2.0.3?

[Andrew Bennetts]

Martin Pool wrote:
[...]
> 
> In looking at the trunk bug fix section, I see a fair number of things
> that, at least looking at the description, really are genuine bug
> fixes (as opposed to performance or testing) and that could qualify
> for 2.0.  I wonder why they weren't proposed for it?  Like the
> following:

Hmm... some of these are from relatively new contributors (thanks to patch
piloting?).  The obvious way for a new contributor to make a fix is to
branch trunk and submit it to trunk.  Our process typically expects the
person proposing the patch to suggest the backport, unless someone makes
noise about backporting in the bug.  So for many of these bugs I suspect
that, quite literally, no-one had thought about whether they should be
applied to 2.0.x as well (until your mail, of course).

That said, here's my thoughts on mine:

> * Terminate ssh subprocesses when no references to them remain, fixing
>   subprocess and file descriptor leaks.  (Andrew Bennetts, #426662)

The patch is pretty short and it seems sane, but it depends on when garbage
collection kicks in.  Also, I don't think many people are affected.  So the
risk/reward tradeoff seems bad for 2.0.  If 2.1.0b4 shows no problems then
perhaps it's worthwhile?

> * The ``--hardlink`` option of ``bzr branch`` and ``bzr checkout`` now
>   works for 2a format trees.  Only files unaffected by content filters
>   will be hardlinked.  (Andrew Bennetts, #408193)

Arguably this is a feature not a bugfix...  I'm also not super-confident
with content-filtering related code, so again I figured it was better to be
paranoid than risk some sort of regression in 2.0.  Again if people report
that it works perfectly for them in 2.1.0b4 then it's probably fine to put
in 2.0 though.

-Andrew.