Hi, IMHO, using a bugtracker system to list bugfix or changes is a good idea. But today, the practice in FOP project is to fill this list directly in the cited page. We can adopt a new policy here and have a change/bugfix list using bugtracker facilities. That could be done when we'll migrate to Jira. IIUC, Bugzilla entries will be migrated to Jira in the future. Waiting this migration, we can today begin to systematically fill a new Bugzilla entry, to keep trace in Jira.
WDYT? 2012/7/12 Glenn Adams <gl...@skynav.com>: > > On Thu, Jul 12, 2012 at 10:47 AM, Vincent Hennebert <vhenneb...@gmail.com> > wrote: >> >> On 12/07/12 16:17, Glenn Adams wrote: >> > If there is no bug entry in bugzilla, then there is/was no bug. Since >> > this >> > is clearly a bug fix, there should be a documentation trail through the >> > bug >> > database. So please create an entry and do so in the future. The >> > status.xml >> > document is only an informal paraphrase of bug database entries, and >> > should >> > not be considered the authoritative list of bugs. >> >> Well it /is/ authoritative, in the sense that its content is used to >> display the list of changes on the website: >> http://xmlgraphics.apache.org/fop/changes.html > > > The fact that status.xml makes reference to bugzilla entries, and not the > other way around shows the latter is more authoritative than the former. > >> >> >> >> As long as it’s the case, duplicating entries in Bugzilla is just >> a waste of time. > > > Not it isn't. It is good time spent to document the work on FOP/XGC, etc. > Others are doing this, so you should follow suit and not be remiss in your > duties. > >> >> >> Once that status.xml has been deprecated and an other mechanism >> implemented to extract the list of changes from Bugzilla and display >> them on the website, I’ll certainly start creating entries in Bugzilla. >> In the meantime, I don’t see the point of doing both. >> >> Unless, again, I’m missing something. > > > It's a matter of following best current practice. Failing to record in BZ is > not following BCP. > -- pascal