Hi Peter, Some very interesting ideas here. Sadly, no, none of these are on the roadmap (we have so much we want to do that we'll easily spend years on our own current TODO lists). However, after the currently in-development 1.6 release, our focus will be on getting our extensions branch into the product. This would enable people to write extensions to do all sorts of things. For example, augmenting the Review Board UI, or the API, or jumping in and analyzing diffs. Much work has been done on this, and it works fairly well so far.
The tricky part is matching up code in the codebase with reviews on diffs. We store diffs of every change (with associated revision), and reviews on the lines in that diff. I don't know exactly how the health determination would be done, but if something were to query Review Board, we'd need a way to match up lines of code with those found in diffs. That means a lot of processing (patching the files pulled from the repository and generating patched copies we can analyze). How do you take a line in a file and ask if there have been reviews on that line when the file may have changed 10 times since the last review? I can see it being difficult, though an interesting problem to tackle. Certainly something to think about. Better bug tracker integration is something we plan to do, however. Some experimental extension work was done on that recently, but it's not ready for general use (and won't be until we get the branch merged in after 1.6). Christian -- Christian Hammond - chip...@chipx86.com Review Board - http://www.reviewboard.org VMware, Inc. - http://www.vmware.com On Sun, Jan 23, 2011 at 1:59 PM, Peter Mounce <p...@neverrunwithscissors.com > wrote: > I just discovered the application today (hi!); I've been looking for > something like it, and come across Atlassian Crucible, Gerrit and Rietveld > so far. > > One of the features I'd want from such a tool would be to have a > code-health report that is derived from the % of code covered by review. > * The code-health of an unreviewed section of code would be very low. > * The code-health of a reviewed ages-ago section of code would be low. > * The code-health of a reviewed-not-long-ago section of code would be > higher. > > It might be nice to support different classes of people having different > affects on code-health in this report - for example, getting code reviewed > and passed by a senior developer would carry more weight than a junior > (where "senior" and "junior" are just class-names; I'm not suggesting that > people automatically get more weight by being employed longer). > > I would also want to be able to assign other scores to such a report from > tools that run during the build process, like code-coverage triggered by > automated tests, static analysis reports, etc. > > I'd also want an API way to integrate with an issue-tracker so that one > could correlate defects between pieces of code affected, which could happen > by knowing which issues are being hit by which source-control commits > (something that's already quite common, I think; each place I've worked has > had that to one degree or another, but stop short of actually analysing that > data). > > Is this kind of thing within the remit (or roadmap!) of reviewboard? > > -- > Want to help the Review Board project? Donate today at > http://www.reviewboard.org/donate/ > Happy user? Let us know at http://www.reviewboard.org/users/ > -~----------~----~----~----~------~----~------~--~--- > To unsubscribe from this group, send email to > reviewboard+unsubscr...@googlegroups.com<reviewboard%2bunsubscr...@googlegroups.com> > For more options, visit this group at > http://groups.google.com/group/reviewboard?hl=en -- Want to help the Review Board project? Donate today at http://www.reviewboard.org/donate/ Happy user? Let us know at http://www.reviewboard.org/users/ -~----------~----~----~----~------~----~------~--~--- To unsubscribe from this group, send email to reviewboard+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/reviewboard?hl=en
