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

Re: status information

From: Greg Stein <gstein_at_gmail.com>
Date: Fri, 18 Jun 2010 17:21:55 -0400

On Fri, Jun 18, 2010 at 14:07, Mark Phippard <markphip_at_gmail.com> wrote:
>...
> I would like to also think that we will not even consider shipping
> 1.7.0 unless it is faster overall than 1.6, and that would likely
> include being faster for the API usage patterns of TortoiseSVN.

Absolutely true. Which is why aspersions against that are so freakin' annoying.

> A couple other points/questions.
>
> Stefan seems to be operating under the impression that the public
> libsvn_wc API is going away in 1.7.  I did not think that was true?  I

It is not true. libsvn_wc will continue to exist and be supported.
Clients may need to call into it, as far as I'm concerned.

Others would like clients to only use libsvn_client functions. *shrug*

We might not add public functions to the API, but we're keeping pretty
much all existing concepts there and available.

>...
> Second, Stefan's idea of an API with flags about what you want to
> retrieve makes sense to me IF we are talking about a client API.  This
> would seem like a good way to keep callers of the API writing to the
> client API and allow us to deal with using the callbacks mentioned
> from within the implementation of the client API based on the flags
> passed.  I do not get the push-back to this idea.

The functions in our API which take flags have proven to be the most
problematic over time, IMO. To hazard a guess, I think it is a cop-out
to thinking harder about the semantics of what the code is really
trying to do. So with the loss of those semantics, we end up with
less-maintainable code.

Cheers,
-g
Received on 2010-06-18 23:22:35 CEST

This is an archived mail posted to the Subversion Dev mailing list.