[Piuparts-devel] what's stuck in my local trees?

[Dave Steele]

On Sat, May 26, 2012 at 12:17 PM, Andreas Beckmann <debian at abeckmann.de> wrote:
> Hi,
>
> this is piuparts stuff I'm currently working on and have lots of dirty
> commits in my local trees:
>


I tend to use github for backup - there is dirty laundry there:


multiple-slaves - running

Patches to slave-run and slave-join for parallel slave runs and
monitoring. I tend to run from this.


config-defaults - works
retest-manager - probably doesn't compile
master-idle - untested

These three, in aggregate, form the basis for my proposed idle/retest
solution. This shows promise - no protocol changes are required. Note
that it removes many of your slave sleep calls, to push the idles to
master, allowing the package retest condition to be detected there.

config-defaults allows section Config variables to be defaulted in the
global section. It is not strictly required here, but allows for much
more concise section definitions.


slave_state - doesn't compile yet

This allows the slave to report operating state. It is on the path for
implementing piupartsd as a true init.d service. This is the reason
that commit 49449e79 considers multiple slaves in one section tree a
failure condition.

    https://github.com/davesteele/piuparts/commit/49449e79bf50ae48f39ed672cd6e18736e84ed0c


piuparts-server
rrdeps
leftover-dir
report-performance

These are submitted.


The rest are not active, and will be deleted when associated issues
are resolved.