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

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

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

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

#rockbox log for 2012-12-07

00:03:32 Quit OthnielGraichen (Quit: CGI:IRC (Ping timeout))
00:04:24***Saving seen data "./dancer.seen"
00:20:45 Quit pamaury (Remote host closed the connection)
00:21:11 Quit the-kyle (Ping timeout: 246 seconds)
00:21:56 Join Gallomimia [0] (
00:24:26 Quit Gallomimia (Client Quit)
00:26:11 Join the-kyle [0] (~kyle@
00:29:28 Quit bertrik (Ping timeout: 264 seconds)
00:45:43 Quit soap (Quit: soap)
00:48:02 Quit Rower85 (Quit: Hmmm...)
00:52:08 Quit sakax (Ping timeout: 252 seconds)
01:01:31 Quit ender1 (Quit: Stupidity cannot be cured. Stupidity is the only universal capital crime; the sentence is death. There is no appeal, and execution is carried out automatically and without pity.-- Robert A. Heinlein)
01:04:31 Join sakax [0] (
01:11:34 Quit SuperBrainAK (Ping timeout: 264 seconds)
01:20:01 Join soap [0] (
01:20:01 Quit soap (Changing host)
01:20:01 Join soap [0] (~soap@rockbox/staff/soap)
02:04:25***Saving seen data "./dancer.seen"
02:04:35 Quit saratoga (Quit: Page closed)
02:04:58 Join krabador [0] (
02:29:25 Quit ruskie (Quit: ...)
02:37:20 Join ruskie [0] (ruskie@sourcemage/mage/ruskie)
02:47:32 Quit the-kyle (Ping timeout: 240 seconds)
03:22:58 Quit krabador (Ping timeout: 264 seconds)
03:29:24 Quit dfkt (Quit: -= SysReset 2.55=- Sic gorgiamus allos subjectatos nunc.)
03:37:11 Join krabador [0] (
03:43:12 Quit froggyman (Ping timeout: 255 seconds)
03:45:03 Join darkham_ [0] (
03:46:35 Quit darkham_ (Client Quit)
03:48:46 Quit krabador (Ping timeout: 264 seconds)
03:55:46 Join Gallomimia [0] (
04:04:27***Saving seen data "./dancer.seen"
04:12:44 Quit amiconn (Disconnected by services)
04:12:45 Join amiconn_ [0] (amiconn@rockbox/developer/amiconn)
04:12:49 Nick amiconn_ is now known as amiconn (amiconn@rockbox/developer/amiconn)
04:13:43 Join pixelma_ [0] (pixelma@rockbox/staff/pixelma)
04:13:43 Quit pixelma (Disconnected by services)
04:13:44 Nick pixelma_ is now known as pixelma (pixelma@rockbox/staff/pixelma)
04:31:10 Quit yuriks (Ping timeout: 244 seconds)
04:34:43 Join jhMikeS [0] (~jethead71@
04:34:43 Quit jhMikeS (Changing host)
04:34:43 Join jhMikeS [0] (~jethead71@rockbox/developer/jhMikeS)
04:37:05 Join TheSphinX^ [0] (
04:38:02 Quit TheSphinX_ (Read error: Operation timed out)
04:38:23 Quit Gallomimia (Quit: Gallomimia)
05:04:38 Join Gallomimia [0] (
05:06:59 Join eckoit [0] (~ryan@
05:26:01 Quit [7] (Disconnected by services)
05:26:09 Join TheSeven [0] (~quassel@rockbox/developer/TheSeven)
05:35:49 Join SuperBrainAK [0] (
06:01:44 Quit ParkerR_ (Excess Flood)
06:04:29***Saving seen data "./dancer.seen"
06:08:27 Join ParkerR [0] (ParkerR@unaffiliated/parkerr)
06:09:07 Nick ParkerR is now known as ParkerR_ (ParkerR@unaffiliated/parkerr)
06:09:42 Join akaWolf [0] (~akaWolf@unaffiliated/akawolf)
06:27:25 Join yuriks [0] (~yuriks@opentyrian/developer/yuriks)
06:35:11 Join mortalis [0] (
06:38:23 Quit SuperBrainAK (Quit: pbly going to sleep /_\)
06:41:26 Quit shamus (Ping timeout: 255 seconds)
06:43:04 Join shamus [0] (
07:17:54 Quit shamus (Ping timeout: 255 seconds)
07:18:59 Quit yuriks (Ping timeout: 248 seconds)
07:24:00 Join shamus [0] (
07:38:44 Join rmmh [0] (
07:41:46rmmhthe Clip+ page says that some models come with Si4702 chips, is there any ways to determine what model number my chip is? the FM info says SI4700
07:42:15rmmhbut debug.c doesn't seem to have any discrimination between SI470x models
07:53:28rmmhfrom a datasheet and the registers on the fm info page, it looks like it actually is a Si4703... which means that this Clip+ should support RDS
07:56:50 Join dongs [0] (
07:56:56dongshow the hell can I configure voice activated recording
07:57:03dongswhat are those dB values and why is there no preview
08:03:58 Quit ps-auxw (Ping timeout: 252 seconds)
08:04:30***Saving seen data "./dancer.seen"
08:05:50 Join ps-auxw [0] (~arneb@2001:470:c807:0:1532:4e5f:2ad3:4123)
08:06:21Mirdongs: dB values are the audio level
08:07:04dongsyea and?
08:07:08dongswhere's the preview of what it is?
08:07:13dongsi have no idea what '35dB' is
08:07:29dongsim talking about the UI for configuring voice activated recording
08:07:36Mirpreview of what what is?
08:08:38dongs'start above -35dB'
08:08:50dongsi would have expected a preview in that same dialog that actually shows the levels
08:08:54dongsand maybe numbers.
08:10:59dongsthere is a graph when previewing recording, but no nubmers in it
08:11:05 Quit Scromple (Quit: Leaving)
08:19:32 Join wodz [0] (
08:20:18wodzpamaury: So what is the consensus? I looked at logs and I don't quite get how should I proceed.
08:22:41wodzgevaerts: Regarding rk27xx usb hw - OF simply don't touch requests handled by hw. It doesn't change internal state for GET ADDRESS and SET CONFIGURATION.
08:22:59dongsso um anyone? how do I configure the trigger limits in voice recorder
08:23:38wodzSo in this case it doesn't need interrupt. I find this awkward but thats how they manage this.
08:23:53wodzdongs: Whats the problem?
08:24:03dongswodz, meaningless numbers without preview
08:24:07dongsfor start/stop triggers
08:24:16dongsI don't know what "-35dB" is and there's no reference/preview
08:24:37wodzHow do you expect the preview to be working?
08:24:54dongsi dont know , perhaps mic bars with actual numbers would be helpful
08:25:02dongsso i could you know, make some noise and see if thats above 35dB or not.
08:25:11wodzyou are welcome to propose the patch
08:25:20dongsthis is under settings/recording/trigger
08:25:35dongsi found references to voice activated recording as old as 2008
08:25:43dongssurely in 4+ years someone had similar idea.
08:26:05wodzIf there is no such option you have been proofed wrong
08:26:14 Quit shamus (Read error: Connection reset by peer)
08:26:23dongsis tehre any way to get actual dB values when recording then?
08:26:28dongsall i see is bars, which also don't mean anything to me
08:26:32dongsas Im not an audio pro
08:26:37dongsare the ticks signifying anything?
08:27:13 Join shamus [0] (
08:28:30 Join melmothX [0] (~melmoth@unaffiliated/melmothx)
08:28:34wodzthe bar represent signal level in either linear or logarithmic scale (depending on setting IMO). right most is 0dB left most is -infinity (or the max attenuation hw supports - don't know)
08:29:06dongsit killed all my USB devices again
08:31:35 Quit shamus (Read error: Connection reset by peer)
08:32:09 Join shamus [0] (
08:32:24dongsannd now its frozen with USB pic.
08:32:28wodzdongs: The point is that -35dB is not meaningless but has very precisely defined meaning.
08:32:46dongswodz, i dont have anything that displays current sound level in dB.
08:33:16wodzthats rather unrelated
08:33:30 Join einhirn [0] (
08:36:35dongsso how does one configure voice activated recording then
08:37:21rmmhI'm trying to enable RDS support on Clip+, and it looks like it's choking because the hardware can come with multiple tuners −− do I need to implement custom tuner_get_rds_info functions that will respond to either one
08:38:24dongsrmmh: is there any way to see current mic sound levels as dB on clip+
08:39:38rmmhdongs: maybe a custom recording theme? (.sbs)
08:39:48dongswhere do i find that?
08:41:08dongsi see something that looks like .ini files there
08:41:15dongswouldnt even know where to start changing as a user.
08:41:20dongsis there anything I can just download and use?
08:43:02dongs this is all nice and well documented, but has the biggest problem for a normal user: all the inputs are in numeric dB, which are not available anywhere in UI as far as I can tell
08:43:35wodzin recording settings you can set min and max for peak meter. So you can set low limit the same as your trigger and you will instantly see if you are above threshold or not
08:44:24wodzI agree that tick could have numerical value displayed.
08:45:48 Join ender1 [0] (
08:46:13copperTorne: I'm getting high stereo crosstalk from my Galaxy Nexus which has a TRRS socket with a TRS jack inserted
08:48:31 Join bertrik [0] (~quassel@rockbox/developer/bertrik)
08:49:04dongsinstalling just themes doesnt work.....
08:49:09dongsin rockbox utility
08:50:10dongsrmmh: i dont even see how to do custom skin in recording.
08:50:42wodzrecording screen does not support skins. Only sbs AFAIK
08:50:51dongswell i dont even know what taht is.
08:51:32dongsi'm a user, who reads manuals.
08:52:44dongshaving to replug every USB device into my computer is getting fucking old too
08:55:36 Join Zagor [0] (
08:55:36 Quit Zagor (Changing host)
08:55:36 Join Zagor [242] (~bjst@rockbox/developer/Zagor)
09:00:57 Join kevku [0] (x@2001:470:28:773::3)
09:01:39 Quit wodz (Quit: Leaving)
09:05:26dongsman... rockbox customWPS manual is like a forest of shit I dont want to / need to know
09:05:37dongsall I want to do is display dB values in recording that cant be SO HARD TO DO
09:06:21dongsi dont understand how every setting related to recording is measured in dB, but there is no way to see the numbers.
09:14:07rmmhhrm, seem to have my radio received RDS data now, but it's not propagating up...
09:15:48dongsi dont think i can even receive a single FM station here
09:15:57dongsi havent listened to any radio in about 20 years
09:16:04dongsis tehre actually places where radio is relevant in 2012?
09:16:10dongssome 3rd world countries i guess?
09:16:50rmmhmost people in cars?
09:21:01 Quit bertrik (Ping timeout: 265 seconds)
09:25:44rmmhis there any sort of kernel logging I can do?
09:26:01 Join lorenzo92 [0] (
09:27:32lorenzo92pamaury (logs): no it doesn't handle all the YP players
09:27:45lorenzo92I have reversed also another YP firmware format ;)
09:27:52lorenzo92one of the recent models
09:29:47lorenzo92pamaury (logs): so till now I know something like 3 different firmware formats: the old YP -> encrypted tar.gz; the middle YP -> R0/R1/*maybe R2/ -> the latter YP -> not encrypted, just play with some offsets
09:31:34 Join petur [0] (~petur@rockbox/developer/petur)
09:32:24lorenzo92kugel pamaury: here you can find my (dirty :D) python script for the latter:
09:32:57lorenzo92I think the best option is to <<create an option>> since generalization cannot be feasible, in my opinion
09:33:57lorenzo92pamaury: btw, I finally modded the rootfs to get rid of the password in Z5 :) HW is really interesting, if you need some information just ask (there are also a lot of diagnosis programs)
09:37:23 Quit kevku (Ping timeout: 264 seconds)
09:43:15 Join LinusN [0] (
09:48:48 Join pamaury [0] (
09:48:48 Quit pamaury (Changing host)
09:48:48 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
09:52:44 Join wodz [0] (
10:04:32***Saving seen data "./dancer.seen"
10:10:33 Quit nosa-j (Ping timeout: 240 seconds)
10:10:41 Quit Gallomimia (Ping timeout: 260 seconds)
10:12:18 Join nosa-j [0] (~m00k@
10:17:19 Quit lorenzo92 (Ping timeout: 244 seconds)
10:18:11 Quit pamaury (Ping timeout: 248 seconds)
10:20:29 Quit nosa-j (Ping timeout: 260 seconds)
10:24:45 Join nosa-j [0] (~m00k@
10:29:42wodzha, I think I know why rk27xx usb exhibits 'no interfaces' - it does not define any type of endpoints in config.h
10:33:27 Quit tchan (Ping timeout: 246 seconds)
10:43:49 Quit nosa-j (Ping timeout: 250 seconds)
10:47:29 Quit einhirn (Quit: Miranda IM! Smaller, Faster, Easier.
10:52:16 Join nosa-j [0] (~m00k@
10:56:19 Join pamaury [0] (
10:56:19 Quit pamaury (Changing host)
10:56:19 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
10:58:11 Quit nosa-j (Ping timeout: 276 seconds)
10:59:07wodzpamaury: ping
11:01:28 Join nosa-j [0] (~m00k@
11:01:41wodz<wodz> ha, I think I know why rk27xx usb exhibits 'no interfaces' - it does not define any type of endpoints in config.h
11:02:20wodzand another thing - how should I proceed with SET ADDRESS and SET CONFIGURATION requests
11:02:22pamaurythe outcome of the discussion is that we keep this design because in theory nothing bad can happen and we allow for direct call of set address and set config so that this kind of controller can be handled. I'll do the change in a minute
11:02:53pamauryby the way, your current patch still has a problem: you need to check for changes in the flag both the tick task and at the beginning of the udc_int
11:03:41wodzyou mean to sort out the boot caused by usb plug?
11:04:24pamauryI don't understand
11:04:28wodzme too
11:05:16wodzwhat you mean by udc_init ?
11:06:10pamauryinterrupt handler
11:07:09 Quit nosa-j (Ping timeout: 260 seconds)
11:07:16pamaurydo you see why ?
11:07:25wodzyes, I think
11:07:38pamaurygreat :)
11:07:52pamauryI'll do the change in usb_core and pastebin it
11:08:14wodzcan't the stack simply ignore this requests and assume the hw will take care? Thats what OF does
11:09:14gevaertsThe reason we care is that we decide which interfaces to present at connect time, and we only kknow about the connection when we see one of these
11:09:54gevaertsChanging that is probably not impossible, but that's more thinking :)
11:10:33pamaurynot to say that this is the only way to properly handle multiple configurations (even though we don't have such a thing)
11:11:27pamauryAnyway this controller is pretty dumb
11:11:52wodzthats out of question :-)
11:12:17 Join nosa-j [0] (~m00k@
11:14:16pamaurygevaerts: about directly handling the set addr/config, there is a slight issue: what if we are in an interrupt context ? I don't think that's dangerous but I don't really like the idea of doing the full init_connection() in an interrupt handler
11:16:47wodzsince this issue is so specific maybe we can live with fake packets crafted and simply take care that no two such packets are issued in single run
11:17:58wodzhmm, this does not allow to check in udc_int though :/
11:18:33pamauryI prefer a direct handling but since it could be called for the interrupt handler we need to make sure that's safe
11:19:33wodzI still can't think of the reason anyone want such hybrid scheme as this core provides. Thats insane.
11:21:51 Quit nosa-j (Ping timeout: 265 seconds)
11:22:26pamaurythat's mostly because we can't do dynamic allocation
11:23:10wodzno I mean in general - whats the purpose of handling some specific setup steps in hw?
11:23:53wodzand without proper irq notify
11:25:46gevaertspamaury: good point
11:26:02gevaertsEspecially the init_connection() bit is dangerous I'd say
11:26:30pamaurygevaerts: we could introduce some new ids in the usb_queue ?
11:26:46gevaertsThat's probably the best way
11:27:31pamaurysomething like USB_NOTIFY_SET_ADDR, which argument is the addr and USB_NOTIFY_SET_CONFIG, which argument is the config
11:27:47 Join nosa-j [0] (~m00k@
11:27:49*gevaerts nods
11:28:04pamauryok, let me come up with an implementation and then we'll see
11:28:56wodzpamaury: maybe we don't need tick_task then and it will be sufficient to check in udc_int?
11:29:43pamauryI think that's sufficient yes. Because if you don't get any int after set config it means you don't need to configure the drivers, same things for set address
11:36:11 Join einhirn [0] (
11:37:59 Quit nosa-j (Ping timeout: 260 seconds)
11:39:32pamauryI'll be away for a moment, let me know if you have comments on this
11:40:55 Quit solarcloud_3scre (Quit: Leaving)
11:58:45 Join webguest52 [0] (
12:04:35***Saving seen data "./dancer.seen"
12:05:48webguest52Hi! I am tihinking about contributing to the Rockbox firmware. First, I had a look at the tracker. There are many submitted patches that were ignored, starting from 2006. Why is that? I don't want my improvements to end like that.
12:06:29Tornewe don't use the bugtracker for patches any more
12:06:39Torneit wasn't ever a particularly good way to keep track of stuff
12:07:14Tornelots of patches never get committed because they aren't very good, or they don't work on all targets, or the author never gave us their real name, or because the author never actually asked an appropriate developer to take a look at it
12:07:27Tornebut, we also never committed lots of thos epatches because we only have finite time to look at them
12:07:39Tornewe use gerrit for patches now, which makes it easier to look at and review patches
12:07:48Tornewhich makes it a bit more likely :)
12:08:10webguest52a bit ;-)
12:08:42Torneyah, a bit. it means it takes a little less time to review/land a patch, which helps
12:08:46Torneit doesn't solve any of the other problems
12:09:11Tornebecause those are mostly the standard problems that always apply to software: sometimes people's changes are not done well, even if the idea is good
12:09:24Tornealso, sometimes people propose changes where we don't think the idea is worthwhile :)
12:09:39wodzspeaking about rotting patches - IMO we should commit g#339 and g#340
12:09:42fs-bluebot3Gerrit review #339 at : Fix FS #12391 : Memory corruption on PP502x after commit_discard_idcache() by Boris Gjenero (changes/39/339/1)
12:09:52Torneif you want to contribute without potentially wasting time and effort, talk to us on irc/forums/mailing lists first about what you want to do and how you are going to do it
12:10:06Torneand hang around here where you can see the process going on
12:10:14Torneand it's far more likely someone will pay attention :)
12:10:46Tornethat's not really rotting, that's from less than six eeks ago :)
12:11:37Tornewodz: i suspect nobody has committed them because dreamlayers has the ability to do so himself, and thus people are assuming he will do it when he thinks it's ready
12:11:49Tornethis may not actually be true, though, so yeah.
12:11:58wodzTorne: preceded by xxxx (don't know exactly) weeks on FS
12:12:34Torneanyway, i hadn't seem those
12:12:44Tornebut they are of interest to me
12:12:52Torneso i'll take a look if i find time
12:13:29wodzFS #12391
12:13:30fs-bluebot 3PP502x commit_discard_idcache() causes memory corruption (bugs, requires)
12:13:39Torneyah, i know the issue
12:13:47gevaertspamaury: looks reasonable. I think I'd call usb_core_do_set_addr() and usb_core_do_set_config() directly from usb.c though, but that's a matter of taste
12:13:49Tornei didn't know there was a finished patch
12:13:55Torneso, now i do :)
12:13:59Tornei'll look when i get a chance
12:17:59 Join Topy44|2 [0] (
12:21:41 Quit Topy44 (Ping timeout: 264 seconds)
12:23:25 Quit petur (Quit: *plop*)
12:33:53 Quit fs-bluebot (Ping timeout: 240 seconds)
12:35:37 Quit bluebrother (Ping timeout: 250 seconds)
12:35:39 Join fs-bluebot [0] (
12:36:27 Join the-kyle [0] (
12:37:26 Join bluebrother [0] (~dom@rockbox/developer/bluebrother)
12:40:53 Quit the-kyle (Ping timeout: 264 seconds)
12:41:49 Join nosa-j [0] (~m00k@
12:47:28webguest52I recently updated from a very old Rockbox version. There were two things I noticed. 1st, the delay between selecting a file and actually hearing a sound increased dramatically. I guess, data is loaded into an increased buffer. Is that buffer size an option I can change somewhere?
12:50:12Tornebuffering happens int he background; we don't wait until the buffer is full before playback starts
12:50:40Torneis it a hard disk based player?
12:50:44Tornedo you have the directory cache enabled
12:50:56webguest52its an old archos ondio
12:51:05Tornealso, did you keep your settings from the old version, or start from scratch?
12:51:50Torneturn the directory cache on, if it's not already enabled. not having it on is usually the cause of slow playback starts on players with spinning disks
12:52:07webguest52its no spinning disks. its mmc
12:52:23Torneit may still help
12:52:40Tornetry it :)
12:53:49webguest52i'll try. still something else must have changed.but that's probably hard to find after many years of development
12:54:30webguest52so is the number of bytes that have to be read before playback starts somewhere configurable?
12:55:44Torneit starts decoding as soon as there's anything there to decode
12:56:03Tornei don't think there's any minimum
12:56:27Tornei don't think your problem is to do with buffering
12:56:37Torneit may be a more general problem with IO speed
12:57:09webguest52okay? I'll try to recover the old versoin from an old backup disk and do a direct comparison.
12:57:59Torneyou could do, but i don't mean that it's not a real problem
12:58:22Tornei mean, the new version is probably either actually slower at accessing the storage, or it's doing some extra work that the old version didn't do that also requires IO
13:06:39webguest52ah, I understand. Once I have time, have installed the toolchain, and that delay bothers me too much, I'll try to find the course
13:07:09webguest52Is there some documentation how I would start debugging? Is there some kind of log or console?
13:14:51webguest52... where I could post debug messages to?
13:23:17pamaurygevaerts: ok, should I commit this now or as part of the rk27xx usb work ?
13:23:46gevaertsI think it can go in separately
13:24:05gevaertsIn fact, I think that will make possible future investigations easier :)
13:24:47 Quit nosa-j (Ping timeout: 255 seconds)
13:32:40 Quit webguest52 (Quit: CGI:IRC (Ping timeout))
13:34:16pamaurylorenzo92, kugel: if there are indeed several formats for the yp firmwares, we should make the samsungtools aware of all known formats. And that's a good of documenting things too
13:38:20 Join nosa-j [0] (~m00k@
13:44:05 Join timo [0] (
13:44:29 Nick timo is now known as Guest50060 (
13:45:09wodzpamaury: So are you going to commit usbstack rework?
13:45:34wodzI don't know if I find time to fiddle with this today though
13:46:53 Quit Guest50060 (Client Quit)
13:47:12pamaurydone. So now all you need is to call usb_core_notify_* on set addr and set config. I won't have time for this today
13:47:29 Quit nosa-j (Ping timeout: 260 seconds)
13:49:05gevaertspamaury: you mean call usb_signal_notify(), right?
13:50:07pamauryno, usb_core_notify, just like the drivers call usb_core_control_request
13:50:35pamauryusb_signal_* functions are used by usb_core to queue
13:50:52gevaertsOh, right. Sorry
13:50:56*gevaerts gets more coffee
13:50:59 Quit jhMikeS (Ping timeout: 248 seconds)
13:51:53pamaurySETUP packet path is: usb_core_control_request -> usb_signal_transfer_completion -> usb_core_handle_transfer_completion
13:51:53pamauryNOTIFY packet path is: usb_core_notify_* -> usb_signal_notify -> usb_core_handle_notify
14:04:38***Saving seen data "./dancer.seen"
14:14:49 Join amayer_ [0] (
14:23:50 Quit dongs (Ping timeout: 252 seconds)
14:29:48 Join dongs [0] (
14:39:03 Quit dongs (Ping timeout: 265 seconds)
14:39:22 Join dongs [0] (
14:43:32 Quit mortalis (Quit: Leaving)
14:50:46 Join solarcloud_3scrn [0] (~solarclou@
14:57:03 Join tchan [0] (~tchan@lunar-linux/developer/tchan)
15:11:43 Quit wodz (Quit: Leaving)
15:22:20 Join kevku [0] (x@2001:470:28:773::3)
15:25:09 Join Robin0800 [0] (
15:35:57 Join timo [0] (
15:36:00 Quit timo (Client Quit)
15:36:08 Join timo [0] (
15:36:21 Nick timo is now known as Guest56048 (
15:37:09 Quit Guest56048 (Client Quit)
15:37:13 Join lebellium [0] (
15:39:43 Join tudortmund [0] (81d98329@gateway/web/freenode/ip.
15:50:26tudortmundHello everybody. I really hope you can help me.
15:50:41tudortmundI'm working at the University of Dortmund (Germany) and we are searching for a technical solution regarding a current research project.
15:50:52 Join the-kyle [0] (
15:50:55tudortmundWe would like to manipulate cheap mp3 players in a way that 1) the participants of the study can't change the content of the player and 2) we are able to save some kind of logfile including detailed usage statistics.
15:51:26tudortmundIs Rockbox able to do that?
15:51:40Zagorsince we are rockbox, #1 is pretty much out of the question
15:52:02Zagorour whole existence is about letting the user modify their player
15:52:03tudortmundMust we program a Plugin or is one available that might be appropriate?
15:52:19gevaertsA plugin definitely won't be enough
15:52:32 Join y4n [0] (~y4n@unaffiliated/y4ndexx)
15:52:39funmanI would write a plugin and use autorun feature
15:52:43 Join dfkt [0] (dfkt@unaffiliated/dfkt)
15:52:45 Join lebellium_ [0] (
15:53:00gevaertsWhat are people supposed to be able to do?
15:53:48tudortmundOK. Cool :)
15:54:41AlexPBy manipulate contents, you mean add/remove tracks?
15:55:12tudortmundThey should be able to listen to the tracks we put on the player, switch between the tracks, but not add or remove tracks, right.
15:55:46funmansolder together the data + and - of usb ;)
15:55:56AlexPRight, so you would want to disable USB and dual boot, but then you won't be able to change back to a version that would allow to change it afterwards
15:56:08Zagoryeah, you're going to need physical modifications for that
15:56:13 Quit lebellium (Ping timeout: 240 seconds)
15:56:24 Nick lebellium_ is now known as lebellium (
15:56:39amayer_...or rewrite the usb drivers to require some kind of authentication after connect
15:56:50Zagorof course it all depends how thoroughly you want to prevent them
15:57:06Zagoramayer_: sure, that's fine for our driver. but many devices can also use the OF driver.
15:57:24AlexPZagor: yeah, you would have to mod a bootloader to remove dual boot too
15:57:24amayer_Zagor: very true
15:57:34pamauryone solution would be to chose a player with a recovery mode and modify rockbox to prevent usb from working except if you enter a PIN code for example
15:58:20pamaurythat way you can still recover but it makes it impossible to modify the player except if you *really* want it
15:58:25amayer_and it couldnt be on an apple product, because they could do emergency recovery and install what ever they want(unless you got the physical modification route)
15:59:00AlexPamayer_: He also said cheap which rules out Apple :)
15:59:12tudortmundsorry, but what do you mean "physical modification"?
15:59:13pamaurytudortmund: when you say that the people can't add/remove track, you mean impossible by accident or really impossible
16:00:04amayer_tudortmund: modify the usb connector to not accept data
16:00:20Zagoror fill it with epoxy
16:00:46Zagorof course that would make it difficult to charge :-)
16:01:15tudortmundour participants are children about 10 years. it's not that i am afraid that they might try to "hack" the players. it's just that they should not be able to put it into their usb-ports and simply replace our tracks by their music. ;)
16:01:23AlexPZagor: You also mod it to allow inductive charging of course :)
16:01:38funmantudortmund: well you can just disable usb data
16:01:41 Quit pamaury (Read error: Operation timed out)
16:02:03funmanand hope 10yo kids don't search about rockbox, dual boot etc..
16:02:23funmanif they manage to write their music to it i'd say they deserve to use it
16:02:24Tornei dunno that our UI is particularly friendly/suitable for 10 year olds, however
16:02:34Tornewe have a lot of menus and options
16:03:08Zagor10 year-olds are pretty device savvy these days
16:03:43amayer_tudortmund: it could be done, but you are looking at a lot of work(either codeing, or physical modifications + coding)
16:03:45tudortmund:) and what about my other point: the logfile?
16:04:11Tornerockbox has a logging system you could use for that, if you just add log statements where you want them
16:04:18tudortmundok... :-?
16:04:39***Saving seen data "./dancer.seen"
16:05:00amayer_well if you are going as far as to do all the work that they cant add/remove songs or boot into OF then adding some lines of code to log the currently playing song are relativly easy
16:05:37 Quit the-kyle (Ping timeout: 276 seconds)
16:08:15 Quit lebellium (Read error: Connection reset by peer)
16:08:16tudortmundso assuming we would hire someone to do that for us: what would you think how long it takes to implement this whole thing?
16:08:32Torneanyway, yeah, everything you want to do can be done if your bar for "security" is relatively low and you doesn't intend to prevent deliberate attempts to hack it by people who know lots of computer stuff
16:08:36tudortmundjust roundabout...
16:08:41 Join lebellium [0] (
16:09:17tudortmund@Torne: sounds good. :)
16:09:42Tornebut if i was doing this i would want to allow plenty of time to test it on some kids
16:09:49Torneto discover if they end up stuck in odd places in the UI
16:09:58Torneso that i could just kill the bits of UI that were a problem
16:10:21Tornewe get plenty of support requests for people stuck using rockbox to do fairly easy and obvious things from adults :p
16:12:33tudortmundjust to be able to calculate this and decide whether it is an option for our project: how many hours of work would you calculate? i really have NO idea... :-P
16:13:24tudortmundAnd just to get that clear: Does somebody know another option / solution that might help me?
16:14:28 Quit lebellium (Ping timeout: 264 seconds)
16:16:07 Join the-kyle [0] (
16:16:21gevaertsSo to be clear, you want logging, no USB storage unless say a PIN is know, and possibly UI tweaks?
16:16:38 Join pretty_function [0] (~sigBART@
16:16:55Torneno usb storage at all is probably enough on many devices, i.e. force dual boot to uninstall it
16:17:00Torneless complciated to do
16:17:25Torneas long as the bootloader for that device deosn't auto-dualboot on usb then it's unlikely that a kid will randomlyu discover it :)
16:17:32tudortmundyes! :)
16:17:45tudortmundthat's an easy one.. he? :-P
16:18:24gevaertsOr you could always alter the bootloader to make dual boot harder to do, such as e.g. requiring a combination of buttons instead of just onme
16:18:39Tornedepending exactly what you want to log, also, the log might be a good start
16:18:47Torneif you just want a playback history then the log is already that
16:18:53Tornethough it logs based on partial playbacks
16:19:00gevaertsIf you're talking about cheap players, I'd assume something like the sansa clip family is what you're looking at?
16:19:02Tornebut could be modified to be more specific
16:19:35gevaertsYes, knowing what exactly you want to log can help a lot in guessing the amount of work :)
16:20:18tudortmundyes. sansa would be one of our choices
16:21:01tudortmundok... the project is in its very beginning, but i'll try it:
16:21:13 Quit the-kyle (Ping timeout: 276 seconds)
16:21:35Torneso if you needed minimal changes to just block usb data and make dualboot harder to invoke, and tweak the existing logging code a bit, someone who actually knew our codebase could probably do taht in a few hours
16:21:43Tornemore extensive logging woudl take longer
16:21:50Tornesomeone starting from scratch in our code would also take longer
16:22:13tudortmundwe would prvide the players to the kids for i.e. 4 weeks and then like to "read" from a logfile how long they heard which track.
16:22:39Tornethat's all you want? so each time a track is played, log how much of it was played, basically?
16:22:48Tornethe logging code already does almost exactly that :)
16:23:00Tornethe normal behaviour is to log tracks that are 50% or more played
16:23:16Tornei.e. if you listen to at least half of it
16:23:23Torneand i think the log alrady includes how much of it you listened to
16:23:25tudortmundit would not be enogh to know that they "started" track 1 five times and that they started track 2 six times. it would be better if we could also know how long they actually listened to that specific track.
16:23:28Torneso, you could just kill the 50% minimum
16:23:53Torneit basically does that already :)
16:24:02Torneso that would actually also be pretty easy
16:25:06tudortmundideally we would see something like: "2012/12/07: 35min overall (track1: 12min, track2: 7min, ...)
16:25:35gevaertsI'd say that's what postprocessing is for :)
16:26:22tudortmundgreat to hear!
16:27:11gevaerts;a=blob;f=apps/scrobbler.c;h=1a0ad9390ed31efa7f19c5da430d5a5e41f8c0de;hb=HEAD#l173 is where the actual log file entries are made
16:27:39tudortmundthank you very much for answering my questions so extensively! :) I'll give that informations to my projet team and maybe in some weeks i'll be back here to ask more specific questions... ;)
16:27:58gevaertsIt looks like it doesn't actually log the played time, but then it also looks like you only need something like a three line change
16:29:28tudortmundCan i usually reach you in here? If not you might wanna shoot me an email to give me your email adress: timo.lueke(at)
16:29:48gevaertsYou should always be able to find some people in here
16:30:29tudortmundBecause i would really like to ask you some more questions, but i have to get to a meeting which already started.... they are waiting for me. :-P
16:30:38gevaertsI'd say the people who actually participated in the discussion are all likely to be around during business hours
16:31:06tudortmundOK. Great. Thank you very, very much! :)
16:33:16 Part tudortmund
16:37:01 Part eckoit
16:44:38 Join lebellium [0] (
16:50:44 Quit apo (Read error: Operation timed out)
16:51:54dfkti checked gerrit and flyspray but didn't find anything relevant - is the RAAA issue known that on some devices (galaxy tab with CM9 in my case) the screen won't turn on again if rockbox is running?
16:52:32dfkti run RAAA with sleep timer, and after rockbox shuts off, the screen can be turned on again
16:53:00 Join apo [0] (
16:56:26 Quit Zagor (Quit: Clint excited)
16:57:57 Quit linuxstb (Quit: This computer has gone to sleep)
16:59:03 Join pamaury [0] (
16:59:03 Quit pamaury (Changing host)
16:59:03 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
17:04:22 Join linuxstb [0] (~linuxstb@unaffiliated/linuxstb)
17:05:01 Quit linuxstb (Client Quit)
17:07:27 Part LinusN
17:33:44 Quit bebna (Quit: Leaving.)
17:40:45 Join efyx [0] (~efyx@
17:44:29 Quit pretty_function (Remote host closed the connection)
17:52:53 Quit Robin0800 (Remote host closed the connection)
17:55:13 Quit einhirn (Quit: Miranda IM! Smaller, Faster, Easier.
18:04:40***Saving seen data "./dancer.seen"
18:11:23 Join pretty_function [0] (~sigBART@
18:12:31 Quit pretty_function (Remote host closed the connection)
18:13:02 Quit FOAD (Quit: I'll be back)
18:13:05 Join linuxstb [0] (~linuxstb@unaffiliated/linuxstb)
18:13:32 Join FOAD [0] (~foad@
18:15:03 Join pretty_function [0] (~sigBART@
18:32:16 Join Rower85 [0] (
18:55:15 Quit ruskie (Excess Flood)
19:08:25 Join froggyman [0] (
19:08:25 Quit froggyman (Changing host)
19:08:25 Join froggyman [0] (~me@unaffiliated/froggyman)
19:11:40 Join theli_ua [0] (
19:15:34 Join Gallomimia [0] (
19:17:57 Quit pretty_function (Remote host closed the connection)
19:23:20 Join ruskie [0] (ruskie@sourcemage/mage/ruskie)
19:24:39 Quit prof_wolfff (Ping timeout: 246 seconds)
19:39:40 Quit Rower85 (Read error: Connection reset by peer)
19:39:55 Join Rower85 [0] (
19:41:47 Join Guest44874 [0] (
19:45:08 Quit Rower85 (Ping timeout: 244 seconds)
20:04:42***Saving seen data "./dancer.seen"
20:09:51 Quit ps-auxw (Quit: leaving)
20:09:54 Join yuriks [0] (~yuriks@opentyrian/developer/yuriks)
20:10:07 Join lebellium_ [0] (
20:11:47 Quit lebellium (Ping timeout: 248 seconds)
20:12:01 Nick lebellium_ is now known as lebellium (
20:19:19 Join SuperBrainAK [0] (
20:43:56 Join bertrik [0] (~quassel@rockbox/developer/bertrik)
20:46:19bertrikrmmh: are you sure you have an RDS capable tuner?
20:46:52bertrikif so, I don't see any barriers to enabling RDS on the clip+
20:47:33 Quit froggyman (Ping timeout: 240 seconds)
20:49:37 Quit Gallomimia (Ping timeout: 250 seconds)
20:53:30 Join stoffel [0] (
20:57:23 Quit yuriks (Ping timeout: 256 seconds)
21:03:30bertrikRDS needs an interrupt to work correctly, perhaps that's not working with the clip+
21:05:34 Quit stoffel (*.net *.split)
21:05:34 Quit shamus (*.net *.split)
21:05:34 Quit linuxguy3 (*.net *.split)
21:05:34 Quit radio12 (*.net *.split)
21:05:35 Quit pystar89 (*.net *.split)
21:06:52amayer_JdGordon: ping
21:07:42 Join ps-auxw [0] (~arneb@2001:470:c807:0:1532:4e5f:2ad3:4123)
21:09:29 Join Gallomimia [0] (
21:10:59 Join stoffel [0] (
21:10:59 Join shamus [0] (
21:10:59 Join linuxguy3 [0] (
21:10:59 Join radio12 [0] (
21:10:59 Join pystar89 [0] (
21:15:10 Quit theli_ua (Quit: Залишаю)
21:18:53 Join theli_ua [0] (
21:21:22 Quit y4n (Quit: PANTS OFF!)
21:22:24 Quit stoffel (Ping timeout: 255 seconds)
21:23:13 Quit theli_ua (Ping timeout: 240 seconds)
21:28:20 Join Gallomimia_ [0] (
21:29:17 Quit akaWolf (Ping timeout: 252 seconds)
21:30:04 Quit Gallomimia (Ping timeout: 264 seconds)
21:30:05 Nick Gallomimia_ is now known as Gallomimia (
21:51:53 Quit bigfoot- (Ping timeout: 252 seconds)
21:53:04 Join yuriks [0] (~yuriks@opentyrian/developer/yuriks)
21:58:05bertrikrmmh: one interesting thing you can try, is to run the diagnostic mode of the OF, it's somewhere in the system menu IIRC
21:58:48bertrikyou can enable diagnostic mode by flashing a "T" version of the firmware (replace the last character of the OF file name with a T)
21:59:06bertrikpossibly you can also enable it by holding all buttons at the same time during boot
21:59:38 Join jhMikeS [0] (~jethead71@rockbox/developer/jhMikeS)
22:04:43***Saving seen data "./dancer.seen"
22:22:50 Join dan_a [0] (~dan@
22:32:09 Join froggyman [0] (
22:32:09 Quit froggyman (Changing host)
22:32:09 Join froggyman [0] (~me@unaffiliated/froggyman)
22:32:13 Join shai [0] (
22:50:12 Join stripwax [0] (~Miranda@rockbox/developer/stripwax)
23:01:31 Quit shai (Quit: Leaving)
23:03:01 Quit yuriks (Ping timeout: 244 seconds)
23:06:31 Part amayer_
23:09:27 Quit melmothX (Quit: bau)
23:17:44 Join _stripwax_ [0] (
23:19:30 Quit stripwax (Ping timeout: 255 seconds)
23:31:44 Quit _stripwax_ (Quit: Miranda IM! Smaller, Faster, Easier.
23:38:09rmmhbertrik: yep. I'm looking at the registers in the FM info screen, and it's the right model and it's also strobing the "RDS info ready" bit. now I'm just looking into how to make the clipplus ISR handle it
23:38:55 Join the-kyle [0] (
23:38:56rmmhbertrik: it's possible the clip+ doesn't have the SI4702 interrupt connected to the CPU, in which case I'll have to do polling
23:42:32 Quit Gallomimia (Quit: I am likely going to change locations)
23:50:31bertrikbut we'll have to poll it quite fast then, RDS has about 11 frames/s
23:51:13rmmhI've done very little embedded stuff, so I'm pretty new to this
23:52:52rmmhthe Toshiba Gigabeat has some RDS_ISR macros for the si4700
23:54:02bertrikI can have a look again at the original firmware to see if there's some radio code for pin setup for an interrupt
23:55:53rmmhit looks like the clipzip and fuzeplus have RDS capabilities, and those are pretty similar hardware (right?)
23:56:19bertrikno, not at all actually
23:56:50rmmharen't they all based on the as3525?
23:57:44bertrikno, not all of them, a lot of them are, but not the fuzeplus

Previous day | Next day