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 2011-09-24

00:00:04 Join factor [0] (~factor@
00:01:01 Quit TheLemonMan (Quit: WeeChat 0.3.5)
00:09:08 Quit bertrik (Quit: And That, My Liege, Is How We Know the Earth to Be Banana Shaped)
00:22:23 Quit lovasoa (Ping timeout: 256 seconds)
00:23:33***Saving seen data "./dancer.seen"
00:23:59 Quit ReimuHakurei_ (Ping timeout: 260 seconds)
00:24:15 Quit dre (*.net *.split)
00:24:15 Quit Farthen (*.net *.split)
00:31:07 Join madskiny [0] (dre@gateway/shell/
00:32:00 Join Farthen [0] (~Farthen@2a01:4f8:101:2a4:0:bc28:b2e1:9)
00:35:22 Quit liar (Ping timeout: 260 seconds)
00:36:23 Join [Saint] [0] (~st.lasciv@
00:38:17 Quit kadoban (Ping timeout: 260 seconds)
01:07:55 Quit n1s (Remote host closed the connection)
01:08:08 Join soap [0] (~soap@rockbox/staff/soap)
01:44:28 Join mamarley [0] (~quassel@2001:470:5:84d:221:6aff:fe19:4d0c)
01:55:17 Quit neferty (Ping timeout: 260 seconds)
01:56:01 Join neferty [0] (~andor@
02:02:34 Join FoolOnHill [0] (
02:05:12 Quit FoH (Ping timeout: 260 seconds)
02:17:51 Join kadoban [0] (
02:23:37***Saving seen data "./dancer.seen"
02:44:43 Quit mamarley (Remote host closed the connection)
02:49:29 Quit pamaury (Remote host closed the connection)
03:34:26 Quit hilbert (Quit: zzzzzzzzz.)
03:39:41 Quit kadoban (Ping timeout: 276 seconds)
04:23:38***Saving seen data "./dancer.seen"
04:32:13 Quit [Saint] (Ping timeout: 260 seconds)
04:36:48 Join [Saint] [0] (~st.lasciv@
04:41:39 Quit amiconn (Disconnected by services)
04:41:39 Quit pixelma (Disconnected by services)
04:41:40 Join amiconn_ [0] (quassel@rockbox/developer/amiconn)
04:41:41 Join pixelma_ [0] (quassel@rockbox/staff/pixelma)
04:41:44 Nick pixelma_ is now known as pixelma (quassel@rockbox/staff/pixelma)
04:42:02 Nick amiconn_ is now known as amiconn (quassel@rockbox/developer/amiconn)
04:46:56 Quit TheSeven (Disconnected by services)
04:47:09 Join [7] [0] (~TheSeven@rockbox/developer/TheSeven)
05:03:29 Join bluefoxx [0] (
05:03:43 Quit bluefoxx_ (Ping timeout: 260 seconds)
05:20:39DEBUGEOF from server (Connection timed out) (snapshot: netstuff.c line 545)
05:20:39***Saving seen data "./dancer.seen"
05:20:42***Started Dancer V4.16
05:20:42***Connected to on port 6667
05:20:42***Logfile for #rockbox started
05:20:42Mode"logbot :+i" by logbot
05:20:47***Server message 501: 'logbot :Unknown MODE flag'
05:20:47 Join logbot [0] (
05:20:47 Join linuxstb [0] (~linuxstb@rockbox/developer/linuxstb)
05:20:47 Join rasher [0] (~rasher@rockbox/developer/rasher)
05:20:47 Join Lalufu [0] (~sun@unaffiliated/lalufu)
05:20:47 Join avacore [0] (
05:20:47 Join polemon [0] (
05:20:47 Join feisar- [0] (
05:20:47 Join markun [0] (
05:20:47 Join sinthetek [0] (~sinthetek@unaffiliated/sinthetek)
05:20:47 Join semitones [0] (
05:20:47 Join bieber [0] (
05:20:47 Join fyrestorm [0] (
05:20:47 Join bluefoxx [0] (
05:20:47 Join [7] [0] (~TheSeven@rockbox/developer/TheSeven)
05:20:47 Join pixelma [0] (quassel@rockbox/staff/pixelma)
05:20:47 Join amiconn [0] (quassel@rockbox/developer/amiconn)
05:20:47 Join [Saint] [0] (~st.lasciv@
05:20:47 Join FoolOnHill [0] (
05:20:47 Join neferty [0] (~andor@
05:20:47 Join soap [0] (~soap@rockbox/staff/soap)
05:20:47 Join Farthen [0] (~Farthen@2a01:4f8:101:2a4:0:bc28:b2e1:9)
05:20:47 Join madskiny [0] (dre@gateway/shell/
05:20:47 Join factor [0] (~factor@
05:20:47 Join fs-bluebot [0] (
05:20:47 Join antil33t [0] (
05:20:47 Join bluebroth3r [0] (~dom@rockbox/developer/bluebrother)
05:20:47 Join Zarggg [0] (
05:20:47 Join Xerion [0] (
05:20:47 Join Horschti [0] (~Horscht@xbmc/user/horscht)
05:20:47 Join ej0rge [0] (~alhaz@
05:20:47 Join ender| [0] (
05:20:47 Join mystica555 [0] (
05:20:47 Join powell14ski [0] (
05:20:47 Join Topy [0] (
05:20:47 Join mikroflops [0] (
05:20:47 Join kevku [0] (x@2001:470:1f15:1d8f::)
05:20:47 Join Rob2223 [0] (
05:20:47 Join tchan [0] (~tchan@lunar-linux/developer/tchan)
05:20:47 Join simonlnu [0] (~simon@unaffiliated/simonrvn)
05:20:47 Join saratoga [0] (9803ec71@gateway/web/freenode/ip.
05:20:47 Join Bagder [241] (~daniel@rockbox/developer/bagder)
05:20:47 Join balintx [0] (
05:20:47 Join evilnick [0] (~evilnick@rockbox/staff/evilnick)
05:20:47 Join bzed [0] (
05:20:47 Join ranmachan [0] (ranma@2a01:4f8:130:9321::2)
05:20:47 Join TBCOOL [0] (
05:20:47 Join user829385 [0] (~aoeu@
05:20:47 Join niekie [0] (~niek@CAcert/Assurer/niekie)
05:20:47 Join alexbobp [0] (~alex@
05:20:47 Join kugel [0] (~kugel@rockbox/developer/kugel)
05:20:47 Join Whopper [0] (~colin@about/cooking/sandwich/whopper)
05:20:47 Join froggyman [0] (~seth@unaffiliated/froggyman)
05:20:47 Join parafin [0] (
05:20:47 Join jae [0] (
05:20:47 Join Guinness [0] (
05:20:47 Join soap_ [0] (~soap@rockbox/staff/soap)
05:20:47 Join ps-auxw [0] (~arneb@2001:470:c807:0:1532:4e5f:2ad3:4123)
05:20:47 Join semitones_ [0] (
05:20:47 Join GeekShadow [0] (
05:20:47 Join user890104 [0] (
05:20:47 Join Galois [0] (
05:20:47 Join Rondom [0] (~rondom@2a01:488:66:1000:b24d:4f2f:0:1)
05:20:47 Join Elfish [0] (amba@2a01:4f8:100:90a1:abc:abc:abc:abc)
05:20:47 Join ukleinek [0] (
05:20:47 Join GodEater [0] (~bibble@rockbox/staff/GodEater)
05:20:47 Join desowin_ [0] (
05:20:47 Join tmzt [0] (~tmzt@
05:20:47 Join eintopf [0] (
05:20:47 Join Torne [0] (~torne@rockbox/developer/Torne)
05:20:47 Join guymann [0] (
05:20:47 Join Zambezi [0] (Zulu@unaffiliated/zambezi)
05:20:47 Join beta2k [0] (
05:20:47 Join jhMikeS [0] (~jethead71@rockbox/developer/jhMikeS)
05:20:47 Join Hadaka [0] (
05:20:47 Join maraz [0] (
05:20:47 Join simabeis [0] (
05:20:47 Join Unhelpful [0] (~quassel@rockbox/developer/Unhelpful)
05:20:47 Join efyx__ [0] (
05:20:47 Join [fred] [0] (
05:20:47 Join scorche|sh [0] (
05:20:47 Join [Sanyi] [0] (thg@gateway/shell/
05:20:47 Join Llorean [0] (~DarkkOne@rockbox/user/Llorean)
05:20:47 Join Utchybann [0] (
05:20:47 Join FOAD [0] (~foad@unaffiliated/foad)
05:20:47 Join zchs [0] (
05:20:47 Join Slasheri [0] (miipekk@rockbox/developer/Slasheri)
05:20:47 Join aevin [0] (eivindsy@unaffiliated/aevin)
05:20:47 Join juri_ [0] (~juri@
05:20:47 Join Guest42312 [0] (
05:20:47 Join nomada [0] (
05:20:47 Join rudi_s [0] (
05:20:47 Join mc2739 [0] (~mc2739@rockbox/developer/mc2739)
05:20:47 Join banan_ [0] (
05:20:47 Join Staphylo [0] (
05:20:47 Join crwl [0] (
05:20:47 Join zu [0] (
05:20:47 Join linuxguy3 [0] (
05:20:47 Join ved [0] (
05:20:47 Join scorche [0] (~scorche@rockbox/administrator/scorche)
05:20:47 Join yosafbridge [0] (
05:20:47 Join gevaerts [0] (~fg@rockbox/developer/gevaerts)
05:20:47 Join preglow [0] (
05:20:47 Join @ChanServ [0] (ChanServ@services.)
05:20:47 Join CIA-14 [0] (
05:20:47 Join knittl [0] (~knittl@unaffiliated/knittl)
05:20:47 Join jacekowski [0] (
05:20:47 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
05:20:47 Join dionoea [0] (~dionoea@videolan/developer/dionoea)
05:20:47 Join Ac-town [0] (~Ac-town@osuosl/staff/actown)
05:20:47 Join krazykit [0] (
05:20:47 Join Barahir [0] (
05:20:55 Join Rob2222 [0] (
05:25:18 Quit Rob2223 (Ping timeout: 260 seconds)
05:34:41 Join ruskie [0] (ruskie@sourcemage/mage/ruskie)
06:31:13 Quit [Saint] (Ping timeout: 260 seconds)
06:33:41 Join [Saint] [0] (~st.lasciv@
06:39:43 Quit powell14ski (Quit: powell14ski)
06:44:54 Quit [Saint] (Ping timeout: 256 seconds)
07:08:15 Quit factor (Read error: Connection reset by peer)
07:10:41 Join factor [0] (~factor@
07:11:50 Join n1s [0] (~quassel@rockbox/developer/n1s)
07:18:55 Quit n1s (Ping timeout: 245 seconds)
07:20:45***Saving seen data "./dancer.seen"
07:43:43 Quit linuxstb (Ping timeout: 248 seconds)
07:56:52 Quit user829385 (Quit: Leaving.)
08:30:42 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
08:40:46 Quit bluefoxx (Ping timeout: 248 seconds)
08:42:48 Join kadoban [0] (
09:06:42 Join bluefoxx [0] (
09:20:46***Saving seen data "./dancer.seen"
09:21:18 Quit kadoban (Ping timeout: 248 seconds)
09:29:10 Quit mikroflops (Read error: Operation timed out)
09:35:01 Join mikroflops [0] (
10:15:10 Quit bertrik (Ping timeout: 245 seconds)
10:22:10 Join user829385 [0] (~aoeu@
10:23:57 Quit Torne (Read error: Operation timed out)
10:25:00 Join [Saint_AndChat] [0] (~Saint]@
10:33:26kugel[Saint_AndChat]: I made some comments about your theme
10:34:16[Saint_AndChat]Ah, cool. Ill have a look when I'm not on my phone.
10:37:26[Saint_AndChat]...I hope they're good comments, I have criticism sensitive lenses in my glasses ;)
10:37:57kugelcan't browse FS from your theme?
10:38:04kugels/theme/phone/ (wtf)
10:38:36 Join liar [0] (
10:38:58[Saint_AndChat]Sure I can, but not whilst driving. :)
10:39:09kugelbut you can chat?
10:39:59[Saint_AndChat]Voice recognition for the win.
10:42:11 Join n1s [0] (~quassel@rockbox/developer/n1s)
10:42:12 Quit n1s (Read error: Connection reset by peer)
10:42:27 Join n1s [0] (~quassel@rockbox/developer/n1s)
10:49:17 Quit [Saint_AndChat] (Remote host closed the connection)
10:50:07 Join [Saint_AndChat] [0] (~Saint]@
10:51:59 Join bertrik [0] (
10:52:00 Quit bertrik (Changing host)
10:52:00 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
10:53:54 Join linuxstb [0] (~linuxstb@rockbox/developer/linuxstb)
11:00:54JdGordonkugel: did you see my ping a couple days ago about malloc backed buflib for raaa?
11:02:11JdGordonI'm going to commit jdgordon/rockbox/commit/adc88c5c86e7275751013c520053d667d9353824"> unless you've got a different solution being worked on?
11:02:37kugeluhm, please not
11:02:56kugelwhy do you need that now already?
11:03:09kugelthat means I cannot test buflib things on my phone
11:03:23JdGordonyou can disable it
11:03:40kugelanswer my question please
11:03:55JdGordon[Saint_AndChat]'s theme eats nearly half the buffer which is pretty shit
11:04:18[Saint_AndChat]It'll get smaller! :)
11:04:34kugelthat's a silly reason for that patch
11:05:02*JdGordon doesnt know why he bothers
11:05:19JdGordonif you're working on a different implementation then good
11:06:29kugelI'm not, not currently
11:06:43kugelI first want to get buflib and its clients working reliably
11:06:56kugeland I want to test on my phone, with svn builds
11:07:46JdGordonok, so disable it in your build
11:08:00JdGordonthats the reason I did it that way anyway, so it would 1000% transparent
11:08:01kugeldid you read svn builds?
11:08:58kugelthere's no point in the malloc patch just yet
11:11:00 Quit kevku (Ping timeout: 240 seconds)
11:11:32kugel[Saint_AndChat]: the memory footprint is indeed quite excessive
11:19:19[Saint_AndChat]There's only so much I can do about that.
11:20:02[Saint_AndChat]It'll get ~1.XMB smaller with string translation.
11:20:31[Saint_AndChat]Fonts are the killer, that I can't change.
11:20:44kugelwow, the titlebar stuff is alone 414k
11:20:49***Saving seen data "./dancer.seen"
11:21:14[Saint_AndChat]That's the specific translation I was talking of.
11:22:06kugelthe titlebar is pretty wasteful
11:22:12[Saint_AndChat]But, its the fonts, really. And no one wants bitmap fonts on a high(Ishmael) Res screen.
11:22:38[Saint_AndChat]And yes, it is...if you want me to say it a third time... ;)
11:22:45kugelyou add the golden thing to the backdrop and just render monochrome bitmaps (or even write it out)
11:24:27[Saint_AndChat]Is there something about development you're missing? I've never said its a: perfect, or b: finished ;)
11:24:39[Saint_AndChat]The look won't change, the guts will.
11:30:12[Saint_AndChat]I mainly focussed on function, appearance second. Now both are pretty much in an acceptable level my current changes focus on reducing its footprint. The titlebar not being drawn in the backdrop was a legacy of not all screens having .sbs support and an empty yellow bar looked odd. Changing it back is a step I'll take (planned for some time) when I do the titles with .Lang strings instead of bitmaps.
11:31:15amiconn'make reconf' is broken for voice builds at least
11:32:01amiconnIt duplicates several options in the CONFIGURE_OPTIONS= line, without obeying proper quoting
11:32:31amiconnIf you run 'make reconf' a second time, it will complain about unsupported options
11:32:55 Join kevku [0] (x@2001:470:1f15:1d8f::)
11:36:03amiconnAlso adding new languages in alphabetical order breaks 'reconf', but that can be tolerated imo
11:37:12kugelconfigure filters out options from cmdline that are passed explicitely via reconf, the remainder is passed as cmdline
11:37:49kugelbasically everything that can be selected interactively needs to be filtered out
11:40:07amiconnIt Atm it duplicates −−language, −−tts and −−tts-options, removing the quoting from the last one
11:40:18amiconn−−tts-options is even non-interactive
11:41:09amiconnAnd since I have e.g. −−tts-options='/Voice:Audrey16 /speed:2' it complains about an unsupported option /speed:2 next time
11:41:26 Join hilbert [0] (
11:41:54amiconnThis used to work att devocn time
11:42:25kugelI changed configure in the meantime, iirc
11:44:47 Quit hilbert (Client Quit)
11:45:23 Join hilbert [0] (
11:46:55 Quit factor (Ping timeout: 256 seconds)
11:48:20 Join pamaury [0] (
11:48:20 Quit pamaury (Changing host)
11:48:20 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
12:00:35 Quit [Saint_AndChat] (Ping timeout: 245 seconds)
12:01:03 Join TheLemonMan [0] (
12:06:51 Join MethoS- [0] (~clemens@
12:13:01 Join factor [0] (~factor@
12:14:28 Join [Saint_AndChat] [0] (~Saint]@
12:17:19bertrikI suppose we don't have .svg's for the usb logos, right?
12:19:11bertrikI need an usb logo of 96 pixels wide to make the sansa clip zip main firmware
12:19:36bertrikIt has a dithered shadow, so I can just resize it
12:21:25 Join benedikt93 [0] (~benedikt9@unaffiliated/benedikt93)
12:26:56amiconnMeh, current rockbox crashes on Ondio FM
12:27:02 Quit ender| (Ping timeout: 260 seconds)
12:27:28amiconnCPUAdrErr at 09014134. That's in tagtree
12:30:56amiconnSame on Recorder, just the address is slightly different: 09015104
12:31:13*amiconn suspects a buflib issue
12:32:30 Join y4n [0] (y4n@unaffiliated/y4ndexx)
12:34:21 Join Jerom [0] (~jerome@2a02:8420:216:a800:f66d:4ff:fe45:790f)
12:35:04amiconnThese addresses are within get_tag(), almost at the end
12:37:52*bertrik is confused by the <driver>-target.h system of #includes
12:38:46[Saint_AndChat]bertrik: no, no svg.
12:39:05bertrikbasically every backlight driver has the prototype for _backlight_init in a target-specific file, but the backlight.c source depends on it returning a bool
12:39:06[Saint_AndChat]Just add some width to the transparent section.
12:39:44bertrikI just resized the biggest one for now, we can fix that later
12:40:18[Saint_AndChat]It'll be off by some probably barely noticeable amount. What ate the deminsions? Ill do it for you this evening.
12:40:38pamaurybertrik: yeah, the <driver>-target.h is a bit weird, and sometimes the real prototype in another file. I think I run into a issue with i2c because of this
12:40:49bertrikThe clip zip has a 96x96x16 display
12:41:22[Saint_AndChat]Do not just a hack by adding transparency I mean. Ill make a new image from the original.
12:41:39[Saint_AndChat]Ill put it on fs if you're not around.
12:42:00 Join ender| [0] (
12:43:40[Saint_AndChat]Iirc, the closest x16 source image to that height is 128?
12:44:54[Saint_AndChat]...meh, ill have a look soon. I'd really love svg master images. That's probably my next "big" project.
12:46:04[Saint_AndChat]Replacing all the images with consistently coloured/sized ones derived from an svg master...dreams are free :)
12:46:08bertrikFor now, I just want the clipzip to compile and don't really care if some logo looks ugly or not, you're welcome to help with a better logo of course
12:46:27*[Saint_AndChat] nods.
12:47:40 Join matze` [0] (
12:49:38 Quit GodEater (Ping timeout: 244 seconds)
12:49:45 Join GodEater [0] (
12:49:51 Quit GodEater (Changing host)
12:49:52 Join GodEater [0] (~bibble@rockbox/staff/GodEater)
12:51:56*amiconn pings kugel for ideas
12:54:27amiconnLooks like the pointer argument (int *tag) is unaligned
12:55:18pamauryany arm expert: can I do a conditional return like "bxe lr" ?
12:56:38amiconnJust the condition 'e' doesn't exist
12:56:43amiconnThey're all two-letter
12:58:18pamauryyeah, eq, just wanted to know if that was possible with the bx call variant
13:00:57amiconnAlmost all arm instructions can be conditionally executed. Everything that has been defined in ARMv4 or earlier definitely can
13:01:51*amiconn hands pamaury a copy of arm_DDI0100I.pdf
13:02:08pamauryI have one, it's just faster to ask than to find it :)
13:02:39amiconnA3.16.7 lists the exceptions
13:03:21n1syeah, the ARM ARM is a very nice reference
13:05:03kugelamiconn: buflib doesn't touch the data
13:05:39kugelso it doesn't get unaligned through buflib usage
13:05:50amiconnDoes buflib align?
13:06:12amiconnAlso I wonder why it crashes right at startup. Didn't you test on hwcodec?
13:06:18kugelthe pointer buflib returns are word aligned es
13:06:25kugeli did
13:07:26amiconnWhat might be the problem then?
13:08:53amiconnLooks like a bisect is due :\
13:11:58 Join mamarley [0] (~quassel@2001:470:5:84d:221:6aff:fe19:4d0c)
13:12:37kugelamiconn: my hwcodec is a pain to use because I don't have the right batteries, I would welcome if you can do this
13:13:55amiconnr30018 is what definitely works. That's pre-buflib
13:20:52***Saving seen data "./dancer.seen"
13:21:03CIA-14New commit by bertrik (r30588): Sansa clipzip: make main build compile (plugins still disabled)
13:22:50pamaurythis fuze+ is a nonsense: if I load 0x100 bytes at 0x0 during boot, OF boots, if I write 0x100 bytes at 0x40000000, OF doesn't boot and in any case if I load the full rockbox bootloader I can run it without any problem
13:23:11CIA-14r30588 build result: All green
13:24:20 Join Stummi [0] (~Stummi@rockbox/developer/Stummi)
13:26:29 Quit soap (Quit: Leaving)
13:28:19pamauryhum, and now if I write at 0x42000000, OF display logo and keeps rebooting :-/
13:36:46 Quit Topy (Ping timeout: 248 seconds)
13:49:02 Quit Zarggg (Ping timeout: 248 seconds)
13:49:28 Join Zarggg [0] (
14:30:16 Quit pamaury (Quit: No Ping reply in 180 seconds.)
14:30:30 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
14:31:55 Quit Stummi (Quit: Bye!)
14:38:01 Quit mamarley (Remote host closed the connection)
14:43:46 Quit matze` (Remote host closed the connection)
14:46:15 Join domonoky [0] (~Domonoky@rockbox/developer/domonoky)
14:46:22 Join mamarley [0] (~quassel@2001:470:5:84d:221:6aff:fe19:4d0c)
14:47:15 Quit bluefoxx (Ping timeout: 245 seconds)
14:49:33 Join bluefoxx [0] (
14:52:40 Join lovasoa [0] (~lovasoa@2a01:e35:8a2e:8080:e2b9:a5ff:fe5b:ca7b)
14:56:22JdGordonthere isnt much point putting CACHEALIGN_ATTR on a sturct item is there?
14:56:37JdGordonespecially if that struct is being alloecd by buflib so could move around?
14:57:21JdGordonshld I add padding to the strcut to get that on a specific offset?
14:57:29kugelit only works for static structs IIUC
14:57:50gevaertsCACHEALIGN_ATTR is a hint to the linker, really
14:58:14gevaertsSo yes, static allocation only
14:59:41 Quit pamaury (Remote host closed the connection)
14:59:53JdGordonis bool a char or int?
15:01:49gevaerts"An object declared as type _Bool is large enough to store the values 0 and 1."
15:02:01gevaerts(bool is defined to _Bool in stdbool.h)
15:02:25JdGordonhmm, shold I bother [putting that in a union with an int to keep the rest word aligned?
15:02:30JdGordonor will gcc do that anyway?
15:03:38 Part lovasoa
15:03:43*kugel wonders what bool has to do with struct alignment
15:03:54n1sprimitive struct memvers are aligned to their size
15:04:44JdGordonso ints will always be word aligneD?
15:07:39pixelmabertrik: you could use one of the grey/mono USB logos for now. I'm not sure if I have an svg original for that one, I'll have a look later. The semi-transparent part at least has to be done 'by hand' to look correctly
15:08:23pixelmahmm no, those are probably too wide too
15:09:29gevaerts"Each non-bit-field member of a structure or union object is aligned in an implementation-defined manner appropriate to its type"
15:12:27 Quit Jerom (Quit: Leaving.)
15:19:31amiconnn1s: Regarding long blocks in vorbis - do you actually mean using a negative quality value -1 or was that a typo?
15:19:35CIA-14New commit by jdgordon (r30589): FS #12273 - use buflib for font storage. thanks to the testers :)
15:20:53***Saving seen data "./dancer.seen"
15:21:16n1samiconn: yes, -1, i think quality 0 will give them too
15:22:01CIA-14r30589 build result: 152 errors, 3 warnings (jdgordon committed)
15:24:32JdGordonbloody checkwps
15:24:47Lalufuhow many preexisting queues are there in rockbox? I know that iap uses button_queue, are there others?
15:24:53*JdGordon even did a charcell build locally to not have to deal with commit errors
15:25:16JdGordondamn nice fake delta though :)
15:25:28JdGordonunless of course you dont use fonts, thne its a real delta :)
15:28:43JdGordonthe font buffer went from statically allocated to dynamic
15:28:48JdGordonso the delta doesnt pick it up
15:29:23CIA-14New commit by jdgordon (r30590): Fix the build errors
15:29:50JdGordonLalufu: there are a few, why do you ask?
15:31:31CIA-14r30590 build result: 0 errors, 96 warnings (jdgordon committed)
15:31:49Lalufucurrently IAP uses the button queue to communicate internal events (command arrived via serial/periodic handler) from the interrupt/timeout handler to the functions doing the real work
15:33:40Lalufuthis interferes (amplified by a change I made) with the code that asks for a reboot after you installed a new bootfile
15:34:23Lalufuthat code sees the IAP internal events and interprets it as "any other key", cancelling the reboot
15:34:24CIA-14New commit by jdgordon (r30591): last warning
15:34:49JdGordonyeah, you shouldnt use the button queue :)
15:35:17Lalufuwhich queue would be appropriate for this?
15:35:24JdGordona new one
15:36:39CIA-14r30591 build result: All green
15:38:04LalufuOK. I'm a bit unclear on how the messages traverse the code, though.
15:38:44Lalufuiap.c queue_post()s a message to the button queue, and somehow it ends up in default_event_handler_ex() in misc.c, which then calls the worker function from iap.c.
15:38:47Lalufuhow does it get there?
15:39:49JdGordonyou probably want a dedicated thread and queue like playback or buffering use?
15:41:26Lalufuso create a private queue and spawn off my own handler thread that reads from the queue and calls the worker functions?
15:44:38LalufuHm, OK. that would be doable, I suppose.
15:45:48amiconnn1s: Here are the votes from the PP jury. PP5020 (ipod photo): vorbis_128 +4.5% (28.24 MHz -> 27.01 MHz), vorbis_40 +20.6% (31.12 MHz -> 25.80 MHz)
15:46:37amiconnPP5002 (ipod 2nd Gen): vorbis_128 +9.9% (30.48 MHz -> 27.73 MHz), vorbis_40 +31.5% (35.65 MHz -> 27.11 MHz)
15:47:01amiconnIt's obvious that PP5002 profits even more, as expected
15:47:42*JdGordon doesnt undertstand buflib_buffer_out/in
15:48:09amiconnWeird: Before the latest optimisation, lower quality needed more CPU power in vorbis.
15:48:18JdGordonis that the only way to resize a buffer (specifically to grow it)
15:48:19kugelJdGordon: it takes buflib buffer away from the front
15:48:53kugelgiving also works, so yes, resize
15:50:24JdGordonignoreing the poor names, how will it not usually fail if the callbacks arent called?
15:50:39kugelnot at all
15:50:46kugelit shouldn't be used in the core
15:51:05JdGordonok, well how can i grow a handle?
15:51:19kugelyou can't
16:03:45n1samiconn: nice
16:11:45 Quit antil33t (Read error: Connection reset by peer)
16:12:11 Join antil33t [0] (
16:16:11 Quit TheLemonMan (Quit: WeeChat 0.3.5)
16:17:39 Join TheLemonMan [0] (
16:24:55 Quit TheLemonMan (Quit: WeeChat 0.3.5)
16:28:37 Join TheLemonMan [0] (
16:30:29 Quit TheLemonMan (Client Quit)
16:32:19 Quit ender| (Ping timeout: 244 seconds)
16:44:32 Join ender| [0] (
16:52:18CIA-14New commit by jdgordon (r30592): Allow fonts to use smaller buffers than the default size. use font_load_ex() to speficiy the buffer size. If the font is already loaded with a smaller ...
16:53:17 Join TheLemonMan [0] (
16:54:44CIA-14r30592 build result: 190 errors, 0 warnings (jdgordon committed)
16:55:48*JdGordon really wants to kill checkwps :)
16:56:20bertrikyou can try it yourself, before committing, right?
16:56:38n1swhere's the fun in that?
16:57:31CIA-14New commit by jdgordon (r30593): fix checkwps
16:57:40*JdGordon 's computer cant do 100+ builds in 90s :)
16:58:05JdGordonwould be nice if "fix foo" commits wernt on the frontpage though
16:59:00bertrikbut the checkwps error is for basically all targets, so you don't have to do 100+ builds, just a few
16:59:46CIA-14r30593 build result: All green
17:01:01 Join bluefoxx_ [0] (
17:01:02 Quit bluefoxx (Ping timeout: 248 seconds)
17:01:48 Quit antil33t (Read error: Connection reset by peer)
17:02:14 Join antil33t [0] (
17:08:15 Join kadoban [0] (
17:09:34 Quit y4n (Disconnected by services)
17:09:53 Join y4n [0] (y4n@unaffiliated/y4ndexx)
17:12:58 Join gbl08ma [0] (
17:14:21gbl08mahello everyone. so now, fonts are stored in dynamic allocated RAM (buflib)? I see, the themes buffer has been reduced...
17:14:30gbl08ma(I read the SVN logs)
17:15:11gbl08maunfortunately, support for multiple fonts keeps broken. Guess I'll have to wait some more time... :)
17:16:30gbl08maand judging by the buflib debug screen, bmps are now also stored in dynamic memory... sweet
17:16:51Lalufuif I call queue_init() to create my own queue, who would/wouldn't I set register_queue?
17:18:45JdGordongbl08ma: why do you say mutliple fonts are broken?
17:20:07 Quit Bagder (Ping timeout: 256 seconds)
17:20:26gbl08mabecause photoSkins WPS still doesn't load correctly, instead it shows the system fallback WPS
17:20:35gbl08mabut hey, I haven't reloaded the theme
17:20:56***Saving seen data "./dancer.seen"
17:21:07gbl08manah, still broken
17:21:50n1sgbl08ma: that sounds like a parse failure
17:22:04Lalufuhm. register_queue gets me global events?
17:26:47gbl08man1s: yeah, but it's happening since some weeks ago, and I haven't changed the themes. people on IRC told me it was because multiple fonts support was broken
17:27:04gbl08maand in fact, it worked with themes that only loaded a single font, such as cabbiev2
17:28:07 Quit Galois (Ping timeout: 244 seconds)
17:34:51 Join Bagder [0] (
17:34:58 Quit Bagder (Changing host)
17:34:58 Join Bagder [241] (~daniel@rockbox/developer/bagder)
17:36:56gevaertsgbl08ma: I know my new wps and sbs use multiple fonts, and they work
17:39:42 Quit kadoban (Ping timeout: 260 seconds)
17:39:56gevaertshm, the backdrop just went away when switching back from the sbs...
17:40:52 Quit n1s (Ping timeout: 260 seconds)
17:42:02 Quit liar (Ping timeout: 260 seconds)
17:43:31 Join liar [0] (
17:44:18 Join powell14ski [0] (
17:47:04[Saint_AndChat]gbl08ma: is this a touch theme, or no?
17:47:55gbl08mait was a touch theme, I removed the touch parts as the nano2g has no touchscreen
17:48:17[Saint_AndChat]That's not needed.
17:48:17 Join Topy44 [0] (
17:48:40[Saint_AndChat]It woit will reject touch tags.
17:48:59 Join ReimuHakurei [0] (~kudo@2001:0:4137:9e76:38af:1ce9:b58f:2bf0)
17:49:49[Saint_AndChat]A tpouchscreen theme will work fine on a non-touch target.
17:52:05[Saint_AndChat]You must also have resized everything, yes? There's no touch target that shares the nano resolution.
17:52:27[Saint_AndChat]Perhaps you have out of bounds viewports still.
17:53:05 Join ReimuHakurei_ [0] (
17:53:31 Quit ReimuHakurei (Read error: Connection reset by peer)
17:53:33[Saint_AndChat]...though, it wouldn't load at all then....hmmm.
17:54:08[Saint_AndChat]I'd still need to see it to rule out author-error.
17:58:08*[Saint_AndChat] wonders if gbl08ma fell out of the conversation ;)
17:58:52[Saint_AndChat]Maybe I was wrong to think there was one :D
18:08:27 Join lovasoa [0] (~lovasoa@2a01:e35:8a2e:8080:e2b9:a5ff:fe5b:ca7b)
18:10:04lovasoaWhy isn't the "Database" menu localized ?
18:12:24 Join Keripo [0] (
18:12:36gbl08ma[Saint_AndChat] I'm just doing other things, and yeah, I wasn't following the conversation
18:12:57gbl08maphotoSkins has been ported and working for long
18:13:01gbl08maon the nano2g
18:13:15gbl08maI ported it myself, with i-audio 's help
18:13:26gbl08mait worked well for months till some weeks ago
18:13:55gbl08mai know touch tags are ignored, I just took them off because I wanted to save file space and to make the code easier to read
18:14:07gbl08ma(ignored on non-touch targets)
18:14:50[Saint_AndChat]Iiuc, it saves no space. The rejected tags aren't kept.
18:15:13[Saint_AndChat] gained ~2kb for music ;)
18:16:35[Saint_AndChat]Touch tags will be spat out at parse time, and not count for the skin buffer usage.
18:16:47gbl08maand then? on computers, having 2KB of disk free is the difference between having Linux booting or not :P
18:17:59[Saint_AndChat]2kb is a really generous estimate, too ;)
18:17:59*gbl08ma was installing software on a limited Linux system one day, and ended with 0 bytes free at the middle of the process = no Linux command works
18:19:19gbl08maplus, if each theme cuts its size for about 400 bytes, then 400 * 100 = 400000 bytes = 400KBytes saved if my brain isn't failing
18:20:03gevaertsIt is
18:20:24gevaertsYou're off by a factor 10, and moreover I don't see where that 100 comes from
18:20:32[Saint_AndChat]Also...where is this saving coming from?
18:20:39[Saint_AndChat]What saves?
18:20:57[Saint_AndChat]The themesite....?
18:21:06*[Saint_AndChat] is lost.
18:21:14gevaertsAlso, if you're talking about file sizes, think about cluster sizes
18:22:27gbl08ma100 = estimate number of themes an owner of an iPod video might have installed
18:22:46gbl08mayes it's true I'm dividing by 1000 and not by 1024
18:22:48[Saint_AndChat]Huge estime.
18:22:56[Saint_AndChat]Estimate even.
18:23:11gevaertsgbl08ma: 400 * 100 = 40000, not 400000
18:23:18gbl08macome on, I have all the themes for the nano installed on mine! why can't an iPod video owner do the same?
18:23:21pixelmalovasoa: if you mean the database "subitems" - those aren't using the Rockbox menu system but a tagnavi file. You can make that look like you want it to (even the number of items, how they are named etc.)
18:23:22gevaertsAnd my estimate is 3
18:23:54gevaertsBut it all doesn't matter, since the filesystem will make each file take 32K or so anyway
18:24:05gbl08magevaerts: I really had the impression that my brain was failing
18:24:26gbl08mawhy don't photoSkins for nano themes display correctly?
18:24:36pixelmanot sure if it would be possible to make the language strings available to tagnavi too, similar to the WPS (or how hard that would be)
18:24:57lovasoapixelma: thank you, I'll translate the menu myself. But why isn't it localized by default ?
18:25:02[Saint_AndChat]gbl08ma: my honest guess?
18:25:08[Saint_AndChat]Bad code.
18:25:11 Join lorenzo92 [0] (
18:25:26 Quit lorenzo92 (Client Quit)
18:25:28gbl08mathen why it loaded correctly before?
18:26:01[Saint_AndChat]The fact it worked until now doesn't mean the code was "good", more that it escaped some edge case.
18:26:30gbl08mawith arboxWidgets it doesn't load the WPS correctly either. I didn't code or modify arboxWidgets
18:26:32[Saint_AndChat]Which it now no longer does, apparently.
18:27:01[Saint_AndChat]Right, and it doesn't mean that code is "good" either.
18:27:17gbl08maI just tried arboxWidgets and iLike. None load correctly.
18:27:27gbl08ma(the WPS)
18:27:31gbl08mathe SBS loads fine
18:27:56[Saint_AndChat]iLike may well be broken too.
18:28:07[Saint_AndChat]I've not touched it in months.
18:28:21gbl08maI haven't touched photoSkins in months either
18:28:35gbl08mathen what happened in Rockbox for the themes to become incompatible?
18:28:55gbl08maI mean, almost every theme except some simpler ones don't have a proper WPS
18:29:03gbl08maat the moment
18:29:19gbl08maso, this means all the code except cabbiev2 and a few others is "bad code"?
18:31:23*gevaerts tries one of the seventeen themes called "photoskins something" on a nano2g sim and doesn't see an issue
18:31:25pixelmalovasoa: it can't switch depending on your language setting at all
18:32:02[Saint_AndChat]If its not "bad code", then, its "very weird shit". No theme, ever, has used more skin RAM than my RaaA themes...this I can guarantee, and they load perfectly. They use 3~5MB of images along, and 5 anti aliased fonts each.
18:32:10lovasoaBut it's a bug, isn't it ?
18:32:21gbl08magevaerts: what's the RB revision on that nano?
18:32:29gevaertsgbl08ma: *sim*
18:32:39gevaertsand current
18:32:48gbl08maI'm using unmodified r30593 on a real device
18:33:23[Saint_AndChat]are you changing between themes, this is known broken.
18:33:46[Saint_AndChat]Does it load correctly if you change a theme, then reboot?
18:34:09gbl08mawell, I changed between themes, but today when I turned on the iPod it didn't show correctly a photoSkins WPS
18:34:27gbl08mabut i'm rebooting again just to check
18:34:44[Saint_AndChat]I don't think the similar is using bufflib, which may be a factor in gevaerts testing.
18:35:09[Saint_AndChat]Bah, s/similar/sim /
18:35:30gevaertsiLike is broken in the sim for me
18:35:53gbl08maIMHO this problem only came due to some buflib change, but I'm not sure and IANAD
18:35:56gevaerts[Saint_AndChat]: why wouldn't it?
18:36:43gbl08maafter reboot, iLike keeps having a broken WPS
18:36:56[Saint_AndChat]gevaerts: I didn't think the Sims did "true" buffering compared to target.
18:37:33gevaertsThey do
18:38:27[Saint_AndChat]I can't think of a syntax change that would break ilike.
18:38:28pixelmaswcodec ones
18:38:45[Saint_AndChat]Can someone define "broken" for me?
18:39:01[Saint_AndChat]Re: iLike
18:39:03gevaerts[Saint_AndChat]: the wps doesn't load, so you get the built-in one
18:39:33[Saint_AndChat]Hmmmm....that's going to be a bastard to debug.
18:39:51[Saint_AndChat]My guess is exceeded buffer.
18:40:15[Saint_AndChat]There should be no syntax changes that hit it.
18:40:45gbl08maphotoSkins never exceeded the buffer, even before buflib came, and I don't think iLike ever did
18:40:47[Saint_AndChat]If the themesite says it works with current, it parses, and should load.
18:41:47gevaertsgbl08ma: the buffer is now a lot smaller, images and fonts aren't part of it any more
18:42:15gevaerts[Saint_AndChat]: iLike works again if I increase WPS_MAX_TOKENS
18:42:33gbl08maI've just checked and iLike doesn't fill even half the buffer
18:42:43[Saint_AndChat]Its rejecting all my comments.
18:43:08LalufuJdGordon: it works!
18:43:15[Saint_AndChat]I bet taking out comment lines fixes things, I do nothing major with tags.
18:43:16gevaertsgbl08ma: how can you check if it doesn't even load?
18:43:37Lalufuwould IAP qualify as a system thread, i.e. could I use PRIORITY_SYSTEM instead of PRIORITY_BACKGROUND?
18:43:54 Join T44 [0] (
18:44:24[Saint_AndChat]It doesn't fill even half the buffer *because* it doesn't load, gbl08ma
18:44:43[Saint_AndChat]And gevaerts just found out why.
18:44:55gevaerts[Saint_AndChat]: iLike uses 88K out of 80K available
18:45:05Lalufuthat's quite a trick.
18:46:18[Saint_AndChat]gevaerts: right,thanks. I bet its an easy fix. Ill remove my comments.
18:46:51gbl08ma"Just" move everything related to themes, even the tags, to a buflib alloc, and forget the size limit on themes (excluding, obviously, the device's ram limit) :D
18:46:57[Saint_AndChat]I thi needed to re-do the current screen code anyway.
18:47:10 Quit Topy44 (Ping timeout: 248 seconds)
18:47:13[Saint_AndChat]*think I
18:47:26gevaerts[Saint_AndChat]: right. 63.7K if I strip all comment lines in the wps, sbs and fms
18:47:41[Saint_AndChat]Thanks gevaerts
18:48:12[Saint_AndChat]gbl08ma: you don't "just do" anything with the buffering system. ;)
18:48:20gbl08masee, why those 2KB I talked about are important? :D and in this case, it was much more than 2KB...
18:48:35gevaertsgbl08ma: can you check if increasing WPS_MAX_TOKENS in apps/gui/skin_engine/wps_internals.h fixes things for you as well?
18:49:43gbl08maI don't have a rockbox build environment set up anymore, sorry
18:49:57[Saint_AndChat]gbl08ma: you obviously didn't read, or understand. Touch tags DONT COUNT here.
18:50:18[Saint_AndChat]They are rejected at parse time.
18:50:21gbl08maI had to reinstall linux on my desktop and lost the build environment, still didn't feel the need to setting it up again
18:51:10[Saint_AndChat]If its not a tag the target can use, its rejected and wont count towards skin ram usage.
18:51:35gevaertsI think the question is why comments are kept
18:51:43[Saint_AndChat]...but, that should be the case for comments.
18:51:51gbl08mathen why does it count comments? shouldn't it reject them
18:51:59gbl08maoops saint came first :)
18:52:02[Saint_AndChat]gevaerts: I just clicked to that too ;)
18:52:40[Saint_AndChat]It certainly used to reject tags it can't use (and comments).
18:52:52[Saint_AndChat]...something broke.
18:53:14gevaertsThe amount saved in the skin buffer by stripping comment lines exactly (well, within 1%) matches the amount of characters on those lines
18:53:50gbl08maso, it's not a problem in the themes themselves. I hope :)
18:55:09[Saint_AndChat]If tags aren't rejected, I wouldn't expect things to work for say, a condition that requires recording, but didn't give a false case.
18:55:34[Saint_AndChat](Not giving a false case is accepted)
18:57:08[Saint_AndChat]And, from what I know, even if I'm wrong about tags being rejected at parse time (I font believe I am), comments REALLY shouldn't be counted as tags and added to the skin buffer usage.
18:59:19*gevaerts submitted FS #12289
18:59:19fs-bluebot Skin comments take up space in the skin buffer (bugs, new)
19:00:21gevaertsJdGordon: do you have ideas on that?
19:01:05[Saint_AndChat]Not until late today, probably.
19:01:30[Saint_AndChat]Or 12 hours or so for you euro's.
19:13:08 Join Topy [0] (
19:16:50 Quit T44 (Ping timeout: 256 seconds)
19:21:00***Saving seen data "./dancer.seen"
19:21:55*[Saint_AndChat] runs from the skin parser shrieking...
19:23:11[Saint_AndChat]I thought I'd check if my theory about unused tags being rejected was true. But I think I just hurt my brain.
19:28:30 Join Galois [0] (
19:30:11 Join Horscht [0] (
19:30:11 Quit Horscht (Changing host)
19:30:11 Join Horscht [0] (~Horscht@xbmc/user/horscht)
19:30:32 Quit Keripo (Ping timeout: 260 seconds)
19:32:25 Join Torne [0] (~torne@rockbox/developer/Torne)
19:32:52 Quit Horschti (Ping timeout: 260 seconds)
19:34:18 Quit factor (Read error: Connection reset by peer)
19:38:43 Join fml [0] (
19:39:19fmlgevaerts: are you here? I need some help with a WPS
19:39:44*[Saint_AndChat] steps up.
19:40:22[Saint_AndChat]I'm free, fml
19:40:28 Join Keripo [0] (
19:41:28 Join T44 [0] (
19:41:45 Join [Saint] [0] (~st.lasciv@
19:41:54 Quit [Saint_AndChat] (Quit: AndChat - Android IRC Client -
19:42:01fmlSaint_AndChat: thanks! This is the first time I actually create a WPS with skinned lists. And I not quite understand how to do it. I created a SBS and defined a custom UI viewport there.
19:42:23[Saint]Right. What part is bothering you?
19:42:28fmlBut I can't make it to really draw the menues, file lists etc.
19:43:19[Saint]pastebin what you have.
19:43:36[Saint]I'll have a look. (no promises, though ;)
19:43:53[Saint]I'm still playing with the skinned lists myself.
19:44:20fml[Saint]: here's my .sbs:
19:45:06fmlWhere should I define the vieport for an item? In sbs? Or in wps?
19:45:18 Quit Topy (Ping timeout: 248 seconds)
19:45:39[Saint]in the .sbs.
19:46:04[Saint]You're declaring a custom UI viewport, but never calling it.
19:46:22[Saint]the %Vi line is unneeded.
19:46:52[Saint]Oh, wait. I'm mixing myself up.
19:50:58 Join factor [0] (~factor@
19:53:18[Saint]the CustomWPS example really is *terrible*
19:54:50kugelimprove it
19:55:40[Saint]If I knew a lot about think I'd be looking at the example syntax? ;)
20:01:47 Quit lovasoa (Ping timeout: 240 seconds)
20:06:28gevaertsfml: you're declaring a list area of 5 by 50 pixels and you then try to put 51x17-sized items in it
20:07:08fmlThis is what I have now: The list title is drawn as expected, but the menu items are not. They are drawn as per default.
20:07:31gevaerts%Lb is wrong
20:08:45gevaertsYour "item" viewport is 0,0,50,16, so you probably want %Lb(item,50,16) or so
20:09:02 Quit liar (Ping timeout: 260 seconds)
20:09:55 Join bluebrother [0] (
20:09:55 Quit bluebrother (Changing host)
20:09:55 Join bluebrother [0] (~dom@rockbox/developer/bluebrother)
20:11:27fmlgevaerts: I changed the dimensions but that didn't help. I don't quite understand why I have to specify the dimension in %Lb when the item size is already defined in the item's viewport.
20:11:38fmlIsn't it a double definition?
20:11:58 Quit fs-bluebot (Ping timeout: 248 seconds)
20:12:21gevaertsNo. (a) the viewport can have an offset, and (b) you're not restricted to one viewport. You can have several viewports (all with the same label) that make up one item
20:13:05 Join fs-bluebot [0] (
20:13:19fmlgevaerts: OK. That I understand. But it still does not work.
20:13:30 Quit bluebroth3r (Ping timeout: 256 seconds)
20:13:47[Saint]It *does* work...but, even I'm struggling a little here.
20:14:28[Saint]I usually need to "see" the code working before I get a hang of it. The CustomWPS example doesn't really help me in this respect ;)
20:15:14 Join Topy [0] (
20:16:26fmlWhat???!!! Is it possible that files are cached in a sim? I edited the file, started the sim anew and non't see the change!
20:16:43fmlI edited the .sbs
20:16:53[Saint]No, they are not cached.
20:16:54bertrikthe .sbs gets renamed IIRC
20:17:04bertrikor at least the .wps does
20:17:21gevaertsbertrik: huh?
20:17:42kugelfml: there's only dircache
20:18:08kugelthat caches, so don't see new/deleted files. but the data itself isn't cached so you should see changes
20:18:22 Quit T44 (Ping timeout: 248 seconds)
20:19:29fmlOk, I edited the wron file (a copy). So that's OK
20:19:50fmlBut the list (main menu) is still not drawn as expected
20:19:50bertrikgevaerts, fml sorry I think I'm confusing things, when working with wps'es some time ago something like this bit me in the back, don't know the details
20:20:58bertrikI think I edited the file in the simdisk, then did a make install, which overwrote my changes, but nevermind
20:23:36gevaertsfml: moving the %Lb line down to just above the list viewport makes it work for me
20:26:17fmlgevaerts: aha! Now the items are drawn OK. But I don't see the selection bar.
20:30:04 Quit Keripo (Read error: Connection reset by peer)
20:30:43gevaertsfml: the bar appears if you move that line around too, but I can't get it to work nicely
20:32:45fmlArgh! %LB is the *scroll* bar! I thought that was the selection bar!!! Shame on me!
20:33:20gevaertsIndeed! :)
20:34:35gevaertsYou have to draw the selection bar yourself, based on what %Lc says
20:36:13pixelmaso [Saint] lied to me when he said the scroll bar wasn't customisable yet? :)
20:36:37gevaertshm, this scroll bar is tricky!
20:36:40pixelmashould have been a ;)
20:36:47*gevaerts now has a private scroll bar for each list item
20:36:59[Saint]pixelma: You can only tell it to draw.
20:37:02[Saint]and where.
20:37:16[Saint]I thought you meant customizable as in "use your own images"
20:37:27pixelmawas there something about its width?
20:37:55[Saint]the width is a system setting, afaik, no?
20:38:20[Saint]gevaerts really knows more about skinned lists than I at present, and probably future ;)
20:38:43pixelmaok, better leave it for the default theme
20:39:05[Saint]I intend to gracefully step back from themes once this cabbie fiasco is over and done with :p
20:39:20[Saint](disclaimer, I probably won't )
20:39:20 Quit Bagder (Ping timeout: 245 seconds)
20:40:51fml%?Lc<%Vf(000000)%Vb(ffffff)|%Vf(ffffff)%Vb(000000)>%LN. %LT <−−- this code does not seem to draw a selected or non-selected item as I would expect
20:41:10pixelmabut basically I didn't like the... hmm, how to put it... sequence of rather "fat" scrollbar, small icons and then fat font. Maybe it also has to with missing spacing between these
20:45:27gevaertsfml: I know I had issues with %Vf and %Vb in combination with %Lc
20:46:42gevaertsOr in combination with %Lb in general I suspect
20:46:53fmlgevaerts: if I define separate viewports for selected and not selected items and conditionally display them in the item's VP, they are all displayed at a fixed position, i.e. not relative to the item's "box"
20:47:39gevaertsYes, I expect that sort of trick won't work. *Only* the item marked in %Lb is "move"
20:49:11gevaerts%Vs(color,000000) works as a replacement for %Vf(000000). That's something that still needs work
20:49:28gevaerts%Vf and %Vb should just work, but something gets confused
20:50:30pixelma%Vf/%Vb working conditionally was something that needed a fix later to make it work when it was introduced - but it was fixed and I use it in my WPS successfully
20:51:18 Quit TheLemonMan (Quit: WeeChat 0.3.5)
20:52:29fmlgevaerts: yesss! This works: %?Lc<%Vs(invert)|>%LN. %LT
20:52:43gevaertspixelma: I think the *conditional* part works fine. What happens is that the %Vf and %Vb as done by the first list item stick for the other list items, so I suspect it's something with this list item viewport being drawn multiple times (which I think isn't done anywhere else)
20:53:37*gevaerts should make a small screen theme with one list item per screen
20:54:26gevaerts[Saint]: haven't you experimented with grid lists yet for touchscreens?
20:55:15[Saint]Only briefly. Enough to get a WIP going.
20:56:34 Join Keripo [0] (
20:57:47[Saint]I ended up getting sidetracked and finding out I could do a bunch of (at the time, only dummy) UI viewports switched conditionally with a tabular menu kinda thing.
20:58:02fmlIt's a pity that %LN yields a 0-based number. It wouldbe more natural to make it 1-based.
20:58:45[Saint]"more natural" compared to...?
20:59:02fmlIt doesn't matter for conditional logic, but it's more natural for numbering list items.
20:59:41[Saint]the icons have always been this way, afaik. 32 icons, 0~31
20:59:43fml[Saint]: 0-based
20:59:52gevaerts[Saint]: this is displayed
21:00:11gevaertsfml: maybe we should just add this arithmetic tag and be done with it...
21:01:14fmlgevaerts: maybe. But I'd still make %LN 1-based. Adding the math tag would require enabling WPS vars for all targets, not only touch screens.
21:01:35gevaertsWould it?
21:01:37[Saint]its convoluted, but, couldn't you make it display the "correct" list number with relative ease with a conditional like %?LN<1|2|3|...etc>?
21:01:56gevaerts[Saint]: you know this can go up to tens of thousands, right?
21:02:14[Saint]sure, I didn't say it'd be pretty! :D
21:02:28gevaertsThere's this skin buffer size!
21:02:29fmlgevaerts: how else would you use the result of the math?
21:03:03gevaertsfml: you'd do something like %arith(+,%LT,1) and it would output the new number
21:03:15*gevaerts likes prefix notation :)
21:03:47fmlgevaerts: OK, that would be possible
21:04:41gevaertsI'm not opposed to making %LT 1-based though
21:05:00fmlI now think that skinned lists are a nice idea, but it's not ready for usage by a mere mortal.
21:05:21fmlThere are too many whistles and some things do not just work.
21:05:56gevaertsIf %LT is always used as text (as opposed to e.g. indexing bitmap strips, and you can't really do that anyway because you don't know how many items there will be), it would be silly to *always* have to add 1
21:09:35CIA-14New commit by gevaerts (r30594): Make %LT 1-based. This tag will likely only be used to print the item number (as opposed to e.g. using it as an index to a bitmap strip, which has the ...
21:09:59gevaertsfml: better? :)
21:10:31[Saint]that's a few of my themes broken.
21:10:57*[Saint] checks he's thinking of the right tag.
21:11:48*gevaerts would be surprised if [Saint] had more than one theme using %LT
21:11:55CIA-14r30594 build result: 0 errors, 88 warnings (gevaerts committed)
21:11:59gevaertsEven the one seems unlikely :)
21:12:02[Saint]AH...I was confusing %Lt
21:12:36[Saint]and even then, I was confusing that tags function with %Li
21:12:38CIA-14New commit by gevaerts (r30595): Fix silly mistake
21:12:56gevaertsIt's actually %LN
21:13:16*gevaerts apologises for the confusion
21:13:28[Saint]add it to CustomWPS, boy!
21:13:33*[Saint] cracks the whip
21:14:33[Saint]Is there, by chance, anyone present with an RaaA target, of the resolution 320x480?
21:14:50kugelI'm here
21:14:52CIA-14r30595 build result: All green
21:15:09fmlgevaerts: yes, that's better.
21:15:18gevaerts[Saint]: done
21:15:51[Saint]kugel: Ah, excellent. I generally avoid pinging people unless I've seen them active in the last while. Didn't see you in *-community 'til now.
21:15:59[Saint]I'll make a .zip for you to poke at.
21:16:32pixelmagevaerts: regarding the %Vf/b issue - does "line needs to scroll" or not make a difference?
21:16:49gevaertspixelma: no
21:20:15kugelwhat did you change?
21:20:28kugelI fixed the ffwd/rew and volume icons in the meantime
21:20:58[Saint]fonts, re: dislike of bold. positioning of some wrong valued items, touch areas.
21:21:04***Saving seen data "./dancer.seen"
21:21:42 Join T44 [0] (
21:22:21fmlThanks to all who helped me to sort things out!
21:22:25 Quit fml (Quit: ChatZilla 0.9.87 [Firefox 6.0.2/20110902133214])
21:23:22kugel[Saint]: sysfont the menus, ffwd/rew still wrong
21:24:03kugelsorry, I think I didn't mention it on the tracker, but those are cropped since they're mistakenly used as a bitmap strip
21:25:34 Join JesusFreak316 [0] (
21:25:34 Quit Topy (Ping timeout: 248 seconds)
21:26:05amiconnkugel: I've now bisected hwcodec (recorder)
21:26:23amiconnUnfortunately there seem to be two problems. First a hang, later the crash
21:26:33amiconnThe hang is caused by r30312
21:26:51bertrikamiconn, there's a similar problem now on flyspray, for the onda vx747
21:26:57[Saint]Ah, whoops. I added the wrong font to the theme. And I just figured where I messed up the FFWD/REW. They're *supposed* to be a bitmapstrip, but I included the wrong images.
21:27:17gevaertsJdGordon: can we get a %L<something> tag that draws the appropriate icon for the item? Or am I missing something and you can already do that?
21:27:20[Saint]kugel: 36pt was too big for the menus/lists?
21:27:38amiconnkugel: r30312 is a talk_init change introduced by you
21:27:39 Quit kevku (Quit: KVIrc 4.0.4 Insomnia
21:27:41[Saint]gevaerts: %Li
21:27:47kugel[Saint]: with your last zip i get sysfont
21:28:13gevaerts[Saint]: %Li and %LI get icon numbers
21:28:20amiconnr30308 was the main buflib commit, and there were several fixups
21:28:22kugelamiconn: I have a script to show links to viewvc for rXXX so you don't need to tell what the commit does :)
21:28:27[Saint]kugel: yeah, that's to be expected. Its looking for a font that isn't there. you mentioned on the tracker, though. Something about the fonts too big for the menus?
21:28:38[Saint]was it just bold that was the problem, or the height also?
21:28:59pixelmahmm, while I do have a vector version of the usb logo, I remember lots of "repairing" pixel by pixel even with the bigger bitmaps made of it
21:29:06amiconnkugel: Did you test your changes on hwcodec with voice enabled and a voice file installed?
21:29:28 Join petur [0] (~petur@rockbox/developer/petur)
21:29:50gevaertshm, maybe bitmap strips are usable here. I somehow thought they wouldn't be
21:30:21[Saint]gevaerts: yeah, sorry. I wasn't very descriptive.
21:30:42[Saint]yes, load the icons as a bitmapstrip, and display with %Li
21:30:56bertrikpixelma, yes, even the largest usb bitmap looks like a rendered vector image, but with some dithering for half-transparency done by hand (I presume)
21:31:37gevaerts[Saint]: I think I tried that and got off by one, so you'd need a separate copy. Not a huge problem since I'd want a different icon size depending on the screen anyway...
21:31:49pixelmabertrik: that for sure but I also think there was fixup for the rest of the graphics
21:31:59bertrikoh ok
21:32:02[Saint]"off by one"?
21:32:14[Saint]I have a theme (somewhere) with this sec.
21:32:23gevaertsyes, at least for %LI I got the wrong icon
21:33:24kugelamiconn: yep
21:33:47amiconnDo you have an idea why it might hang?
21:33:48[Saint]gevaerts: It's 0~32, 0 == "no icon"
21:33:55kugeland it exposed quite some bugs, but it worked shortly before committing
21:34:28kugelactually, IIRC, I had r30312 done before in my tree to fix some hwcodec bug
21:34:30amiconnBtw, hang means bootscreen is shown forever
21:34:36 Join gbl08ma_ [0] (
21:34:42[Saint]gevaerts: So, you likely don't want a declaration in the first %Li case.
21:34:48[Saint]hence your "off by one"
21:34:54amiconnIt also happens with 'voice menus' disabled (but voice files still present)
21:35:13kugelthe sim shows nothing I'm afraid?
21:35:33amiconnI didn't try a sim at all
21:35:40gevaerts[Saint]: a declaration? You mean as in build a huge conditional?
21:36:22[Saint]Yep. ;)
21:36:33gevaerts[Saint]: I want to use %xd(i, %LI), not %LI<kilometers of text>!
21:36:48 Quit gbl08ma (Ping timeout: 256 seconds)
21:36:52gevaertsAnd apart from this off-by-one, that works fine!
21:37:09[Saint]that won't work, afaik, as that doesn't know about things that don't have a first case. like %Li
21:37:21gevaertsSo I guess I'll just copy the icon strips :)
21:37:35gevaerts(and add an empty box)
21:37:47[Saint]that does it ;)
21:37:53[Saint]just c/p your identifier.
21:38:05gevaertsI *refuse* to do that!
21:38:32gevaertsIn fact, maybe I'll add this arithmetic tag just for this!
21:38:43[Saint]my way might be /long/, but its *right* dammit!
21:38:44amiconnHmm, now I tried building a sim. 'make install' doesn't work at all
21:38:51amiconnsimdisk/ is completely empty
21:39:09[Saint]adding a whole other iconset and adding a blank icon is a hack! :P
21:39:19gevaerts[Saint]: you're probably loading the iconset a second time anyway!
21:39:38*[Saint] hates to think of the amount of dirty skin hacks he's done.
21:39:48gevaertsI'm pretty sure the skin code isn't smart enough to say "Oh, wait, this is the active icon set, I don't need to load this"
21:39:52[Saint]gevaerts: Yes, but, it's not a copy of the main iconset.
21:40:12[Saint](in my case, at least)
21:40:19gevaertsSo you prefer unreadable code over *one* extra image space?
21:40:46 Join lovasoa [0] (
21:40:54[Saint]Bah, its like Jd, and now you, that call that unreadable ;)
21:41:10[Saint]your way, I can't see how many cases I'm dealing with without looking at the image.
21:41:16amiconn'make fullinstall' doesn't work either
21:44:04amiconnHow is this supposed to work nowadays?
21:44:55gevaertsamiconn: make fullinstall worked for me earlier today
21:45:12amiconnWell, here it doesn't, without any error message
21:45:20amiconnThis is r30312 though
21:47:30amiconn'make install V=1' reveals that is being run. It builds the whole .rockbox dir, and then immediately removes it again without copying into simdisk/
21:54:08[Saint]gevaerts: %xl(identifier)" and "%?LT<%?Li<identifier,%Li>>"
21:54:26[Saint]*identifier is you font
21:55:06[Saint]shave off the %?LT if you *always* want icons, and not for just lists that have titles
21:55:38*[Saint] remembered that this "find out how many cases a tag has" stuff is already in SVN
21:56:29 Join dan [0] (
21:56:35[Saint]so, "%xl(identifier)" and "%?Li<identifier,%Li>" would also work.
21:56:57 Nick dan is now known as Guest48954 (
21:56:58Guest48954can someone help me please
21:57:11[Saint]Guest42312: Only if you say what the problem is.
21:57:17[Saint]And only if its about Rockbox ;)
21:57:20Guest48954That would help.
21:57:31Guest48954I have a clip+
21:58:05Guest48954and i installed rockbox fone but when i plugged it in it restored it to the old proper firmware that came on the device
21:58:25[Saint]This is expected.
21:58:41[Saint]It didn't "restore", there's just no USB support in Rockbox for your target.
21:58:49[Saint]So we use the original firmware instead.
21:59:04Guest48954oh,how can i put files on without it reseting?
21:59:13[Saint]Read the manual regarding dual-boot, I don't know the key combination for this player off the top of my head.
21:59:46[Saint]Oh, whoops. You can't. It will always use the default firmware for USB.
21:59:57[Saint](until USB is enabled in ROckbox for your player)
22:00:02Guest48954then reset it on next boot, which is annoying?
22:01:00Guest48954even when i plug it into my computer to charge it resets to the original firmware
22:01:01[Saint]after its booted the OF, and you've transferred your files, turning the player off and then on again should boot Rockbox unless I'm mistaken.
22:01:15 Part lovasoa
22:01:34[Saint]and if you don't want it to enumerate, hold a key (mentioned in the manual), and it will enter a "charge only" mode.
22:01:42 Join n1s [0] (~quassel@rockbox/developer/n1s)
22:01:50[Saint]reading the manual, is a very good this. All this is described ;)
22:02:00Guest48954haha its so long xD
22:02:31 Join Bagder [0] (
22:02:31 Quit Bagder (Changing host)
22:02:31 Join Bagder [241] (~daniel@rockbox/developer/bagder)
22:02:47[Saint]It has a search function (the manual) if I'm correct.
22:02:58[Saint]you should be able to search for what you want to know about.
22:03:15Guest48954oh yeah rockbox does sort itself out when booting after transferring.
22:03:27Guest48954thanks somuch :)
22:03:50amiconngevaerts: I've solved the mystery
22:04:05amiconnSome time ago I must have done a symlinkinstall
22:04:16[Saint]Guest42312: No problem.
22:04:41amiconnAfter that, ordinary 'make install' won't work - and the symlink will be broken because ordinary install deletes the source (where the link is pointing to)
22:06:54amiconnkugel: Sim doesn't hang
22:09:03amiconnThat is, r30312 doesn't hang. Current svn *does* crash
22:09:34 Quit Guest48954 (Quit: CGI:IRC)
22:11:39amiconnr30595 to be specific. It crashes in strcat(), and the backtrace is unhelpful :(
22:13:34 Join dfkt|n [0] (
22:13:34 Quit dfkt|n (Changing host)
22:13:34 Join dfkt|n [0] (dfktn@unaffiliated/dfkt)
22:17:38amiconnkugel: r30595 sim only crashes if there's a voice file present for the currently selected language
22:18:05 Quit dfkt|n (Client Quit)
22:19:05amiconnBacktrace of a linux sim is more helpful
22:19:22 Join dfkt|n [0] (
22:19:22 Quit dfkt|n (Changing host)
22:19:22 Join dfkt|n [0] (dfktn@unaffiliated/dfkt)
22:19:30 Quit y4n (Quit: The world around me drops away, replaced by worlds being created and destroyed by my imagination.)
22:23:09amiconnIt's definitely voice related. The sim crashes in main.c:init_tagcache() which tries to talk
22:23:26 Join lovasoa [0] (~lovasoa@2a01:e35:8a2e:8080:e2b9:a5ff:fe5b:ca7b)
22:23:53amiconnkugel ^^
22:24:47lovasoaI just updated my iPod to the latest SVN, and the cover don't display anymore, with any skin.
22:24:54lovasoaIs it normal ?
22:25:41[Saint]No, its not.
22:26:15[Saint]I cannot say why its happeneing, but I *can* tell you it shouldn't happen.
22:27:06[Saint]when you say "with any skin", I assume you've checked cabbiev2 as well?
22:27:13lovasoaI meant : do you all have the same problem ?
22:27:33[Saint]You're the first one to report it.
22:27:36lovasoaYes, I tried cabbiev2
22:28:18[Saint]There was some other theme issues reported earlier, but the album art was not one of them.
22:28:24lovasoaIs there anything I can do to help you solve the problem?
22:29:08[Saint]Do you recall the last version you had on your player where album art did display?
22:29:14[Saint]that would help.
22:29:33[Saint]If you can build your own Rockbox binaries, it would help even more.
22:30:07[Saint](as then you'd be able to mention the exact revision (through trial and error) that does work)
22:30:30lovasoaActually, I am building my own binaries
22:31:04 Join Topy [0] (
22:32:38lovasoaI am currently rebuilding from scratch, after a make clean
22:33:50 Quit T44 (Ping timeout: 248 seconds)
22:34:54[Saint]do you know how to bisect a known good revision?
22:35:16lovasoaNo, I don't
22:35:30lovasoaIs it long ?
22:35:30*gevaerts has seen album art on his gigabeat F running r30593
22:35:54lovasoaBecause I'll have to go to bed in a short time...
22:36:35[Saint]pick a revision to build, try it, and label it good or bad, depending on the result. If it is bad, go back forther, if it is good, pick a revision halway between the revision you just built and the current (known bad) revision, build that, and label it good or bad, depending on the result.
22:36:39[Saint]etc. etc.
22:36:41pixelmamy phone running 30582 shows album art too
22:37:03lovasoaIt seems logical
22:37:03pixelmaerr... I read gevaerts' revision number wrong
22:37:30lovasoaBut I don't know how to do that with svn...
22:38:13[Saint]svn up XXXXX (where XXXXXX is the desired revision) (iirc)
22:38:14bertriksvn update -r<revision>
22:38:23[Saint]ah, I forgot the -r
22:38:28*amiconn wonders how this *ever* worked
22:39:12amiconninit_tagcache() is called before audio_init() (which in turn calls talk_init()) so voice is uninitialized when init_tagcache() is running
22:39:47bertriksheer coincidence then?
22:40:24lovasoaI'm gonna try that
22:41:47lovasoaI can start with r30589, as it's the first recent revision that touches the skin engine...
22:41:56amiconnIt's strange that swcodec doesn't crash there (same init order)
22:43:04Lalufuprotip: check out the git-tree and use git-bisect
22:43:34lovasoaMy compilation just finished
22:43:40amiconnHmm, actually init_tagcache() only tries to talk on swcodec
22:43:42lovasoaAnd it works
22:43:47amiconnSo the problem must be sth else
22:43:54[Saint]from 30589, assuming its good, it'll take like...what ~5 builds, max?
22:44:24lovasoaI just recompiled from scratch (after a make clean) and it works
22:44:30kugel"amiconn wonders how this *ever* worked" <−− I wondered this many times
22:44:40[Saint]lovasoa: *facepalm* ;)
22:45:25[Saint]Heh, we'va all done it, I'm sure.
22:45:58[Saint]But, next time some "very weird s**t" happens, and you're building your own binaries...try one of the ROckbox provided binaries forst.
22:46:11[Saint]*first, too.
22:46:32lovasoaBut there are some themes that still don't work...
22:48:21amiconnStill the sim only crashes when a voice file is present for the currently selected language
22:49:10amiconnOh, and this test revealed another bug: Voice doesn't work when the built-in english is used, only when 'english.lng' is selected explicitly
22:49:23amiconnHwcodec sim to be precise
22:49:39amiconn(the latter bug probably affects all targets and sims)
22:49:51pixelmawasn't there also backdrop corruption or so reported with voice enabled?
22:54:00amiconnIt almost looks like something is overwritten that causes the sim to crash in a system function, which is being called from sleep(). The full backtrace is here (player sim, but the bt looks the same for recorder sim apart from the actual addresses):
22:56:07 Quit benedikt93 (Quit: Bye ;))
22:56:19amiconnHmm, probably false alarm. Does pausing a thread send a SIGUSR1 when running under gdb?
22:57:51amiconnI've now continued a few times, and then it finally crashes with SIGSEGV elsewhere:
22:58:33amiconnThis is in tagtree - where current svn crashes on actual hwcodec target as well. Not in exactly the same place though
22:59:18 Quit Keripo (Ping timeout: 276 seconds)
22:59:30kugelamiconn: yes, gdb pauses on SIG*
23:00:11amiconnNo, I mean whether pausing (from within thread_unix.c) sends a SIGUSR1
23:00:16gevaertsamiconn: are you using configure −−sdl-threads? If not, maybe try it. Native threads and gdb can be a bit annoying if you don't set up everything right
23:00:42kugelamiconn: thread creation sends this yes
23:00:44amiconnAnyway, the SIGSEGV is probably the more interesting one
23:03:04amiconngevaerts: I tried now. It doesn't send SIGUSR1 this way. The SIGSEGV backtrace is the same as with native threads
23:03:50 Part lovasoa
23:05:28kugeldo "handle SIGUSR1 pass nostopp" in gdb to ignore it
23:10:09 Quit juri_ (Ping timeout: 260 seconds)
23:10:27 Quit mystica555 (Ping timeout: 260 seconds)
23:11:21kugelamiconn: I could observe the behavior that voice only works after selecting english.lng
23:11:54 Join mystica555 [0] (
23:12:04amiconnIt also works if you select deutsch.lang and have a deutsch.voice etc
23:12:19 Quit JesusFreak316 (Remote host closed the connection)
23:12:28amiconnThis also used to work. It's not really as annoying as the crash though, and probably not too difficult to fix
23:14:00amiconnUnfortunately the backtrace doesn't tell the parameters strcpy() is called with
23:15:28kugelNULL or out-of-bounds most likely
23:15:30amiconnIt does point towards problems with memory allocation in tagtree since the buflib move though
23:16:03amiconnIiuc tagtree_strdup() allocates some memory and stores the string there
23:16:57 Join T44 [0] (
23:18:38 Quit Topy (Ping timeout: 248 seconds)
23:19:06amiconnbreak tagtree.c:278 -> print dest -> $2 = 0x158 <Adddress 0x158 out of bounds>
23:21:00 Join pamaury [0] (
23:21:00 Quit pamaury (Changing host)
23:21:00 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
23:21:07***Saving seen data "./dancer.seen"
23:22:34amiconnNo out-of-bounds in an swcodec sim, as expected
23:22:51 Join gbl08ma__ [0] (
23:23:34kugeltagtree shouldnt be different between those
23:25:35 Quit gbl08ma_ (Ping timeout: 245 seconds)
23:26:17kugel[Saint]: do you have another zip?
23:26:36amiconnApart from a few SWCODEC checks for autoresume it isn't
23:26:52amiconnSo somehow it is getting a wrong address space from buflib on hwcodec
23:28:47[Saint]kugel: this should fix the icons, and the FFWD/REW "buttons(?)"
23:28:59-->"ods 200" received from cgthayer (
23:29:11[Saint]a dedicated iconset for the port is still needed.
23:29:19amiconnIt's also voice related, because it doesn't happen if there is no voice file for the current language
23:30:26[Saint]maybe, for the time being, I should include the (possibly slightly too large) iconset from the 480x800 port.
23:31:26[Saint]I'm not 100% certain what "sane" icon sizes are for each port, though. 18x18, ~3-x3-, and ~4-x4-
23:35:19kugel[Saint]: main menu font is too large for my test (however, I'm using my list spacing patch) and I don't like bold generally
23:35:44kugelI dont think a UI should be in bold throughoutly
23:35:51[Saint]It shouldn't be bold.
23:36:00 Join mystica555_ [0] (
23:36:04[Saint]it isn't, in fact.
23:36:15kugelah it's so big it looks bold
23:36:19kugelbut on the wps there's bold
23:36:22[Saint]perhaps you need to re-load the .cfg
23:36:41[Saint]Yeah, the bold in the .wps is intentional.
23:36:56amiconnAh, talk_init() *is* called before init_tagcache() (within settings_apply() ) - this explains how talking in init_tagcache() works
23:37:11[Saint]and, a 36pt font will look *terrible* with your list patch no doubt.
23:37:14amiconnIt also explains how voice can affect tagtree_init() at all
23:37:19amiconnkugel ^
23:37:52kugelI guess that means I should work on the patch so it's default before your theme goes in then :)
23:38:05[Saint]good luck with that ;)
23:39:44kugelwhen going back from the playlist viewer the pop doesnt work anymore
23:40:09[Saint]in needs 4 seconds to go false.
23:40:19[Saint]I can (possibly) reduce that value.
23:40:45kugelhm why is that?
23:41:08[Saint]conditions are rechecked when the screen is re-entered.
23:41:14[Saint]I could maybe work around it.
23:41:48kugelthe pop up creates an cleared line at the bottom of the aa
23:42:10[Saint]Gah! that bug again. thankyou.
23:42:29[Saint]that's a skin variable bug...which I have NFI how it affects AA.
23:43:07[Saint](well, worked around with an ugly hack)
23:43:30[Saint]the "fix" for that is Jd's territory. I've just side-stepped it.
23:43:37LalufuI'm a bit puzzled here. If the GUI is in the main menu, with the cursor on "Files
23:43:47kugelmoving X a few lines down?
23:43:59[Saint]kugel: correct.
23:44:08Lalufu", and I send a BUTTON_RC_PLAY event, the player changes into the file menu, as if I had pressed Select (all this on an iPod)
23:44:11Lalufuany idea why?
23:44:31kugel[Saint]: that's a long standing bug and has to with the draw order I think
23:44:41Lalufusending BUTTON_RC_PLAY while in playback mode pauses/resumes as expected
23:44:48kugelimages are drawn last, except albumart, IIRC
23:45:13kugelAA is only redrawn once in a while
23:45:35kugelI don't know why those empty lines are created though
23:45:41[Saint]kugel: It only seems to be *one* line that does this, the skin variable check/set/increment declarations.
23:45:59[Saint]It took me a *very* long time to convince Jd it wasn't my bug ;)
23:47:02[Saint]I seem to remember he knew why it was happening, but he forgot about it I think, and so did I, because I discovered how to work around it.
23:47:35[Saint]which is "draw more blank lines than the total height of the viewport its declared in"
23:48:19[Saint]so, the draw bug stil happens, but, it happens OOB so you can't see it.
23:57:27 Quit domonoky (Read error: Connection reset by peer)
23:57:49 Join Keripo [0] (

Previous day | Next day