Marc Sherman wrote:
> Greg Stein wrote:
>
>> For something like version control, I would tend to think slower
>> cycles also, and would support a six-month cycle. What kind of
>> confidence can we give people if we are continually making new
>> releases. "Woah? What was wrong with the last one? They're making a
>> new release again?!"
>>
>> Though hearing from "real" admins like Lieven, that can certainly
>> alter that viewpoint.
>>
>
> Ok, if that's a request for more input from admins, I'll pitch in. I'm
> the svn admin for my company, and I'm eagerly awaiting 1.4 for the wc
> improvements -- they'll be a big win for all of our developers. I'm
> less interested in serf; we're currently using neon, and definitely
> experiencing performance pain with it, but I plan to alleviate that pain
> by getting svnserve access to our repositories set up within the next
> month, so serf won't be a big win for us.
>
> While 6 months seems to me to be the sweet spot for minor releases, IMO
> the current state of trunk is ripe for an exception to that rule.
>
Well, you could call _me_ a "real" admin, too, since I introduced SVN as
the VC tool of choice in our company. My story is a little different; we
won't use svnserve in the foreseeable future, because our setup depends
on authenticating against our Active Directory server, so that we have a
single centralised accounts database. We use Apache with a (locally
modified) mod_auth_kerb and stock mod_(dav|authz)_svn. Serf _might_ help
us performance-wise, but even if it does go into an (early) 1.4, I
certainly wouldn't use it in production until it's had some time to
prove itself. So in my case, I'd rather see serf in a somewhat later
1.4, after it's had time to soak for a bit.
-- Brane
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Sat Jan 28 16:53:32 2006