| 00:00.08 | Christopher | rwhitby-web: I also had to do a couple things different than template.mk because there is no configure script for nethack |
| 00:06.00 | rwhitby-web | no problem - just keep the structure and variables the same - you should only have to change some commands in the .configured target, right? |
| 00:06.15 | rwhitby-web | [there are other packages which don't use configure] |
| 00:06.25 | Christopher | and I have to pass flags directly to make |
| 00:15.05 | rwhitby-web | yep - you can still use TARGET_CONFIGURE_OPTS to do most of that passing of flags |
| 00:16.28 | jacques | anyone know why my unslung cvs build is failing with: /home/packages/nslu2/nslu/unslung-20041027/builds/slingbox/applets/busybox.c:5:19: errno.h: No such file or directory |
| 00:17.06 | rwhitby-web | nope, is this after a full distclean (as lots of stuff changed last night)( |
| 00:17.34 | jacques | it's after a new checkout |
| 00:17.54 | jacques | but never mind. it's happening with older checkouts too - I dunno what's wrong |
| 00:18.39 | jacques | i probably forgot something |
| 00:19.46 | jacques | tho I am curious if it's currently working for anyone |
| 00:24.06 | jacques | errno.h is in my toolchain |
| 00:25.49 | jacques | WTF? |
| 00:25.50 | rwhitby-web | make distclean ; make is working for me with latest CVS (that was our criteria for finishing up last night on the staging dir policy changes) |
| 00:26.05 | jacques | errno.h is in include/linux |
| 00:26.12 | jacques | but busybox is looking in just include |
| 00:26.42 | jacques | if I change #include <errno.h> to #include <linux/errno.h> it gets past it |
| 00:27.19 | jacques | so what is the deal? |
| 00:27.29 | rwhitby-web | glc: have you done a make distclean? [glc is reporting can't compile bugs on stuff that does compile for us] |
| 00:27.52 | jacques | hmm interesting |
| 00:28.19 | jacques | there is in errno.h in include on my host, but not in my toolchain |
| 00:28.28 | jacques | why have I not hit this before - same toolchain ? |
| 00:28.57 | *** join/#nslu2-linux Tiersten (tman@ideal.trejan.com) |
| 00:29.08 | ka6sox-laptop | $path? |
| 00:29.14 | ka6sox-laptop | er $PATH |
| 00:29.42 | jacques | ok my toolchain is screwed |
| 00:29.46 | ka6sox-laptop | or whatever its called in the build chain. |
| 00:29.54 | jacques | older toolchains have errno.h in include |
| 00:30.18 | jacques | maybe I never tried this particular chain - I have like 8 |
| 00:30.19 | jacques | :-\ |
| 00:30.28 | rwhitby-web | do people think I should make the bugtracker send emals to the list? |
| 00:30.37 | rwhitby-web | s/emals/emails/ |
| 00:31.06 | ka6sox-laptop | list or the owners of the bug? |
| 00:31.07 | rwhitby-web | for bugs that are assigned to nslu2-linux (i.e. for packages that don't have a defined maintainer) |
| 00:31.24 | ka6sox-laptop | yes that would be good to see who rises to fix them. |
| 00:33.37 | rwhitby-web | so the only trouble is that if I set the email address for the nslu2-linux account to the group address, then someone could tell it to email the password. |
| 00:34.12 | Tiersten | Modify the bugtracker to not send the password for that account |
| 00:34.45 | ka6sox-laptop | oops |
| 00:34.57 | ka6sox-laptop | let me look at the php for that. |
| 00:36.24 | Tiersten | Shouldn't take more than a few minutes to hack that in |
| 00:38.35 | rwhitby-web | yeah - we were hoping not to have to hack source, so we could upgrade to newer versions easily |
| 00:39.05 | rwhitby-web | I could just make that account non-priviledged, so then it didn't matter if someone changes the password. |
| 00:39.15 | rwhitby-web | [and we'll all see if it happens] |
| 00:39.45 | ka6sox-laptop | that might be a good idea |
| 00:39.51 | ka6sox-laptop | better than me hacking the pw |
| 00:40.01 | ka6sox-laptop | (allows for upgrades too |
| 00:45.05 | CIA-3 | 03kas11 * 10anyu2/source/nslu2-io.c: reverting nslu2-io.c to a version that builds |
| 00:55.58 | kas11 | GPIO8 turns the power off on the NSLU2 |
| 00:56.14 | Tiersten | Uh huh.... |
| 00:56.18 | Tiersten | It's in the wiki |
| 00:56.35 | kas11 | haven't looked lately |
| 00:56.46 | ka6sox-laptop | bye for now |
| 00:57.26 | Tiersten | I traced it out last week and asked Allan to confirm it |
| 00:57.38 | kas11 | at least I can power down now without using the powr cord :) |
| 00:57.52 | Tiersten | Yup |
| 00:58.10 | Tiersten | Does mean that we don't have any spare GPIOs at all now though |
| 00:58.27 | Tiersten | Everything has been accounted for |
| 00:58.29 | kas11 | yup...all used up |
| 00:58.52 | Tiersten | Wonder if VoodooZ managed to reuse one of the LED ones to replace his blown I2C line |
| 01:00.01 | kas11 | he got i2c working...his rtc was wedged, I guess...a comlete powerdown was called for |
| 01:00.12 | Tiersten | I thought that was rwhitby? |
| 01:00.40 | Tiersten | I know he somehow locked up the RTC and needed to remove the battery to fix it |
| 01:00.48 | *** join/#nslu2-linux _chris_ (~chris@pcp09944349pcs.hyatsv01.md.comcast.net) |
| 01:00.50 | _chris_ | re |
| 01:00.53 | Tiersten | Hi |
| 01:00.58 | kas11 | I suggested it for VoodooZ...but as I recall rwhitby thought that might help him too |
| 01:01.17 | Tiersten | Ah |
| 01:01.32 | Tiersten | Bit of a crap design if locking up the RTC causes it to affect the I2C bus as well! |
| 01:01.50 | Tiersten | It was the clock line as well from what I remember |
| 01:01.57 | Tiersten | The RTC shouldn't ever be driving that |
| 01:02.23 | kas11 | there may be issues with the oscillator quitting too...we may want to play with the digital padder |
| 01:02.49 | Tiersten | The what? |
| 01:03.27 | Tiersten | I'll ask VoodooZ next time he's around if he managed to get it fixed anyway |
| 01:03.51 | *** join/#nslu2-linux mickg10_ (~ZOG@ool-18bacfea.dyn.optonline.net) |
| 01:04.01 | *** part/#nslu2-linux mickg10_ (~ZOG@ool-18bacfea.dyn.optonline.net) |
| 01:04.18 | kas11 | there is an adjustable digital capacitance padder in the x1205 |
| 01:04.25 | rwhitby-web | kas11: BTW, thanks for the RTC battery removal tip :-) |
| 01:04.49 | rwhitby-web | my production slug now has a working clock again |
| 01:05.10 | kas11 | rwhitby ...glad it worked for ya ;) |
| 01:05.24 | Tiersten | WTF is a capacitance padder? |
| 01:05.37 | Tiersten | You talking about the oscillator trimming feature? |
| 01:06.44 | kas11 | adds or subtracts capacitance across the oscillator tank to increase or decrease feedback as required for a stable frequency |
| 01:07.15 | Tiersten | aka the oscillator compensation trim capacitor... |
| 01:07.17 | rwhitby-web | bbiab |
| 01:08.13 | kas11 | the capacitance does pull the frequency up or down a bit...it is primarily intended to adjust the clock's accuracy...but if a crystal won't oscillate, you add a little feedback to get it to play |
| 01:08.56 | Tiersten | Yeah but the defaults should be fine. It's a defective crystal if it doesn't work |
| 01:10.02 | Tiersten | I've never heard of anybody refer to the trim cap as a capacitance padder *shrug* |
| 01:11.08 | kas11 | all crystals aren't created equal..and adding some padding beats having a dead rtc...and since it can be under software control... |
| 01:11.31 | Kaijuu | How can you know if your crystal is accurate anyways? |
| 01:11.42 | Tiersten | 32768hz crystals are mass produced to an insane degree. I've never seen a dud one |
| 01:11.48 | Tiersten | They're in nearly everything with a RTC |
| 01:11.53 | Tiersten | Kaijuu: umm... if it loses time? :) |
| 01:12.19 | Kaijuu | Ya, but how do you know if it is your clock that's loosing time or the crystal? |
| 01:12.37 | Tiersten | You don't. You use a known good time source |
| 01:12.49 | Tiersten | A GPS or whatever the equiv of a radio clock is in your country |
| 01:13.06 | Tiersten | Or NTP |
| 01:13.09 | Kaijuu | Hum.... GPS is a good idea |
| 01:13.35 | glc | I have too many freaking windows open/ |
| 01:13.36 | _chris_ | did the eclipse start yet? |
| 01:13.48 | Tiersten | If you use a GPS and you want really precise time measurements then you must take into account that it's UTC and not GMT |
| 01:14.04 | glc | everything compiles now. |
| 01:14.20 | Kaijuu | 10:20 I think. |
| 01:14.21 | Tiersten | eclipse? where? |
| 01:14.44 | Kaijuu | Can't see a thing... cloudy outside |
| 01:15.29 | glc | Lower left of moon just being touched now. |
| 01:17.31 | peteru | Clear skies here, but Sun is shining, so still can't see it. |
| 01:17.33 | Kaijuu | Hum... in PST time zone... start at 6:14 last until 9:54 |
| 01:17.49 | Kaijuu | maximum at 8:04 |
| 01:18.41 | glc | Is this an unexpected TOTAL ECLIPSE OF THE MOON? |
| 01:20.16 | Kaijuu | my bad, I can actually see the moon today |
| 01:24.22 | jacques | i bet it's cloudy here |
| 01:24.32 | jacques | ~weather kcvo |
| 01:24.37 | glc | I really like SlugBug. Now everyone who has the same problem can easily how it was resolved without wading through 20 days worth of logs. |
| 01:25.12 | Tiersten | Might still be a good idea to condense it down into a summary and link it into the wiki |
| 01:25.15 | jacques | yep - totally overcast |
| 01:26.45 | glc | I would hate to start cluttering up the wiki with tonnes of problem reports and resolutions when the wiki already points to SlugBug. |
| 01:26.59 | Kaijuu | It already started... I can see it start to clip on the edge... |
| 01:26.59 | Tiersten | True |
| 01:27.21 | _chris_ | east coast moon is goin under |
| 01:29.43 | glc | East coast of where? I am in Illinois, and it is still on the rise here. |
| 01:30.13 | _chris_ | glc: maryland. right near the DC cesspool |
| 01:31.21 | glc | Do you mean going under clouds? I would think it would be almost straight overhead there. |
| 01:31.36 | _chris_ | glc: the eclipse is beginning here :) |
| 01:32.37 | glc | Ok. At first I thought you meant the moon was going under the horizon. That's why I asked where you were. |
| 01:34.00 | Kaijuu | It sure takes a long time to build the crosstool... two hours already. |
| 01:34.16 | _chris_ | Kaiju: it took me about 45 mins. what kind of hardware do you have? |
| 01:36.30 | Kaijuu | P200 with 96meg of ram 2gig of hd. Debian with 2.4.25 kernel + gcc 3.0.4.... |
| 01:36.30 | CIA-3 | 03blunck2 * 10unslung/make/gkrellm.mk: updated to look in sources/gkrellmd for control file and start script |
| 01:36.48 | CIA-3 | 03blunck2 * 10unslung/sources/gkrellmd.rc: removed. renamed to sources/gkrellmd/rc.gkrellmd |
| 01:36.50 | jacques | well that explains that :-) |
| 01:37.06 | Kaijuu | I guess I use the wrong machine.... |
| 01:37.08 | CIA-3 | 03blunck2 * 10unslung/sources/gkrellm.control: removed. renamed to gkrellmd/control |
| 01:37.23 | jacques | I would recommend against trying oe with that machine :-) |
| 01:37.37 | CIA-3 | 03blunck2 * 10unslung/sources/gkrellmd/control: added. control file for gkrellmd. |
| 01:37.50 | CIA-3 | 03blunck2 * 10unslung/sources/gkrellmd/rc.gkrellmd: added. renamed from gkrellmd.rc |
| 01:40.08 | jacques | hmm rc37 is NOT working for me |
| 01:40.15 | jacques | someon on the wiki said no problems |
| 01:40.34 | jacques | /tmp/ccrbPm49.s: Assembler messages: |
| 01:40.34 | jacques | /tmp/ccrbPm49.s:5934: Error: bad immediate value for offset (4100) |
| 01:40.34 | jacques | make[2]: *** [/home/packages/crosstool/crosstool-0.28-rc37/build/armv5b-softfloat-linux/gcc-3.4.2-glibc-2.2.5/build-glibc/locale/ld-collate.o] Error 1 |
| 01:40.45 | jacques | tried two different binutils versions |
| 01:40.46 | *** join/#nslu2-linux [g2] (~g2@rdu57-6-117.nc.rr.com) |
| 01:42.01 | jacques | hmm this concurs: http://kegel.com/crosstool/crosstool-0.28-rc37/buildlogs/0.28/ |
| 01:42.06 | jacques | concurs with me that is |
| 01:42.41 | jacques | http://kegel.com/crosstool/crosstool-0.28-rc37/buildlogs/0.28/armv5b-softfloat-gcc-3.4.2-glibc-20040827.log.txt |
| 01:42.58 | jacques | ah well, today is not my day for things working :-( |
| 01:43.24 | _chris_ | hmm. |
| 01:43.33 | _chris_ | does glib build for anyone? (in the unslung module) |
| 01:44.00 | peteru | I have built crosstool-0.28-rc37 no probs |
| 01:44.10 | rwhitby-away | hasn't been converted to new staging policy yet. compare against template |
| 01:44.16 | jacques | peteru what exactly did you build? |
| 01:44.18 | _chris_ | rwhitby-away: lol |
| 01:44.35 | _chris_ | rwhitby-away: matching the template? bah! it doesn't even configure for me! ;) |
| 01:45.00 | _chris_ | rwhitby-away: i'm hoping someone pokes fun at me because i wrote it! :) |
| 01:45.37 | rwhitby-away | yep, check STAGING_DIR references - need to add .../opt/... |
| 01:45.53 | peteru | eval `cat armv5b-softfloat.dat gcc-3.3.4-glibc-2.2.5.dat` sh all.sh --notest |
| 01:48.03 | jacques | peteru oooh ok - you're building different gcc version |
| 01:48.07 | peteru | my dev machine is running gentoo with sys-devel/binutils-2.15.92.0.2-r1 |
| 01:48.34 | rwhitby-treo | go for 3.x |
| 01:48.41 | jacques | actually something is going right - I can see the moon after all :-) |
| 01:48.41 | peteru | oh. I missed the 3.4.2 part |
| 01:48.59 | jacques | i have build 3.4.2 just fine in the past |
| 01:49.13 | jacques | with older crosstool - I just wanted to see if this new version would fix my include problem |
| 01:49.25 | jacques | maybe I will just manually copy errno.h to the right place :-\ |
| 01:51.29 | CIA-3 | 03kas11 * 10anyu2/ (Makefile source/nslu2-io.c): beware...this unconditionallys kill the power when you press the button |
| 01:54.11 | [g2] | jacques, I built r37 just yesterday |
| 01:55.00 | rwhitby-treo | _chris_: any package that is in TO_BE_TESTED hasn't been converted yet, so please convert them as you go. |
| 01:55.32 | kas11 | [g2] what is in anyu2 cvs ought to be buildable now |
| 01:58.37 | [g2] | kas11 thx |
| 01:58.58 | [g2] | did you nail down buildroot ? |
| 02:00.59 | kas11 | operator error...inittab was getting cp'd to the wrong place in the script |
| 02:01.06 | Kaijuu | If you kill the monitorpowerbutton process, does the power button still works? |
| 02:02.12 | Kaijuu | * it is the CheckPowerButton process.... let me see. |
| 02:02.15 | _chris_ | brb |
| 02:02.17 | _chris_ | mood check |
| 02:02.21 | _chris_ | s/mood/moon |
| 02:02.26 | _chris_ | althought mood check would work too |
| 02:02.57 | Kaijuu | haha.... kill the process makes the power button useless... |
| 02:03.11 | _chris_ | still there |
| 02:03.46 | kas11 | there isn't anything doing the monitoring here at this point...the power button just pulls the plug...this isn't intended for endusers at this point since there isn't any real organized startup stuff yet |
| 02:05.17 | [g2] | kas11, the other night I pulled a fresh cvs and ran into buildroot2. Are you using buildroot2 ? |
| 02:06.06 | Kaijuu | moon is almost gone.... |
| 02:06.18 | kas11 | yes...pulled the latest...altho I see rc3 is still on the BB login message...thought that was a full release now |
| 02:07.03 | [g2] | kas11, when I went to build with 2.6.9 headers there were issues |
| 02:07.43 | kas11 | I have replaced/patched a header or 2...they are in the patches directory |
| 02:08.20 | kas11 | the Makefile will grab them for you tho if you run it |
| 02:09.08 | kas11 | skbuff.h was the primary offender IIRC |
| 02:10.16 | kas11 | you can pull the one from 2.6.7,,,they just dropped a member from a union that ixp425_eth.c was using |
| 02:12.20 | kas11 | the .S files are needed to conform to the newer layout DSaxena is using and to get the NSLU2 recognized but the boot code |
| 02:14.47 | rwhitby-away | kas11: this all still works with std redboot, right? |
| 02:16.06 | *** join/#nslu2-linux McTails (~mcasadeva@user-0cev0tp.cable.mindspring.com) |
| 02:16.22 | kas11 | there is a newer patch for ixp425_eth.c because I never could get the suppled ones to apply correctly...1 hunk was dupped...since it is apt to be downloaded with the patches, I decided it was just better to supply my own patch until it gets squared away upstream...since the rest of the csr isn't gpl'd. I didn't expect to get much sympathy anywhere |
| 02:17.10 | kas11 | yepper...no changes to Redboot...it is all very mcuh like your original anyu2...just updated as necessary |
| 02:17.18 | McTails | hi |
| 02:17.24 | _chris_ | hi |
| 02:18.49 | rwhitby-treo | McTails: can you remove the unused stuff from the sf.net nslu web area? |
| 02:18.55 | kas11 | rwhitby and dyoung did all the heavy lifting...I just rode their coat tails |
| 02:19.33 | rwhitby-treo | kas11: you have gone far beEyond what we started! |
| 02:20.39 | kas11 | if you hadn't started it, it wouldn't be where it is today...I learned a great deal by following in your footsteps |
| 02:21.28 | McTails | what unused stuff ? |
| 02:21.33 | kas11 | but darn, it is fascinating when you get into this stuff...;) |
| 02:22.16 | rwhitby-treo | McTails: have a look |
| 02:22.44 | kas11 | sorta kewl to keep a 3200 A64 pegged for 30 minutes watching it all scroll by |
| 02:23.22 | kas11 | [g2] yep..and I am hooked ;) |
| 02:23.28 | rwhitby-treo | McTails: pretty much anything other than the wiki and ipkg feed |
| 02:24.09 | rwhitby-treo | remove any unused stuff that has your name in it |
| 02:24.51 | rwhitby-treo | only wiki and cvs remain on sf.net now |
| 02:25.09 | *** join/#nslu2-linux dyoung-nslu2-log (~dyoung@cpe-24-94-83-226.hawaii.rr.com) |
| 02:25.09 | *** join/#nslu2-linux dyoung-logger (~dyoung@cpe-24-94-83-226.hawaii.rr.com) |
| 02:25.47 | McTails | rwhitby-treo: How about the archives of the bulletion boardand the irclogfiles folder ? |
| 02:26.02 | CIA-3 | 03blunck2 * 10unslung/make/glib.mk: |
| 02:26.02 | CIA-3 | moved glib.cache and glib.control into "glib" directory. |
| 02:26.02 | CIA-3 | also added --disable-rebuilds option to configure |
| 02:26.02 | CIA-3 | added prefix. trying to be more conformant to standards |
| 02:26.13 | rwhitby-treo | do they have your name on them? |
| 02:26.20 | CIA-3 | 03blunck2 * 10unslung/sources/glib.cache: removed. relocated to sources/glib/glib.cache |
| 02:26.31 | McTails | Nope |
| 02:26.33 | CIA-3 | 03blunck2 * 10unslung/sources/glib.control: removed. relocated to sources/glib/control |
| 02:27.02 | rwhitby-treo | then don't mess with them :-) |
| 02:27.07 | McTails | ok, done |
| 02:27.08 | CIA-3 | 03blunck2 * 10unslung/sources/glib/control: added. glib control file. snatched from ../glib.control |
| 02:27.27 | rwhitby-treo | McTails: thanks |
| 02:27.30 | McTails | welcome |
| 02:27.50 | CIA-3 | 03blunck2 * 10unslung/sources/glib/glib.cache: |
| 02:27.50 | CIA-3 | added. snatched from ../glib.cache. |
| 02:27.50 | CIA-3 | also added glib_cv_use_pid_surrogate=no as an argument to avoid a |
| 02:27.50 | CIA-3 | configure-time error |
| 02:28.25 | _chris_ | can someone do me a favor and check out the glib package? |
| 02:28.40 | McTails | rwhitby-treo: Vaguely curious, do you think the NSLU could handle Apache, PHP, MySQL, and Mailman for a development enviroment ? |
| 02:29.07 | rwhitby-treo | will it fit in 32Mb? |
| 02:29.09 | Tiersten | kas11: When you say they dropped a member from the union, are you sure that you're referring to a valid variable now that you've put it back? |
| 02:29.23 | McTails | rwhitby-treo: No idea |
| 02:29.53 | Tiersten | Fiddling with the struct and not checking the corresponding code is dangerous. If this variable no longer exists and that chunk of memory is used for something else then you might be stomping on something important |
| 02:30.10 | Tiersten | i.e. They changed it for a reason. Do you know the reason? |
| 02:30.14 | *** part/#nslu2-linux McTails (~mcasadeva@user-0cev0tp.cable.mindspring.com) |
| 02:31.41 | kas11 | Tiersen well, it is a union...a large one at that...it just appears that no one else was using that member...there no space issue if that is what you mean |
| 02:32.05 | Tiersten | Yes but if that member was dropped then it might mean they're no longer updating it in the other parts of the code |
| 02:32.30 | Tiersten | What exactly is the change? |
| 02:32.55 | kas11 | and since Intel is sorta on thin ice with the status of their code (ixp425_eth.c is gpl but the rest of csr isn't), I doubt anyone else cared if they noticed |
| 02:33.26 | Tiersten | I believe what Intel are doing is legal. It's not to the spirit of the GPL but it's legal |
| 02:33.50 | Tiersten | It's similar to what the Lucent Winmodem drivers did. There was a binary only chunk and a source wrapper |
| 02:34.48 | kas11 | do I expect it to break anything in userland...no |
| 02:35.09 | Tiersten | I'm not asking whether it'll break anything in userland |
| 02:35.29 | Tiersten | I'm asking whether changing the union back to the old one is the correct way to "fix" it |
| 02:36.41 | kas11 | rwhitby I am not a zealot...but I don't expect sympathy from upstream...Alan Cox's name is on the header in question...I am not going to try to persuade him to change it back *VBG* |
| 02:36.48 | Tiersten | If just adding back the member to the struct fixes it then it means the ixp425_eth code referring to it is redundant and can be removed |
| 02:37.28 | Tiersten | However, if it's some flag or something and the default behaviour is good for 99% of cases then adding back the member may be making it use an uninitialised value |
| 02:37.37 | Tiersten | It'll be good for 99% of the time but blow up randomly |
| 02:38.36 | Tiersten | rwhitby: Ya. Companies have found dubious loopholes in it. It technically follows the GPL but doesn't follow the spirit of it. Like these semi binary only drivers |
| 02:39.06 | [g2] | nod. Linus has made his thoughts clear on that |
| 02:39.09 | Tiersten | They make a stub open source driver which interfaces with the rest of the kernel and a massive binary only bit which talks to their stub |
| 02:39.29 | Tiersten | [g2]: Yeah. We'd be a lot better off without them |
| 02:39.52 | [g2] | except for co's like nvidia he didn't think it should be there |
| 02:39.56 | kas11 | ixp425_eth.c isn't bad to look at...but the rest of the stuff...ye gods |
| 02:40.05 | rwhitby-treo | yeah, but if they couldn't use that loophole, then they wouldn't use linux, and we wouldn't be able to play with them. |
| 02:40.47 | Tiersten | ah yes. the open source position on that is if they'd be "forced" to make an open source driver for that |
| 02:40.59 | kas11 | anyhow, the union problem needs to be addressed...but for now it isn't going to break anything...just pita |
| 02:41.01 | Tiersten | The real life position which most companies will take is "Screw you" |
| 02:41.04 | rwhitby-treo | cause they sure enough aren't going to give their competitors the source to those drivers |
| 02:41.12 | Tiersten | Yup |
| 02:41.46 | rwhitby-treo | so we live with a non-ideal situation ... |
| 02:41.53 | Tiersten | In the case of the huge nVidia driver, even if they didn't care about their competitors they wouldn't be allowed to. Alledgely it contains technology licensed from other companies |
| 02:42.54 | Tiersten | kas11: *shrug* I've not had the time to look into it closely but I still think replacing skbuff.h with the old version is not the right decision. It'll come back to bite us this decision. |
| 02:43.05 | rwhitby-treo | yep, cutting edge technology and open source drivers are usually disjoint sets |
| 02:43.14 | kas11 | There may be a patch already...I haven't made an exhaustive search...and I was disgusted enough with some of the other patches not applying that I said rude words and took the easy way out |
| 02:43.25 | *** join/#nslu2-linux dyoung-web (~80ab9a96@cpe-24-94-83-226.hawaii.rr.com) |
| 02:43.37 | Tiersten | rwhitby: Intel are fairly good about it actually. Their wifi gear has a open source project which they're helping out with |
| 02:44.22 | rwhitby-treo | yep, no longer cutting edge |
| 02:44.51 | Tiersten | a/b/g support isn't cutting edge enough for you? ;) |
| 02:45.11 | rwhitby-treo | talk 'n' to me |
| 02:45.13 | Tiersten | Intersil/Harris/whoever-they-are-now also have a open source driver. Not sure if they help out however |
| 02:45.27 | Tiersten | Broadcom aren't good though |
| 02:45.27 | ka6sox | I like the intel G driver. |
| 02:45.46 | ka6sox | Broadcom frankly are anti-opensource |
| 02:45.48 | ka6sox | :P |
| 02:45.59 | Tiersten | rwhitby & [g2] should know all about Broadcom because of the WRT54G :) |
| 02:46.24 | dyoung-web | I like Kaijuu's handle. ;-) |
| 02:46.27 | [g2] | maybe the new CEO from Phillips or Siemens will change that a little |
| 02:46.41 | Kaijuu | :P |
| 02:46.42 | [g2] | hey dyoung-web ! |
| 02:46.53 | dyoung-web | Hi, |
| 02:46.54 | Tiersten | dyoung: huh? care to explain to somebody that doesn't understand? :) |
| 02:47.25 | rwhitby-web | http://www.angelfire.com/vamp/kyuuketsuki0/Kaijuu.html |
| 02:47.27 | dyoung-web | Kaijuu. You hear it in all the japanese 'godzilla' genre of films. It means monster. |
| 02:47.35 | Tiersten | ah |
| 02:47.37 | Kaijuu | haha.... |
| 02:48.01 | ka6sox | I was wondering. |
| 02:48.44 | *** join/#nslu2-linux wiredduns (~duns@donszelmann-pbdsl1.Stanford.EDU) |
| 02:48.45 | [g2] | dyoung are you on land ? |
| 02:48.52 | Kaijuu | Secect revealed.... |
| 02:48.54 | dyoung-web | I'm on solid land now. |
| 02:49.01 | [g2] | wb |
| 02:49.06 | Tiersten | Yeah. wb |
| 02:49.07 | dyoung-web | thanks. |
| 02:49.24 | ka6sox | you call those Islands "solid land"? |
| 02:50.06 | dyoung-web | Compared to that liquid stuff I was on for the past weeks, YES. |
| 02:50.06 | rwhitby-web | [g2]: how's the port of kas11's stuff to OE going ? :-) |
| 02:53.32 | wiredduns | ? |
| 02:54.02 | dyoung-web | kas11, the one dumped hunk in the IXP425_eth.c code was for a comment; I thought we documented it properly... |
| 02:54.27 | [g2] | rwhitby-web, I'm trying to reproduce the regular cvs first |
| 02:54.36 | [g2] | then move to oe |
| 02:54.45 | rwhitby-treo | cool |
| 02:54.46 | [g2] | call me old-fashioned |
| 02:55.07 | rwhitby-treo | tried out 2.5 yet? |
| 02:55.09 | kas11 | Tiersen take a look...I'm no C expert...but I'll be darned if I see how it can possibly break anything |
| 02:55.40 | Tiersten | I explained this earlier. If you're using an uninitialised variable then it may blow up |
| 02:55.43 | rwhitby-treo | dyoung: can you torture test 2.5? |
| 02:55.55 | [g2] | kas11: make[1]: Leaving directory `/home/tom/dev/nslu2/anyu2/buildroot/build_armeb/busybox' |
| 02:55.55 | [g2] | cp /home/tom/dev/nslu2/anyu2/patches/.config.bb /home/tom/dev/nslu2/anyu2/buildroot/build_armeb/busybox/.config |
| 02:55.55 | [g2] | cp: cannot stat `/home/tom/dev/nslu2/anyu2/patches/.config.bb': No such file or directory |
| 02:55.55 | [g2] | make: *** [/home/tom/dev/nslu2/anyu2/buildroot/.built] Error 1 |
| 02:56.21 | dyoung-web | Um, I like torturing Unslung. ;-) |
| 02:56.33 | [g2] | rwhitby-treo, it's building. I had my packages parent pointing at openembedded since that last push |
| 02:56.42 | dyoung-web | I forgot to boot 'builder' this mroning, so its going to have to wait a few hours. |
| 02:56.54 | Tiersten | There are several possiblities for the change. 1. It was unused and they removed it. 2. They changed it and now it's part of something else 3. They don't want anybody using that variable since it's supposed to be private |
| 02:57.06 | [g2] | dyoung, what no remote power-on ? |
| 02:57.14 | Tiersten | What was changed in skbuff.h? Was the variable deleted, moved or renamed? |
| 02:57.17 | [g2] | dyoung-web, slacker :) |
| 02:57.22 | kas11 | dyoung ...this version of patch that I have seems partially borked...it wasn't honoring the -f or -N either...after messing with it for a while, I took the easy way...just applied them manually and did my own diff and used that as the patch |
| 02:58.27 | dyoung-web | Dammit jim! I didnt get around to the Remote Power-on applicatcion yet! |
| 02:58.52 | Kaijuu | It actually takes more time to build the cross tool than the kernel... what a supprise... |
| 02:58.58 | dyoung-web | I misplaced my NSLU2 power supply; so I gotta find that first. |
| 02:59.17 | jacques | if I wanna use arbitrary USB devices, should I assume I will be using openslug ? |
| 02:59.45 | dyoung-web | (Or, I could do Something Ugly(tm) ) |
| 02:59.55 | kas11 | Tiersen struct ethhdr *ethernet was removed from the mac union |
| 03:00.01 | rwhitby-treo | unslung just needs modules added, right, |
| 03:00.39 | rwhitby-treo | dyoung got a usb module to work |
| 03:01.08 | kas11 | if you want to do the cast and watch the compiler complain, you could use *raw...but tit won't make any difference |
| 03:02.06 | jacques | rwhitby-treo, I thought usb was crippled in the stock kernel - does re-enabling it violate the principles of unslung ? |
| 03:02.37 | rwhitby-treo | tiersten asserts that it's not cripPled |
| 03:02.41 | [g2] | kas11, there's no patches directory for anyu2! |
| 03:02.47 | rwhitby-treo | just missing modules |
| 03:02.57 | dyoung-web | jacques, I believe adding USB stuff makes it into SlugPlus. |
| 03:03.18 | jacques | I was thinking of making a printserver |
| 03:03.24 | jacques | for my usb2.0 laser |
| 03:03.29 | kas11 | ok...I thought cvs and I were getting along these day...let me look |
| 03:03.35 | rwhitby-treo | I hope to have usb-serial package for unslung |
| 03:03.47 | dyoung-web | I built that one too. It works. |
| 03:04.12 | rwhitby-treo | soon as my hw flow control cable from ka6sox arrives |
| 03:04.32 | dyoung-web | You have a USB Flow Control cable coming? |
| 03:04.51 | rwhitby-treo | so I can move the dial-in modem from the wrt to the slug and have hw flow control |
| 03:04.54 | [g2] | I've got a usb-to-usb cable |
| 03:05.19 | rwhitby-treo | ka6sox is modding the RS cable |
| 03:05.31 | rwhitby-treo | the chip has the signals |
| 03:05.46 | dyoung-web | Hmm.. |
| 03:05.50 | dyoung-web | Good point. |
| 03:06.00 | ka6sox | got the idea from you dyoung-web! |