Rich Burridge wrote: > PSARC members, I'm looking for a +1 endorsement for this project before I > can close out this case as approved. > > There was discussion on whether this project should combine with the > existing bzip2 application, but we understand that because these two > programs behave differently [1][2][3] and this combined functionality > doesn't already exist, the majority of those responding [2][4][5][6] agreed > it wasn't a requirement of the project team to make these changes and that > the project team could provide pbzip2 as a new separation integration. > Disagreement came from one replier [7]. > > Can we get the +1 endorsement for the specification as originally provided > or does the project team need to update their materials in some way?
Not from me. From an ARC interfaces view point bzip2 and pbzip2 are the same and thus there should be only one of them. The issues with pbzip2 not being able to replace bzip2 all appear to me to be implementation issues. If this was purely Solaris developed code rather than two divergent upstream F/OSS projects (well actually one is a fork of the other) I doubt we would be considering allowing pbzip2 when bzip2 already exists. I would much rather see one binary bzip2 in the filesystem, if that means that project team wanting to integrate pbzip2 needs to do some work then so be it. The rules aren't that you get to dump stuff in unmodified - if they were why even bother with ARC review. Now having said all of that I don't feel strongly enough to derail and write an opinion about it because I think that is a waste of ARC resources (take that more as Abstain rather than a Deny vote if you like). I do however very strongly encourage the project team to work out how they can ship a single bzip2 binary that doesn't suffer from the problems that were brought up during this review. -- Darren J Moffat
