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 2008-12-01

00:02:00 Quit BHSPitLappy ("Ex-Chat")
00:05:17 Quit miepchen^schlaf ()
00:17:27 Quit guyjones ("Leaving")
00:23:22 Quit Nico_P (Remote closed the connection)
00:24:21jhMikeSkugel: the backlight brightness adjustment isn't working correctly in mpegplayer now (happening on FX so far).
00:25:50 Join BHSPitLappy [0] (n=BHSPitLa@unaffiliated/bhspitmonkey)
00:27:37 Join __lifeless [0] (n=lifeless@
00:31:47 Quit Schmogel ("Miranda IM! Smaller, Faster, Easier.")
00:33:01kugeljhMikeS: not working correctly as in?
00:33:47kugeljhMikeS: oh I see
00:33:51jhMikeSkugel: fading strangely and not properly setting brightness even thought it should still be using the helper.c functions
00:34:23*kugel did nearly forget that mpegplayer has such a feature
00:36:22kugeljhMikeS: actually, it seems to use the setting.
00:36:28kugelJust the menu not
00:37:01kugelelefants dream definitely respects my setting, the mpegplayer menu is acting weird though
00:37:05jhMikeSit fails to reset after exit. it used to behave properly, not that I can pin it on your commit yet
00:38:19jhMikeSI just caught it while working on power management code
00:38:32kugeljhMikeS: can you double check that? the video is definitely using the mpegplayer brightness for me
00:40:01kugeljhMikeS: I don't experience weird fading though. How can I reproduce that?
00:42:07 Quit faemir (Remote closed the connection)
00:44:04 Quit _lifeless (Read error: 110 (Connection timed out))
00:46:04 Join aarcane [0] (
00:46:22kugeljhMikeS: IIUC it's working proper, as the brightness should only be applied in "video mode"
00:46:29 Quit ender` (" Where python would say you can skin a cat with a knife, so adding other methods would just make the language hard to learn,")
00:47:04 Quit Bensawsome ("The awsome is gone :(")
00:47:51kugelI'm looking at line 628 in mpegplayer.c
00:48:11 Quit __lifeless (Read error: 110 (Connection timed out))
00:49:13JdGordonyay sound!
00:49:25*JdGordon sends belated applkause to funman
00:49:32kugelJdGordon: yay
00:49:36kugelnot for us though :(
00:49:47kugelwas quite a party you missed :)
00:49:52JdGordonstupid bloody buttons
00:50:14kugelJdGordon: we definitely need to mess with interrupts
00:50:45JdGordonhow do i make a of dissadembly?
00:51:03kugelthe other ams's work because the buttons don't use interrupts. Our buttons work in bootloader because the bootloader doesn't use interrupts _and_ not the tick task
00:51:38kugelJdGordon: you mean disassembly? I used objdump. I didn't found something about buttons in the firmware block though
00:52:09JdGordonwhats the command to do the dump?
00:52:40jhMikeSkugel: I have backlight brightness lower in menus than in mpegplayer
00:53:01kugelJdGordon: arm-elf-objdump -D -b binary -m arm -M force-thumb file.bin > file.txt
00:54:31JdGordon236MB :)
00:54:52kugeljhMikeS: Ok. I tried with global settings to minimum, mpegplayer set to maximum brightness. The menu is using the low brightness and the video the high brightness
00:54:59kugelJdGordon: have fun at reading
00:55:00jhMikeSif exiting while playing, the brightness doesn't re-dim until a key is pressed and the fading between levels fails
00:55:19kugelit does for me
00:55:43kugeljhMikeS: which target do you use?
00:56:08kugelI assume the S, which doesn't use my fading, but only the setting
00:56:08jhMikeSI just noticed on Gigabeat F/X
00:56:31kugelthe F/X does neither use the sw fading
00:57:24jhMikeSMust be some other change along the line. I'll check e200.
00:57:27kugelI'm not entirely sure if it's related to my patch.
00:57:47kugelIt's definitely working on my e200
00:58:29 Join blkhawk [0] (
00:58:37kugeljhMikeS: If, then r19224 is causing it. I doubt it though
00:58:46jhMikeSkugel: looking like probably not. e200 is fine here.
00:59:47 Join _lifeless [0] (n=lifeless@
01:01:31 Join blkhawk- [0] (
01:01:38 Quit shadearg ("goodbye.")
01:02:21jhMikeSThough if I have Backlight on Hold set to "off", I can get some occasional flashes of full brightness when turning hold off again (if I flick it on and off rapidly)
01:03:21 Quit BHSPitLappy (Remote closed the connection)
01:03:36kugeljhMikeS: on the e200?
01:03:37 Join shadearg [0] (
01:04:28kugeljhMikeS: that's most likely due to lcd_enable, which happens to cause flashes (try disabling fading and you'll notice it everytime you turn the backlight on)
01:06:01kugelseems the backlight gets confused if you do that too often
01:07:07jhMikeSThat's a different effect from turning the bl on a little too soon at times. This I can see even with the lamp plugin.
01:07:16kugelbut unless you intend to destroy the hold switch you're not gonna notice that
01:07:47kugelyea, I just tried, it's not lcd_enable
01:07:55jhMikeShehe. well, I do tend to test anything abusively like that. :)
01:08:26kugelno idea how to fix that though
01:09:22kugelfixing that will likely need sanity checks, and thus cost more then it brings
01:09:41 Quit SoapWork ("CGI:IRC")
01:10:18kugelrapidly entering and exiting lamp just crashed my sansa :o
01:12:08jhMikeSusing voice in the splash?
01:12:26kugelI think voice is on, but I don't have voice fules
01:13:13kugeljhMikeS: another topic. Are you good with irq stuff? e200v2 and Fuze needs to implement them for button reading. And I have now idea how that works
01:13:53jhMikeSI don't know how to set them up on that SoC but they should be like other I'd imagine.
01:14:29 Join EspeonEefi [0] (i=eefi@STRATTON-TWO-EIGHTY.MIT.EDU)
01:14:34kugelwell, we have a datasheet, so irqs are documented. But this is absolutely the first time in my life I get in touch with it
01:15:22jhMikeShmmm, you have a VIC there too.
01:15:30kugeland copying from other button driver is hard since the macro names are different, so I don't know which I need to copy to
01:15:44kugelyea, VIC
01:16:02 Join blkjawk[work] [0] (
01:16:40*jhMikeS opens his as3525 ds copy
01:17:28kugelvic is on page 24, the GPIO stuff at page 48+
01:17:43 Quit herrwaldo ("Konversation terminated!")
01:17:51 Join Bensawsome [0] (n=Bensawso@unaffiliated/bensawsome)
01:17:56 Join SUSaiyan [0] (
01:18:14kugelI already tried a bit, but it doesn't work. I got a panic even at enabling INTERRUPT_GPIOB in the VIC
01:18:33jhMikeSwhere's the register list?
01:18:42kugelwhich register list?
01:18:57jhMikeSfor the controller.
01:19:05kugelvic interrupt sources are on page 25
01:19:06 Quit blkhawk (Connection timed out)
01:19:23jhMikeSyou know, VIC_IRQ_STATUS, etc.
01:19:42 Nick blkjawk[work] is now known as blkhawk (
01:21:12kugeljhMikeS: I would have to search the datasheet myself. But the registers are also written down in firmware/export/as3525.h
01:21:38kugelthe GPIO registers are incomplete regarding irq
01:22:42jhMikeSI don't see a regmap for the vic, just the IRQ list
01:24:00 Join Strife89 [0] (n=michael@
01:24:17kugeljhMikeS: line 381
01:24:33kugel#define VIC_IRQ_STATUS (*(volatile unsigned long*)(VIC_BASE+0x00))
01:24:52jhMikeSah, hmmm, that's an ARM module and they give a link
01:30:04kugeljhMikeS: you fount it? Any idea how to get it working?
01:30:26jhMikeSyou specified a handler routine?
01:30:39kugelI don't think so
01:31:56 Quit blkhawk- (Read error: 110 (Connection timed out))
01:32:41kugelthat's what I've done so far (please don't laugh ;) ):
01:32:53jhMikeSit looks like you should just specify a function void __attribute__((interrupt("IRQ"))) xxx(void) {}
01:33:07jhMikeSxxx = name of vector
01:33:52kugel"void irq_handler(void) __attribute__((interrupt ("IRQ"), naked));", that's in system-as3525.c
01:35:10 Quit mofux (Read error: 110 (Connection timed out))
01:35:39 Join mofux [0] (
01:36:07kugeljhMikeS: is this what you mean?
01:36:23jhMikeSyes, it's manual prologue/epilogue. I guess you just declare a normal function then.
01:40:43kugelis what I've already done entirely wrong?
01:42:15jhMikeSexample: (anywhere in the code) "void INT_GPIOA(void) { <handle event here> }
01:43:28kugeland <handle event here> would contain the button reading? or setting up the interrupts like I've done in the paste
01:45:30jhMikeSit would handle any events enabled for that GPIO port and clear them (buttons or otherwise)
01:46:49 Join uep [0] (
01:48:09 Quit MethoS-- (Remote closed the connection)
01:49:43uepI'm having a problem where my h10 20gb is saying that the file system is read-only.
01:50:33jhMikeSkugel: The intersection of set & pending. Each GPIO needs the particular pins set to have ints enabled, serviced and cleared in the routine. You probably want edge sensitivity for buttons (dual-edge perhaps).
01:52:29kugeljhMikeS: like this?
01:52:55 Part pixelma
01:53:02kugelI don't really know what the difference between "edge" and "level" interrupts are
01:53:24 Join pixelma2 [0] (n=marianne@rockbox/staff/pixelma)
01:54:40***Saving seen data "./dancer.seen"
01:54:47uepI would think that edge means the rising or falling edge of the signal, level being anytime it's "high".
01:56:52uepIn the context of buttons, I think it would mean either getting interrupted when the button is pressed or released (edge), or constant interrupts while it's being pressed (level). Don't take my word for it, because I think I missed the bulk of the conversation.
01:57:47jhMikeSyou want to detect changes, so dual level look like it would be the setting for them
01:58:29kugeljhMikeS: what I just pasted doesn't work
01:58:36kugelbut the fuze boots at least
01:59:25kugeland the "Unhandled IRQ" panicf isn' there anymore
02:01:24uepDoes anybody have any insight into my "read-only" filesystem problem?
02:01:45uepI assume I can correct it and that it's not a fatal problem since I can still at least browse the files on the drive and connect to it via usb.
02:03:13uepIt actually just seems like ubuntu might be mounting it as read-only, but I'm not sure how to fix it because it's being automatically mounted.
02:06:10jhMikeSkugel: for each button line, IS bits should be 0, IBEs 1, IEs 1, read the MIS to obtain the changed lines, write 1s in the handler to the ICs to clear the events
02:07:07 Quit Bensawsome (Nick collision from services.)
02:07:37kugeljhMikeS: I assume I only need to set the bits in IS, IBE etc for that single pin I'm going to read?
02:08:13 Join Guest69525 [0] (
02:08:35 Quit AndyI (Read error: 110 (Connection timed out))
02:08:44 Join AndyI [0] (i=AndyI@
02:08:44 Quit AndyIL (Read error: 110 (Connection timed out))
02:08:54 Join AndyIL [0] (i=AndyI@
02:09:56kugeljhMikeS: I read the change on the pin through MIS? So I don't need if(GPIO_PIN){ btn |= BUTTON_} at all?
02:10:53jhMikeSyou can filter what you bother to read
02:10:54 Nick Guest69525 is now known as Bensawsome (
02:11:32kugelI mean, I don't need to read the pin directly?
02:11:38jhMikeSbut, xxx_IC = xxx_MIS will clear all enabled events
02:12:02jhMikeSyou should read the pin directly
02:12:38 Join LambdaCalculus37 [0] (
02:14:49 Quit uep (Remote closed the connection)
02:17:26kugeljhMikeS: so what do I have to do with MIS again? I understood that that indicates if the irq has already been generated
02:17:31jhMikeSI don't think this VIC handler is written correctly
02:18:00jhMikeSIt shouldn't need manual dispatching since it has a vector table built in
02:18:24jhMikeSMIS gives the status of unacked events that are also unmasked
02:19:07jhMikeSso writing those bits from xxx_MIS to xxx_IC will ack them in one assigment (assuming I'm reading this right)
02:21:10kugeljhMikeS: what about RIS? Is that uninvolved?
02:21:48ThundercloudI just found out something interesting
02:21:59jhMikeSit's a raw status reg that doesn't respect the masking register and reads the status, masked or not
02:22:02ThundercloudApparently the Cowon D2's processor can run java bytecode natively
02:22:30UnhelpfulThundercloud: a number of later ARM procs can, but many of the details are undocumented
02:22:36jhMikeSThundercloud: I believe the Gigabeat S can as well
02:22:43ThundercloudUnhelpful: That's... unhelpful
02:22:48Thundercloud(Undocumented that is)
02:23:06LloreanIt's not like there's any real benefit to core Rockbox in that ability anyway
02:23:07jhMikeSUndocumented != not doable :)
02:23:13ThundercloudSeriously if I could get a boot-over USB going for the D2 and a reasonable set of instructions I could really have fun with that.
02:23:16Unhelpfulyou have to *implement* a full JVM, and have it properly interfaced with a system the CPU uses to call your code when unhandled java opcodes cause exceptions.
02:23:16ThundercloudLlorean: No but it'd be amusing
02:23:23jhMikeSLlorean: A Java VM plugin? :)
02:23:25kugeljhMikeS: so, does this look good?
02:23:33LloreanjhMikeS: Plugin != Core
02:23:42ThundercloudUnhelpful: Er
02:23:52ThundercloudOr alternatively try and get some kind of documentation as to what works
02:23:55ThundercloudOr reverse-engineer stuff
02:24:04kugelI've taken everything out of the button_init_device function, except for VIC_INT_ENABLE = INTERRUPT_GPIOB
02:24:10ThundercloudThere's no point running a full JVM on something which can handle java natively right?
02:24:16UnhelpfulThundercloud: yes, that's also an alternative, to know which opcodes aren't handled.
02:24:39ThundercloudI'm willing to bet the D2's firmware is running java
02:24:53Unhelpfulyou still need to have your JVM, be it full or partial, be properly set up to be called when the CPU fails some java bytecode.
02:25:04ThundercloudMy guess is either J2ME or a subset of it
02:25:08jhMikeSkugel: sorry, pastebin seems to always be slow
02:25:14LloreanThundercloud: Why are you making that assumption?
02:25:17ThundercloudUnhelpful: It'd be a nice side project.
02:25:29ThundercloudLlorean: J2ME's designed for mobile phones
02:25:32ThundercloudOr mobile devices in general.
02:25:38ThundercloudIt would make sense to have a mobile processor support it
02:25:39*jhMikeS waits for pastebin to show the frickin' thing already
02:25:39LloreanThundercloud: That doesn't mean you run your firmware in it, though.
02:25:42kugeljhMikeS: just to let you know while you're waiting on pastebin, it doesn't work
02:25:42ThundercloudOr at least a subset of it.
02:26:03 Join BHSPitMonkey [0] (n=stephen@unaffiliated/bhspitmonkey)
02:26:11Unhelpfulmaybe, but i don't know of any public docs on handled vs unhandled parts of java, or how to interface your jvm to catch the unhandled stuff. :/
02:26:11kugelis pastebin generally slow to usa people? LambdaCalculus37 also has problems with it afair
02:26:18ThundercloudOr, alternatively, just thought
02:26:30ThundercloudSun microsystems developed a device called a SUNspot
02:26:34ThundercloudWhich also ran native java code
02:26:39 Quit mofux (Remote closed the connection)
02:26:45LambdaCalculus37kugel: I never had problems with pastebin.
02:26:55ThundercloudWhatever happens, it seems to have at least canvas/AWT support going on
02:27:13Unhelpfulpfft, old stuff. native high-level hardware's been done before. never heard of a lisp machine?
02:27:18Unhelpfulgetting OT, it seems, btw.
02:27:29ThundercloudUnhelpful: :P
02:27:32jhMikeSkugel: this thing needs nice vic_enable/disable_interrupt function that handle all the crud. This controller is more like the imx31 AVIC in operation.
02:27:33ThundercloudThing is, it's quite exciting
02:27:36ThundercloudI'm a Java man by profession
02:27:43ThundercloudHoping to get into the mobile industry
02:27:51ThundercloudThis would just be a damn cool eyeopener
02:27:57 Quit Strife89 ("Shower!")
02:28:19kugeljhMikeS: is VIC_INT_ENABLE |= INTERRUPT_GPIOB; not nice enough?
02:29:10kugeljhMikeS: also, do you mean, that the code *should* work, if the code was correct?
02:29:32kugelwith the latter code I mean the irq code, not the one I wrote
02:29:46 Quit HellDragon (Client Quit)
02:32:31jhMikeSkugel: that will enable interrupts in the VIC but you have to enable on the pins and handle those too
02:33:22kugeljhMikeS: yea. I already was curious about that. I've looked into other drivers and it seemed easier/less code
02:38:31jhMikeSWhy's this using simple (slow) rather than vectored handling (fast)?
02:39:20kugelno idea
02:41:03 Join johnm [0] (
02:42:54 Join toffe82 [0] (n=chatzill@
02:43:35kugeljhMikeS: do I need to call INT_GPIOB or with that be done by the irq handler?
02:46:17 Join nicfer [0] (
02:47:15kugeljhMikeS: Thanks a lot for your help. I tried it for one button now and it worked, although that button also works without irq stuff. Maybe you can help me further later
02:47:22 Join funman [0] (
02:47:28*kugel goes to bed
02:47:41funmanjhMikeS: because I didn't understand another way to use the VIC :o
02:48:38funmanI used the method described in the 'Programmer model' section of the datasheet
02:49:23ThundercloudOne thing i've never quite understood, what do you guys actually program rockbox in? ASM?
02:49:40funmankugel: nice that you work on using interrupts
02:49:42kugelfunman, jhMikeS: no matter how they work. We'll need it working for the fuze and e200v2
02:49:45Unhelpfuland C
02:49:53ThundercloudUnhelpful: I see
02:49:56kugelfunman: well, I need to :/
02:49:58ThundercloudMore of which though?
02:49:58funmankugel: it works fine already (it's used for DMA)
02:50:19kugelfunman: then you might guide me through it. But not today anymore.
02:50:49kugelbut I guess you won't guide anyone in your sound code addiction :p
02:50:50funmanThundercloud: C
02:50:55Thundercloudfunman: Cool
02:51:34funmanThundercloud: from the "contributing" doc: "Write all code in C. Sometimes assembly is faster, but C is always more readable and maintainable."
02:51:48Thundercloudfunman: Okay cool
02:52:21nicferHmmm, would be possible to run rockbox in those chinese portable players?
02:53:01kugelThundercloud: the entire apps/ code (except for codecs), i.e. UI, plugins, wps etc is in C
02:53:14kugelonly parts of the drivers in firmware/ are in ASM
02:53:21funmanjhMikeS: if i understand correctly, the addresses of each isr should be written to the VICVECTADDR register, and the arm irq vector should be ignored?
02:54:48*kugel goes to bed now, really
02:54:56 Quit kugel ("ChatZilla 0.9.84 [Firefox 3.0.4/2008111318]")
03:00:29jhMikeSfunman: pretty much. The ARM doc shows the procedure for setting that up.
03:02:02jhMikeSfunman: it's look in basic operation much like what's in avic-imx31.c (which would be a better model overall for this)
03:02:36funmanthanks, i add that to my TODO list :)
03:03:25krazykit`nicfer, it would be, should someone do the (very hard) work of porting.
03:03:30jhMikeSvectored interrupts are much easier for bootloaders too
03:04:26funmaneasier in which way?
03:04:53nicferBecause it's kinda hard to find sansa, irivers or all those rockbox capable devices here in my country
03:06:52jhMikeSfunman: because they're out of the way of loading the firmware and the default ARM "vectors"
03:08:28 Quit LambdaCalculus37 ("Ka-chunka")
03:11:16 Quit jhulst (Read error: 110 (Connection timed out))
03:11:18 Join Giulio [0] (
03:11:23 Join JdGordon_ [0] (n=jonno@rockbox/developer/JdGordon)
03:11:56 Nick Giulio is now known as Davide-NYC (
03:13:45 Part nicfer
03:15:26 Join Darksair [0] (n=user@
03:18:18 Quit JdGordon_ ("Miranda IM! Smaller, Faster, Easier.")
03:19:06 Join HellDragon [0] (
03:19:16 Part Davide-NYC
03:23:26 Join perrikwp [0] (i=18ac0c41@gateway/web/ajax/
03:30:14 Quit Thundercloud (Remote closed the connection)
03:35:37 Quit HellDragon (Client Quit)
03:36:26 Join HellDragon [0] (
03:44:41 Quit nuonguy ("This computer has gone to sleep")
03:54:43***Saving seen data "./dancer.seen"
03:58:59 Join jhulst [0] (n=jhulst@unaffiliated/jhulst)
03:59:53 Quit fdinel (Read error: 110 (Connection timed out))
04:02:22 Quit m0f0x (Read error: 110 (Connection timed out))
04:04:15 Join blkhawk- [0] (
04:13:13 Quit blkhawk (Read error: 145 (Connection timed out))
04:14:12 Nick blkhawk- is now known as blkhawk (
04:14:13Unhelpfulis the iaudio X5 remote the M3 remote again?
04:14:25Unhelpfulie, as with the iaudio M5 remote?
04:27:33 Nick fxb is now known as fxb__ (
04:56:43 Quit jhulst (Read error: 110 (Connection timed out))
04:59:03 Join nuonguy [0] (
05:00:01 Quit nuonguy (Client Quit)
05:00:25DEBUGReceived signal 15 (SIGTERM), terminating (snapshot: dancer.c line 124)
05:00:25***Saving seen data "./dancer.seen"
08:46:05***Started Dancer V4.16
08:46:05***Connected to on port 6667
08:46:05***Logfile for #rockbox started
08:46:08***Server message 501: 'logbot :Unknown MODE flag'
08:46:08Mode"logbot :+i" by logbot
08:46:08 Join logbot [0] (n=bjst@gateway/web/cgi-irc/
08:46:08 Join lasser [0] (
08:46:08 Join ender` [0] (
08:46:08 Join GodEater [0] (i=c2cbc962@gateway/web/ajax/
08:46:08 Join Zagor [0] (n=bjorn@rockbox/developer/Zagor)
08:46:08 Join Crackerizer [0] (
08:46:08 Join kugel [0] (n=chatzill@unaffiliated/kugel)
08:46:08 Join stoffel_ [0] (
08:46:08 Join denes_ [0] (
08:46:08 Join planetbeing [0] (
08:46:08 Join __lifeless [0] (n=lifeless@
08:46:08 Join esthar [0] (
08:46:08 Join n17ikh|Lappy [0] (
08:46:08 Join pixelma2 [0] (n=marianne@rockbox/staff/pixelma)
08:46:08 Join guyjones [0] (
08:46:08 Join amiconn_ [50] (n=jens@rockbox/developer/amiconn)
08:46:08 Join jhulst [0] (n=jhulst@unaffiliated/jhulst)
08:46:08 Join blkhawk [0] (
08:46:08 Join HellDragon [0] (n=jd@Wikipedia/HellDragon)
08:46:08 Join perrikwp [0] (i=18ac0c41@gateway/web/ajax/
08:46:08 Join Darksair [0] (n=user@
08:46:08 Join funman [0] (
08:46:08 Join johnm [0] (
08:46:08 Join BHSPitMonkey [0] (n=stephen@unaffiliated/bhspitmonkey)
08:46:08 Join AndyIL [0] (i=AndyI@
08:46:08 Join AndyI [0] (i=AndyI@
08:46:08 Join Bensawsome [0] (n=Bensawso@unaffiliated/bensawsome)
08:46:08 Join SUSaiyan [0] (
08:46:08 Join shadearg [0] (
08:46:08 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
08:46:08 Join jhMikeS [50] (n=jethead7@rockbox/developer/jhMikeS)
08:46:08 Join advcomp2019 [0] (n=advcomp2@unaffiliated/advcomp2019)
08:46:08 Join gromit` [0] (
08:46:08 Join soap [50] (n=soap@rockbox/staff/soap)
08:46:08 Join goffa [0] (n=goffa@
08:46:08 Join l7 [0] (
08:46:08 Join leachim6 [0] (
08:46:08 Join nplus [0] (n=nplus@141.25.Globcom.Net)
08:46:08 Join Bagder [241] (n=daniel@rockbox/developer/bagder)
08:46:08 Join Llorean [0] (n=DarkkOne@rockbox/administrator/Llorean)
08:46:08 Join agaffney [0] (n=agaffney@gentoo/developer/agaffney)
08:46:08 Join reacocard [0] (
08:46:08 Join JdGordon [0] (n=jonno@rockbox/developer/JdGordon)
08:46:08 Join Rob2222 [0] (
08:46:08 Join desowin_ [0] (n=desowin@
08:46:08 Join BlakeJohnson86 [0] (
08:46:08 Join krazykit` [0] (n=kkit@
08:46:08 Join Acksaw [0] (
08:46:08 Join Zarggg [0] (
08:46:08 Join midkay [0] (n=midkay@rockbox/developer/midkay)
08:46:08 Join MrDuck [0] (
08:46:08 Join HBK [0] (
08:46:08 Join sbhsu [0] (
08:46:08 Join daurn [0] (n=daurnima@unaffiliated/daurnimator)
08:46:08 Join daurnimator [0] (
08:46:08 Join synergist [0] (
08:46:08 Join _Auron_ [0] (
08:46:08 Join tchan [0] (n=tchan@lunar-linux/developer/tchan)
08:46:08 Join pabs [0] (
08:46:08 Join blithe [0] (
08:46:08 Join martian67 [0] (i=lol3izer@about/linux/regular/martian67)
08:46:08 Join tvelocity [0] (
08:46:08 Join ameyer [0] (
08:46:08 Join fyrestorm [0] (
08:46:08 Join markun [50] (n=markun@rockbox/developer/markun)
08:46:08 Join rasher [50] (n=rasher@rockbox/developer/rasher)
08:46:08 Join Bjoern-Erik [0] (
08:46:08 Join thegeek [0] (
08:46:08 Join axionix [0] (
08:46:08 Join balou [0] (
08:46:08 Join Xerion [0] (
08:46:08 Join andrew__ [0] (
08:46:08 Join notplus_M [0] (
08:46:08 Join stu8ball [0] (
08:46:08 Join @ChanServ [0] (ChanServ@services.)
08:46:08 Join J-23 [0] (
08:46:08 Join puzzles [0] (n=dan@xmms2/developer/puzzles)
08:46:08 Join TMM [0] (
08:46:08 Join ze [0] (i=ze@
08:46:08 Join idshark [0] (
08:46:08 Join buk_ [0] (
08:46:08 Join Dieterbe [0] (
08:46:08 Join Jabone [0] (
08:46:08 Join jfc^3 [0] (
08:46:08 Join tim__b [0] (
08:46:08 Join Zambezi [0] (
08:46:08 Join tarbo [0] (n=me@unaffiliated/tarbo)
08:46:08 Join DataGhost [0] (i=dataghos@unaffiliated/dataghost)
08:46:08 Join beta2k [0] (
08:46:08 Join FOAD [0] (
08:46:08 Join shodanX [0] (
08:46:08 Join Unhelpful [0] (
08:46:08 Join Kopfgeldjaeger [0] (
08:46:08 Join scorche [0] (i=Blah@rockbox/administrator/scorche)
08:46:08 Join Trista109 [0] (
08:46:08 Join cg [0] (
08:46:08 Join jchillerup [0] (
08:46:08 Join jon-kha [0] (
08:46:08 Join Tuplanolla [0] (n=jani@unaffiliated/tuplanolla)
08:46:08 Join preglow [0] (
08:46:08 Join DaCapn [0] (
08:46:08 Join amigan [0] (i=dcp1990@unaffiliated/amigan)
08:46:08 Join courtc [0] (n=court@unaffiliated/courtc)
08:46:08 Join dionoea [0] (n=dionoea@videolan/developer/dionoea)
08:46:08 Join flux [0] (
08:46:08 Join freqmod_qu [0] (i=quassel@2001:700:300:1430:213:d3ff:fee9:5ed0)
08:46:08 Join scorche|sh [50] (n=scorche@rockbox/administrator/scorche)
08:46:08 Join parafin [0] (
08:46:08 Join feisar [0] (
08:46:08 Join suom1 [0] (
08:46:08 Join larstobi [0] (n=larstobi@
08:46:08 Join crashd [0] (
08:46:08 Join Galois [0] (
08:46:08 Join ch4os [0] (n=ch4os@gentoo/user/ch4os)
08:46:08 Join Slasheri [0] (i=miipekk@rockbox/developer/Slasheri)
08:46:08 Join lastebil [0] (
08:46:08 Join crwl [0] (
08:46:08 Join Neovanglist [0] (i=Neovangl@
08:46:08 Join avacore [0] (
08:46:08 Join Zom [0] (n=zom@reactos/tester/Zom)
08:46:08 Join fred_2 [0] (
08:46:08 Join Hadaka [0] (
08:46:08 Join liiwi [0] (
08:46:08 Join maraz [0] (
08:46:08 Join lostlogic [50] (n=lostlogi@rockbox/developer/lostlogic)
08:46:08 Join fxb__ [0] (
08:46:08 Join Nimdae [0] (
08:46:09CtcpVersion from freenode-connect!freenode@freenode/bot/connect
08:46:14 Join pondlife [50] (n=Steve@rockbox/developer/pondlife)
08:46:22JdGordon#if !defined(BOOTLOADER) && 1 <- WTF?
08:46:43funmanit can be easily disabled
08:46:54Zagorsite is back up now
08:47:05kugelJdGordon: gets removed if it starts working
08:47:25pondlifeZagor: 1st of the month?? :)
08:47:59Zagorya. /me forgot to disable the automatic reboot...
08:48:10kugelfunman: it doesn't detect and/or read buttons at all
08:48:17 Join Rob2223 [0] (
08:48:32 Quit jhulst (Remote closed the connection)
08:48:38funmankugel: perhaps gpio(b/c)_afsel needs to be positioned on GPIO for interrupts to happen
08:48:40JdGordonthere is still the assumption that you can use interupts with the afsel anyway...
08:49:01kugelJdGordon: ??
08:49:28JdGordonmaybe we should switch the logic so the LCD always assumes they are set to the gpio instead of the dbop
08:49:54JdGordonseen as the lcd updates will only happen in the main thread context (while buttons are in a ticktask)
08:50:42kugelJdGordon: we need interrupts regardless of the lcd
08:51:26JdGordonyes, but if the gpio's are always set the the dbop they may not trigger interrupts (or if they do, they might do it incorreclty)
08:51:53 Join franky [0] (
08:54:04 Join Bagderr [0] (n=daniel@gateway/web/cgi-irc/
08:54:32 Nick Bagderr is now known as B4gder (n=daniel@gateway/web/cgi-irc/
08:55:26JdGordonhmm.. your also not touching the afsel in GPIOC
08:56:19kugelJdGordon: yea, I already noticed. Fixing didn't help
08:57:01JdGordondoes the button light definatly work?
08:57:12*JdGordon doesnt tihnk he as seen it work since it was commited
08:59:06 Join gevaerts [0] (n=fg@rockbox/developer/gevaerts)
09:00:12funmandoesn't work in multivolume build, since CCU_IO is set to SD setting and not GPIOD
09:00:59kugelJdGordon: switching the logic gives me white screens
09:01:28kugelJdGordon: no, it doesn't work. it needs a little fix
09:01:49*JdGordon back in 20
09:02:05kugelif I get the chance to test if SD works with that xpd mixed mode...
09:02:32 Join Jaykay [0] (
09:03:21funmanI think there is no way since the datasheet mention MemoryStick and not SD
09:04:00markunCrackerizer: we don't have support for combining glyphs right now (which you need for proper thai support)
09:05:28 Quit Rob2222 (Read error: 110 (Connection timed out))
09:07:27Crackerizermarkun: is this the kernel problem?
09:09:55Crackerizermarkun: and is it complicated to support combining glyphs?
09:10:16markunI don't know how difficult it would be to implement.
09:10:46markunit's a font rendering problem, but I don't think that should still be called part of the "kernel"
09:12:05markunwe would need to store a list of all the combining characters and when rendering, combine it with the previous one.
09:12:24markunand there can be multiple combining characters of course
09:12:56Crackerizermarkun: I thing i saw where the font rendering code is..
09:14:18markunlet me check. I don't think the actual drawing is done in firmware/font.c
09:14:19Crackerizerand i think it is possible to combine characters without storing any characters list.
09:14:38Crackerizermarkun: it's in lcd driver
09:15:38CrackerizerI guess this one. lcd_putsxyofs(int x, int y, int ofs, const unsigned char *str)
09:18:23markunI have to go now, so good luck with it
09:18:50Crackerizerhave a nice day markun.
09:18:57markunthanks, you too!
09:29:51kugelfunman: wouldn't it be cleaner to let sd_enable do nothing in the bootloader instead of #ifdefs at various places?
09:30:34funman#ifdef BOOTLOADER #define sd_enable(a) #endif would have the same effect
09:31:27kugelis it available in the bootloader for the pp flash targets?
09:33:37 Quit MrDuck (Read error: 113 (No route to host))
09:38:45 Join beta2k_ [0] (
09:39:35 Quit beta2k (Read error: 104 (Connection reset by peer))
09:42:13 Quit kugel (Remote closed the connection)
09:43:01 Quit Jaykay ("ChatZilla 0.9.84 [Firefox 3.0.4/2008102920]")
09:45:52 Quit lasser (Remote closed the connection)
09:47:53JdGordonhmm... is there no way to force the buttonbar to not show (other than global_status.buttonbar?
09:48:39 Join einhirn [0] (
09:50:55 Join n1s [0] (n=nils@rockbox/developer/n1s)
09:53:46JdGordonn1s: you asked about fs#9260 yesterday? no I'm not really plannig on working on it any time soon...
09:54:30n1sJdGordon: ah, ok, is it correct though that this could cause dir operations to fail because of too many open dirs?
09:55:39JdGordonyes, if ft_open() isnt used
09:56:22funmancan I request that cabbiev2 top 2 lines are shifted bottom by 1 pixel to fit perfectly into the yellow area of the Clip's screen ? (I don't know if this change can be made model specific)
09:57:44 Join lasser [0] (
09:57:52JdGordonfunman: the clip will either need a new "screen type/size" setup, or maybe special handling to drop those pixels down in the wps?
09:58:42*JdGordon guesses our archos users have the buttonbar disabled?
09:59:02 Join LinusN [0] (n=linus@gateway/web/cgi-irc/
09:59:03funmanwell the progress bar and the title fit in 14 pixels, 16 pixels would be perfect
09:59:10 Part LinusN
09:59:20 Join LinusN [0] (n=linus@rockbox/developer/LinusN)
09:59:49*JdGordon really wants some sort of status/button bar manager....
10:00:07 Join nuonguy [0] (
10:00:11pondlifeJdGordon: If the user wants to enable them, they should appear
10:00:45pondlifeAlthough, the current exception for the WPS is reasonable until we can theme them.
10:00:55amiconn_funman: Nice solution to the mask problem. 0xffffffff would probably have worked too though, and that you can set with a single instruction.
10:01:04JdGordonright, but I've just fixed a redraw bug with it that has been there since the menu recode last year :D (although it probably didnt hit the ondios which may be why noonw noticed?)
10:01:56amiconn_The Ondios don't have the button bar of course, as they have no soft buttons
10:02:02 Nick amiconn_ is now known as amiconn (n=jens@rockbox/developer/amiconn)
10:02:04JdGordonoh right :p
10:02:16funmanamiconn_: ah right, mov instruction has a 'not' bit?
10:02:16JdGordonalthough the reason i meant it didnt hit the ondio was no quickscreen :/
10:02:28amiconnfunman: 'mvn' instruction
10:02:56*pondlife would like to add "Browse CFG files" as a quickscreen option
10:02:56amiconnJdGordon: That too (due to lack of buttons)
10:03:15 Join kugel [0] (n=chatzill@unaffiliated/kugel)
10:03:27amiconnThe quickscreen has no button bar, btw
10:04:04JdGordonyeah, thats the problem.. the menu doesnt ever redraw the bb, so its obvious when you come out of the quickscreen
10:05:09JdGordonpondlife: that would be good... but would it load the cfg as soon as it changes? or on exit?..
10:05:29*amiconn has the button bar disabled, as he knows what the buttons do and prefers the extra screen space, but the button bar is useful for beginners, or if you don't use the device often
10:05:44pondlifeJdGordon: Immediately, I think
10:05:45kugelJdGordon: I suppose the option would just get you into the browser
10:06:16pondlifeThe most common config change I make is to select car.cfg or dj.cfg..
10:06:18kugelor list every cfg in the quickscreen?
10:06:23 Join petur [50] (n=petur@rockbox/developer/petur)
10:06:46JdGordoncfg browsing was something which I've wanted to add for a while, but ti will be "complicated"
10:07:15pondlifekugel: Well, ideally I'd just map the long setting button to browse CFG files and forego the quickscreen... (for my particular usage pattern)
10:07:19JdGordonor... like kugel said.. it could just point to the browser
10:08:27pondlifeNot a biggie though - it's not many button presses at the moment: MENU - DOWN - LONG SELECT
10:08:56funmanI'm still looking why my Clip deadlocks when playing audio files .. I was watching the stack levels but they keep constant (codec stack used at 91% however)
10:11:05amiconnYea, libmad uses a lot of stack, so the usage will be >90%, but that's okay
10:11:46funmanI also notice frequent skips when decoding mp3
10:12:28funmanI should run test_codec over various bitrates/codecs
10:13:37*JdGordon is going to commit the time/date screen patch unless there is some major yelling....
10:13:47funmannote: the clip deadlocks (I see the scrolling on screen stop) but the PCM buffer continues to be played (until the current DMA transfer finishes I suppose)
10:13:57pondlifeJdGordon: YELL! ;p
10:14:06pondlifeNah, don't let me stop you
10:15:00JdGordonI *think* i've fixed all the drawing issues
10:15:13*kugel notices JdGordon tends to commit stuff when most devs aren't active
10:15:25*JdGordon slaps kugel
10:15:36*pondlife hands kugel a fresh trout
10:15:50*kugel expects a post-commit discussion in about 4h
10:16:06JdGordonat least i can point to the ml for this one!
10:16:35n1sthis one was at least brought up on the ml several days before commit though
10:16:44kugelhave you fixed the voice?
10:17:02JdGordonyes, checking now.. voice works goodly
10:17:27kugelthen I think it's worth the upcomming discussion
10:18:22kugelmaybe there even won't be a discussion. iirc there was no discussion about my backlight fading patch (and it was announced in the ml too)
10:19:21kugelJdGordon: is it still a top level item? I'd like that, so I could disable the statusbar, which I mostly only watch at for the time
10:19:31JdGordonno, its the first item in system
10:21:01 Quit GodEater (Remote closed the connection)
10:21:02 Quit perrikwp (Read error: 104 (Connection reset by peer))
10:21:56 Join Xerion_ [0] (
10:23:04pondlifeJdGordon: Does it also remove the time/date from the info screen?
10:24:09JdGordonand fixes the voice problem there which was the origional motivation
10:26:26 Quit Xerion (Read error: 60 (Operation timed out))
10:26:26 Nick Xerion_ is now known as Xerion (
10:27:16 Quit funman (Read error: 54 (Connection reset by peer))
10:28:13 Join funman [0] (
10:31:00funmanhm now I just got a data abort in memset()
10:31:09 Join GodEater [0] (i=c2cbc962@gateway/web/ajax/
10:31:38*Zagor looks at the wikipedia plugin
10:31:47funmanI'd like to write a coredump on the internal storage, but I don't know how to get the values of the registers when this happens
10:31:55 Quit GodEater (Remote closed the connection)
10:33:11 Join Genre9mp3 [0] (n=yngwiejo@rockbox/contributor/Genre9mp3)
10:33:41 Quit Genre9mp3 (Client Quit)
10:36:23 Join MrDuck [0] (
10:37:02n1sfunman: hack something into an exception handler?
10:37:20 Quit BHSPitMonkey (Remote closed the connection)
10:38:36funmanhum it seems the registers aren't modified after all
10:38:59funmani'm just unsure if i still can use SD now that a thread has gone into a corrupted state
10:39:37n1sJdGordon: are you going to update the manual?
10:40:21JdGordonI fail at TeX... but I can give you plaintext ig you want
10:40:42JdGordonanyone think 8814 is worth trying to do anything about?
10:42:17n1sI think we can remove the splash, it's not very useful
10:42:32*kugel can't get the interrupts to work
10:44:05 Quit funman ("leaving")
10:44:08JdGordonn1s: I'm not really sure what to write for it... it seems the alarm options arnt in the manual at all? and the text for rockbox info wasnt updated for the time/date anyway
10:45:07JdGordonthe only important bit is pressing context-menu will voice the time
10:45:35 Join GodEater [0] (i=c2cbc962@gateway/web/ajax/
10:46:02n1swrite a short description of the screen mentioning the voice thing, if the other things are missing in the first place you can add them if you want to :)
10:46:07***Saving seen data "./dancer.seen"
10:47:04kugelJdGordon: What about having a "Speak time" menu item (which is only there when voice is set on), so that it's more obvious?
10:47:05 Join AJG983 [0] (n=621b8510@gateway/web/cgi-irc/
10:47:27JdGordonkugel: I tried doing that, but couldnt get it to work properly
10:47:38JdGordonthe voice for that option would get spoken instead of the time
10:47:49*kugel has some doubts that visually impared which "read" the manual
10:47:52AJG983Anyone know where I can find out the current status on the development of the build for a cowon d2?
10:48:34AJG983I did a search on the main rockbox site and only results that came up were two topics on patches
10:48:39n1sJdGordon: wake-up alarm is in the manual
10:48:41kugelat least not so intensively, so they don't even get to the time & date screen part
10:48:53JdGordonah yes, found it
10:49:00scorcheAJG983: check the corresponding thread under the New Ports forum
10:49:37 Quit AJG983 (Client Quit)
10:49:55n1skugel: visually impaired and blind users often use screen readers or Braille displays for reading and can indeed read lots of text
10:52:53kugeljhMikeS: ping
10:53:53 Quit n17ikh|Lappy ()
10:57:58JdGordonn1s: something like ?
11:00:00n1sJdGordon: looks fine, \opt{!foo} doesn't work however, use \nopt{foo} instead
11:00:15*JdGordon isnt doing a manual commit...
11:00:22JdGordonit will ruin my reuptation :D
11:00:35GodEaterJdGordon: not if it breaks the manual ;)
11:00:40JdGordonand that \Button was a guess
11:00:48n1sok, I'll do it then :P
11:01:55n1syay, 4 out of 4 hunks failed
11:04:02kugelGodEater: and it apparently would have broken it :)
11:04:20n1shaha, that ubuntu paste things pads all lines with a bit of whitespace for fun
11:04:41kugeln1s: it also adds CRs
11:05:16n1sthis is monumentally stupid
11:05:20kugeln1s: you should probably click on "download as text" at the top
11:05:33n1skugel: that's what i did :)
11:05:45kugellol, they're even there
11:06:12kugelI never had that
11:06:27*kugel tends to blame JdGordon for that
11:07:11JdGordonok, i wont use that pastebin ever again!
11:07:23kugelonly the first 2 thirds have white spaces?
11:08:10 Join skipper [0] (n=skipper@
11:13:25 Quit nuonguy ("This computer has gone to sleep")
11:18:31 Quit GodEater (Remote closed the connection)
11:19:24 Quit kugel (Remote closed the connection)
11:24:27 Join GodEater [0] (i=c2cbc962@gateway/web/ajax/
11:37:25 Join mofux [0] (
11:37:39 Nick fxb__ is now known as fxb (
11:41:35Zagormuh, where was this -Dread=sim_read defined again?
11:42:29*B4gder doesn't recall where...
12:01:38 Join kugel [0] (n=chatzill@unaffiliated/kugel)
12:08:47 Join Nico_P [50] (n=nicolas@rockbox/developer/NicoP)
12:23:02 Join bmbl [0] (n=Miranda@unaffiliated/bmbl)
12:32:11 Quit GodEater (" ajax IRC Client")
12:32:19 Join GodEater [0] (i=c2cbc962@gateway/web/ajax/
12:32:51 Join JoelanJeffreys [0] (n=792cc1d0@gateway/web/cgi-irc/
12:36:38 Join tyfoo [0] (
12:38:18 Quit JoelanJeffreys ("CGI:IRC (Ping timeout)")
12:40:21 Join Thundercloud [0] (
12:40:56 Quit kugel (Nick collision from services.)
12:40:58 Join kugel_ [0] (n=chatzill@
12:41:14 Nick kugel_ is now known as kugel (n=chatzill@
12:41:34 Quit kugel (Remote closed the connection)
12:43:51 Quit tyfoo ("Carpe diem")
12:46:09***Saving seen data "./dancer.seen"
12:49:14 Quit skipper (Remote closed the connection)
12:49:55 Join TheSphinX^ [0] (
12:54:16n1sjhMikeS: should FS #8270 be closed?
12:56:28 Quit TheSphinX^ ("XChat@Linux")
13:00:18*n1s thinks the "Operating system" <-> "Player type" change got lost in the FS upgrade
13:00:42n1sZagor: ping (see previous line)
13:01:15 Quit AndyIL ()
13:08:17 Join tyfoo [0] (
13:12:18 Quit tyfoo (Client Quit)
13:13:06lasserhi there. At first: congratulations to devs for sound on sansav2! Great Job!
13:14:55 Join Schmogel [0] (
13:15:31 Join HyperHorse_ [0] (
13:15:51HyperHorse_ello ello???
13:16:26lasserBut for now there seems to be a problem on my e280v1 running Rb r 19284. Whenever I turn it on it stucks in boosted mode. So tells me the "View buffering thread" in the debug menu.
13:17:23JdGordonlasser: a temporary workaround is go into the cpu boost option in the debug and use the wheel to drop the count to 0
13:17:30ZagorHyperHorse_: We will stick out like a carrot in an omelet.
13:18:08HyperHorse_hmm.. what?
13:18:09lasserJdGordon: yes, I noticed that. Is this a bug?
13:18:21HyperHorse_Zagor: what was the point of saying that?
13:18:23 Join skipper [0] (n=skipper@
13:19:30JdGordonlasser: playback is known to have some difficult to trace issues which can lead to a stuck boost...
13:19:53HyperHorse_Zagor: that series was a classic
13:20:09HyperHorse_anyone seen "Network" ??
13:20:28JdGordonthis is a on-topic channel...
13:20:34ZagorHyperHorse_: come into #rockbox-community for idle chat
13:20:53HyperHorse_so there's no support yet for the creative zen???
13:21:11CrackerizerI'm hacking the rockbox's code for fully thai support and i'm about 50% done.
13:21:17ZagorHyperHorse_: it is being worked on, but not yet suitable for public consumption
13:21:36HyperHorse_Crackerizer: you're working on the zen?
13:22:10CrackerizerHyperHourse: No, i'm working on ipod video. but i thing it should work on all model.
13:22:28n1sCrackerizer: how intrusive are the needed changes? and how much binary size increase and ram usage increase do you expect?
13:22:29lasserJdGordon: hm, but these difficulties are introduced in the last weeks, aren't they? Before I never noticed a behaviour like that...
13:23:02JdGordoni doubt that... iirc not much work has happened on playback for a while.. so its probably much older
13:23:19JdGordonCrackerizer: what sort of support? just translation?
13:23:29Crackerizern1s: not much. just a few condition checks. and bitmap building function calls.
13:23:42n1ssounds good
13:23:54CrackerizerJdGordon: no, I have translate rockbox to thai while ago.
13:24:26CrackerizerI'm trying to make it display thai alphabets and vowels correctly.
13:25:29CrackerizerIt seems like there are 2 functions for putting an alphabet on the screen.
13:26:19*JdGordon would like to get RTL going but already has a really long todo list :p
13:27:19Nico_PJdGordon: did you get results with the PCM mixer?
13:27:40JdGordonno, gave up on it for now... I do want to start that again though
13:28:04CrackerizerSome thai text displayed correctly but some not. for example, the highlight menu items display correctly but the normal menu items dont.
13:33:25 Join rvvs89 [0] (
13:36:10 Join robin0800 [0] (
13:40:41*JdGordon has a quick groan at chessclock which obviously hasnt had an update since it was first commited
13:40:53JdGordon2 line display custom everything :)
13:43:33gevaertsJdGordon: you can also make it crash with less than 5 keypresses (I think even 2, but I'm not sure)
13:43:52JdGordonyour bug report is what made me look at it
13:44:06JdGordonI tihnks its less work to rewrite the whole thing than to try to find the bug
13:45:51 Join kugel [0] (n=chatzill@unaffiliated/kugel)
13:48:11JdGordonhas anyone seen rediculous runtime estimates in the rockbox info screen? (FS #8945)
13:50:31 Quit Nico_P (Remote closed the connection)
13:51:20kugelJdGordon: wasn't that fixed?
13:51:43JdGordonaccording the the old comments it was still happening... but that was from months ago
13:51:44kugelI remember it being a signed <-> unsigned issue
13:52:29kugelor rather a unsigned counter counts still down after 0
13:55:08 Join omgun [0] (n=79b45a2d@gateway/web/cgi-irc/
13:56:16jhMikeSn1s: Depends if it's a bug or not. I don't have a nano to assess the issue. The one reporting never even posted a comment there after opening it.
13:57:32amiconnIn some cases extreme runtime estimates are expected (namely on Ondio when running from USB power)
13:58:33CrackerizerI'm out of idea. any suggestion please..
13:58:46JdGordonCrackerizer: show us your patch...
13:59:16CrackerizerJdGordon: how? send it here?
13:59:35JdGordonor go straight to our patch tracker
14:00:18 Quit HyperHorse_ ("be back later")
14:01:49 Quit bmbl (Read error: 110 (Connection timed out))
14:02:01Crackerizeri made some changes to this function.
14:02:32JdGordonfirst of all... you should do "svn diff > patch.diff" instead
14:03:32JdGordonand whats the problem?
14:03:36CrackerizerBut i did's check out the source from svn...
14:03:50linuxstbCrackerizer: Some comments would be helpful as well ;)
14:03:50 Quit omgun ("CGI:IRC (EOF)")
14:04:26Crackerizerlinuxstb: thanks for the advice :)
14:04:41linuxstbYou're welcome.
14:06:35Crackerizerok...I added condition checks on the unicode. if the next alphabet is a vowel, it will be displayed at the same position as the first alphabet.
14:07:15 Join bmbl [0] (n=Miranda@unaffiliated/bmbl)
14:07:29Crackerizerthe problem is, it works only when the text is highlighed.
14:07:51JdGordonwhat happens when it doesnt work?
14:07:57JdGordonyou just get the vowel?
14:08:05CrackerizerjdGordon: yes
14:09:00 Quit kugel (Remote closed the connection)
14:09:55 Join kugel [0] (n=chatzill@unaffiliated/kugel)
14:10:57JdGordonI would probably go about this differently... hack font_get_bits() to merge the 2 glyhps instead of drawing twice
14:11:08JdGordonthat means you dont need to modify every lcd driver
14:14:57 Join pondlife1 [0] (
14:16:43B4gdertracker cleanup!!
14:17:03CrackerizerJdGordon: Yes, that the first thing i tried. There is a trade off between simplicity and performance.
14:17:19 Quit jhMikeS (Nick collision from services.)
14:17:21JdGordonCrackerizer: no idea if this will work... (blind very tired guess) updating your paste...
14:17:25 Join jhMikeS [50] (n=jethead7@rockbox/developer/jhMikeS)
14:17:48pixelma2I think (hope) funman reads logs... I'm looking into cabbiev2 for the Clip. Currently it uses the wps file which was made for the Iriver remotes (same display size but used with a larger font since you can't have two different font, except sysfont + userfont, at the same time) and is "optimised" for that larger font but I think it is possible to have 2 different WPSs with the current build system but needs some experimentation. On a related no
14:17:48pixelma2te: I'm using your patch to make the Clip sim more realistic and it has a problem with scrolling in the lower part - it scrolls without the 2px offset which is there for the static display
14:17:50CrackerizerJdGordon: I'll test it :)
14:17:58pixelma2oops, sorry
14:18:15n1sB4gder: talking tracker cleanup, what do you think about FS #9338 ?
14:18:37JdGordonCrackerizer: i added the 3 lines with vp->drawmode
14:18:46JdGordonlike i said, very blind half asleep guess
14:19:31B4gdern1s: I think the gmake thing is fine, the rest I'm not so sure about
14:20:20B4gderI like the script to download files first, then unpack them, then build them
14:20:50B4gderand I made it fail if the build dir already exist to prevent mistakes to happen if you accidentally run it again on the same dir
14:20:54n1sB4gder: could you commit the fine part and close it? ;)
14:21:11B4gderI will
14:21:15pixelma2funman: just realising that sysfont is 8 pixels as is the default 8-rockfont so maybe the Iriver remote could use that and the layout would be simiilar (although sysfont is wider)
14:21:16n1sgreat :)
14:21:31CrackerizerJdGordon: is vp global?
14:22:31CrackerizerJdGordon: i changed it to current_vp instead..
14:22:32JdGordoncurrent_vp not vp
14:24:27Crackerizerreturn segmentation fault
14:24:47 Join Strife89 [0] (n=michael@
14:25:29Strife89I see you guys have sound on the Sansa AMS targets! :)
14:25:36CrackerizerI thought every text display will call this function..
14:26:18CrackerizerI'm very new to rockbox development. just started it today..
14:26:47JdGordonCrackerizer: ok, change that DRAWMODE define to DRMODE_BG or if that doesnt work... DRMODE_FG
14:30:59JdGordonno that prob wont work
14:32:16Crackerizerthe DRMODE_FG works.
14:32:23JdGordonok, cool :)
14:32:39amiconnIt won't always work, depending on what drawmode the text output functions are called with
14:33:19JdGordonfixing the font code to merge the glyphs is still probably the best way to do it
14:33:21CrackerizerWill this affect the other language?
14:33:40 Quit pondlife (Read error: 110 (Connection timed out))
14:33:43n1samiconn: would you call -c3000 usable on coldfire targets now?
14:33:56CrackerizerjdGordon: Yes, that way probably faster..
14:34:07amiconnn1s: It is usable at least with hw tone controls (i.e. irivers). Didn't try iAudios
14:34:42amiconnBuffering while playing a -c3000 track (i.e. starting with it) took ~30 sec on my H180
14:34:59JdGordonCrackerizer: these letters would ALWAYS need the vowels? and its always the same vowel?
14:35:33n1samiconn: maybe it's a bit tight to call it usable in the manual then
14:35:47JdGordonif it is you probably dont even need extra buffers to merge them.. overwriting the glyph cache wouldnt be a prolem
14:36:37CrackerizerJdGordon: no, not always.
14:36:48amiconnCrackerizer: I didn't look at the patch. I don't think it would affect other languages, but you need to be very careful about modifying the drawmode. At the very least you have to change it back before returning, and the modified drawmode needs to depend on the original one
14:37:30JdGordonCrackerizer: your patch will always put the vowel on?
14:38:02amiconnJdGordon: Overwriting is not an option, as (1) the glyph cache isn't reloaded all the time and (2) a small enough font is loaded completely and doesn't use lru.
14:38:22CrackerizerJdGordon: It will put on only if the unicode is matched.
14:38:43Crackerizerthere are 2 levels of vowel in thai
14:38:46JdGordonright, but if the same glpyh always needs the vowel, then overwriting that glyphs cache isnt bad?
14:39:36amiconnWell, if it does, why isn't there a combined glyph in the first place?
14:39:54CrackerizerJdGordon, amiconn: It doesnt
14:39:56 Quit kugel (Read error: 110 (Connection timed out))
14:40:15amiconnYeah, that's what I thought
14:40:24JdGordonbut you are unconditionally adding the vowel?
14:41:05CrackerizerJdGordon: I dont understand..
14:41:08amiconnI wonder whether there are code points with the appropriate precomposed glyphs though, similar to how e.g. the arabic presentation forms exist.
14:41:17amiconnThen a little preprocessing would solve the problem
14:42:19amiconnJdGordon: Btw, copying the font bitmaps into some buffer before drawing is an old idea of mine, which I still consider useful.
14:42:43JdGordonI thought it was abandoned because it wouldnt speed things up?
14:42:51amiconnIt would allow several things. (1) composing glyphs - not only for thai, but also for diacritics in latin-based languages.
14:43:16JdGordonwould a 64 byte buffer be enough for any usable font glyph?
14:43:25amiconn(2) Switch to a different mono bitmap format on targets where it would make sense (e.g. horizontal packed mono on the greyscale ipods)
14:43:48amiconn(3) Allow glyph postprocessing, e.g. algorithmic emboldening
14:43:57CrackerizerI'll be back. out for dinner..;-)
14:45:37amiconnJdGordon: No, it wouldn't, considering large fonts. And my idea was to buffer a whole line of text as a mono bitmap before drawing.
14:46:10***Saving seen data "./dancer.seen"
14:48:09JdGordonok, but for a single glyph... it would just be a matter of having a large enough static buffer and oring the bytes right?
14:53:59 Join LambdaCalculus37 [0] (i=44a04303@gateway/web/ajax/
14:57:15jhMikeSWhat's the proper manner to delete a .lang string for some targets? Some compel charging from USB by hardware and having an option for it is pointless.
14:57:51JdGordonset *: "" and fix the feature for it for the rest
14:58:07JdGordonunless of course *: is none already... then things get messy
14:59:57jhMikeSLANG_USB_CHARGING is the string. usb_charging:"...", *: none, *:none
15:01:19JdGordonI guess add the targets you want removed under the usb_charging line
15:01:29JdGordonunless you want to remove the string totally?
15:02:25jhMikeSno, it's fine for ones where can be decided by software like H300. Gigbeat F/X and H10 handle it with hardware. Wish I had a complete list.
15:02:32 Join Horscht [0] (n=Horscht@xbmc/user/horscht)
15:02:59jhMikeSOr is H10 5GB different than 20GB. amiconn?
15:04:01pondlife1Anyone able to suggest how I could merge the PCF50605/50606 RTC code - is as synced as I could get, but need to make slightly different low-level calls.
15:04:20pondlife1e.g. disable_irq_save()/pcf50606_read_multiple()/restore_irq() vs. pcf50605_read_multiple() etc.
15:04:22amiconnjhMikeS: Different in what respect?
15:04:43amiconnn1s: Doest {gigabeat} include all gigabeats? Do we already have a 'best' manual?
15:05:02jhMikeSamiconn: does H10 5GB handle all charging with hardware, USB or otherwise, like H10 20GB?
15:05:16amiconnHow would I know?
15:05:48n1samiconn: yes, it includes both f/x and s but the s manual doesn't build anyway
15:05:49jhMikeSH10 20GB will charge from any source regardless if it's turned on
15:06:15jhMikeSNo software is required to engage it
15:06:23amiconnThe small H10 always charges when plugged into USB. No idea about external charger as I have none.
15:06:48amiconnThere's a little LED on the DC plug of the H10 cable that indicates whether it's charging or not
15:06:49jhMikeSOn my H10, I just use the charger from the H120 (same one)
15:07:28jhMikeSok, sound like the same thing then
15:08:56JdGordonjhMikeS: ok, your options are rework the feature to exclude the targets (will break exsiting lang/voice files though) or add the target line and mark it somehow as deprecated so if/when deprecated strings are removed that will be fixed up
15:08:58n1sjhMikeS: i didn't quite get what you wanted to do with the lang string but if you change which targets match 'none' the string/clip order in lng and voice fiels for those targets will break
15:09:30amiconnThe H1x0 charger plug does indeed fit into the H10 cable. Never tried that before...
15:10:03 Quit Strife89 ("Huzzah!")
15:12:29 Nick JdGordon is now known as JdGordon|zzz (n=jonno@rockbox/developer/JdGordon)
15:13:13Zagorugh. the nes enumlator depends on ++ evaluation order.
15:13:26Zagorenumlator is a great word btw
15:14:17amiconnstructlator, lol
15:15:07jhMikeSn1s: Just removing the string does break everything. I wanted to make it as unnoticed as possible. You can mark "deprecated" for particular targets then?
15:15:41amiconnNot in a way that's detected automatically on next cleanup
15:16:02amiconnYou'd need to set it to "" for the affected targets, and add a note to the description field
15:18:34amiconnn1s: Just in case you or someone else will make the beast manual build - the beast can handle ape up to -c4000 nicely
15:18:50*jhMikeS needs to know which targets besides H10 and GBF/X handle the charging automatically. Any archos?
15:19:19amiconnArchos doesn't charge from USB apart from fm/v2 recorders
15:19:34amiconnI don't know whether this is fully automatic or not
15:19:57amiconnn1s: Also note that the speeds given are only valid for 44.1kHz 16bit stereo with current bitstreams (v 3.99)
15:20:42n1samiconn: true, i'm pretty sure the majority of people use 44.1/16 though
15:20:57amiconnMono is a little more than twice as fast. Different sample rates will have the expected effect
15:21:27n1sjhMikeS: i don't _think_ we can do deprecated for some targets only, but deprecating a string is technically the same as setting it to ""
15:21:32amiconnAlso, older bitstrams (3.97) are faster at the same -c level (but that -c level means less compression as the same level for v 3.99 streams)
15:21:43n1s(which we can do without breaking the order)
15:22:16jhMikeSn1s: so long as it gets cleaned later
15:22:58amiconn24bit is probably a lot slower - didn't test that yet but will soonish
15:24:07CrackerizerI'm back..
15:25:09Crackerizeramiconn: I think adding pre-postprocess should be a feature not just a patch..
15:25:33 Quit einhirn (Read error: 54 (Connection reset by peer))
15:26:23 Join nicfer [0] (n=usuario@
15:26:29n1sJdGordon|zzz: I think FS #9507 was confirmed
15:27:10n1sDid anyone ever see the problem where mpegplayer audio would not play if music was paused when starting the plugin? (specifically on ipod video)
15:27:42nicferhmmm would someday those chinese 'mp4' players be supported by rockbox?
15:28:07B4gdernicfer: specifc ones sure if someone wants it bad enough
15:28:10n1snicfer: if someone works on it. maybe
15:28:45nicferI'm from Argentina, those are very popular here
15:28:59B4gderwell "those" are generally not the same one
15:29:10B4gder"mp4 players" tend to lump everything from china under one label
15:29:22B4gderwhile they are different in reality
15:29:34jhMikeSn1s: It seems like the mute gets stuck on or something. Audio isn't paused since video is playing and audio is the clock.
15:30:12n1sjhMikeS: so you have seen this hapen?
15:30:46nicferthere are also some phillips and sony players
15:31:07nicferlet me think other brands available here
15:31:07B4gdernicfer: there exist players by many brands yes
15:31:18jhMikeSn1s: I just did it on gigabeat F
15:31:26B4gdernicfer: I'm sure even Argentinian people have internet => ebay etc
15:32:04B4gdernicfer: but the answer is the same disregarding of brand: if someone wants to they work on a port
15:32:06n1snicfer: developing support for one player is a very long and laborious process and is done by people who own those players and have the necessary skill and determination
15:32:22n1sjhMikeS: ok, i'll reopen the bug report
15:32:43jhMikeSn1s: if fade on stop/pause is enabled, it does it if stopped or paused
15:32:58 Quit Thundercloud (Remote closed the connection)
15:33:27jhMikeSI would suspect other plugins that play sound would be affected too
15:33:51n1sjhMikeS: please add any comments to FS #9507
15:34:02jhMikeSjust checked. metronome too
15:35:56n1smaybe pcm_play_data() should make sure that sound is not muted?
15:36:23jhMikeSI think the muting in pcmbuf.c should go. e200, where it's disabled has to problem.
15:36:41n1ssounds like a simpler solution, yes :)
15:37:18 Join Thundercloud [0] (
15:38:25jhMikeSI think there was a reason that it was left on gigbeat F/X but later PCM driver work may have made it unnecessary
15:38:57n1sthe bug was initially reported for ipod video, does it do the same as the gigabeat?
15:40:00jhMikeSYes, because PCMBUF_MUTING is left #defined for those. I was checking with specific audio codecs and disabling it for those but I've never really found one it did any good on.
15:43:34 Join kushal_12_27_200 [0] (n=kushal@
15:54:24 Quit kushal_12_27_200 (Read error: 145 (Connection timed out))
15:54:35 Join kugel [0] (n=chatzill@unaffiliated/kugel)
15:54:57 Quit MrDuck (Read error: 110 (Connection timed out))
15:55:45 Quit n1s ()
15:58:11 Join {phoenix} [0] (
16:01:41nicferhmmm, of all the devices supported by rockbox, what's the best (ipods not including)
16:02:38nicferoh I forget to visit that
16:04:32daurnlulz, none in production
16:05:23 Join blahrus [0] (n=blahrus@
16:05:53daurnnicfer: gigabeat, iaudiox5 in hdd
16:06:20daurnnicfer: otherwise the sansa are good for flash
16:12:39CrackerizerWhat is the target build for creative zen?
16:16:00nicferI can't found in ebay neither the sansa c200 or the e200
16:16:12nicferthere are only e280
16:16:27B4gdere280 is a unit in the "e200 series"
16:17:55 Part pondlife1
16:17:58nicferI don't know if they are v1 or v2
16:19:28 Join jgarvey [0] (
16:19:32 Quit bmbl (Read error: 104 (Connection reset by peer))
16:25:11Unhelpfulfigured out the m3 remote problem last night. need to rework scale_nearest more than a little, but i think i see how to fix it now... it was to do with the fb_data and fb_remote_data types being defined per-platform
16:26:37 Part B4gder
16:33:20Zagorhow do I build rbutilqt?
16:33:57Zagorah, RockboxUtilityDevelopment#How_To_Compile
16:35:32 Join tyfoo [0] (
16:37:28 Quit Darksair ("Do you hear that? This is the sound of inevitability. This is the sound of your death, Mr. Anderson.")
16:37:54 Join gromit`` [0] (
16:38:08kugeljhMikeS: ping
16:38:53 Join toffe82 [0] (
16:40:57 Quit {phoenix} (Remote closed the connection)
16:41:05jhMikeSthis LANG_USB_CHARGING stuff was added only to support it on H300 in r12169 :\
16:41:25 Join Darksair [0] (n=user@
16:43:18kugeljhMikeS: two questions regarding irq: a) the datasheet reads for all gpio interrupt register "All bits are cleared by a reset". What's the reset here? Powering of the unit or setting GPIO*_IC
16:43:34 Quit Darksair (Client Quit)
16:44:06jhMikeSprobably powering or activating the RESET line on the SoC
16:44:36kugelsecond: I think I only want interrupts when the edges raises (for the power button as high means button pressed)/falls(for the other buttons as low means button pressed), not both, right?
16:44:39 Quit Crackerizer (Read error: 110 (Connection timed out))
16:45:02kugelIBE (which you told me to activate) enables "both edge detection" though
16:45:35 Join Crackerizer [0] (
16:45:36 Join Darksair [0] (n=user@
16:45:47jhMikeSkugel: I'd say you'd still want it for both edges otherwise you only get interrupts for one action. Pressed/not pressed is decided by reading the pin level.
16:46:12***Saving seen data "./dancer.seen"
16:46:13kugelalso, I'm still wondering what I need to set in INT_GPIO (i.e. at every interrupt event) and what in the button init
16:47:12 Quit gromit` (Read error: 110 (Connection timed out))
16:47:19jhMikeSFor each GPIO with a button, set the detection type and unmask for each corresponding pin.
16:48:34jhMikeSFirst, read any buttons in init, clear any pending interrupts on the lines, then unmask the right pins.
16:50:17jhMikeSunmask approx. = allow interrupt to fire. otherwise the event is not triggered.
16:51:48 Join XavierGr [0] (n=xavier@rockbox/staff/XavierGr)
16:53:22kugeljhMikeS: that'd be _IE then
16:55:02kugeland that all in INT_GPIO
16:57:10 Quit Thundercloud (Remote closed the connection)
17:01:12kugeljhMikeS: I think it worked
17:01:50 Join pondlife1 [0] (
17:01:51kugelI just got the power button
17:03:01 Join J-23_ [0] (n=moarc@
17:03:14 Quit skipper (Remote closed the connection)
17:06:06 Quit J-23 (Remote closed the connection)
17:08:18 Nick J-23_ is now known as J-23 (n=moarc@
17:09:49 Quit mofux (Read error: 104 (Connection reset by peer))
17:11:03 Join toffe82_ [0] (
17:11:33kugelnow it locks upon boot when I add a single button on the other gpio port
17:11:43 Join obo [0] (n=obo@rockbox/developer/obo)
17:14:32 Quit Zagor ("Client exiting")
17:15:45 Quit toffe82 (Read error: 60 (Operation timed out))
17:15:47 Nick toffe82_ is now known as toffe82 (
17:19:47 Quit XavierGr ()
17:21:19*pondlife1 wonders what the "Album Art:No" line in Rockbox Info refers to...
17:22:49 Join XavierGr [0] (n=xavier@rockbox/staff/XavierGr)
17:23:07obothe size of album art in the current WPS, if it is used?
17:23:48linuxstbShouldn't it say "None" ?
17:24:01 Join Tristan [0] (
17:25:35pondlife1Why under Rockbox Info though?
17:25:50pondlife1It varies per track, I'd think
17:25:53 Quit Darksair ("To Arch or Gentoo? That is the question...")
17:26:02pondlife1Or at least per WPS
17:26:10linuxstbYes, it varies based on the WPS
17:26:22linuxstbI can't think of anywhere better for it.
17:26:52pondlife1Or checkwps?
17:26:55linuxstbIt's not a debug feature - it's so users know what size their art needs to be.
17:26:58 Join mofux [0] (
17:27:08 Join Darksair [0] (n=user@
17:27:44linuxstbMaybe we can just get rid of it when album-art resizing is committed though.
17:28:05 Join t0mas [0] (n=tomas@rockbox/developer/t0mas)
17:28:52kugelpondlife1: it's basically just reading the parameter of the %Cl tag
17:29:34kugellinuxstb: well, why? It'll still be interesting which size the wps expects (and resizes to)
17:29:53 Quit stoffel_ ("Reconnecting")
17:30:03 Join stoffel_ [0] (
17:30:05linuxstbIf you're interested, open the .wps. Users won't need to care about it any more.
17:30:40pondlife1That's how I feel about it anyway, resizing or not... ;)
17:31:21 Join herrwaldo [0] (
17:32:04kugellinuxstb: does it hurt binsize/ram/performance so much that it needs removing an existing feature?
17:32:27pondlife1Nah, not really - it's just a new thing to me
17:32:37pondlife1Seems pointless though IMHO
17:33:19pondlife1But we can all argue about features, so I won't
17:33:23linuxstbkugel: If Rockbox automatically resized album-art, I strongly doubt that info would have been added. Therefore, when Rockbox does resize, it makes sense (IMO) to remove it.
17:33:24kugelso. I now need to wait until the fuze runs out of battery again :( and it's fully charged
17:34:12pondlife1No way to pop the case open and cut power?
17:34:13kugellinuxstb: either way, I don't feel strong about it. I've hardly watch in the info screen (probably because I use resizing since ages :S )
17:34:16krazykit`kugel, or pop it open and take out the battery?
17:34:32kugelnoe, no way
17:34:49kugelnot without resoldering the battery cables to the board
17:35:00 Quit Darksair ("(define zero (lambda (f) (lambda (x) x)))")
17:35:01Unhelpfuli'm working on it. gray scaling needs a rewrite. :/
17:35:51 Part pondlife1
17:35:58 Join pondlife1 [0] (
17:36:19 Part pondlife1
17:36:47 Join looping_ [0] (
17:36:50 Join pondlife [50] (n=Steve@rockbox/developer/pondlife)
17:37:13 Part looping_
17:38:34kugeloh damnit
17:38:40kugelI wanted to work on the buttons :S
17:38:50kugelat least the display is on
17:39:04kugelso discharging maybe probably not take the full 25h
17:39:29 Quit XavierGr ()
17:39:31krazykit`kugel, sounds like a good time to wire in a switch :)
17:39:34Unhelpfulthere's no paperclip button, or a battery disconnect like on gigabeat?
17:40:18gevaertsNo power button to hold for a minute?
17:41:22kugelgevaerts: no :(
17:42:43kugelit might even be boosted
17:42:56kugelif I'm lucky I can continue developing tomorrow
17:43:08kugeltomorrow morning that is
17:43:16 Quit petur ("*plop*")
17:43:26kugelI had that once. But without display. I thought it was broken
17:44:09 Quit robin0800 (Read error: 110 (Connection timed out))
17:46:21 Quit herrwaldo (Remote closed the connection)
17:46:44 Join miepchen^schlaf [0] (
17:47:43 Quit Trista109 (Connection timed out)
17:50:40 Join Darksair [0] (n=user@
17:51:04 Join perrikwp [0] (i=18ac0c41@gateway/web/ajax/
17:52:17 Quit Crackerizer ("Leaving.")
17:52:44kugelhaha, this will be the first battery bench on the fuze
17:53:44 Join Jaykay [0] (
17:58:24 Join XavierGr [0] (n=xavier@rockbox/staff/XavierGr)
18:00:18Jaykaywhere can i post/write suggestions for the tracker cleanup week?or is nobody allowed to do this?^^
18:02:13*jhMikeS is just going to blow this USB charging option out for anything that doesn't have it implemented (only H300 does anything with it)
18:02:32amiconnjhMikeS: Btw, the atomic gpio bit manipulation also works for GPIOx_OUTPUT_EN and GPIOx_ENABLE, not only for GPIOx_OUTPUT_VAL (regarding e.g. r19214)
18:03:10kugeljhMikeS: my fuze is totally locked now, since I forgot to remove the #if 0 before the INT_GPIO function
18:03:11jhMikeSamiconn: I had asked you about that before but you said you didn't know at that time :)
18:03:22*kugel shouldn't have commented out the panicf
18:03:27amiconnThat must have been a looong time ago....
18:03:30LloreanJaykay: Did you know of some bugs that are fixed and should be closed?
18:03:45jhMikeSamiconn: right after you added the bitwise macros
18:05:00Jaykayno..... not yet. i know some patches that could be committed. and if i have too much time ill look also for alredy fixed bugs^^
18:05:15kugelI suppose my fuze is caught in the isr, which even disables the hardware power off
18:05:20Jaykaybut i dont know where to write this
18:05:26 Join m0f0x [0] (
18:05:33LloreanJaykay: "Patches that could be committed" are decisions that developers need to make.
18:05:46 Quit Darksair ("People who are zhuangbility want to show their niubility but only reflect their shability.")
18:05:51LloreanIf you think a patch should be committted *and* are willing to work on the flaws if a developer decides it needs fixing, bring it up in here or on the mailing list.
18:06:06amiconnjhMikeS: I used them this way e.g. in ata-sd-pp.c
18:06:28amiconnThey also work for GPIOx_INT_EN
18:07:26jhMikeSkugel: you are acking the interrupts correctly?
18:07:43 Join Darksair [0] (n=user@
18:08:15kugeljhMikeS: I forgot the remove an #if 0, so there's no INT_GPIO function
18:08:28kugeland since I commented out the panicf...
18:09:08kugelnot sure if you read it, but I think I got the power button to work with irq. I had problems on the other GPIO port though
18:09:53JaykayLlorean: so only the developers are allowed to help in the "tracker cleanup week", nobody is allowed to help them by suggesting some patches?
18:10:49*kugel decides to update his customlist patch before it gets closed during that week
18:11:02gevaertsJaykay: suggesting patches _is_ allowed. Only, if you suggest a patch, be prepared to work on improving it when asked
18:11:06 Quit Jaykay (Read error: 104 (Connection reset by peer))
18:11:30jhMikeSgevaerts: are USB ports required to supply at least a certain level of current?
18:11:38gevaertsjhMikeS: 100mA
18:11:53 Join Jaykay [0] (
18:12:13Jaykaysorry i was thrown out what did you write?
18:12:18gevaertsYou can ask for 500mA, but you're not guaranteed to get it
18:12:25gevaertsJaykay: there are logs on the website
18:12:52Jaykayi see, sorry
18:13:02LloreanJaykay: Also, I never said you're not allowed to do anything.
18:13:15LloreanI clarified when you should choose to help, so that you aren't wasting your own time.
18:13:15 Join MethoS- [0] (
18:13:18amiconngevaerts: Aren't even the 100mA not guaranteed?
18:14:01amiconnI mean, a port can be shut down by the host, and then the device must not draw more than a few mA (something around 2.5mA iirc)
18:14:07 Join MrDuck [0] (
18:14:10 Join herrwaldo [0] (
18:14:25JaykayLlorean: you said ""Patches that could be committed" are decisions that developers need to make.". it was a bit like "you are not allowed"
18:14:49 Join bluebrother [0] (n=dom@rockbox/staff/bluebrother)
18:15:00Jaykayanyway, where should i suggest these patches?
18:15:11*bluebrother wonders why the "restarting playback" splash is pointless ...
18:15:28bluebrotherit at least indicates that the restart was intentional and is expected
18:17:23LloreanJaykay: I already told you where discussion should happen, in the very next line I spoke after saying the one you just quoted...
18:17:30 Join faemir [0] (
18:17:46LloreanJaykay: It was said four minutes before you told me you thought I was saying you weren't allowed.
18:17:54gevaertsamiconn: that's when suspended IIRC, but I'm not too familiar with that part of the spec. If the port is powered off (also possible), you just get nothing at all (i.e. no 5V, so you can't even try to draw current)
18:20:05LloreanJaykay: Any patch that is currently in the tracker *almost* certainly needs some work.
18:20:19LloreanSo there's really very little point in trying to bring it forward unless you're willing to take responsibility for the work that needs done.
18:21:50Jaykaybut when i detect a bug thats already fixed i can simply write it her?
18:22:12LloreanBringing it up on the mailing list is the best way.
18:22:30LloreanIf you bring it up here, the people who can check on it may be asleep or away, so it's less likely to be acted on.
18:24:26 Join domonoky [0] (n=Domonoky@rockbox/developer/domonoky)
18:26:24 Join Seed [0] (
18:31:08 Join Nico_P [50] (n=nicolas@rockbox/developer/NicoP)
18:32:00 Quit Darksair ("(define (add-1 n) (lambda (f) (lambda (x) (f ((n f) x)))))")
18:33:20 Join ap0 [0] (
18:33:26 Quit obo (Read error: 145 (Connection timed out))
18:35:55 Part pondlife
18:40:41 Join bmbl [0] (n=Miranda@unaffiliated/bmbl)
18:40:48 Quit Jaykay ("ChatZilla 0.9.84 [Firefox 3.0.4/2008102920]")
18:45:12 Join obo_ [0] (n=obo@rockbox/developer/obo)
18:45:27 Join casainho [0] (
18:46:15***Saving seen data "./dancer.seen"
18:46:31casainhohello :-) −− I would like help in understand the minimum and maximum values that enters on "void tick_start(unsigned int interval_in_ms)"
18:46:58casainhowhat are the min and max of interval_in_ms ?
18:47:19casainhoon file kernel-_target_.c
18:48:22 Quit cg (Remote closed the connection)
18:48:25 Join bertrik [0] (
18:49:01LinusNcasainho: that depends on your target
18:49:14 Quit stoffel_ (Read error: 113 (No route to host))
18:49:16 Quit miepchen^schlaf ()
18:49:33 Join miepchen^schlaf [0] (
18:49:37casainhobut what values can people choose to use on that interval_in_ms?
18:50:24casainhoor should I try to use the max values of unsigned int? and by the way, is 16 bits or 32 bits?
18:50:54LinusNcasainho: all targets to this date use (1000/HZ), where HZ is 100
18:51:14LinusNwhich means 10ms
18:52:01LinusNso you might never have to support other values than 10ms
18:52:04casainhookok - so I think I will try to use 16 bits resolution
18:52:33casainhookok - thanks, it's clear for me now :-)
18:52:37LinusNas long as your tick timer is exactly 10ms, you can use any resolution you want
18:53:36casainhoand the same question but about this: bool __timer_set(long cycles, bool start)
18:53:46casainhocycles are 32 bits?
18:54:16 Join stoffel_ [0] (
19:00:48LinusNcasainho: yes
19:01:12 Join skipper [0] (
19:01:54*LinusN spots lots of code that should be moved to the target tree (firmware/timer.c)
19:02:10casainhoI don't understand what is for that timer_set fucntion, ca n you please explain and gine me one example?
19:04:08 Join BigBambi [0] (n=Alex@rockbox/staff/BigBambi)
19:04:20*kugel notices his customlist patch has some major flaws
19:04:49 Join karashata [0] (n=karashat@
19:06:21casainhoso, I think __timer_set configures one timer with cycles number... but I would like to know if cycles are multiple of any value, like 1ms...
19:07:14 Quit Horscht (Remote closed the connection)
19:09:15 Join Strife89 [0] (n=michael@
19:09:19 Join {phoenix} [0] (
19:10:07 Join Horscht [0] (n=Horscht@xbmc/user/horscht)
19:16:16LinusNcasainho: you are supposed to define TIMER_FREQ to the true frequency of the time base
19:16:43casainhoah, okok :-)
19:17:03casainhothanks :-) −− I will start writing the code for kernel_init();
19:17:04LinusNfor example, if your timer clock (after prescaling) is 24MHz, then set TIMER_FREQ to 24000000
19:18:16casainhoso, what should be the minimum TIMER_FREQ?
19:18:24casainhothe smallaer possible?
19:19:30casainhoor, the higher possible?
19:19:30LinusNthe higher the resolution, the better, i guess
19:19:42casainhookok :-)
19:20:02LinusNit all depends on how your timers work on your MCU
19:20:16casainhohow can I know If an did kernel did initialized correclty?
19:20:54domonokyif it doesnt crash, its atleast not completly wrong :-)
19:21:04LinusNsomething like that
19:22:17 Quit perrikwp (" ajax IRC Client")
19:23:06 Join perrikwp [0] (i=18ac0c41@gateway/web/ajax/
19:24:09casainholike, If kernel_init() returns, at least did not hang... but, after, can I test it with something, like an LED?
19:24:58domonokyblinking a led is a good test..
19:25:19domonokyjust place the code to blink the led, after the init in apps/main.c
19:25:38domonokyor into your bootloader, if you are working on that.
19:27:20LinusNcasainho: i often use digital i/o ports to output waveforms that i can watch using an oscilloscope or logic analyzer, perfect for checking the tick timer frequency
19:27:49casainhoI am working on the bootloader −− but, what parts of kernel can I use to blink the LED, the sleep(int ticks)??
19:28:09casainhoI have a serial debug port, it's working now :-)
19:28:17LinusNthat's prefect
19:28:24LinusNperfect, even :-)
19:28:38domonokysleep needs interrupts working, so it maybe better to use delay loops in the beginning.
19:29:01domonokyor if you have serial working, just output something to serial, after the kernel_init() ?
19:29:03casainhoI have the LED flashing using interrupts... but not the kernel
19:29:45casainhookok - I can see if kernel_init() returns using the serial port...
19:31:16 Join nuonguy [0] (
19:32:16 Quit casainho (Read error: 113 (No route to host))
19:34:29 Join casainho [0] (
19:34:47casainhosorry - my computer did hang
19:37:33casainhoI have aonther question about kernel: To have kerner_init() working, is needed to put timer-_target_.c working, like the __timer_register()??
19:37:40 Quit stu8ball (Read error: 60 (Operation timed out))
19:39:25LinusNcasainho: no, timer_register() is not necessary
19:39:36LinusNonly the tick timer
19:39:42casainhoah, so it's very simple then... :-)
19:39:49LinusNyup :-)
19:40:39LinusNgtg, good luck
19:40:43 Part LinusN
19:41:14 Join rotator [0] (n=e@rockbox/developer/rotator)
19:41:37 Join stu8ball [0] (
19:42:27casainhoso, what functions of this list can I use, to flash the LED using the kernel? what would be the simple cdoe?
19:43:59Bagderwhile(1) { sleep(HZ); toggleit() }
19:45:00domonokywhile(1){int delay=0x20000; while(delay > 0) delay−−; toggleLed() }
19:45:29Bagderright, but that doesn't really use the kernel
19:46:00domonoky:-) yes, that makes it more reliable :-)
19:46:23casainhook Bagder, I think that is really easy and I understand that I will use the "sleep();" :-)
19:46:46casainhoI will start coding now ;-)
19:49:29 Nick __lifeless is now known as _lifeless (n=lifeless@
19:50:48kugeldomonoky: lol, how fast do you want to toggle the led?
19:51:20domonokykugel: depends on the cpu freq :-)
19:53:12casainhothanks anyway Domonoky ;-)
19:53:47*domonoky goes to hear music on his m200v4 :-)
19:56:10*bertrik apparently uses a different test wav than funman, because my wav sounds like it's sung by chipmunks
19:56:55domonokymine sounds ok. but i use the svn bootloader.
19:58:18bertriksurely the bootloader should not influence playback speed
19:59:30domonokybertrik: i think we set the freq only in the bootloader (see system-as3525.c)
19:59:39 Join ringo999 [0] (
19:59:58 Quit stoffel_ (Read error: 113 (No route to host))
20:00:06ringo999is it normal that the usergroup of the files on my ipod is bound to be root?
20:00:30Lloreanringo999: That really doesn't have anything to do with Rockbox.
20:00:52domonokybertrik: i mean the plla frequency, and m_clock for i2s_out is divided from that pll freq.
20:01:13domonokyringo999: fat32 doesnt know any usergroup attribute...
20:02:01ringo999domonoky: mmm...i was trying to use rsync and its complaining because of the group. but there is probably some rsync parameter that neglects that...
20:02:27 Quit J-23 ("leaving")
20:02:46*domonoky tries to move ringo999 to #rockbox-community
20:05:31 Quit obo_ (Read error: 60 (Operation timed out))
20:07:06*bluebrother has small caps working.
20:07:13bluebrotherPalatino is the one to blame ...
20:07:56amiconnIsn't Palatino a font?
20:08:26bluebrotherthe manual specifies Palatino to be used. But it seems this one is lacking "real" small caps, so LaTeX fakes them by using different font sizes
20:10:38 Join J-23 [0] (
20:11:14gevaertsSo we can expect a new look for the manual?
20:11:58*bluebrother restores palatino to recheck
20:14:17bluebrotheryep. Just remove palatino and it works. Let's see if that also fixes the issue with "heavy html" ...
20:14:41 Quit casainho ("ChatZilla 0.9.84 [Firefox 3.0.4/2008111318]")
20:17:20bluebrotherhmm, interestingly it doesn't fix that
20:20:59 Part nicfer
20:22:00 Join kronflux [0] (
20:23:01kronfluxLambdaCalculus37: any recent breakthroughs with the nano2g?
20:23:02kronfluxAnyone else: any further progress with the Sandisk Sansa Fuze?
20:23:24bluebrotherkronflux: the breakthrough is: still nobody working on it
20:24:00Lloreankronflux: Please, just read the form threads and wiki pages for progress.
20:24:00bluebrotheror was there something special going on I missed?
20:24:22LloreanUnless you're planning to work on it yourself, you can't do anything until it works, and when it does work it'll most certainly be announced.
20:25:47kronfluxI have read threads. LambdaCalculus37 has been working on it, so just because theres no progress announced, does not mean there is none. also, I have worked on it as well.
20:26:21LloreanNo progress has been announced just means no progress has been announced.
20:26:30LloreanBut we ask people not to come in here looking for status updates.
20:27:02 Quit stu8ball (Read error: 60 (Operation timed out))
20:27:33kronfluxwell, as I said, I've been working on it too. so the question for the nano2g was directed at a fellow developer and the person I've worked with directly. so it wasnt a status update. it was a question of developer interest. if there was progress, or breakthroughs, its more likely I'd be able to do more.
20:28:44bluebrotherdevelopment discussion happens in this channel, at least usually ...
20:28:44CtcpPing from gevaerts!n=fg@rockbox/developer/gevaerts
20:29:05Lloreankronflux: You asked for status updates about the fuze.
20:29:25 Quit Zarggg (Excess Flood)
20:29:44CtcpPing from gevaerts!n=fg@rockbox/developer/gevaerts
20:29:52 Join Zarggg [0] (
20:30:45 Join stu8ball [0] (
20:31:16 Join dberg918 [0] (
20:31:44kronfluxyes, yes I did. and it was mere interest. am I not allowed to be curious as well as have developer interest? the question still applies to development, does it not? and once again, if there was further progress with the fuze, its more likely I could get more involved. obviously not everything is said in the forums and wiki pages. this is a channel related to development. so.. I'm not sure what your problem is with what I'm asking.
20:32:07Lloreankronflux: As I said before, we ask people not to come in here and just ask for status updates.
20:32:21LloreanIf you're involved with the development, sure, ask questions about what you're working on.
20:32:32LloreanBut if you're just looking to be a tester or something, when they need testers they'll say so.
20:33:27kronfluxand as I said, its not just a status update. I'm not even involved in the development of the fuze. but I'd like to be. asking questions leads to discussions which leads to involvement. I dont want to be some beta tester.
20:33:35dberg918Is anyone else having trouble with lcd brightness?
20:33:41*domonoky thinks the sansa fuze should have sound now, but the buttons dont work, sou can start the playback :-)
20:34:00domonokys/sou can/ so you can NOT/
20:34:19bluebrotheryou can't? Just because of some stupid buttons? ;-)
20:34:31dberg918On my gigabeat f40, when I try to change the setting, it automatically reverts to some kind of default brightness after I hit a key
20:34:34gevaertsYou can. You just have to build your own playlist control file and edit config.cfg a bit
20:34:55Lloreankronflux: Then ask more specific questions than "Any progress?" because that's clearly a request for a status update.
20:35:05bertrikor work harder on the button code :P
20:35:19*domonoky doesnt have a fuze...
20:35:20Lloreankronflux: For example, you could _read the logs_ then ask questions about the most recently discussed progress.
20:36:06Lloreankronflux: As the channel guidelines say, we expect you to search first, so showing some knowledge of the most recently discussed public status would go a long way to making it clear you were trying to get involved, rather than being one of the people who just shows up from time to time asking if we're working on a player they're interested in.
20:36:28LambdaCalculus37kronflux: No progress yet with the nano2g. I'm having a hard time figuring out the encryption and not enough free time to work on it.
20:37:23Lloreanbertrik: Which AMS Sansas are closest to "usable" these days?
20:37:25dberg918I know that software fading has recently been put in SVN (if I remember correctly), is it possible that this has somehow affected the lcd brightness setting?
20:38:15kronfluxLlorean: you never give up, do you? :p I really don't care what you say. I'm not going to argue with you like a couple of 8 year olds. I asked a queston. you didnt like the question. you dont like it? ignore it. dont answer it. it wasnt directed at just you, otherwise, I would have said your name first. I dont always have time to skim through all the logs. a lot of days go by between visits to the channel. I'm not new here. you can assume
20:38:15kronfluxI just want a status update all you want. but what I ask is strictly developer interest. too bad if my wording isnt to your liking. as I said, if I want your opinion, I'll ask.
20:38:31*domonoky thinks clip and m200v4 are nearly in the same state (although clip has already pluggin button maps)
20:38:50Lloreankronflux: Or, I kick you from the channel and ban you for ignoring channel guidelines. So how about you say "I'll search first" now, instead of antagonizing me further?
20:39:08bertrikLlorean, I agree with what domonoky said
20:39:45Lloreanbertrik, domonoky: Interesting. I have an m200v4 and in theory a clip if it ever ships. Are the button maps based on similar targets?
20:39:50kronfluxLlorean: and once again be acting like an 8 year old. I'm not new here. I know the guidelines. if you take what I say the wrong way and assume things rather than asking, thats your perogative.
20:39:59bertrikclip is mostly based on c200
20:40:37Lloreankronflux: If you're not new here, follow the guidelines. If you insult me again, I'll "act like an 8 year old." Now, it's a yes/no question: Will you attempt to read the logs first in the future, rather than asking for general progress updates?
20:40:41domonokyLlorean: m200v4 button map isnt really ready, i just filled up as much as was needed to get it useable..
20:40:59gevaertsdberg918: you could try to isolate this a bit more by trying older daily builds
20:40:59CtcpPing from gevaerts!n=fg@rockbox/developer/gevaerts
20:41:05Lloreandomonoky: Is it complete for core, or still even needs work there?
20:41:51kronfluxLlorean: I usually do. and if you'd actually read what I said to you, you'd know I'm not asking for general progress updates. if I wanted general progress updates, I'd say "hey! is it done yet?:D:D:D:D"
20:42:08kronfluxLlorean: as I said, I dont always have the time to read the logs. but I usually do.
20:42:15domonokyLlorean: its ok for the main screens, ie filebrowser and wps. Didnt really care about the other screens. Would be nice if you could help to improve them :-)
20:42:17Lloreankronflux: "any further progress with the Sandisk Sansa Fuze" sounds like a request for general progress.
20:42:42kronfluxLlorean: as I said, thats not how I meant it. and look. it had the effect I was looking for. it sprung a conversation about it.
20:42:50Lloreandomonoky: Since I have a bit of a thing about keymaps, I probably will later this week.
20:42:59dberg918gevaerts: you wouldn't happen to remember which revision introduced the software lcd fading, would you?
20:43:17Lloreankronflux: I'm being absolutely serious here. In the future, ask specific questions. Don't ask general "How is progress on Player X going?"
20:43:27*domonoky offers to help Llorean with any problems he has with getting the m200v4 up and running.. :-)
20:43:32dberg918...wait, I think the date got recorded in the MajorChanges wikipage
20:44:12Lloreandomonoky: Is it just a matter of using the ams bootloader merging tool on an OF and copying it over? (I very briefly skimmed the wiki page but only glanced at the beginning of the install instructions)
20:44:13domonokykronflux: at leats take a look into the forum thread, before asking for status...
20:44:23kronfluxLlorean: and I'm being absolutely serious too. if I really wanted to know if it was "done" or "working yet" or whatever, I'd say "hey! is rockbox available for player x yet?!
20:44:27bertrikLlorean, I would really like have someone take a good look at the keymaps for new targets, rather than just blindly copy other targets
20:44:31domonokyLlorean: yes.
20:44:40kronfluxeither way, i'm out.
20:44:49Lloreankronflux: You asked if there was any progress. I'm telling you not to ask for _progress_ reports, not "completion status"
20:44:52LloreanAsk specific questions.
20:45:01Llorean"Did you get buttons working?" "Did you decrypt the flash image" etc
20:45:05CtcpPing from gevaerts!n=fg@rockbox/developer/gevaerts
20:45:44LambdaCalculus37kronflux: I'm about the only guy that was doing anything related to the nano2g. I've also got a lot of "real life" stuff to worry about. I haven't decrypted the flash image, nor have I even written any specific code yet.
20:46:17***Saving seen data "./dancer.seen"
20:46:24 Quit Acksaw (Connection timed out)
20:46:28kronfluxyeah, me too. which is why I'm saying I dont always have the time to read the logs. I have a job and a life too.
20:46:30 Quit lasser ("ChatZilla 0.9.84 [Iceweasel 3.0.3/2008092816]")
20:46:31bluebrotherkronflux: well, it created a lot of unnecessary (and distracting) noise. Seems like this is what you wanted?
20:46:37Lloreanbertrik: Well, I'll try to set aside some time later this week and have a look at it then for the m200v4, and then clip whenever I have one physically to play with.
20:46:57Lloreankronflux: Make time to read the log before you ask other people to make time to explain the status to you.
20:47:14bluebrotheras in "it sprung a conversation" ...
20:47:32CtcpPing from gevaerts!n=fg@rockbox/developer/gevaerts
20:47:35CtcpPing from gevaerts!n=fg@rockbox/developer/gevaerts
20:47:47kronfluxbluebrother: not quite. actually I'm leaving to prevent any more noise. I just dont appreciate being treated differently because of the way I word things. yes, I worded it wrong. I didnt mean to ask for a progress report. but I'll just shut up now to save any further dispute as to what I said.
20:47:53 Quit kronflux ("Leaving")
20:48:04domonokyLlorean: a problem with the m200v4 is, that it doesnt have much buttons. And we probably have to "miss-use" the volume-keys in some screens.
20:48:17Lloreandomonoky: I'm perfectly happy with misusing volume keys if we must.
20:48:33LloreanI really like the idea of universal volume control, but I'd rather "have all the features normally accessible available"
20:48:33bluebrotherdomonoky: can I just build bootloader and Rockbox for m200ams from svn and it runs? Or are there other adjustments necessary?
20:48:36 Quit BigBambi (Read error: 104 (Connection reset by peer))
20:48:45bluebrotherI guess I need to mkamsboot it after building?
20:49:18dberg918I'm checking out r19226, this is the build before backlight fading got split into two separate features
20:49:25domonokyyes, build bootloader. build mkamsboot. merge them. build a main build and copy all to the dap.
20:50:25LloreanIt looks like the only difference in physical buttons is the positioning of the "Power" button.
20:50:35domonokyto get sound apply the patch in svn after making the bootloader (you should use a svn bootloader), and build the main-build again.
20:50:36LloreanAny reason it couldn't just directly share the c200 keymap (like the H100/h300 share one?)
20:50:50bertrikbluebrother, FS #9592 is the latest and greatest patch to apply for dma sound playback
20:51:53*domonoky checks the c200 keymap..
20:52:23 Quit nplus (Remote closed the connection)
20:52:25bluebrotherdomonoky: I need bootloader.bin for mkamsboot?
20:52:30bertrikthe clip doesn't have a REC button, I don't know about the m200
20:52:48domonokybluebrother: bootloader-xxx.sansa
20:53:00bluebrothersorry, just figured it myself :)
20:53:18LloreanAh, right, recording.
20:53:31LambdaCalculus37bertrik: The m200 has a Play/Pause, Back, Forward, Repeat/AB, Vol Up/Down, and Power/Menu buttons.
20:53:49LloreanWell, the recording button isn't used for core Rockbox functions. Could we work around that fact to reuse the keymap, conditionally removing that bit?
20:54:10LloreanI mean, we're not going to assign a button to its functions anyway, I should think.
20:54:29Lloreanbertrik: Yeah, no recording button on the clip either. I forgot the c200 had it.
20:55:24 Join lasser [0] (
20:55:30domonokyhm.. maybe just copy the keymap and replace the button_names ?
20:56:14 Quit johnm (Remote closed the connection)
20:56:15LloreanWhy are the button names different?
20:56:18*Llorean hasn't looked yet.
20:56:30domonokyor rename the buttons in the button-target.h ..
20:56:35LloreanPhysically, they're the same buttons except the missing record button?
20:56:36bertrikwhat button names? VOLUP vs. VOL_UP?
20:56:52domonokyup/down is playpause/repeatab
20:57:21 Quit pixelma2 ("-")
20:57:27domonokyand power is menu. the different writing for VOL should probably be changed :-)
20:57:33 Join pixelma [50] (i=pixelma@rockbox/staff/pixelma)
20:57:41linuxstbHas pixelma's patch for the c200 keymap been committed?
20:57:50*domonoky checks how the c200 looks like.
20:57:52Lloreanlinuxstb: Not yet.
20:57:58*linuxstb recalls hating the c200 keymap when he first got a c200...
20:58:18Lloreandomonoky: It looks very much like the m200, but with the power button on the face instead of the side.
20:58:32LloreanIn terms of positions, at least
20:58:41LloreanHow they're constructed is very different.
20:59:23domonokyyes, from the position they are very similar. So we should probably copy it.
20:59:50*LambdaCalculus37 thinks that's a good idea
20:59:51pixelmaI'd like the parts I'm most happy with soonish - resume/stop in tree on the volume keys, virtual keyboard (with a slight modification though) and WPS context menu on long select (the latter is perhaps better for Llorean to commit, he's used to taking the blame ;) )
21:00:41*Llorean thinks those changes sound good.
21:00:45*pixelma adds a "commit" to her sentence above
21:00:45*domonoky thinks context menus should always be on long select... its like right-click
21:00:55Lloreandomonoky: That's what I said, oh so long ago. :)
21:01:50*bertrik agrees
21:02:29*domonoky thinks what would be better, copy the c200 keymap, or have m200 targets which use SANSA_C200_PAD ?
21:02:35pixelmaI don't have much time this week though
21:03:07Lloreandomonoky: If it can be done without the missing record button being a problem, I'd think having it use the SANSA_C200_PAD might be better
21:03:19LloreanThen any future keymap changes (adjustments to plugins, etc) would be simpler.
21:04:08*bertrik is still surprised that often about half of the source code lines in plugins are dedicated to keymaps
21:04:16domonokyLlorean: the rec key could be worked around with a simple #ifdef at least for actions. Plugins migth be more difficult.
21:04:39LloreanPlugins are always more difficult. I'm most concerned about Core anyway.
21:04:54dberg918okay, r19226 has the same issue, now I'm trying r19220, which is before software fading was implemented for certain targets
21:05:29pixelmabertrik: yes, but changed that turned out to not be really usable
21:05:43domonokywe could even place a #define HAVE_REC_BUTTON (or HAS ? ) into the config-<target>.h
21:06:21 Quit bluebrother (Nick collision from services.)
21:06:26 Join bluebrother [0] (n=dom@rockbox/staff/bluebrother)
21:06:46domonokyLlorean: yes, but plugins might get more difficult if a check for XXX_PAD isnt enough anymore..
21:06:53dberg918there's another minor issue that seems to go hand in hand with this brightness thing
21:07:53dberg918when I switch themes, the line selector and the status bar stay as they are after the theme is switched, until I get to the root menu
21:07:58dberg918then everything is fine
21:08:33Lloreandberg918: If you can track down which revisions started each problem, please file bug reports (separate ones, for each problem)
21:08:57dberg918no prob :)
21:09:46bluebrotherdomonoky: did I miss something special about the firmware upgrade? Wasn't it just dropping the bin file and disconnecting?
21:10:18domonokybluebrother: it has the correct filename ? droping and disconnecting should be enough.
21:10:27pixelmadomonoky: my first reaction (not well thought through) was that it could be more flexible to copy the c200 keymap because in the case that the different placement of the power button and the missing rec button make a real difference to the usability of the controls you can do more about it
21:10:48bluebrotherm200p.bin, or m200p-e.bin.
21:11:01bluebrothertired both, neither worked for me
21:11:13pixelmaand have an own M200_PAD, I mean
21:11:47*domonoky always uses m200p.bin. But have had this problems already earlier, if i remember correctly ?
21:12:02bluebrotheryep, I had this issue before.
21:12:27domonokyyou used the same OF version as worked before ?
21:12:44bluebrothercan't remember which one that was :(
21:12:45bertrikthe clip begins upgrading the firmware almost immediately after unplugging, then powers off when the upgrade is finishd
21:12:48 Quit Strife89 ("Got work to do....")
21:13:03domonokysame behaviour as on m200
21:13:16domonokybut it may just shutoff, if you have hold enabled.
21:13:51domonokybut the upgrade on the next boot. It also refuses to flash, if the battery is too empty.
21:14:00bertriknice, ogg playback on the clip is stable enough to finish an entire song :)
21:16:38dberg918looks like r19220 doesn't have the brightness issue, but it still has the status bar problem, I'll check r19221 in a little while to make sure it's software fading causing the problem
21:17:04domonokybluebrother: which version do you have now ? (i use 4.1.08 at moment).
21:19:03 Quit miepchen^schlaf ()
21:19:03 Quit kugel (Read error: 104 (Connection reset by peer))
21:20:00 Quit nuonguy ("This computer has gone to sleep")
21:20:38 Join kugel [0] (n=chatzill@unaffiliated/kugel)
21:21:04*LambdaCalculus37 is having some troubles rolling a new build right now... can someone roll him a Clip bootloader and build with FS #9592 applied?
21:21:06 Join jhulst [0] (n=jhulst@unaffiliated/jhulst)
21:21:22 Join miepchen^schlaf [0] (
21:21:56 Join nuonguy [0] (
21:22:25 Quit kugel (Remote closed the connection)
21:24:18 Quit nuonguy (Client Quit)
21:25:25 Join petur [50] (n=petur@rockbox/developer/petur)
21:25:27bluebrotherdomonoky: 4.01.08-A
21:25:42 Join n1s [0] (n=nils@rockbox/developer/n1s)
21:26:37 Join tyfoo2 [0] (
21:26:40 Join aarcane [0] (
21:26:56 Quit beta2k_ (Read error: 110 (Connection timed out))
21:27:23domonokybluebrother: and if you boot into the OF and the insert USB the file is still there ?
21:28:16 Join pondlife [50] (n=Steve@rockbox/developer/pondlife)
21:28:27 Quit pondlife ("Leaving.")
21:28:44 Join Acksaw [0] (
21:30:10bertrikLambdaCalculus37, I can make one for you, how do you want it? a m300.bin and a rockbox.tar.gz?
21:30:22bluebrothereven tried an unmodified version from Bagders site.
21:32:00domonokydumb question: you are copying it in MSC mode ? :)
21:33:23bluebrotheryes ... but lets check if that reset itself
21:33:34bertrikLambdaCalculus37, clip bootloader is here:
21:34:07bertrikclip rockbox is here:
21:34:59 Join ibseco [0] (
21:35:02LambdaCalculus37bertrik: Thanks. :)
21:36:07 Part franky
21:36:49 Join mofux_ [0] (
21:37:28 Quit mofux (Read error: 60 (Operation timed out))
21:38:38 Join funman [0] (n=5a32d034@gateway/web/cgi-irc/
21:38:45 Quit stu8ball (Read error: 104 (Connection reset by peer))
21:38:49 Join stu8ball [0] (
21:39:26domonokybluebrother: or maybe try if the official sansa updater works..
21:40:55 Quit tyfoo (No route to host)
21:41:17funmandid you try "m200.bin" ?
21:42:40funmani still have crashes on the clip, a few times I see a data abort in buffering.c ; I added some debug to UIE() but now I'm waiting for something to happen instead of just see sound and display "freeze"
21:44:30bluebrotherm200.bin didn't work either. Will give up for today.
21:45:49n1sbluebrother: great news on the small caps!
21:45:58pixelmafunman: did you find me lengthy comment in the logs?
21:46:37 Quit {phoenix} (Remote closed the connection)
21:46:45LambdaCalculus37bertrik: The file for the Rockbox build doesn't open for extraction; 7-zip is saying the tar is "broken".
21:47:03funmanpixelma: no sorry, at about which time ?
21:47:36n1shmm, the mrobe 100 is not even in MAINTAINERS, who was it that ported rockbox to that player?
21:47:46LambdaCalculus37lowlight did IIRC.
21:47:59n1sLambdaCalculus37: ah, thanks
21:48:04pixelmaand roolku did a lot there too
21:48:12pixelmafunman 14:17
21:48:14bertrikLambdaCalculus37, it opens fine here, maybe something went wrong during the upload, better be careful with that bootloader too.
21:48:36n1sIt's a bit troubling having "supported" targets which no active devs have...
21:48:38LambdaCalculus37bertrik: Can you email them to me?
21:50:24bertrikLambdaCalculus37, I can try
21:50:31 Join shotofadds [0] (n=rob@rockbox/developer/shotofadds)
21:50:42LambdaCalculus37bertrik: PM
21:51:43n1sdoes the mrobe have any kind of disk mode?
21:51:56funmanpixelma: so perhaps just a font change in the wps can show nicer on the Clip?
21:52:23funmanfor the Clip simulator I'll have a look when my laptop's power supply work again :'(
21:53:13LambdaCalculus37n1s: Not sure; I don't have one.
21:53:42LambdaCalculus37n1s: domonoky and gevaerts both have mrobes.
21:53:56 Quit Horscht ("Snak 5.3.3 Unregistered copy. Evaluation period is over. Program will now quit. Thanks for using Snak.")
21:55:18domonokyn1s: what do you mean with disk-mode ? it reboots into the OF for USB like all pp-players.. :-)
21:55:46n1sdomonoky: so if you inser usb while it's off it goes into the OF?
21:56:09*n1s owns 0 pp players
21:56:29pixelmafunman: it might but currently you only have one user font for menus and WPS so it would change the menus too (everything else than an 8-* font won't fit so nicely in the upper part and is probably a waste as you will have less lines overall)
21:56:49domonokyn1s: jup
21:57:05*domonoky tests...
21:57:13pixelman1s: but got one borrowed, I heard
21:57:19domonokyah, no it goes into a OF "usb-mode" ..
21:57:56n1spixelma: yes, that one goes into OF when usb is connected while it's of and if usb is connected while in rockbox it freezes
21:58:31funmanpixelma hm the default is 8 pixels now I thinks : so the problem is the 6 pixels high progress bar
21:58:38pixelman1s: the latter depends on your build, it works as it should with some builds
21:59:09 Join Horscht [0] (n=Horscht@xbmc/user/horscht)
21:59:30n1spixelma: it never worked for me, i have tried at least 10 different builds (accidentallY)
21:59:48pixelmafunman: yes, but pushing it down won't look nice on the Iriver remote
22:00:19funmanhm this is weird I just got an undefined instruction in (likely) plasma plugin
22:00:25n1spixelma: btw i really like the change to having context-menu on long select :)
22:01:19pixelmait's not there yet... need to borrow Llorean's flame suit :)
22:01:53n1sdomonoky: can you take a look at FS #9597, i just saw he can't turn the player off apparently
22:02:17funmanpixelma: right, so perhaps we would need 2 different wps
22:02:20*LambdaCalculus37 plays his first MP3 on his Sansa Clip!
22:02:37funmanlambda: great, you're a man now :)
22:03:29domonokyn1s: will do. There is a reset hole on mrobes..
22:03:36LambdaCalculus37For the record, the first song I picked was Peter Howell's arrangement of the Doctor Who theme (from the 4th Doctor to the 6th Doctor).
22:04:27*LambdaCalculus37 now plays his second song... he's clearing enjoying this too much :P
22:05:35LambdaCalculus37funman: Sound playback works well enough, but the WPS keymapping on the Clip needs to be cleaned up. I can't stop playback, and it seems to "stick" for a few seconds after a file's done playing.
22:05:43 Quit Horscht ("I got raided by the FBI and all i got is this lousy quit message")
22:06:06funmanlamba: i don't know muck keymaps .. bertrik started giving a go to Clip keymap
22:06:42funmanperhaps "sticking" is due to < 1 song buffer size
22:07:10LloreanThe Clip has the same buttons as the m200, right?
22:07:19bertrikI just fixed a couple of obvious keymap problems
22:07:21LambdaCalculus37Llorean: Yes.
22:07:44LambdaCalculus37Although the Power button is a slider switch with a "lock" position for hold.
22:07:57LloreanProbably a good candidate for another copy of the c200-minus-recording keymap
22:08:02LambdaCalculus37I think the c200 keymap would work on the Clip as well.
22:08:22funmanthat's what I used for plugins, but for the core I copied e200 keymap
22:08:45Lloreanc200 probably makes more sense, since there's no wheel.
22:09:22pixelmafunman: yeah, and a clever way to make the build system distinguish them to put the right thing into the zip. Currently it uses the screen's resolution which is part of the filename - that's the same for those, have a look into the "wps" folder. There already is a way to let the system put a different WPS into the package for the H300 and the H100 which can use the same remote but due to the different size of the main screen's can possibly use a d
22:09:23LambdaCalculus37Llorean: I'm finding it a little weird because on the c200, I'm used to a long play press to stop audio playback.
22:09:41LloreanLambdaCalculus37: With pixelma's changes, that'll be one of the volume buttons in the future.
22:09:49LloreanOh, wait, not in the WPS though
22:10:18pixelmaonly in addition to stop in the browsers or WPS
22:10:36pixelmaerrr... s/WPS/menu
22:10:50funmandomonoky: can you play mp3 now ? (with 24MHz pclk)
22:11:18 Join BigBambi [0] (n=Alex@rockbox/staff/BigBambi)
22:12:12funmanwell i'm stupid : wav files would require much more SD work than mp3 and they played fine ..
22:13:42domonokymp3 doesnt really work, the decoder is too slow, or not working correctly. i get small parts of sound, with looong pauses inbetween.
22:14:06LloreanHow fast are you running these players at right now?
22:14:12LambdaCalculus37domonoky: On your m200v4?
22:14:13domonokybut i havent done much experimentation since i got the sound wokring again.
22:14:34domonokyLambdaCalculus37: yes, but with very expermiental pcm code.
22:14:46domonokywav works pretty good.
22:14:57funman31 MHz / 248MHz (there is a 83MHz freq but i'm not sure how it's used)
22:15:09LambdaCalculus37MP3s are playing well here on my Clip.
22:15:25funmanlambda : but the peripheral clock is higher than on m200
22:15:40LambdaCalculus37funman: Ahh, that explains it.
22:15:53funmani'm not sure how that changes anything but SD transfers however
22:17:34*domonoky should do some experimentation on this frequency thing on m200v4. But not today anymore.
22:18:47funmanwell pclk also feeds the memory controller
22:18:53*LambdaCalculus37 just got a horrible squealing sound out of his Clip
22:20:05funmananother data abort in buffering.c (in linked lists access)
22:20:16LambdaCalculus37funman: Yep. How'd you guess? ;)
22:20:53bertrikthe ams sansa codec communication currently expects a PCLK of 24 MHz, so it's a bit overclocked right now
22:21:25funmanlamba: ah, for you too? ><
22:21:36LambdaCalculus37funman: It seems to be happening with one specific MP3 file, though.
22:21:51funmanbertrik, isn't there a pclk clock divider for the codec clock ?
22:21:55 Join Rondom [0] (
22:22:18funmanoh .. I see what you mean, the divider must be adjusted
22:22:46funmanperhaps we should use defines for pclk/plla so we don't forget a peripheral when we change them
22:23:13bertrikfunman, yeah something like that
22:23:56bertrikfunman, what clocks does the OF use? and how does that compare to the clocking recommendations on page 112 of the datasheet?
22:24:26funmanbertrik various freqs I think, I'll have a look again
22:24:55funmanthere is a table of different PLLA settings, but I don't remember if they use all of them
22:25:24 Quit karashata ("G'bye everyone!")
22:26:50 Join {phoenix} [0] (
22:27:00 Join Horscht [0] (n=Horscht@xbmc/user/horscht)
22:28:56*LambdaCalculus37 will test his Clip out a little more on the bus ride home
22:29:38 Quit LambdaCalculus37 (" ajax IRC Client")
22:30:50 Quit faemir (Read error: 110 (Connection timed out))
22:31:47 Join faemir [0] (
22:32:15 Quit tyfoo2 ("Carpe diem")
22:41:30 Quit funman ("CGI:IRC")
22:44:42 Join safetydan [0] (n=deverton@rockbox/developer/safetydan)
22:46:18***Saving seen data "./dancer.seen"
22:52:17 Join kugel [0] (n=chatzill@unaffiliated/kugel)
22:54:56 Quit {phoenix} (Remote closed the connection)
22:55:01 Join Thundercloud [0] (
22:59:31 Join Zagor [242] (
23:00:16 Quit Thundercloud (Remote closed the connection)
23:01:08 Quit Rondom ("Ex-Chat")
23:01:09 Join Thundercloud [0] (
23:02:36 Quit XavierGr (Read error: 104 (Connection reset by peer))
23:04:47 Quit bertrik ("Leaving")
23:05:00 Quit Bensawsome (Read error: 110 (Connection timed out))
23:09:15kugelok. I have rewritten my customlist patch. Now I need to find a way to actually update the list upon settings change
23:09:19 Quit skipper (Connection timed out)
23:10:43 Quit leachim6 (Read error: 104 (Connection reset by peer))
23:11:25 Quit ap0 (Read error: 104 (Connection reset by peer))
23:12:49 Join XavierGr [0] (n=xavier@rockbox/staff/XavierGr)
23:13:11 Join skipper [0] (
23:14:43 Quit domonoky (Read error: 104 (Connection reset by peer))
23:16:55 Join grndslm [0] (
23:18:04 Quit t0mas ("good night!")
23:19:41 Quit Schmogel (Read error: 104 (Connection reset by peer))
23:20:19 Join Schmogel [0] (
23:21:36 Join DerDome [0] (
23:23:02 Quit skipper (Read error: 60 (Operation timed out))
23:26:33 Quit ibseco (Read error: 113 (No route to host))
23:32:30 Quit Schmogel (Read error: 104 (Connection reset by peer))
23:33:29*Bagder mentions tracker cleanup week!
23:33:54*Zagor works on a rbutilqt bug!
23:34:00Bagdermaybe we should offer a prize for most closed entries during the week!
23:34:08BigBambiI reckon I can win that
23:34:16Bagderfixing is overrated, closing is what we're after ;-)
23:34:18BigBambiOr do you want poper easons and stuff too?
23:34:35BigBambiand *proper
23:35:06*bluebrother spots Zagor working on rbutil
23:35:41Zagorbluebrother: yeah I thought I should ping you. I'll post a patch for the "out of space"=="unzip error" bug soon.
23:36:14bluebrotherah. I was looking into a figure-free-disk-space solution a couple of days ago. Slightly different but still related ...
23:36:48 Join LambdaCalculus37 [0] (
23:36:51bluebrotheras usual, every OS uses a different syscall :(
23:37:42Zagormuh, silly me was only thinking of posix... :-(
23:37:52Zagordo we really have to support windows? is anybody using that?
23:38:05bluebrothernah, lets scrap support for it ;-)
23:38:05BagderI saw a page on the internet that said so
23:38:20Bagderthis internet thing seems to be really cool
23:38:33bluebrotherBagder: you shouldn't look at about that ;-)
23:38:41 Join skipper [0] (
23:39:01zei thought nt-based windows had a posix layer
23:41:49Zagorze: generally, anything posix-like microsoft produces is either hopefully outdated or doesn't work
23:42:03zeZagor: ah, guess thats no surprise
23:42:13Zagorit's almost like they don't want people to use it
23:43:32LloreanZlib itself isn't capable of determining why extracting failed?
23:43:49bluebrothertalking about tracker cleanup: someone want to do something about FS #9554?
23:45:50 Quit Thundercloud (Remote closed the connection)
23:47:39pixelmabluebrother: amiconn preferred the way it is on M3 and rather wanted to "port" its keymap to X5, M5
23:48:09pixelmaand wasn't there already another patch like that?
23:48:18bluebrotherwell, I don't care too much about it but IMO we should either reject that patch or commit it.
23:48:51bluebrotherI believe that was this patch ;-) I just stumbled across it again
23:49:14pixelmayeah, a search doesn't turn up a second one
23:51:28amiconnImo that fs entry should be rejected. The way it is mapped now might differ from the OF, but it makes mapping consistent between main and remote
23:53:24amiconnThat is, Vol+ moves up, Vol- moves down on main & remote. Volume has rocker switches on both, and the on on the main unit is aligned vertically
23:54:38amiconnThe M3 vs. M5/X5 remote mapping differences are another matter. Iirc mainly Vol+/Vol- and Rew/Ffwd are switched on X5/M5 vs. M3 remote
23:55:39amiconnThe way it is on M3 makes it consistent with the main M3 buttons, and (at least to me) feels more natural. It's actually debatable, as both are horizontal rocker switches on the remotes (one at the top, the other at the bottom)
23:55:59kugelswitch-case is faster than if-else if-else if-...-else if there's many possible branches, correct?
23:57:08n1skugel: it should be, the compiler might optimize if-else into a jump table anyway though
23:57:27kugelmight, yea
23:57:40amiconnOn X5/M5, consistency between main and remote isn't an issue, because the keypads are vastly different. But I'd want the X5/M5 remote mappings to be consistent with the M3 remote mappings, which in turn are closely related to the M3 main keypad.
23:57:47kugelI'd rather not count on that
23:57:48n1siiuc placing the most common cases early in the switch is more imporant
23:58:16n1s(or early in the if-else cascade)
23:58:31amiconnn1s: If it's a jump table, order doesn't matter. Gcc will reorder them anyway as it sees fit.
23:58:53n1samiconn: ah

Previous day | Next day