Brian Havard wrote:
> On Wed, 10 Nov 2004 20:50:33 +0000, Julian Foad wrote:
>>Well, I'm torn between protecting any OS/2 users that we might already have,
>>and trying to get out of keeping this work-around forever.
>
> Don't worry about it too much. If I need to write an implementation
> for OS/2 I'll do so. Actually, the unix implementation looks almost
> portable so I can probably adapt that.
Excellent. Thanks. I don't suppose there's any chance you could implement it
in APR before we release a version of Subversion that requires it?
I don't want to upset people by my habit of concentrating most of my energy on
little warts like this and perhaps distracting attention from bigger issues.
Karl asks why I don't just leave it alone, and I suspect several other people
feel the same way. One of the main attractions of working with all of you on
Subversion is the niceness - cleanness - correctness - of the source code. I
just happen to be a person who takes pleasure in cleaning up little things like
this - witness how many of my commits have been just fixing spelling mistakes,
or removing redundant code, rather than adding features.
You could persuade me not to remove this work-around yet, by arguing that we
don't want to break Subversion for OS/2 users, but in that case we must make
some effort to resolve the long-term situation by getting it implemented in APR
first. We have the difficult situation at the moment of there being little
motivation to fix APR. Removing the work-around would be a bit of a violent
way of providing that motivation, but I get the impression from Brian and
Justin that it might be acceptable.
Thus I will remove the work-around soon if no-one vetoes or persuades me out of it.
- Julian
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Thu Nov 11 18:34:58 2004