| 00:07.12 | CIA-6 | 03rwhitby * r3778 10optware/trunk/Makefile: Demoted wxbase on oleg - failes to build on nudi |
| 00:36.34 | CIA-6 | 03bzhou * r3779 10optware/trunk/ (Makefile make/netcat.mk): added netcat |
| 01:41.11 | rwhitby | Uploading new ddwrt, oleg and fsg3 optware feeds now ... |
| 01:41.31 | rwhitby | http://ipkg.nslu2-linux.org/feeds/optware/ddwrt/cross/unstable/ |
| 01:41.57 | rwhitby | ditto for "oleg" (wl500gx with Oleg's firmware and updated uclibc libraries) and "fsg3" (Freecom FSG-3) |
| 01:59.11 | *** join/#nslu2-linux yvasilev (n=yvasilev@201.144.25.101) |
| 02:08.20 | eno-away | rwhitby, this is really great |
| 02:17.59 | rwhitby | oleo: I had to demote some uclibc packages to push the feeds through - we can work through the failures and see if any can be repromoted. |
| 02:45.11 | *** join/#nslu2-linux nm (n=nm@222.252.56.221) |
| 02:55.38 | rwhitby | eno-away: ping |
| 02:59.54 | CIA-6 | 03rwhitby * r3780 10optware/trunk/make/buildroot.mk: buildroot: dirclean should not touch the toolchain dir - that's what the top-level toolclean target is for |
| 03:56.02 | eno-away | rwhitby, pong |
| 03:57.19 | rwhitby | eno: see /msg |
| 04:04.32 | *** join/#nslu2-linux ZvQ (n=Sinclair@p54A0A022.dip0.t-ipconnect.de) |
| 04:10.54 | *** join/#nslu2-linux nailbags|laptop (n=neil@c220-237-123-137.randw1.nsw.optusnet.com.au) |
| 05:01.08 | *** join/#nslu2-linux iDc (n=Sinclair@p54A0ADD9.dip0.t-ipconnect.de) |
| 05:57.14 | *** join/#nslu2-linux jacques (n=jacques@cpe-66-69-157-96.houston.res.rr.com) |
| 06:21.06 | *** join/#nslu2-linux pepijn (n=pepijn@195.159.157.158) |
| 06:27.29 | rwhitby | we might even get a clean autobuild this time around ... |
| 06:35.25 | CIA-6 | 03bzhou * r3781 10optware/trunk/make/py-buffet.mk: py-buffet: upstream upgrade to 1.0 |
| 06:42.35 | *** join/#nslu2-linux l0c4lh0st (n=a@f178138.upc-f.chello.nl) |
| 07:19.05 | *** join/#nslu2-linux nm (n=nm@222.252.51.21) |
| 07:20.43 | *** join/#nslu2-linux marceln (n=marceln@pion.xs4all.nl) |
| 08:30.01 | rwhitby | clean autobuilds. Let's keep them that way :-) |
| 09:16.21 | NAiL | Anyone got any good recommended filesystem choices? I've been burned with ext3 too many times now :-( |
| 09:16.34 | NAiL | (This is for my desktop) |
| 09:17.34 | NAiL | Yesterday, my filesystem was completely destroyed for no apparent reason |
| 09:18.41 | eFfeM | there is ext4 but I have no idea how mature it is |
| 09:18.53 | pepijn | how about jfs? |
| 09:19.01 | eFfeM | alternative could be reiser but again no info how this compares to ext3 (but reiser is mature) |
| 09:19.13 | eFfeM | ntfs :-) |
| 09:19.31 | eFfeM | i've never really had problems with ext3 |
| 09:19.52 | NAiL | I've had problems with xfs and ext2/3 |
| 09:20.07 | NAiL | No idea what causes it... been on a bunch of machines |
| 09:23.03 | eFfeM | afaik unix/linux fs issues have mostly one of the following reasons |
| 09:23.12 | eFfeM | - someone writing to the raw device |
| 09:23.20 | eFfeM | - turning off the system without sync |
| 09:23.27 | eFfeM | - bad disk |
| 09:23.49 | pepijn | that is in reverse order of likelyhood |
| 09:23.56 | NAiL | yeah |
| 09:24.14 | eFfeM | what also has happened to me (also on slug) is that I lost some files, but that again could be due to turn off without sync or ipkg or so making a mess out of things |
| 09:24.36 | NAiL | I suspect it was caused by a power outage which fsck didn't catch at bootup, and it then propagated through the entire filesystem until it was broken beyond repair |
| 09:25.06 | NAiL | or, fsck did catch it, but it just replayed the journal... didn't check the rest of the fs. |
| 09:25.43 | eFfeM | fsck should catch it, as there is a marker on the disk that is erased when you cleanly shutdown |
| 09:25.48 | NAiL | I know |
| 09:26.01 | eFfeM | however, I know about one other issue that I've seen once (and actually had strike on me) |
| 09:26.54 | NAiL | My issue with XFS was that if I lost power, any files open were overwritten with ^@ |
| 09:27.13 | eFfeM | if you have a fairly new disk it'll come with e.g. 8mb of cache, now if just before shutting down you have written a lot to the disk (e.g. in the final sync) it may be that the disk has accepted the data, but not have written them to disk yet when the system actually powers off |
| 09:27.19 | NAiL | (eg. xf86config) |
| 09:28.53 | kolla | I use ext3, but with noatime,sync for my /home |
| 09:30.16 | NAiL | good point with the sync atleast :) |
| 09:31.29 | kolla | on the laptop I use sync everywhere but /var and /tmp |
| 09:31.49 | kolla | and only remount / and /usr async when upgrading |
| 09:32.01 | kolla | and everything noatime |
| 09:32.21 | NAiL | I think I'll try that |
| 09:32.22 | NAiL | thanks |
| 09:32.49 | kolla | there is a performance penalty though :) |
| 09:33.09 | NAiL | I don't care *that* much about it right now |
| 09:33.13 | kolla | but I grew up with computers that I could turn off by using the power switch, and I kinda like that :) |
| 09:33.23 | NAiL | I need a filesystem that will work for more than a couple of weeks |
| 09:33.33 | kolla | :) |
| 09:33.37 | NAiL | (atleast until I can set up a backup of some sort) |
| 09:33.57 | NAiL | I lost a fair bit of stuff and work yesterday :( |
| 09:34.25 | kolla | slug related also? :/ |
| 09:34.29 | kolla | hehe |
| 09:35.04 | NAiL | yup |
| 09:35.24 | NAiL | I had set up my dev env and was working on a couple of the bugs in the tracker :( |
| 09:36.13 | kolla | oh :/ |
| 09:37.37 | kolla | I'll try to build my own kernel with usb-serial console |
| 09:38.29 | kolla | my "gentoo slug" never gets online, so I need a console on it :) |
| 09:38.53 | kolla | it startet when I upgraded the openslug on the flash |
| 09:43.11 | NAiL | bbl |
| 09:48.06 | eFfeM | cu l8r |
| 11:31.40 | NAiL | *sigh* |
| 11:31.41 | NAiL | > > Nvidia Nforce4 chipset with Maxtor SATA II drives with certain firmware |
| 11:31.41 | NAiL | > > revisions cause data corruption and system instability when under |
| 11:31.41 | NAiL | > > moderate to heavy I/O load. |
| 11:50.43 | *** join/#nslu2-linux gda_ (n=gda@dslb-082-083-128-122.pools.arcor-ip.net) |
| 11:52.01 | gda_ | rwhitby: who can help me with the cross perl thing, eno told me he has not enough time |
| 11:52.23 | gda_ | I need only someone to discuss from time to time |
| 11:53.07 | gda_ | currently I have working a cross perl 5.8.8 that uses its own host perl |
| 11:54.06 | gda_ | I use this host perl successfull to build cross packages like per-archive-tar, but have to verify that it still works native on the nslu2 |
| 11:54.14 | gda_ | but I have no nslu2 |
| 12:02.15 | *** join/#nslu2-linux bullet (n=bullet@248.219.203.62.cust.bluewin.ch) |
| 12:04.34 | *** join/#nslu2-linux vmaster_ (i=vmaster@p549B758B.dip.t-dialin.net) |
| 12:39.58 | *** join/#nslu2-linux EvilDevil_ (n=miau@p54A6DE7A.dip.t-dialin.net) |
| 12:58.41 | *** join/#nslu2-linux emm_is (n=chatzill@i-83-67-201-219.freedom2surf.net) |
| 13:06.55 | *** join/#nslu2-linux marcusb (n=marcus@67.152.216.81.static.g-va.siw.siwnet.net) |
| 13:17.46 | *** join/#nslu2-linux kolla (n=kolla@250.80-203-244.nextgentel.com) |
| 13:21.34 | *** join/#nslu2-linux blomma (n=blomma@c-1b9b72d5.014-202-6e6b701.cust.bredbandsbolaget.se) |
| 13:40.29 | *** join/#nslu2-linux blomma (n=blomma@c-1b9b72d5.014-202-6e6b701.cust.bredbandsbolaget.se) |
| 13:42.15 | *** join/#nslu2-linux fleetfoot (n=fleetfoo@c-1b9b72d5.014-202-6e6b701.cust.bredbandsbolaget.se) |
| 13:55.38 | *** join/#nslu2-linux fleetfoot (n=fleetfoo@c-1b9b72d5.014-202-6e6b701.cust.bredbandsbolaget.se) |
| 14:21.59 | *** join/#nslu2-linux kolla_ (n=kolla@250.80-203-244.nextgentel.com) |
| 14:57.45 | *** join/#nslu2-linux yvasilev (n=yvasilev@201.144.25.101) |
| 15:07.58 | *** join/#nslu2-linux EvilDevil (n=miau@p54A6DE7A.dip.t-dialin.net) |
| 15:25.07 | *** join/#nslu2-linux EvilDevil_ (n=miau@p54A6DE7A.dip.t-dialin.net) |
| 16:17.22 | eno | gda_, i just replied your email on -developers |
| 16:17.49 | eno | i would move only ds101g to cross perl 5.8.8 for now |
| 16:18.51 | eno | it is ok in some special cases, to have different platforms with diff versions |
| 16:19.16 | eno | with the goal to have same version in the end |
| 16:22.11 | eno | of course, make/perl.mk would be a bit ugly |
| 16:30.41 | gda_ | eno: sorry I am new to this, I am used to use a completely different build environment |
| 16:31.10 | gda_ | eno: how can I have two perl.mk files for different platform? |
| 16:31.27 | gda_ | s/platform/platforms/ |
| 16:35.33 | eno | that's why i say make/perl.mk will be a bit ugly, sprinkled with ifdef's |
| 16:35.49 | gda_ | I see, starting with: |
| 16:35.53 | gda_ | ifeq ($(OPTWARE_TARGET),ds101g) |
| 16:36.02 | gda_ | PERL_VERSION=5.8.8 |
| 16:36.04 | gda_ | else |
| 16:36.09 | gda_ | PERL_VERSION=5.8.6 |
| 16:36.11 | gda_ | endif |
| 16:36.12 | gda_ | ? |
| 16:36.28 | eno | i think sources/perl/Cross/perl.mk is basically backward (native) compatible |
| 16:36.35 | eno | yes |
| 16:36.42 | gda_ | okay |
| 16:37.08 | eno | i understand it is ugly, but i want to make one change at a time |
| 16:37.23 | gda_ | I know, you don't have time, but could you verify that it is native compatible? |
| 16:38.02 | eno | i will do that |
| 16:38.08 | gda_ | another question |
| 16:38.41 | gda_ | I have started to use your perl-hostperl inside of packages like per-compress-zlib ... |
| 16:39.19 | CIA-6 | 03bzhou * r3782 10optware/trunk/Makefile: promoted dspam except on wl500g |
| 16:39.27 | gda_ | so it is possible to build for perl 5.8.8 without having that installed, is this not helpful for native builds too? |
| 16:39.47 | gda_ | what problem with dspam? |
| 16:40.11 | eno | daemon.o: In function `daemon_listen': |
| 16:40.26 | eno | <PROTECTED> |
| 16:40.49 | eno | only on wl500g |
| 16:41.21 | gda_ | oh, I patched out the recognition of the right place for pthread because it doesn't work cross |
| 16:42.11 | eno | to native compilation, it is two folded, on the one hand, host-perl takes longer |
| 16:42.12 | gda_ | does wl500g don't have it, or only in another lib then expected, look inside my patch, you will know what I mean? |
| 16:42.41 | gda_ | s/then/than/ |
| 16:42.46 | eno | OTOH, it helps eliminate one build dependency |
| 16:43.15 | eno | but i would not worry too much about native host-perl, eventually we want to go cross |
| 16:44.54 | gda_ | the problem is, if all platforms use perl-hostperl, I need only some not platform specific changes to do for the perl-* packages |
| 16:46.54 | eno | i see your point. |
| 16:48.07 | gda_ | for dspam or perl? |
| 16:48.09 | eno | i'm thinking maybe we should have Cross/perl.mk stay where it is, even the couple of initial cross perl-* packages should go inside Cross/ |
| 16:48.42 | eno | just use ifdef to include them |
| 16:49.13 | gda_ | <PROTECTED> |
| 16:49.20 | eno | it won't be long before we switch to cross |
| 16:49.32 | eno | this is only for the time being |
| 16:49.57 | gda_ | make -C $(CROSS_DIR)/perl.mk |
| 16:50.00 | gda_ | endif |
| 16:50.10 | gda_ | inside perl.mk? |
| 16:50.40 | gda_ | or how to do it? |
| 16:51.05 | eno | something like that, but s|make -C $(CROSS_DIR)/perl.mk|include $(PERL_SOURCE_DIR)/Cross/perl.mk| |
| 16:51.38 | eno | with an "else" include the existing content of perl.mk |
| 16:51.52 | gda_ | will try that |
| 16:52.24 | gda_ | I like the idea |
| 16:54.01 | eno | i will try on the nslu2 native -> cross front in the mean time, maybe not today, but soon |
| 16:54.06 | gda_ | can you tell me whether wl500g really has no pthread_create, maybe in a lib in another path? |
| 16:55.08 | eno | for dspam, my philosophy is mark it as broken for wl500g now, if someone wants to use the package, they will try to fix it |
| 16:55.28 | gda_ | because of the armv5b-softfloat-linux -> armv5b-linux, it is a small change in the config.sh-arm* file |
| 16:55.30 | eno | i'm not that familiar with wl500g either |
| 16:55.46 | gda_ | change archname |
| 16:56.14 | eno | hope it's just a small change |
| 16:56.58 | gda_ | I had to do it too, because it is ppc-linux and not powerpc-603e-linux on the ds101g |
| 16:57.58 | eno | i c, thanks for the pointer, will definitely try that |
| 16:58.02 | gda_ | archname is only one line, not too heavy change |
| 16:59.03 | gda_ | bbl |
| 16:59.05 | eno | for the dspam compilation problem, if you want to fix it for wl500g, you need to setup wl500g toolchain |
| 17:00.07 | eno | alright, gda_, i'm glad we can proceed in parallel |
| 17:14.15 | gda_ | eno: sorry, but without the hardware I lack a little bit motivation to setup the toolchain |
| 17:16.14 | CIA-6 | 03caplink811 * r3783 10optware/trunk/sources/clamav/postinst: don't use chgrp anymore, because it's doesn't exist in the unslung fw, use chown owner:group instead |
| 17:22.29 | eFfeM | hm, doesn't nudi build ucslugc ? |
| 17:22.35 | eFfeM | can't find it in the log |
| 17:23.34 | eno | eFfeM, http://groups.yahoo.com/group/nslu2-developers/message/1051 |
| 17:24.17 | eFfeM | ah, ok, thanks |
| 17:24.52 | eFfeM | actually there was a ticket with my name on it for too long and I just decided to give it attention |
| 17:25.05 | eFfeM | actually patched one issue with uclibc |
| 17:25.29 | eFfeM | it is now compiling with gcc 3.4.4; want to try with gcc 4 later |
| 17:25.37 | eFfeM | will ask rod to re-enable if I get things working |
| 17:26.16 | CIA-6 | 03gda * r3784 10optware/trunk/ (make/perl.mk sources/perl/Cross/perl.mk): perl: 5.8.8 gets build cross for ds101g, all other gets unchanged 5.8.6 |
| 17:30.33 | CIA-6 | 03gda * r3785 10optware/trunk/Makefile: enabled build of perl for ds101g |
| 17:33.02 | EvilDevil | hmm... anybody knows where to find the Intel NPE + Accesslibrary 2.1 ? They released 2.3 a few days ago and I can't find the older ones |
| 17:33.55 | NAiL | http://www.intel.com/design/network/products/npfamily/download_ixp400.htm |
| 17:33.58 | NAiL | under "archived" |
| 17:35.15 | CIA-6 | 03gda * r3786 10optware/trunk/make/perl.mk: perl: stupid bug |
| 17:36.17 | EvilDevil | thanks NAiL |
| 17:40.58 | *** join/#nslu2-linux caplink811_log (n=caplink8@dslb-088-073-049-145.pools.arcor-ip.net) |
| 17:41.01 | *** join/#nslu2-linux AdamBaker (n=aab@userch028.dsl.pipex.com) |
| 17:46.56 | eFfeM | anyone actaully already working on this 2.3 version ? |
| 17:50.14 | CIA-6 | 03bzhou * r3787 10optware/trunk/ (Makefile make/dspam.mk): dspam: libpthread problem fixed, promoted for wl500g |
| 17:53.35 | CIA-6 | 03bzhou * r3788 10optware/trunk/Makefile: demote dspam on ds101, indirectly depend on libstdc++-stage, broken on ds101 |
| 18:02.48 | *** join/#nslu2-linux caplink811_log (n=caplink8@dslb-088-073-049-145.pools.arcor-ip.net) |
| 18:09.35 | *** join/#nslu2-linux caplink811_log (n=caplink8@dslb-088-073-049-145.pools.arcor-ip.net) |
| 18:11.49 | CIA-6 | 03bzhou * r3789 10optware/trunk/make/pound.mk: pound: upstream upgrade to 2.1 |
| 18:13.50 | CIA-6 | 03caplink811 * r3790 10optware/trunk/sources/miau/paths.patch.0.6.1: reflected update of mainstream version (0.6.2) |
| 18:17.10 | CIA-6 | 03caplink811 * r3791 10optware/trunk/make/miau.mk: reflected update of upstream version (0.6.2), bumped ipk-version |
| 18:26.35 | *** join/#nslu2-linux caplink811_log (n=caplink8@dslb-088-073-049-145.pools.arcor-ip.net) |
| 18:33.22 | CIA-6 | 03bzhou * r3792 10optware/trunk/make/perl.mk: small improvement so that we can list PERL_CROSS_TARGETS (not necessarily all enabled in optware/Makefile) |
| 18:43.50 | *** join/#nslu2-linux _kaenat (n=ringlej@hrrsnbrg-bluewave1-69-161-6-246.chvlva.adelphia.net) |
| 18:48.44 | *** join/#nslu2-linux Eiffel (n=x@pD9573404.dip0.t-ipconnect.de) |
| 18:53.01 | *** join/#nslu2-linux caplink811-log (n=caplink8@dslb-088-073-050-058.pools.arcor-ip.net) |
| 19:07.49 | CIA-6 | 03marceln * r3793 10optware/trunk/ (8 files in 2 dirs): mysql5: Compiles but doesn't work yet. |
| 19:07.49 | *** join/#nslu2-linux kaenat_ (n=ringlej@hrrsnbrg-bluewave1-69-161-6-246.chvlva.adelphia.net) |
| 19:09.51 | *** join/#nslu2-linux bullet_ (n=bullet@78.25.79.83.cust.bluewin.ch) |
| 21:04.42 | gda_ | cool! nudi has just build the first cross perl for the ds101g! Maybe even the first cross perl at all? |
| 21:06.15 | eno-away | gda_, don't say that. Openslug had cross perl for quite some time |
| 21:06.43 | gda_ | okay, so only for ds101g |
| 21:06.58 | gda_ | I have thought you are away |
| 21:07.17 | gda_ | because you are eno-away |
| 21:07.18 | eno-away | i was away, and will be away again |
| 21:08.34 | eno-away | going to bleeding-edge festival, see yo la tengo performing this evening |
| 21:09.20 | gda_ | eno-away: I don't understand a single word, I assume you can't eat it |
| 21:09.54 | eno-away | "yo la tengo" is a cool band |
| 21:10.32 | gda_ | for me cool means "metallica" or "nirvana" |
| 21:10.55 | gda_ | but I guess I am a litte bit older |
| 21:12.00 | eno-away | well, actually ylt probably started earlier than nirvana, they have the same root as nirvana => punk |
| 21:13.25 | gda_ | I see, what about Led Zepplin of Manfred Mann |
| 21:14.04 | gda_ | s/of/and/ |
| 21:14.50 | eno-away | i like led zepplin too |
| 21:15.24 | eno-away | but recently listen more to alternatives, like velvet underground, sonic youth, mission of burma |
| 21:16.13 | eno-away | bleeding edge is alt.music fest |
| 21:20.05 | gda_ | currently I prefere nickelback and opera arias :) |
| 22:01.18 | *** join/#nslu2-linux npt-laptop (n=npt@12-218-7-27.client.mchsi.com) |
| 22:01.38 | *** join/#nslu2-linux npt (n=npt@12-218-7-27.client.mchsi.com) |
| 22:14.39 | *** part/#nslu2-linux marceln (n=marceln@pion.xs4all.nl) |
| 22:41.46 | *** join/#nslu2-linux npt_ (n=impulse@12-218-7-27.client.mchsi.com) |
| 23:23.08 | *** join/#nslu2-linux kfm (n=kfm82@p54BED9DF.dip.t-dialin.net) |