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

Re: rawhide report: 20060210 changes (part 1/2)



On Fri, Feb 10, 2006 at 01:17:05PM -0500, Jeremy Katz wrote:
> On Fri, 2006-02-10 at 19:06 +0100, Pozsar Balazs wrote:
> > On Fri, Feb 10, 2006 at 07:57:44AM -0500, Build System wrote:
> > [...]
> > > - rebuilt for new gcc4.1 snapshot and glibc changes
> > [...]
> > 
> > I am wondering long ago about these changelog entries... Why are 
> > rebuilds mentioned in the changelogs? Strictly speaking, these are not 
> > changes, at least not to the .src.rpm.
> 
> Because otherwise, if I have foo-1.2.1-1 installed and see foo-1.2.1-2,
> how do I know what changed without diffing src.rpms?  That's what the
> entire purpose of the changelog is.

If foo-1.2.1-1.src.rpm is rebuilt, it could be versioned as 
foo-1.2.1-1.1, foo-1.2.1-1.2, foo-1.2.1-1.3 and so on.


> > And by the way, I also fail to see why rebuilds are so special. Why 
> > aren't all packages rebuilt periodically by the build system?
> > (Say, once a week or fortnight.)
> > I think it can be easily seen that it would be a nice regular qa test, 
> > and also it would make sure that all gcc/glibc or any other 
> > library/compiler toolchain element changes/improvements would be 
> > propageted in regular and short timebase.
> 
> There are more regular builds of everything for testing, but pushing
> things out just ends up causing a lot more bandwidth usage than its
> worth.  Most packages get churned enough in the development tree that
> any changes get propagated to packages quickly enough.

Is bandwidth really a problem for fedora developers/testers?


-- 
pozsy


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