> (Good thing you put 'usually' in there, because I've got a patch in the > system that fixes a bug in the *test suite* that's been hanging for over a > week now)
It's not on the verified patches list, so it's unlikely to get much attention. http://dev.rubyonrails.org/report/12 > Is it the policy of the Rails project that defect reports without a > verifiable test case are invalid and will be closed? No, defect reports without test cases aren't supposed to be automatically closed. A great way for people to help is to find defect reports, then attach failing test cases, or at the very least confirm that they experience the same behaviour. However I believe you're overstating the impact of closing a defect. There are two possibilities when someone reports a bug: a) The bug was invalid, and caused either by the user's system, the alignment of jupiter's moons or some other freak occurrence b) The bug is valid, someone else will be able to verify it Prematurely closing defects has no adverse impact on the first case. In the second case, the submitter can simply reopen them. Now even if you assume that closing the ticket has so completely enraged the submitter that they've gone off to write web frameworks in ocaml, we're still ok. Someone else will hit the same bug, and will either reopen the original report (unlikely because trac's search is shit) or submit a new one. The new ticket can then be fixed, or we can repeat this process again. The reality is that there are bugs in every system, and if there's a defect which is so rare that only one person has ever hit it, then I'm not going to lose any sleep over continuing to ship that bug. > > The problem is we have tons of unverified (and bogus) > > defect reports with no easy way to prove them. > > So mark them "more info required" with a note that they need to have more > info attached or they'll be closed in N days/weeks/months. That's exactly the same as closing them, as they're unlikely to continue to receive any attention. I think lying to our submitters by leaving an issue 'open' when no one intends to even look at it is worse than closing it prematurely. > <advocate type="devil"> > That cuts both ways, though -- I, as a developer, don't have time to debug a > problem in Rails. That problem isn't mine -- I didn't write the buggy code > that's causing the hassle. It's *your* fault, dammit. > </advocate> > > In reality, I know all about the OSS project philosophy, I'm a great > believer in it myself. If you want support, take out a support contract, > and all that. That's a complete mischaracterisation of the situation we're in, I'm more than willing to help people fix the problems they have. I spend a good portion of my time doing just that. We have too many reports, so many that it's impossible to make a dent in them. Reporters hear nothing back about their bug report, get disenchanted and leave. The patch queue fills up to the point we don't know where to start. If we get better turnaround time on patches and encourage contributors, and in the process prematurely close a few edge-case defects. So be it. > Going back to my original point, though, about actively *closing* tickets > that document *real* bugs that are trivially verifiable[1] simply because > the patch isn't sufficient is only reasonable in one circumstance -- if > defect reports without test cases are summarily closed. I haven't seen that > policy documented anywhere (and I would imagine something that it would want > to be in big letters on the front page), and in fact the exact opposite is > stated on the dev front page -- "Tickets are fine". I'm just reporting an > irritating inconsistency in the handling of tickets, biased against > well-meaning but insufficiently clued contributors (which are the sort of > people I'd be wanting to encourage, not discourage, if it were my project). Indeed, while reports without failing test cases aren't as likely to get attention, they're still valuable information, and could provide indications of where to start contributing for people who are keen to help out. Please don't close them unless you've tried to reproduce the bug, and are unable to do so. However I don't buy the sky is falling hypothesis of closing a ticket prematurely, just reopen it, explain it's still a bug. If the disagreement continues, start a thread here. -- Cheers Koz --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "Ruby on Rails: Core" group. To post to this group, send email to [email protected] To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/rubyonrails-core?hl=en -~----------~----~----~----~------~----~------~--~---
