carlsonj writes: > Josh Hurst writes: > > > Or that code review comments must > > > consist of diffs? > > This may be nice > > > > > > > > Or perhaps something else? What would be "fair?" > > Fair may be to start submitting some patches. Just > some. Just to avoid > > the impression that this project was done using > slave labor > > Ah, I see. In that case, I think you may be mistaken > about the role > that a "reviewer" plays in the development process. > > Perhaps other open source projects insist that review > comments must be > accompanied by diffs. Or that there's some sort of > reciprocity > involved: "I'll consider your suggestions only if > you'll shoulder some > of the project burden." That has simply never been > the case with > Solaris development, as it places far too high a > burden on reviewers. > It's hard enough to get people to do decent reviews > of the work of > others, and placing additional commitments on those > who volunteer to > do that work means that reviews will be meager at > best.
There are some that think that the only comments/suggestions there's an obligation to consider are those accompanied by code. Note: I'm _not_ one of them; occasionally I'll have time to write something, but usually not. So in those environments, I end up simply throwing down the toilet most input I might otherwise provide. That doesn't make sense, IMO. >From what I've seen, most of the comments here have at least included constructive elements (sketches of alternative approaches); granted that code _might_ sometimes move things along faster, but so what? As long as the project folks aren't overly frustrated, such that they hopefully come back for some of the followup, all the back and forth can be regarded as an investment in their training, that will pay off in the long run. (All IMO, as a non-participant observer) I know I've been waiting for a proper modern ksh93 on Solaris for many years. Happens I can either build my own or arrange things so the AT&T binaries work until such time as it's properly integrated; a few more months one way or the other, after years, isn't an issue as long as it's done right so that it will be kept reasonably current and in sync, while at the same time not causing any reasonably avoidable breakage. Patience doesn't suck as long as progress is being made... This message posted from opensolaris.org
