[svn.haxx.se] · SVN Dev · SVN Users · SVN Org · TSVN Dev · TSVN Users · Subclipse Dev · Subclipse Users · this month's index

Re: [VOTE]: Default http-client for 1.7 Serf or Neon

From: C. Michael Pilato <cmpilato_at_collab.net>
Date: Thu, 04 Aug 2011 16:50:24 -0400

On 08/04/2011 03:36 PM, Greg Stein wrote:
> On Thu, Aug 4, 2011 at 15:20, Mark Phippard <markphip_at_gmail.com> wrote:
>> ...
>> vote for their preference. I am not saying that the simple majority
>> decision of this vote should become the default.
> Not until I'm convinced to remove my veto. We've discussed several
> times that the future direction lies with serf, and I believe it is a
> wrong choice to move us away from that. The issue tracker has not had
> any serf-related bugs for a while, so there is nothing to indicate
> "not ready".


It's obvious that you firmly believe that the future lies with Serf. And
I'm aligned with you there (though with some caveats you'd likely disagree
with). But nobody is asking about that future -- we're asking about the
present. We're talking about taking what many believe to be a step backward
in terms of the relative stability of Subversion as a whole for questionable
(if any) gain. Sure, Serf's stability will improve the more it's used *and
its bugs are reported* -- no question about it. I'm just not convinced that
users will take the time to report an issue against Serf when an immediate
workaround for pretty much any of their Serf-specific problems already
exists ("I'll just switch to Neon"). Further working against this is the
fact that our single most common binary package producer, TortoiseSVN, plus
others are saying that for the sake of their user base, they'll switch to
Neon by default anyway. Who then, will be using Serf and reporting to us
its specific problems?

I'm concerned about our effectively transmitting this message from the
Subversion developers to its users: We've made the decision to
intentionally offer you, by default, a product that is known to be less
stable than it could be in exchange for ... nothing, really, except the
benefit that *we* get (and then you'll get, by extension, some time in the
future) by forcing you to be our QA department. That doesn't strike me as
the embodiment of "community before code".

All that said, if your veto hinges on the presence/absence of
1.7.0-milestoned bugs filed against Serf alone, that does seem fair to me.
After all, it's practically impossible to "fix" what cannot be clearly (and
reliably) demonstrated as broken.

C. Michael Pilato <cmpilato_at_collab.net>
CollabNet   <>   www.collab.net   <>   Distributed Development On Demand

Received on 2011-08-04 22:50:58 CEST

This is an archived mail posted to the Subversion Dev mailing list.

This site is subject to the Apache Privacy Policy and the Apache Public Forum Archive Policy.