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

Re: [PATCH] Introduce per-instance filesystem UUIDs

From: Evgeny Kotkov <evgeny.kotkov_at_visualsvn.com>
Date: Tue, 19 Aug 2014 00:43:49 +0400

Hi Stefan,

> Here at the SHF hackathon, we briefly talked about how to proceed. I quite
> happy with the current implementation wrt to its technical approach. However,
> we identified another operation that should bump the instance ID, fs_recover.
> And there might be more.
> None of that is complicated to solve in any way but we should have a full,
> reviewed list of places that need to update or possibly use the instance ID
> before releasing it. So, the correct way to to that is having a branch for
> the full feature - as suggested by Brane.
> Given that many of us are the same room for this week, progress on that branch
> as well as its review can be quick. Unless there is something fundamentally
> broken with it, there is a good chance to have it in 1.9.


> :D To me, you came across as fighting desperately to get
> the change into the 1.9 release.

Actually, I do not have concerns about this change becoming or not becoming a
part of 1.9. What I am really concerned about is that this internal little fix
could ultimately become an unnecessary complicated, full-stack and user-visible
feature. I doubt that we want users to know something about the way we solve
our *internal* problems. Why would they ever need to know? Probably not the
best analogy, but to me, it would look the same way if we, for instance, added
a command(s) to do some sort of reference counting that Subversion requires
underneath. Or, maybe, to deallocate memory. Or to close unused file handles.

As long as you are together on the hackathon, and everyone thinks that
reverting-and-branching-and-voting is a better way here, I would not insist on
doing it somehow else. However, I suspect that this approach requires a proper
revert justfication, and, unfortunately, I cannot really come up with that sort
of thing myself.

> More specifically, you seemed to have a hard time seeing any reason not to
> just apply that simple enough fix to for specific problem.

That is true, it became a lot easier with the provided patch.

> I really appreciate this. Written communication is the most lossy, opening
> things to bias and lots of second guessing. I have to remind myself of that
> fact once in a while.

Thank you for these kind words.

Evgeny Kotkov
Received on 2014-08-18 22:44:41 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.