"Peter N. Lundblad" <peter@famlundblad.se> wrote on 03/23/2006 12:26:54
PM:
> Paul Burba writes:
> > > I can't help thinking that this parameter is redundant. Why can't
the
> > > "frompage" or "topage" parameter to svn_utf_cstring_*_utf8_ex() be
> > > used as the
> > > cache key?
> >
> > Hmmm, yes, why can't we do that? Except for the hassle of an API
change,
>
> Yes, I've been thinking the same. The only think that could speak
> against this would be the performance penalty of building up the
> unique cache key (we need to include both to and from encoding and a
> unique part, since this key is put into the users' pool sometimes). I
> don't believe it should make any difference, but I don't know. When
> we do utf8 conversion it is most often in conjunction with some I/O
> operations that should be much more expensive than copying 25 bytes
> into a buffer...
>
> We don't even need an API change, since the cache_key parameter does
> not have any semantic significance. Just update the docstring to state
> that it is ignored. Or we could rev the API to get rid of the bogus
> extra parameter.
>
> Any takers?
Hi Peter,
I'm glad to do it, all I need to know is if folks want to do it without
the API change or want svn_utf_cstring_*_utf8_ex revved.
Paul B.
_____________________________________________________________________________
Scanned for SoftLanding Systems, Inc. and SoftLanding Europe Plc by IBM Email Security Management Services powered by MessageLabs.
_____________________________________________________________________________
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Thu Mar 23 22:26:04 2006