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 2008-12-17

00:02:57 Join Outerlimit [0] (
00:05:02 Join QUICKSTART [0] (
00:06:26 Quit Llorean ("Leaving.")
00:08:34 Quit _MiniGod ("ChatZilla 0.9.84 [Firefox 3.0.4/2008102920]")
00:10:54 Quit __lifeless (Remote closed the connection)
00:13:09 Join _lifeless [0] (n=lifeless@
00:13:36 Part baio
00:18:22 Join Aurix_Lexico1 [0] (
00:22:22 Quit Seed ("cu, Andre")
00:25:02 Join BigBambi [0] (n=alex@
00:25:02 Quit goffa (Read error: 104 (Connection reset by peer))
00:25:58 Join goffa [0] (n=goffa@
00:28:42 Join mc2739 [0] (
00:30:42 Quit bertrik ("Leaving")
00:32:18 Quit Schmogel (Read error: 104 (Connection reset by peer))
00:35:15 Quit Aurix_Lexico (Read error: 110 (Connection timed out))
00:35:51 Join webguest41 [0] (n=4f837793@gateway/web/cgi-irc/
00:36:14 Quit webguest41 (Client Quit)
00:37:30 Quit mofux (Read error: 110 (Connection timed out))
00:39:24***Saving seen data "./dancer.seen"
00:40:07 Quit XavierGr (Read error: 110 (Connection timed out))
00:49:38 Part linuxstb ("Leaving")
00:50:15 Join Lynx_ [0] (
00:50:16 Quit bmbl ("Woah!")
00:51:51 Quit parafin (Remote closed the connection)
00:52:54 Quit MarcGuay (" ajax IRC Client")
00:53:10 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
00:56:06 Join parafin [0] (
00:59:48 Part linuxstb ("Leaving")
01:00:39 Quit gregzx ("ChatZilla 0.9.84 [Firefox 3.0.4/2008102920]")
01:01:05 Join Thundercloud [0] (
01:01:08 Quit tyfoo (No route to host)
01:04:06 Join tyfoo [0] (
01:04:24 Quit tyfoo (Read error: 104 (Connection reset by peer))
01:04:37 Join Llorean [0] (
01:05:25 Quit moos ("Rockbox rules the DAP world")
01:08:46 Quit n1s ()
01:08:53 Quit jhMikeS (Read error: 104 (Connection reset by peer))
01:15:24 Nick JdGordon|zzz is now known as JdGordon (n=jonno@rockbox/developer/JdGordon)
01:16:37 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
01:18:35lucentFS #7831 looks ripe to commit
01:18:41lucentbugfix for FLAC support
01:27:00rasherSounds like something linuxstb forgot about
01:27:31 Quit PaulJam (".")
01:30:12hillshum_how about adding a list of recent forum posts to the homepage
01:30:52Lloreanhillshum_: It's pretty full as it is.
01:31:02LloreanThe front page of the forum has a list of the most recent forum posts anyway.
01:34:50 Quit herrwaldo ("Konversation terminated!")
01:37:10 Join jhMikeS [50] (n=jethead7@rockbox/developer/jhMikeS)
01:37:31 Part toffe82
01:38:17hillshum_but having all info (svn, twiki, forums, etc) on one page would be nice
01:41:13 Quit Rob2222 (Read error: 104 (Connection reset by peer))
01:41:54 Join Rob2222 [0] (
01:44:36 Quit Thundercloud (Remote closed the connection)
01:53:46 Quit MethoS- (Remote closed the connection)
02:11:27 Join LambdaCalculus37 [0] (n=rmenes@rockbox/staff/LambdaCalculus37)
02:11:53 Join mofux [0] (
02:24:06 Nick fxb is now known as fxb__ (
02:31:56 Quit lasser (Read error: 60 (Operation timed out))
02:39:27***Saving seen data "./dancer.seen"
02:42:17 Quit mofux (Remote closed the connection)
02:42:51 Quit Lynx_ (Remote closed the connection)
02:43:37 Quit QUICKSTART (Remote closed the connection)
02:44:18 Join QUICKSTART [0] (
02:45:20 Join RockRabbit [0] (n=3aac9a01@gateway/web/cgi-irc/
02:46:18RockRabbitBagder: Would like to get hold of a copy of the AMS datasheet. Im working on the Sansa c250. Any chance?
02:57:36 Join agaffney_ [0] (n=agaffney@gentoo/developer/agaffney)
02:57:52 Quit agaffney (Nick collision from services.)
02:58:02 Nick agaffney_ is now known as agaffney (n=agaffney@gentoo/developer/agaffney)
03:03:11 Quit JdGordon (Remote closed the connection)
03:06:44 Join JdGordon [0] (
03:16:41 Join admin1lbo [0] (
03:17:18 Quit aneqrs ()
03:17:38 Join wpyh [0] (n=william@
03:17:48 Part wpyh
03:18:20 Join wpyh [0] (n=william@
03:28:09 Quit QUICKSTART (Read error: 113 (No route to host))
03:29:10 Nick admin1lbo is now known as QUICKSTART (
03:48:01LambdaCalculus37wpyh: I think any information from MTABLE.SYS should be on the SansaV2 wiki page.
04:02:00 Join dfkt_dt [0] (i=dfkt@unaffiliated/dfkt)
04:05:13 Quit rvvs89 (Read error: 60 (Operation timed out))
04:05:16daurnimatorwpyh: I agree with your conclusion
04:05:45daurnimatorwpyh: in my old zaurus, in recovery mode, it switched from an sd driver to an mmc driver, and it had a max of 1gb cards
04:05:58daurnimator(or at least, thats what memory says)
04:06:05 Join blkhawk- [0] (
04:06:11 Quit LambdaCalculus37 ("Ka-chunka")
04:06:23daurnimatorotherwise, I think there was storys that an sd driver that only used up to 1gb took up alot less space....
04:06:30 Join rvvs89 [0] (
04:06:37daurnimatormaybe cause it can be made in 16bit instead of 32 or something??? I don't really know
04:09:14LloreanMy personal guess is that the MMC: labeling predates the hardware of these players, and is probably just a relic. It may be a clue, but it's also a file that's not something users are supposed to be looking into, so I wouldn't put too much weight behind what's in it until some other test is made.
04:20:16 Quit dfkt (Read error: 110 (Connection timed out))
04:21:02 Quit blkhawk (Read error: 110 (Connection timed out))
04:21:03 Nick blkhawk- is now known as blkhawk (
04:25:37daurnimatorLlorean / wpyh:
04:25:48daurnimatorIn older 1.x implementations the standard capacity block was exactly 512 bytes. This gives 4096 x 512 x 512 = 1 gigabyte of storage memory
04:27:56 Quit wpyh (Read error: 110 (Connection timed out))
04:27:58Lloreandaurnimator: Which would suggest it's just a bad assumption in the current SD driver.
04:28:22daurnimatoror well, not a bad assumption, but a programming shortcut
04:28:39daurnimatorjust set something as a constant rather than reading it and making it a var :P
04:29:07 Join wpyh [0] (n=william@
04:29:41wpyhanyway, daurnimator has summed it up nicely
04:30:13wpyhhowever, IIUC we need to use 512-byte sectors for the high capacity (>2GB) MMC cards as well.
04:30:23LloreanWhat daurnimator has pointed out is that we may be improperly implementing the SD standard.
04:30:34wpyhA good way to test it would be to try to read the EXT_CSD register from the "card" and parse its contents
04:30:39wpyhyes, I'm reading the logs
04:30:57wpyhdaurnimator: yes, supporting a 2GB card would add code to support a 1024-byte sector size, so the driver would be a bit larger, which is the case in your Zaurus
04:31:11wpyhFrom the MMC v4.3 spec, it says that the address for memory >2GB uses 512-byte sectors. This explains why we can't access >1GB on the models with >2GB of internal storage.
04:33:51wpyhThe wikipedia entry that daurnimator points out has information on the compatibility between SD and MMC:
04:35:44wpyhBasically, the SPI and 1-bit protocols can be used to access both SD and MMC cards. However, the 4-bit protocol is not. Also, I'm not sure that SDHC and high-capacity MMC are compatible. Since the internal storage on my Fuze doesn't report itself as SDHC (the command fails somehow), I guess they are not compatible.
04:38:32 Quit RockRabbit ("CGI:IRC (EOF)")
04:39:08daurnimatorwpyh: or it may just not be conforming to the standard
04:39:12daurnimatorwhich is highly possible
04:39:29***Saving seen data "./dancer.seen"
04:42:09wpyhwell, that's one possible explanation, but I would not think so. I think the storage inside <=2GB players use the old (standard-capacity) MMC standard which can address up to 2GB, and the storage inside >2GB players use the new (high-capacity) MMC standard, which can address more than 2GB. The new standard uses 512B sectors (IIUC from the MMC v4.3 standard, you can get it free from JEDEC), so we can only address the first 1GB. Which makes sense :)
04:45:31daurnimatorwpyh: you have ?
04:45:43daurnimatorit pretty much has everything, including sd and sdhc
04:48:13 Join Chronon [0] (
05:02:20 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
05:04:04 Quit QUICKSTART (Remote closed the connection)
05:04:47 Quit wpyh (Read error: 110 (Connection timed out))
05:04:55 Join wpyh [0] (n=william@
05:05:21wpyhdaurnimator: it's the SD spec, not SDHC
05:05:46daurnimatorwpyh: it includes both AFAICS
05:06:32daurnimatorwpyh: see page 14
05:06:47daurnimatorStandard Capacity SD Memory Card: Up to and including 2 GB
05:06:47daurnimatorHigh Capacity SD Memory Card: More than 2GB (This version of specification limits
05:06:47daurnimatorcapacity up to and including 32GB)
05:07:06wpyhhm... you're right
05:07:20wpyhI was searching for SDHC and EXT_CSD and found none :p
05:09:03 Join Rob2223 [0] (
05:10:26daurnimatorwpyh: just search for "High Capa"
05:13:31wpyhI think the SD and MMC standards are similar even for high-capacity cards: the size of a data block is fixed to 512 bytes. Hm... but we can't identify the internal storage as SDHC, the send_cmd() call on line 268 in ata_sd_as3525.c fails
05:16:17 Quit Rob2222 (Read error: 60 (Operation timed out))
05:24:07 Quit Chronon (Read error: 110 (Connection timed out))
05:35:50 Join zoyd [0] (n=vinay@
05:42:07 Join pixelma_ [0] (n=pixelma@rockbox/staff/pixelma)
05:42:25zoydanyone around owns a cowon D2?
05:45:08 Quit Aurix_Lexico1 (Read error: 110 (Connection timed out))
05:45:40 Quit pixelma (Read error: 60 (Operation timed out))
05:46:57 Quit mc2739 ("ChatZilla 0.9.84 [Firefox 3.0.4/2008102920]")
05:56:31 Quit JdGordon (Read error: 104 (Connection reset by peer))
05:57:19lucentwpyh: I own an 8GB fuze and have microSD cards of 8GB capacity, I can run tests if you need someone to
05:59:51 Quit amiconn (Read error: 110 (Connection timed out))
05:59:55 Join amiconn [50] (n=jens@rockbox/developer/amiconn)
06:04:13 Join JdGordon [0] (
06:06:00 Join _Auron_ [0] (
06:19:41 Join Chronon [0] (
06:21:49wpyhlucent: yes, please. One very useful test would be to put a few music files on the internal storage, a few on the microSD card, and then boot to the OF. It would find the files and write to MTABLE.SYS . Then please upload that MTABLE.SYS file somewhere for us to see.
06:22:24wpyhlucent: the internal storage will be referred to as "mmc:0", but I'm not sure what the microSD card will be referred to as.
06:26:20lucentwpyh: I can certainly do this, one minute please
06:26:40wpyhlucent: don't worry −− I have to go out for a few hours anyway :)
06:27:49lucentwpyh: how do we find this though, special filenames that I'm saving to internal and external?
06:29:03wpyhthis is all contained in the MTABLE.SYS file −− it seems to be a simple database (with visible strings). So you can do "strings MTABLE.SYS | less" and see all the pathnames. However, it would be very helpful if you could upload it somewhere
06:29:57lucentwpyh: the external card is referenced as mmc:1
06:31:13wpyhthis seems to support Llorean's idea that "mmc" is a leftover relic.
06:31:33wpyhI'm still curious though
06:31:50lucentMMC being a subset of SD, I'm not surprised about the naming
06:31:53*wpyh has to go away for a while. hopes kugel/funman will notice
06:32:03lucentthanks wpyh
06:32:05zoydi have a Philips SA60xx series player, wondering what are the rockbox chances on it?
06:32:27wpyhlucent: thanks to _you_ :)
06:34:06lucentwpyh: I pretty much confirm that rockbox doesn't run if it's loaded anywhere above 1GB offset on the internal memory of this device, and panics with an SD timeout message accessing files internally above 1GB
06:34:26 Quit perrikwp (" ajax IRC Client")
06:34:33 Join perrikwp [0] (i=18ac0c41@gateway/web/ajax/
06:35:04 Join Acky [0] (
06:35:36lucentI don't know the exact offset which it fails, I could run some more tests, but the thing I discovered is that it panics if I load a lot of data onto the internal memory and then copy .rockbox/* over
06:36:17lucent(panics if I run rockbox firmware after doing this)
06:39:34***Saving seen data "./dancer.seen"
06:52:15 Quit Acksaw (Connection timed out)
06:53:56 Quit wpyh (Read error: 110 (Connection timed out))
07:34:01vitjahey could iaudio7 have 5.6 V when charged?
07:39:44 Join Darksair [0] (n=user@
07:49:18lucentergg... why doesn't the SD command to send CID work
07:53:01 Part zoyd
08:02:39 Join ender` [0] (
08:03:01 Join bertrik [0] (
08:04:52lucenthey um is ata_sd_as3525.c going unused?
08:06:07lucentmaybe I meant to ask about ata-sd-pp.c
08:13:59 Join Rob2222 [0] (
08:14:48 Quit Darksair ("Use the Force, Luke!")
08:19:39 Join Darksair [0] (n=user@
08:21:12 Join wpyh [0] (n=william@
08:30:06 Quit Darksair (Read error: 54 (Connection reset by peer))
08:32:29 Quit Rob2223 (Read error: 110 (Connection timed out))
08:39:35***Saving seen data "./dancer.seen"
08:58:10 Join n1s [0] (n=nils@rockbox/developer/n1s)
08:58:38 Nick fxb__ is now known as fxb (
09:05:25 Join B4gder [241] (n=daniel@rockbox/developer/bagder)
09:07:26 Join petur [50] (n=petur@rockbox/developer/petur)
09:09:25 Quit bertrik (Read error: 104 (Connection reset by peer))
09:11:18 Nick pixelma_ is now known as pixelma (n=pixelma@rockbox/staff/pixelma)
09:12:30 Join kugel [0] (n=chatzill@unaffiliated/kugel)
09:14:29 Quit wpyh ("Leaving.")
09:14:36 Join Zagor [0] (n=bjorn@rockbox/developer/Zagor)
09:20:28 Join ameyer [0] (
09:20:32 Quit BHSPitLappy (Remote closed the connection)
09:25:43 Quit jhulst (Read error: 60 (Operation timed out))
09:26:12kugellucent: what was this talk with wpyh? "hc" is definitely not the issue
09:27:33kugellucent: did you get your card to work?
09:28:47*kugel needs to go again
09:28:49 Quit kugel ("ChatZilla 0.9.84 [Firefox 3.0.4/2008111318]")
09:37:01 Join lasser [0] (
09:45:15 Quit kachna|lappy (Read error: 113 (No route to host))
09:45:25 Join Thundercloud [0] (
09:58:20 Quit alexbobp (Read error: 54 (Connection reset by peer))
10:00:09 Join alexbobp [0] (n=alex@
10:07:48 Join LinusN [0] (n=linus@rockbox/developer/LinusN)
10:27:07 Quit Thundercloud (Read error: 104 (Connection reset by peer))
10:39:38***Saving seen data "./dancer.seen"
10:46:55 Quit ameyer (Read error: 110 (Connection timed out))
10:47:17 Quit petur (Read error: 104 (Connection reset by peer))
10:48:30 Join petur [50] (n=petur@rockbox/developer/petur)
10:58:37 Join kachna|lappy [0] (
11:01:54 Join mofux [0] (
11:37:25 Quit kharo (Read error: 60 (Operation timed out))
11:38:05 Join kharo [0] (
11:40:08Unhelpfulamiconn: FS #9458 has the improved greyscale work up if you're interested. no greylib support yet, but very read to add it, just a matter of a couple extra things to pass to image loaders.
11:49:43 Join dany_21a_ [0] (
11:51:23dany_21a_hi lucent, just read the log regarding the 1GB limit of the sansa AMS...
11:52:22dany_21a_did some tests resently, and i found that the sector it fails its always the same.. and its nearly less then 1GB (even including the offset calculated in the driver)
11:52:52 Join Nibbler [0] (
11:53:03dany_21a_but I was able to read up to 2GB on my 2GB external MicroSD
11:54:21 Join Darksair [0] (n=user@
11:59:48dany_21a_on my 8GB fuze it fails while reading the sector 0x1DAE00 - which is the a bit beneath the first 1GB: (0x1dae00+61440)*512/(1024*1024*1024)=0,956787109375
12:03:55 Nick fxb is now known as fxb__ (
12:12:40 Join moos [0] (i=moos@rockbox/staff/moos)
12:15:43dany_21a_to test where it fails i used following quick hack... also good to test the driver, if someone wants to debug it:
12:21:54 Join robin0800 [0] (
12:22:40 Quit B4gder (Remote closed the connection)
12:23:02 Join B4gder [241] (n=daniel@rockbox/developer/bagder)
12:26:13 Part dany_21a_
12:34:53 Quit Chronon ("User pushed the X - because it's Xtra, baby")
12:39:39***Saving seen data "./dancer.seen"
12:53:23 Quit Nibbler (Read error: 110 (Connection timed out))
12:57:58 Join PaulJam [0] (
12:58:04 Quit blithe (Read error: 60 (Operation timed out))
13:01:08 Join bodymind [0] (n=bodymind@
13:02:08 Quit plus_M (Remote closed the connection)
13:10:19 Join scorche|1h [0] (
13:10:28 Join blithe [0] (
13:12:13 Quit scorche|sh (Read error: 104 (Connection reset by peer))
13:14:18 Nick scorche|1h is now known as scorche|sh (
13:14:50stsquadAre the MENU_* macros all meant to be used in the global context?
13:15:18 Join plus_M [0] (
13:23:31 Join kugel [0] (n=chatzill@unaffiliated/kugel)
13:25:01 Quit AndyI (Read error: 145 (Connection timed out))
13:25:11bodymindhei, i just get an error on my rockbox for sansa e260v2... i compiled last svn bootloader... i get an error: "No partition found"
13:25:18 Join AndyI [0] (i=AndyI@
13:25:35_Auron_v2 isn't compatible yet I believe?
13:26:19bodymindin here it says it's on develop... but the partition should be detected no?
13:27:49moosB4gder and others: Is the freeze suposed to end today, and then branch etc....?
13:29:38 Quit daurnimator ("Cyas later...")
13:31:43 Quit kugel (Nick collision from services.)
13:31:46 Join kugel_ [0] (
13:31:53 Nick kugel_ is now known as kugel (
13:33:09moosthe freeze is suposed to be 1 week, right?
13:33:35 Join daurnimator [0] (
13:34:00Zagormoos: yes
13:34:31moosZagor: then the freeze is over? :)
13:34:53 Join daurn [0] (n=daurnima@unaffiliated/daurnimator)
13:35:08ZagorI'd say the freeze is over when the branch is made
13:35:53moosIs anyone still working on any bugfix that we need in this release?
13:36:36ZagorI don't know. I'm preparing a commit of new default settings as discussed in the mailing list.
13:37:24moosNice, with all those posts, missed some of them :) One consensus was reached?
13:37:48Zagorit seems most people feel 15 seconds is good
13:38:04moos+1 here
13:38:37kugelZagor: 15s is a good compromise I think
13:38:39amiconn15 sec for colour targets, imo. For mono/grey, 5 sec is fine.
13:40:14Zagoramiconn: I think 5 is always too short for a first-time user
13:40:44B4gderon the mono/grey targets, the screen does remain readable though
13:41:04scorche_Auron_: you have a question?
13:41:21_Auron_wondering what you mean by delay; delay for what?
13:41:22ZagorB4gder: if you are in bright ambient light, yes. I still think 15 is a better default value.
13:41:29scorche_Auron_: read the mailing list...
13:41:39B4gderprobably, I don't really have any opinion in this matter
13:41:49 Join fredddy [0] (
13:42:14ZagorI want to the default to focus on being "friendly", not necessarily runtime-optimised
13:43:12kugelZagor: I'm wondering about gather runtime data
13:43:32kugelI've heard many users being confused that rating and stuff doesn't work
13:43:42moosZagor: user-friendly isn't our first advantage ;)
13:43:51 Quit fredddy (Read error: 54 (Connection reset by peer))
13:43:57pixelmagather runtime data costs some RAM - if enabled it should only be on high mem targets
13:44:02*n1s agrees with Zagor
13:44:08pixelmaor maybe I'm confusing this now
13:44:28pixelmaenabled by default that is
13:44:35 Join fredddy [0] (
13:44:44_Auron_ah.. backlight timeout.
13:45:42ZagorI know nothing about ratings. I never tried it.
13:46:11_Auron_ratings seem silly to me
13:46:18*_Auron_ shrugs
13:46:46 Quit fredddy (Read error: 104 (Connection reset by peer))
13:47:38Zagorkugel: what do you mean by "trackgain on suffle"? replaygain_type = REPLAYGAIN_SHUFFLE?
13:48:15pixelmaseems "gather runtime data" is not as RAM costly, mixed it up with log
13:48:33Zagor"noclip off" doesn't feel very friendly to me
13:49:07Zagoror does it have implications I don't know about?
13:49:35 Join fredddy [0] (
13:50:12LloreanZagor: I agree that 5 is too short for a first time user. I think 15 is on the very edge of short enough, but doable.
13:50:31Zagor15 it is then
13:50:46LloreanFor replaygain, I'd have to say the option that results in Album gain if shuffle is off, Track gain if shuffle is on is the behaviour almost anyone using the feature will expect.
13:50:51_Auron_15 sounds goods.. 5 seconds is a bit brief.
13:51:02pixelmaI've set it myself to 15s on M5 and c200, 5s on my Ondio
13:51:38n1sI still don't think RG by default is very nice, it's altering sound and can be pretty annoying if you hav not RG'ed your tracks yourself
13:52:12Lloreann1s: It usually doesn't introduce any more variation than the normal mastering process has over the years.
13:52:25n1sLlorean: i disagree
13:53:39Zagorit's a touch call I'd say. RG is a nice feature we want people to get automatically. but not if we think there are significant side-effects for casual users.
13:53:48Zagortough call
13:53:52Lloreann1s: I have music in my collection that Replaygain has increased the gain of, and decreased the gain of. Mathematically, at least for me, if I left in the widest ranging songs no Replaygained song would be further from the furthest songs than they would be from each other.
13:54:08LloreanI don't see how there could be "serious" side effects.
13:54:23LloreanAnyone who listens to non-RGed music from different time periods is going to be used to volume jumps between albums already.
13:54:39ZagorLlorean: yes, that is my thinking aswell
13:54:58pixelmabacklight is quite a battery waster there. Did some benchmarks back then 6:45h (backlight on) vs. 8:45h (backlight off), that's probably not really comparable with the LED backlight + lithium batteries most of the other targets use
13:55:16n1swell, yes the only side effect is volume jumps, it's not terribly serious i just don't think it's a nice default
13:55:50Lloreann1s: I think with a truly mixed collection, the volume jumps are bigger anyway.
13:56:15LloreanI have +11 db and -9db on albums, which suggests they're 20db apart without Replaygain.
13:56:35ZagorI think we should try it and see if we get any comments about it
13:56:43LloreanI think people who don't know what it is and have it on are just going to assume the usual, that their albums are mastered at different levels.
13:57:03n1sit also depends on listening habits, i.e. do you make a playlist of the haunted and enya without RG it will differ a lot but mixing Lamb of god with the haunted will not
13:58:25n1sbut sure, i dont' feel terribly strongly about it
13:59:27Zagorany strong opinions about "backlight off on button hold"? I'm split myself.
13:59:41LloreanI'm personally in the crowd that uses it.
13:59:48*kugel too
13:59:52LloreanI'm not really sure it needs a default though.
14:00:06 Quit fredddy (Remote closed the connection)
14:00:14LloreanI don't really see any sense in leaving it on while on hold though
14:00:19kugelmost OFs do this too (i.e. it won't be some unexpected behavior)
14:00:25LloreanYou need user interaction to light up the screen, and they can turn hold on AFTER they're done looking.
14:00:47pixelmaI don't use it and there are some themes which show different info on hold (not sure if one of the SVN ones does, I doubt it)
14:00:55 Join fredddy [0] (
14:01:15LloreanBasically, I think most users are either going to want hold to turn off the light, or to turn it permanently on (for use say, in a dock where volume is controlled some other way, and they just want the lit screen to be visible from a distanct)
14:01:27 Quit fredddy (Read error: 54 (Connection reset by peer))
14:01:29LloreanAnd "on" definitely isn't a good default.
14:01:35ZagorI'll set it to "off during hold"
14:03:51Zagorwhat was the idea behind a separate buttonlight timer anyway?
14:04:20 Quit jhMikeS (Nick collision from services.)
14:04:26 Join jhMikeS [50] (n=jethead7@rockbox/developer/jhMikeS)
14:04:37pixelmaI like it
14:04:47n1sthat we wanted more settings?
14:05:13amiconnpixelma: There are targets where backlight is *really* wasteful, e.g. ipod mini. Runtime is cut to about 1/3 when running with backlight always on
14:05:27LloreanZagor: Well at least it lets you turn them completely off.
14:05:39pixelmausually I switch the buttonlight on c200 off (or set it to a 1s) because the blue LEDs are very bright and so close to the screen
14:07:07kugelZagor: hm, yea, I'm not too sure why there's seperate *timer* too, but I definitely see use of a different option; maybe only "off, on, same as backlight"?
14:07:38Zagorkugel: yes I think that would be better
14:07:52pixelmathat sounds more complicated than now
14:07:58Zagor(but I'm naturally not going to change anything now)
14:08:10kugelbut that's probably rather a post-freeze one, so we'd still decide on a default timeout value, I suppose?
14:08:14 Join tyfoo [0] (
14:08:19 Join maffe [0] (
14:08:30Zagorkugel: yes, I'm setting it 15 like the backlight
14:08:31kugelpixelma: more complicated?
14:08:35LloreanI think if there's always going to be an option, it may as well have its own timer.
14:08:43 Quit maffe (Client Quit)
14:09:12ZagorLlorean: possibly with the added option "as backlight"
14:09:26 Join XavierGr [0] (n=xavier@rockbox/staff/XavierGr)
14:09:27pixelmayes, new options instead of "reusing" the existing ones
14:09:30kugelLlorean: hm, maybe, but then at least add the "same as backlight"?
14:09:35LloreanZagor: Well, I can see some people wanting the backlight permanently on while the buttonlight to help clarify registered keypresses.
14:10:04ZagorLlorean: yes, I just meant an addition option in the setting: off, on, as backlight, 1,2,3,4,5,6...
14:10:12LloreanAh, yes.
14:10:18LloreanThat makes more sense. :)
14:11:23kugelLlorean: yea, now that you mention it, I'm usually also one of the guys having backlight always on, and then using the hold as a light-switch
14:11:48*kugel haven't used that for a while though, that's why he forgot about it :)
14:12:04ZagorI'm inclined to agree with kugel about "show path in browser". I think it qualifies as a "newbie friendly" option.
14:12:16_Auron_I just set it to something like 5s and let it turn off when I put it back in my pocket
14:13:18stsquadAre there any text-mode simulator builds? Running SDL over SSH over X is a bit slow for testing
14:13:22LloreanZagor: I second that.
14:13:52B4gderstsquad: mount the fs with sshfs instead and run it locally!
14:14:36 Join aneqrs [0] (
14:17:39Zagorwhat is the negative impact of enabling "gather runtime data"?
14:17:41 Quit DataGhost (Nick collision from services.)
14:17:49 Join DataGhost [0] (i=dataghos@unaffiliated/dataghost)
14:20:02 Join fredddy [0] (
14:20:53pixelmaI would hope that it's fixed but there was a time when enabling this option caused disk spinups at track change, can't stress this enough: I'm not sure that's valid anymore
14:21:01kugelZagor: I'm not sure, I think Slasheri once said if you enabled it, enabling load to ram is recommended
14:21:16PaulJampixelma: i think this was fixed
14:22:07kugelZagor: even though I don't know why. All data which is gathered is saved in the master index file of the database, and that one is always loaded into ram (no matter of which node you browse)
14:22:37pixelmakugel: not that I know of because it seems to work correctly on my Ondio (which has no "load database to RAM")
14:22:53kugelpixelma: I said recommended, not required :)
14:23:22kugelpixelma: but then I even wonder more about the costs
14:24:13*kugel mixes the word order of his previous sentence a bit so that it sounds correct
14:25:52kugelZagor, pixelma: Maybe it was just formed to an option since it implies writing to the database files after initializing
14:26:16kugelfile rather, only the master index
14:29:48 Quit kugel ("ChatZilla 0.9.84 [Firefox 3.0.4/2008111318]")
14:30:06 Join kugel [0] (n=chatzill@unaffiliated/kugel)
14:33:12 Join pondlife [50] (n=Steve@rockbox/developer/pondlife)
14:34:21pondlifeZagor: Are we planning on having Show Path set to "Current Directory Only" ? That would be my choice...
14:34:42Zagorpondlife: yes, that's my choice too
14:35:16*pondlife is happy he'll have less options to reset on a clean install
14:35:51 Join wpyh [0] (n=william@
14:39:40***Saving seen data "./dancer.seen"
14:40:52kugelfor me it zero's out due to the warning and backlight ones
14:41:04 Join mib_p99fcqq7 [0] (i=d8ef4481@gateway/web/ajax/
14:41:13 Nick mib_p99fcqq7 is now known as MarcGuay (i=d8ef4481@gateway/web/ajax/
14:41:18B4gderI copy the options when I upgrade ;-)
14:41:36*Llorean keeps a .cfg file around he copies to new players.
14:46:18 Join Aurix_Lexico [0] (
14:48:04*amiconn has quite a bunch of non-standard settings
14:49:15 Quit linuxstb (Read error: 110 (Connection timed out))
14:49:37 Join LambdaCalculus37 [0] (i=44a04303@rockbox/staff/LambdaCalculus37)
14:51:20*JdGordon is a bit late....
14:51:25JdGordoncan we discuss removing show files from the QS?
14:51:49JdGordonthere has been quite a few problems because users changed that accidently
14:52:40MarcGuayJdGordon: Any replacement candidates in mind?
14:52:41JdGordonor am I too late?
14:52:52ZagorI don't even know how to access the QS on my c200 and clip :-)
14:52:58MarcGuayI didn't hear a bell ring.
14:53:02Zagorso I don't really have an opinion
14:53:17JdGordonI dont have any real suggestions for a replacement.. just that it should go
14:54:22kugelJdGordon: the problem with accidentely change those is probably that on some playes pressing the up butto also changes the bottom item
14:54:36pixelmaZagor: IIRC on the c200 the quickscreen is only half present - accessible only in the WPS (on others you can get it from the menus too)
14:54:46kugeland users don't expect that, but rather e.g. that they can return to wps
14:54:48LloreanJdGordon: Is Equalizer off/on an option?
14:54:51kugel(that's the case on e200)
14:54:57MarcGuaykugel: That's a good point.
14:54:58JdGordonLlorean: yeah should be
14:55:05 Quit Darksair ("Emacs = ESC-Meta-Alt-Ctrl-Shift")
14:55:12LloreanYou'd still have to set it manually, but it's the sort of thing OFs like to put close to hand.
14:55:22kugelchanging quickscreen just moves the problem to another setting
14:55:28pixelmanot on hwcodec
14:55:32JdGordonnot if its a toggle setting
14:55:53LloreanEspecially not if it's a toggle setting like EQ that doesn't do anything if you haven't set it up elsewhere first. :)
14:56:06JdGordonsw/hwcodec dont *have* to have the same settings on the screen
14:56:53JdGordonpixelma: we can have nothing by default on the hwcodecs.. which might be nice anyway with their small screens
14:57:13kugelLlorean: if eq is to be picked i'd rather pick crossfeed or crossfade. those actually change behavior without needing to visit the respective settings menu
14:57:28Lloreankugel: They're also mostly set-once options.
14:57:42JdGordonwell. some might say eq is also
14:57:48JdGordoneq _should_ be set and forget
14:57:49kugeluhm, not crossfeed imho
14:57:51pixelmaJdGordon: the lower option is not a real problem screen estate wise
14:57:59LloreanYeah, but more people argue EQ is not a set-once option than argue, for example, crossfade is.
14:58:13MarcGuayAuto-Change Dir Yes/No/Random is relatively harmless if changed... Unless Random *requires* a setup with the plugin, can't remember.
14:58:23JdGordonit does
14:59:14kugelI change crossfeed rather frequently (more than eq), e.g. I turn it off when connected to an audio systems
14:59:19*JdGordon would really like a .cfg browser on the QS but couldnt do it without major hacking
14:59:31kugelalso it doesn't fit for some genres
14:59:36Lloreankugel: But the average user doesn't even know what Crossfeed is.
14:59:43LloreanThis isn't for *your* habits.
15:00:00JdGordonwhat about fade on pause?
15:00:08kugelLlorean: I'm just arguing that crossfeed isn't generally a set once option
15:00:13markunmaybe we should rename crossfeed to SpeakerSim(TM)
15:00:15LloreanJdGordon: Probably not something that is changed often.
15:00:16MarcGuayJdGordon: Sounds good.
15:00:16JdGordonor skip length?
15:00:21Lloreankugel: It is for anyone who doesn't hook it up to multiple things.
15:00:37MarcGuaySkip length could also lead to unnecessary support questions.
15:00:48MarcGuaySo could breathing, though.
15:00:50JdGordonor recursivly insert directories?
15:00:59JdGordonwhich could actually be sueful there
15:01:02LloreanJdGordon: Skip length is useful there, but likely to cause the same problems as file view.
15:01:13LloreanRecursively insert isn't something you often toggle, is it?
15:01:15markunor "ROCK WOW"
15:01:22JdGordonmove shuffle to the bottom and skip length to the left?
15:01:24kugelanother note on skip length: Am I the only one that finds it annoying that you cannot skip the track if the remaining time is less than the skip length? Is that even intended?
15:01:52JdGordonkugel: that sounds like a bug
15:01:55JdGordonhint hint :D
15:02:02LloreanJdGordon: I'd put repeat on the bottom if I were going to do that.
15:02:13LloreanJdGordon: I think the best fix, though, is to make "Up" leave the screen, rather than also cycling the bottom option.
15:02:26LloreanI think for about 2/3 of people who accidentally change it, it's because they hit Up and didn't realize it also changed it.
15:02:27kugelLlorean: I agree
15:02:32JdGordonwell, shuffle is toggle so it fixes kugel's problem before about having up and down do the same
15:02:36kugelthat's causing problems on any setting
15:03:07LloreanJdGordon: The layout of the screen suggests that only three directional buttons will do anything. So, it's probably true that it should actually work that way.
15:03:18JdGordonok, so UP should exit the screen and put skip lenght instead of show files?
15:03:38LloreanI think it's safe to leave "show files" at that point.
15:03:44pixelmawasn't there some technical (code) reason for this?
15:03:48kugelJdGordon: it should do whatever the up buttom does imho (and leave if it's not applicable)
15:04:03kugelpixelma: no, it's a pure keymap problem
15:04:03LloreanI think skip length is actually less useful, overall, than show files.
15:04:22kugelpixelma: or do you mean skip length?
15:04:29JdGordonskip length probably has too many possible values to be usable
15:05:02LloreanMaybe something like "Car adapter mode"
15:05:15pixelmaI meant "up" also changing... I believe it had something to do with flipping the screen and some button actions... if I could only remember...
15:05:19LloreanSomething you don't set too often, but you also don't want to spend three minutes browsing for the few times you want to change it.
15:05:32*JdGordon thinks it should be "rand()%nb_settings"
15:05:53MarcGuayQuick, indeed.
15:05:55JdGordonpixelma: thats possible, but shouldnt be a problem since actions
15:06:19JdGordonactualy.. no that doesnt sound right seen as the screen flipping for the buttons happens way down in the button driver
15:06:51*JdGordon is happy with car adaptor mode
15:06:55 Join DrMoos [0] (
15:06:58kugel this little diff should fix most quickscreen problems
15:07:11kugelexample for e200, should be done on any keymap
15:07:43JdGordonno, we should keep the keymap because I would like to eventually add a top item
15:07:48 Quit moos (Read error: 104 (Connection reset by peer))
15:07:49 Nick DrMoos is now known as moos (
15:08:01kugelpixelma: hm, flippen screen, indeed a problem. good call
15:08:32kugelJdGordon: "eventually", revert it when you do so, as of now it's only causing problems
15:08:37*MarcGuay votes for Bass or Treble as the bottom option and goes to make coffee...
15:09:08JdGordonkugel: there is an easier fix... comment out the QS_DOWNINVERSE case in quickscreen.c
15:09:30JdGordonMarcGuay: again.. too many possible values with no way of going "backwards"
15:09:46LloreanMarcGuay: The problem with one of those, is that invariably it should be paired with the other. I thought of that, but I think there's again, too many options and hard to pick between (though I'd lean that more would want bass)
15:09:52kugelJdGordon: but that won't make the top button do anything. I think it's better to do the top button do what it always does (e.g. showing the wps on e200)
15:09:53 Quit ch4os ("leaving")
15:10:09JdGordonok, so add a return GO_TO_PREVIOUS under that case
15:10:17Lloreankugel: I think *any* button except one of the three directions should simply return to the previous screen.
15:10:43LloreanThe quickscreen doesn't work like other screens. In fact, ideally it was supposed to vanish on release of the quickscreen button on targets where you could do button combos.
15:11:03 Join ch4os [0] (n=ch4os@gentoo/user/ch4os)
15:11:04kugelthat's an option too
15:11:44pixelmabut didn't work since button actions
15:11:57 Join Schmogel [0] (
15:12:04*JdGordon has always said that could be fixable on the targets which it was possible anyway
15:16:30 Quit bxcracer_ (Read error: 60 (Operation timed out))
15:17:06bodymindhei, i get an error on my rockbox for sansa e260v2... i compiled last svn bootloader... i get the error: "No partition found" on top of the rockbox image boot. the SD driver is not completely done? where can i start, to make it work? :) tkx...
15:18:27 Quit parafin ("So long and thanks for all the fish")
15:19:47kugelZagor: so we don't enable gather runtime data because we're not entirely aware of the costs?
15:20:03pondlifeZagor: Now I read some logs, I see that the dynamic playlist destruction warning is to be enabled by default? If so, are you sure about that one?
15:20:28kugelpondlife: it was generally agreed on on the ml
15:20:36 Quit kachna|lappy (Read error: 110 (Connection timed out))
15:20:42pondlifekugel: Not by me...!
15:20:46Zagorkugel: I felt we needed a bit more discussion/research anyway. I just committed what I felt was non-controversial.
15:20:50JdGordonso disable it...
15:20:59LloreanShouldn't gather runtime data do nothing if your database isn't enabled?
15:21:01pondlifeI actually use it
15:21:08pondlifeBut I don't think it's n00b-friendly
15:21:17JdGordonhaving warn on playlist is n00b friendly
15:21:24Lloreanpondlife: It's n00b-friendlier than allowing their playlists to silently vanish into nothing
15:21:29kugelpondlife: not by me either, but we're the only ones :) It was definitely more who voted in favor of it
15:21:43*Llorean doesn't see how it's really unfriendly at all.
15:21:58pondlifeIt's unfriendly if you're not expecting it to stop and pop up
15:22:00LloreanYou can turn it off after the first time you see it, and it showing up even once makes you aware Rockbox won't automatically save changes.
15:22:15kugelwell, you select a file and you don't get the music instantly, that's not friendly imho
15:22:24pondlifeMy point exactly
15:22:32LloreanA few seconds lost is MUCH better than 20 minutes of playlist creation lost.
15:22:44pondlifeThat screen has single-handedly made mrs pondlife give up on Rockbox
15:22:46LloreanOn the balance of unfriendliness, I'd say it's much in favor of showing the option at least once.
15:22:55kugelbut silently destroying the current playlist isn't so friendly either, it's a tough decision
15:23:21pondlifeShe didn't want to have to understand the playlist stuff, just to play a track
15:24:02LloreanDoes it display if you've just clicked on a file in a folder, then click on another file in another folder without ever using "Insert"?
15:24:03pondlifeI agree I might be being OTT, but I don't think it's a friendly thing for non-technical users
15:24:35pondlifeLlorean: I don't think it's meant to, but it might be there's a bug
15:24:41Lloreanpondlife: How is a nuisance that you can disable *more* friendly than losing an entire playlist because you expected it to silently save for you?
15:24:44kugelwell, it's about are we expecting playlist oriented user or user who simply want to play music
15:24:51pixelmadoes someone know off the top of their heads what the current default for "max files in directory" is?
15:24:54LloreanEr, *more* unfriendly
15:24:56bodymindhei? :(
15:25:15Lloreanbodymind: Read the source, track down bugs, fix bugs?
15:25:22pondlifeLlorean: Don't worry, I get your point, and agree
15:25:24Lloreanpixelma: I think it may be 1000?
15:25:36Lloreanpondlife: I do think it could be improved by "if you've never used the Insert options, don't prompt"
15:25:37LambdaCalculus37bodymind: Or wait until the port is ready, whenever that is?
15:25:49pondlifeI'm just passing on some non-tech-user input
15:25:53kugelsadly my proposal to "yes and don't show again" wasn't received well :/
15:26:00bodymindi'm a programmer i'm really on to helping...
15:26:05LloreanBasically, queued files are supposed to vanish anyway, and if all the files are from one folder (you pressed 'select') there's no reason to prompt, because no changes have been made.
15:26:07 Quit bodymind ("Ex-Chat")
15:26:40LloreanBut that's more a "needed fix" if it doesn't do it already, than a reason to make it trivial to lose playlists.
15:27:28MarcGuayLlorean: "Does it display even if you don't modify the playlist?" No.
15:27:33pondlifeThe wording might be the problem, maybe reword it as "You haven't saved your playlist. Press SELECT to save, OFF to continue and lose the playlist."?
15:27:44kugelLlorean: how about flagging the dynamic playlist as "virtual" playlist, and once you insert something the flag is removed
15:28:04MarcGuayI think it's safe to say that if you've gotten into the playlist menu and inserted tracks or whatnot, you're capable of changing the setting to meet your needs.
15:28:15LloreanMarcGuay: I know it doesn't if you don't modify m3u files. But if you click "Select" on a file in a folder, then do the same in another folder replacing the dynamic playlist, does it?
15:28:23JdGordonspeaking of playlists... pondlife, the wiki is back.. wanna make a start on playlist/playback/buffering interaction?
15:28:33MarcGuayLlorean: IIUC, no.
15:28:33pondlifeJdGordon: Next year, yes!
15:28:41 Nick fxb__ is now known as fxb (
15:28:54JdGordonbshanah habaah..!
15:29:14JdGordonnext year.. in the wiki! (in hebrew)
15:29:16LloreanMarcGuay: I think it's probably sane enough then.
15:29:20pondlifeI haven't started Christmas planning yet
15:29:28pondlifeBut that's OT
15:29:30 Quit mofux (Remote closed the connection)
15:29:50kugelMarcGuay: I'd count this as an argument against the warning
15:30:08MarcGuayLlorean: I still think the options/wording could use a rework.
15:30:38MarcGuaykugel: Sounds like it, doesn't it...
15:31:05LloreanCount what as an argument against the warning?
15:31:20kugelI just think having this warning will just frustrate new users
15:31:45kugelsince I think they're not going to do insane playlisting stuff on the first boot, but just play music for some time
15:31:52LloreanIt should only pop up when they'll actually LOSE something by ignoring it.
15:31:52pondlifeIt's mainly that it's badly worded. the word "dynamic" is probably best avoided...
15:32:13LloreanIt shouldn't pop up unless they actually *DO* playlist stuff.
15:32:19LloreanIn which case, it's a good warning.
15:32:30kugelonce they do playlisting (which I consider as rather advanced) they'll have get known to the settings as well
15:32:34LloreanThe only time it's a problem is if it pops up when there's nothing to lose (playing a folder without changes, etc) in which case, that should be fixed.
15:32:55pondlifeHeh, I managed to get it to display
15:32:57MarcGuayLlorean: My tests show that isn't a problem.
15:33:02pondlifeIt is :/
15:33:17Lloreanpondlife: Did you make use of the "insert" option, or just click a file?
15:33:42pondlife1) Play an album from the database
15:33:49pondlife2) Play a second album
15:33:57pondlifeWarning, but no inserts
15:34:04kugelLlorean: I agree, but it's post-release stuff, we have to decide based on what rockbox does now
15:34:20MarcGuayAh, database. Doesn't occur in file browser, same actions.
15:34:20Lloreanpondlife: Then that should be fixed, but I don't see it as a reason to disable this user-protection until then.
15:34:29pondlifeLlorean: I agree
15:34:37LloreanI do agree it could be reworded.
15:34:37pondlifeI'll log the bug
15:34:49Llorean"Your playlist is unsaved" is much easier to understand.
15:35:12 Join kushal_12_27_200 [0] (
15:35:16MarcGuaykugel: Have you noticed more screen flickering on the e200 since backlight fading?
15:35:26kugelif we're optimising the defaults for new users, I still think, new users in particular don't benefit from the warning, since they don't fear about playlists for the first days of using
15:35:38kugelMarcGuay: not at all
15:35:55MarcGuayMaybe it's the difference between "almost-faded-out" and "on"...
15:36:19kugelwhat flickering? While fading?
15:36:22Lloreankugel: A lot of people install Rockbox specifically for the dynamic playlisting features their original firmware didn't offer.
15:36:27 Quit kushal_12_27_200 (Remote closed the connection)
15:36:44 Join kushal_12_27_200 [0] (
15:37:05kugelI don't think that's the majority
15:37:14MarcGuaykugel: I think it's it you do an action right before the backlight goes off, it feels like it's flickering when it jumps back to full.
15:37:28kugelthe majority of new users are the ones who install it since they're curious or because they saw a cool theme
15:37:44kugelthat's my impression
15:37:46Lloreankugel: Well then they can disable the warning.
15:37:54LloreanAgain, it's better to lose a couple seconds of time, than several minutes of work.
15:38:26MarcGuaykugel: Actually, when the backlight springs on from off (no fading), it flickers. Could always have been that way.
15:39:03pixelmathat's probably the lcd enable
15:39:05LloreanThey don't know their playlist will be lost until they see that warning. If it's off, they don't know their playlist will be lost until it's actually lost.
15:39:08kugelcan you please exactly describe when you feel it flickering?
15:39:13pondlifeOK, logged the database problem -
15:39:26 Join Lynx_ [0] (
15:39:51MarcGuaykugel: When it turns on without fading. I'm starting to think it's always been like that.
15:40:05kugelMarcGuay: with fading disabled?
15:40:19kugelit flashes, that's due to lcd_enable, and it was always like that
15:40:27MarcGuaykugel: Yes.
15:41:18 Join Jaykay [0] (
15:41:22kugelthat could be fixed with a small delay, like I've done here
15:42:13kugelbut then again, given that you don't have with with fading enabled, it would just (slightly though) delay the turning
15:42:33kugel"given that you have fading enabled"
15:43:49JdGordonpondlife: are you sure thats the expected behaviour?
15:43:53MarcGuaypondlife: I'm glad we narrowed that down. I was starting to wonder if I was losing my mind.
15:44:22JdGordonhmm... reads a bit of the log I missed
15:44:33pondlifeJdGordon: It is to me... and I can't see why Database selections and file selections should be different
15:44:51pondlifeBoth browsers are just playlist generators
15:45:15JdGordonSlasheri: please please please try to put some time into the database playlist handling... having each file added seperatly causes so many different issues
15:45:17MarcGuaypondlife: If you selected "All tracks" and got a selection of several albums... Or selected tracks from 1976... Would you expect the warning?
15:45:44pondlifeOnly if I was about to lose manual playlist changes
15:45:50MarcGuayBecause it's passive playlist creation...?
15:45:56MarcGuayMakes sense.
15:45:58pondlifeSo is file browser selection
15:46:43pondlifeI mean, you could add a file to a folder, then reselect a track in that folder... the playlist changes, but you wouldn't want a warning
15:46:45LloreanI think you shouldn't get the warning until you've made use of "Insert"
15:46:51pondlifeI agree
15:46:52LloreanOr any of the "Insert *" functions
15:47:05MarcGuayThumbs up from me.
15:47:23MarcGuayYou need to prove that you know what a playlist is before being asked questions about it.
15:47:35 Quit kushal_12_27_200 ("Leaving")
15:48:12LloreanWell, if you haven't used "insert", in theory your old playlist can be recreated with a single keypress of "select" on the right screen, so it's more or less the same as having a .m3u files somewhere to click on.
15:48:24 Quit rasher (Read error: 110 (Connection timed out))
15:48:27MarcGuaySo where does the open bug leave the default setting...?
15:48:46pondlifeFix the bug ;)
15:49:00pondlifeIf it's unfixable, please don't enable that by default
15:49:22LloreanI'm not sure we should disable it because database doesn't work right with playlists yet.
15:49:25pondlifeNon-tech-users will likely prefer the database in my experience
15:49:27LloreanIt also doesn't work right with bookmarks, etc.
15:49:38LloreanDatabase is, to an extent, broken for several features now it seems.
15:49:48pondlifeOnly bookmarks, really
15:49:54LloreanAnd this one.
15:49:59MarcGuayI think that's fair. Or come up with nicer options/wording for the warning before enabling it by default.
15:50:05JdGordonbookmarks and db never played nicley together
15:50:06LloreanIt definitely needs a nicer wording.
15:50:25JdGordonwell, not since I've been here anyway
15:50:37MarcGuay"Save current playlist?"
15:51:09MarcGuaySince it's not true in the database that you've made changes, that would do.
15:51:29kugelJdGordon: want me to have a go at the quickscreen up button thing? I'd also fix a display issue (icon vp's are slightly wrong, but that issue is only visible when using a non-fullscreen parent)
15:52:13JdGordonfix the button how? we didnt come to a consensus
15:53:02LloreanMarcGuay: "Unsaved playlist" with options "replace" and "cancel"
15:53:34JdGordonLlorean: shouldnt that also have "save and restart"?
15:53:36kugelJdGordon: the consensus was that the up button is causing problems
15:53:37MarcGuayLlorean: No choice to resave ThugRapz1 as ThugRapz2?
15:53:41kugelor am I wrong?
15:53:54pondlifekugel: So we make UP do nothing, I guess...?
15:54:05pondlifeReserving it for a 4th option in the future (where possible)
15:54:19kugelno, I'd rather say make every other button exit to previous screen, as proposed by Llorean
15:55:09 Nick JdGordon is now known as JdGordon|zzz (n=jonno@rockbox/developer/JdGordon)
15:55:23JdGordon|zzzdo a patch... it will be looked at... probably
15:55:37LloreanMarcGuay: Those can be added later, but probably shouldn't be added for 3.1 since it's a new and untested feature to save the playlist from that option.
15:56:46amiconnJdGordon|zzz: "show files" is about the only option which I actually use in the quickscreen, so I'd vote against removing it. It's an option that makes a lot of sense to have quickly available
15:57:00pondlifeI'd also vote to keep Show Files
15:57:17pondlifeI didn't know it was considered for kicking though
15:57:20LloreanI think with the "up" issue resolved, show files will get triggered accidentally a lot less often.
15:58:34amiconn"up" isn't an issue, it's a requirement (on the recorders, and only for really-quick mode, which is *still* broken)
15:58:53Lloreanamiconn: "Up" changing the value that "Down" also changes?
16:01:08amiconnThis is because in flip mode, logical "down" becomes physical "up", and "up" and one of the F keys cannot be recognised simultaneously
16:01:33amiconnThis is also the reason why there is no "up" in the quickscreens
16:01:49LloreanSo it's necessary as a special case for one target in a mode that no longer works anyway?
16:01:54 Join bxcracer [0] (n=bxcracer@
16:01:58LloreanI don't see that as a reason not to fix it on every other target, though..
16:02:21 Join bodymind [0] (n=bodymind@
16:02:26*kugel agrees
16:02:38*pondlife notes that onplay.c add_to_playlist() has some strange logic... why should it be doing playlist_create()?
16:03:28amiconnLlorean: I didn't even know that "up" is mapped on non-recorders. Of course that's unnecessary
16:03:48Lloreanamiconn: It's mapped for most targets, it seems. Causing people to frequently change "File View" without being aware of it.
16:04:04LloreanIn many cases "Up" is a cancel button in other screens, so they try to use it to leave the quickscreen too.
16:04:14LloreanSince obviously "left" won't work.
16:05:14MarcGuayWe were talking a while ago about how the go-to-WPS button should be universal... This is a case where it falls outside that "rule".
16:06:17pondlifeMarcGuay: Whch target?
16:06:30*MarcGuay finally understands what Llorean meant by "replace and cancel"....
16:06:40pondlifeOn H300, PLAY should still go to the WPS from the QS, but it does nothing
16:06:45kugelpondlife: he was probably saying quickscreen
16:07:14LloreanI think the Go-To-WPS button should be universal from any *list*
16:07:21Lloreanbasically, anywhere where the keymapping is "normal"
16:07:40pondlifeAny screen where it's not doing anything else (plugins excepted), I reckon.
16:07:46kugelI don't feel very strong about it. As long as it is consistent
16:08:08pondlifeSTOP and PLAY should be as universal as possible, IMHO
16:08:20kugeland I think both "all other buttons exit from quickscreen" and "go-to-wps button always goes to wps, even in quickscreen" is consistent
16:08:51Lloreanpondlife: The idea with the quickscreen, originally, was that it was only visible while you held the button for it down.
16:09:08 Part LinusN
16:09:18LloreanI think "any button not used to change it, clears it" is closer to consistent with this behaviour since possibly one day it'll return, and you won't be able to use those shortcuts from it at that point anyway.
16:09:33pondlifeLlorean: That would be nice, but I guess we have restrictions on which multi-buttons we can detect
16:10:07pondlifei.e. on H300 we can't detect MENU+another button,.. IIRC
16:10:20pondlifeSo it always be a dedicated screen
16:10:24pondlifeQS, I mean
16:11:32LloreanWell, in that case it's still just two presses of play to get back to the WPS (if you called it from a non-WPS screen) anyway.
16:11:47pondlifeIndeed, but consistency is good.
16:12:16pondlifeIf operating blind, I sometimes pause it rather than playing (even/odd number of presses)
16:12:55pondlifeGiven that PLAY does nothing at all in there, it would be better if it went to the WPS
16:13:39pondlifeBut just a minor point
16:14:35MarcGuayLlorean: Regarding the wording of the warning, "replace" could be interpreted as "resave with the changes i've made" or "replace the old playlist with this new one". Perhaps "Continue without saving" and "Cancel" would be clearer options.
16:15:11LloreanMarcGuay: Or "Discard"?
16:15:29*Llorean likes short messages where possible.
16:16:03MarcGuayThat works, too.
16:16:21 Join kushal_12_27_200 [0] (
16:23:17wpyhkugel: I have read dany_21a_'s comments −− what does he mean by "the offset calculated in the driver"?
16:24:01kugelwpyh: offset to skip the OF firmware
16:24:21wpyhhm... where can I find that?
16:24:40kugelit's not on a seperate partition, but we don't want to show up the OF files. If I remember right, it's also the only way to get it working
16:25:17wpyhI mean, the value of the offset?
16:26:28 Join jgarvey [0] (
16:26:39kugelwpyh: in ata_sd_as3525.
16:27:15wpyhI've tried his code and got the error at the exact same position (0x1DAE00)
16:27:40kugelLlorean, pondlife: so what now? what to go for with quickscreen? global go-to-wps button or every button except the three directional ones exit it?
16:27:48wpyhso, you think this is not related to a high-capacity MMC card issue? I see that sdhc is false inside sd_init_card()
16:28:04kugelwpyh: it is
16:28:12pondlifekugel: Just fix one thing at a time.
16:28:14kugelthe problem is, that the internal memory is NOT hc
16:28:32pondlifeThe UP button should do nothing (on most targets) is one fix.
16:28:36kugelthey get more than 2GB using banks (like the v1 do)
16:28:50pondlifeThe PLAY button should go to WPS (on some targets) is another.
16:29:17wpyhhm... if it's not hc, could it possibly be multiple "cards"? but card 1 is already the external storage...
16:29:24Lloreanpondlife: not a 'fix' as much as a behaviour change.
16:29:25kugelpondlife: yes, and I'm asking if the play should go to the wps or exit [or do noting]
16:29:38kugelon the e200 e.g. the wps button is the up button
16:30:00pondlifeI think the former mod is less contentious, so just do that for now
16:30:06wpyhkugel: how can we see if it's HC or not? I was guessing that it's a HC MMC card as opposed to an SDHC card...
16:30:30pondlifeI can only speak for my own targets though
16:30:44kugelyou've just mentioned yourself
16:31:09pondlifeI know when I tried to use an e200 I found the keymap truly horrid
16:31:22pondlifeThis was months ago though
16:31:51wpyhkugel: forgive my ignorance, but is SDHC compatible with high-capacity MMC? I think they are different, but I'm not sure since I'm not familiar with hardware
16:32:01kugelI don't know
16:32:06kugelwhy do you think there's mmc inside?
16:32:24wpyhso I was guessing that the internal memory is high-capacity MMC, thus it doesn't identify as SDHC
16:32:40wpyhI saw the strings "mmc:0:\MUSIC\..." inside MTABLE.SYS
16:33:00Lloreanwpyh: it also says mmc:1 for the SDHC card though.
16:33:18wpyhyes, Llorean. therefore I'm not so sure now
16:33:19kugelwpyh: again, the v1 sansas have different banks (imagine it as raid'ed LC cards), and it's just logical that the ams sansas do it also
16:33:24 Quit Jaykay (Read error: 110 (Connection timed out))
16:33:30wpyhhm.. ok
16:33:52Lloreanpondlife: Was this before or after my keymap changes that you tried the e200? As in, is "months" sometime this year, or sometime in '07?
16:34:01kugelwpyh: and the same code works for external SD(HC) cards
16:34:11pondlifeLlorean: This year
16:34:24pondlifeThe main problem was the use of UP to exit seemed odd
16:34:49pondlifeI'm used to UP/DOWN/LEFT/RIGHT/PLAY/STOP/MENU
16:34:54kugelLlorean: btw, I wonder if it's intended that the power button only stops playback if you're in the root dir in the filebrowser
16:34:57pondlifewhere STOP = exit, normally
16:35:06wpyhok, so they should be similar then
16:35:11kugelif you're in a folder it just goes upwards, until you reached the root
16:35:20MarcGuaypondlife: That doesn't sound familiar on the e200. (up = exit)
16:35:30pondlifeAh, sorrry - c200
16:35:44Lloreanpondlife: Oh, that's very different.
16:35:47MarcGuayAhh. Big diff.
16:35:55Lloreankugel: Yes, that sounds like a bug in the keymapping.
16:36:00MarcGuayThe c200 sounds brutal to navigate.
16:36:13pondlifeIt is, but it need not be
16:36:26kugelLlorean: that's since your keymap change, and I saw no mention of this issue yet, so I thought it's intended
16:36:30Lloreanpondlife: Have you tried Pixelma's patch by chance?
16:36:43pondlifeYes, I did, at the time
16:36:53bodymindis it very dangerous to mess up with HAVE_MULTIVOLUME? in export/config-e200v2.h
16:37:00pondlifeIt didn't go far enough for my liking
16:37:02Lloreankugel: I'm pretty sure "stop" used to work everywhere, not just in the root. Could you file a bug report on it, and I'll try to check on it at some point?
16:37:05kugelLlorean: I think it's very useful to stop if you're e.g. going to enter pictureflow being in the plugins folder, but then it doesn't stop but leads you one level up
16:37:10pondlifeBut I don't have the c200 any more
16:37:46kugelbodymind: what do you want to do?
16:38:08kugelbodymind: btw; your partition not found is likely to be caused by the "we cannot access more than 1GB on >2GB targets"
16:38:31bodymindhm, ok, finally an answer :) tkx
16:38:37kugelI suggest formatting your player and put rockbox on it very first
16:38:57bodymindmy partition is 4gb.. should i create a 1gb partition?
16:38:58kugelto make sure rockbox is within the first GB
16:39:16 Part pondlife
16:39:18kugelyou can do that too. But pick a bit less than 1GB (0,95)
16:39:41kugeler wait
16:39:45***Saving seen data "./dancer.seen"
16:39:47kugeldon't change partitions!
16:39:54bodymindnop, i won't do that :p
16:40:01kugelbodymind: really, don't!
16:40:05MarcGuaykugel, Llorean: The stop button seems to act as expected in the database. Actually, it does both: stops playback and returns one level.
16:40:19kugeluse mkfs.vfat to create a filesystem on it with the mentioned ~1Gb
16:40:29bodymindi was thinking a second time... it would brick it for sure :|
16:40:39kugelMarcGuay: still not what I consider as expected
16:41:08MarcGuayErrr... Same in the browser/lists...
16:41:26 Join toffe82 [0] (
16:42:02MarcGuaykugel: Seems the plugins list is a bit different from others.
16:42:37kugeloh yea, it does stop
16:43:18MarcGuayPlugins, Browse fonts, etc, don't stop, only return.
16:43:26MarcGuayOthers do both.
16:43:27bodymindfdisk -l /dev/sdb doesn't report any partition in /dev/sdb :s
16:43:30MarcGuayA bit weird.
16:43:53kugelweird, and surely inconsistent
16:43:58kugelit should just stop, nothing more
16:44:19kugelbodymind: the partition layout is quite weird
16:44:32kugelusing mkfs.vfat is safe though
16:45:03bodymindby safe you mean: mkfs.vfat /dev/sdb ... just to be sure :)
16:45:08Zagorbodymind: isn't it a "superfloppy" like many other flash players?
16:45:38wpyhit's a whole-disk filesystem
16:45:48 Quit _lifeless (Remote closed the connection)
16:46:08bodymindDisk /dev/sdb: 4077 MB, 4077912064 bytes
16:46:08bodymind126 heads, 62 sectors/track, 1019 cylinders
16:46:08bodymindUnits = cylinders of 7812 * 512 = 3999744 bytes
16:46:08DBUGEnqueued KICK bodymind
16:46:08bodymindDisk identifier: 0x00000000
16:46:14bodymindthis is what fdisk reports
16:46:27Zagorit's pretty common. you don't see a partition because there is none. the flash is unpartitioned, just like a floppy.
16:46:38bodymindit won't brick the OF ?
16:46:45bodymindif i make a partition
16:46:52ZagorI don't know
16:47:02MarcGuaykugel: Stop works in the H300 in the plugins list. Doesn't act as a back-button as well. Seems to be some overlapping in the e200 setuup.,
16:47:11kugelbodymind: I tried to with gparted, it nearly bricked my fuze
16:47:24bodymindgparted is dangerous.. :x
16:47:29kugeleither don't touch it at all and put rockbox on it first, or use mkfs.vfat
16:47:34Zagorbodymind: is there a particular reason you want a partition?
16:47:37kugelboth works
16:47:47bodymindok, i will try without mkfs first ;)
16:49:33bodymindZagor: because i can't boot rockbox, i get an error: "Partition not found"
16:49:54Zagorbodymind: aha
16:50:04wpyhkugel: so, have you tried switching to a second bank and reading data from there?
16:50:17kugelit's not implemented yet on ams sansas
16:50:38kugelI'll soon'ish have a look at it though
16:53:08bodymindkugel: i think that i didn't quite understand what you mean with put rockbox first... what do you mean with first? before what?
16:53:20kugelbefore music
16:53:24kugelafter formatting
16:58:04 Join Horscht [0] (n=Horscht@xbmc/user/horscht)
16:58:15bodymindi just want to make sure i'm not building this wrongly.... but what i do is: ../tools/configure
16:58:16 Join parafin [0] (
16:58:18bodymindchoose normal
16:58:32bodymindand then: make bin
16:59:05bodymindand to make the bootloader.. i make another directory where i do: ../tools/configure -> bootloader
16:59:39kugelbodymind: make bin won't be sufficient for the very first installation
17:00:37kugeli mean, it'll boot and run, but you surely want to have codecs and cabbiev2, don't you?
17:00:42Llorean'make bin' generally shouldn't be used unless you know a specific reason why you should be using it.
17:01:11bodymindok, i used it because before it gave an error because it wasn't full implemented
17:03:04 Quit Horscht ("Verlassend")
17:04:38 Quit Zagor ("Client exiting")
17:06:13kugelLlorean: so, this should make the up button don't do anything:
17:06:20kugel(regarding quickscreen)
17:06:22 Join jfc^2 [0] (
17:07:04kugeloh wait, we needed a special case for recorder?
17:07:32 Join Horscht [0] (n=Horscht@xbmc/user/horscht)
17:08:10 Quit Horscht (Client Quit)
17:08:43 Join Horscht [0] (n=Horscht@xbmc/user/horscht)
17:09:23kugelso, this one rather
17:13:21 Quit Horscht (Client Quit)
17:13:24bodymindi get the same thing "No partition found. Insert USB cable and fix it" :\ i've formatted the player with the OF > Settings > Format
17:13:48 Join Horscht [0] (n=Horscht@xbmc/user/horscht)
17:13:55 Quit kushal_12_27_200 ("Leaving")
17:14:22 Quit bxcracer ("Ex-Chat")
17:15:02kugelhave you used mkfs.vfat now or not?
17:15:34kugelmaybe you need a few tries to get it within the first GB
17:15:41 Join bxcracer [0] (n=bxcracer@
17:15:44kugeland, I used windows to format
17:15:58bodymindok, i will do it later ;) tkx for the help
17:19:34 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
17:21:03 Quit Horscht ("I got raided by the FBI and all i got is this lousy quit message")
17:21:33 Quit jfc^2 (Read error: 104 (Connection reset by peer))
17:22:00 Join Horscht [0] (n=Horscht@xbmc/user/horscht)
17:22:01 Join jfc^2 [0] (
17:22:05 Quit Horscht (Read error: 104 (Connection reset by peer))
17:22:18MarcGuaysoap: re: CoLinux. Is it not necessary, then, to resize the image "until/unless you need to run"?
17:22:40 Join Horscht [0] (n=Horscht@xbmc/user/horscht)
17:23:04 Quit Horscht (Read error: 104 (Connection reset by peer))
17:23:28 Join Horscht [0] (n=Horscht@xbmc/user/horscht)
17:26:24 Quit BigE[ssh1 ("Lost terminal")
17:32:31 Join evilnick [0] (i=0c140464@gateway/web/ajax/
17:34:28 Quit jfc^2 (Read error: 104 (Connection reset by peer))
17:37:39 Join jfc^2 [0] (
17:38:31 Quit moos ("Rockbox rules the DAP world")
17:40:37MarcGuaysoap: I updated the page to match changes they seem to have made to the setup. Review at your own risk.
17:48:03 Join Strife89 [0] (n=michael@
17:48:18 Join {phoenix} [0] (
17:48:59 Join Seed [0] (
17:51:17 Quit MarcGuay (" ajax IRC Client")
17:54:34 Quit bodymind ("Ex-Chat")
17:59:23 Join Jaykay [0] (
18:04:41 Quit jfc^2 (Remote closed the connection)
18:07:28 Quit Lynx_ (Read error: 145 (Connection timed out))
18:07:38 Join Lynx_ [0] (
18:11:06 Join miepchen^schlaf [0] (
18:11:38 Quit petur ("work->home")
18:13:49 Join rasher [0] (
18:13:59 Join stoffel_ [0] (
18:17:15soapahh, he left
18:18:53 Join kachna|lappy [0] (
18:19:47 Join MethoS- [0] (
18:24:18rasherWhy is there a find_albumart and search_albumart_files?
18:24:54rasherfind_albumart is only used in playback.c, and doesn't seem to do a complete search
18:25:46rasherOkay, find_albumart calls search_albumart_files, so the search is correct. Now the question remains: why?
18:27:15rasherOkay. I think I get it. Ignore me.
18:27:46 Join karashata [0] (n=karashat@
18:28:18 Join faemir [0] (
18:29:57 Join mib_hnbveix1 [0] (i=d8ef4910@gateway/web/ajax/
18:30:02 Nick mib_hnbveix1 is now known as MarcGuay (i=d8ef4910@gateway/web/ajax/
18:30:11MarcGuaysoap: Rebooting.
18:30:18MarcGuayErr. Rebooted.
18:30:44soapok, the changes look fine - have you tested them I assume?
18:31:06MarcGuayThey've changed quite a bit since you set up that page. I'm having trouble with the network connection right now, but the installation seems successful.
18:31:09soapi think we should (we = I) try to keep the image up-to-date and not ask the end users to ./
18:32:03soapI will have a more consistent Windows box here in a day or two once I move more functions of this machine over to my new "server" and will be able to maintain that a little better.
18:33:09soapYou are not the first one to complain of network issues, I was able to walk two people through them in the last 12 months, but one person I was unable to help - that points to a fundamental flaw in my networking setup somewhere.
18:33:21MarcGuaysoap: In your experience, having the main network connection set to allow others to use it... Ah... :)
18:33:26soapI also think we need to emphasize that CoLinux is rarely better than VMWare.
18:33:50MarcGuayI lost my VMWare setup in a wave of crap and thought I'd try something new.
18:34:28soap_I_ was using it because I had a CPU at the time which was not supported by VMWare, I have a feeling the number of people with hardware pre-Pentium III is diminishing rapidly.
18:34:48 Join herrwaldo [0] (
18:35:07MarcGuayI found VMWare rather flaky as well. It often wouldn't let me access the shared drive until I rebooted.
18:37:10 Join _lifeless [0] (n=lifeless@
18:39:46***Saving seen data "./dancer.seen"
18:40:21 Quit Horscht (Read error: 104 (Connection reset by peer))
18:40:45 Quit Strife89 ("Going to eat some lunch.")
18:40:53 Join Horscht [0] (n=Horscht@xbmc/user/horscht)
18:46:28 Quit Horscht ("I am root. If you see me laughing, you better have a backup")
18:47:51 Quit stoffel_ (Read error: 113 (No route to host))
18:52:20 Join stoffel_ [0] (
18:53:20 Quit linuxstb (Read error: 113 (No route to host))
18:58:54lucentkugel: bodymind's error is the same error that happens when I access ##MUSIC# folder from within rockbox firmware, it is a different error than the SD timeout when I load .rockbox/* above 1GB
18:59:17 Join neddy1 [0] (n=john@nat/sun/x-b6ac52904cad4d1d)
18:59:54 Quit neddy1 (Client Quit)
19:00:01kugellucent: I can access that folder
19:00:45lucentkugel: I can browse the files in it, and when I attempt to play music that was pre-loaded on my Sansa Fuze from ##MUSIC#, there's an error about needing to reattach USB
19:01:47lucent"No partition" "found." "Insert USB cable" "and fix it."
19:02:03kugelso it's probably after 1GB
19:02:08CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
19:02:08*rasher thinks Bagder should get around to making the build system keep handing out builds
19:03:32 Join petur [50] (n=petur@rockbox/developer/petur)
19:03:49 Quit karashata ("G'bye everyone!")
19:03:52 Join Horscht [0] (n=Horscht@xbmc/user/horscht)
19:04:48lucentkugel: maybe you knew this already, but the error that happens when I insert my 8GB microSD card, it happens because we don't get the CID... but just before that happens it is correctly identified as sdhc = true
19:05:09lucentkugel: the internal memory I'm guessing, it is identified as sdhc = false
19:05:41lucentkugel: I guess this by sticking a panicf in there only when sdhc is true, and it only does that panicf when I insert my sdhc card
19:06:00kugelsdhc isn't the issue
19:06:04lucentif I have the panicf in there anytime sdhc = false, then it panicf's on boot
19:06:04kugelmy microsd is hc too
19:06:05lucentoh okay
19:06:41lucentkugel: I'm curious to know what sd-ata-pp.c does and if it's being used at all?
19:07:10kugelwell, it's for the PP targets
19:07:14lucenterr I mixed up the name, it is actually ata-sd-pp.c
19:07:18kugeldoesn't have much to do with our targets
19:07:25lucentFuze is not PP, oh okay
19:07:55lucenthave we any targets that do work properly with 8gb SD?
19:08:11kugelyes, sansa v1
19:08:37kugelwell, "properly" is slightly exaggerated I guess
19:09:03hillshum_still better than what SanDisk did
19:09:13kugelwriting to those can cause corruptions very rarely
19:09:28Lloreankugel: But that's to non-HC SD as well.
19:09:33lucentI'm watching the source code, I see:
19:09:35lucentsd_command(SD_ALL_SEND_CID, 0, currcard->cid, 2);
19:09:52lucentsend_cmd(drive, SD_ALL_SEND_CID, 1, MCI_RESP|MCI_LONG_RESP|MCI_ARG, card_info[drive].cid)
19:11:19lucentwait though, I was modifying a bit
19:11:26lucentthe '1' is my blind work
19:12:28 Quit neddy (Connection timed out)
19:14:36lucentabout that last '2' parameter though, the code goes on:
19:14:44lucentwords = (type == 2) ? 9 : 3;
19:14:59lucentso hm there's some magic going on for PP targets
19:15:34*kugel doesn't really understand what lucent wants to tell him
19:15:55lucentsorry I'm just thinking out loud
19:16:28 Quit Horscht ("")
19:16:45lucentI read through the MMC v4.2 and SD simplified specs
19:17:24lucentsomething is broken because we don't get the CID from an inserted card on my player
19:17:36lucentkugel: do you get the CID from when you insert a card on your player?
19:20:40kugellucent: looks like
19:20:59kugelI haven't double checked, but as I don't get the panic, it seems the card is properly initialized
19:23:41kugellucent: type is just the response length, which is covered by MCI_LONG_RESP for our driver
19:25:57*lucent nods
19:26:14lucentthat's clever
19:26:32 Join Horscht [0] (n=Horscht@xbmc/user/horscht)
19:27:47kugellucent: I have personally no idea why cid fails on your card
19:28:18lucentwhat's the preferred way to get debug output without panic'ing?
19:34:06B4gderlogf() ? splash?
19:35:14gevaertsmorse on backlight?
19:37:08lucentI'm not familiar with logf, does it output via USB cable or...
19:37:57gevaertsIt can on portalplayer, but the usual way is via a "viewer" in the debug menu
19:39:54lucentokay thanks
19:41:30lucentkugel: what does your debug view partitions have to say with and without the card inserted?
19:43:17 Quit havien (Read error: 104 (Connection reset by peer))
19:43:38 Join havien [0] (
19:44:22 Join karashata [0] (n=karashat@
19:44:32 Quit MarcGuay (" ajax IRC Client")
19:45:20kugellucent: not much
19:45:29kugel4 partitions, each 0, in both cases
19:45:45kugeldisk info is more interesting for external storage
19:46:03 Join bertrik [0] (
19:48:33 Quit BigBambi (Remote closed the connection)
19:49:04lucent is mine
19:50:16kugelit shows 0 for all here
19:50:18lucentthat's different right? you said yours are 0
19:50:35kugelmight be due to that I used mkfs.vfat though
19:51:05*kugel found buttonlight code in the OF
19:51:27 Quit stoffel_ (Read error: 60 (Operation timed out))
19:52:18*J-23 learns to play guitar
19:52:20 Join stoffel_ [0] (
19:52:21bertrikkugel, you got radio to work on the fuze too, right?
19:52:27J-23argh, wrong channel ;)
19:52:41kugelbertrik: yes
19:52:42 Quit stoffel_ (Client Quit)
19:53:08kugelbertrik: are you going to commit?
19:54:11kugelDoes anyone know do address of the kind 0x21c54 mean in the ams sansa OF?
19:54:20bertrikthe fmradio_i2c stuff was committed yesterday, the si4700 needs a little more discussion with n1s and the codec LINEIN2 stuff also needs more thought before committing
19:55:33kugelbertrik: uhm, did I get something wrong? Is the radio already in svn?
19:55:56kugelwhat's the clip-radio6.patch for then?
19:55:56lucentin sansa fuze OF the "lock image" appears when hitting hold, wonder if it can be backtracked by finding that
19:56:21kugellucent: sure, point us to the image :)
19:58:17J-23remove the LCD and capture signals going to it, and convert them to bitmap :P
19:58:49bertrikkugel, that patch contains some ugly hacks to demonstrate that radio works, but that should not be committed
19:59:25lucentbertrik: would just look for 3x3 black color in bitmap format, unless it's compressed
19:59:47lucentthe center of the "keyhole" graphic is all one color black value
19:59:55lucentnot seen anywhere else in images that I look at
20:00:34lucentmightn't be possible hm
20:00:49lucentother thing to look at is all cases where the backlight is being turned on
20:01:09 Join gregzx [0] (
20:01:14kugellucent: I bet there're dozens of images a black 3x3 part
20:01:36lucentwith "hold" engaged, when a key is pressed, the backlight is turned on and that graphic is displayed, and then disappears after a wait
20:02:51 Join Zagor [242] (
20:05:29 Join Thundercloud [0] (
20:13:12kugelI think I found buttonlight(bool on)
20:14:21kugellcd init calls it with on=true
20:15:08lucentmakes sense, pressing power button in OF -> backlight turns on, buttonlight turns on
20:15:08 Join Horschti [0] (
20:15:25 Quit Horscht (Nick collision from services.)
20:16:29kugellcd init != backlight on
20:19:49kugelthat's the function
20:20:04kugellucent: I'm fighting with the buttonlight, since it interferes with the microsd
20:20:16kugelso I need to look into the disassembly
20:21:00lucentkugel: I hope you are successful with the task :)
20:21:21*kugel too
20:24:24 Join jhulst [0] (n=jhulst@unaffiliated/jhulst)
20:32:56kugelbertrik: ping
20:33:17 Join bluebrother [0] (n=dom@rockbox/developer/bluebrother)
20:33:49 Join __lifeless [0] (n=lifeless@
20:35:37*bertrik is still here
20:36:36kugelbertrik: I'm curious what those adresses could mean: 0x300a70b0, 0x21c54, 0x24504
20:37:48 Join bmbl [0] (n=Miranda@unaffiliated/bmbl)
20:38:28bertrikthe 0x2XXXX addresses are in internal RAM, the 0x3XXXXXXX is in DRAM
20:39:48***Saving seen data "./dancer.seen"
20:39:49kugelbertrik: ah, ok (I already expected the latter one). both are used by rockbox right? I.e. reading those because the OF does isn't helpful
20:40:16kugelbertrik: here's my buttonlight disassembly, maybe you have an idea what those could mean
20:41:11kugelwhoops, the second last line is wrong
20:42:29*amiconn wonders who thought defaulting replaygain to 'on' is a good idea
20:42:40*kugel hides
20:42:54*Jaykay thinks it was a good idea
20:42:55*amiconn even wonders who invented replaygain in the first place :((
20:43:58kugelI did not invent it!
20:44:06*Jaykay hides to
20:44:18*Jaykay cant write english properly
20:44:56Jaykayamiconn:whats so bad about replaygain?
20:45:24amiconnIt's annoying. I don't want any player/playing software to mess with playback related things behind my back
20:46:13Jaykaywhy is it annoying? i thing the changes are in most cases subtle
20:47:05Zagoramiconn: the default settings are not for you
20:47:21pixelmait doesn't do anything if you don't have any replaygain tags, does it?
20:47:35amiconnWith the new defaults, I have to set even more things to non-default than with the old ones
20:47:57amiconnAnd afaik precut is always applied, tags present or not
20:48:00Zagoruse a .cfg file
20:48:35*amiconn thinks we should follow a strict opt-in policy for things like this
20:49:10Zagor"things like this"?
20:49:12kugelprecut isn't changed
20:49:41 Quit _lifeless (Read error: 110 (Connection timed out))
20:50:07kugelbertrik: do you have a few minutes to look at it?
20:50:34amiconnZagor: Fancy things which change output without the user noticing (if he doesn't check settings)
20:51:06bertrikkugel, to be honest, I'm not very motivated to look at the button light code tonight. I've seen it in the sansa clip code too and it confused me.
20:51:16amiconnAnd while I don't use and don't want replaygain, I cannot guarantee that all my tracks are replaygain tag free
20:51:29kugelbertrik: have you even clicked at it? I annotated it already
20:51:45lucentGPIOD hmm
20:52:51kugelamiconn: then batch remove the tags. having it on on default is a good thing
20:53:01 Join Lear [0] (i=chatzill@rockbox/developer/lear)
20:53:09kugelit doesn't alter/cripple the sound, it only changes the volume
20:54:28lucentyeah and it only does it once per track, not continuously
20:55:47 Join akur [0] (
20:55:49 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
20:56:04amiconnChanging the volume is changing the sound imo. And it does it in a way that can potentially cause distortion
20:57:01kugelamiconn: turn it off then. replaygain is for people that want t avoid distortion, and thus it should be on
20:57:10pixelmacan there be positive replaygain values?
20:57:18lucentamiconn: there's a desire for features which can normalize the volume of a wide ranging recording style, older albums are "quieter" with more headroom, newer albums are compressed to sound "louder"... but they want this change without modifying the original rip
20:57:38kugelpixelma: if the level is below 89db yes
20:57:47lucentreplaygain is sufficient and widely used
20:57:59gevaertslucent: The objection is to the *by default* thing, not about the existence of the option
20:58:00pixelmaso yeah, there is one downside
20:58:11 Join webguest18 [0] (n=44af5d91@gateway/web/cgi-irc/
20:58:14*bertrik considers renaming as3525-codec.c to ascodec-as3525.c to bring it in line the convention of the other as3525 driver code
20:58:25kugelpixelma: why is this a downside?
20:58:32Zagorpixelma: it's no more a downside than poorly ripped encoded files are
20:58:39kugelpositive values don't automatically mean distortion
20:58:45amiconnlucent: I know this, and what replaygain does. But I much prefer adjusting the volume manually than have the playback software manipulate it behind my back
20:59:12kugeland if it causes noticeable distortion, then the files aren't replaygained
20:59:20pixelmae.g. if you use your player on car/home stereo and explicitly set it to 0dB, you can get clipping if the rg value is positive and you don't know about it
20:59:41 Quit webguest18 (Client Quit)
20:59:42Zagorpixelma: no. in that case the replaygain value is wrong
21:00:43 Quit Jaykay ("ChatZilla 0.9.84 [Firefox 3.0.4/2008102920]")
21:01:41kugelamiconn: but the other 99% of the people prefer to use existing possibilities to avoid manually adjusting the volume for every single track
21:01:57amiconnZagor: Afaiu replaygain normalizes the perceived volume. So for faint track with a few sharp pulses, applying the replaygain value can lead to the pulses clipping. The precut value is there to prevent this, and it needs to set low enough to actually do.
21:02:01lucentkugel: most people are not deejays
21:02:04*kugel surely exaggerated, but still
21:02:12amiconnHow much this 'low enough' is depends on the music...
21:02:26lucentkugel: most people want their toaster oven to make toast, instead of them sitting there with a shirt iron to make their own toast the perfect way by hand
21:02:48pixelmaZagor: understood now, took a while
21:02:54LearPre-cut isn't to prevent distortions; that's what the peak values are for.
21:05:20lucentamiconn: it's a matter of convenience
21:05:44*amiconn is fundamentally annoyed by this replaygain
21:06:03lucentif ReplayGain is inconvenient for you, then why not improve ReplayGain?
21:06:19Zagoramiconn: then keep it disabled.
21:06:33amiconnSince rockbox supports it, and it is technically possible to implement it on hwcodec, it should be implemented on hwcodec
21:07:18Zagoramiconn: no, we implement things we want to implement. if you want to add it to hwcodec, feel free to do so.
21:07:38amiconnThere is a patch for this, but the patch doesn't take the mas buffer delay into account. It is possible to do this, but it's not trivial, and since replaygain is one of those features I never use, I'm not motivated to work on it....
21:07:41Unhelpfulhow would that work, read the RG tags, and change the global gain in the packets before feeding to the codec?
21:08:03amiconnUnhelpful: No, simply adjust the output volume accordingly
21:08:41amiconnThis can even be done without adjusting the global analog volume, by using the stereo matrix in the mas. This matrix also works for s/pdif output
21:09:22Unhelpfulthe other way would be equivalent to what mp3gain "normally" does, and would not involve having to account for a decode delay.
21:09:24amiconnBut fiddling with the output volume reqires taking the mas buffer delay into account
21:09:27 Quit gevaerts (Nick collision from services.)
21:09:36 Join gevaerts [0] (n=fg@rockbox/developer/gevaerts)
21:09:42amiconnUnhelpful: Yeah, but it would only work for mp3, not for mp2, and neither for pcm
21:09:44 Join tyfoo2 [0] (
21:10:26Unhelpfulhrm, mp2 doesn't have per-packet gain values?
21:12:31 Nick hillshum_ is now known as Hillshum (
21:14:28Unhelpfulthat would be a problem then. extending the "volume adjust" method to swcodec targets would also mean that we don't need DSP for RG if there are no other DSP effects being used, and that we're sending the sound to the audio hardware unaltered in that case, for audiophiles who care about that.
21:15:12 Join vertic23 [0] (
21:15:39amiconnSwcodec has a potentially higher and less predictable delay between feeeding samples into the pcm buffer and actually outputting them
21:16:44Unhelpfulwould the hardware allow an interrupt on reaching a certain point in the buffer, or similar?
21:16:55amiconnWhat we basically need to do on hwcodec is calculating how many frames are in the mas' internal buffer, in order to calculate the delay. This can be done by looking at the size of every frame we feed itno the mas
21:17:22 Quit Thundercloud (Remote closed the connection)
21:18:18amiconnThe mas requests new data as soon as there is some free space, and we can check the timer tick regularly. I think it should be possible to get the estimate accurate to about +/- 10 ms (1 tick)
21:19:17amiconnThen we should push gain reductions a little, and delay gain increases a little to always stay on the quite side
21:20:29amiconnThe mas buffer is fixed size, but is able to buffer ahead a significant amount of time. With 8kbps mp3 it's >1 second
21:20:33Unhelpfuli can see things being a little dodgy in shuffled mode, if we're going from or to a track that mixes to the adjacent one on the album it came from. would fading the volume across a few ticks maybe be a good idea?
21:21:14amiconnIf you use album gain, nothing would change between those tracks
21:22:20Unhelpfulamiconn: right, but if we're transitioning to such a track while shuffling, it could hit quite loud or quiet compared to the previous track.
21:22:25amiconnAnd you cannot really fade - setting the volume isn't very fast, because it has to go through the mas' i2c
21:22:55amiconnUnhelpful: Hence my idea to delay gain increases, and push gain decreases
21:23:21 Quit akur (Read error: 145 (Connection timed out))
21:23:33LearUnhelpful: ReplayGain is supposed to adjust samples, to reduce the likelyhood of clipping (e.g., from lossy encoding).
21:23:38Unhelpfulamiconn: i didn't follow what you were saying before, but i see that it's exactly what i was about to tell you.
21:23:52amiconnThis would mean that the sound ends up too quiet for a few milliseconds - imo much better than too loud
21:24:00 Quit tyfoo (Connection timed out)
21:26:57 Join BigBambi [0] (
21:27:50Unhelpfuli'm inclined to agree, if the best we can hope for is "adjust volume on a tick near the actual track change"
21:28:07 Join domonoky [0] (n=Domonoky@rockbox/developer/domonoky)
21:28:43 Quit perrikwp (" ajax IRC Client")
21:32:49 Quit vertic39 (Read error: 110 (Connection timed out))
21:33:20 Join akur [0] (
21:33:26 Part akur
21:33:40 Quit LambdaCalculus37 (" ajax IRC Client")
21:34:02 Join LambdaCalculus37 [0] (i=44a04303@rockbox/staff/LambdaCalculus37)
21:36:15 Join toffe82_ [0] (
21:40:23 Join blkhawk- [0] (
21:44:23 Join saratoga [0] (n=9803c6dd@gateway/web/cgi-irc/
21:45:17saratogareplaygain induced clipping is virtually unheard of
21:45:30 Join Hillshum_ [0] (i=4ba5f199@gateway/web/ajax/
21:46:13 Quit Hillshum_ (Client Quit)
21:46:14lucentI've heard it
21:46:22lucentand it was completely my fault
21:46:58lucentsong-by-song RG on a collection with some really bad encodes in the pile
21:47:24 Join neddy [0] (n=john@nat/sun/x-a6a5e8f8b583ab86)
21:49:44 Join dfkt [0] (i=dfkt@unaffiliated/dfkt)
21:51:40 Join Rondom [0] (
21:53:59 Quit tyfoo2 ("Carpe diem")
21:55:45 Join Thundercloud [0] (
21:57:23 Join massiveH [0] (
21:57:23 Quit massiveH (Client Quit)
21:57:58 Quit blkhawk (Read error: 110 (Connection timed out))
21:58:21 Nick blkhawk- is now known as blkhawk (
21:59:39 Quit LambdaCalculus37 (" LUG Xmas party!")
21:59:50 Quit dfkt_dt (Read error: 60 (Operation timed out))
22:00:32 Quit toffe82 (Connection timed out)
22:00:52 Quit gregzx (Remote closed the connection)
22:01:22 Quit faemir (Operation timed out)
22:06:25 Join massiveH [0] (
22:06:42 Quit massiveH (Read error: 104 (Connection reset by peer))
22:11:05 Join dfkt_dt [0] (i=dfkt@unaffiliated/dfkt)
22:15:47saratogakugel: has anyone requested the USB datasheet yet for the AMS chips?
22:16:15kugelsaratoga: Not that I know
22:17:38 Quit dfkt (Read error: 60 (Operation timed out))
22:21:07lucenthow to delete files from FAT so that the data is zero
22:21:25lucentelse it is hard to compress a filesystem image
22:21:41 Join gregzx [0] (
22:22:35saratogacreate a file thats all zeros the size of the disk
22:23:01 Quit {phoenix} (Remote closed the connection)
22:23:30lucentcopying from fuze internal is slow for me :/
22:23:41lucenttaking an 1+ hour to copy 8GB
22:24:24 Quit Horschti ("User was distributing pornography on server; system seized by FBI")
22:26:13 Join tyfoo [0] (
22:27:55 Quit fredddy (Remote closed the connection)
22:30:24 Quit karashata ("G'bye everyone!")
22:30:47 Join Horscht [0] (n=Horscht@xbmc/user/horscht)
22:35:41 Quit B4gder ("It is time to say moo")
22:39:51***Saving seen data "./dancer.seen"
22:50:21kugeldoes anyone have an idea what "Rod control" might be?
22:54:32 Quit Lear ("ChatZilla 0.9.84 [Firefox 3.0.5/2008120122]")
22:59:11 Join faemir [0] (
22:59:52linuxstbkugel: resistor open drain control?
23:01:19kugellinuxstb: no idea what that is :/ but it's set after activating the buttonlight in the OF
23:01:49linuxstbkugel: It may help if you gave some context to your question.
23:01:52*kugel made a disassembly since buttonlight isn't trivial due to interferes with microsd
23:03:17kugelSD_MCI_BASE bit 7 is "Rod control", as per datasheet
23:03:20 Quit XavierGr (Nick collision from services.)
23:03:31 Join XavierGr [0] (n=xavier@rockbox/staff/XavierGr)
23:05:21kugeland I fount that some kind of buttonlight_init within lcd_init_device
23:08:46kugellinuxstb: do you have an idea what that is?
23:10:35*bluebrother wonders if it would help to put the autodetection button at the top of the window
23:13:17linuxstbkugel: No.
23:13:46 Quit Thundercloud (Remote closed the connection)
23:17:37 Quit domonoky (Read error: 104 (Connection reset by peer))
23:18:31 Join JdGordon [0] (
23:27:40bluebrotherhmm. Displaying a busy dialog instead of the selection is a bit strange if the detection is rather quick −− i.e. if no slow external drive is involved.
23:29:32 Quit bertrik ("Leaving")
23:30:25 Quit Rondom ("Ex-Chat")
23:31:19 Quit JdGordon|zzz (Read error: 110 (Connection timed out))
23:33:50PaulJamhmm, does anyone else find it strange that now the "backlight (when plugged in)" timeout is shorter than the backlight on battery timeout?
23:34:47soapgood catch - change it!
23:38:36 Join ameyer [0] (
23:42:03 Quit faemir (Read error: 104 (Connection reset by peer))
23:42:42 Join faemir [0] (
23:45:23 Quit petur ("Zzzzz")
23:46:40 Join RockRabbit [0] (n=3aac9a01@gateway/web/cgi-irc/
23:47:00 Quit HellDragon (Read error: 54 (Connection reset by peer))
23:48:20 Quit RockRabbit (Client Quit)
23:49:00 Join HellDragon [0] (
23:49:07 Join RockRabbit [0] (n=3aac9a01@gateway/web/cgi-irc/
23:50:02RockRabbitWhen trying to compile rockbox, im getting "../tools/configure: 2044: arm-elf-gcc: not found WARNING: The compiler you must use (arm-elf-gcc) is not in your path! " trying to run ../tools/configure.
23:50:18lucentRockRabbit: are your devel tools set up?
23:50:33RockRabbitI have already run to get the development tools and that seemed to work fine.
23:51:24bluebrotherdo you have put the install path into your PATH variable?
23:51:30lucentRockRabbit: you need to set your path to get to those tools where they are installed, I think
23:51:32bluebrother*have you put
23:53:31RockRabbitecho $PATH shows "/usr/local/arm-elf/bin:/usr/bin/perl:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/games "
23:53:58lucentdid you export PATH or just set it ?
23:54:00RockRabbitHowever the appears to be no arm-elf-gcc anywhere on the hard drive.
23:54:08lucentoh heh
23:54:19lucentRockRabbit: what prefix did you install to?
23:54:53RockRabbitWhat do you mean by prefix?
23:55:19RockRabbitThe line in my bashrc file is "export PATH=/usr/local/arm-elf/bin:/usr/bin/perl:$PATH"
23:55:20bluebrotherhow did you install the compiler?
23:55:59 Quit evilnick (" ajax IRC Client")
23:56:09RockRabbitI can build an early version of the bootloader no problems
23:56:33lucentum... okay
23:56:45lucentRockRabbit: which OS are you using?
23:56:57RockRabbitUbuntu 8.10 AMD64
23:57:13lucentlet's walk through the process together, okay?
23:57:35lucentI'll privmsg you, one minute please
23:57:53 Quit ender` (" The likelihood of Perl being involved in a system is directly proportional to the length of time the system has been in mai")

Previous day | Next day