> > 4. A sound confidence in that there are so many Evolution bugs that have > > been outstanding for a very long time 12 months + > The picture can be wrong if you look only in a number of bugs.
Agree, 115%. Looking at the bugzilla of a project to ascertain something meaningful about it is almost entirely useless and perhaps even misleading. (1) Some projects promote problem/bug/enhancement requests via Bugzilla more aggressively than others. In some projects only the developers actively use Bugzilla, in others users are encouraged to "contribute" to bugzilla. (2) More popular projects [and Evolution is about as popular as they get] will..... have more issues reported. (3) Different projects use Bugzilla differently - some use it just to track bugs, other use it to keep track of any kind of reported problem (even likely spurious ones). Others use it as a way to track "wishlist" item (that maybe only one user is interested in). (4) Some use bugzilla in a more passive way, where if you file a bug you also have to jangle a chain on the mailing list, etc... To gather any real meaning from Bugzilla you have to *really* know the psychology of a project; and the only people who know that are people who have been *involved* with the project for some time. In short - it is a useless tool for anything other than what it is. Bugzilla also misses (as does tools like Ohloh) all development around a project; for evolution this would be plugins and backends. Non-core development for many projects is quite significant. In short, as with all things in life, substantive findings are hard and grueling work. > When you look in bug reports, there is a lot of bugs that is not easy to tell > what to do as problem is reported by one person with not so common hardware > or software configuration that no one else confirmed, bugs waiting for > response from reporter for a long time, bugs waiting on response from > upstream developers, This is all true; I see bugs all the time where I think "Huh?" because I have lots of users using the software-in-question right under my nose and have never seen anything like the issue being reported. This is especially bad when you have people using alpha/beta release of distributions, fringe distributions, etc... One has to get down to what libraries (and version) they are linking against, compiler version, etc.... Ugh! Most of the time it just isn't worth the effort - when there are allot of both bugs and interesting issues relevant to the large mainstream. > overloaded developers not having time to check old bug > reports, etc. Yes, and as a developer who spends a fair amount of time in bugzilla - BORING! <aside>The number of people who file bugs and then vanish or just don't respond to further inquiries is pretty significant. I've even seen people submit *patches* along with reports, but never surface either in bugzilla or mail lists again. Both weird and frustrating.</aside> > I guess that being busy is overall problem as you have time to check number > of > bugs, but not what they are, so you make your decision on facts that you have > time to capture. I think calling them "facts" is dubious, more like "impressions"; which may very well be false. "Facts" are hard won things, they are almost never [maybe absolutely never], derived from doing a scan or overview of data. > BTW, I'm not using Evolution, but 2 other posts in this thread that say they > are satisfied with, made me wonder how can happen that Evolution has many > open bugs, but still works good. Because: a bug is not "a bug". It is much more complicated then that. That and to say that Evolution lacks "Continual Development and support" is preposterous. It gets updated all the time, and certainly with almost every GNOME release (every ~6 months). And "Development vision"??? Sounds nice, but totally a straw dog. Software has goals and purposes, people talking about vision don't belong within light years of a code repository. -- Adam Tauno Williams Network & Systems Administrator Consultant - http://www.whitemiceconsulting.com Developer - http://www.opengroupware.org -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
