On 01-07-2009 12:02 PM, Thorsten Behrens wrote:
Ultimately, I believe development work is already fragmented enough
with the existing number of projects, so I'd suggest stripping down
the scope to measuring performance (as QA-ing performance again
seriously overlaps with the QA project).
I interpreted the proposal as directly addressing the fragmentation that
indeed seems to exist, by creating a unifying project centered around
performance issues. Such a group could maintain a broader overview of
reported problems with performance, and this might be very helpful in
coordinating and communicating information to and between the existing
programming groups. When it comes to the actual refactoring/rewriting of
code, certainly those who already have the strongest familiarity with
the 0's and 1's "under the hood" would be most qualified to make
performance-oriented mods.
(caveat: I say this as an outside observer who, in my real life, is a
project manager and strategic consultant on organizational/product
issues just like this one. I am not programmer, and who have only a
beginner's understanding of the structure of OOo groups)
Think of it as a sort of task force to support successful fulfillment of
a very important product acceptance criterion.
...
On 01-07-2009 6:00 PM, Dr. Bernhard Dippold wrote:
Perhaps it might be possible to modify the scope from
"Members of the project will improve the performance in
these areas,"
to
"Identified areas with insufficient performance will be worked on
together with the dedicated projects for these areas."
Absolutely, and to help coordinate the vetting of performance-problem
reports, the writing of test cases, QA recruiting, and so on that would
be necessary to support the programming effort.
Same with QA - setting up a test case for performance testing is
great. It might be included in the QA testing routine afterward.
This would mean that developer interested in performance optimization
work will have to subscribe to more than one mailing list (depending
on the part of OOo they work on).
In my experience "more than one list" is already a reality even for
persons who, like myself, are only marginally involved ... and IMO is
not a deterring issue.
I don't know if this is a problem for some of them - but the
Performance project will only be able to come into life and stay alive
if it can be seen as binding link between the projects.
Amen.
kazar
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
