**** BEGIN LOGGING AT Sun Nov 13 02:59:57 2005 Nov 13 03:29:49 łorning Nov 13 03:47:23 morning all Nov 13 03:48:09 All the patches I requeued have been applied by rmk. This should be good news for Zaurus 2.6.15 :) Nov 13 03:49:09 yay! Nov 13 03:50:49 That adds PM support for c7x0 and cxx00 models into mainline so mainline is just missing sound... Nov 13 04:37:57 Hi. Everytime I install an ipkg that adds something to /etc the permission on /etc are being reset to drwx------ and this is breaking programs that run as non-root. Any idea's what is causing this? Nov 13 05:02:04 hi all Nov 13 05:02:16 hey hrw Nov 13 05:05:39 I have to trace one error in collie fstab Nov 13 05:11:30 hi hrw Nov 13 05:11:56 hrw: Those patches were accepted by rmk. Fingers crossed Linus isn't awkward about them... Nov 13 05:11:58 on collie with other then 64/0 kernel mtdram is not mounted by default with 3.5.4-rc Nov 13 05:12:25 RP: which ones? Nov 13 05:15:03 hrw: c7x0 and cxx00 power management and akita machine + io expander support Nov 13 05:15:13 ah. great Nov 13 05:16:09 RP: I could host the cxxx images on ewi546 if you want Nov 13 05:17:14 ntpdate is b0rken is our images Nov 13 05:17:33 koen: That might be an idea, thanks. I need to talk to mickeyl about what we're going to do about 2.6 on the cxx00's and then I'm sure we can work something out Nov 13 05:18:40 hr it is? Nov 13 05:18:54 root@h2200:~# ntpdate -b time.apple.com Nov 13 05:18:54 Looking for host time.apple.com and service ntp Nov 13 05:18:54 host found : time0.apple.com Nov 13 05:18:55 13 Nov 12:18:45 ntpdate[20315]: step time server 17.254.0.26 offset -5.093466 sec Nov 13 05:19:14 root@collie:~# ntpdate ntp.task.gda.pl Nov 13 05:19:15 Looking for host ntp.task.gda.pl and service ntp Nov 13 05:19:15 host found : ntp.task.gda.pl Nov 13 05:19:15 31 Oct 15:00:33 ntpdate[2385]: setsockopt() IPV6_V6ONLY failed: Protocol not available Nov 13 05:19:32 same for time.apple.com Nov 13 05:20:12 http://pastebin.com/427640 Nov 13 05:20:13 I dont have ipv6 net here. Nov 13 05:20:15 that's my setup Nov 13 05:20:32 I don't have ipv6 to the outside net Nov 13 05:21:08 http://pastebin.com/427643 is mine Nov 13 05:21:36 it could be missing a kernel option Nov 13 05:22:29 maybe - but I dont know which and why Nov 13 05:32:32 is it worth to get ipv6 over ipv4 tunnel? will it give me something other then hard to remember address? Nov 13 05:32:32 hrw: Does it work if you force ipv4? ie: "ntpdate -4b time.apple.com" Nov 13 05:32:56 v8jlene: no - result is the same Nov 13 05:33:27 rebooting to get rid of ipv6 module Nov 13 05:34:39 hrw: Very odd... looking at the code it shouldn't be doing anything ipv6 related with -4... and that erorr is from some IPv6 only code ;( Nov 13 05:36:38 ~curse wifi Nov 13 05:36:39 May the fleas of a thousand camels infest your most sensitive regions, wifi ! Nov 13 05:37:30 I dont know.. does it my AP or my wifi card is so shitted Nov 13 05:37:52 all what I have now is blinking led on wifi card Nov 13 05:38:39 I went through a few AP's before I found one that was stable ;) Nov 13 05:39:30 v8jlene: mine is wrt54gs Nov 13 05:40:08 It appears that directories in the ipkg files are set to permission based on the umask you have when you build? Do I need to change my umask to stop all the directories permissions being reset to owner only when I install ipkgs? Nov 13 05:41:34 hrw: I've always had good success with the wrt54g's... Nov 13 05:43:06 v8jlene: when I dont have ipv6 module loaded ntpdate works ok Nov 13 05:53:32 hrw: No idea then... the source code says you should only get that error if you force IPv6 and don't have IPv6 in the kernel. Could be some kernel and/or headers problem for that version of the kernel. Probably need to get someone else on the same version to check? koen had no trouble and is running a 2.6.12 and I have no trouble and am running a 2.6.14 kernel. Nov 13 05:54:08 I run 2.4.18 and it wont change on that machine Nov 13 05:54:50 unless kergoth writes an SD driver Nov 13 05:55:40 I lost that faith Nov 13 05:59:36 Arghhh... the umask you have when you build does dictate the permission on directories ;( I just re-build and re-installed my ipk with a umsk 0 and /etc has now been reset to drwxrwxrwx... (I've gone from one extreme to the other now...) And those permissions are reset with eery package you install... Nov 13 06:05:06 v8jlene: that is clearly a bug in the package you are building Nov 13 06:05:28 arguably, it is also a bug in ipkg, but I'm not sure that this is so clear-cut. Nov 13 06:06:26 pb_: How so? I created the directories with the correct permissions in do_install, but bitbake changes them when it copies them into the individual package directories. It lease the permissions on files alone, but not on directories. Nov 13 06:07:43 altboot looks nice - I will push it for collie Nov 13 06:10:10 v8jlene: oh, right. in that case, the bug must be in package.bbclass. Nov 13 06:10:46 I'm a bit surprised that nobody else has noticed the problem before, if it really is destroying the mode on all directories, but I guess anything is possible. Nov 13 06:15:32 pb_: I just installed openswan and /etc was reset from drwxrwxrwx back to drwx------ (the umask I had on yesterdays build) just to verify that it really is every package. I only noticed it because quagga changes to the quagga user before trying to read it's configuration files out of /etc/quagga. Each time I re-installed it the deamons failed to start because the permission on /etc were reset again! Maybe I'm the only person on the planet with umask=07 Nov 13 06:15:32 7 or running programs that actually care? Nov 13 06:16:25 v8jlene: virtually all programs will care about reading at least some files in /etc, even if just stuff like resolv.conf, ld.so.conf and passwd. Nov 13 06:16:39 I guess it's possible that you are the only person on the planet with an 077 umask, but I doubt it. Nov 13 06:20:27 hail zecke Nov 13 06:20:32 03hrw 07org.oe.dev * r803460a5... 10/: Nov 13 06:20:33 collie.conf: added altboot to base system Nov 13 06:20:33 - this add possibility to: Nov 13 06:20:33 - boot from SD, CF, NFS Nov 13 06:20:33 - drop into emergency shell Nov 13 06:20:33 All those 'boot on SD' scripts/hacks should die now. Nov 13 06:21:05 koen: patch added to #350 Nov 13 06:22:48 Where do I send new .bb files to? The mailing list? Nov 13 06:22:54 bugtracker Nov 13 06:23:16 bugtracker will send info to mailing list Nov 13 06:24:27 hrw: Ok, thanks. Nov 13 06:30:46 hrw: don't forget to poke mickeyl to make .tar and .ext2 images for collie Nov 13 06:31:34 koen: good point Nov 13 06:31:56 NOTE: package libxml2-2.6.22: completed Nov 13 06:32:03 * koen sighs of relief Nov 13 06:32:38 ~lart libxml2 for having broken regression tests Nov 13 06:32:38 * ibot chops libxml2 in half with a free Solaris 7 CD for having broken regression tests Nov 13 06:42:50 03koen 07org.oe.dev * r2fa74870... 10/: libxml2: add 2.6.22 with a patch to remove the regression tester that fails to compile (the irony!) Nov 13 06:48:39 koen, hrw: Why not set collie.conf to generate all those image types? Nov 13 06:49:31 RP: hmm.. Nov 13 06:49:49 we already have it in few machines Nov 13 06:51:09 having an altboot req in a machine isn't nice either Nov 13 06:51:19 it would be better to do it with overrides in oz.conf Nov 13 06:52:16 koen: or with override in machine itself Nov 13 06:54:30 anyway user can uninstall it if not like it Nov 13 06:54:58 if not used it only take 3s during start (and that dealy is configurable in altboot.cfg) Nov 13 07:39:14 cu Nov 13 08:24:49 anyone familiar with 8bit BDF fonts? Nov 13 09:23:46 Hello all. Nov 13 09:24:27 I struggled again building a x86 bootstrap image. Nov 13 09:25:01 portmap can not be build, as obviously the name is not matching any standard. Nov 13 09:25:05 error is ... Nov 13 09:25:17 ERROR: log data follows (/data/scdev2/build-x86-rpm/tmp/work/portmap-5-9-r5/temp/log.BUILDSPEC.4109) Nov 13 09:25:17 | error: line 2: Illegal char '-' in version: Version : 5-9 Nov 13 09:25:17 | Building target platforms: i586-linux Nov 13 09:25:25 Any idea how to fix that ? Nov 13 09:26:16 change pv to 5.9 Nov 13 09:26:31 koen: In the bb file ? Nov 13 09:26:35 yes Nov 13 09:26:53 Is it necessary to rename the file ? Nov 13 09:27:21 it's against our new guidelines for pv Nov 13 09:27:34 renaming isn't necessary Nov 13 09:27:51 koen: So you mean PV = "5.9" ? Nov 13 09:28:12 yes, and change src_uri too Nov 13 09:29:39 koen: Will do. Nov 13 09:29:51 Next error : psmisc does not build. Nov 13 09:30:56 configure complains tgetent not found in tinfo, ncurses, termcap. Which package does provide tgetent ? Nov 13 09:32:59 terminfo? Nov 13 09:33:56 * koen|770 usually uses packages.debian.org to search Nov 13 09:34:55 koen: No terminfo package in org.openembedded.dev/packages Nov 13 09:39:08 koen: May I ask you a question about bash scripting ? Nov 13 09:40:21 sure Nov 13 09:40:43 Ihave a problem on one PC to get '' Nov 13 09:40:52 mkdir -p ${RPMBUILDPATH}/{SPECS,RPMS/{i386,i586,i686,noarch,ppc,mips,mipsel,arm},SRPMS,SOURCES,BUILD} Nov 13 09:40:59 extracted. Nov 13 09:41:23 On my "normal" host it works, but not on ma laptop. Any ideas ? Nov 13 09:42:33 koen|770: I thought our new guidelines were going to allow - :-/ Nov 13 09:42:54 uv1: no idea on that Nov 13 09:43:09 RP: only for kernels, right? Nov 13 09:44:21 koen|770: If we can handle it for kernels, we can probably handle it for most things but I can see what you're saying... Nov 13 09:44:44 koen|770: We're going to drop it from the preferred CVS PV format? Nov 13 09:45:24 I guess so Nov 13 09:45:41 judging from the lack of rrsponses Nov 13 09:46:08 I don't have a problem with the dashes. We just need a policy saying one way or the other... Nov 13 09:46:56 koen, Is there a way to select the appropriate compiler flags based on gcc version? Nov 13 09:47:06 this tune thing seems to be a problem all around Nov 13 09:47:27 looking at the rpm error above it seems oe can't handle it Nov 13 09:48:13 Crofton__: probably, but I don't know how Nov 13 09:48:23 koen|770: Was that an OE or an rpm error? Nov 13 09:48:28 well hopefully someone explains it to us :) Nov 13 09:48:35 and will it also break on kernels? Nov 13 09:48:55 Which rpm error ? Nov 13 09:49:20 the portmap one Nov 13 09:49:37 uv1: " | error: line 2: Illegal char '-' in version: Version : 5-9" Nov 13 09:49:50 Is that rpm specific ? Nov 13 09:50:01 Maybe ;-) Nov 13 09:50:08 uv1: That's what I'm asking... ;-) Nov 13 09:52:29 RP: uv1 tries that again (rebuilding original portmap package) ... Nov 13 09:53:54 RP: Yep rpm specific. Nov 13 09:54:49 uv1: I'd be tempted to add something to translate the "-" character in PV into some other character... Nov 13 09:54:59 (for rpms) Nov 13 09:56:37 I could do more changes on package_rpm.class. Will try, if someone tells me how s/-/\./ works in python ;-) Nov 13 09:59:11 uv1: usually, I'd just call on sed... Nov 13 09:59:38 split and rejoin? Nov 13 10:01:51 re.sub(regex, replacement, subject) ;-) Nov 13 10:02:15 But then als movefile (installation) have to be fixed. Trying ... Nov 13 10:05:36 Seems to work ... Nov 13 10:05:44 Third patch on package_rpm. Nov 13 10:06:06 Should I put taht into 431 bug ? Nov 13 10:09:45 uv1: you can also use string.replace(old,new), which is probably cheaper than re for simple replacements. Nov 13 10:11:14 hi gentlemen Nov 13 10:11:39 pb_: So ${PV}.replace("-", "." ) ? Nov 13 10:11:49 are you talking about using rpm format under OE based distribs ? Nov 13 10:13:22 uv1: pretty much, yeah Nov 13 10:13:37 alan|xchat: Yes. Nov 13 10:13:48 pb_: Ok, will try. Nov 13 10:15:03 uv1: is that just for curiosity or do you really plan using rpm packages ? (Oo) Nov 13 10:16:29 heh Nov 13 10:16:37 I guess someone must use rpm. Nov 13 10:18:33 pb_: My company wants, as we have several embedded linux test systems already managed by an rpm repository. Nov 13 10:18:49 They dont want to through that away ... Nov 13 10:19:55 seems fair Nov 13 10:20:51 pb_: BTW, replace is mentioned "deprecated" in python lib doc. Nov 13 10:24:08 uv1: afaik, only the "string.replace()" library function is deprecated, not the object method. Nov 13 10:25:10 Ok replace works as well. Should I provide the patch to bug 431 ? Nov 13 10:25:30 i.e. string.replace(${PV}, "-", ".") would be deprecated, but ${PV}.replace("-", ".") would not. Nov 13 10:25:38 yeah, might as well Nov 13 10:27:09 alan|xchat: We really plan to use, though we might have to patch more things ;-) Nov 13 10:38:49 int Nov 13 10:43:24 rpm has a problem with ints? that does sound unfortunate. Nov 13 10:43:55 pb_: I have mistaken this channel with maple ;) Nov 13 10:44:14 understandable Nov 13 10:44:47 mathematics is certainly a dominant theme in this channel Nov 13 10:45:37 this assignment would make me nuts... Nov 13 10:46:28 oh dear Nov 13 10:46:43 if it wasn't too late for that... Nov 13 10:47:14 heh Nov 13 10:49:30 crumbs, it's cold outside today Nov 13 10:49:38 must be winter or something Nov 13 10:49:42 pb_: and does it rain? Nov 13 10:49:55 actually no Nov 13 10:50:17 did it at least rain? Nov 13 10:50:34 hey zecke Nov 13 10:51:18 not since I've been back in the country, which I guess is about eleven hours. Nov 13 10:51:25 so, I guess we are due for rain at any minute Nov 13 10:52:03 hehe Nov 13 10:56:21 pb_: where are you? Nov 13 10:56:38 here it started raining at last two weeks ago, after more than one year... Nov 13 10:57:10 Bernardo: in England Nov 13 10:57:44 where, as zecke will tell you, it rains approximately all the time Nov 13 10:57:48 ok, a few miles straight north from here Nov 13 10:57:56 lol Nov 13 10:58:15 in ireland nobody would believe me on my third visit that I had always had sun... :) Nov 13 10:58:27 guys, i just flashed a homemade gpe image. Current version of Vi doesn't implement file writing (???). Anyone notices that ? should i report this as a bug ? Nov 13 10:58:54 alan|xchat: that does sound like an egregious bug Nov 13 10:59:02 what exactly happens when you try to write a file? Nov 13 11:01:12 ":exit" : ''exit is not implemented Nov 13 11:01:37 nothing writen with :x nor with :wq Nov 13 11:02:24 :quit works, though Nov 13 11:02:51 saving is overrated anyway Nov 13 11:03:15 alan|xchat: is that the busybox vi or vim? Nov 13 11:03:37 busybox vi, i think. Nov 13 11:03:45 how can i check ? Nov 13 11:03:56 "ls -l /bin/vi" Nov 13 11:03:59 alan|xchat: is vi a symlink to busybox? Nov 13 11:04:09 alan|xchat: or as pb_ told you ;) Nov 13 11:04:53 alan|xchat: are you sure the location you are trying to save into is writeable? busybox vi does have a bug where it will just quietly do nothing if the file can't be created, rather than printing any kind of error message. Nov 13 11:04:56 vi -> busybox Nov 13 11:05:45 pb_: i use it as root on /etc/ipkg.conf... this should be writable ;) Nov 13 11:08:29 yeah, one would hope Nov 13 11:13:23 ok. works with vim. Nov 13 11:21:43 Bootstrap-image for x86 builded including rpm repository. Thanks for help to all. See you ... Nov 13 11:31:56 uv1: did you test your sed magic on BSD? Nov 13 11:42:13 what file rules the windowmaker used under GPE ? i would like to use e-wm... Nov 13 11:43:34 it would usually be ~/.xinitrc, but not here... Nov 13 11:45:07 have a look in /etc/Xsession.d/ Nov 13 11:45:25 you want the file with the largest number, probably 99window-manager or something Nov 13 11:47:42 zecke: sed magic, BSD ? Nov 13 11:48:09 uv1: in your rpm patch Nov 13 11:48:40 zecke: You mean the spec file listing ? Nov 13 11:49:46 zecke: Have no BSD to test ;-( Nov 13 11:53:38 I would welcome if more 'native' python is used instead of spawning shells Nov 13 11:54:07 we know python is pretty portable, and I also know people love the GNU extensions (including myself) Nov 13 11:54:36 uv1: though this is a 'minor' issue Nov 13 11:55:01 zecke: Understood, but as I am not a python expert (at least I "learend" that the last weekend), it may be better to use bash scripts for me ;-) Nov 13 11:55:35 Nevertheless, I can try to do in python, but not before next weekend. Nov 13 11:56:58 uv1: that is the point, one of my BSD boxes doesn't even have a bash ;) Nov 13 11:57:38 zecke: I promise for next weekend ... Nov 13 11:57:40 Bye Nov 13 11:58:19 take care Nov 13 11:58:21 uv1: wait Nov 13 11:58:27 waiting ... Nov 13 11:58:39 uv1: did anyone volunteer to push your rpm changes? Nov 13 11:59:10 zecke: No I was just told by koen to put that into bugzilla. Nov 13 11:59:22 Bug is 431 Nov 13 12:00:29 uv1: if it doesn't get too late I will review and push them tonight Nov 13 12:00:39 (depends on how fast I finish my asisgnments though) Nov 13 12:01:17 zecke: Thanks and by. Nov 13 12:01:28 me hands zecke maple-tabcompletion Nov 13 12:02:12 koen|770: well I would prefer a 'solve text assignments' method Nov 13 12:21:14 Ciao all Nov 13 12:23:36 hey pigi Nov 13 12:23:46 hey koen|770 ! Nov 13 12:25:48 hehehehe !!! i have e17 on my c750 !!! Nov 13 12:25:52 * alan|xchat happy Nov 13 12:35:30 can anyone tell me why "iwconfig wlan0 mode master" fails?? I assume this was not intended Nov 13 12:39:20 any number of reasons, I guess. most likely is that your WLAN card, or its driver, doesn't support AP mode. alternatively, it might be that you are typing that command into someplace other than the shell, or you don't have a wifi card at all, or wireless-tools is not installed, or your card is not named wlan0, or... Nov 13 12:40:33 heh Nov 13 12:42:21 pb_: I know I'm using the command correctly, ie, in a shell, I have a wireless card, my wireless card is wlan0. You wouldn't happen to know if Linksys WCF12 supports AP mode? Nov 13 12:42:51 Zero_Chaos: it should do if you are using the hostap driver Nov 13 12:43:58 I think those cards have at least two alternative drivers, though, so there is at least some risk you are not using hostap. Nov 13 12:44:24 pb_: I would imagine it supports it. And I assure you that I'm using hostap :-) Nov 13 12:44:39 Do you have hostapd installed and running? Nov 13 12:44:57 installed, but I don't know if it's running... Nov 13 12:45:26 That might be the problem. I think the driver may refuse to let you select master mode if it isn't running. Nov 13 12:45:49 alternatively, it's possible that your driver was somehow compiled without AP support. It's a build-time option, though it ought to be enabled. Nov 13 12:46:26 What's the actual error message you get from that iwconfig command? Nov 13 12:50:15 pb_: well it stopped erroring, but void11 still doesn't seem to be working :-( Nov 13 12:54:56 oh dear Nov 13 12:55:00 is there anything in the logs? Nov 13 12:55:04 (either dmesg or logread) Nov 13 12:56:29 pb_: doesn't appear to be anything useful... Nov 13 12:57:59 how odd Nov 13 12:58:14 you might need to talk to mickey|dinner when he finishes his dinner. He is our hostap expert. Nov 13 12:59:09 pb_: thanks Nov 13 13:04:11 pb_: Opening raw packet socket for ifindex 18. ioctl[PRISM2_IOCTL_HOSTAPD] Invalid argument Nov 13 13:05:37 bbl Nov 13 14:11:48 mickey|dinner: how is that dinner? Nov 13 14:13:07 03rwhitby 07org.oe.dev * rc7a65bbb... 10/: nas100d-kernel: Removed noirqdebug from cmdline Nov 13 14:13:11 03rwhitby 07org.oe.dev * r14567c9c... 10/: nas100d-kernel: more updates to i2c and ide and defconfig from azummo Nov 13 14:13:15 03koen 07org.oe.dev * r81893f63... 10/: qof: add qof 0.6.0-pre1 Nov 13 14:17:01 LoL, mickey has been having dinner for more than two days... Nov 13 14:17:22 crazy germans Nov 13 14:17:46 koen|770: anyone else in here a hostap expert? or just him? Nov 13 14:18:10 no idea Nov 13 14:18:17 Zero_Chaos: just him Nov 13 14:19:03 zecke: eh, my luck :-) Nov 13 14:36:45 yeah, mickey is famously gluttonous Nov 13 14:53:59 monotone isn't too bad Nov 13 14:55:11 the rosterbranch should improve a lot Nov 13 14:55:41 what is that? Nov 13 14:56:23 monotone's future Nov 13 14:56:25 I made a seperate branch for my local work, so I can do stupid stuff and still take updates from main dev branch Nov 13 14:56:49 u using the keyboard or handwriting recognition? Nov 13 14:57:01 vkb Nov 13 15:01:09 3333;aaaaaaaaa Nov 13 15:06:01 ggilbert: yeah, I hate 3333 as well Nov 13 15:19:06 nite all Nov 13 15:54:11 03koen 07org.oe.dev * r29ae825e... 10/: qof: update to 0.6.0 release Nov 13 17:41:12 can someone tell me if the feeds are broken by default (i.e. right after install) under 3.5.3? I seem to be able to update the package list but when I try to install anything either from the command line or from the Opie package manager tool, it hangs. after about 5 minutes it'll say "Terminated." on the command line, but the package management tool will just sit there forever. Nov 13 17:41:48 (bear in mind this is a fresh install of 3.5.3 on an SL-5600 Nov 13 17:41:49 ) Nov 13 17:42:59 er, 55 Nov 13 18:54:30 well, I figured out what the problem was. there is apparently no room in root Nov 13 18:54:33 1.4KB Nov 13 18:54:44 from the get-go Nov 13 19:15:39 StarKruzr: Hehe... very annoying that I imagine ;) Nov 13 19:17:13 dude Nov 13 19:17:16 check this out Nov 13 19:17:30 http://www.oesf.org/forums/index.php?showtopic=16091 Nov 13 19:17:56 * StarKruzr cannot imagine why root would be filled to 91% capacity straight from flashing Nov 13 19:18:18 I think on OZ 3.2 it was like 59%? something like that, maybe less Nov 13 19:24:16 No idea... I'm not working with any of the standard builds/devices. Nov 13 19:25:21 v8jlene: someone theorized that flashing does not actually remove data from root. do you know of a way to wipe out root explicitly? Nov 13 19:25:25 cd /mnt Nov 13 19:25:25 ls Nov 13 19:25:27 oops Nov 13 19:25:28 heh Nov 13 19:30:02 root is jffs2, right? Nov 13 19:30:13 is there a command to initiate garbage collection manually, maybe? Nov 13 19:30:27 (I tried sync, but it laughed at me) Nov 13 19:31:35 StarKruzr: As I said I'm using a different device. Mine has a translation layer over the flash to make it look like a normal disk (so I just use fdisk to setup partitions and then run ext3 on it). So I can't really offer much help.. Nov 13 19:32:43 That would be a beautiful, beautiful thing. Nov 13 19:33:10 sadly I cannot afford a new Z Nov 13 19:33:30 StarKruzr: And there's usually not many other's around here at this time of day (or night I guess!). It starts to get busy in about 8hrs time... Nov 13 19:34:49 * StarKruzr has tried coming in here many times with very little success Nov 13 19:35:08 StarKruzr: The downside is that it does more writing to the flash and will wear it out quicker. I think I would have preffered jffs2 over mtd! Nov 13 19:35:54 v8jlene: I heard somewhere that it would take ~7 years even with fairly active writing to wear it out Nov 13 19:36:57 (with a memory corruption one can do it in 10 minuntes) :S Nov 13 19:37:04 heh Nov 13 19:37:24 disclaimer: that was the statistic for "normal usage" Nov 13 19:40:00 Could be, can't say I've really looked ;) I worry about ext3 writing update timestamps and stuff all the time! And I'm building a device to go into remote locations for transmisison equipment monitoring - 7yrs sounds a bit short for such a project ;) Nov 13 19:40:14 Really looked at the exact figures that is... Nov 13 20:08:47 \me Wonders why there are "net", "network" and "console/network" sections... Nov 13 22:11:11 hey kergoth. you available at the moment? Nov 13 22:12:21 somewhat. if you have an oz or oe question, dont bother asking me. i havent touched oz in over a year, and oe in a good 6 months Nov 13 22:26:14 okay. Nov 14 00:09:03 03jbowler 07org.oe.dev * r5696f581... 10/: ixp400-eth: make building for 2.6.14-mm in 1.4 Nov 14 00:09:08 03jbowler 07org.oe.dev * rd91acc33... 10/: Nov 14 00:09:08 nslu2-kernel: upgrade to 2.6.14.2 Nov 14 00:09:08 nslu2-kernel: add (only semi working) 2.6.15-rc1 Nov 14 00:09:11 03jbowler 07org.oe.dev * r826007c3... 10/: slugos.conf: add masqmail to build, version 0.2.20 Nov 14 00:30:29 morning Nov 14 00:33:01 morning alan Nov 14 01:26:28 morning all! Nov 14 01:28:36 morning koobla Nov 14 01:37:01 morning Nov 14 01:44:20 morning xora Nov 14 01:45:00 yop XorA Nov 14 01:45:17 XorA: e17 is on my 750 ^^ Nov 14 01:49:46 anyone had a look at the GettingStarted wiki page? Nov 14 01:49:54 seems it got spammed Nov 14 01:50:05 is oembedded.org back online ? Nov 14 01:50:19 no / looked at oe.handhelds.org Nov 14 01:51:00 koobla : it cannot to connect. Nov 14 01:51:25 http://oe.handhelds.org/cgi-bin/moin.cgi/GettingStarted Nov 14 01:51:59 looks like it has been hacked... Nov 14 01:52:11 looking at http://oe.handhelds.org/cgi-bin/moin.cgi/RecentChanges Nov 14 01:52:27 should get all changes Nov 14 01:53:00 pffff.... Nov 14 01:53:21 google has some pages chached? Nov 14 01:53:46 theres a way to fix it, but I cant remeber how :-( Nov 14 01:53:48 i think so Nov 14 01:54:00 you can revert the pages, but I dont know how to drive moinmoin Nov 14 01:55:03 in the recent changes page, theres an info button Nov 14 01:57:17 seems there have been several rollbacks recently :( Nov 14 01:57:34 morning all Nov 14 01:57:46 hI! Nov 14 01:58:07 hey RP Nov 14 01:58:12 morning RP. Nov 14 01:58:42 RP:I dont know whether the SL-C3000 support hard disk device cache. Nov 14 01:59:23 ( spitz | oz-3.5.4 | kernel-2.6.14 | SL-C3000 ) Nov 14 01:59:47 RP:any ideas.? Nov 14 02:03:44 hi Nov 14 02:03:59 hey hrw|work Nov 14 02:05:55 hi hrw|work Nov 14 02:06:09 power outage in building == no logs from weekend Nov 14 02:06:10 ;) Nov 14 02:06:22 freethinker: The Linux ide code should just work with the cache Nov 14 02:06:52 hrw|work: You missed me getting annoyed by rmk (again) ;-) Nov 14 02:07:03 hrw|work: but its worked out in the end... Nov 14 02:07:29 good Nov 14 02:10:20 getting to work after 3 day weekend is ... argh Nov 14 02:13:09 RP:But my hard disk only has 8M per second. Nov 14 02:13:33 It's on an ISA bus effectively. Nov 14 02:13:59 (At least, that's what I understand CF to be similar to - original PCMCIA, right?) Nov 14 02:14:24 freethinker: 8M/s in what kind of device? Nov 14 02:14:45 freethinker: Did you get a much faster speed under 2.4? Nov 14 02:14:45 hrw|work: spitz | oz-3.5.4 | kernel-2.6.14 | SL-C3000 Nov 14 02:14:49 (From before you joined) Nov 14 02:16:40 8M is lot. I have much less from cf card in c760 Nov 14 02:17:56 hrw: in my CF card Nov 14 02:18:44 I remember 0.6M -> 1.3M -> 2.5M steps when using Amiga with ATA disc.. Nov 14 02:18:54 RP: it has almost same speed under 2.4kernel. Nov 14 02:21:32 freethinker: I think its likely to be running as fast as its ever going to then... Nov 14 02:21:49 I'm sure it will be using the cache unless you can prove otherwise? Nov 14 02:22:21 morning koen Nov 14 02:22:45 hey RP Nov 14 02:22:48 RP:thanks :) Nov 14 02:23:42 hrw: thank you a lot Nov 14 02:25:11 hi koen Nov 14 02:29:43 collie discussions are in KernelTraffic Nov 14 02:30:27 hey Nov 14 02:30:35 hi zecke Nov 14 02:30:39 *idiots* ... Nov 14 02:30:45 hey hrw|work & zecke Nov 14 02:30:46 searching in a Makefile for .exe... Nov 14 02:31:14 *idiots*... they should not take a system security class if they do not even know some bits of Unix... Nov 14 02:34:14 zecke: students? Nov 14 02:34:31 yes ;) Nov 14 02:34:36 but moron students Nov 14 02:34:48 "now we need to find out how to create a file..." Nov 14 02:35:13 actually I shut up as I was too dumb to finish my math assignment... Nov 14 02:45:23 morning Nov 14 02:51:17 morning greentux Nov 14 02:52:26 hrw|work: which machine i have to use in BB for borzoi? can u sen dme your local.conf? Nov 14 02:52:33 morning all Nov 14 02:52:48 morning Nov 14 02:58:21 greentux: borzoi Nov 14 02:59:33 hrw|work: ok and distr = openzaurus-3.5.4 (for latest) Nov 14 02:59:47 iirc yes Nov 14 02:59:54 hi Dirk **** ENDING LOGGING AT Mon Nov 14 02:59:57 2005