incubator-general mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Michael Meeks <>
Subject Re: Proposal for OpenOffice Incubator strategy
Date Fri, 03 Jun 2011 14:05:31 GMT
Hi Rob,

On Thu, 2011-06-02 at 21:26 -0400, wrote:
> Finally, I think we're exaggerating the difficulty of getting out a 
> release of OpenOfice.  LibreOffice did it very quickly.  And so did IBM 
> with Symphony.  This is not rocket science. 

	I am impressed by your optimism. Let us see how quickly you personally
manage a windows build yourself of what ends up inside the initial
Apache code-base (incidentally, I'm eagerly awaiting that myself, when
do we see it ? only after acceptance of the podling?). We could even
have a small race :-)

> As for infrastructure, we are blessed with an amazing Apache 
> Infrastructure Team.  I have full confidence in their capabilities.

	Last I looked OO.o 3.3.1 is 70+Gb of binary data to compile and
distribute to a mirror infrastructure. Shipping source is quite easy,
binaries are more troublesome to both create, and distribute. No doubt
it can be done - but it is less than trivial. I suspect that Symphony
does not cater for the diversity of locales that the community will
expect (as one data point).

> As for continuity of OpenOffice releases, there was a full stable
> release of OpenOffice in January and a preview 3.4.0 release in April.
> It is very reasonable for the new ApacheOffice project to start up,
> and even while in incubation produce a release.

	It is unclear to me whether you can release binaries with all the
copy-left dependencies bundled as Apache. If that is so - easy enough.
If not, life will be harder, more development will be required, and the
result will be much less feature-full.

	Which makes me wonder - who would provide the binaries for the
project ? is that typically done by a single individual ? or a company ?
(I assume not) or is it a requirement that anyone can do that ?

	I ask because - it is a non-trivial amount of work to setup a Windows
(particularly) build environment capable of compiling OO.o, and clearly
diversity is useful there. Then again, if you have Eric and IBM that can
compile there perhaps that is robust enough.

	As for the general quality of the 3.4.0 preview release, it seems to me
that there is some considerable work left to make that release-able.

> Will there be a longer-than-user delay between releases as we
> produce our first release? Of course. But I'm not particularly
> troubled by this.

	My take is (and some of this depends on Apache policy, and what code
arrives out of the process) that it could take a handful of months to
get back to the OO.o position. But of course, I look forward to being
proved wrong.



--  <><, Pseudo Engineer, itinerant idiot

To unsubscribe, e-mail:
For additional commands, e-mail:

View raw message