:45:57preglowstripwax: i wouldn't accept it, no
22:46:14stripwaxpreglow - what can I do to encourage the patch to be accepted?
22:46:39preglowstripwax: well, what can you do? if the point is to make it possible to play pacman while playing music, it'll break in the future
22:46:42stripwaxlinuxstb seemed interested (or, at least, gave me the idea to try it out..)
22:47:00jhMikeSpreglow: have I commented on it? (will look in one moment)
22:47:03stripwaxpreglow - the cpu and cop will just get switched around, no?
22:47:23preglowjhMikeS: not in the least, the bug is unrelated to mpegplayer too, so i'll close it, just wondering if there's some info i can give, but it's pretty vague...
22:47:31stripwaxCurrently cpu=codec, cop=pacbox, in the future it will be cpu=pacbox, cop=codec
22:47:35jhMikeSpreglow: one sec.
22:47:37preglowstripwax: sure
22:47:50preglowstripwax: how will this go together with pacman sound, though?
22:48:01stripwaxpreglow - not sure I understand. pacman sound is not implemented
22:48:11stripwaxif/when it is, it can run on the same cpu as the codec, presumably
22:48:18 Join male [0] (
22:48:19jhMikeSloading that SVN log in my browser is causing some nice pagefile activity :)
22:48:27stripwaxunless you think ppl want codec *and* pacman sound?
22:48:37preglowstripwax: that's what i'm saying, what do you plan to do when someone comes aliong and implements it? add a "sound on/off" setting?
22:48:46 Join safetydan [0] (n=safetyda@rockbox/developer/safetydan)
22:48:59stripwaxpreglow - whoever 'implements it' will add a sound on/off setting themselves, surely?
22:49:15preglowdo any other games currently allow you to choose between playing music and playing game sound?
22:49:16stripwaxwe've already talked about it being possible to play pacbox + codec on gigabeat
22:49:25 Nick BilleniumZzZ is now known as Billenium (
22:49:33jhMikeSit is possible to do
22:49:39preglowwell, i guess it's a valid setting anyway, so sure, i don't have much against being able to listen to music and play at the same time
22:49:55jhMikeSgigabeast is also fast enought to play SPC+pacbox at 30fps
22:50:04preglowspoiled people...
22:50:14stripwaxis gigabeat fast enough to play pacbox + spc + pacbox emulated sound mixed in? :-)
22:50:35linuxstbIf not, the gigabeast definitely will be...
22:50:39 Quit Domonoky (Read error: 104 (Connection reset by peer))
22:50:40jhMikeSstripwax: I'd say quite likely yes
22:50:54stripwaxok. mixing is also something else I'm not planning to implement :-p
22:51:31*stripwax puts the finishing touches to sliding_puzzle..
22:52:21*preglow wonders why nobody just did a native pacman
22:52:30preglowi really don't see the point in emulating it, heh
22:52:32stripwaxOk so. Original question - is there anything I can do to help get the pacbox patch committed or is it fatally incompatible with future plans?
22:52:41stripwaxpreglow - nerdy accuracy :-)
22:52:45 Quit linuxstb__ (Read error: 110 (Connection timed out))
22:52:56jhMikeSpreglow: I'd say that last comment in 8245 is related to actually doing proper YUV conversion
22:52:57*linuxstb was about to say that, but slightly differently ;)
22:53:04preglowstripwax: make some interested person commit it :)
22:53:07stripwaxApple didn't bother emulating the arcade roms (.. obviously ..), so I'd imagine they have no problem running 60fps wiht audio
22:53:20preglowjhMikeS: it wasn't proper before?
22:53:31stripwaxpreglow - hehe. I prefer 'encourage' rather than 'make' :)
22:53:53jhMikeSthe JPEG conversion made it less blocky because it hid it from not adding the correct offset
22:54:02stripwaxI'd love to know if it works on nano btw.
22:54:20jhMikeSnow you see all the nice image artifacts as they were meant to be seen :)
22:55:25stripwaxpreglow - actually I did make a start on a native pacman but the same week the initial commit of pacbox came along :)
22:55:42preglowjmspeex: jpeg conversion?
22:56:14 Join homielowe [0] (
22:56:19*jhMikeS wonders where jmspeex is
22:56:21preglowjhMikeS: jpeg conversion? :>
22:56:38jmspeexpreglow: what about jpeg?
22:56:45preglowjmspeex: wrong nick, sorry
22:56:49preglowmy tab completer sucks ass
22:57:01jhMikeSJPEG conversion = use the JPEG yuv conversion. JPEG uses the entire 0-255 range for all components - video does not.
22:57:39preglowjhMikeS: really? i thought that was a broadcast only thing
22:57:52 Quit barrywardell ()
22:58:02 Quit moos ("Rockbox rules the DAP world")
22:58:47jhMikeSCCIR6601? No, Y=16-235, Cr/Cb=16-240 for MPEG video.
22:59:36preglowyeah, i thought that was a broadcast limitation
22:59:53preglowno reason to comply to that when we're just encoding for decoding on a dap?
23:01:06jhMikeSyou'll compress the color dynamic range then
23:01:09*preglow wonders why mpegplayer takes so long in pausing
23:01:50*jhMikeS need to finish his mpegplayer revisions and everything will be "instant".
23:04:10preglowwhat other numbers are displayed when displauy fps is on?
23:04:29jhMikeSdropped frames and sequence number
23:04:30 Quit l7 ("Lost terminal")
23:05:26 Quit nicktastic ("Leaving")
23:05:42 Part przemhb
23:06:04 Quit MethoS- (Remote closed the connection)
23:08:16 Quit Redbreva (Read error: 104 (Connection reset by peer))
23:08:33 Nick parafin is now known as parafin|away (
23:12:02 Join stewball [0] (n=WTFOMGBB@
23:13:44 Quit J3TC- (Read error: 110 (Connection timed out))
23:14:38***Saving seen data "./dancer.seen"
23:15:40 Join donutman25 [0] (n=chatzill@
23:20:16 Quit donutman25 (Client Quit)
23:25:55 Join darkapostrophe [0] (
23:31:28 Join kri [0] (i=5ae3840f@gateway/web/cgi-irc/
23:33:01 Quit linuxstb (Remote closed the connection)
23:33:24stripwaxDoes anyone have time to try out the sliding_puzzle patch? Or anything else needs doing now before it could be committed?
23:33:46stripwaxI've just updated it with a few more tweaks and using the HAVE_ALBUMART as discussed previously
23:35:06 Quit kri (Client Quit)
23:37:35 Quit homielowe (Read error: 110 (Connection timed out))
23:39:22 Join homielowe [0] (
23:39:22 Quit Dark_Apostrophe (Connection timed out)
23:39:22 Nick darkapostrophe is now known as Dark_Apostrophe (
23:41:58 Quit Arathis ("Bye, bye")
23:43:06 Quit bertrik ("bye")
23:45:35 Quit Thundercloud (Remote closed the connection)
23:47:09 Quit HellDragon (Client Quit)
23:47:26PaulJamlittle question, with the new settings system for the backlight timeout, would it be possible to set the backlight timeout to several hours via the .cfg file?
23:48:14 Join HellDragon [0] (i=jd@unaffiliated/helldragon)
23:50:44 Join barrywardell [0] (
23:51:54 Join kri [0] (
23:53:32 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")
23:53:43 Join Kei-o [0] (n=Kei-o@
23:53:50DrCrow_Does anyone here know anything about the zvmport?
23:54:06 Join goffa_ [0] (n=goffa@
23:55:11XavierGrGigabeat MES30/60VK is supposed to be Gigabeat S?
23:56:47 Join MarcGuay [0] (i=d8ef4c65@gateway/web/cgi-irc/
23:57:15toffe82XavierGr: yes, not sure what 60vk means but mes30 is a gigabeat s with a 30gb hdd
23:57:44 Join EspeonEfi [0] (i=espeonee@STRATTON-FORTY-SEVEN.MIT.EDU)
23:58:01 Quit EspeonEfi (Client Quit)
23:58:30 Quit MarcGuay (Client Quit)
23:58:36 Join EspeonEefi [0] (i=espeonee@STRATTON-FORTY-SEVEN.MIT.EDU)
23:58:43 Quit Kei-o ("Leaving")
23:58:45toffe82XavierGr: the mes60vk is the gigabeat S with a 60 Gb hdd

Previous day | Next day