incubator-general mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Dirk-Willem van Gulik <>
Subject Hardware, build farms and the like
Date Tue, 07 Jun 2011 12:03:31 GMT
So code is one thing. Open office does also come with quite a need for build farms, automated
test and so on.

It would be good to understand this early - and understand wether this matters. I.e. Can normal
development continue with the basics (svn, bug tracking, mailing lists, archives and a delivery
CDN of released products) - or is it paramount that there is more on day one ?

And secondly - is it crucial that such happens very near to the code ? Or can it happen at
a distance ? E.g. at companies or at specific community focal points ? 

Apache is all over the board - some projects just test, tag and release source tarballs, others
rely on volunteers (usually a mix of companies and individuals) to contribute compiled binaries
while others actually use ASF hardware to build these*. While others almost completely rely
on others to take their raw products and integrate it into products.

Or is there currently in the OpenOffice world a distinct lack/loss of capability associated
with this transition - and it would be important to ensure that that capability is part of
the incubator proposal ? Do we know how much hardware that is ?

And secondly - does this cover it - or are their subtle other bits of infrastructure (end
user help/document servers, core/crash-dump receivers, special DNS setups, template servers
or what not) which are needed too ? Is that documented sufficiently ?


*: And yes - we're fine with that - in apache we're process oriented - and if the PMC has
clear oversight of a well managed process to get zips and exe's 'released' - then they can
use whatever appropriate magic.
View raw message