| 00:03.52 | *** join/#oe marcan_ (1337@80.26.156.166) |
| 00:14.14 | *** join/#oe chouimat|ibook (~dieu@r2351064.cidc.net) |
| 00:18.21 | *** join/#oe chouimat|ibook (~dieu@r2351064.cidc.net) |
| 00:34.23 | jbowler | ping kergoth |
| 00:34.48 | jbowler | ~seen kergoth |
| 00:34.49 | ibot | kergoth is currently on #uclibc (3d 20h 10m 53s) #nslu2-linux (3d 20h 10m 53s) #oe (3d 20h 10m 53s) #elinux (3d 20h 10m 53s) #handhelds.org (3d 20h 10m 53s) #ol (3d 20h 10m 53s) #tacobeam (3d 20h 10m 53s). Has said a total of 130 messages. Is idling for 3h 46m 56s |
| 00:34.59 | *** join/#oe chouimat_ (~dieu@r2351064.cidc.net) |
| 01:04.53 | *** join/#oe noclouds (~mhfan@60.166.164.101) |
| 01:21.54 | *** join/#oe [g2] (~g2@g2.nslu2-linux) |
| 01:27.19 | *** join/#oe gnufan (~mhfan@60.166.165.192) |
| 01:33.23 | *** join/#oe aphistic (~aphistic@CPE-65-30-217-110.wi.res.rr.com) |
| 01:34.15 | *** part/#oe Laibsch (~leggewie@G17c8.g.pppool.de) |
| 01:34.25 | *** join/#oe chouimat|ibook (~dieu@r2351064.cidc.net) |
| 01:48.13 | *** join/#oe chouimat|ibook (~dieu@r2351064.cidc.net) |
| 02:08.55 | *** join/#oe gnufan (~mhfan@60.166.166.73) |
| 02:17.34 | *** join/#oe chouimat|ibook (~dieu@r2351064.cidc.net) |
| 02:18.35 | *** join/#oe chouimat_ (~dieu@r2351064.cidc.net) |
| 02:38.31 | *** join/#oe keturn (~kevint@c-24-21-141-200.hsd1.or.comcast.net) |
| 02:41.20 | *** join/#oe CIA-9 (~CIA@flapjack.navi.cx) |
| 02:57.12 | *** join/#oe W8TVI (~noel@69.39.75.252) |
| 03:49.49 | *** join/#oe _fishynet (fishy@adsl-215-224-10.aep.bellsouth.net) |
| 04:05.00 | *** join/#oe _fishynet (fishy@adsl-19-125-215.asm.bellsouth.net) |
| 04:31.43 | *** join/#oe _fishynet (fishy@adsl-19-125-57.asm.bellsouth.net) |
| 04:42.49 | *** join/#oe fishynet (fishy@adsl-34-55-183.asm.bellsouth.net) |
| 04:48.24 | *** join/#oe treke|home (~ggilbert@69-175-32-221.ventca.adelphia.net) |
| 05:08.20 | *** join/#oe _fishynet (fishy@adsl-34-55-254.asm.bellsouth.net) |
| 05:38.15 | *** join/#oe raster (~raster@YahooBB219198104088.bbtec.net) |
| 05:39.42 | emte | heya raster |
| 05:39.48 | emte | hows the weekend? |
| 05:41.01 | raster | grrr |
| 05:41.08 | raster | fread is just as fast as mmap |
| 05:41.11 | raster | grrrr |
| 05:41.17 | raster | there goes that idea |
| 05:41.28 | raster | time to do the much nastier speedusp which involve format changes |
| 05:41.51 | emte | ? |
| 05:44.18 | emte | using ptmalloc? |
| 05:44.54 | raster | no |
| 05:45.02 | raster | trying to see if i can get eet to speedup image loads |
| 05:45.07 | raster | from eet archives |
| 05:45.24 | emte | ah |
| 05:45.44 | emte | you sure its not just a resolution problem? |
| 05:45.46 | raster | i just tried an mmap vs fread experiment wiht the c ode |
| 05:45.50 | raster | poerted the reading code to use mmap |
| 05:45.53 | raster | instead of fread |
| 05:45.55 | *** join/#oe mithro (~tim@dsl1-83.gw1.adl1.airnet.com.au) |
| 05:45.58 | raster | did some speed comparisons |
| 05:46.02 | emte | ah |
| 05:46.09 | raster | i can load 149 images /sec using mmap |
| 05:46.15 | raster | and 147 with fread |
| 05:46.22 | raster | not worth the added complexity :( |
| 05:46.31 | emte | these are all 600x800 72dpi? |
| 05:57.01 | *** join/#oe zap (~zap@217.170.93.196) |
| 06:23.54 | raster | hahahah |
| 06:23.56 | raster | dpi> |
| 06:23.58 | raster | ahhaha |
| 06:24.02 | raster | why does dpi matter? |
| 06:24.05 | raster | :) |
| 06:24.15 | raster | and no they arent |
| 06:25.11 | raster | they are varying sizes |
| 06:25.18 | raster | about half are 425x412 |
| 06:25.33 | raster | and half are 187x247 |
| 06:25.52 | raster | that all have alpha channels too |
| 06:26.12 | raster | so effectively its loading 2 images per image (the rgb then the alpha) |
| 06:31.23 | emte | well its a waste to use images over 75 dpi on a monitor |
| 06:34.29 | raster | its a monitor |
| 06:34.31 | raster | its pixels |
| 06:34.34 | raster | dpi is irrelevant |
| 06:34.40 | raster | Nx N pixels is N x N pixels |
| 06:34.55 | emte | dpi affects file size |
| 06:35.29 | raster | no it doesnt |
| 06:35.33 | raster | number of pixels does |
| 06:35.49 | emte | they are inter-related |
| 06:35.57 | raster | (and if compressed how well compression deals with removing data if its lossyc ompression ala jpeg or if it can find patterns) |
| 06:36.02 | raster | they are completely unrelated |
| 06:36.13 | raster | dpi is completely irrelevant |
| 06:36.27 | raster | except for expressiont he resolution of an output device ( a screen or printer) |
| 06:36.41 | raster | as far as software goes in filling the pixels itsd 100% irrelevant :) |
| 06:36.55 | emte | a 200 pixel image at 75dpi is ~ 3k a 200pixel image at 600dpi is ~ 5Mb |
| 06:37.03 | raster | the app doesnt know (or care) how many inches/cm's their window is |
| 06:37.09 | raster | it knwos its N pixels x M pixels |
| 06:37.14 | emte | correct |
| 06:37.20 | raster | no |
| 06:37.24 | raster | thats 100% incorrect |
| 06:37.28 | emte | but it has to parse exta nonessential data |
| 06:37.28 | raster | a 200pixel image is 200x200 pixels |
| 06:37.30 | raster | it has no dpi |
| 06:37.40 | raster | its the same size irrespective of dpi |
| 06:37.54 | raster | no its |
| 06:37.56 | raster | not non-essential |
| 06:38.21 | raster | dpi is a function of ONLY the final output device |
| 06:38.37 | raster | OR an app that wants a display element to be PHYSICALLY a defined size |
| 06:38.51 | raster | eg it wants to display a diagram that is 3x3 inches and make it 3x3 inches on a display |
| 06:38.55 | raster | that is very rare |
| 06:39.01 | raster | and almost never done |
| 06:39.08 | raster | displays are done in pixels |
| 06:39.16 | raster | your output is N x M pixels |
| 06:39.23 | raster | your input data is A x B pixels |
| 06:39.29 | raster | dpi doesnt come into it |
| 06:39.50 | raster | the more pixels you have the more information you can display |
| 06:39.58 | raster | with either soruce or destination |
| 06:40.04 | raster | if its the screen that has more pixels |
| 06:40.04 | emte | hmm |
| 06:40.12 | raster | then you end up with a "better image" |
| 06:40.21 | raster | either more pixels are packed into the same physicla size |
| 06:40.24 | raster | thus it looks "sharper" |
| 06:40.38 | raster | or its physically larger (a 21" display) so u just have mroe pixels u can display |
| 06:40.44 | raster | as far as an app goes |
| 06:40.49 | raster | its still a display of NxM pixels |
| 06:41.00 | raster | dpi is irrelevant |
| 06:41.09 | raster | i have this argument every time with "printing people" |
| 06:41.14 | raster | and "graphic art people" |
| 06:41.17 | raster | at printign shops |
| 06:41.23 | raster | who go "but how many dpi is it?" |
| 06:41.26 | raster | it doesnt matter |
| 06:41.33 | raster | i have a source imageitx N xM pixels |
| 06:41.44 | raster | it will be streatched to fill that sticker/brouchure/pamphlet |
| 06:41.47 | raster | i knwo what will happen |
| 06:41.51 | raster | just print the shit |
| 06:41.52 | raster | :) |
| 06:42.13 | raster | anyh dpi info set on an image file (jpeg, tiff etc.) |
| 06:42.28 | raster | is simply a hint at to the PHYSICAL size it may want to be produced at |
| 06:42.39 | raster | it 2cm x 3cm |
| 06:42.41 | emte | yes and no |
| 06:42.55 | raster | but thats about it |
| 06:43.21 | emte | dpi stands for "dots per inch" its a north american shorthand for "lines per inch" |
| 06:43.38 | raster | yes |
| 06:43.41 | raster | i know :) |
| 06:43.42 | emte | refering to ink points on a press sheet |
| 06:43.57 | raster | itsd a measure of image elements per physicla measurment unit |
| 06:44.04 | raster | u coudluse dpmm |
| 06:44.13 | raster | it can be used to measure anything |
| 06:44.16 | emte | only in canadfa or eurpoe |
| 06:44.17 | raster | if a dot is a pixel on a screen |
| 06:44.26 | raster | u can say a crt or lcd has N dpi |
| 06:44.39 | raster | but thats just a imperial vs metric thing |
| 06:44.42 | emte | thats incorrect tho |
| 06:45.02 | raster | the concept is that u meansure display device resolution in terms of physical attributes of the device |
| 06:45.10 | raster | how is it incorrect? |
| 06:45.28 | raster | if you consider 1 pixel to be a dot |
| 06:45.33 | emte | if you take a digital image of a fixed size, displayed at a fixed size as a human you will not see the difference between 75 and 20,000 |
| 06:45.34 | raster | which most people would |
| 06:46.15 | raster | for output pixels |
| 06:46.16 | raster | yes |
| 06:46.25 | raster | as all u do is stretch the same input pixels over more output |
| 06:46.32 | raster | having to invent more pixels |
| 06:46.43 | raster | tho it depends how many pixels u have as a source |
| 06:46.48 | raster | if its more than you can display |
| 06:46.51 | raster | you need to scale down |
| 06:46.54 | emte | so for taht image anything abouve 75 is a waste and only serves to increase the image's data size |
| 06:46.55 | raster | thus the higer the dpi the better |
| 06:46.57 | raster | ie |
| 06:47.00 | raster | better resolution |
| 06:47.11 | raster | no |
| 06:47.16 | raster | it depends on the soruce size |
| 06:47.34 | raster | at the end of the day |
| 06:47.41 | raster | u can just trhwo dpi out of it entirely |
| 06:47.43 | emte | the fixed size being referd to is it's intended usage size |
| 06:47.46 | raster | its an irrelevant bit of info |
| 06:47.54 | raster | you have a soruce image 100x100 pixels |
| 06:48.01 | raster | you have a display device 1000x1000 pixels |
| 06:48.19 | raster | displaying the 100x100 image at full size on the output device |
| 06:48.24 | raster | means u simply invent pixels |
| 06:48.35 | raster | so u just end up producing lots of redundant data |
| 06:48.44 | raster | i know |
| 06:48.47 | raster | but dpi isnt part of that |
| 06:48.53 | raster | thats just image scaling |
| 06:49.07 | emte | that analysis is correct for vector but wrong for raster images |
| 06:49.08 | raster | thats how things work |
| 06:49.16 | raster | no |
| 06:49.19 | raster | its correct for raster too |
| 06:49.20 | raster | :) |
| 06:49.23 | raster | trust me |
| 06:49.30 | raster | i've been doing this for a long time :) |
| 06:49.38 | emte | scaling a raster image up 100% from it's source size will make it look like shit |
| 06:49.54 | raster | that depends on your algorithm |
| 06:50.05 | raster | pixel replication will not change the data at all |
| 06:50.09 | raster | just make it blockier |
| 06:50.19 | emte | correct |
| 06:50.23 | raster | as it woudl if you moved your head closer to the screen so the image filled your field of fision |
| 06:50.26 | raster | vision |
| 06:50.36 | raster | interpolation smooths it out |
| 06:50.37 | emte | sort of |
| 06:50.42 | raster | in place of blockiness its blurriness |
| 06:50.45 | emte | mm no |
| 06:51.03 | raster | but you dont transmit more information as your source is still the same |
| 06:51.05 | raster | mind you |
| 06:51.14 | raster | when you COMBINE images in layers |
| 06:51.18 | emte | within 10-20% its reasonable but after that the gaps are too large |
| 06:51.19 | raster | this can be very useful |
| 06:51.32 | raster | yes |
| 06:51.35 | raster | you will notice them |
| 06:51.38 | raster | so what do u do?> |
| 06:51.41 | raster | u have 100x100 pixels |
| 06:51.45 | raster | u need to fill the screen with it |
| 06:51.47 | raster | so what do u do? |
| 06:52.02 | emte | you buy a proper image for the intended purpose :) |
| 06:52.02 | raster | you have no choice but to scale and fill in the "gaps" |
| 06:52.09 | raster | you have no choice |
| 06:52.13 | raster | u have been GIVNE the image |
| 06:52.19 | raster | u cant go invent a new one |
| 06:52.25 | raster | ie u downloaded it form a web page |
| 06:52.36 | raster | or it comes form a digital camera whose ccd is only N mpixels |
| 06:52.36 | emte | yes you can invent a new image |
| 06:52.41 | raster | u cant invent data that isnt there |
| 06:52.44 | raster | no u cant |
| 06:52.50 | emte | i cant say that was my fun part of the job |
| 06:52.55 | raster | an applicatin cant invent a new image |
| 06:53.04 | emte | sure you can |
| 06:53.12 | raster | no you cant |
| 06:53.32 | emte | then i should tell adobe and macromedia thier applications dont exist? |
| 06:53.33 | raster | you can apply algorithms that can get incresingly complex to try extrapolate and guess what info may hav ebeen there |
| 06:53.37 | raster | (fractal...) |
| 06:53.48 | raster | but its still a guess at best |
| 06:54.24 | raster | u have loaded anb image into memory and decoded |
| 06:54.26 | emte | if your talking machine repair i agrewe |
| 06:54.29 | raster | u now have N pixels of data |
| 06:54.41 | raster | thats all the info u have |
| 06:54.48 | raster | u have to make the most of it |
| 06:55.17 | emte | correct, so as a person handling it you rebuild/create |
| 06:56.15 | emte | but comming full circle there is no reason for you to have a 600Mb image as your desktop background |
| 06:57.32 | emte | it will only take your system longer to load it especially if its starved for ram |
| 07:00.53 | emte | but i suppose back to my original question about dpi, there really is no reason to use images abouve 1600 @ 75dpi since that is the current max resolution is it not? |
| 07:01.22 | raster | yes and no |
| 07:01.29 | raster | what if you dont KNOW the target resolution? |
| 07:01.33 | raster | it coudl be 640x480 |
| 07:01.38 | raster | coudl be 1600x1200 |
| 07:02.12 | raster | could be 3840x2880 |
| 07:02.25 | raster | it could be 320x240 |
| 07:02.34 | raster | or 1920x1200 |
| 07:02.38 | raster | or 1024x768 |
| 07:02.44 | raster | or 10248x768... |
| 07:03.01 | raster | so you provide the best resolutiuon input data you can |
| 07:03.08 | raster | so it looks best at the highest res |
| 07:03.29 | raster | and then the display algorithms remove data - or reduce it howvere they see best |
| 07:06.44 | *** join/#oe [cc]smart (~smart@gw.ptr-62-65-149-158.customer.ch.netstream.com) |
| 07:07.01 | emte | yeah |
| 07:11.03 | ljp | ya, or 20x20 |
| 07:21.09 | raster | eww |
| 07:21.10 | raster | 20x20 |
| 07:21.27 | raster | the again |
| 07:21.33 | raster | u'll get that on some devices |
| 07:21.42 | raster | liek cellphones and their secondary display (on the outside) |
| 07:21.49 | raster | doesnt give u much room to move tho |
| 07:56.50 | *** join/#oe jacques_ (~jacques@c-67-171-198-193.hsd1.or.comcast.net) |
| 08:01.36 | *** join/#oe zedkatuf (~dfutak@82-32-58-45.cable.ubr08.azte.blueyonder.co.uk) |
| 08:01.55 | *** join/#oe maqt (~matthew@82-35-248-11.cable.ubr06.dals.blueyonder.co.uk) |
| 08:37.18 | koen | good morning all |
| 10:10.38 | *** join/#oe ibot (ibot@apt.bot.TimRiker.active.supporter.pdpc) |
| 10:10.38 | *** topic/#oe is OpenEmbedded Developer Lounge | http://bugs.openembedded.org/ (l/p guest/guest) | This is not a distribution support channel | The OE mailing list is back, but everyone needs to resubscribe. | We have a bitbake-dev mailing list for discussions on the core. If you're interested in BitBake-Ng, please subscribe. | LCA 2005 Embedded Miniconf - http://www.openembedded.org/miniconf |
| 10:46.05 | *** join/#oe mithro (~tim@dsl1-83.gw1.adl1.airnet.com.au) |
| 11:06.11 | Luke-Jr | um |
| 11:06.22 | Luke-Jr | How do I get my gpe-image to use 2.6.12-bla? |
| 11:07.09 | Luke-Jr | or is that a bad idea? |
| 11:16.00 | *** join/#oe bluelightning (~bluelight@203-173-244-208.akl.ihugultra.co.nz) |
| 11:16.36 | bluelightning | hello all |
| 11:25.49 | koen | hey bluelightning |
| 11:26.01 | bluelightning | hi koen |
| 11:26.08 | bluelightning | koen: how was your holiday? |
| 11:26.15 | koen | very good |
| 11:27.24 | koen | hungary always reminds me of the fact that the netherlands lacks mountains |
| 11:31.55 | Luke-Jr | lol |
| 11:32.10 | *** join/#oe qfh (~qfh@ppp-62-245-211-234.mnet-online.de) |
| 11:34.37 | koen | I wanted to visit greg72 in gyongyos, but the downpour on thursday and friday messed up the "schedule" |
| 11:50.14 | *** join/#oe Jean-MAr1 (~jmau@pD9EB73F5.dip0.t-ipconnect.de) |
| 11:53.10 | *** part/#oe Jean-MAr1 (~jmau@pD9EB73F5.dip0.t-ipconnect.de) |
| 11:59.49 | *** join/#oe qfh (~qfh@ppp-62-245-211-234.mnet-online.de) |
| 12:09.41 | *** join/#oe CIA-9 (~CIA@flapjack.navi.cx) |
| 12:34.09 | mickeyl | morning |
| 12:35.36 | koen | hey mickeyl |
| 12:35.42 | koen | good job on the OOOOOO newsletter |
| 12:35.47 | mickeyl | thanks :) |
| 12:35.51 | mickeyl | welcome back |
| 12:36.04 | koen | thanks |
| 12:38.27 | *** join/#oe andersee (~andersee@codepoet.org) |
| 13:09.18 | *** join/#oe Timelord (~TL@64-147-65-148.access-4.intertex.net) |
| 13:14.09 | lardman | Where in the build process does a binary get stripped? Is the one in the build directory unstripped for example? (wanting to track down why Quake segfaults) |
| 13:14.25 | mickeyl | during packaging |
| 13:15.13 | lardman | but is it stripped in ${S}, then copied, or copied then stripped in ${WORKDIR}/image (or whatever the name is)? |
| 13:16.02 | mickeyl | read package.bbclass |
| 13:16.06 | lardman | ok |
| 13:16.32 | mickeyl | the one in the build directory is always untouched |
| 13:16.40 | lardman | cool, thanks |
| 13:17.01 | mickeyl | you can just inhibit it globally though |
| 13:17.22 | mickeyl | echo INHIBIT_PACKAGE_STRIP = "1" >>conf/local.conf |
| 13:17.42 | lardman | It's only for the one package, so I was just going to copy the file over and run strace/gdb (I'm learning here :) ) on it and see what I can find |
| 13:20.10 | koen | lardman: I noticed a cset for the new cfu1u driver, is it working yet? |
| 13:20.34 | lardman | koen: yes, after a fashion, though you should ask B0Ti as he has done all the work on it |
| 13:20.51 | lardman | koen: It still has issues - it doesn't actually do much |
| 13:21.15 | lardman | koen: for example mass-storage devices appear (to me at least) to be unpowered - the light comes on, then goes off |
| 13:21.46 | lardman | koen: but it does detect and use my powered hub correctly (can see the mass-storage device through it, same as if I plug in directly - still doesn't work though) |
| 13:22.47 | koen | sounds like progress |
| 13:23.03 | lardman | koen: yes, B0Ti has made real progress on it |
| 13:43.35 | lardman | To enable usbfs in 2.6.11, the menu entry in 'make menuconfig' under modules > usb is not there - is there a reason for its absence? |
| 13:44.49 | lardman | and another random one: any ideas how I should solve this build problem: http://pastebin.ca/11186 ? |
| 13:45.16 | bluelightning | night all |
| 13:45.22 | lardman | I've already added the correct path so it's passed when the linker is called |
| 13:45.33 | lardman | night bluelightning |
| 13:46.48 | mickeyl | show log.do_compile |
| 13:46.51 | *** join/#oe B0Ti (~boti@82.79.98.49) |
| 13:50.33 | lardman | http://pastebin.ca/11228 though this is with -rlink-lib added (just adds another warning at the beginning, the error is the same), give me a tick and I'll clean & rebuild |
| 13:52.39 | lardman | http://pastebin.ca/11229 - without rpath-link in the .bb file |
| 13:53.02 | lardman | .bb here: http://students.bath.ac.uk/enpsgp/Zaurus/emul/emul |
| 13:53.29 | lardman | and the reason why here: http://www.oesf.org/forums/index.php?showtopic=3953&hl= |
| 13:56.49 | mickeyl | lardman: your option is wrong |
| 13:57.14 | mickeyl | it's not a CFLAGS option, it's a LDFLAGS option |
| 13:57.57 | mickeyl | if you want to pass this option to the linker via gcc |
| 13:57.59 | mickeyl | then you have to say |
| 13:58.07 | mickeyl | -Wl,-rpath-link,... |
| 13:59.31 | lardman | oh, okay |
| 14:00.30 | lardman | Actually I did have it in LDFLAGS, hang on, let me run it again and update the version which is up on the web |
| 14:00.40 | lardman | though I didn't have -Wl |
| 14:03.08 | lardman | ha, LDFLAGS was after -lemul, and so wasn't found - plus I didn't have -Wl - thanks mickeyl |
| 14:08.33 | mickeyl | np |
| 14:15.56 | *** join/#oe mithro (~tim@dsl1-83.gw1.adl1.airnet.com.au) |
| 14:26.01 | koen | does anybody know what's wrong with openlogging? |
| 14:30.55 | mickeyl | broken atm. |
| 14:31.08 | mickeyl | i hope it'll be up soon |
| 14:31.44 | mickeyl | we should give 'em a weekday to spot the problem though |
| 14:32.05 | koen | has anyone used bk support on it yet? |
| 14:32.22 | mickeyl | don't think so |
| 14:32.29 | mickeyl | are we entitled to use bk support with the free version at all? |
| 14:32.46 | koen | no idea, but it tells me to do so :) |
| 14:33.02 | mickeyl | go ahead then |
| 15:21.43 | B0Ti | what kind of an usb host controller does the C3000 and SL-6000 have? |
| 15:22.49 | koen | iirc the c3000 uses the pxa270 host controller |
| 15:22.54 | koen | no idea on the 6k |
| 15:22.59 | mickeyl | the 6000 tc6393 |
| 15:25.01 | B0Ti | So why isn't host controller support enabled in 2.6.12-rc's defconfig? |
| 15:25.19 | mickeyl | 2.6.12-rc's defconfig for what machine ? |
| 15:25.54 | B0Ti | Mine is a c860. |
| 15:26.04 | mickeyl | uhm |
| 15:26.08 | mickeyl | then c860 has no usb host |
| 15:26.10 | mickeyl | s/then/the7 |
| 15:26.14 | mickeyl | s/the7/the/ |
| 15:26.36 | B0Ti | doesn't come with a host controller though, but with the ratoc card it does :P |
| 15:27.04 | mickeyl | options aren't enabled just because of a mood |
| 15:27.17 | mickeyl | options are enabled when there is demand and they're tested |
| 15:27.42 | lardman | Right, Quake works (I get to the console anyway, note quite sure how to play), Mickeyl, could you apply the changes from here when you have time please: http://students.bath.ac.uk/enpsgp/Zaurus/quake/quake/ |
| 15:28.02 | mickeyl | lardman: i'll incorporate them into my tree. bk is broken atm. |
| 15:28.15 | lardman | mickeyl: Okay, thanks |
| 15:30.26 | lardman | I've got to agree with B0Ti - it would be nice if the modules were available without needing to mess about with the defconfigs |
| 15:30.34 | lardman | as I also have a ratoc card of course |
| 15:31.22 | gints|home | lardman: may be bb package may depend on libsdl-virtual or whatever.. |
| 15:31.37 | gints|home | to use libsdl-qpe too |
| 15:32.04 | lardman | gints|home: I tried that to begin with, but it failed, perhaps because my local.conf wasn't set up correctly |
| 15:32.22 | koen | isn't that the x11quake source? |
| 15:32.37 | B0Ti | Trying to compile 2.6.12: sound/oss/poodle_wm8731.c:346: error: `GPIO_HP_IN' undeclared |
| 15:32.42 | lardman | koen: pocket quake, modified I think, might have come from x11-quakwe |
| 15:33.04 | lardman | gints|home: Try it and see, it'd be nice if it would work for both |
| 15:33.33 | lardman | Have a look at this though: http://students.bath.ac.uk/enpsgp/Zaurus/quake/readme.txt |
| 15:35.13 | gints|home | lardman: please add thease links as reference to http://handhelds.org/moin/moin.cgi/FamiliarGames |
| 15:36.42 | lardman | gints|home: I'm not allowed to edit the page it tells me |
| 15:36.52 | *** join/#oe mithro (~tim@dsl1-83.gw1.adl1.airnet.com.au) |
| 15:37.12 | koen | lardman: goto userpreferences and create a login |
| 15:41.25 | lardman | bbiam - I've added Quake1, but no subpage yet - I'll need to read how to when I get back |
| 15:49.49 | B0Ti | looks like the corgi sound stuff cannot be compiled as modules. |
| 15:50.32 | *** join/#oe mallum (~mallum@host81-155-187-182.range81-155.btcentralplus.com) |
| 16:09.25 | koen | lardman: http://handhelds.org/moin/moin.cgi/Quake1 |
| 16:17.27 | *** join/#oe _CoreDump|home (~mhentges@hentges.net) |
| 16:21.15 | *** join/#oe cdbot (~cdbot@hentges.net) |
| 16:33.28 | koen | mickeyl: did the localtime thing work out? |
| 16:33.56 | mickeyl | ? |
| 16:34.15 | koen | the timestamp in /etc |
| 16:34.19 | mickeyl | ah |
| 16:34.23 | mickeyl | yeah, it works very good |
| 16:34.28 | mickeyl | we should use it on all images |
| 16:34.32 | mickeyl | dunno if i already did that |
| 16:34.41 | koen | can gpe share in its wealth too? |
| 16:36.31 | mickeyl | sure |
| 16:36.33 | mickeyl | I'll add it to gpe-image |
| 16:37.03 | koen | can it be in bootstrap? |
| 16:37.34 | mickeyl | yes, but that wouldn't make it appear in gpe or opie |
| 16:37.46 | mickeyl | the image files don't include each other |
| 16:38.09 | koen | bootstrap_extra_depends? |
| 16:38.17 | mickeyl | no |
| 16:38.22 | mickeyl | it's not something you need to add via depends |
| 16:38.28 | mickeyl | it's already in initscripts |
| 16:38.33 | mickeyl | you need to add the initial timestamp |
| 16:38.35 | mickeyl | which is a command |
| 16:38.38 | koen | aha |
| 16:38.40 | mickeyl | IMAGE_PREPROCESS_COMMAND = "create_etc_timestamp;" |
| 16:38.57 | mickeyl | so just add that to bootstrap-image and gpe-image and it'll work |
| 16:39.02 | mickeyl | ~lart openlogging |
| 16:39.03 | koen | thanks |
| 16:39.14 | *** join/#oe Crofton|laptop (~balister@66-207-66-26.black.dmt.ntelos.net) |
| 16:39.18 | mickeyl | np |
| 16:54.04 | lardman | koen: thanks |
| 16:55.52 | *** join/#oe man-di (~man-di@dyndsl-080-228-194-042.ewe-ip-backbone.de) |
| 17:02.00 | *** join/#oe pb_ (~pb@2002:5168:d330:1:a00:1fff:fe06:93c) |
| 17:02.36 | pb_ | afternoon |
| 17:02.44 | lardman | How does one get a line break in the handhelds wiki? |
| 17:02.56 | koen | [br] iirc |
| 17:03.23 | lardman | nope |
| 17:03.43 | koen | [[br]]? |
| 17:03.59 | lardman | yep, just found that in the help page |
| 17:04.48 | lardman | thx |
| 17:06.02 | mickeyl | greetings pb_ |
| 17:06.09 | pb_ | hi mickeyl |
| 17:06.12 | koen | hey pb_ |
| 17:06.17 | pb_ | hi koen |
| 17:08.40 | *** join/#oe jbowler (~jbowler@67.139.207.38) |
| 17:08.56 | jbowler | ~seen kergoth |
| 17:08.57 | ibot | kergoth is currently on #devlounge #uclibc #nslu2-linux #oe #elinux #handhelds.org #ol #tacobeam |
| 17:19.29 | *** join/#oe molivier2 (~mac@f01v-22-201.d4.club-internet.fr) |
| 17:47.36 | *** join/#oe SuN (~Sun@195-241-12-117-mx.xdsl.tiscali.nl) |
| 17:51.39 | *** join/#oe dkey (~dkey@193.170.48.234) |
| 17:54.11 | *** join/#oe exastra (~go@l5.org) |
| 18:07.49 | Crofton|laptop | I hadn't even noticed |
| 18:07.55 | Crofton|laptop | thanks |
| 18:32.42 | CoreDump|home | ~lart openlogging |
| 18:48.11 | *** join/#oe lardman (lardman@pc-138-38-248-150.bath.ac.uk) |
| 18:48.58 | lardman | mickey|tv: oops, slight mistake in one of the quake1 patches, means that a game can't be started, I'm just testing the fix now and will update tomorrow assuming it works |
| 18:49.07 | lardman | night all |
| 19:05.46 | *** join/#oe beewoolie-akf (~beewoolie@florence.buici.com) |
| 19:08.56 | Cwiiis | Can anyone give me any pointers on how to connect to my zaurus via serial cable, to read kernel debug messages and such? |
| 19:14.20 | CoreDump|home | Cwiiis: a collie? i thought that would be USB |
| 19:14.58 | Cwiiis | CoreDump|home: kernel messages go to /dev/ttyS0 |
| 19:15.05 | CoreDump|home | hmm |
| 19:16.12 | CoreDump|home | if it's a real serial connection then it's as easy as pointing your terminal emulator to the serial device on your desktop |
| 19:18.51 | Cwiiis | ah, cool, let me see how to do that then :p hehe |
| 19:19.59 | CoreDump|home | i've used the serial console to work on my desktop from my notebook in the past. works great :) |
| 19:20.28 | Cwiiis | :) |
| 19:20.42 | Cwiiis | Just a case of 'xterm /dev/ttyS0'? (or similar) |
| 19:20.57 | CoreDump|home | Cwiiis: ummm *scratches head* |
| 19:21.05 | CoreDump|home | I've used minicom |
| 19:23.10 | Cwiiis | hmm |
| 19:23.55 | *** join/#oe chouimat|outside (~dieu@r2351064.cidc.net) |
| 19:28.10 | *** join/#oe CIA-9 (~CIA@flapjack.navi.cx) |
| 19:41.49 | *** part/#oe jbowler (~jbowler@67.139.207.38) |
| 19:55.14 | Cwiiis | CoreDump|home: Any tips on connecting to a serial terminal with minicom? :) |
| 19:55.28 | koen | 115200 8n1, no flow control |
| 19:55.38 | CoreDump|home | sounds about right |
| 19:55.57 | koen | and you'll probably need a getty in your inittab for ttyS0 |
| 19:56.23 | Cwiiis | No idea what any of this means right at the moment :/ |
| 19:56.24 | CoreDump|home | right, not for viewing the kernel messages though |
| 19:56.31 | koen | true |
| 19:56.38 | CoreDump|home | launch mincom as root with "minicom -s" |
| 19:57.09 | Cwiiis | ah, that seems a bit easier :) |
| 19:58.02 | Cwiiis | After doing that, should running minicom just display the kernel messages? |
| 19:58.49 | koen | iirc you need to set it up first |
| 19:59.03 | koen | ctrl-a s |
| 19:59.07 | koen | (again, iirc) |
| 19:59.22 | Cwiiis | hmm, that appears to be upload? |
| 19:59.38 | Cwiiis | ctrl-a o perhaps |
| 19:59.40 | koen | ctrl-a z? |
| 20:00.18 | Cwiiis | I did the setting up, as far as I know, should that be it? Can I echo messages to /dev/ttyS0 on the Z to test it? (if you know) |
| 20:00.33 | CoreDump|home | that should work, yes |
| 20:00.59 | CoreDump|home | 115200 8n1, no flow control <- I dunno if that is the kernel default though |
| 20:01.12 | Cwiiis | Did stty on the Zaurus, it would seem so |
| 20:01.18 | Cwiiis | (well, stty -F /dev/ttyS0 of course) |
| 20:01.32 | koen | ~whaleslap angelboot |
| 20:01.33 | ibot | ACTION slaps angelboot upside and over the head with one freakishly huge killer whale named hugh |
| 20:02.04 | Cwiiis | ah brilliant, works, cheers guys :D |
| 20:02.41 | *** join/#oe SuN (~Sun@195-241-12-117-mx.xdsl.tiscali.nl) |
| 20:02.41 | *** join/#oe zap (~zap@217.170.93.196) |
| 20:02.41 | *** join/#oe stealth (~stealth@user-0c6sdm3.cable.mindspring.com) [NETSPLIT VICTIM] |
| 20:02.41 | *** join/#oe ljp (~lpotter@203.94.178.46) |
| 20:08.04 | *** join/#oe stealth (~stealth@24.110.54.195) |
| 20:09.13 | Cwiiis | hmm... |
| 20:09.47 | Cwiiis | I guess you'd have to network first or something... |
| 20:09.49 | koen | yes |
| 20:09.54 | koen | it's called ppp |
| 20:18.19 | nhorlock | evening all |
| 20:18.56 | nhorlock | Anyone had problems with libiberty when using bitbake to build oe??? |
| 20:21.31 | nhorlock | I'm getting "undefined reference to '__builtin_expect' in function pexec. |
| 20:23.09 | Luke-Jr | What's the proper way to select a kernel version (2.6.12-bla)? |
| 20:27.36 | nhorlock | Hmmm looks like the __builtin_expect problem was common with gcc 2.95 but we're meant to be using a much later compiler. |
| 20:29.26 | Luke-Jr | hmm |
| 20:29.38 | Luke-Jr | my gpe-images refuse to sleep after a few sleep/resumes |
| 20:37.40 | koen | mickey|tv: bitmover contacted me for more details on the openlogging thing as it seems we're the only project suffering from it |
| 20:42.31 | *** join/#oe jacques (~jacques@jacques.nslu2-linux) |
| 20:55.15 | nhorlock | Does anyone else find building things with bitbake as painful as I do. It seems that I can never get to actually do any development of my own as I have to spend evening after evening trying to figure out stupid build problems and dealing with broken mirrors. There has to be a better way. |
| 21:01.52 | emte | there are quite a few issues at the moment which are causing problems |
| 21:02.16 | emte | hopefully they will get sorted out in the near future |
| 21:04.53 | Luke-Jr | What's the proper way to select a kernel version (2.6.12-bla)? |
| 21:05.47 | koen | nhorlock: how are broken mirrors an OE problem? |
| 21:13.07 | nhorlock | koen: Trying to build opie-image and initially got ftp.tr.debian.... failing, for "fakeroot", this filaed over to oesources.org and got 404 not found. then I edited local.conf to force a UK mirror instead of using the pool and this is failing ('cos I need to give a better local path. |
| 21:13.40 | nhorlock | ^local^UK mirror local^ |
| 21:14.09 | nhorlock | It is not OE that is broken as such it is the time wasted trying to get to a workable position |
| 21:22.53 | nhorlock | It gets better, bitbake is now succesfully getting to all the mirrors but fakeroot-1.2.2 is not on any of them. I've checked manually and its correct. |
| 21:23.09 | nhorlock | 1.2.2. simply is no longer there. |
| 21:23.45 | nhorlock | Ho hum, I've had it for now, another weekend of development wasted. |
| 21:25.31 | *** join/#oe CIA-9 (~CIA@flapjack.navi.cx) |
| 21:32.13 | mickeyl | it won't get you anywhere to complain to the few people working with it |
| 21:32.25 | mickeyl | if at all, complain to not enough people maintaining packages, not to us |
| 21:32.25 | mickeyl | heh |
| 21:32.48 | pb_ | heh |
| 21:36.44 | nhorlock | mickeyl: I realise that bitching to you is not going to help. But I am willing to try to maintain some packages, I really want to help bolster the SL6000 support but even the most basic apps won't build for me, I'm obviously missing something here, it seems like sarcasm is the only support those willing to try and help are going to get. |
| 21:37.32 | nhorlock | mickeyl: Sorry, that was a bit harsh, its been a long night. |
| 21:38.10 | mickeyl | if the reason that those basic apps - whatever these may be - don't build, you may be better off just asking for a particular package or a more fully populated source mirror |
| 21:38.25 | mickeyl | s/build/build is missing sources/ |
| 21:39.36 | mickeyl | for instance, you may try to use www.vanille.de/sources/ as a CVS_TARBALL_STASH while it's online |
| 21:41.46 | nhorlock | mickeyl: I'll have a look at that, thanks. I'm also seeing why there is a dependency on fakeroot_1.2.2 as it would seem that debian has dropped it, so the right thing seems to be to move to a newer version. |
| 21:42.11 | pb_ | yeah, that happens |
| 21:42.21 | mickeyl | debian drops things frequently |
| 21:42.32 | mickeyl | someone should change to using their archive server instead of their pool server |
| 21:42.33 | pb_ | it sucks a bit, but there's no alternative upstream |
| 21:42.56 | pb_ | yeah, trouble with archive.debian.org is that it _only_ holds old versions. |
| 21:43.07 | pb_ | kind of like the Attic thing we have with rxvt-unicode. |
| 21:43.21 | pb_ | (not that I suppose you'd have encountered that one, heh) |
| 21:43.25 | mickeyl | :D |
| 21:43.38 | mickeyl | i rarely clean my sources dir |
| 21:43.42 | mickeyl | i know what happens when i do :D |
| 21:45.18 | pb_ | hehe |
| 21:47.03 | nhorlock | pb_: maybe archive.debian.org is where I should try for fakeroot then.......I'll be back....As an aside it would be useful to have a list of packages that are no longer easily obtained on the Wiki, I'll have a go at listing them as I find them. Though I expect to grind to a halt on the __builtin_expect issue again not long after I resolve this issue. |
| 21:48.58 | pb_ | What's "the __builtin_expect issue"? |
| 21:49.48 | pb_ | mickeyl: maybe you should have Greg rsync your vanille.de/sources directory onto oesources.org |
| 21:49.59 | mickeyl | pb_: right. treke|home ? |
| 21:51.24 | nhorlock | before trying the opie-image target I was simply building/failing on nano . When it gets to libiberty it gets a symbol not found __builtin_expect, which is a gcc builtin for conditional jump prediction hinting |
| 21:52.48 | nhorlock | how much space is needed for a vanille.de mirror? I have a server with some space even if I can't provide the lot I may be able to provide some. |
| 21:52.51 | mickeyl | d'oh. never hit that one |
| 21:53.00 | mickeyl | you sure you didn't build a gcc snapshot per accidant ? |
| 21:54.28 | mickeyl | koen: good. did you send 'em the output of bk log -d ? |
| 21:55.06 | pb_ | afaik, the only way you'd end up without __builtin_expect would be if you somehow built gcc 2.95.3 by accident. |
| 21:55.23 | pb_ | which would be extraordinarily unlikely, since most people can't even build gcc 2.95.3 when they try. |
| 21:55.54 | pb_ | so, that's certainly a bit weird. but, nobody else has reported it, so I guess it must be some bizarre local problem. |
| 21:56.10 | nhorlock | mickeyl:don't think so. Does a bitbake -c clean kill of any local build/tmp builds etc. as I've clean built a few times just in case. |
| 21:56.46 | mickeyl | alias |
| 21:56.53 | mickeyl | alias bkclean='rm -rf tmp' |
| 21:56.55 | mickeyl | :D |
| 21:57.08 | nhorlock | pb__:I agree, my googles turned up nothing after 2.95. |
| 21:57.32 | pb_ | this wasn't a native libiberty or anything like that? |
| 21:58.42 | pb_ | well, if this was nano, I guess it must have been. |
| 21:58.46 | pb_ | what version is your host compiler? |
| 21:59.22 | nhorlock | As far as I know I've done nothing but pulled down the OE/bitbake stuff according to the Wiki. I do have old Sharp and opie tool chains on this box so I am suspecting they might have to be scythed as well :-( |
| 21:59.51 | pb_ | So long as they aren't in your $PATH, they should be harmless. |
| 22:00.13 | pb_ | If they _are_ in your $PATH then they will cause all kinds of devastation. You'd definitely need to avoid that. |
| 22:01.27 | nhorlock | Hmmmm, well done pb__ I think you may have hit the nail on the head. On Saturday night I was farting around with an old build I bet this shell has a bad gcc on the path. Doh!!! |
| 22:02.00 | mickeyl | mickey@gandalf:/$ du -h -s /local/pkg/sources |
| 22:02.00 | mickeyl | 6,8G /local/pkg/sources |
| 22:02.17 | nhorlock | Reading specs from /opt/Embedix/native/lib/gcc-lib/i686-pc-linux-gnu/2.95.3/specs |
| 22:02.17 | nhorlock | gcc version 2.95.3 20010315 (release) |
| 22:02.23 | mickeyl | tsss |
| 22:02.33 | *** join/#oe imk (~dj@69-175-55-137.ontrca.adelphia.net) |
| 22:02.35 | mickeyl | ~curse embedix toolchains |
| 22:02.37 | ibot | May you be reincarnated as a Windows XP administrator, embedix toolchains ! |
| 22:02.51 | nhorlock | haha |
| 22:06.20 | nhorlock | pwd |
| 22:06.47 | nhorlock | whoops wrong window, gcc -v now returns gcc version 3.4.2 20041017 (Red Hat 3.4.2-6.fc3) |
| 22:06.56 | pb_ | very good |
| 22:09.55 | nhorlock | It is stupid things like that, that always trip you up. Typically it is now time I was asleep, but at least I can go to bed knowing that I've made a step forward. |
| 22:10.55 | mickeyl | good idea |
| 22:10.57 | mickeyl | g'night folks |
| 22:11.08 | nhorlock | mickeyl: I have about 27G of space on my webserver |
| 22:11.40 | nhorlock | mickeyl: looks like I should be able to help. |
| 22:12.43 | *** join/#oe Flyinglord^ (~qfh@ppp-62-245-211-234.mnet-online.de) |
| 22:12.44 | nhorlock | mickeyl: goodnight. Catch you later. |
| 22:13.29 | nhorlock | pb__: thanks for your help. I'll be back tomorrow hopefully with some good news. |
| 22:43.26 | *** join/#oe Geo_KM (~keith@CPE-61-9-204-1.nsw.bigpond.net.au) |
| 22:59.47 | *** join/#oe RR22 (~ircap@154.Red-217-127-168.pooles.rima-tde.net) |
| 23:00.29 | *** part/#oe RR22 (~ircap@154.Red-217-127-168.pooles.rima-tde.net) |
| 23:18.55 | *** join/#oe [g2] (~g2@cpe-066-057-008-035.nc.res.rr.com) |
| 23:39.39 | *** join/#oe macmac (~mac@f01m-33-254.d4.club-internet.fr) |
| 23:41.42 | *** join/#oe Geo_KM (~keith@CPE-61-9-204-1.nsw.bigpond.net.au) |