LocutusOfBorg | s390x is on strike again | 07:27 |
---|---|---|
LocutusOfBorg | s390x is on fire! | 09:02 |
cjwatson | LocutusOfBorg: back now | 11:41 |
=== cpaelzer_ is now known as cpaelzer | ||
realtime-neil | I'm looking for a tool that will produce a text dump of everything in a given team. Does such a thing exist? If so, what is it and where can I get it? | 14:02 |
cjwatson | that's a very fuzzy concept really. can you narrow it down at all? | 14:03 |
cjwatson | (the answer is probably not as a single thing, but it might be possible to help with certain aspects of it) | 14:03 |
realtime-neil | okay, consider the scope narrowed to just PPAs | 14:08 |
cjwatson | realtime-neil: I think the answer is going to be that there's no such tool but it's relatively easy to build whatever you need using the API here - LP does enough different things that we intentionally prefer to build mechanisms that people can use to do what they need rather than prepackaged tools. For example `team.ppas` (in launchpadlib, assuming that `team` is an object for a team that you've | 14:28 |
cjwatson | already fetched) will give you a collection of all that team's PPAs, and for each PPA you can use e.g. https://launchpad.net/+apidoc/devel.html#archive-getPublishedSources and https://launchpad.net/+apidoc/devel.html#archive-getPublishedBinaries to see what's published in it | 14:28 |
realtime-neil | Okay, thanks. I'm not much for python, but give it a go. | 14:29 |
realtime-neil | sed 's/give it a go/I will give it a go/' | 14:30 |
cjwatson | It's not mandatory to use Python to use the Launchpad API, but having bindings is useful. | 14:33 |
cjwatson | Saves dealing with all the encoding/decoding etc. | 14:34 |
realtime-neil | I'd be a fool to ignore the body of work that already exists. | 14:34 |
LocutusOfBorg | realtime-neil, https://xkcd.com/353/ | 16:13 |
jugmac00 | realtime-neil: This is a starter: https://gist.github.com/jugmac00/a77af7e4554de3042440fda8cfc10ce9 | 17:11 |
realtime-neil | thanks, jugmac00 | 17:12 |
cjwatson | (passing cachedir is unnecessary BTW, and the script is more naturally reusable by others if you leave that out) | 17:15 |
jugmac00 | cjwatson: thanks, updated | 17:21 |
sarnold | unfortunately, a few hundred comics later, https://xkcd.com/353/ | 18:33 |
realtime-neil | sarnold, I think you pasted the wrong link | 18:52 |
sarnold | ARGH | 18:54 |
sarnold | https://xkcd.com/1987/ | 18:54 |
sarnold | apparently I still haven't adapted to firefox's new urlbar select behaviour :( | 18:54 |
realtime-neil | That's always the one I remember when someone mentions `pip`. | 18:54 |
realtime-neil | This is happening: https://code.launchpad.net/~realtime-robotics/+archive/ubuntu/builds/+recipebuild/3397823 dependency wait on `libocct-foundation-dev (< 7.6)`; the target PPA "builds" is already depending on the PPA "freecad" providing the needed packages. https://launchpad.net/~realtime-robotics/+archive/ubuntu/builds https://launchpad.net/~realtime-robotics/+archive/ubuntu/freecad | 19:49 |
realtime-neil | I've run `mark-suite-dirty --suite=focal` on both PPAs. | 19:50 |
realtime-neil | I've retried the recipe. | 19:50 |
realtime-neil | Error message mentions this: "builddeps:cadexsdk : Depends: libocct-foundation-dev (< 7.6) but 1:7.5.2+dfsg1-0~202107020155~ubuntu20.04.1 is to be installed" | 19:51 |
realtime-neil | Meanwhile, the recipe for Jammy succeeds: https://code.launchpad.net/~realtime-robotics/+archive/ubuntu/builds/+build/24191755 | 20:01 |
cjwatson | realtime-neil: Well, that's true - 1:7.5.2... is less than 7.6 | 22:09 |
cjwatson | realtime-neil: Maybe you meant "(<< 1:7.6)" ? | 22:10 |
cjwatson | err is GREATER than 7.6, I meant to say | 22:10 |
cjwatson | sorry, sometimes IRC would be more comprehensible with an edit function | 22:11 |
realtime-neil | I gave up on version associations. I think I remember discovering that focal devscripts (or something) doesn't do version associations very well. | 22:47 |
cjwatson | realtime-neil: I don't understand. What's a "version association"? | 22:48 |
cjwatson | (Not a term I've heard in ~23 years of doing Debian packaging.) | 22:49 |
realtime-neil | the stuff in the parenthesis | 22:50 |
cjwatson | realtime-neil: It is EXTREMELY unlikely that focal devscripts fails to parse versioned dependencies correctly, and much more likely that you mistyped them in some way. | 22:50 |
cjwatson | realtime-neil: What you quoted above was definitely mistyped. | 22:50 |
realtime-neil | What worked on jammy failed on focal. If I mistyped something I would have expected both failures. | 22:51 |
cjwatson | Versioned dependencies are completely ubiquitous through the packaging system. They work, I promise you. | 22:53 |
cjwatson | So the problem is not that. | 22:53 |
cjwatson | So where's this 1:7.5.2+dfsg1-0~202107020155~ubuntu20.04.1 version coming from, I wonder ... | 22:54 |
realtime-neil | It _was_ in the "builds" ppa before I deleted it. | 22:55 |
cjwatson | Oh right, I see it now | 22:55 |
realtime-neil | Deleting it didn't help me. | 22:55 |
cjwatson | It will do, you may just not have waited for the deletion to complete | 22:55 |
cjwatson | That was exactly and precisely the problem | 22:55 |
realtime-neil | Ah. spooky invisible actions | 22:55 |
cjwatson | The reason for the problem is that 1:7.5.2+dfsg1-0~202107020155~ubuntu20.04.1 is not less than 7.6 - the epoch at the start means that it's greater than any version without the epoch | 22:56 |
realtime-neil | And that's fine. The desired versions are in my freecad PPA | 22:57 |
cjwatson | (epoch = 1:, 2: etc. as a prefix to the version, generally used for getting out of situations where the version numbering sequence of a package changes so that a later version doesn't actually compare later) | 22:57 |
realtime-neil | unless I'm not understanding | 22:57 |
cjwatson | Right, so it should all be just fine if you revert your Build-Depends change and let the recipe build again. | 22:58 |
realtime-neil | I might try that tomorrow. | 22:58 |
cjwatson | The deletion has completed properly now, so it won't see that 1:7.5.2... version and be confused by it. | 22:58 |
realtime-neil | thanks very much! | 22:58 |
realtime-neil | How can I tell if a deletion has completed? | 22:58 |
cjwatson | Look at https://code.launchpad.net/~realtime-robotics/+archive/ubuntu/builds/+packages and see if the package is still there | 22:59 |
cjwatson | When the deletion completes, it'll go away (unless you filter by "Any status" - the default filter is "Published") | 22:59 |
realtime-neil | Ah, okay. thanks again | 22:59 |
cjwatson | When you request a deletion, that's queued for the next publisher cycle | 22:59 |
cjwatson | Same as uploads and completed builds | 22:59 |
cjwatson | realtime-neil: Probably something that I've internalized so much by this point I don't even think that it might not be obvious, TBH! | 23:02 |
cjwatson | I wonder if we should change the page notification you get when you delete packages to briefly explain that the packages will be deleted soon, to cue people into the fact that it isn't instant. | 23:03 |
cjwatson | I think at the moment it just says "Source and binaries deleted by <name>" and a list of the affected packages | 23:04 |
sarnold | I don't hit that button often but I thought it said something to suggest it might not happen immediately, but the visual feedback also suggests that in *this* case it happened immediately | 23:24 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!