[petsc-dev] petsc-dev on bitbucket

[Matthew Knepley]

On Fri, Feb 10, 2012 at 10:08 AM, Satish Balay <balay at mcs.anl.gov> wrote:

> On Fri, 10 Feb 2012, Matthew Knepley wrote:
>
> > > until we move to bitbucket - can you continue to push to petsc.cs.iit
> > > - and not bitbucket? [for now - pushes to petsc.cs.iit are
> > > automatically pushed to bitbucket]
> > >
> >
> > I thought we were switched already :) I cannot understand why we would
> not
> > give up this maintenance burden (and directories is not a good enough
> > reason).
>
> For my part - I generally don't like the way the change is pushed.
>
> - one prblem => suggest change to a new system [not evaluating all the
> features we currently have vs what we loose]
>

I thought I did that in several mails. What do you think I have no
addressed?


> - and I really get annoyed when a problem is pointed out with the new
> sytem - and the solution to that is: the current model is workng - but
> antique - lets change it to fix for this new system we are pusing
>

By this you mean not having directories? I agree that is a shortcoming. My
explicit
comment was that this is far outweighed by the positives. Everything is a
balance,
and here I believe the balance is clearly on the side of BB. You disagree
on this
I see.


> Wrt current admin overhead - its not that much. Most of the work is
> adding new users [not modifying user keys]. This part of admin work
> will is exist with bitbucket.
>

Every time the user changes keys, we have to do something. I cannot believe
Barry
is not screaming about this maintenance nightmare.


> The other systrem admin part is not much. Jsut reboot the machine when
> there is a kernel update.
>

We clearly had this situation this week, and "just reboot" sucks pretty
badly.


> Also the reliability aspect of such sites is also verblown. I'm sure
> some of them had their outages. [maybe not bit bucket - but some of
> them do].


Out outages are more frequent. There is no need to exaggerate. However, I
listed
several improvements in maintenance in addition to uptime.

    Matt


>
> Satish
>



-- 
What most experimenters take for granted before they begin their
experiments is infinitely more interesting than any results to which their
experiments lead.
-- Norbert Wiener
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-dev/attachments/20120210/a68a4ecf/attachment.html>