[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

Re: FESCo Meeting Summary for 2007-02-08



On Thu, 15 Feb 2007 09:00:19 -0500, Jesse Keating wrote:

> > mktemp -ud also returns a new path with every invocation. A buildroot that
> > becomes a moving target (even when overriding %_tmppath) is impractical for
> > --short-circuit builds or anything where you would want to examine the
> > buildroot contents.
> 
> Without a unique  buildroot for each build, you're back to the problem of 
> overwriting or picking something up that you didn't put there in the first 
> place. 

No, because every spec file does "rm -rf $RPM_BUILD_ROOT" at the beginning
of %install. All that is needed is making %_tmppath point to the user's
private space. And that is highly recommended for many years.

>  I do believe the patches that were sent upstream to do BuildRoot 
> internally within RPM make use of unique directories per build.

Raises the question why there is a draft on making a specific BuildRoot
mandatory?


[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]