| *** mSher [~mike@82-70-136-246.dsl.in-addr.zen.co.uk] has joined #baserock | 05:53 | |
| *** zoli_ [~zoli_@linaro/zoli] has joined #baserock | 05:54 | |
| *** zoli_ [~zoli_@linaro/zoli] has quit [Remote host closed the connection] | 06:14 | |
| *** zoli_ [~zoli_@linaro/zoli] has joined #baserock | 06:35 | |
| *** dabukalam [~dabukalam@87.113.195.160] has quit [Ping timeout: 246 seconds] | 06:50 | |
| *** dutch [~william@82-70-136-246.dsl.in-addr.zen.co.uk] has joined #baserock | 07:00 | |
| *** tiagogomes [~tiagogome@82-70-136-246.dsl.in-addr.zen.co.uk] has joined #baserock | 08:08 | |
| *** rdale [~quassel@82-70-136-246.dsl.in-addr.zen.co.uk] has joined #baserock | 08:11 | |
| *** rdale [~quassel@82-70-136-246.dsl.in-addr.zen.co.uk] has quit [Ping timeout: 244 seconds] | 08:20 | |
| *** rdale [~quassel@82-70-136-246.dsl.in-addr.zen.co.uk] has joined #baserock | 08:20 | |
| *** jonathanmaw [~jonathanm@82-70-136-246.dsl.in-addr.zen.co.uk] has joined #baserock | 08:34 | |
| *** ssam2 [~ssam2@82-70-136-246.dsl.in-addr.zen.co.uk] has joined #baserock | 08:42 | |
| *** logger_ [~ubuntu@85.199.252.110] has quit [Quit: leaving] | 09:07 | |
| pedroalvarez | any operator around? | 09:12 |
|---|---|---|
| *** br_logger [~ubuntu@85.199.252.110] has joined #baserock | 09:16 | |
| *** dutch [~william@82-70-136-246.dsl.in-addr.zen.co.uk] has quit [Ping timeout: 260 seconds] | 09:16 | |
| *** CTtpollard [~tom@82-70-136-246.dsl.in-addr.zen.co.uk] has quit [Ping timeout: 272 seconds] | 09:17 | |
| *** dutch [~william@82-70-136-246.dsl.in-addr.zen.co.uk] has joined #baserock | 09:20 | |
| *** CTtpollard [~tom@82-70-136-246.dsl.in-addr.zen.co.uk] has joined #baserock | 09:21 | |
| *** br_logger [~ubuntu@85.199.252.110] has quit [Quit: leaving] | 09:23 | |
| *** br_logger [~ubuntu@85.199.252.110] has joined #baserock | 09:33 | |
| Mode #baserock +o Kinnison by ChanServ | 09:40 | |
| pedroalvarez | Kinnison: ca we change the topic to remove the warning about g.b.o and put that the channel is being logged? | 09:41 |
| Kinnison | is there a URL you want me to put for log access? | 09:45 |
| pedroalvarez | for now they are being published here: http://85.199.252.110/ | 09:48 |
| pedroalvarez | Kinnison: thanks | 09:48 |
| *** zoli_ [~zoli_@linaro/zoli] has quit [Remote host closed the connection] | 09:49 | |
| *** zoli_ [~zoli_@linaro/zoli] has joined #baserock | 09:49 | |
| Mode #baserock +v pedroalvarez by ChanServ | 09:50 | |
| Mode #baserock +v ssam2 by ChanServ | 09:51 | |
| *** jjardon_ [~jjardon@demo.convos.by] has joined #baserock | 09:55 | |
| Mode #baserock +o ssam2 by ChanServ | 09:57 | |
| ssam2 | cool! | 09:58 |
| persia | Kinnison: Do you have the magic authentication tokens to the DNS API? | 09:58 |
| Kinnison | I *may* | 09:58 |
| * Kinnison looks | 09:58 | |
| pedroalvarez | I'd like to add a /ignore rule in the irssi logger to ignore messages with "nologplease" or similar | 09:58 |
| Kinnison | heh | 09:58 |
| straycat | If you wanted to do that you'd want to store state for each nick or somethink like that, so that all messages from a certain nick are filtered out and possibly replaced with a message noting that the original message has been omitted. | 10:00 |
| straycat | But you can't really do that properly here because there's no requirement that people joining this channel have registered nicks. | 10:01 |
| persia | Could send messages to unregistered nicks that they either need to be registered to toggle logging or preface each statement with a magic token | 10:02 |
| pedroalvarez | that sounds like a bot, not like irssi logging :) | 10:03 |
| Kinnison | persia: It appears I do have access. Whether I *should* or not is another question :-) | 10:04 |
| persia | irssi supports sufficient scripting that the distinction is hard to draw, but yes, that reaches well into intelligent automation territory :) | 10:04 |
| persia | Kinnison: The key bit is that *someone* should. If you are not comfortable with it being you, I suggest a call for volunteers on baserock-dev@ to find a replacement. | 10:05 |
| persia | For now, I think it would be cool to add names for some of the services pedroalvarez has published (mason, irc logs, etc.) | 10:06 |
| Mode #baserock +v richard_maw by ChanServ | 10:06 | |
| Kinnison | There is testgerrit | 10:06 |
| Kinnison | do we want testmason, testirclogs, etc. ? | 10:06 |
| Mode #baserock +o richard_maw by ChanServ | 10:06 | |
| Kinnison | pedroalvarez: ^^ | 10:06 |
| pedroalvarez | +1 to testirclogs | 10:08 |
| pedroalvarez | about mason, doesn't sound like has to have the "test" prefix | 10:09 |
| Kinnison | so mason.basrock.org should have what address? | 10:10 |
| pedroalvarez | I'd add the arch in the url | 10:10 |
| SotK | does that mason run tests? | 10:10 |
| pedroalvarez | cause I was going to set up a armv7 mason | 10:10 |
| pedroalvarez | SotK: no | 10:11 |
| Kinnison | pedroalvarez: mason-x86-64 | 10:11 |
| Kinnison | pedroalvarez: IP? | 10:11 |
| pedroalvarez | 85.199.252.101 | 10:11 |
| Kinnison | Added. I believe the typical propagation delay is one hour | 10:11 |
| pedroalvarez | Kinnison: tvm :) | 10:13 |
| * Kinnison has also put in-train a discussion about getting the DNS moved somewhere where the community can manage it | 10:14 | |
| * persia believes that "Kinnison is a member of the community and has a bag of magic tokens" counts as "the community can manage it" | 10:19 | |
| * Kinnison believes that "the community can manage it" means more than one member thereof | 10:20 | |
| Kinnison | hence today I spread ops for #baserock to more community members | 10:20 |
| persia | Although we should probably have an infra page on the wiki, with indicators of who has tokens (and Ideally have more than one person able to handle requests for each role). | 10:20 |
| persia | Kinnison: Right. | 10:20 |
| * Kinnison wants to roll baserock.org DNS, MX, listmaster etc into one account somewhere separable from where it all is right now | 10:20 | |
| persia | Kinnison: Out of curiosity, is "baserock" a registered prefix with staff? | 10:20 |
| Kinnison | No, it's pure channel config for now | 10:21 |
| Kinnison | If you want to organise proper registration in freenode, that'd be cool | 10:21 |
| persia | I'd rather multiple roles. Folk who do mail moderation, IRC admin, and infra management tend to have different attitudes. | 10:21 |
| * persia will put that on the list | 10:21 | |
| Kinnison | Okay | 10:21 |
| *** jjardon_ [~jjardon@demo.convos.by] has quit [Quit: jjardon_] | 10:25 | |
| pedroalvarez | persia: I've been thinking this morning about the idea of an infra wiki page. | 10:40 |
| pedroalvarez | good to know that it's not only in my head | 10:41 |
| persia | Minds of similar calibers follow similar grooves :) | 10:41 |
| richard_maw | persia: I like that wording of the phrase. It's less smug. | 10:44 |
| straycat | it also explains why i still spend my evenings watching children's cartoons | 10:44 |
| persia | richard_maw: Well either we're both forward-thinking visionaries, or we're both so forgetful we need a wiki page to track stuff everyone knows. It's hard to say on which side the argument for such documentation lies :) | 10:45 |
| radiofree | how come, when using bash in baserock, my delete key becomes "~" | 10:51 |
| radiofree | have to set some keyboard layout somewhere? | 10:52 |
| persia | Lack of sufficient terminfo | 10:52 |
| persia | Also, weak locale support | 10:52 |
| persia | If you happen to configure your local terminal and locale to "linux" and "C", I think it kinda works. | 10:53 |
| ssam2 | radiofree: try copying in the /etc/inputrc from a non-Baserock system | 10:55 |
| radiofree | ssam2: yay that worked, thanks | 10:56 |
| ssam2 | I believe that file should be installed by the GNU Readline chunk, or something | 10:57 |
| ssam2 | but I've never had time to investigate fully | 10:57 |
| petefoth | <@ssam2> radiofree: try copying in the /etc/inputrc from a | 11:05 |
| petefoth | non-Baserock systemwould go well on a 'BAserock Tips and Tricks' page | 11:05 |
| * petefoth realises he may have just volunteered for something :) | 11:06 | |
| jjardon | radiofree: add "\e[3~": delete-char in ~/.inputrc | 11:18 |
| jjardon | oh!, or what Sam said | 11:19 |
| pedroalvarez | i found this regarding imputrc: http://www.linuxfromscratch.org/blfs/view/5.1/postlfs/inputrc.html | 11:25 |
| * jjardon reading about readline just remembers him to send a patch to upgrade readline as the current in baserock is quite old (and makes NetworkManager master no to compile) | 11:28 | |
| pedroalvarez | jjardon: we are on that version because of the license implications for Genivi | 11:29 |
| *** liw [liw@xvm-166-37.ghst.net] has quit [] | 11:53 | |
| *** tiagogomes [~tiagogome@82-70-136-246.dsl.in-addr.zen.co.uk] has quit [Ping timeout: 272 seconds] | 11:59 | |
| Kinnison | GPLv3 :-( | 12:00 |
| rjek | wtf does network manager need readline?! | 12:04 |
| Kinnison | to encourage lennart to add network stuff to systemd | 12:04 |
| rjek | Ah, understood. | 12:05 |
| * richard_maw rolls eyes | 12:05 | |
| richard_maw | there's a command line client | 12:05 |
| rjek | Will he be enveloping `ip` and making its command line syntax even more arcane? :) | 12:05 |
| * richard_maw sighs and seeks biscuits | 12:06 | |
| rjek | `ip` should really be called `if`, but it seems that name was already taken | 12:07 |
| persia | Is there any strong reason to keep "readline" at a single version? Why not give it two chunk names, one the new version, and one the old, and use different ones for different systems? | 12:08 |
| * rjek went to the effort of writing "fakeline", which is ABI compatible with readline but the only call that actually does anything just uses read() | 12:08 | |
| Kinnison | persia: readline is pretty deep in the stack, we'd have to have a way to parameterise strata from above before we could do that effectively | 12:10 |
| richard_maw | nah, we could make the co-installable | 12:10 |
| * straycat wants cpython to continue to be built with curses and readline | 12:10 | |
| richard_maw | and we'd only use the old readline for stuff which goes to genivi | 12:10 |
| persia | This would be a great time to dispose of the strata concept in favor of arbitrary hierarchies :) | 12:10 |
| straycat | just as I add a command that expects a stratum? >.> | 12:11 |
| * straycat disappears | 12:11 | |
| persia | straycat: Of course: if we all had the same goals, everything would work, but life would be boring :) | 12:12 |
| rjek | Provides: ? | 12:12 |
| *** tiagogomes [~tiagogome@82-70-136-246.dsl.in-addr.zen.co.uk] has joined #baserock | 12:13 | |
| * rjek wonders if it would be nice for that. If you have a missing dependancy, morph build spits out a list of chunks you could use to satisfy it. | 12:13 | |
| persia | rjek: Doesn't help, because of caching: we need to be able to differentiate the same tree hash for source A compiled against the two different readlines in a way that can be correctly identified when assembling a system. | 12:13 |
| ssam2 | franred: hey, you might already know about this, but turns out there's an 'all of OpenStack' git repo | 12:14 |
| ssam2 | https://github.com/openstack/openstack | 12:14 |
| * ssam2 just discovered it | 12:14 | |
| * persia would also prefer to avoid needing to define policies for various "Provides" that cause confusion later: as an example, there are many tools in Debian that both consume and produce SMTP, but only a small subset bother to provide "mail-transport-agent" | 12:15 | |
| ssam2 | ""Any commit in this repo will get a collection of commits in the other repos that have explicitly been tested with each other, if that sort of thing is important to you."" | 12:15 |
| persia | The benefits of pre-commit CI :) | 12:16 |
| franred | yeah, but that does not include all the python dependencies for all the modules :/ | 12:19 |
| franred | in any case, ssam2, thanks for the link :) | 12:21 |
| franred | and yes, rjek, that would be great but you have autotools, ruby, python, .... which implies to create a big set of tools to import/create new chunks - I think we have started to do this for ruby and python | 12:23 |
| Mode #baserock +o pedroalvarez by ChanServ | 12:42 | |
| straycat | persia, I think that depends on how broad (or not) the goals are. :) | 13:20 |
| wikicat | Wiki change: move 'Moc OS with Qemu' down the list, as it ie probably less commong than using VirtualBox http://source.baserock.branchable.com/?p=source.git;a=commitdiff;h=c73169d | 13:31 |
| wikicat | Wiki change: Rationalise the VirtualBox stuff. Remove duplication. USe constent names for the downloaded image' http://source.baserock.branchable.com/?p=source.git;a=commitdiff;h=88e6bb1 | 13:31 |
| wikicat | Wiki change: clarify we are settining up a development image http://source.baserock.branchable.com/?p=source.git;a=commitdiff;h=3da6654 | 13:31 |
| wikicat | Wiki change: S11704 Top-level reqs / Stories for current Baserock work http://source.baserock.branchable.com/?p=source.git;a=commitdiff;h=62ef507 | 13:31 |
| wikicat | Wiki change: Removed recommendation to use British spelling only per instructions from Paul Sherwood. http://source.baserock.branchable.com/?p=source.git;a=commitdiff;h=52b5098 | 13:35 |
| wikicat | Wiki change: typos http://source.baserock.branchable.com/?p=source.git;a=commitdiff;h=ac792e5 | 13:49 |
| *** genii [~quassel@ubuntu/member/genii] has joined #baserock | 13:49 | |
| straycat | richard_maw, I don't really know when I'm supposed to use generators over lists. Nothing I'm writing is using much memory. | 13:58 |
| persia | I always thought it was a shortcut optimisation: that one used a generator when one didn't expect to be exhaustive, and used a list when one did want to be exhaustive (unless the list would be annoyingly long). | 14:01 |
| richard_maw | I generally use generators when I only need an iterable to be iterated once. If I need to use it again then I use a list comprehension instead, or pass the generator to a set | 14:03 |
| richard_maw | partly I just think it looks nicer to pass a generator to a set constructor, but there's also speed and memory optimisations available | 14:03 |
| * persia hopes there is an "unless I'm dealing with several billion members of the set" exception there | 14:03 | |
| richard_maw | persia: yeah, if that's going on you either have to filter out the stuff you're interested in, or face the fact that you'll have to generate the iterable multiple times, rather than re-using it | 14:05 |
| persia | For that class of thing, if I can't parallelise (lack of available resources), I usually just iterate over a generator, without ever constructing a list or a set. | 14:05 |
| richard_maw | there's itertools.tee if you want to pass a generator to multiple consumers, but you have to make sure the users of the iterators consume it in parallel, or you end up storing the whole thing as a list internally | 14:06 |
| persia | Right. At that point it becomes a hardware question :) | 14:07 |
| * richard_maw sometimes wants a statically typed language with lazy evaluation with the syntax of python | 14:08 | |
| * persia does not intend that to be taken as a suggestion to add hardware capability probing into software to select alternate algorithms except in the most specific of circumstances | 14:08 | |
| jmacs | richard_maw: I'd like that too. | 14:09 |
| petefoth | Can anyone give me the correwtc spell (to put in the wiki - if it were for my own use I'd trey to work it out myself) fr using sshfs to mount the guest vm's file system in the host? | 14:11 |
| ssam2 | petefoth: I use `sshfs -o idmap=user root@IP:/src /vm-src' | 14:12 |
| ssam2 | to mount /src in my VM to /vm-src in my desktop | 14:12 |
| ssam2 | idmap=user translates things owned by 'root' in the VM to be owned by my user account in my desktop | 14:13 |
| petefoth | ssam2: perfect! Are you using KVM or VB for thr virutalisation? | 14:13 |
| ssam2 | KVM, but should be the same for either | 14:13 |
| richard_maw | we've encountered differences in the past | 14:14 |
| ssam2 | with how the ssh daemon inside the VM behaves ??? | 14:14 |
| ssam2 | ouch! | 14:14 |
| richard_maw | I think we ended up configuring VB to proxy the ssh port locally, so you ended up with an sshfs command as `sshfs -o idmap=user,port=12345 root@localhost:/src /vm-src` | 14:15 |
| straycat | richard_maw, How come you didn't submit the yarn changes upstream? | 14:16 |
| ssam2 | ok. any baserock-in-virtualbox users able to advise petefoth on an approach for sshfs that they know works ? | 14:17 |
| * straycat advises not using virtual box | 14:17 | |
| * radiofree suggests the sshfs line from ssam2 be added to petefoth's "tips and tricks" page | 14:17 | |
| richard_maw | straycat: because I knew upstream was unable to devote time to reviewing it, and I've previously gotten a wait-and-see response from upstream | 14:18 |
| straycat | Ahh | 14:18 |
| ssam2 | radiofree: i believe that's the reason petefoth asked for the line in the first place ;) | 14:18 |
| petefoth | radiofree: I'm in the process of adding it to the 'Setting up your VM' page | 14:19 |
| petefoth | straycat: I like VirtualBox because it is cross-platform whereas KVM and QEMU are less so. | 14:20 |
| straycat | petefoth, I dislike it because I've written software that uses and so know how terrible it is. | 14:20 |
| petefoth | straycat: yes, but it works :) (at least for me) | 14:21 |
| straycat | *uses it | 14:21 |
| * Kinnison dislikes VirtualBox because its networking is (deliberately) much less capable than KVMs | 14:22 | |
| straycat | It also has various commands that will fail but exit 0 return status. | 14:24 |
| robtaylor | so recenly i discovered that samsung have been implemented a drm capable backend for qemu | 14:26 |
| robtaylor | just got it working with i386 emulated | 14:27 |
| * petefoth scrolls backwards to find the sshfs spells :) | 14:27 | |
| robtaylor | and now with kvm | 14:30 |
| robtaylor | woo | 14:30 |
| radiofree | robtaylor: awesome! | 14:31 |
| Kinnison | robtaylor: that's pretty cool | 14:32 |
| radiofree | link? | 14:32 |
| Kinnison | robtaylor: so 3D acceleration is plausible for kvm now? | 14:32 |
| straycat | richard_maw, If your yarn changes work I think we should merge them into baserock/morph, it's up to upstream to review those changes and decide whether they want them or not. | 14:33 |
| richard_maw | straycat: yep, that was my plan | 14:33 |
| * straycat nods | 14:34 | |
| richard_maw | robtaylor: is that related to http://lwn.net/Articles/611212/ ? | 14:34 |
| *** pdar [~patrickda@access.ducie-dc1.codethink.co.uk] has quit [Ping timeout: 245 seconds] | 14:36 | |
| *** pdar [~patrickda@access.ducie-dc1.codethink.co.uk] has joined #baserock | 14:36 | |
| wikicat | Wiki change: Add to the 'using sshfs section. It's not complete, but it's better http://source.baserock.branchable.com/?p=source.git;a=commitdiff;h=6d96cc2 | 14:37 |
| wikicat | Wiki change: Rationalise the stuff about ssh http://source.baserock.branchable.com/?p=source.git;a=commitdiff;h=8a17aae | 14:37 |
| petefoth | Hmm wikicat seems to have problem with the oerder changes are made | 14:39 |
| straycat | It's probably reverse order. | 14:41 |
| robtaylor | richard_maw: nope | 14:42 |
| robtaylor | richard_maw: and its massivly undocumented and undiscussed | 14:42 |
| robtaylor | Kinnison: yes, appears so. I just ran a webgl demo in a qemu with -enable-kvm | 14:43 |
| robtaylor | I'll push everything up to my github so others can play | 14:43 |
| Kinnison | robtaylor: awesome | 14:43 |
| *** zoli_ [~zoli_@linaro/zoli] has quit [Remote host closed the connection] | 15:23 | |
| pedroalvarez | moing to glibc is going to be more difficult than I expected | 15:40 |
| pedroalvarez | s/moing/moving/ | 15:41 |
| richard_maw | what's causing you problems? | 15:41 |
| pedroalvarez | building linux-api-headers (stage3) fails when cross-bootstraping because it tries to use bash, and bash is broken because it needs ncurses | 15:44 |
| pedroalvarez | hmm, I guess that adding a stage2-ncurses will fix that | 15:44 |
| richard_maw | this also requires you put a bash in build-essential | 15:45 |
| Kinnison | pedroalvarez: IIRC linux-api-headers will actually work with sh, but you need a dodgy script to emulate bash being around | 15:45 |
| richard_maw | I don't suppose you can patch the linux-api-headers chunk? | 15:45 |
| Kinnison | rjek encountered this when he was trying to bootstrap mips64 -- he may have input | 15:47 |
| rjek | hmm wah whatnow? | 15:47 |
| pedroalvarez | but, we are building linux-api-headers without bash, and it works, aren't we? | 15:47 |
| richard_maw | then it'd be something in glibc that for some reason now requires bash | 15:48 |
| rjek | Yes. | 15:48 |
| pedroalvarez | I thought that the problem here is that when builing the stage3 in cross-bootstrap I can't say what build dependencies I need | 15:48 |
| richard_maw | or you've updated linux-api-headers to a version which uses bash | 15:48 |
| rjek | I had a shell script called "bash" that just invoked busybox sh | 15:48 |
| pedroalvarez | I haven't updated linux-api-headerd, but having bash in the system when installing l-a-headers makes the installation behave different | 15:52 |
| ssam2 | i'd much prefer a 'stage2-fake-bash' chunk to adding GNU Bash and ncurses and whatever else GNU Bash grows as a dependency in future to stage2 of the bootstrap ... | 15:52 |
| * richard_maw wonders if it's because eglibc had a patch to make ldd work with /bin/sh rather than /bin/bash | 15:53 | |
| ssam2 | the only shell script that we need for producing the linux-api-headers chunk is http://git.baserock.org/cgi-bin/cgit.cgi/delta/linux.git/tree/scripts/headers_install.sh?h=baserock/v3.8 | 15:54 |
| ssam2 | which has at the top: #!/bin/sh | 15:54 |
| ssam2 | so I don't see why that'd need GNU Bash anyway | 15:54 |
| Kinnison | we run make mrproper | 15:55 |
| Kinnison | that's possibly not necessary | 15:55 |
| richard_maw | pedroalvarez: can you check whether the ldd script generated by the glibc has #!/bin/bash at the top? | 15:55 |
| pedroalvarez | richard_maw: if bash is present when building glibc, yes | 15:56 |
| pedroalvarez | otherwise it will have #!/bin/ | 15:56 |
| pedroalvarez | sorry | 15:56 |
| pedroalvarez | it will have "#! no" | 15:56 |
| richard_maw | eww | 15:56 |
| pedroalvarez | :) | 15:57 |
| * richard_maw wonders why the patch from eglibc wasn't taken | 15:57 | |
| richard_maw | on the other hand, I can't see why it would need ldd to build | 15:57 |
| richard_maw | the only script I can find in the linux repo that calls ldd is scripts/ver_linux | 15:58 |
| richard_maw | and that's only referenced by the documentation | 15:58 |
| richard_maw | pedroalvarez: can you send a paste of the output of the linux-api-headers build that fails? | 15:59 |
| pedroalvarez | I think I'm failing to explain what the problem is :( | 15:59 |
| *** zoli_ [~zoli_@0x5e91887a.adsl.cybercity.dk] has joined #baserock | 16:01 | |
| *** zoli_ [~zoli_@0x5e91887a.adsl.cybercity.dk] has quit [Changing host] | 16:01 | |
| *** zoli_ [~zoli_@linaro/zoli] has joined #baserock | 16:01 | |
| pedroalvarez | glibc when building needs bash present. It uses the path of bash to replace some placeholders, and if it's not present it will put "#! no" in some scripts | 16:02 |
| pedroalvarez | so i decided to add a stage2-bash, so glibc depends on it, and then it can find bash and do the "right thing" | 16:02 |
| ssam2 | pedroalvarez: would a stage2-fake-bash chunk achieve the same thing ? | 16:03 |
| pedroalvarez | this works ok when building a system, but the problem is when doing a cross-bootstrap | 16:03 |
| pedroalvarez | ssam2: it could | 16:03 |
| rjek | .win 50 | 16:04 |
| wikicat | rjek: Error: "win" is not a valid command. | 16:04 |
| pedroalvarez | because actually glibc doesn't use it, only checks if it's there | 16:04 |
| rjek | GAH GAH GAH | 16:04 |
| rjek | wikicat: Silence! | 16:04 |
| wikicat | rjek: Error: "Silence!" is not a valid command. | 16:04 |
| pedroalvarez | ssam2: when you mean stage2-fake-bash you mean rjek's trick to call busybox sh? | 16:11 |
| rjek | exec /bin/sh $@ or whatever it was. | 16:11 |
| Kinnison | "$@" | 16:11 |
| rjek | yes | 16:12 |
| pedroalvarez | i'll continue using that | 16:13 |
| richard_maw | printf '#!/bin/sh\nexec /bin/sh "$@"\n' | install -D /proc/self/fd/0 "$DESTDIR/bin/bash" | 16:13 |
| rjek | eugh | 16:14 |
| Kinnison | well done richard_maw | 16:14 |
| Kinnison | that's hideous, and clever | 16:14 |
| Kinnison | but missing a -m0755 | 16:14 |
| richard_maw | nah, that's the default | 16:14 |
| Kinnison | it is? eww | 16:14 |
| * richard_maw wonders if he can get inline chunk morphologies and null source repositories done in a day | 16:20 | |
| richard_maw | it'd be a nicer way of adding the fake-bash chunk | 16:21 |
| persia | Doesn't the default depend on current ulimit (022 for 755), rather than being constantly 755? | 16:21 |
| richard_maw | busybox install always chmods to 0755, and there's a comment here saying it needs to do this for GNU coreutils 6.10 compatibility | 16:23 |
| persia | Ah, busybox install. | 16:25 |
| * persia should probably get familiar with busybox implementations rather than just being randomly annoyed by them someday | 16:26 | |
| richard_maw | a quick reading of the GNU coreutils source code leads me to believe that the install from that always chmods too, rather than obeying umask | 16:33 |
| *** tiagogomes [~tiagogome@82-70-136-246.dsl.in-addr.zen.co.uk] has quit [Quit: Leaving] | 16:33 | |
| *** mSher [~mike@82-70-136-246.dsl.in-addr.zen.co.uk] has quit [Quit: Leaving] | 16:45 | |
| *** thecorconian [~thecorcon@136.1.1.102] has joined #baserock | 16:47 | |
| *** thecorconian [~thecorcon@136.1.1.102] has quit [] | 16:48 | |
| ssam2 | surely 'ln -s /bin/sh /bin/bash' is a simpler way of doing it | 16:48 |
| rjek | No. | 16:48 |
| Kinnison | No | 16:48 |
| ssam2 | perhaps not for Busybox | 16:48 |
| ssam2 | why not? | 16:48 |
| Kinnison | busybox | 16:48 |
| rjek | Because /bin/sh is busybox | 16:48 |
| ssam2 | I guess so | 16:49 |
| rjek | And busybox will go "eh, I don't know what bash is" | 16:49 |
| rjek | (it uses what you execute it as to decide how to behave) | 16:49 |
| Kinnison | pedroalvarez: to meet the glibc needs you might need to set and export BASH=/bin/bash in your bash shim script | 16:49 |
| rjek | Perhaps we could have an ugly hack where we change busybox to accept "bash" as a synonym for "sh" | 16:49 |
| *** ssam2 [~ssam2@82-70-136-246.dsl.in-addr.zen.co.uk] has quit [Quit: Leaving] | 16:52 | |
| jmacs | That sounds worse than making /bin/bash a script that calls sh | 16:52 |
| robtaylor | Kinnison: richard_maw: https://github.com/robtaylor/qemu-wayland-runner | 16:53 |
| pedroalvarez | Kinnison: true | 16:53 |
| robtaylor | ~ I wanna know what bash is ~ | 16:54 |
| robtaylor | ~ I want you to show meeeee ~ | 16:54 |
| * rjek looks at robtaylor. | 16:54 | |
| pedroalvarez | hahah | 16:54 |
| * rjek narrows eyes. | 16:54 | |
| Zara | robtaylor: that had just started playing in my head too; really don't know why | 16:55 |
| robtaylor | Zara: its all rjek's fault | 16:57 |
| rjek | It often is. | 16:57 |
| richard_maw | robtaylor: interesting, and does it require anything special in the userland? | 16:57 |
| richard_maw | and the kernel too for that matter | 16:58 |
| robtaylor | richard_maw: only on the guest side | 16:58 |
| richard_maw | that's the side I was pondering, so that's a yes. | 16:59 |
| richard_maw | if the work goes upstream, I'd be very happy to see it in Baserock | 17:00 |
| persia | Is it not in upstream qemu? | 17:00 |
| robtaylor | i doubt itll go upstream without effort | 17:00 |
| robtaylor | these guys seem to be in their own little world | 17:00 |
| robtaylor | half of the links on https://wiki.tizen.org/wiki/Emulator require a login | 17:01 |
| robtaylor | https://lists.tizen.org/pipermail/dev/2014-March/002118.html | 17:01 |
| *** jonathanmaw [~jonathanm@82-70-136-246.dsl.in-addr.zen.co.uk] has quit [Quit: Leaving] | 17:02 | |
| persia | Hrm. | 17:02 |
| radiofree | robtaylor: i suppose there's nothing special about the version of weston their using there? | 17:06 |
| radiofree | is it just weston with the drm backend? | 17:06 |
| radiofree | s/their/they're | 17:06 |
| robtaylor | radiofree: i beleive so | 17:07 |
| * robtaylor adds some reading to his README | 17:07 | |
| radiofree | i suppose there's some driver in the rootfs? | 17:07 |
| robtaylor | there's a kernel vigs drm driver, i think | 17:08 |
| robtaylor | and then some yagl gl libraries | 17:08 |
| robtaylor | radiofree: have a look at the readme in https://github.com/robtaylor/qemu-wayland-runner now | 17:09 |
| robtaylor | https://review.tizen.org/git/?p=sdk/emulator-yagl.git;a=tree;h=8f237e3be1d7f1376ba37b20bfe95e5a622f2729;hb=8f237e3be1d7f1376ba37b20bfe95e5a622f2729 | 17:09 |
| robtaylor | one commit, with everything in | 17:09 |
| persia | heh | 17:10 |
| radiofree | :) | 17:10 |
| robtaylor | but still, this emulator they are putting together has some nice things going on | 17:10 |
| *** dutch [~william@82-70-136-246.dsl.in-addr.zen.co.uk] has quit [Quit: Quit] | 17:11 | |
| robtaylor | does stuff like androvm with a daemon on the guest for controlling simulated hardware | 17:11 |
| robtaylor | (lights, geo, compass, etc) | 17:11 |
| radiofree | need some patched libdrm as well? * libdrm - VIGS has it's own part in libdrm | 17:11 |
| radiofree | https://lists.tizen.org/pipermail/tizen-submit/2014-April/035488.html | 17:12 |
| *** CTtpollard [~tom@82-70-136-246.dsl.in-addr.zen.co.uk] has quit [Quit: Ex-Chat] | 17:12 | |
| robtaylor | radiofree: seems like it | 17:13 |
| radiofree | i can never unwind yocto recipes, no idea where that git repo is located | 17:14 |
| radiofree | platform/upstream/libdrm.git apparently | 17:14 |
| robtaylor | radiofree: https://review.tizen.org/git/?p=platform/upstream/libdrm.git;a=summary | 17:14 |
| radiofree | anyway it's pretty cool, what's the performance like | 17:14 |
| robtaylor | yup | 17:14 |
| radiofree | nice! | 17:14 |
| robtaylor | give it a try! | 17:15 |
| robtaylor | it seems ok, though mouse pointer feels sluggish, that could just be accerator settings | 17:15 |
| pedroalvarez | before I leave: glibc build failure with fake-bash | 17:23 |
| pedroalvarez | http://pastebin.com/jqJgNF87 | 17:24 |
| pedroalvarez | (this was doing cross-bootstrap, I'm testing now with a normal build | 17:24 |
| pedroalvarez | ) | 17:24 |
| *** pdar [~patrickda@access.ducie-dc1.codethink.co.uk] has quit [Ping timeout: 245 seconds] | 17:34 | |
| pedroalvarez | it wokrs... -.-' | 17:35 |
| *** ridgerunner [~robjones@access.ducie-dc1.codethink.co.uk] has quit [Remote host closed the connection] | 17:38 | |
| *** jmacs [~jimmacart@access.ducie-dc1.codethink.co.uk] has quit [Remote host closed the connection] | 17:38 | |
| *** mwilliams_ct [~mikewilli@access.ducie-dc1.codethink.co.uk] has quit [Read error: Connection reset by peer] | 17:38 | |
| *** benbrown_ [~benbrown@access.ducie-dc1.codethink.co.uk] has quit [Read error: Connection reset by peer] | 17:38 | |
| *** SotK [~adamcoldr@access.ducie-dc1.codethink.co.uk] has quit [Read error: Connection reset by peer] | 17:38 | |
| *** Zara [~zarazaime@access.ducie-dc1.codethink.co.uk] has quit [Read error: Connection reset by peer] | 17:38 | |
| *** paulsherwood [~paulsherw@access.ducie-dc1.codethink.co.uk] has quit [Read error: Connection reset by peer] | 17:38 | |
| *** petefoth [~petefothe@access.ducie-dc1.codethink.co.uk] has quit [Read error: Connection reset by peer] | 17:38 | |
| persia | So it's about the differences of cross-bootstrap that breaks things. Annoying that, but good to understand. | 17:38 |
| *** ridgerunner [~robjones@access.ducie-dc1.codethink.co.uk] has joined #baserock | 18:07 | |
| *** zoli_ [~zoli_@linaro/zoli] has quit [Read error: Connection reset by peer] | 18:37 | |
| *** juergbi [~juerg@vserver.paldo.org] has quit [Quit: Coyote finally caught me] | 20:03 | |
| *** juergbi [~juerg@vserver.paldo.org] has joined #baserock | 20:06 | |
| pedroalvarez | erm.. actually I think I was being dumb | 21:26 |
| pedroalvarez | I believe I have to mount some partitions (dev proc sys and tmp) before running the native building of cross-bootstrap | 21:28 |
| pedroalvarez | after doing that... it works :) | 21:44 |
| pedroalvarez | with fake-bash | 21:46 |
| pedroalvarez | as a result, I should update this page: http://wiki.baserock.org/guides/how-to-cross-bootstrap/ | 21:50 |
| pedroalvarez | because some things are missing on the point 3 | 21:51 |
| *** genii [~quassel@ubuntu/member/genii] has quit [Remote host closed the connection] | 23:23 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!