On 12/04/2006 3:16 a.m., Adrian Chadd wrote:
I'd like to see a number of squid-2.5 related projects brought over
into monotone to make merging and managing them easier. I'd also like
to merge in changes into a 'head' branch so mature projects can be
merged in and then incorporated into the work of others.
This is all starting to sound (again) more and more like 2.6 branch should be
opened up. Unfortunately we're back where we started a few months ago in
October whereby we had this discussion, a decision was made to stabilise 3.0 but
alas we're not a whole lot closer to a release or stability with 3.0 now than we
were back then. There was talk at the time about squid-3 being possibly a
matter of months away.
In other words, it seems that that many of Henrik's comments have turned out to
be true (Sorry Alex and Rob):
http://www.squid-cache.org/mail-archive/squid-dev/200511/0054.html
I think we're getting to the point where we just have to do a 2.6 release else
the whole situation of us being stuck far from any sort of release or motivation
risks dragging on for months or years longer. A line has to be drawn somewhere
at which point we decide that current strategy is not working and to make a move
in a different direction. My perception based on mails to -dev and -cvs is that
3.0 has stalled yet again, so lets not continue to wait indefinitely for
progress which could be best described as "slow".
Aside from that, I'm seriously in favour of anything which gets away from any
sort of dependency on Sourceforge. The entire SF anoncvs access has been down
for over a week with no indication of when it will be back, and this is not the
first time this year there have been major problems with it. I'll let Adrian
and Rob fight the specifics of what we move to instead, as long as it's not CVS
on sourceforge I'm happy ;-)
Reuben
