| 01:10.45 | calculus | ooo, glad I didn't try plugging things in yet... beaglebone is safe for 1.8v for adc (although I can't find this in the trm) and gpio is 3.3v safe |
| 03:27.21 | *** join/#edev dijenerate (~dijenerat@199.47.52.87) |
| 04:09.39 | *** join/#edev TheSeven (~quassel@rockbox/developer/TheSeven) |
| 05:33.48 | *** join/#edev contempt (~contempt@unaffiliated/contempt) |
| 07:41.25 | *** join/#edev pvanhoof (~philip@codeminded.be) |
| 07:41.43 | *** join/#edev pvanhoof_ (~philip@codeminded.be) |
| 08:20.05 | *** join/#edev clio (~andrej@85.159.109.222) |
| 09:15.19 | *** join/#edev grummund (~user@unaffiliated/grummund) |
| 09:39.33 | *** join/#edev florian_kc (~fuchs@Maemo/community/contributor/florian) |
| 10:04.30 | *** join/#edev jackmitchell (~Thunderbi@195.171.99.130) |
| 10:13.11 | *** join/#edev ccole (~cole@coledd.com) |
| 10:14.12 | *** join/#edev bzb (~bzb@69-196-174-80.dsl.teksavvy.com) |
| 10:51.52 | *** join/#edev av500 (~vladimir@b2b-46-252-131-98.unitymedia.biz) |
| 11:33.24 | *** join/#edev g1powermac (~g1powerma@unaffiliated/g1powermac) |
| 11:40.02 | *** join/#edev jackmitchell (~Thunderbi@195.171.99.130) |
| 12:07.50 | *** join/#edev gustavoz (~gustavoz@host42.181-14-143.telecom.net.ar) |
| 12:38.38 | *** join/#edev jackmitchell (~Thunderbi@195.171.99.130) |
| 14:39.22 | *** join/#edev Crofton (~balister@81.240.254.28) |
| 15:49.14 | *** join/#edev dimz (957e6a13@gateway/web/freenode/ip.149.126.106.19) |
| 16:00.11 | *** part/#edev dimz (957e6a13@gateway/web/freenode/ip.149.126.106.19) |
| 16:08.59 | *** join/#edev kergoth (~kergoth@covenant.kergoth.com) |
| 16:12.09 | *** join/#edev av500 (~vladimir@46.252.131.98) |
| 16:15.11 | *** join/#edev prpplague (~danders@192.94.92.14) |
| 16:19.13 | *** join/#edev landley (~landley@136.162.34.1) |
| 16:19.14 | *** mode/#edev [+o landley] by ChanServ |
| 16:19.58 | landley | xchat is buggy. |
| 16:20.15 | landley | Open new sever window, from there /join #channel |
| 16:20.23 | landley | The new tabs show up in the old window, attached to... nothing? |
| 16:20.44 | landley | Acually right now #musl says it belongs to the server #edev. That's gonna end well. |
| 16:54.23 | *** join/#edev aupiff (~roy@199.87.82.114) |
| 17:09.19 | g1powermac | landley, well now thats interesting |
| 17:10.54 | landley | g1powermac: hmmm? |
| 17:11.13 | g1powermac | your xchat bug |
| 17:11.17 | landley | prpplague: what have you been up to in the past month? |
| 17:11.23 | landley | g1powermac: eh, I break stuff. |
| 17:11.47 | g1powermac | yea, always fun |
| 17:11.51 | landley | I detached the windows. Couldn't get 'em to attach to the new one, wound up closing it. |
| 17:12.02 | landley | If xchat crashes... oh well. |
| 17:12.12 | landley | It's got dangling participles. |
| 17:12.21 | g1powermac | heh |
| 17:28.14 | *** join/#edev NightMonkey (~NightrMon@pdpc/supporter/professional/nightmonkey) |
| 17:51.10 | *** join/#edev contempt (~contempt@unaffiliated/contempt) |
| 17:53.14 | calculus | landley: have you seen this, http://starchlinux.org/ ? |
| 18:00.28 | prpplague | landley: job hunting |
| 18:00.33 | prpplague | landley: panda is dead |
| 18:02.20 | kergoth | panda's dead? |
| 18:04.24 | prpplague | kergoth: yea basically, TI isn't going to really say it is dead, just not do anything more with it |
| 18:04.34 | prpplague | kergoth: no future designs |
| 18:04.38 | kergoth | ahh |
| 18:04.39 | prpplague | kergoth: no support |
| 18:04.40 | kergoth | that's too bad |
| 18:04.51 | prpplague | kergoth: i am the last of the people at TI working on it |
| 18:04.58 | prpplague | kergoth: and i'll be gone at the end of the month |
| 18:07.00 | calculus | :_( |
| 18:07.05 | calculus | cries for the panda |
| 18:08.06 | prpplague | a real shame too since we have an OMAP5 panda ready |
| 18:09.19 | landley | calculus: I'd heard somebody was working on one but didn't have the link. |
| 18:10.03 | landley | Eventually we'll want musl-linked versions of debian, ubuntu, fedora, suse, gentoo... |
| 18:10.17 | landley | If I had the darn bootstrapping stuff working this would be easy-ish. :) |
| 18:11.23 | landley | (The bootstrapping stuff is "from a chroot with busybox and a toolchain, build a new chroot with your distro's base root filesystem in it, make your repository work, then build all the packages to populate the repo". |
| 18:11.44 | landley | In theory, quite straightforward. In practice: gentoo assumes it's building under gentoo, fedora under fedora, debian under debian... |
| 18:11.59 | landley | And they're full of undocumented assumptions that just get sucked along from the previous version and were last actually _set_ sometime in the 90's. |
| 18:13.25 | *** join/#edev florian (~fuchs@Maemo/community/contributor/florian) |
| 18:13.33 | landley | prpplague: I'm up in minnesota on a 6 month contract for Cray. |
| 18:13.36 | landley | Going where the work is... |
| 18:13.45 | prpplague | landley: indeed |
| 18:13.47 | landley | (That technological hotbed: St. Paul.) |
| 18:13.55 | prpplague | landley: unfortunately i have critters to worry about |
| 18:14.40 | calculus | prpplague: you should sneak one of the said omap5 pandas and hand it to me |
| 18:14.46 | calculus | :) |
| 18:14.58 | prpplague | calculus: are you attending ELC? |
| 18:15.08 | landley | prpplague: I'm giving a talk at ELC. |
| 18:15.29 | prpplague | landley: hehe i know, hehe, i am program committee for ELC, hehe |
| 18:15.37 | landley | There you go then. :) |
| 18:15.55 | calculus | prpplague: probably just the wednesday demo night |
| 18:16.17 | prpplague | calculus: i can't guarantee anything, but i'll have a few boards with me |
| 18:16.21 | landley | I'd wondered why they approved one of the proposals I submitted. Threw three random ideas in there the last day of the CFP, one of 'em stuck. |
| 18:16.25 | landley | Nepotism: explains everything. |
| 18:17.28 | landley | Possibly if I get a few more quiet evenings I can convert aboriginal linux over to musl. |
| 18:18.00 | landley | toybox+musl gives us a fully bsd-licensed userspace. Then we just need a toolchain... |
| 18:22.28 | *** join/#edev vendu (~vendu@unaffiliated/vendu) |
| 18:22.31 | vendu | hello |
| 18:22.37 | vendu | anyone seen raster online lately? :) |
| 18:22.49 | wmat | landley: llvm perhaps? |
| 18:23.30 | landley | wmat: llvm, pcc, open64, tinycc/qcc, and there's a couple other candidates. |
| 18:23.59 | landley | llvm is the one Apple threw money at when gcc switched licenses to CDDL v2. |
| 18:24.51 | landley | Alas it's in C++, has a bunch of prerequisite packages, first built a subset of the linux kernel in october 2010 (still doesn't build the whole thing out of the box)... |
| 18:25.09 | landley | llvm started as an incremental gcc replacement and is still weaning itself off of the gnu/dammit toolchain. |
| 18:25.22 | wmat | indeed |
| 18:25.30 | landley | llvm was a code generator backend using the gcc front-end, then clang replaced that but they still used binutils and libstdc++ from gcc... |
| 18:25.43 | landley | They apparently have a libstdc++ now. I don't remember if they need lex and yacc... |
| 18:26.04 | landley | I liked gcc4.2.1 plus binutils 2.17 as two standalone packages with no external dependencies. |
| 18:26.16 | landley | Current gcc is like 5 packages before you get to binutils. |
| 18:26.38 | landley | tinycc is one package, but Fabrice moved on and the current guys have been going "we're going to get a release out any day!" for 4 months now. |
| 18:27.11 | landley | Not that this release will build any new packages the old one didn't (no kernel, no busybox). |
| 18:28.03 | landley | They just haven't managed to get a release out yet this decade so people think the project's dead. |
| 18:28.43 | landley | But at least they finally got the link to the mailing list updated on the web page, that only took a year. |
| 18:29.12 | wmat | heh |
| 18:29.28 | wmat | doesn't instill much confidence in the project |
| 18:29.55 | landley | (I need to do qcc. But toybox 1.0 comes first, and I think switching aboriginal to musl comes before either.) |
| 18:30.27 | landley | Script finished, day job beckons... |
| 18:39.13 | slashbeast | aboriginal to musl? |
| 18:39.29 | slashbeast | I bet that would make a lot of things not even build, because incompatibility with musl |
| 19:19.51 | landley | slashbeast: not really. I've been feeding bugs to dalias for a while now, and so have other people. |
| 19:20.19 | landley | And keep in mind aboriginal's uClibc right now. |
| 19:20.41 | landley | If you want a glibc chroot, you build one natively. That's part of the reason I've got a bootstrap-lfs control image. Does it for you. |
| 19:21.25 | landley | This means "get an s390 target working" and "get package X working on target" are completely separated. |
| 19:21.45 | landley | I believe musl already builds linux from scratch, although I'm still doing 6.8 and need to update my script. |
| 19:26.58 | slashbeast | landley: well I bet I will fail to build at least lvm2 and mdadm with musl at this very moment |
| 19:34.29 | *** join/#edev mrcan_ (~mrcan@unaffiliated/mrcan) |
| 19:45.20 | *** join/#edev grummund (~user@unaffiliated/grummund) |
| 19:51.32 | prpplague | landley: hehe yea we debated on which one of yours to accept |
| 19:51.47 | prpplague | landley: we generally only accept one per person |
| 19:54.58 | landley | slashbeast: I thought I'd finally fixed that mdadm thing you had lying around forever. It's not on my current todo list... |
| 20:02.01 | slashbeast | landley: yeah, it was missing symbols and I found them, but thats not what I am talking about |
| 20:02.17 | slashbeast | a lot of apps have switches for said libc, like -DHAS_UCLIBC etc |
| 20:07.22 | landley | slashbeast: I've poked dalias about that. |
| 20:08.07 | landley | I believe he's open to the possibility of adding one after 1.0, but until then any package that doesn't build against musl should either result in a change in musl or a patch to push upstream to the package. |
| 20:09.38 | slashbeast | if you are going to switch to musl same goes for better-initramfs so it will be fun. |
| 20:11.48 | landley | slashbeast: sabotage is a musl distro that's building lots of packages. I think the list is at https://github.com/rofl0r/sabotage/tree/master/pkg |
| 20:12.21 | prpplague | hmmm |
| 20:12.37 | landley | We have a nefarious scheme to make a big splash with the 1.0 release. |
| 20:12.43 | prpplague | i have 3 offers on the table, but the one i really want hasn't made an offer yet |
| 20:12.58 | landley | http://www.openwall.com/lists/musl/2012/11/29/7 |
| 20:12.58 | prpplague | wonders how to politely request that they make a decision |
| 20:13.17 | landley | Until then we're being... not exactly secretive, but making lists of things to post rather than posting 'em. |
| 20:13.35 | slashbeast | landley: last time I checked it used extreme outdated gcc |
| 20:13.43 | landley | prpplague: let them know you you have three other offers on the table, that you're very interested in the position but have a family to feed. |
| 20:14.00 | landley | slashbeast: I use an extremely outdated gcc. dalias does not. |
| 20:14.01 | prpplague | landley: yea, that is basically what i am writing up |
| 20:14.15 | landley | slashbeast: it works with both. |
| 20:14.40 | landley | In fact I had to update my binutils to a git snapshot to get a flag the musl build wanted. |
| 20:15.17 | calculus | prpplague: any in the bay area? |
| 20:15.34 | prpplague | calculus: yea one |
| 20:15.36 | prpplague | calculus: :( |
| 20:16.03 | calculus | prpplague: and probably not the one you are dying to hear from |
| 20:16.30 | prpplague | calculus: yea |
| 20:16.42 | prpplague | calculus: the one i want is in DFW with TI |
| 20:17.24 | prpplague | calculus: it really is funny how things are, places like google and lab126 dont care if you have a degree or not |
| 20:17.46 | landley | slashbeast: 90% of musl development actually happens on #musl here, the mailing list is a secondary channel. |
| 20:17.47 | prpplague | calculus: but places like TI, hitachi and panasonic, barely will talk to you if you don't have one |
| 20:18.20 | landley | prpplague: I pointed you at my old three waves spiel years ago, didn't I? |
| 20:18.44 | prpplague | landley: yea, i think i still have it bookmarked, hehe |
| 20:19.10 | landley | A degree is certification. Reduce risk by having an external authority provide a paper trail documenting the fact you can't be blamed for selecting this one. |
| 20:19.24 | landley | Bureaucracies love certifications, even meaningless ones. |
| 20:19.42 | landley | They cover their ass with them. |
| 20:19.47 | slashbeast | landley: whatsoever I will try to get it working as soon as you release aboriginal on musl |
| 20:20.05 | landley | slashbeast: not this weekend, but probably this month. |
| 20:23.24 | landley | slashbeast: sh4rm4 says lvm works with musl. |
| 20:24.26 | slashbeast | thats progress, I tried it like year? ago, maybe less, and I coudn't build lvm nor mdadm |
| 20:24.47 | landley | Keep in mind that musl is only a couple years old. It has been _hugely_ active in the past year. |
| 20:26.46 | slashbeast | don't worry, I am prepared for issues and such. |
| 21:02.49 | *** join/#edev FrostEyes (~frosteyes@users33.kollegienet.dk) |
| 21:08.07 | *** join/#edev grummund (~user@unaffiliated/grummund) |
| 21:13.56 | *** join/#edev contempt (~contempt@unaffiliated/contempt) |
| 21:22.42 | *** part/#edev grummund (~user@unaffiliated/grummund) |
| 21:29.54 | landley | slashbeast: actually sh4rm4 is testing mdadm uder musl as we speak. |
| 21:30.09 | landley | (I know this because he's complaining about insane gcc compilation time over in #musl.) |
| 21:30.28 | landley | -O0: 3 seconds, -Os 7 secs, -Os -s -flto -fwhole-program 30 seconds. |
| 21:30.32 | landley | That's gcc for you. |
| 21:31.42 | landley | dalias is currently auditing the thread locking functions so we can put the link time optimization blob in musl.a and let link time optimization penetrate the libc stuff. |
| 21:32.12 | landley | (Which is apparently bad if the function call stops being a memory barrier for the compiler and a locking function hasn't got an explicit memory barrier.) |
| 21:34.00 | *** join/#edev Posterdati (~antani@host107-26-dynamic.59-82-r.retail.telecomitalia.it) |
| 21:41.00 | CosmicPenguin | prpplague: you rang like 3 days ago? |
| 21:42.19 | prpplague | CosmicPenguin: hehe indeed |
| 22:12.54 | prpplague | CosmicPenguin: see /msg |
| 22:24.52 | *** join/#edev GPSFan (~kenm@64.92.145.112) |
| 22:24.52 | *** mode/#edev [+o GPSFan] by ChanServ |
| 22:34.39 | *** join/#edev contempt (~contempt@unaffiliated/contempt) |
| 23:35.24 | *** join/#edev contempt (~contempt@unaffiliated/contempt) |