On Wed, 31 May 2006, Peter N. Lundblad wrote:
> > Modified: trunk/CHANGES
> > URL: http://svn.collab.net/viewvc/svn/trunk/CHANGES?pathrev=19863&r1=19862&r2=19863
> > ==============================================================================
> > --- trunk/CHANGES (original)
> > +++ trunk/CHANGES Tue May 30 19:30:28 2006
> > @@ -137,6 +137,16 @@
> > - new support for svn_client_info (r18758)
> > - minor corrections to SVN::Fs (r19312)
> > * javahl bindings:
> > + - APIs to get the major/minor/patch/tag for the version (r17604 -07)
> What is "the version". Suggest "new APIs to get native library version"
> or something.
I've improved the wording.
> > + - cleanups to and consolidation of READMEs
> Do we really need to mention this? I didn't in general when writing
> the rest of CHANGES.
I added this only because we mentioned cleanups to README with the
1.0.0 release notes. I'm fine with no longer doing so -- wiped!
> > + - C++/Java code refactoring, cleanup, and consolidation
> Does this change anything from an outside perspective, or does it only
> improve maintainability? If the latter, please drop.
If you're using the C++ code directly, it matters. However, I assume
such usage to be an edge case. If we care to mention only the "end
point" of the bindings implementation (the APIs commonly used by
bindings programmers), this could be collapsed to "improved JavaDoc"
or some such.
> > + - fixed: cleanse gcc compilation warnings (r16436)
> ^ typo?
> > + - fixed: run unit tests against SVNClientSynchronized (r18054)
> I don't think we need to mention individual test script fixes either.
Malcolm Rowe followed up with:
> On Wed, May 31, 2006 at 01:29:26PM +0200, Peter N. Lundblad wrote:
> > email@example.com writes:
> > > + - fixed: Visual Studio 2005 compilation error (r16436)
> > These two don't need mentioning here, IMO.
> This one might be appropriate if it prevents compilation otherwise.
> We do normally list fixes in terms of the bug, so:
> - fixed: Unable to compile using Visual Studio 2005 (rXXX)
We mention fixes for compilation errors elsewhere in CHANGES, too. I
left this one in.
> (though I note that r16436 appears to be a GCC4 warning fix, and
> doesn't mention VS2005 at all).
Yeah, had my revnums mixed up on a couple lines, corrected.
Thanks for the review, guys.
Received on Wed May 31 19:01:52 2006
- application/pgp-signature attachment: stored