[petsc-dev] Fwd: [petsc-maint #119133] petsc-dev configure crash

[Satish Balay]

On Thu, 7 Jun 2012, Jed Brown wrote:

> On Thu, Jun 7, 2012 at 3:30 PM, Satish Balay <balay at mcs.anl.gov> wrote:
> 
> > >  subrepository sources for config/BuildSystem differ
> > > use (l)ocal source (eadaa286b5ca) or (r)emote source (6ab6f1ae4aff)?
> >
> > Ok  - I have to choose 'r' here for what I want [wrt diff/auto-merge].
> >
> > [balay at maverick petsc-buildsystem-dev]$ hg stat -S
> > M .hgsubstate
> > M config/BuildSystem/config/packages/BlasLapack.py
> > M config/examples/asterix/asterix64.py
> > [balay at maverick petsc-buildsystem-dev]$
> >
> > But for some reason '.hgsubstate' is not updated [to correspond to 'hg
> > pull -u' version].
> 
> 
> Had you changed the .hgsub?

Nope.

> In any case, the right solution when there is a conflict in BuildSystem is
> always to go into BuildSystem and resolve it (by merging or rebasing), then
> commit the result to the petsc repository.

This is a usage without commits - only local changes/edits - and then 'pull -u'

> As for the (now long past) request for implicitly tracking the latest
> versions of subrepos, this quote from Matt Mackall is apropos.
> 
> *We're always telling people "no, subrepos are NOT for automatically
> getting the latest versions of other projects, that's not Version Control™"*

Sure - but I'm trying to understand all the extra control flow paths
that come up with subrepos. Also the above case doesn't fall in the
category of 'automatically getting the latest version of other
projects'

Satish