On Thu, 15 Aug 2013 11:07:20 -0500, Phil Steitz wrote:
On Aug 15, 2013, at 7:31 AM, Gilles <gil...@harfang.homelinux.org>
wrote:
On Wed, 14 Aug 2013 23:40:58 +0100, sebb wrote:
On 14 August 2013 23:34, Gilles <gil...@harfang.homelinux.org>
wrote:
At this point, I'd tend to think that creating a copy of trunk in
the
Commons's "sandbox" part of the repository will be more
productive.
Why sandbox? Just use a temporary branch in the existing math repo.
Has Evan the permissions to commit there?
No, but same is true of the Sandbox unless he is already an ASF
committer. Sebb's point is that temp branches in svn can be used for
this.
Hmm; isn't there a place where no committer could be given privileges
to write to the repository?
It's also fine to just look at his stuff in GitHub for now, IMO.
As we discussed in the few last posts, the "stuff" is not complete;
it is only the realization of the "potentialities" that would it
worth replacing the currnet code.
Personally, I was expecting this issue to be moved forward, i.e. evolve
the design until it actually (i.e. in code) meets all the expected
improvements.
But I won't have time to fill in all the blanks (Javadoc, formatting,
naming, unit tests porting) for this to be committed in trunk.
I was hoping that we can work in real-time with Evan; meaning: he
would create the bulk of the code, then I (and anyone interested)
can comment on those aspects that would IMHO need polishing, then
he could commit additions and modifications.
So that in the end we can commit top trunk the whole contribution
without fear that some tedious and time consuming work will be needed
afterwards.[1]
Gilles
[1] Cf. "BOBYQAOptimizer"; (efficient and maintainable) porting is
still far from finished even after tens of hours spent on it.
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org
For additional commands, e-mail: dev-h...@commons.apache.org
