On Sun, Mar 20, 2011 at 01:30:37PM +0000, sebb wrote: > On 20 March 2011 10:12, Gilles Sadowski <gil...@harfang.homelinux.org> wrote: > > Hi. > > > >> Gilles, > >> > >> This should have been done in three separate commits. > >> > >> The first one, at least, should have been separated. It is easier > >> for reviewers and makes the commit log clearer if we separate > >> formatting / javadoc cleanup commits from those that update or > >> change the code. The JIRA reference will pull all of these diffs > >> under the referenced issue. It is better if the commits that > >> reference the issue are directly related to the issue. > > +1 > > > Yes, it would be better. But I find it not so bad, as a compromise between > > reviewers's (potential) work and my (actual) work. > > However there are generally several reviewers, each of whom will have > to do extra work. > > And it makes it much easier for future maintainers when looking at past > history. > > Or indeed if one aspect of the commit has to be reverted.
I know all of these arguments; thank you for repeating them. I've also repeated several times that it is not up to other people to decide that I should spend <x> minutes more on an issue because they want to double-check each line which I modify. The more so if the changes are so trivial (like that issue + Javadoc + Junit 4) that it will be quite unlikely they would need to be reverted. [Is that the case, here? Shall a "revert" be needed in case I introduced a typo in the Javadoc comments? I don't think so.] I seriously doubt anyone here now or in the future will spend his time looking at several months old diffs... If he would, he could as well deal with the issue that raises such useless discussions. Regards, Gilles --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org For additional commands, e-mail: dev-h...@commons.apache.org
