| *** klausade has joined #maemo | 00:00 | |
| *** povbot has joined #maemo | 00:11 | |
| *** rev has joined #maemo | 00:13 | |
| *** sxpert has quit IRC | 00:20 | |
| *** luck has quit IRC | 00:22 | |
| *** sxpert has joined #maemo | 00:23 | |
| *** spect has joined #maemo | 00:23 | |
| *** andrunko has quit IRC | 00:26 | |
| *** etrunko has quit IRC | 00:26 | |
| *** _shawn_ has joined #maemo | 00:27 | |
| *** povbot` has quit IRC | 00:30 | |
| *** mallum has quit IRC | 00:32 | |
| *** __shawn has quit IRC | 00:34 | |
| *** saerdnaer has quit IRC | 00:35 | |
| *** bstock_ has quit IRC | 00:36 | |
| arj | still looking for ppp people :D | 00:37 |
|---|---|---|
| Tak | You down with PPP? | 00:39 |
| arj | I'm having a little trouble getting it running | 00:41 |
| sxpert | ok. | 00:42 |
| * sxpert has now a brand new debian system... | 00:42 | |
| koen | 'brand new' and 'debian' in the same sentence? | 00:44 |
| ssvb | sxpert, i hope it does not have a brand new kernel ;-) | 00:45 |
| sxpert | ssvb, apt-get install scratchbox installs 1.0.7 or something | 00:45 |
| sxpert | koen, brand new as in, just ripped the package :D | 00:45 |
| ssvb | by the way, is it possible to upgrade gcc for use with maemo sdk to something better than 3.4.4 (release) (CodeSourcery ARM 2005q3-2)? | 00:49 |
| ssvb | it seems a bit too buggy for me | 00:49 |
| sxpert | you probably can, but you may have issues with c++ though | 00:50 |
| *** kb7sqi has joined #maemo | 00:50 | |
| *** k-s has quit IRC | 00:53 | |
| *** ajturner has quit IRC | 01:05 | |
| *** florian has joined #maemo | 01:11 | |
| florian | re | 01:14 |
| Tak | mote! | 01:15 |
| sxpert | ok. progress. | 01:15 |
| *** chenca has left #maemo | 01:15 | |
| sxpert | now, I can't seem to start xephyr... | 01:15 |
| sxpert | it complains about DISPLAY not being set | 01:15 |
| Tak | are you using the PC profile? | 01:16 |
| *** bergie has quit IRC | 01:16 | |
| sxpert | I think I did. | 01:17 |
| *** fab has quit IRC | 01:17 | |
| * sxpert tries again | 01:19 | |
| *** koen has quit IRC | 01:21 | |
| sxpert | Tak, can't get it to work | 01:25 |
| *** matt_c has quit IRC | 01:25 | |
| Tak | do you, in fact, have DISPLAY set? | 01:25 |
| sxpert | before logging into scratchbox, yes | 01:25 |
| Tak | have you tried resetting it once you're logged in? | 01:26 |
| sxpert | if I manually set DISPLAY, I get permission errors from the server | 01:26 |
| Tak | try `xhost +localhost` or `xhost +` (outside of scratchbox) | 01:26 |
| sxpert | xhost + did it | 01:27 |
| sxpert | is there something I could hack ? | 01:28 |
| Tak | what do you mean? | 01:28 |
| sxpert | ok. I hacked my .profile to copy .Xauthority where it needs copied :D | 01:30 |
| Tak | ah | 01:31 |
| sxpert | I'm all set :D | 01:34 |
| sxpert | guess I can start hacking on my app now :D | 01:34 |
| pvanhoof | florian, ping | 01:52 |
| florian | pvanhoof: yessss | 01:52 |
| nomis | pvanhoof: I am Simon btw. :) | 02:03 |
| pvanhoof | ok | 02:03 |
| pvanhoof | :) | 02:03 |
| pvanhoof | hi :) | 02:03 |
| * nomis waves | 02:03 | |
| pvanhoof | florian, you are of course allowed to forward that mail to nomis | 02:03 |
| florian | pvanhoof: ok, great | 02:04 |
| pvanhoof | just don't start forwarding those prices/rates to just anybody :p | 02:04 |
| pvanhoof | :) | 02:04 |
| nomis | florian: I think I'll visit KC tomorrow afternoon. | 02:05 |
| florian | heh | 02:06 |
| florian | nomis: even better :-) | 02:06 |
| |tbb| | i wish i can resize those applets by myself ;( | 02:07 |
| *** terral has quit IRC | 02:08 | |
| *** matt_c has joined #maemo | 02:12 | |
| pvanhoof | nomis, did I already replied your mail? | 02:12 |
| pvanhoof | it's marked as important, but I forgot whether I already replied it or not | 02:13 |
| pvanhoof | if so, the reply is in the sent-box of my email client at my current customer | 02:13 |
| pvanhoof | hrmm .. sucks , having two email clients | 02:13 |
| pvanhoof | anyway, if I didn't .. florian now knows some stuff :) | 02:14 |
| nomis | pvanhoof: I got a reply, thanks. | 02:14 |
| nomis | pvanhoof: I however did not yet reply in return. | 02:14 |
| pvanhoof | oh right | 02:14 |
| pvanhoof | I found it | 02:15 |
| *** matt_c has quit IRC | 02:18 | |
| *** qgil has left #maemo | 02:22 | |
| *** matt_c has joined #maemo | 02:23 | |
| *** skorianez has joined #maemo | 02:23 | |
| *** |tbb| has quit IRC | 02:24 | |
| pvanhoof | florian, are you kc guys checking out fosdem in februari? | 02:27 |
| pvanhoof | hey, we Belgians also visit LinuxTag and GUADEC/Stuttgart :), you should! :) | 02:28 |
| florian | pvanhoof: i'm pretty sure we'll be there... | 02:28 |
| pvanhoof | cool | 02:29 |
| nomis | for me it depends a bit on university schedule, but I believe it will be possible. | 02:29 |
| florian | pvanhoof: the only reason i missed the last one was Jonas :-) | 02:29 |
| pvanhoof | we don't eat babies in Belgium. You can bring him :) | 02:29 |
| pvanhoof | well, most Belgians don't | 02:29 |
| *** sp3000 has quit IRC | 02:29 | |
| pvanhoof | not afaik, but maybe they are keeping me out of their secret community of baby eaters | 02:29 |
| florian | pvanhoof: heh | 02:30 |
| florian | well.. time to get a little rest | 02:32 |
| florian | good night | 02:32 |
| pvanhoof | good night | 02:32 |
| *** florian has quit IRC | 02:33 | |
| * nomis heads for the bed as well. Night. | 02:39 | |
| *** buck68 has left #maemo | 02:41 | |
| *** pvanhoof has quit IRC | 02:51 | |
| *** arj has quit IRC | 02:57 | |
| *** rev has quit IRC | 03:18 | |
| *** mndctrl has quit IRC | 03:31 | |
| *** skorianez has quit IRC | 03:32 | |
| *** mproctor has quit IRC | 03:37 | |
| spect | good night | 03:37 |
| *** spect has quit IRC | 03:38 | |
| *** Ryback_ has quit IRC | 03:43 | |
| *** tko_ has quit IRC | 03:56 | |
| *** tko has quit IRC | 03:59 | |
| *** tko has joined #maemo | 04:01 | |
| *** _handful_ has joined #maemo | 04:38 | |
| *** maddler has quit IRC | 04:49 | |
| *** _handful_ has quit IRC | 05:35 | |
| *** marceoli has joined #maemo | 05:40 | |
| *** jsun_ has joined #maemo | 05:55 | |
| *** mat__ has joined #maemo | 05:57 | |
| *** mat has quit IRC | 05:57 | |
| *** mat__ is now known as mat | 05:58 | |
| jsun_ | hi - | 06:05 |
| jsun_ | I just followed the maemo 2.1 tutorial to install sdk pc | 06:05 |
| jsun_ | the tutorial says Xephyr is preinstalled. | 06:06 |
| jsun_ | but I can't seem to find it. | 06:06 |
| jsun_ | anybody knows what is wrong here: me or the tutorial? :) | 06:06 |
| *** rkaway3 has joined #maemo | 07:15 | |
| *** marceoli has quit IRC | 07:26 | |
| *** _shawn_ has quit IRC | 07:30 | |
| *** rkaway2 has quit IRC | 07:31 | |
| *** papa_booiiing has quit IRC | 07:40 | |
| *** booiiing has joined #maemo | 07:44 | |
| *** jsun_ has quit IRC | 07:51 | |
| *** skodde has quit IRC | 08:10 | |
| *** __shawn has joined #maemo | 08:42 | |
| *** bergie has joined #maemo | 08:44 | |
| *** ab has joined #maemo | 08:59 | |
| *** bergie has quit IRC | 09:10 | |
| *** bergie has joined #maemo | 09:38 | |
| *** koen has joined #maemo | 09:42 | |
| *** ssvb has quit IRC | 09:48 | |
| *** klausade has quit IRC | 10:06 | |
| *** sp3000 has joined #maemo | 10:09 | |
| tigert | what distro? | 10:15 |
| tigert | if you can not find it, check that your commandline has the correct case for it, also it wasnt installed at least in ubuntu. make sure you are not trying to run it from scratchbox | 10:17 |
| tigert | since its not in there afaik. | 10:17 |
| dpb | Who are you talking to? jsun_? The person who's not here? | 10:18 |
| Jaffa | Morning, all | 10:19 |
| dpb | Xephyr was atleast in the 1.1 x86 maemo rootstrap | 10:19 |
| tigert | oh heh | 10:34 |
| tigert | i filter out joins and parts | 10:34 |
| tigert | less noise, but that is the side-effect ;) | 10:35 |
| *** dralex has joined #maemo | 10:51 | |
| *** Guardian has joined #maemo | 10:55 | |
| *** Milhouse has quit IRC | 11:05 | |
| *** phil|work is now known as phil|sleep | 11:17 | |
| *** fab has joined #maemo | 11:18 | |
| *** ssvb has joined #maemo | 11:21 | |
| *** keesj has joined #maemo | 11:21 | |
| keesj | Hi | 11:22 |
| keesj | I want to try the fast memcopy. can anyboy help with me with that? | 11:24 |
| *** AD-N770 has joined #maemo | 11:33 | |
| *** Fatal has quit IRC | 11:42 | |
| *** Fatal has joined #maemo | 11:44 | |
| *** sKaBoy has joined #maemo | 11:52 | |
| *** sp3000 has quit IRC | 11:53 | |
| *** sKaBoy has quit IRC | 11:57 | |
| *** Guardian has quit IRC | 11:59 | |
| *** bergie has quit IRC | 12:02 | |
| *** jobi has quit IRC | 12:28 | |
| *** bilboed has joined #maemo | 12:29 | |
| *** qgil has joined #maemo | 12:41 | |
| *** bergie has joined #maemo | 13:03 | |
| *** florian_kc has joined #maemo | 13:04 | |
| *** morty has joined #maemo | 13:06 | |
| florian_kc | hi all | 13:08 |
| *** morty has quit IRC | 13:10 | |
| ssvb | hi keesj, what kind of help do you need with fast memcopy? | 13:18 |
| *** AD-N770 has quit IRC | 13:18 | |
| *** qgil has quit IRC | 13:19 | |
| *** mgedmin has joined #maemo | 13:21 | |
| *** jobi has joined #maemo | 13:25 | |
| *** lardman has joined #maemo | 13:35 | |
| *** Ryback_ has joined #maemo | 13:44 | |
| *** X-Fade_ is now known as X-Fade | 13:48 | |
| *** Ryback_ has quit IRC | 13:56 | |
| *** Ryback_ has joined #maemo | 13:56 | |
| *** dwd has joined #maemo | 14:16 | |
| *** koen has quit IRC | 14:16 | |
| dwd | So... If I want to make a fool of myself by asking basic [py]GTK questions, should I ask here or am I better off in #pygtk? | 14:16 |
| *** lardman has quit IRC | 14:19 | |
| * mgedmin would guess #pygtk | 14:19 | |
| mgedmin | unless these questions are specific to maemo | 14:19 |
| mgedmin | uh, there's no #pygtk on freenode | 14:20 |
| dwd | mgedmin: It's hiding away on gimpnet. | 14:21 |
| myren | whats your client? | 14:21 |
| *** etrunko has joined #maemo | 14:27 | |
| *** bilboed has quit IRC | 14:32 | |
| keesj | re | 14:33 |
| keesj | ssvb, I am not a asm guru :) How do I know if the new memcpy is used | 14:34 |
| keesj | I have included the .h file in my project and replaced memcpy with memcpy32. but I would like to build SDL and SDL_gfx with the memcpy feature | 14:35 |
| ssvb | keesj: memcpy32 is not a replacement for memcpy (it is a variant that deals with 32-bit values instead of bytes), you just need to use arm9_memcpy instead of memcpy | 14:42 |
| ssvb | also probably ARM optimized version of SDL is needed, I did not check what version of SDL is used in maemo and if it has any patches, but upstrem SDL did not seem to have any ARM optimizations the last time I checked it | 14:44 |
| keesj | ssvb, I am currently using it to clear the screen (copy image data to the main buffer) | 14:44 |
| *** |tbb| has joined #maemo | 14:44 | |
| ssvb | keesj, do you see any performance improvement using it? | 14:45 |
| keesj | no , not at all | 14:45 |
| ssvb | can I have a look at your sources? | 14:46 |
| keesj | I still have the same framerate , | 14:46 |
| keesj | you want everything? just the copy code? | 14:46 |
| keesj | (I am working on a xmoto port) | 14:47 |
| ssvb | everything would help to see the whole picture, but just the copy code should be enough | 14:47 |
| ssvb | can you provide some link to it? | 14:48 |
| keesj | i can make a tar | 14:48 |
| *** waite has joined #maemo | 14:48 | |
| keesj | hmm :) | 14:53 |
| ssvb | keesj: by the way, did you try to profile xmoto to see what parts of it are the most performance critical and need optimizations the most? what is the current framerate? | 14:54 |
| keesj | ssvb, yes i did | 14:54 |
| keesj | but I only can profile on my machine (i386) | 14:56 |
| keesj | I have a framerate of 4.44 | 14:56 |
| ssvb | keesj, in this case, screen copy should not take that much time and bottleneck is most likely somewhere else | 14:58 |
| ssvb | I also find profiling on x86 quite useful, and valgrind (with callgrind tool) is excellent for this | 14:59 |
| keesj | there are heavy parts that use the ODE library and floating point , but I don't want to change that part | 14:59 |
| ssvb | my athlon xp 2400+ seems to run in valgrind even a bit faster than the same program natively on Nokia 770 :) | 14:59 |
| keesj | idem here | 15:00 |
| keesj | ssvb, this is the current profiler output http://www.keesj.dds.nl/prof.png | 15:01 |
| keesj | I am working on the hlineColor . (takes 33% right now) . | 15:02 |
| keesj | it's a loop that I want to replace with a few memcpy's http://paste-it.net/706 | 15:04 |
| *** mat|work has quit IRC | 15:05 | |
| keesj | if I remove that critial part the next thing that take time is the clearScreen (5%) | 15:08 |
| keesj | So I guessed that a fast memcpy would improve speed a lot. | 15:09 |
| *** mat|work has joined #maemo | 15:09 | |
| ssvb | keesj, that loop is very inefficient (also getting remainder is very expensive on arm, it does not have a dividion instruction) | 15:11 |
| ssvb | if you rewrite it without the use of % operator, it will become a lot faster | 15:11 |
| keesj | so that might be the bottleneck | 15:12 |
| waite | keesj, Take a look at http://216.239.51.104/search?q=cache:zFL2GPyvcHwJ:graphics.stanford.edu/~seander/bithacks.html+c+hack+modulus+without+%25&hl=en&gl=us&ct=clnk&cd=1&ie=UTF-8 | 15:14 |
| waite | There are a few versions of Computing modulus with a division operator | 15:15 |
| waite | s/with/without/ | 15:15 |
| keesj | your all a great help! | 15:16 |
| waite | The routine is really short so make it a function and just substitute the % with modulus(z, (sdl_gfxTexture->surface->w )) | 15:17 |
| pokute_ | hmm. if sdl_gfxTexture->surface->w is a constant for the duration of the loop, you can replace it with a if (z >= sdl_gfxTexture->surface->w) z = 0 ; Could be a bit faaster | 15:17 |
| ssvb | keesj, did you try memcpy that is currently commented out in that code? | 15:17 |
| tigert | hmm | 15:17 |
| keesj | pokute_, I just did that | 15:18 |
| ssvb | keesj, do you have any estimate how many pixels are usually copied in that loop? | 15:19 |
| pokute_ | Gah, I was slow... | 15:19 |
| *** orduz has joined #maemo | 15:19 | |
| keesj | this method is used in a filledPolygon method | 15:20 |
| keesj | the last step is to draw on the screen so the value is something between 1 and 800 | 15:21 |
| waite | keesj, here: http://paste-it.net/707 | 15:21 |
| keesj | the code used to draw a single color, but I have changed it so it draws textures | 15:21 |
| ssvb | keesj, the fastest code for this loop would be to use arm9_memcpy16 function from https://garage.maemo.org/plugins/scmsvn/viewcvs.php/trunk/fastmem-arm9/fastmem-arm9.S?root=mplayer&rev=82&view=markup | 15:22 |
| ssvb | it was specifically designed to copy 16-bit pixel data | 15:22 |
| keesj | waite, perhaps even inline? | 15:23 |
| waite | Sure. It would be fne inline also | 15:23 |
| keesj | waite, and that is less cpu then the if statement? | 15:23 |
| waite | It relpaces this stmt: ( z % (sdl_gfxTexture->surface->w ) | 15:24 |
| ssvb | keesj, in this particular code 'if ' statement should be faster | 15:24 |
| waite | I was focusing on line 6 of your paste | 15:24 |
| keesj | waite, pokute_ proposed if (z >= sdl_gfxTexture->surface->w) z = 0 | 15:24 |
| waite | I agree | 15:25 |
| keesj | because it's constant? | 15:25 |
| pokute_ | keesj: another proposal. | 15:25 |
| pokute_ | having separate start and z is useless. | 15:25 |
| pokute_ | or wait, nope, won't work. :-/ | 15:26 |
| keesj | so what the code actually is trying to do is to is to draw a line using a texture. | 15:27 |
| waite | Iti is early and I have not had a full coffee, but how does: if (z >= sdl_gfxTexture->surface->w) z = 0 satisfy the algo? | 15:27 |
| pokute_ | ahh. maybe if you make. z = z + start before the loop. then have [z] instead of [start + z] and have if (z >= (precalculated_sdl_gfxTextureSurfaceWPlusStart) z = start ; | 15:29 |
| pokute_ | get rid of one addition calculation per loop run. | 15:29 |
| pokute_ | Or forget that. :-D | 15:30 |
| pokute_ | I got a more clear proposition. | 15:30 |
| keesj | the problem is that it doesn't start a 0 | 15:31 |
| pokute_ | have (Uint16 *) precalculatedpixelarray = (Uint16*) sdl_gfxTexture->surface->pixels[start] ; before the loop | 15:32 |
| pokute_ | that way you can use precalculatedpixelarray[z] inside the loop | 15:32 |
| pokute_ | after all, start is a constant variable, it's useless to have inside the loop. | 15:33 |
| pokute_ | argh. paste-it doesn't seem to allow for adding annotations to existing pastes | 15:35 |
| keesj | no that will not work this is an Image of what I am trying to do http://www.keesj.dds.nl/te.gif . I need start inside the loop | 15:35 |
| pokute_ | hmm? Why not? | 15:36 |
| keesj | start is the start "x" value of texture. but z need to be reset before texture->w iterations | 15:37 |
| pokute_ | (void *) a = 0xc0debeef ; (void *) b = a [20] ; In this case a [20 + x] and b [x] point to the same position, no? | 15:38 |
| keesj | alright if (z == texture->w) then z -= texture->w | 15:38 |
| pokute_ | no need for that. :-) as z never rises more than 1 at a time. | 15:38 |
| pokute_ | Look at the latest proposal, I already said that the second proposal was more cleaner. :-) | 15:39 |
| pokute_ | http://paste-it.net/708 | 15:41 |
| *** orduz_ has joined #maemo | 15:41 | |
| *** orduz has quit IRC | 15:41 | |
| pokute_ | And there's still more to optimize. :-) | 15:42 |
| pokute_ | But that would require some more obscure code. | 15:44 |
| pokute_ | depends on the size of sdl_gfxTexture->surface->w. If it's more than 50 then more optimization would make sense. | 15:45 |
| keesj | pokute_, it depends on the texture some are smaller and other are larger | 15:48 |
| *** mgedmin has quit IRC | 15:49 | |
| pokute_ | well, it would be a lot more complicated, so better try to optimize somewhere else first. :-) | 15:49 |
| keesj | I have a bit of information overdose right now! | 15:49 |
| ssvb | keesj, you mentioned that original code did not use textures, maybe it is better to try it first and add texturing if you have spare cpu power :) | 15:50 |
| keesj | ssvb, it was opengl :) | 15:50 |
| keesj | So I have rewritten most of the calls to SDL_gfx | 15:51 |
| keesj | this was the first run , it was not using filled or textured surfaces http://wiki.xmoto.free.fr/index.php?title=Image:Xmoto_n770_2006_11_18.jpg | 15:52 |
| ssvb | keesj: I see, anyway, I will be around tomorrow and will have much more spare time on weekend, so I can help with optimizing some critical code using ARM assembly | 15:52 |
| keesj | that would be great! | 15:52 |
| keesj | thanks for you time | 15:53 |
| pokute_ | assemblör <3 | 15:53 |
| pokute_ | keesj: can you test the code I pasted? | 15:58 |
| pokute_ | tho I'm surprised that the line to draw is axis-aligned on the texture. | 16:03 |
| pokute_ | makes everything so much easier. | 16:04 |
| pokute_ | and that much weirder | 16:05 |
| |tbb| | is it possible to play divx on n770 smoothly? | 16:35 |
| Jaffa | |tbb|: yes, DivX is the best format for playback on the 770 | 16:38 |
| Jaffa | |tbb|: http://maemo.org/maemowiki/VideoEncoding | 16:39 |
| *** dot_j has joined #maemo | 16:41 | |
| adoyle | |tbb| - I've just managed to get great video results with the Perl script here: http://www.bleb.org/software/770/#encode - I used the 2nd example command on a Qucktime .mov and produced a great looking .avi | 16:42 |
| Jaffa | adoyle: glad you found it useful :) | 16:42 |
| adoyle | Is that your page, Jaffa? | 16:42 |
| Jaffa | adoyle: indeed it is | 16:42 |
| adoyle | Well, then, thanks! | 16:42 |
| Jaffa | My pleasure :) | 16:43 |
| adoyle | Here's a 1.5 minute clip of Doc Edgerton talking about strobe lights - http://museum.mit.edu/mwow/files/tour-content/test/test-005.avi | 16:43 |
| *** benzea has joined #maemo | 16:54 | |
| *** dralex has quit IRC | 17:01 | |
| *** matt_c has quit IRC | 17:08 | |
| keesj | pokute_, I tried the code for some reason it all doesn't help much on i386 , | 17:15 |
| keesj | perhaps its the function that is just called to often | 17:15 |
| pokute_ | keesj: make it inline? | 17:17 |
| pokute_ | But optimization on a processor with high-level out-of-order execution is a whole different beast than optimization for an embedded processor. | 17:18 |
| keesj | Yes , will try it now on the n770 | 17:18 |
| pokute_ | some arm instruction set clock cycle reference chart would help. | 17:22 |
| pokute_ | For example, could it be faster to have the z count down towards zero and use use if (z == 0) z = somefunvalue; | 17:23 |
| pokute_ | probably if z--; sets the zero-flag. :-) | 17:23 |
| keesj | whel it helped the framerate to 6 on the n770 | 17:25 |
| pokute_ | what was it previously? | 17:25 |
| keesj | 4.44 | 17:25 |
| pokute_ | and compared to what code? | 17:25 |
| keesj | the code I first posted | 17:25 |
| *** xan has joined #maemo | 17:26 | |
| keesj | first http://paste-it.net/706 now http://paste-it.net/709 | 17:26 |
| *** obergix[home] has joined #maemo | 17:27 | |
| pokute_ | So if it took originally 33% then now it takes a significantly less... Time to find optimization target somewhere else now. :-) | 17:27 |
| keesj | thanks a lot , | 17:28 |
| pokute_ | Or just to profile it again. :-) | 17:28 |
| pokute_ | I remember that cachegrind had some wonderful per-instruction profiling or something as ridicilously useful. | 17:29 |
| dwd | pokute_: cachegrind doesn't do ARM, though, does it? | 17:29 |
| pokute_ | Curses, but maybe there's some profiling stuff on an arm emulator. | 17:30 |
| *** marceoli has joined #maemo | 17:30 | |
| pokute_ | an arm emulator would be about the best places to put an profiler. :-) | 17:30 |
| *** marceoli has quit IRC | 17:31 | |
| pokute_ | keesj: Do you mind profiling the program again with this code? :-) | 17:31 |
| pokute_ | depending on the results, I'll give make an even better optimized version. :-d | 17:32 |
| Tak | keesj: are you testing with or without sound? | 17:33 |
| *** booiiing has quit IRC | 17:37 | |
| Tak | I gained a ton of performance going from sdl audio to esd (for raw pcm) | 17:37 |
| Tak | and if it's using any kind of compressed audio, a gstreamer interface will probably help massively | 17:38 |
| ssvb | keesj, I'm glad you could improve performance somewhat, as I told you earlier, integer division is slow on arm (it has no special instruction for that and needs software emulation), that's why it makes such a huge difference | 17:39 |
| *** matt_c has joined #maemo | 17:39 | |
| ssvb | Tak: gstreamer is probably not a very good choice for sound effects in games | 17:40 |
| *** booiiing has joined #maemo | 17:40 | |
| * Tak shrugs | 17:40 | |
| *** bergie has quit IRC | 17:40 | |
| Tak | depends how they're doing the sound effects | 17:41 |
| ssvb | it may work for background music though | 17:41 |
| ssvb | gstreamer has quite a large and unpredictable latency on nokia 770 | 17:42 |
| Tak | IME that's true of audio in general on 770 | 17:42 |
| *** skodde has joined #maemo | 17:44 | |
| ssvb | keesj, if you have lots of floating point math, that affects performance quite significantly and profiling on x86 does not provide a whole picture | 17:46 |
| *** benzea has quit IRC | 17:46 | |
| ssvb | I wonder if it is easy to assign somewhat higher cost to floating point instructions in callgrind to make the statistics look close to what would be observed on arm | 17:47 |
| *** kender has joined #maemo | 17:48 | |
| kender | good evening | 17:48 |
| Tak | I still have a whole friday to get through :-/ | 17:50 |
| ssvb | keesj, you can also try to compile the game with -pg gcc option and use gprof to get statistics from running it on arm | 17:50 |
| ssvb | unfortunately when I tried this with mplayer, it did not work unless all the optimizations are disabled :( | 17:51 |
| ssvb | keesj: mplayer just crashed, but some small simple code samples worked ok, maybe you will have more luck with xmoto | 17:51 |
| *** fab has quit IRC | 17:51 | |
| *** klausade has joined #maemo | 17:53 | |
| sxpert-work | tigert: have an LD-3W linked question. have you tried putting it under your hat and roam around with the 770 in hand ? | 18:12 |
| kender | sxpert, xDD | 18:15 |
| pokute_ | Probably not really needed now, but once it becomes colder. | 18:15 |
| kender | xD | 18:17 |
| kender | so, you are nokia 770adict? | 18:17 |
| *** marceoli has joined #maemo | 18:20 | |
| *** fab has joined #maemo | 18:23 | |
| tigert | sxpert, i have an arm pocket on my jacket | 18:25 |
| tigert | works fine for the gps | 18:25 |
| keesj | ssvb, I will try the -pg gcc , option , there are many floating point calucations but I remove them the game will be a totaly different game | 18:27 |
| pokute_ | keesj: the point is to use fixed points, but replacing all instances in a existing program is going to be a pain. | 18:28 |
| keesj | yes ,and ode used it , and i don't want to rewrite ode :) | 18:29 |
| pokute_ | But you could try to use pixel doubling for drawing the backgrounds. Then switch to normal mode to redraw the polygon edges and the player. :-) | 18:29 |
| keesj | I din;t know this was possible | 18:30 |
| ssvb | pokute_, did you try it? | 18:30 |
| *** florian_kc has quit IRC | 18:30 | |
| keesj | pokute_, on i386 it all did not make such a big difference | 18:31 |
| pokute_ | keesj: what made/didn't make difference? | 18:32 |
| ssvb | keesj, using pixel doubling, you effectively reduce screen resolution, so you program works faster as it needs to handle less pixels | 18:32 |
| ssvb | that may be an interesting option | 18:33 |
| ssvb | though I doubt that you can easily get pixel doubled background and normal size sprites | 18:33 |
| keesj | pokute_, the %usage of the hlineColor was 33% and now 31% . But I don't remember the framerates | 18:33 |
| keesj | I hope that I can draw the background with one memcpy option it already works | 18:34 |
| pokute_ | keesj: figures that the difference on x86 would be almost the same. | 18:34 |
| keesj | it doesn't :) | 18:36 |
| pokute_ | :-) | 18:36 |
| ssvb | keesj, by the way, what is stored in 'pixx' variable in your sample code? | 18:37 |
| *** marceoli has quit IRC | 18:38 | |
| keesj | pixx = dst->format->BytesPerPixel; | 18:38 |
| keesj | so 2 in this case | 18:38 |
| keesj | but I will never get a better performance then memset(xxxxx,xxx,xxx) right? | 18:39 |
| ssvb | i see, performance of this function can be doubled in this case | 18:39 |
| ssvb | right | 18:40 |
| ssvb | to get better performance, you should write not 16-bit values to memeory, but combine pairs and write them as 32-bit values | 18:41 |
| keesj | by using int32? | 18:41 |
| ssvb | you will have to unroll loop for that | 18:41 |
| pokute_ | ssvb: and maybe take account the alignment | 18:42 |
| pokute_ | A gain of 5 percent could be archieved for having textures in memory for both alignments. :-) | 18:43 |
| ssvb | anyway, it is interesting to know how much time does this function actually take on arm in order to know whether it is worth optimizing it more | 18:43 |
| keesj | pokute_, the texture data is created on the same machine, I think I could just use int32 on either platforms | 18:44 |
| ssvb | right | 18:44 |
| ssvb | but you can just use optimized memcpy and not bother about that | 18:45 |
| pokute_ | ssvb: that doesn't work very well with texture wrapping. | 18:46 |
| ssvb | so your loop will be replaced by a few memcpy calls, you will have to calculate how many consecutive pixels can be copied and call memcpy on each iteration | 18:46 |
| pokute_ | Tho it would be feasible to replace the loop with a loop running memcpy's | 18:47 |
| keesj | ssvb, Yes that was my first idea | 18:47 |
| keesj | but I did learn a lot today! | 18:47 |
| pokute_ | And I really would like to know how much performance penalty there is for switching pixel doubling on/off and whether if affects the existing framebuffer. | 18:48 |
| ssvb | pokute_, that's exactly what I tried to suggest, maybe my english is not good enough :) | 18:49 |
| * Jaffa bahs at blogs which don't allow comments (e.g. Devesh's) unless you're also a blogger on the same system. | 18:49 | |
| ssvb | switching pixel doubling should cost nothing | 18:49 |
| ssvb | that's just a flag that is used in screen update kernel ioctl | 18:50 |
| ssvb | if i remember this correctly, need to check omapfb sources again | 18:50 |
| ssvb | but anyway, you will have to turn it on/off only once per frame, so it does not matter | 18:51 |
| pokute_ | Yeah, I was thinking of switching the pixel doubling for filling the backgrounds. | 18:52 |
| keesj | without the code the framerate is 6.26 so there is still something to be gained , but I will have to look elsewhere | 18:52 |
| ssvb | you will have problems with blitting sprites in this case | 18:52 |
| pokute_ | keesj: Without the code? You mean texture copying code? | 18:53 |
| keesj | yes , i removed the whole loop | 18:53 |
| pokute_ | ssvb: Just use mipmapped one. :-) | 18:53 |
| pokute_ | 5/4 of memory usage is not bad for vast improved speed | 18:54 |
| pokute_ | If you want to retain both versions of the texture. | 18:54 |
| ssvb | keesj, so i guess optimizing this function more you can gain about 0.1 fps improvement | 18:55 |
| pokute_ | Indeed. No use in optimizing it more. | 18:56 |
| *** minra has joined #maemo | 18:57 | |
| pokute_ | keesj: comment out entirely the level drawing and then report the result. :-) Should give a lot more info on where to fix. | 18:57 |
| minra | ive been using the 770 for my only internet computer (no computer at home) and it is truly incredible | 18:57 |
| pokute_ | minra: It rocks, I don't really miss my pc when I'm visiting at my father's place. Have irc, videos, web and internet radio. Tho when I run out of videos to watch... | 18:59 |
| minra | is the 770 graphics a typical ARM simple framebuffer type device? or is some kind of acceleration available? | 18:59 |
| Tak | agreed - the only thing I need is a wifi antenna | 18:59 |
| minra | yeah usb hard drive with battery would be nice | 18:59 |
| Tak | there's not enough wifi here | 18:59 |
| pokute_ | minra: there's supposed to be some kind of graphics acceleration and the comments I've heard is that it sucks. | 19:00 |
| minra | europe has limited wifi power. in the usa 1000mw is allowed, so we are at a disadvantage | 19:00 |
| Tak | ah | 19:00 |
| Tak | "here" is "Texas" in my case ;-) | 19:00 |
| ssvb | keesj: ... or try to run -pg build and get gprof statistics :) | 19:00 |
| minra | if we could get 2x scaling accelerated then the emulators (mame, fceu) would be much more usable | 19:00 |
| Tak | agreed, although fceu is very usable right now | 19:01 |
| pokute_ | minra: well, 2x (double pixel mode) is available. | 19:01 |
| minra | i am in contact with the xmame maintainer... if i can help him with info on 2x accel, that would be great | 19:01 |
| pokute_ | And I've understood that 2x is without performance penalty whatsoever. | 19:02 |
| minra | btw. i am playing spacwar, gravitar, joust with no problems. robotron is also ok | 19:02 |
| pokute_ | minra: The maemo wiki has stuff on that. | 19:02 |
| minra | that would be awesome | 19:02 |
| minra | ok will look thanks | 19:02 |
| *** sp3000 has joined #maemo | 19:04 | |
| ssvb | minra: it is here - https://maemo.org/maemowiki/GameDevelopment | 19:05 |
| * Tak is looking at it now | 19:06 | |
| * Tak is the xmame/fceu/vba maintainer ;-) | 19:07 | |
| pokute_ | ooh! | 19:07 |
| *** tko_ has joined #maemo | 19:10 | |
| *** orduz_ is now known as orduz | 19:10 | |
| *** orduz has left #maemo | 19:10 | |
| *** benzea has joined #maemo | 19:11 | |
| *** sab_osdl has joined #maemo | 19:11 | |
| minra | what do i need to install to get "ping"? | 19:12 |
| *** phil|sleep is now known as philipl | 19:15 | |
| ssvb | brb | 19:15 |
| *** ssvb has quit IRC | 19:15 | |
| keesj | I generated a gmon.out but it doesn't contain information file is missing call-graph data . I will need to learn gprof :) | 19:18 |
| Tak | is there no libXsp on debian/i386 ? | 19:18 |
| *** dottedmag has joined #maemo | 19:19 | |
| *** buck68 has joined #maemo | 19:35 | |
| *** Guardian has joined #maemo | 19:41 | |
| minra | i am trying to find the .deb for links web browser... and "links" is a TERRIBLE search engine term. any help? | 19:46 |
| buck68 | lynx? | 19:46 |
| *** buck68 has quit IRC | 19:50 | |
| *** minra has quit IRC | 19:54 | |
| *** GreySim has joined #maemo | 19:58 | |
| *** minra has joined #maemo | 19:59 | |
| *** ssvb has joined #maemo | 19:59 | |
| minra | does nokia show any interest in integrating your contributions into the stock distro? | 20:00 |
| *** marceoli has joined #maemo | 20:00 | |
| minra | oh oh oh i got a question. can i backup and restore the builtin flash with dd ? | 20:01 |
| minra | i mean, theres no alternate boot device afaik. if i hose the root fs, i have to send it in. | 20:02 |
| *** marceoli has quit IRC | 20:05 | |
| *** marceoli has joined #maemo | 20:05 | |
| *** marceoli is now known as _handful_ | 20:07 | |
| sxpert | tigert, ok. | 20:07 |
| tigert | minra: you can reflash it yourself | 20:09 |
| sxpert | tigert, I'm looking into creating maemo-surveyor :D | 20:10 |
| sxpert | which would be similar to those trimble windows CE based surveying tools | 20:10 |
| *** _handful_ is now known as _Handful_ | 20:10 | |
| *** minra has quit IRC | 20:13 | |
| *** _Handful_ has quit IRC | 20:14 | |
| *** marceoli has joined #maemo | 20:14 | |
| *** minra has joined #maemo | 20:15 | |
| *** marceoli is now known as _Handful_ | 20:15 | |
| minra | ok ctorrent kills my 770 | 20:15 |
| *** _Handful_ has quit IRC | 20:16 | |
| *** _Handful_ has joined #maemo | 20:16 | |
| minra | hi _Handful' | 20:17 |
| _Handful_ | hi minra | 20:17 |
| minra | i am still learning what not to do with the 770 | 20:18 |
| _Handful_ | :D | 20:18 |
| minra | ctorrent is definitely out | 20:19 |
| |tbb| | dont take a shower with him | 20:19 |
| _Handful_ | what have you been doing with it that had negative consequenses? | 20:19 |
| _Handful_ | :D | 20:19 |
| minra | i just got the screen to draw really strange horizontal noise pixels. never seen anything like it... might have been low battery | 20:20 |
| minra | btw i was online with bluetooth keyboard for six hours before i needed to plug it in... | 20:21 |
| minra | that is way over spec. | 20:21 |
| minra | low brightness | 20:21 |
| |tbb| | minra which bt keyboard do you own? | 20:22 |
| _Handful_ | very nice | 20:22 |
| _Handful_ | today I will run some battery tests with regular audio player and canola player | 20:22 |
| minra | thinkoutside stowaway | 20:22 |
| _Handful_ | I'm just looking for the newest 770 here | 20:22 |
| _Handful_ | 2 of them actually | 20:22 |
| minra | all new keyboard layouts are painful, but i am managing quite well with this one. the tactile response is great. | 20:23 |
| minra | build precision is a bit low, but it hasnt broken yet and i am getting about 5 days from 2x AAAs | 20:24 |
| _Handful_ | better than the ones I've tested | 20:24 |
| minra | i couldnt imagine using a 770 without a keyboard | 20:25 |
| minra | might be fun to get synergy working, to run kbd/mouse from a pc | 20:26 |
| |tbb| | port it minra port it :) | 20:26 |
| *** sp3000 has quit IRC | 20:29 | |
| GreySim | Can you use a mouse with the 770? I mean, even synergy aside (though I <3 synergy), could you use, say, a bluetooth mouse? | 20:30 |
| *** _Handful_ has quit IRC | 20:32 | |
| minra | am i the only person here who sees 4 minute lags on freenode? | 20:32 |
| tigert | I found one issue since installing canola | 20:33 |
| tigert | the daemon seems to take just enough memory, | 20:33 |
| tigert | that it prevents me from unmounting the memory card (swap on mmc enabled) | 20:33 |
| tigert | and since there seems to be no way to stop the background process, it's a tad annoying :) | 20:34 |
| robtaylor | what is canola? | 20:35 |
| roope | Hi. | 20:36 |
| roope | thanks tigert for the template. i'll try something out with it soon. A white text-dark-bg skin. | 20:37 |
| roope | soon = next week. :) | 20:37 |
| etrunko | robtaylor: http://maemo.org/maemowiki/ApplicationCatalog2006#head-d6a4f55c7636804666d28f271bfce2a5232f7e3a | 20:38 |
| *** GreySim has quit IRC | 20:43 | |
| |tbb| | do you all have also the problem not to configure canola while you are not connected to the internet? | 20:44 |
| minra | i will never understand amarok / canola type people :) | 20:47 |
| *** obergix[home] has quit IRC | 20:53 | |
| *** mallum has joined #maemo | 20:55 | |
| keesj | minra, syngery works on the 770. | 20:55 |
| tko_ | maybe I'm forgetting what amarok looks like, but for me comparing canola to amarok seems weird | 20:55 |
| keesj | and what keyboard do you use? | 20:55 |
| |tbb| | keesj, was this a question? | 20:55 |
| |tbb| | the first one | 20:56 |
| keesj | |tbb|, no it's a fact ,synergy run on the 770 | 20:56 |
| |tbb| | do u have a deb file for me? | 20:57 |
| minra | keesj deb deb deb? | 20:57 |
| *** ab has quit IRC | 21:00 | |
| keesj | hmm | 21:00 |
| keesj | I have a deb here , perhaps it's a it2005 release | 21:02 |
| keesj | I will recompile it | 21:03 |
| |tbb| | why there is no entry on maemoapplication list till now? | 21:06 |
| keesj | I don't know , when i started playing with it there was a running project | 21:09 |
| keesj | I wanted to add a software mouse so i createda package two | 21:10 |
| ssvb | are there any news about a better application catalog? | 21:10 |
| keesj | I don't see a very good solution , | 21:11 |
| keesj | I know debian used to build the package themselves for the 3rd parties but stopped | 21:11 |
| keesj | and as a user I don't want to create packages for different versions | 21:12 |
| minra | seems many people do a port and then leave it without registering the work to the world | 21:13 |
| ssvb | yes, i think it is partially solved by garage now | 21:13 |
| ssvb | at least if the work is uploaded there, it will not be completely lost | 21:14 |
| ssvb | now from the user perspective, a convenient application catalog would be a good idea | 21:14 |
| ssvb | http://maemo-apps.org/ looks nice, but it is not much used | 21:14 |
| keesj | that's 3rd party | 21:15 |
| keesj | but it does look nice | 21:16 |
| *** sp3000 has joined #maemo | 21:17 | |
| keesj | i don't even run debian , I don't think I can upload packages | 21:18 |
| ssvb | also it would be good to have some place for application porting requests, so that users might vote for their favourite applications and developers could see what it needed | 21:18 |
| ssvb | a few weeks ago i have seen a user who was happy with his Nokia 770 but was missing 'whois' package (he could not find it) | 21:18 |
| ssvb | compiling a package for him was a matter of just a few minutes | 21:18 |
| ssvb | maybe there are lots of users with similar problems which are easily fixable | 21:18 |
| ssvb | keesj, at least you can build .deb files in scratchbox and upload them somewhere | 21:19 |
| |tbb| | yeah macchanger | 21:19 |
| |tbb| | i would like to have it | 21:19 |
| *** rev has joined #maemo | 21:20 | |
| keesj | uploading right now | 21:20 |
| *** alump_ is now known as alump | 21:20 | |
| |tbb| | how to look at the source code of a site with the n770? | 21:21 |
| rev | |tbb|: use wget? | 21:21 |
| rev | or telnet? | 21:21 |
| |tbb| | i need to get an url out of the embedded video | 21:21 |
| rev | yeah | 21:21 |
| rev | get wget or something | 21:21 |
| rev | and use that | 21:21 |
| rev | then open up the downloaded html file | 21:21 |
| |tbb| | wget then the site after open the html site? no easier way? | 21:22 |
| ssvb | |tbb|: what kind of embedded video are you trying to get? | 21:22 |
| |tbb| | its a wmv | 21:22 |
| ssvb | there are existing scripts for yourube for example, http://www.arrakis.es/~rggi3/youtube-dl/ | 21:23 |
| ssvb | it works on nokia 770 :) | 21:23 |
| |tbb| | i have a few other site not youtube | 21:23 |
| tigert | there was the 770-encoder web interface | 21:24 |
| tigert | that, while it takes a moment to do it, accepts a youtube url and spits out a .avi | 21:24 |
| tigert | that the video player can play | 21:24 |
| |tbb| | i know bleb.org do something like that | 21:24 |
| |tbb| | rev what you mean with telnet? | 21:25 |
| rev | |tbb|: you can telnet to port 80, and say GET /videos/whatever.php?v=12419 | 21:25 |
| rev | |tbb|: then it'll spit the HTML at you | 21:26 |
| rev | |tbb|: but that wouldn't be convenient for what you are descriing | 21:26 |
| rev | |tbb|: also, i'm sure there are webpages that would do this for you- an HTML validator or pretty printer would... something where you'd paste the URL into a box, hit go and then see the source. that'd be less of a pain than using wget and editting it | 21:26 |
| |tbb| | hmmh good idea html validator | 21:27 |
| keesj | not tested yet http://box.mmapps.net/~keesj/synergy/ | 21:28 |
| keesj | brb | 21:30 |
| *** keesj has quit IRC | 21:30 | |
| rev | also |tbb| | 21:30 |
| rev | with a little javascript and one of these html validators, you could write a handy bookmarklet | 21:31 |
| rev | that'd be slick | 21:31 |
| *** keesj has joined #maemo | 21:32 | |
| rev | ahhh | 21:32 |
| rev | here we go | 21:32 |
| *** jsun has joined #maemo | 21:32 | |
| jsun | hi - | 21:32 |
| jsun | does anybody know why I can't find Xephyr after installing 2.1 SDK? | 21:33 |
| jsun | the tutorial says xephyr is pre-installed.. | 21:33 |
| keesj | prefix=/scratchbox/users/${LOGNAME}/targets/SDK_PC/usr | 21:33 |
| keesj | export LD_LIBRARY_PATH=${prefix}/lib; export LD_LIBRARY_PATH | 21:33 |
| keesj | exec ${prefix}/bin/Xephyr :2 -host-cursor -screen 800x480x16 -dpi 96 -ac | 21:33 |
| |tbb| | http://validator.aborla.net/ | 21:33 |
| rev | |tbb|: http://www.codingforums.com/archive/index.php?t-18318.html | 21:33 |
| jsun | keesj, the script complains about not finding xephyr. :) | 21:34 |
| minra | hooray for keesj... i will download and test | 21:34 |
| keesj | minra, I has worked in the past, I gui would be nice | 21:34 |
| jsun | fwiw, I tried twice just to make sure I did not do anything wrong. :) | 21:35 |
| jsun | btw, am I the first one to discover this? :) | 21:37 |
| * jsun feels deeply hornored.... | 21:37 | |
| keesj | jsun perhaps you did not start the scratchbox | 21:37 |
| jsun | keesj, I did! | 21:37 |
| keesj | and /scratchbox/login works? | 21:38 |
| jsun | yes | 21:38 |
| jsun | the host is FC5 (with vdso disabled, obviously). | 21:38 |
| jsun | I looked the output very carefully. did not see any error report. | 21:38 |
| *** _Handful_ has joined #maemo | 21:39 | |
| keesj | minra, I just tested , there are two problems | 21:40 |
| jsun | in 2.1 SDK, there are two packages: rootstrap and rootfs. | 21:40 |
| jsun | when am I supposed to use rootfs package? | 21:41 |
| jsun | the tutorial only talks about extracting from rootstrap. | 21:41 |
| jtra | does anybody know where to get or has a source for osso-xterm 0.13 ? I can only find source for 0.11 on http://770.fs-security.com/xterm/ | 21:41 |
| jtra | I want to implement a sane virtual keyboard for it | 21:41 |
| keesj | first you need to install in command line because the deb doesn't contain the right section, secondly of course you dont' see the mouse | 21:41 |
| jtra | the normal one is not very suitable for using modifiers | 21:42 |
| etrunko | jtra: http://repository.maemo.org/pool/scirocco/free/source/o/osso-xterm/ | 21:43 |
| keesj | but it works | 21:43 |
| jtra | etrunko: thanks | 21:44 |
| etrunko | jtra: apt-get source is your friend ;) | 21:44 |
| jtra | but that would require to have proroper deb-src option in sources.list which I don't have :-( | 21:45 |
| jtra | proper | 21:46 |
| etrunko | jtra: ah, ok | 21:46 |
| jtra | and I was also confused because I expected the source archive to be in the same pool directory as .debs are, like in debian | 21:48 |
| *** hume has joined #maemo | 21:48 | |
| _Handful_ | 80 | 21:48 |
| _Handful_ | */-+ | 21:48 |
| hume | hi...i have a problem with the memory card in my nokia 770 - i cannot empty the trash bin (device usb connected to linux laptop) | 21:50 |
| *** _Handful_ has quit IRC | 21:50 | |
| *** _Handful_ has joined #maemo | 21:55 | |
| *** minra has quit IRC | 21:56 | |
| *** xan is now known as xan-afk | 22:00 | |
| *** __shawn has quit IRC | 22:01 | |
| *** mgedmin has joined #maemo | 22:11 | |
| tko_ | hmm, I'd join the amd64 desktop raffle here if it weren't for getting it home in one piece | 22:17 |
| Tak | mail it? ;-) | 22:17 |
| *** __shawn has joined #maemo | 22:23 | |
| keesj | does my Package.gz file need to contain the Section element? | 22:24 |
| keesj | if I have my own repository | 22:24 |
| tko_ | keesj, IIRC application installer displays only packages in user/* sections | 22:25 |
| keesj | yes, but dpkg-scanpackages doesn't add the Section in the Package.gz | 22:26 |
| tko_ | strange.. is the Section: in debian/control for the Source: or Package: section ? | 22:27 |
| tko_ | you might want to try apt-ftparchive too | 22:28 |
| keesj | for Source | 22:28 |
| tko_ | does it show if you do 'less package.deb' ? (dpkg -c package.deb or something I think) | 22:30 |
| keesj | no it is not in the package | 22:30 |
| *** hume has quit IRC | 22:30 | |
| tko_ | ok then dpkg-scanpackages can't find it. apparently it's not inherited for binary packages and needs to be declared for all Package: sections | 22:31 |
| tko_ | I'm sure it's explained in some tutorial or the very least in debian packaging guides | 22:31 |
| keesj | I just tried that | 22:31 |
| *** kender has quit IRC | 22:33 | |
| *** xan-afk is now known as xan | 22:33 | |
| *** tigert has quit IRC | 22:37 | |
| keesj | /me wrong,when I don't use an override file it works | 22:39 |
| *** etrunko has quit IRC | 22:43 | |
| *** sab_osdl has quit IRC | 22:47 | |
| *** benzea has quit IRC | 22:48 | |
| *** benzea has joined #maemo | 22:48 | |
| *** qgil has joined #maemo | 22:55 | |
| dot_j | anyone know of a lumines clone for the 770? | 23:03 |
| _Handful_ | no :/ | 23:03 |
| MDK | re | 23:05 |
| pokute_ | I know that this is old, but http://www.bbspot.com/toys/slashtitle/index.html?reload=20551051294579d3a515af80.04738296 | 23:07 |
| *** buck68 has joined #maemo | 23:08 | |
| MDK | hmm, sardine seems not to rebuild packages | 23:10 |
| *** kb7sqi has quit IRC | 23:13 | |
| *** kb7sqi has joined #maemo | 23:23 | |
| *** dot_j has quit IRC | 23:31 | |
| *** mgedmin has quit IRC | 23:34 | |
| |tbb| | anybody using bloglines? | 23:36 |
| |tbb| | is it worth to install and if so where to get the latest version? | 23:36 |
| *** spect has joined #maemo | 23:42 | |
| *** buck68 has left #maemo | 23:50 | |
| jobi | hej | 23:53 |
| jobi | dpkg: error processing /var/cache/apt/archives/x11-common_1%3a7.0.14osso1_armel.deb (--unpack): | 23:53 |
| jobi | subprocess pre-installation script killed by signal (Segmentation fault), core dumped | 23:54 |
| jobi | does it ring any bell? | 23:54 |
| MDK | hah, had similiar prob i think | 23:56 |
| MDK | while trying to update some stuff via application manager | 23:57 |
| jobi | I was finally attempting a sardine update on my personal 770 :) | 23:57 |
| MDK | same thing here | 23:58 |
Generated by irclog2html.py 2.15.1 by Marius Gedminas - find it at mg.pov.lt!
