2008-09-06 02:29 bleh 2008-09-06 02:30 the mailserver on sf.net rejects senders who don't have a postmaster@ address 2008-09-06 02:33 sending/subscribing from my other address 2008-09-06 03:35 -!- cdk(~chinmay@121.246.36.66) has joined #tux3 2008-09-06 03:43 -!- cdk(~chinmay@121.246.36.66) has left #tux3 2008-09-06 05:01 -!- stargazr5(~gauravstt@59.95.4.235) has joined #tux3 2008-09-06 07:39 -!- tim_dimm(~timothyhu@cpe-76-90-98-247.socal.res.rr.com) has joined #tux3 2008-09-06 09:04 -!- tim_dimm(~timothyhu@cpe-76-90-98-247.socal.res.rr.com) has joined #tux3 2008-09-06 11:34 -!- stargazr5(~gauravstt@59.95.21.222) has joined #tux3 2008-09-06 14:34 -!- MaZe(~MaZe@c-24-6-86-168.hsd1.ca.comcast.net) has joined #tux3 2008-09-06 15:12 nearly sk8 oclock 2008-09-06 15:12 test earlier these days 2008-09-06 15:12 gets earlier 2008-09-06 15:13 the french girls go back to their hotels earlier too 2008-09-06 15:13 temperature drops slightly below bikini degrees, centigrade 2008-09-06 15:25 we have a response 2008-09-06 15:26 -!- MaZe(~MaZe@c-24-6-86-168.hsd1.ca.comcast.net) has joined #tux3 2008-09-06 15:30 For example autoconf that tests for the number of parameters. 2008-09-06 15:30 See http://tinyurl.com/6k6hnm on how this is done. 2008-09-06 15:30 Then again, FUSE 2.5.3 is almost 2 1/2 years old.. 2008-09-06 15:30 """ 2008-09-06 15:40 response? 2008-09-06 15:40 oh 2008-09-06 15:40 the fuse list :-) 2008-09-06 15:40 autoconf is banned from tux3 2008-09-06 15:41 solutions that avoid autoconf are welcome 2008-09-06 15:41 for example, whatever test autoconf uses 2008-09-06 15:41 we cut & paste 2008-09-06 15:41 stripping out the fluff 2008-09-06 15:42 2.5 years is short on the unix timescale 2008-09-06 15:42 got to think longer term than that 2008-09-06 15:43 nice piece of autoconf wanking 2008-09-06 15:43 has nothing to do with fuse 2008-09-06 15:43 right, I don't like autoconf either 2008-09-06 15:43 hopefully a more clueful response is in the pipe :-) 2008-09-06 15:44 anyway, we have solved it, I hacked my fuse.h 2008-09-06 15:44 alright 2008-09-06 15:44 it's good you are on the fuse list 2008-09-06 15:45 I suspect there are many other things to complain about, most of them more important 2008-09-06 15:47 note that fuse has a parallel mode 2008-09-06 15:47 I actually tried to use fi->fh earlier 2008-09-06 15:47 so once we have it basically working 2008-09-06 15:47 ah 2008-09-06 15:47 but it was failing on the second read 2008-09-06 15:47 maybe I was shoving the wrong pointer in it 2008-09-06 15:47 need to get the fuse source and compile 2008-09-06 15:47 and debug the two together 2008-09-06 15:48 in uml even 2008-09-06 15:48 ok, I can cook up a recipe for that 2008-09-06 15:48 just turning on fuse event tracing in the kernel would be a huge help 2008-09-06 15:48 I wonder if there is an easy way to do that 2008-09-06 15:49 let's see what's in the fuse kernel code ;-) 2008-09-06 15:49 http://lxr.linux.no/linux+v2.6.26.3/fs/fuse/ <- fuse kernel code 2008-09-06 15:50 http://lxr.linux.no/linux+v2.6.26.3/include/linux/fuse.h <- and here 2008-09-06 15:51 http://lxr.linux.no/linux+v2.6.26.3/+ident=14862186 <- FOPEN_KEEP_CACHE for example 2008-09-06 15:52 http://lxr.linux.no/linux+v2.6.26.3/fs/fuse/dir.c#L755 <- /* Directories have separate file-handle space */ 2008-09-06 15:53 FUSE_GETATTR_FH 2008-09-06 15:53 konrad, better than a pointer is the inum, for now 2008-09-06 15:54 call open_inode 2008-09-06 15:54 we don't really care about performance at this point 2008-09-06 15:54 just working right 2008-09-06 15:54 k 2008-09-06 15:55 the libfuse stuff is important too 2008-09-06 15:55 more important that the kernel I think 2008-09-06 15:55 so we should compile it with -g 2008-09-06 15:55 to set breaks and see how it screws up ;-) 2008-09-06 15:55 let's see, how do you get a debug build in debian 2008-09-06 15:56 weakness of debian 2008-09-06 15:56 this is where gentoo is good 2008-09-06 15:56 but I can just apt-get remove libfuse2 2008-09-06 15:56 and build from tarball like a proper hacker 2008-09-06 16:01 sudo apt-get remove --purge fuse-utils libfuse-dev libfuse2 2008-09-06 16:05 all the pain of autoconf is coming back to me now 2008-09-06 16:05 including library path skew 2008-09-06 16:13 ok, installed, working 2008-09-06 16:13 needlessly painful 2008-09-06 16:14 ld.so.config... not documented in man ld :-P 2008-09-06 16:14 ld.so.conf I meant 2008-09-06 16:14 see? pain 2008-09-06 16:15 aha! it was a fuse bug 2008-09-06 16:15 now does not claim to be busy on umount 2008-09-06 16:16 heh 2008-09-06 16:16 and tux3fs does not exit file file not found when you unmount 2008-09-06 16:16 what version of fuse are you using now? 2008-09-06 16:16 um 2008-09-06 16:16 2.7.4 2008-09-06 16:16 is it leet? 2008-09-06 16:17 heh 2008-09-06 16:17 more recent anyways 2008-09-06 16:17 I decided not to track their unstable 2008-09-06 16:17 any more than they should track ours 2008-09-06 16:17 for their home dirs ;-) 2008-09-06 16:18 kay, got to get serious about skating 2008-09-06 16:18 tux3 fuse is looking good 2008-09-06 16:32 it's about time to create the version table 2008-09-06 16:33 will be inode number 2, maybe 2008-09-06 16:33 or maybe it deserves to be number 0 2008-09-06 16:33 even more important than bitmap 2008-09-06 16:33 which is after all a redundant structure 2008-09-06 16:34 0: version 1: bitmap 2: extentmap 2008-09-06 16:34 maybe 2008-09-06 16:35 the rule is: inums below 0x10 do not have dirents 2008-09-06 16:35 they are special tux3 files never seen by user space 2008-09-06 18:09 ok, sk8 oclock, really 2008-09-06 18:51 hey 2008-09-06 19:54 hi bh 2008-09-06 20:44 -!- ChanServ changed mode/#tux3 -> +o flips 2008-09-06 20:44 -!- flips changed topic to "Tux3 list members just hit 100! ~ http://tux3.org" 2008-09-06 20:44 -!- flips changed topic to "Tux3 list membership just hit 100! ~ http://tux3.org" 2008-09-06 20:44 -!- flips changed mode/#tux3 -> -o flips 2008-09-06 23:17 -!- stargazr5(~gauravstt@59.95.22.62) has joined #tux3