Re: [AMBER-Developers] more on patch_amber.py

From: Jason Swails <jason.swails.gmail.com>
Date: Fri, 10 Feb 2012 14:01:24 -0500

I've checked in changes to patch_amber.py. interrupt signals (SIGINT) are
now returned as -1, which does not register as an unapplied patch (rather
than 1, which did).

I've now added a timeout when trying to connect to http://ambermd.org to
determine internet access, so in case the website is down, it will only
wait 10 seconds to tell you so (this timeout is tunable through a CL flag
to patch_amber.py, but we should set the default to something intelligent).

I also indicated how to use the program "yes" to automate a workflow with
the top-level configure/Makefile to always apply patches (or never apply
them, if that's what you want).

All the best,
Jason

On Wed, Feb 8, 2012 at 8:23 PM, Scott Brozell <sbrozell.rci.rutgers.edu>wrote:

> Hi,
>
> Perhaps off base, but when i ran configure from a fresh
> git master last night, ie, 3am est today, configure also spewed
> some stuff about 1 un-applied patch.
>
> So should patch more elegantly handle this (no need to patch the repo)
> or
> ...
> lots of possible or's omitted for my convenience.
>
> scott
>
> On Wed, Feb 08, 2012 at 06:52:58PM -0500, Jason Swails wrote:
> > Yea, I sent you an email about this privately. Patch_amber.py handles
> lack of Internet just fine. What it doesn't handle is the website being
> down, since it isn't being connected even though it has Internet access.
> I'll look into the API some to see if I can trigger some type of timeout
> behavior.
> >
> > The reason it's registering an unapplied patch is because it uses the
> return code to determine the number of unapplied patches available. The
> ctrl-c killed the script with an error code of 1, which is why you got that
> result.
> >
> > I'll put something in to catch a sigint so this doesn't happen.
> >
> > On Feb 8, 2012, at 5:39 PM, "David A. Case" <case.biomaps.rutgers.edu>
> wrote:
> > > Just to note that casegroup.rutgers.edu is down right now; but we
> should
> > > make the configuration procedure robust to such events.
>
> _______________________________________________
> AMBER-Developers mailing list
> AMBER-Developers.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber-developers
>



-- 
Jason M. Swails
Quantum Theory Project,
University of Florida
Ph.D. Candidate
352-392-4032
_______________________________________________
AMBER-Developers mailing list
AMBER-Developers.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber-developers
Received on Fri Feb 10 2012 - 11:30:03 PST
Custom Search