Re: More on possible problem with Neon 0.25.0 [was Re: 1.3.0 RC1 ready to be rolled]

Greg Hudson wrote:
> I'm not sure you can be +0 on something if you think it's a "wise
> policy" not to do it. :)

I'm saying that the general principal of being strict about dependencies
can be relaxed if our project is comfortable with the management of
those other projects. I have a suspicion that any time that a Neon
release candidate is made, at least a couple of people on this list try
it out.

The problem is, as I found out, that any ad hoc testing of Neon with
Subversion was not enough. At least on _my_ machine, 0.25.0 fails when
compression is enabled and that failure was known to the Neon
developers, but that information never penetrated into the INSTALL file.
  I'm trying to prevent my issues from coming up in the future, rather
than just relying on good behavior on the part of our dependencies.

> Anyway, your attitude is decidedly unfriendly to upstream maintainers.
> How would we feel if we found out that people weren't upgrading from svn
> 1.3.0 to svn 1.3.1 because Eclipse was only working with svn 1.3.0 due
> to a similar policy?

If there were any serious problems that 1.3.1 fixes, that would act as a
pressure on the Eclipse team to make a new release. However there are a
couple of things that your scenario ignores:

A) Eclipse would still work with 1.3.0 and indeed works without
Subversion installed at all.

B) Subversion, on the other hand, cannot use ra_dav with 0.25.0 (at
least on my machine) in the default configuration.

A is a minor annoyance; B is breakage (even though there is a workaround).

John

-- 
John Peacock
Director of Information Research and Technology
Rowman & Littlefield Publishing Group
4501 Forbes Boulevard
Suite H
Lanham, MD  20706
301-459-3366 x.5010
fax 301-429-5748
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org