[opam-devel] removing opam 1.0 from the travis test matrix

Thomas Gazagnaire thomas.gazagnaire at gmail.com
Tue Nov 19 16:30:43 GMT 2013


> Note that the data shown on opam.ocaml.org is agregated, and reports accesses to both opam.ocaml.org and opam.ocamlpro.com. That is, until the last time I rsync'd the stats, which needs to be done manually since I didn't bother setting up some secure pipe between the servers.

I think this is a problem as this means we have regular stat bumps (ie. ~1200 yesterday and ~2100 today) which looks like a bit weird (as the only meaningful semantics we can attach to these stats is their evolution).
>  
> I've just added a run of opam2web on the ocamlpro.com logs only at http://opam.ocaml.org/1.0/, for easy comparison ; it's a quick hack but enables to compare stats.
> Keep in mind that the latter is only updated when I manually rsync the logs (last time, just now)

seems that the CSS is missing.

> Hope this helps ; does anyone have a better idea ?

Would be nice to expose anonymous stats at some points ...

>  
> Louis
>  
> Le lundi 18 novembre 2013 09:15:16 Anil Madhavapeddy a écrit :
> > The opam.OCamlPro.org server will have this information. It would be most useful to keep an eye on this and ensure that we migrate people to the new repository. Louis, I think you're the only one with access...?
> >
> > -anil
> >
> > On 18 Nov 2013, at 09:12, Amir Chaudhry <amc79 at cam.ac.uk> wrote:
> >
> > > Just curiosity (and more of a side note, so feel free to ignore): Is there some way to check how may people may be on OPAM 1.0? I'm assuming the logs may be able to tell how frequently `opam update` gets used.
> > >
> > > Amir
> > >
> > > On 18 Nov 2013, at 16:53, Anil Madhavapeddy <anil at recoil.org> wrote:
> > >
> > >> Now that the OPAM 1.1 repository has split, are there any objections to me removing 1.0 from the Travis testing matrix for the 1.1 repository (i.e. github.com/ocaml/opam-repository) ?
> > >>
> > >> I plan to rotate in 1.2 Debs to the live test matrix just as soon as the repository diverges enough to make such testing worthwhile, so we need to retire 1.0 sooner rather than later.
> > >>
> > >> My only point of concern is to ensure that Debian has an OPAM 1.1 package update soon, as otherwise users of Debian Testing will only see the old package set.
> > >>
> > >> -anil
> > >> _______________________________________________
> > >> opam-devel mailing list
> > >> opam-devel at lists.ocaml.org
> > >> http://lists.ocaml.org/listinfo/opam-devel
> > >
> >
> > _______________________________________________
> > opam-devel mailing list
> > opam-devel at lists.ocaml.org
> > http://lists.ocaml.org/listinfo/opam-devel
> _______________________________________________
> opam-devel mailing list
> opam-devel at lists.ocaml.org
> http://lists.ocaml.org/listinfo/opam-devel

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.ocaml.org/pipermail/opam-devel/attachments/20131119/ade7d4c9/attachment.html>


More information about the opam-devel mailing list