[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Consuming excuses.yaml?



On 28/12/18 at 21:04 +0000, Niels Thykier wrote:
> Cyril Brulebois:
> > Hey folks,
> > 
> > I haven't been following release team things lately, so I might have
> > missed something: do we have a tool that parses/consumes excuses.yaml
> > yet? A few years back, I had a tool look at update_excuses.html and
> > generate graphs/clusters of packages failing to migrate, with their
> > close friends.
> > 
> > Digging into puppetdb manually[1] made me think about possibly porting
> > it to use excuses.yaml instead, but maybe something else was developed
> > in the meanwhile?
> > 
> >  1. https://twitter.com/CyrilBrulebois/status/1078514866397286401
> > 
> > Maybe it would make sense to generate those graphs and include them
> > somewhere (release.d.o or tracker.d.o) to make it easier for people
> > to figure out what packages are involved in a given migration, instead
> > of having to click through possibly many packages?
> > 
> > (cc's welcome, thanks)
> > 
> > 
> > Cheers,
> > 
> 
> Hi KiBi,
> 
> I think that would be a splendid idea to have such a tool. :)
> 
> I am tempted to say it should live in the britney2 repo, so we can
> include it in our test cases and remember to update it if/when the
> format changes. :)
> 
> I am less sure of a place to host the generated graphs (we can host them
> on release.d.o if the size/bandwidth requirements are low, but it would
> surprise me with the scale of packages in sid vs. testing)

Hi,

You could also generate the graphs dynamically, similarly to what is
done for popcon graphs (script running on
https://qa.debian.org/popcon.php?package=foo). If that helps, I'm now
importing the excuses.yaml data into UDD, so you could read it from
there instead of parsing the whole file (but parsing the whole file is
quite fast anyway). Or this could be a UDD CGI script that takes the
package name as parameter.

Lucas

Attachment: signature.asc
Description: PGP signature


Reply to: