Re: [Nautilus-list] Re: [Nautilus-test] Enabling "QA Contact"/"Verifier" in Bugzilla. Thoughts?



On Mon, 16 Oct 2000, Eli Goldberg wrote:

> Just so everyone knows, Darin has turned on the QA Contact field.
> 
> I agree that right now the bonobo bustage everywhere makes it less desirable for
> people to specialize, as you noted, it's only a short-term blocker. A list of
> components and owners for Nautilus is at
> http://bugzilla.eazel.com/describecomponents.cgi?product=Nautilus
> 
> ...so, right now, it's me, Albert, jfleck, Josh, and 'nobody'. ;)
> 
> I agree that it would be fantastic if non-Eazel component owners wanted to post
> regular test status to nautilus-test. Right now, I'll gladly settle for 'takes
> component, verifies & pre-screems a lot of bugs, and maybe other stuff that they
> feel like'. ;)

Ummm - looking at the above URL I seem to be the only non-Eazel component 
owner. Anyway, it appears that jfleck is my designated tester so he 
should be the one on nautilus-test ;)
 
> Thanks!
> 
> Caveats:
> 
> * Only new bugs get a QA Contact. I need to go through and bulk-reassign several
> thousand bugs. This would result in a huge amount of spammage. Will connect with
> Darin for a plan. (One solution would be for me to do it from 7:00-8:00 AM on
> Wednesday,  when nobody is working, so that everyone can just delete their
> Bugzilla messages during that time.)
> 
> * It's currently called "QA Contact"; this will eventually change to "Verifier".
> 
> 
> 
> will lashell net wrote:
> 
> > I think this is a great idea, because it gives the engineers an easy way
> > to check on fixes, and to get some feedback.
> >
> > However, here are a couple of issues that exist currently
> > that would prevent this from happening easily right now.
> >
> > The bonobo breakage that happened with this latest merge is
> > widespread enough, and frequent enough that its hard to even
> > get to one thing to test it without some other glaring bug
> > rearing its ugly head. For instance i just had nautilus crash
> > when I was changing a directory. And menu's are broken, etc etc.
> >
> > This makes it hard to test many different features and opportunities
> > to crash nautilus. Now, I love wrecking nautilus as much as the next
> > guy, but I wouldn't feel right at this point only feeling truely obligated
> > to a certain feature set, or even putting extra emphasis when this is
> > happening.
> >
> > I realize this is only temporary as John, Darin, Maciej etc all beat the
> > crap out bonoboish bugs and make nautilus as stable (relative) as it was
> > before the merge.
> >
> > If we do decide to implement this structure eli, it would be really valuable
> > I think to have weekly(?) status reports from the testers responsible for
> > a component or feature set, that gets posted to nautilus-test. The reasoning
> > behind this is that if I'm testing a lot on my component, I would still like
> > to see how things are going etc.
> >
> > One side benefit that having noted testers being responsible for a component
> > would be that testing contributors have someone they -know- will be able to
> > check and try to verify the reproducibility of a bug with them, or maybe answer
> > questions etc.
> >
> > This being said, I would love to feel "responsible" for the testing of a
> > component, but I'm having a hard time picking one? *grin* Any ideas for
> > that?
> >
> > Will
> >
> > On Fri, Oct 13, 2000 at 10:39:34AM -0700 or thereabouts, John Sullivan wrote:
> > > Hi Eli,
> > >
> > > Your reasoning seems sound to me. I particularly like the fact that you
> > > mentioned both pros and cons, enabling the reader to make an informed
> > > opinion. The main reason we didn't turn it on originally was just because we
> > > didn't know how we would use it. I don't have any objections to turning it
> > > on now that we have QA people who can gain benefit from it.
> > >
> > > John
> > >
> > > on 10/13/00 9:21 AM, Eli Goldberg at eli eazel com wrote:
> > >
> > > > Hi!
> > > >
> > > > Unlike most Bugzilla installations that I've seen, Nautilus hasn't
> > > > used the "QA Contact" field. Originally, this was because until 8 weeks
> > > > ago, there weren't any QA contacts on the project. ;)
> > > >
> > > > The "QA Contact" field is a person who owns the testing (or
> > > > sometimes, just the bug-handling) of a particular feature, just as an
> > > > engineer owns the coding of the feature. So, a QA contact would be
> > > > responsible for verifying bugs in their component, as well as helping
> > > > the developer, such as with bug reproduction.
> > > >
> > > > Now, there are two people testing services (Beraj & Albert), along
> > > > with Josh Barrow (Search & builds), myself, Will (a bit of everything
> > > > ;), and a lot of other people testing the client whose names are eluding
> > > >
> > > > me, along with Victor (who does a bit of testing of everything.)
> > > >
> > > >
> > > > I'd like to suggest that we enable the QA Contact/Verifier field on
> > > > the following basis:
> > > >
> > > > * It's currently difficult for people to keep up with bugs in the
> > > > components in which they specialize, since bugs are often reported
> > > > without their knowing.
> > > >
> > > > * Sometimes, engineers need help on a bug, and testers don't know
> > > > about it, since we're not automatically CC'd.
> > > >
> > > > * It provides an entry point for people interested in testing to
> > > > plant a stake in Nautilus and get have personal ownership and
> > > > responsibility for a feature area.
> > > >
> > > > e.g. after Joe Tester is involved for a few weeks and decides he's
> > > > having fun and would like to proceed beyond playing with Nautilus and
> > > > verifying occasional bugs, he could take over de facto ownership of a
> > > > feature or two he wished to test.
> > > >
> > > >
> > > > Reasons why this might not be such a good idea to do yet:
> > > >
> > > > * One can just as easily keep up with new bugs in particular
> > > > components by using queries. We don't need an extra form field for that.
> > > >
> > > > * If an engineer really needs help, they know to CC: a tester.
> > > >
> > > > * There's still only one full-time tester on Nautilus. Most people
> > > > interested in testing Nautilus are still getting their feet wet, and
> > > > trying to build the darned thing, let alone actually work with it. ;)
> > > > Perhaps we should wait another month or two?
> > > >
> > > > Thanks,
> > > >
> > > > Eli
> > > >
> >
> > -----------------------------------------------
> >
> > "Doh.. STUPID STUPID STUPID"
> >
> > --an excerpt taken from every developers cvs
> >   commit message at some point
> > -----------------------------------------------
> >
> > _______________________________________________
> > Nautilus-test mailing list
> > Nautilus-test lists eazel com
> > http://www.eazel.com/mailman/listinfo/nautilus-test
> 
> 
> 
> _______________________________________________
> Nautilus-list mailing list
> Nautilus-list lists eazel com
> http://www.eazel.com/mailman/listinfo/nautilus-list
> 





[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]