| *** gtristan has quit IRC | 05:15 | |
| *** ctbruce has joined #baserock | 07:40 | |
| *** toscalix has joined #baserock | 07:54 | |
| *** toscalix has quit IRC | 08:06 | |
| *** toscalix has joined #baserock | 08:07 | |
| *** toscalix has quit IRC | 08:09 | |
| *** toscalix has joined #baserock | 08:12 | |
| *** paulwaters_ has joined #baserock | 08:13 | |
| *** toscalix has quit IRC | 08:16 | |
| *** jonathanmaw has joined #baserock | 08:19 | |
| *** toscalix has joined #baserock | 08:20 | |
| *** toscalix has quit IRC | 08:22 | |
| *** toscalix has joined #baserock | 08:42 | |
| *** toscalix has quit IRC | 08:52 | |
| *** tiagogomes has quit IRC | 08:52 | |
| *** tiagogomes has joined #baserock | 08:54 | |
| jonathanmaw | jjardon, paulsherwood: can either of you review https://gitlab.com/baserock/ybd/merge_requests/329 ? | 08:58 |
|---|---|---|
| *** CTtpollard has joined #baserock | 09:05 | |
| paulsherwood | jonathanmaw: done | 09:09 |
| jonathanmaw | paulsherwood: I don't suppose you can give some insight into why we use run_sandbox_with_redirection()? | 09:11 |
| paulsherwood | let me see if i can remember | 09:11 |
| jonathanmaw | it looks like ssam was responsible for it originally (commit 66cf5ba0b20ec4f6e3c1d99e1b420b455203cef7) | 09:12 |
| paulsherwood | jonathanmaw: almost certainly ssam2 will know better... | 09:12 |
| *** ssam2 has joined #baserock | 09:13 | |
| *** ChanServ sets mode: +v ssam2 | 09:13 | |
| paulsherwood | jonathanmaw: is your question 'why don't we just run run_sandbox' ? | 09:14 |
| *** toscalix has joined #baserock | 09:14 | |
| jonathanmaw | paulsherwood: yep, why not run_sandboxed | 09:15 |
| jonathanmaw | ssam2: you implemented that code originally. Can you remember the decision behind using run_sandbox_with_redirection()? | 09:15 |
| ssam2 | hmm | 09:16 |
| ssam2 | you mean the name or the functionality ? | 09:16 |
| jonathanmaw | ssam2: the functionality | 09:16 |
| ssam2 | I think so that logs are "live" | 09:16 |
| jonathanmaw | ssam2: what does that mean? | 09:17 |
| ssam2 | it's annoying when the command inside the sandbox gives no output for hours, then at the end of the build dumps the whole log to dis | 09:17 |
| ssam2 | *disk | 09:17 |
| ssam2 | it's better when the log is written "as it happens" | 09:17 |
| ssam2 | so it may have been to solve that problem | 09:17 |
| jonathanmaw | ssam2: I'm not sure how that works. looking at the code (sandboxlib/sandboxlib/chroot.py), run_sandbox_with_redirection just calls run_sandbox and discards out and err | 09:18 |
| ssam2 | that's because out and err are written to a file | 09:19 |
| ssam2 | it expects you to pass in file handles for stdout and stderr | 09:19 |
| jonathanmaw | stdout and stderr passed in as arguments to run_sandbox | 09:19 |
| ssam2 | yeah | 09:20 |
| ssam2 | it seems pretty pointless, looking back at the code now | 09:20 |
| jonathanmaw | ok, so that's some confidence that this change isn't breaking things in strange ways, then | 09:20 |
| ssam2 | I guess the idea of having a function with a different name was to make it clear when a call to run_command() wouldn't return stdout or stderr | 09:20 |
| ssam2 | it's up to the user of the library to keep it that way | 09:21 |
| ssam2 | if that's the case I never documented it in __init__.py, but that's not impossible | 09:21 |
| ssam2 | but yeah, it was all about whether stdout & stderr are written to a pipe in memory, or written to disk | 09:22 |
| ssam2 | and ideally it would support doing *both*, like Morph did, but I never got time to implement that | 09:22 |
| ssam2 | I think morph did it just by modifying the commandline though, which wasn't reusable | 09:22 |
| *** toscalix has quit IRC | 09:35 | |
| *** locallycompact has joined #baserock | 09:57 | |
| *** toscalix has joined #baserock | 10:41 | |
| *** toscalix_ has joined #baserock | 12:58 | |
| *** toscalix has quit IRC | 13:00 | |
| *** toscalix_ has quit IRC | 14:10 | |
| *** paulwaters_ has joined #baserock | 15:02 | |
| *** edcragg has joined #baserock | 15:41 | |
| *** ctbruce has quit IRC | 16:16 | |
| jjardon | paulsherwood: seems ybd managed to build systems/armv7lhf-cross-toolchain-system-x86_64.morph? https://gitlab.com/baserock/definitions/builds/13062773 | 16:16 |
| *** jude_ has quit IRC | 16:17 | |
| paulsherwood | that's a good thing, isn't it? :) | 16:17 |
| *** jude_ has joined #baserock | 16:43 | |
| *** jonathanmaw has quit IRC | 17:16 | |
| *** locallycompact has quit IRC | 17:39 | |
| *** ssam2 has quit IRC | 18:07 | |
| *** jude_ has quit IRC | 18:13 | |
| *** lchlan has quit IRC | 19:28 | |
| *** lchlan has joined #baserock | 19:29 | |
| *** locallycompact has joined #baserock | 19:51 | |
| *** jude_ has joined #baserock | 20:47 | |
| *** locallycompact has quit IRC | 21:10 | |
| *** jude_ has quit IRC | 21:41 | |
| *** locallycompact has joined #baserock | 22:02 | |
| *** locallycompact has quit IRC | 22:09 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!