On Wed, Oct 8, 2008 at 2:24 PM, zooko <[EMAIL PROTECTED]> wrote: > On Oct 8, 2008, at 15:01 PM, Jason Dagit wrote: > > One final note, I would prefer if the user-defined attributes had their >> own place in the patch format, but I seem to be the only person who feels >> strongly about it. So, I guess putting it into the comment field is OK for >> now. >> > > I would, too. They are not, to my mind, "comments" or parts of comments. > > However, older versions of darcs (including, I suppose, darcs 2.1.0) would > fail if asked to process a patch which had an unrecognized header... > > Wouldn't it? > > I haven't tested this or read the relevant source code -- I just assumed > that it is true.
As far as I know this is true yes. On the other hand, I don't know how we deal with introducing new patch types either. I don't really understand why we get so hung up making sure older darcs understands newer darcs. I can understand the other direction as being important. I think we're over constraining ourselves if we worry about compatibility in both directions. We have to manage risk and weigh trade offs, and here is a place where I think ensuring older darcs understands newer darcs is not as important as the proposal under consideration. My understanding is that, we have a chance to break forwards compatibility to make the patch format more flexible. With all that said, I'd rather see people working on performance issues and known bugs. Jason
_______________________________________________ darcs-users mailing list [email protected] http://lists.osuosl.org/mailman/listinfo/darcs-users
