| 00:02.18 | *** join/#edev prpplague (~prpplague@ppp-70-242-114-182.dsl.rcsntx.swbell.net) |
| 01:16.53 | *** join/#edev prpplague (~prpplague@ppp-70-242-114-28.dsl.rcsntx.swbell.net) |
| 01:31.41 | *** join/#edev liuyq (~liuyq0307@221.123.128.226) |
| 01:32.40 | *** join/#edev liuyq (~liuyq0307@221.123.128.226) |
| 01:57.13 | *** join/#edev prpplague (~prpplague@ppp-70-249-146-111.dsl.rcsntx.swbell.net) |
| 03:39.54 | *** join/#edev aholler_ (~aholler@p57B20A29.dip0.t-ipconnect.de) |
| 03:46.13 | *** join/#edev TheSeven (~quassel@rockbox/developer/TheSeven) |
| 04:26.30 | *** join/#edev JoeLlama (~snork@unaffiliated/joellama) |
| 04:53.33 | *** join/#edev MonMotha (~monmotha@2001:470:c5f2:1:224:7eff:fedc:893e) |
| 06:11.36 | *** join/#edev florian (~fuchs@Maemo/community/contributor/florian) |
| 06:45.13 | *** join/#edev _Lucretia_ (~munkee@pdpc/supporter/active/lucretia) |
| 06:46.22 | *** join/#edev aholler (~aholler@p57B20A29.dip0.t-ipconnect.de) |
| 06:46.31 | *** join/#edev TheSeven (~quassel@rockbox/developer/TheSeven) |
| 06:50.11 | *** join/#edev bsdfox (~Bob@c-71-197-83-152.hsd1.ca.comcast.net) |
| 06:50.11 | *** join/#edev bsdfox (~Bob@unaffiliated/bsdfox) |
| 07:01.25 | *** join/#edev liuyq (~liuyq0307@221.123.128.226) |
| 07:02.44 | *** join/#edev liuyq (~liuyq0307@221.123.128.226) |
| 07:47.52 | *** join/#edev bizhanMona (~bizhan@nat/cisco/x-ljsoelazioedzlah) |
| 08:10.48 | *** join/#edev florian (~fuchs@port-217-146-132-69.static.qsc.de) |
| 08:10.49 | *** join/#edev florian (~fuchs@Maemo/community/contributor/florian) |
| 09:43.12 | *** join/#edev gandhijee_ (~akp@50.12.169.99) |
| 09:50.20 | slashbeast | Anyone known aboriginal build process? I wonder how as user it can create a root-filesystem image with root-owned files |
| 09:50.27 | slashbeast | It does not seems to be using fakechroot |
| 10:37.35 | *** join/#edev aholler_ (~aholler@p57B2147E.dip0.t-ipconnect.de) |
| 10:49.33 | *** join/#edev aholler (~aholler@p57B20AD7.dip0.t-ipconnect.de) |
| 10:58.58 | *** join/#edev spacekitteh (~spaec@unaffiliated/traumapony) |
| 11:01.02 | *** join/#edev soafee-chan (~spaec@unaffiliated/traumapony) |
| 11:04.30 | *** join/#edev soafee-chan (~spaec@unaffiliated/traumapony) |
| 13:22.41 | *** join/#edev bizhanMona (~bizhan@nat/cisco/x-jiomnykfydadusut) |
| 13:22.42 | *** join/#edev JoeLlama (~snork@unaffiliated/joellama) |
| 14:09.41 | *** join/#edev prpplague^2 (~danders@192.94.92.11) |
| 14:27.06 | *** mode/#edev [+o prpplague] by ChanServ |
| 14:27.13 | *** join/#edev prp^2 (~prpplague@ppp-70-249-146-111.dsl.rcsntx.swbell.net) |
| 14:37.38 | *** join/#edev keesj (~keesj@oeeo.few.vu.nl) |
| 14:46.03 | *** join/#edev spacekitteh (~spaec@unaffiliated/traumapony) |
| 15:05.31 | *** join/#edev sndcrb (~sndcrb@dsl-hkibrasgw2-fef8de00-126.dhcp.inet.fi) |
| 15:18.23 | *** join/#edev landley (~landley@140.242.26.2) |
| 15:18.25 | *** mode/#edev [+o landley] by ChanServ |
| 15:19.19 | *** join/#edev JoeLlama (~snork@unaffiliated/joellama) |
| 16:10.51 | *** join/#edev BabakKia (~babak@mail.pemicro.com) |
| 16:11.09 | kergoth | slashbeast: you don't need to use fakechroot to create an image with root owned files. you can either use fakeroot (NOT fakechroot) or a manifest based image creation tool (though they dont exist for all filesystems) |
| 16:17.40 | slashbeast | kergoth: correct me if I am worng but as it does use LD_PRELOAD, when I compile and run static binary, the fakeroot will break, right? |
| 16:18.06 | slashbeast | landley: can you tell me how exacly are you creating root-owned rootfs images? I cant find this very code in aboriginal. |
| 16:26.30 | landley | slashbeast: with cpio? |
| 16:26.59 | landley | http://landley.net/hg/aboriginal/file/6840ac8ad2d9/root-image.sh#l36 |
| 16:27.16 | landley | Ok, I'm not using the host cpio I'm using the kernel's gen_init_cpio.c but it's doing the same thing. |
| 16:27.21 | landley | I need to add cpio support to toybox. |
| 16:27.46 | landley | It's basically "an ancient archive format from 1976, like tar but less so". |
| 16:28.02 | SpeedEvil | at |
| 16:28.04 | landley | Which Al Viro chose for initramfs because it's darn simple. |
| 16:28.09 | SpeedEvil | ar |
| 16:28.23 | SpeedEvil | sigh. silly autocorrect. |
| 16:29.24 | landley | I put a cpio invocation that creates an initramfs in http://kernel.org/doc/Documentation/filesystems/ramfs-rootfs-initramfs.txt |
| 16:29.59 | landley | cd dir && find . | cpio -o -H newc | gzip > filename.cpio.gz |
| 16:31.13 | landley | The problem with that is I can't specify a file of ownership/permission stuff like I can with gen_init_cpio.c. Meaning I can't tell it "and how hallucinate up a /dev node with these properties", it has to actually _exist_ in the filesystem for cpio to put it in the archive. |
| 16:31.19 | landley | s/how/now/ |
| 16:31.30 | landley | purl: you can stop now. |
| 16:31.51 | landley | slashbeast: did that help? |
| 16:33.00 | slashbeast | landley: yes thank you |
| 16:34.06 | landley | Creating target filesystem images as a non-root user always involves feeding in overrides to tell the archiver "put this information in the target file that isn't there in the host directory you're reading from". |
| 16:34.08 | slashbeast | landley: I wonder, you seems to love hg, do you really not use branches there? |
| 16:34.17 | landley | I had to ask Philip Lougher to add this to makesquashfs. :) |
| 16:34.18 | slashbeast | I found branching in mercurial to be extreme unfriendly |
| 16:34.20 | landley | genext2fs already had it. |
| 16:34.38 | landley | slashbeast: I just clone repositories. |
| 16:35.00 | landley | I've currently got 3 toybox ones, although I'm draining two of them into the third. |
| 16:35.28 | landley | (This is more or less the way I use git too. I"m aware there's git "shelve" and such, but I don't use it.) |
| 16:36.28 | slashbeast | git was the first VCS I really used so I look at everyting from git point of view |
| 16:36.58 | slashbeast | and just can't stand that in hg I need to create local copie (clone) of repo and treat it as a branch |
| 16:43.41 | landley | slashbeast: *shrug* |
| 16:43.54 | slashbeast | and the mdadm thingy, can I help you make it faster? ;p |
| 16:44.37 | landley | My first source control system was some internal IBM 4-letter horror OS/2 was stored in circa 1996. I was forced to use CVS to contribute to busybox, then he switched to svn which at least sucked less than cvs. |
| 16:45.28 | landley | In 2006 I went to OLS and talked to the designers of both Git and Mercurial trying to figure out which to switch BusyBox to. Mercurial had the much better design internally, and much better user interface externally. |
| 16:45.51 | landley | Unfortunately, Git was written by Linus Torvalds and it's what Linux is in, so it attracted 10x the developer base due to the killer app of being the source control system of linux. |
| 16:46.14 | landley | slashbeast: you could debug the whole thing yourself if you wanted to. |
| 16:46.56 | landley | A system image is built with a uClibc .config and a kernel .config. The problem is adding some symbol to one of those configs, |
| 16:47.03 | slashbeast | landley: I have no idea how, I can provide strace for both mdadm, but I have absolutly no knownlage about kernel nor libc |
| 16:47.17 | landley | The kernel's the easier one to fiddle with because you can build that separately, you don't have to rebuild the root filesystem. |
| 16:47.29 | landley | Just replace the bzImage file in the system-image directory with the new kernel you built. |
| 16:47.31 | slashbeast | I bet thats not kernel as I can build under buildroot chroot and it works |
| 16:47.52 | landley | Then it's a uClibc config symbol. |
| 16:48.58 | landley | On... friday? I did a diff between the buildroot uCLibc config you sent and the aboriginal one, and there were something like a dozen different symbols, of which maybe 8 looked interesting. |
| 16:49.12 | slashbeast | so compare buildroot and aboriginal config and start with adding the missing symbols and testing? |
| 16:49.18 | landley | (Did I send you that email or is it a notes.txt file on my netbook? Said netbook isn't with me at work today so I literally can't work on it here if I wanted to.) |
| 16:49.31 | landley | Yup. There's also a version skew issue though. |
| 16:49.54 | slashbeast | I think you didnt sent me it but I could just not remember it |
| 16:49.55 | landley | The aboriginal one is 0.9.33.2 and the buildroot one is 0.9.32.2. Although your bug report predates that change, so that wasn't it. |
| 16:50.03 | slashbeast | first I will try with buildroot's config and see if it works |
| 16:50.26 | landley | Go for it. |
| 16:50.45 | slashbeast | Roger |
| 16:50.49 | landley | I think you can avoid the miniconfig creation step, there's an override symbol in aboriginal's config... |
| 16:51.08 | slashbeast | I was thinking about putting it as baseconfig anyway |
| 16:51.27 | landley | Or that works too. :) |
| 16:51.30 | slashbeast | I will let you know if I will find anything useful about this issue |
| 16:51.50 | landley | Heh, I need to clean up aboriginal's config. I still gives impactlinux.com as the example preferred mirror. |
| 16:51.54 | aholler | hmm, you forgot good old plain rcs ;) |
| 16:52.05 | landley | (That went away in... 2010?) |
| 16:52.12 | landley | aholler: I never used rcs. |
| 16:52.21 | landley | I used zip files on dos, and then tarballs on linux. |
| 16:52.38 | landley | Every once in a while, tar up your source code with today's date, and keep a directory full of old tarballs. |
| 16:53.16 | landley | Worked surprisingly well. I had a notes.txt and todo.txt in there with notes on what I was doing at the time. I could extract two versions and diff 'em... |
| 16:53.33 | landley | Hideously _inefficient_ of course. The problem with that approach is it doesn't _scale_. But it did work. |
| 16:54.10 | landley | (I'm glossing over "zip files on OS/2 as well". I really treated that as 32-bit dos. info-zip had OS/2 support just fine...) |
| 16:54.21 | landley | (Uphill, both ways, in the snow, of course.) |
| 16:54.55 | aholler | git is just great. |
| 16:56.18 | landley | aholler: git's user interface is horrid. Its internal data representation goes out of its way not to be human-parseable. |
| 16:56.53 | aholler | I'm fine with gitk, don't need anything else |
| 16:57.00 | landley | Git had wrapper scripts (gitorious I think) to make it suck slighytly less, and then it changed its UI completely in the 1.5 release. |
| 16:57.04 | aholler | (except command line git) |
| 16:57.16 | landley | It then has third party wrappers on top (gitk and the web thing), where mercurial just got it right bundled in. |
| 16:57.42 | landley | Random example: in mercurial the user visible repository is http://landley.net/hg/aboriginal and I pull from http://landley.net/hg/aboriginal (I.E. the same URL). |
| 16:58.12 | landley | And I can automatically get tarballs and such of any version there, and there's RSS links where people can follow individual files or git an rss feed of tags or whatever they like. |
| 16:58.15 | kergoth | the only thing i miss about git when using mercurial is the branching, and the handling of remote refs in their own namespace, but that has its own problems, it means that different clones of a repo aren't equal citizens, depending on where tehy cloned from, as they end up with just the branches that were local to the clone source |
| 16:58.29 | kergoth | uses both on a daily basis and likes them both, most days :) |
| 16:58.41 | landley | In git: different URLs for user visible and pull-from, and often no way at all to get the rest of it. |
| 16:59.10 | landley | I use git, but I don't have to like it. :) |
| 16:59.26 | kergoth | https://github.com/kergoth/oe-core -- both user visible and pull-from. |
| 16:59.32 | aholler | I recently had to deal with something where one had to make tons of stuff just to commit something localy and to build a proper patch |
| 16:59.43 | kergoth | but of course thats not true for those hosted elsewhere |
| 16:59.51 | aholler | not sure what it was. I ended up using just patch |
| 16:59.57 | landley | In theory there are "mirror a git repo from mercurial, pull mercurial from a git repo" converters that go both ways. In practice I find system administration fiddly. |
| 17:00.15 | landley | kergoth: cool. It didn't used to be able to do that. |
| 17:01.05 | aholler | ah, it was bazar |
| 17:01.18 | landley | The local crawling horror in dayjobland is called <strike>approxirev</strike> accurev. |
| 17:01.47 | aholler | whatever bazar that does. I didn't found it very usable ;) |
| 17:02.36 | landley | I've also been forced to use Microsoft SourceSafe (an oxymoron, it ate the repo I was working on at the server, that was fun), and at Dell they had "Rational" from IBM. |
| 17:03.03 | landley | The developers at Dell were convinced the IBM salesbeing had bought the Dell VP who made that purchasing decision a _lot_ of dances at the strip club down the street. |
| 17:03.31 | aholler | sourcesafe eats source |
| 17:03.36 | aholler | at least it did |
| 17:03.39 | landley | (Well, ok, they _knew_ that, they were convinced that was why he'd bought such a crappy software suite and imposed it upon the developers.) |
| 17:03.57 | aholler | and that ibm-thingy is a pain to use |
| 17:04.13 | aholler | I forgot how it is called ;) |
| 17:04.24 | landley | Huh, when was I using that... first time was Trilogy in 1999. And then I think Ion Geophysical was using it a few years back. |
| 17:04.43 | landley | I keep confusing "Rational" with "Rational Rose". It's some company IBM acquired. |
| 17:05.14 | aholler | clearcase it is |
| 17:05.16 | landley | Basically http://www.pbs.org/cringely/pulpit/1999/pulpit_19990826_000399.html |
| 17:05.46 | landley | (You know how Microsoft was terrified of "Growing up to be daddy" and did anyway? Yeah, IBM's still got it. It being a bad thing.) |
| 17:06.00 | landley | Or at least had it about a decade ago. |
| 17:07.50 | aholler | anyway, git is fine. and it supports submodules. does mercurial know such too? |
| 17:08.20 | *** join/#edev Crofton (~balister@205-168-82-250.dia.static.qwest.net) |
| 17:08.31 | landley | aholler: submodule support was so nice that Android invented Repo. |
| 17:08.43 | landley | The source reposession system. Sneaks in and drives it away in the middle of the night... |
| 17:10.08 | aholler | invented repo? |
| 17:11.34 | aholler | I haven't played much with repo, but it just looked like something which manages a bunch of git-repos. |
| 17:12.01 | aholler | of course, submodules aren't perfect and they had some problems some years ago, but they work |
| 17:12.19 | aholler | at least to some extend |
| 17:13.06 | aholler | I think they started with repo before submodules worked |
| 17:17.12 | aholler | and doesn't google use internally still something else? I think is something which starts with p |
| 17:17.17 | aholler | +use |
| 17:20.28 | aholler | I've recently worked for a company where people still think svn is something usable. |
| 17:24.37 | landley | most proprietary companies do. They _like_ having the source centrally managed and individual developers not having complete copies of the history. |
| 17:24.51 | landley | Git doesn't let you get access to just _part_ of a repository. |
| 17:25.55 | aholler | thats stupid, I can invoke git on the svn-repo to retrieve the whole history |
| 17:26.31 | aholler | and nothing stops you to create a git-repo without history |
| 17:27.59 | aholler | so either you offer a history or you don't |
| 17:28.13 | aholler | no reason to use svn |
| 17:29.03 | landley | aholler: subversion is popular with proprietary development because they can unplug a single server and take the source code offline. |
| 17:29.28 | landley | People who believe their source is their "crown jewels" and that letting it out of the company is bad like keeping developers on a leash, which subversion does by design. |
| 17:29.32 | aholler | if nobody used git before to retrieve the history ;) |
| 17:29.45 | _av500_ | or svn |
| 17:29.57 | landley | aholler: you don't need git to do that, I had a script that turned a subversion repo into a series of patches back in the busybox days. |
| 17:30.10 | _av500_ | for i in .... |
| 17:30.17 | aholler | you can use svn to create a backup of svn-repos too, sure |
| 17:30.22 | *** join/#edev bsdfox (~Bob@c-71-197-83-152.hsd1.ca.comcast.net) |
| 17:30.23 | *** join/#edev bsdfox (~Bob@unaffiliated/bsdfox) |
| 17:31.02 | _av500_ | if you want to prevent people from having source code you have to make them use paint on a server to edit it |
| 17:31.07 | _av500_ | and prevent screenshots |
| 17:31.10 | landley | But "a developer can explicitly violate their contract" and "the normal workflow creates a zillion copies of the entire history which then get discarded when we drop a flash key in the parking lot" are two different things. |
| 17:31.15 | aholler | _av500_: exactly |
| 17:31.45 | _av500_ | landley: svn would lose code too, just not the whole history |
| 17:31.59 | slashbeast | landley: correct, mdadm.static bulit in aboriginal with new uclibc config does work. |
| 17:32.16 | slashbeast | landley: do you remember what was the 8 symbols that could affect it so I can start binary search? |
| 17:33.02 | _av500_ | aholler: and i had a CEO ask me how to prevent contract programmers from having our source code..... |
| 17:33.23 | aholler | _av500_: I know such discussion too :/ |
| 17:34.02 | aholler | give them a box without outside network and without usb and such. |
| 17:34.18 | aholler | everything else doesn't work |
| 17:35.31 | landley | slashbeast: more/test.sh i686 getconfig uClibc > temp.txt && cd build/packages/uClibc && make allnoconfig KCONFIG_ALLCONFIG=../../../temp.txt |
| 17:35.55 | landley | That should expand the miniconfig for i686 into the uClibc .config in the current directory when you're done. |
| 17:36.21 | Spida | aholler: and shoot them afterwards to erase their memory? |
| 17:36.27 | landley | (I should probably make an expander script. And the above is disgusting because it pollutes the package cache, you'd want to rm -rf build/packages/uClibc after doing that.) |
| 17:36.54 | Spida | aholler: either they trust their dev to do professional work (which includes obeyeing NDAs) or they don't. |
| 17:37.04 | Spida | aholler: imo there is not much in between. |
| 17:37.07 | aholler | Spida: maybe, I think some ceo's would like that |
| 17:37.51 | landley | slashbeast: actually "more/test.sh i686 "setupfor uClibc && getconfig uClibc > mini.config && make allnoconfig KCONFIG_ALLCONFIG=mini.config && cat .config" would be the clean-sh way of doing it. |
| 17:37.53 | aholler | e.g. those which are patenting rectangle |
| 17:37.55 | aholler | s |
| 17:38.23 | landley | Anyway, then what I did was make a miniconfig out of your config and out of the baseconfig, and diffed the miniconfigs. The symbols pretty much jump out at you when you do that. |
| 17:40.46 | *** join/#edev NightMonkey (~NightrMon@pdpc/supporter/professional/nightmonkey) |
| 17:41.24 | landley | slashbeast: I note that if you completely ignored my build system, /usr/src in the system images has the linux, uClibc, and busybox configs used to build it. You can just grab that (or grab the root-filesystem tarball and grab it from there.) |
| 17:41.35 | landley | All this messing around with miniconfigs is my idiosyncratic workflow. :) |
| 17:45.14 | slashbeast | landley: I do use your buiild system, and to see whatever is enabled in 1st config and not in 2nd I have a simple awk script |
| 17:45.28 | slashbeast | landley: however you said something about 8 possible symbols that could trigger it |
| 17:47.16 | landley | I made a list on friday and then was busy with other things all weekend. |
| 17:48.00 | landley | That's not an exact count, that's a "there were a dozen or so changes and I could eliminate 1/4 or so of them just by looking. The rest were an "it could be anything" type thing and I just needed to try them. |
| 17:48.17 | landley | It's probably just one symbol, but it might be more than one. |
| 17:49.19 | landley | Where I actually left of was "I need to set back up a kernel with RAID support switched on". My test environment for that got blanked when I rebuilt everything fresh for the 1.20 release. |
| 17:49.23 | landley | Um, not 1.20. |
| 17:49.32 | landley | 1.2.0 |
| 18:13.20 | aholler | btw. man git-filter-branch has some informations how to reduce the history |
| 18:16.11 | landley | aholler: "all these stupid people who still use windows, don't they know about grep? All these stupid people who still write code in Cobol, here's a list of marvelous things in the Java standard library..." |
| 18:16.18 | landley | It is not a useful discussion. |
| 18:17.17 | aholler | you"ve comed up with the arguments for svn |
| 18:26.55 | aholler | s/comed/came/, sorry, but I"ve heard them too often |
| 18:26.55 | landley | aholler: I don't consider using svn a good idea. I understand why they're doing it. I understand that the things you're bringin up won |
| 18:27.00 | landley | won't persuade them to stop. |
| 18:28.20 | *** join/#edev ccole (~cole@coledd.com) |
| 18:32.36 | aholler | I think the problem is some people just don"t like changes, so they try to find arguments. And the problem is some managers just don"t know to whome they should believe, the one who tells them svn is the right tool or the one who tells them git or mercurial does everything better |
| 18:40.12 | *** join/#edev uartie (~uartie@134.134.139.76) |
| 18:40.24 | *** part/#edev uartie (~uartie@134.134.139.76) |
| 18:44.36 | *** join/#edev CosmicPenguin (~nobody@soa.codeaurora.org) |
| 18:44.36 | *** mode/#edev [+o CosmicPenguin] by ChanServ |
| 18:46.39 | _av500_ | even if you are forced to use svn, people can just use git-svn for their stuff |
| 18:46.57 | _av500_ | I know somebody who uses git on top of company clearcase |
| 18:47.52 | *** join/#edev flo_lap (~fuchs@Maemo/community/contributor/florian) |
| 18:53.58 | *** join/#edev roxfan (dunno@91.180.163.170) |
| 19:08.32 | *** join/#edev CosmicPenguin (~nobody@soa.codeaurora.org) |
| 19:08.32 | *** mode/#edev [+o CosmicPenguin] by ChanServ |
| 19:11.41 | *** join/#edev CosmicPenguin (~nobody@soa.codeaurora.org) |
| 19:11.41 | *** mode/#edev [+o CosmicPenguin] by ChanServ |
| 19:22.11 | *** join/#edev bizhanMona (~bizhan@nat/cisco/x-psrgqtbtdwekpewk) |
| 19:30.12 | *** join/#edev GPSFan (~kenm@64.92.145.112) |
| 19:30.13 | *** mode/#edev [+o GPSFan] by ChanServ |
| 19:31.06 | *** join/#edev CosmicPenguin (~nobody@soa.codeaurora.org) |
| 19:31.06 | *** mode/#edev [+o CosmicPenguin] by ChanServ |
| 19:41.02 | slashbeast | wee just 24 symbols between aboriginal and buildroot config |
| 19:41.08 | slashbeast | let the binary search begins! |
| 19:46.49 | *** join/#edev grummund (~user@unaffiliated/grummund) |
| 20:10.34 | *** join/#edev CosmicPenguin (~nobody@soa.codeaurora.org) |
| 20:10.35 | *** mode/#edev [+o CosmicPenguin] by ChanServ |
| 20:29.16 | *** join/#edev CosmicPenguin (~nobody@soa.codeaurora.org) |
| 20:29.17 | *** mode/#edev [+o CosmicPenguin] by ChanServ |
| 21:09.34 | *** part/#edev BabakKia (~babak@mail.pemicro.com) |
| 21:26.37 | *** join/#edev bsdfox (~Bob@c-71-197-83-152.hsd1.ca.comcast.net) |
| 21:26.37 | *** join/#edev bsdfox (~Bob@unaffiliated/bsdfox) |
| 21:33.37 | *** join/#edev NightMonkey (~NightrMon@pdpc/supporter/professional/nightmonkey) |
| 21:57.21 | *** join/#edev prp^2 (~prpplague@ppp-70-249-147-127.dsl.rcsntx.swbell.net) |
| 22:00.40 | *** join/#edev NightMonkey (~NightrMon@pdpc/supporter/professional/nightmonkey) |
| 22:39.19 | *** join/#edev mrcan (~mrcan@unaffiliated/mrcan) |
| 22:43.12 | *** join/#edev aholler (~aholler@p57B20AD7.dip0.t-ipconnect.de) |