| *** astrophys has quit IRC | 00:16 | |
| *** rjek has quit IRC | 00:50 | |
| *** rjek has joined #baserock | 00:50 | |
| *** gtristan has quit IRC | 02:49 | |
| *** gtristan has joined #baserock | 03:16 | |
| paulsherwood | baserock/python2.meta: - usr/lib/python2.7/ensurepip/_bundled/setuptools-15.2-py2.py3-none-any.whl | 07:38 |
|---|---|---|
| *** ssam2 has joined #baserock | 07:47 | |
| *** ChanServ sets mode: +v ssam2 | 07:47 | |
| *** tiagogomes has joined #baserock | 07:47 | |
| *** anahuelamo has joined #baserock | 08:05 | |
| *** toscalix has joined #baserock | 08:14 | |
| *** anahuelamo has quit IRC | 08:14 | |
| *** anahuelamo has joined #baserock | 08:15 | |
| *** anahuelamo has quit IRC | 08:16 | |
| *** anahuelamo has joined #baserock | 08:16 | |
| *** locallycompact has quit IRC | 08:37 | |
| *** anahuelamo has quit IRC | 08:54 | |
| *** anahuelamo has joined #baserock | 08:58 | |
| *** franred has joined #baserock | 09:03 | |
| *** locallycompact has joined #baserock | 09:31 | |
| *** fay_ has joined #baserock | 09:37 | |
| *** rdale has joined #baserock | 09:48 | |
| franred | anahuelamo, do you really need everything is in the MPC repo to be copied to /opt/MPC? https://gerrit.baserock.org/#/c/2131/3/strata/corba/MPC.morph | 10:15 |
| anahuelamo | franred, yes, I need those files to generate Makefiles for ace_tao components | 10:16 |
| franred | all of it or only pls and some folders? http://git.baserock.org/cgit/delta/MPC.git/tree/ -- bear in mind that you are going to install that in your final system | 10:19 |
| franred | if you add the corba stratum to any system | 10:21 |
| anahuelamo | I'll review it and I'll use it just the things needed. Thanks for the review! | 10:21 |
| franred | no probs, maybe richard_maw or ssam2 has some solution about how to add that only on build time and it does not go to the system (if it is only required for building other stuffs) | 10:23 |
| ssam2 | if it's source code, maybe install it into /usr/src | 10:27 |
| ssam2 | then we can delete all of /usr/src when making images that shouldn't have such things inside it | 10:27 |
| anahuelamo | ok, thank ssam2 ! | 10:28 |
| *** fay_ has quit IRC | 11:25 | |
| paulsherwood | someone just asked about go in baserock... looks like it didn't get merged.... http://git.baserock.org/cgit/baserock/baserock/definitions.git/commit/?h=baserock/ps/docker&id=2047757886509d4b2530a6f1d0c32be4a6758ca0 | 11:58 |
| ssam2 | i remember that working when I tested it | 12:39 |
| ssam2 | so maybe that would even be good enough to bootstrap a more modern version of the Go compiler | 12:40 |
| locallycompact | go > 1.4 requires go jsyk | 12:51 |
| locallycompact | would someone be so kind as to check these lorries https://gerrit.baserock.org/#/c/2133/ | 12:53 |
| SotK | should those ".morph"s be ".lorry"s? | 12:54 |
| locallycompact | ahh, brainfail | 12:54 |
| locallycompact | sorry | 12:54 |
| locallycompact | fixd | 12:56 |
| paulsherwood | locallycompact: fix the whitespace, then i expect someone will +2 | 12:58 |
| locallycompact | ack | 12:59 |
| locallycompact | is https or http preferred/required for github? | 12:59 |
| locallycompact | or no matter | 13:00 |
| rjek | https is preferred by me | 13:00 |
| persia | HTTPS provides checksum validation to ensure the contents match the request. HTTP does not | 13:02 |
| rjek | It's very much more robust as well as secure | 13:05 |
| paulsherwood | anyone fancy fixing kbas to do https? | 13:05 |
| rjek | kbas should be a CGI/FastCGI, and the network protocol handled by a web server, IMO | 13:06 |
| paulsherwood | locallycompact: merged | 13:29 |
| locallycompact | tyvm | 13:29 |
| paulsherwood | rjek: that wouldn't work for some use-cases (eg, i'm a developer... i've built some stuff.. i want to serve it for my colleagues) | 13:30 |
| rjek | Your developer can't run a web server? | 13:32 |
| rjek | That, or perhaps they should all share a kbas | 13:32 |
| paulsherwood | rjek: the ad-hoc usecase has proved useful a few times already | 13:34 |
| ssam2 | there's a python module in the standard library that can provide a web server | 13:38 |
| ssam2 | https://www.linuxjournal.com/content/tech-tip-really-simple-http-server-python | 13:38 |
| ssam2 | a simple one suitable only for testing/hacks | 13:38 |
| rjek | adhoc use sounds like a hack :) | 13:38 |
| paulsherwood | ssam2: ok, but kbas is not *just* ad-hoc... it handles simultaneous uploads, downloads, is atomic etc | 13:40 |
| paulsherwood | i'm happy with it bar the fact that its security is woeful | 13:41 |
| ssam2 | but for non-adhoc cases you can put it behind a real webserver | 13:42 |
| rjek | A much simpler CGI plugged into a regular web server could be all of those things, except insecure :) | 13:42 |
| rjek | Probably faster because it could abuse the sendfile() header some web servers support | 13:42 |
| paulsherwood | this is all very well... but show me the code, please :) kbas is already serving artifacts for real. i'd be happy to improve it, but i don't have the requisite knowledge to close the gap between your statements... and an actual implementation that anyone can installe | 13:44 |
| rjek | (You put an X-Sendfile: header in your reply, and the web server disregards any body you provide and just asks the kernel to squirt the given file down the socket) | 13:44 |
| paulsherwood | rjek: patches, please :-) | 13:45 |
| paulsherwood | https://github.com/devcurmudgeon/ybd/blob/master/kbas/__main__.py is the whole of kbas | 13:46 |
| rjek | Actually, I'd probably just have the CGI for handling uploads, and create a symlink farm that points to the actual file on disc, and map that into the URL namespace and have the web server handle it entirely | 13:46 |
| paulsherwood | locallycompact: did you miss a couple? Submodule 'src/compiler-rt' (https://github.com/rust-lang/compiler-rt.git) registered for path 'src/compiler-rt' | 13:49 |
| paulsherwood | Submodule 'src/jemalloc' (https://github.com/rust-lang/jemalloc.git) registered for path 'src/jemalloc' | 13:49 |
| paulsherwood | Submodule 'src/liblibc' (https://github.com/rust-lang/libc.git) registered for path 'src/liblibc' | 13:49 |
| paulsherwood | Submodule 'src/llvm' (https://github.com/rust-lang/llvm.git) registered for path 'src/llvm' | 13:49 |
| rjek | paulsherwood: Is the protocol/API documented anywhere? | 13:49 |
| paulsherwood | Submodule 'src/rt/hoedown' (https://github.com/rust-lang/hoedown.git) registered for path 'src/rt/hoedown' | 13:49 |
| paulsherwood | Submodule 'src/rust-installer' (https://github.com/rust-lang/rust-installer.git) registered for path 'src/rust-installer' | 13:49 |
| paulsherwood | rjek: sadly, no. but i'll document it if there's a chance you'd supply patches :-) | 13:50 |
| rjek | There's a chance of anything. | 13:51 |
| * paulsherwood is unconvinced | 13:52 | |
| locallycompact | paulsherwood, it's a straight mirror I reckon | 13:52 |
| locallycompact | we have the original on g.b.o | 13:53 |
| paulsherwood | locallycompact: ack | 13:55 |
| * locallycompact hits something | 14:08 | |
| locallycompact | the build in ybd just seems to stop here, doens't have the same problem locally in baserock https://paste.fedoraproject.org/363384/25437411/ | 14:09 |
| *** cyndis has quit IRC | 14:09 | |
| paulsherwood | locallycompact: there seems to be an actual error in that build? | 14:11 |
| paulsherwood | line 578 | 14:11 |
| paulsherwood | it's calling out to the internet | 14:11 |
| locallycompact | ofc it is | 14:11 |
| paulsherwood | interesting that a project that is aiming for reliability is happy to do that during a build | 14:12 |
| * locallycompact toddles off to irc.mozilla | 14:12 | |
| *** astrophys has joined #baserock | 14:18 | |
| anahuelamo | ssam2, I tried to do what you told me before: "ssam2> if it's source code, maybe install it into /usr/src | 14:29 |
| anahuelamo | <ssam2> then we can delete all of /usr/src when making images that shouldn't have such things inside it" I don't know if I'm supposed to remove the files from /usr/src/ after I use them or if that's not needed? | 14:29 |
| ssam2 | it's not really possible to do that at build time | 14:30 |
| ssam2 | i'd leave it there, and worry about it at deploy time | 14:30 |
| franred | ssam2, regarding https://gerrit.baserock.org/#/c/2131/4/strata/corba/ace_tao.morph and if anahuelamo has tested it, it seems to be possible to remove /usr/src/MPC on install-commands :? | 15:01 |
| anahuelamo | franred, I tested, but failed. I just push it again without removing the files at the end | 15:02 |
| franred | anahuelamo, oh, ok. try to prune a little bit MPC directory then if possible and sensible | 15:03 |
| tiagogomes | you want `rm -rf $DESTIR/usr/src/MPC` and post-install-commands would fit better for that | 15:03 |
| franred | tiagogomes, it won't work, because it is in a different chunk when anahuelamo wants to remove that file | 15:04 |
| franred | s/file/folder | 15:04 |
| franred | / | 15:04 |
| ssam2 | that definitely can't work | 15:04 |
| franred | you are right | 15:05 |
| anahuelamo | I can prune MPC directory then, but if I can remove the files at deploy time will be better | 15:05 |
| ssam2 | a build tool constructs a system rootfs by appending together a bunch of .tar files | 15:05 |
| franred | ssam2, would work as a system integration command? | 15:05 |
| ssam2 | oh, as a system-integration-command it would | 15:05 |
| ssam2 | seems a bit of a misuse of those, but not so bad | 15:05 |
| franred | anahuelamo, ^^ | 15:05 |
| anahuelamo | franred, ssam2 any help with the system-integration-command? anywhere I can look at to see how to do that? | 15:06 |
| ssam2 | if you grep in definitions.git you'll find some examples | 15:07 |
| ssam2 | i think it's actually called 'system-integration' | 15:07 |
| anahuelamo | thanks for all the help! | 15:07 |
| franred | anahuelamo, ssam2, in ace_tao.morph add in the end something like http://paste.baserock.org/gokivekaho | 15:10 |
| franred | anahuelamo, I haven't test it and you won't see the action until you have these strata in a system and boot the system, though | 15:11 |
| paulsherwood | locallycompact: what's https://github.com/locallycompact/rust-ybd about, then? :-) | 15:12 |
| locallycompact | Hard to say | 15:13 |
| locallycompact | It might be ybd in rust or it might be some random scribbles | 15:14 |
| paulsherwood | :) | 15:14 |
| locallycompact | builds with cargo build / cargo run | 15:15 |
| anahuelamo | thanks franred! I'll put it in the chunk and test now :) | 15:16 |
| * locallycompact will have to accost steve klabnik about the rust boostrap thing | 15:18 | |
| paulsherwood | locallycompact: before you go too far, please let's address the naming problem | 15:22 |
| paulsherwood | chunk is ok for me... stratum not so. i believe we need some noun for a set of (set of) chunks to include what we currently think of as clusters, systems, strata | 15:23 |
| *** astrophys has quit IRC | 15:24 | |
| locallycompact | paulsherwood, sure | 15:27 |
| *** astrophys has joined #baserock | 15:27 | |
| paulsherwood | group was suggested at one point | 15:27 |
| locallycompact | no | 15:27 |
| paulsherwood | i don't like layer, really | 15:27 |
| rjek | metapackage :D | 15:27 |
| paulsherwood | you're trolling :) | 15:28 |
| locallycompact | group has a single meaning across a lot of fields, including comp sci | 15:28 |
| paulsherwood | ok | 15:28 |
| * paulsherwood => meeting | 15:28 | |
| rjek | metachunk! | 15:28 |
| *** gtristan has quit IRC | 15:29 | |
| *** gtristan has joined #baserock | 15:39 | |
| *** ssam2 has quit IRC | 16:12 | |
| *** tiagogomes has quit IRC | 16:14 | |
| *** dabukalam has quit IRC | 16:25 | |
| *** dabukalam has joined #baserock | 16:25 | |
| *** jmacs has quit IRC | 16:28 | |
| *** SotK has quit IRC | 16:28 | |
| *** SotK has joined #baserock | 16:28 | |
| *** jmacs has joined #baserock | 16:28 | |
| paulsherwood | megachunk. chunks. chunki. chunkae. none of these :) | 16:44 |
| *** toscalix has quit IRC | 16:45 | |
| *** franred has quit IRC | 16:56 | |
| *** locallycompact has quit IRC | 17:03 | |
| *** anahuelamo has quit IRC | 17:04 | |
| *** rdale has quit IRC | 18:50 | |
| *** gtristan has quit IRC | 19:18 | |
| *** gtristan has joined #baserock | 19:27 | |
| *** jjardon_ has joined #baserock | 19:46 | |
| *** jjardon has quit IRC | 19:48 | |
| *** jjardon_ is now known as jjardon | 19:48 | |
| *** gtristan has quit IRC | 20:58 | |
| *** paradigm has quit IRC | 21:03 | |
| *** gtristan has joined #baserock | 21:12 | |
| *** rdale has joined #baserock | 21:50 | |
| *** rdale has quit IRC | 22:17 | |
| *** astrophys has quit IRC | 22:19 | |
| *** astrophys has joined #baserock | 23:10 | |
| *** gtristan has quit IRC | 23:13 | |
| *** gtristan has joined #baserock | 23:14 | |
| *** astrophys has quit IRC | 23:19 | |
| *** cyndis has joined #baserock | 23:20 | |
| *** astrophys has joined #baserock | 23:21 | |
| *** astrophys has quit IRC | 23:43 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!