On Tue, 15 Nov 2005 16:46:02 -0000 "Paul A Bristow" <pbristow@hetp.u-net.com> wrote:
Well I'm not sure it has been in the past? IMO a few minor changes are OK - and desirable even. I don't remember a review when NO changes were required. Nor a product that was never modified - that's why Boost code is good?
Right. However, we have had problems in the past where the original is modified. That's one reason that the original is not supposed to change at all. Mods are almost always needed (compiler workaround or doc updates), but the original submission must stay. We've also had the argument in the past as to what is being reviewed, the newly modified work, or the original. I'm not trying to put a new face on it, just stating that the original submission is always supposed to stay constant (that's what I think the original question was about, anyway).
IMO, as far too often, far too many fundamental issues are being raised far too late in the whole development and review process.
However, we do not have a pre-review stage, so it is difficult to see how these issues would have been raised in the past. I believe our history shows that a submission will be rejected if it is not satisfactory, but when the author has considered the points raised in the review, it can be resubmitted. I guess this is the best we can do, and it seems to be fairly well received.