Mark,
As it got pointed out to me, this release time has been difficult due to the
need for software to be in place that is required for more pressing issues.
I suppose this made exact dates hard to determine this time. The next
release should be more organised I hope, especially now that I have a
complete overview of the procedures. Hopefully the classic starlink will be
under CVS control for next time, which should make thing easier for everyone
if we have the procedures for dealing with CVS in place.
Steve.
-----Original Message-----
From: Mark Taylor [mailto:[log in to unmask]]
Sent: 13 June 2003 09:41
To: [log in to unmask]
Subject: Re: Repository status
On Fri, 13 Jun 2003, Rankin, SE (Stephen) wrote:
> All,
>
> I will tag (USSC218) the repository today if there are no more problems.
But
> in reality, it should be a mutual agreement between everyone when a tag is
> worth doing. Are there changes since the freeze that should not go on the
> CD? Come to think of it, a freeze should be a freeze, no further updates
to
> the repository until further notice. So I think the rule is, TAG is not
> equal to freeze.
Thanks for the clarifications Steve, that seems fair enough.
There are more formal strategies for this based on branches
(see Norman's comments at the bottom of the wiki CvsTagging page),
or by checking out at the freeze date and using CVS tag rather than
rtag, but I'm not necessarily saying those are better.
> Did people want me to downgrade the JRE, JDK and JAI or is everyone happy?
I will do some more tests today, but personally I think I'm happy
with the new versions. However if others (Peter) want to go back
to the earlier versions I don't object to that either.
I have a final request for the future (assuming we've got one etc),
which is that timetables get announced to programmers as early as
they are known, and that we are notified of changes to the schedule
if events cause these to occur. I don't know when the schedule for
this CD was decided on, but for instance I didn't know a deadline
was imminent for the Starlink classic collection until it was
announced as 7 May on the programmers' meeting on 24 April.
If there are formal testing periods these should be announced
too, so we know when it is and isn't too late to provide bugfixes.
Thanks,
Mark
--
Mark Taylor Starlink Programmer Physics, Bristol University, UK
[log in to unmask] 0117 928 8776 http://www.star.bris.ac.uk/~mbt/
|