John Peacock wrote:
> David Glasser wrote:
>> What's the current status? Is it in a state where it would make sense
>> to merge it back to trunk (even if it's not necessarily done) and
>> continue working there? Especially with the recent big changes to
>> trunk (sparse directories, merge tracking), it would be great to get
>> these improvements tested with the latest and greatest as soon as
>> possible.
>
> I just had a conversation with Nik about this very topic. He's swamped
> (he started a new job three weeks ago), and I had a bunch of things hit
> at the same time, so I haven't been as active as I had hoped.
>
> However, if 1.5.0 doesn't branch at the end of the month, I may have
> enough time to review what Nik has already done and get it into a state
> where it could be merged with trunk (for example, I don't think that the
> branch has been refreshed in a while). I'd rather get something merged
> to trunk and possibly part of 1.5.0, rather than wait until we have
> everything done.
The grunt work of actually wrapping the SVN::Client functions in a
manner that permits named parameters is mostly done. There are going to
be a few methods where the auto-generated parameter specifications are
incorrect, but they should hopefully be easy to track down.
What's missing is:
a) Updated documentation that shows (1) how to call all the methods that
were previously wrapped using named parameters (the old positional
parameter calling mechanism still works), (2) new documentation for all
the methods that have been wrapped that weren't previously wrapped.
b) Tests that cover the new methods.
Note also that the methods wrapped are those that were in svnclient when
the branch was cut, and won't include (yet) any that have been newly added.
In a couple of weeks time I should have invoiced enough to go and get a
laptop, at which point, I'll be able to carry on this work during the
daily commute. For the time being though I'm pretty much swamped.
I suspect that planning to merge this back for 1.5.0 is probably a bad
idea, given the amount of work involved in (a) and (b) above (it's not
difficult work, it's just that there's quite a lot of it). If there was
going to a be 1.5.1 for (say) around June then that's probably much more
realistic.
There are a handful of bug reports about the Perl bindings, where people
have submitted patches to wrap specific new functions. I haven't
reviewed them in detail, but they're good candidates for 1.5.0.
N
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Wed Mar 28 16:48:05 2007