[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Condor-devel] github revisited.
- Date: Thu, 13 Sep 2012 14:35:32 -0500
- From: Zachary Miller <zmiller@xxxxxxxxxxx>
- Subject: Re: [Condor-devel] github revisited.
On Thu, Sep 13, 2012 at 02:48:18PM -0400, Tim St Clair wrote:
> Ok seriously... what needs to be done to make this transition?
someone needs to clearly lay out the positives and negatives of switching
so we have a well-thought out plan for switching and good, concrete reasons
for doing so.
i'm definitely not against it. but "i got frustrated with AFS(*)" alone
doesn't sell me at all, nor does any amount of andecdotal complaints without a
plan as stated above. saying "has feature X and Y" or "makes Z easier" would
go a lot further.
i know almost nothing about github, so like i said, i'm not for or against
going there. i just would like to see all the pros/cons gathered together in
one place so we can discuss how it will make our lives better and have an
informed discussion about whether or not to make the switch. right now it
feels like just a handful of people are cheerleading for moving (and they may
very well be right) but that we (we being UW-flightworthy) can't make that
decision without understanding why we should switch.
(or maybe i'm the only one and i'll just shut up :)
so: anyone volunteering to write up said document?
cheers,
-zach
(*) which by the way can be worked around
> * tstclair channels willb... "If only..." there were a utility like github which didn't have AFS & space constraints... wait.
> <willb_> heh
> <mattf> error: insufficient permission for adding an object to repository database ./objects
> <mattf> fatal: failed to write object
> <mattf> error: unpack failed: unpack-objects abnormal exit
> <mattf> To ssh://pinguino.cs.wisc.edu/p/condor/repository/CONDOR_SRC.git
> <mattf> Greg, ^^
> <mattf> tstclair, Greg, willb, anyone know if gittrac can track github?
> <tstclair> I don't see why not it's just loc.
> <tstclair> url
> <mattf> umm, wow -- https://github.com/search?q=condor&type=Repositories
> <mattf> 101 repositories
> <mattf> https://github.com/gilesc/lein-condor
>
> Cheers,
> Tim
>
>
> ----- Original Message -----
> > From: "Erik Erlandson" <eje@xxxxxxxxxx>
> > To: "Alan De Smet" <adesmet@xxxxxxxxxxx>
> > Cc: "Tim St Clair" <tstclair@xxxxxxxxxx>, "Condor Developers" <condor-devel@xxxxxxxxxxx>
> > Sent: Friday, July 27, 2012 6:22:28 PM
> > Subject: Re: [Condor-devel] github revisited.
> >
> > On Fri, 2012-07-27 at 16:48 -0500, Alan De Smet wrote:
> > > Timothy St. Clair <tstclair@xxxxxxxxxx> wrote:
> > > > To be more direct, "What are the limiters/detractors to moving
> > > > the code base to github?".
> > >
> > > I'd like it, but I just like GitHub. But as far as I know it's
> > > incompatible with our security policy of embargoing code. If we
> > > keep such changes local only it means that the local repo is the
> > > "real" one and there isn't a lot of incentive to to do the extra
> > > work to push some, but not all, things up to GitHub.
> >
> >
> > I'm not sure what embargoing code refers to, but a public repo on
> > github
> > still allows the repo owner to grant particular users commit privs,
> > which seems like the crucial thing. Anybody can read the repo, fork
> > it,
> > make pull requests. But only users authorized by the owner can
> > commit.
> >
> >
> >
> _______________________________________________
> Condor-devel mailing list
> Condor-devel@xxxxxxxxxxx
> https://lists.cs.wisc.edu/mailman/listinfo/condor-devel