On 07/09/2013 03:03 PM, Todd Lyons wrote: > 2. My thoughts were a little less extravagant. I was a member of the
> Samba build farm for a few years. Participants' machines synced to
> the master samba buildfarm server (multiple repos), built multiple
> configs when new commits were found, ran samba torture (essentially
> make test), and results were submitted back up to a master samba
> buildfarm server, where it was parsed and made available for
> display/analysis. I'd like to consider doing the same for Exim.
> There are a couple of public projects that use a distributed build
> farm, and we could simply model off of their systems.
>
> The two ideas aren't mutually exclusive, but they do have common
> portions which make it somewhat duplicative to do both. Comments,
> feedback?
The Samba build farm sounds somewhat like the Postgresql one:
