pcplantdb

[John Schinnerer <john@eco-living.net>]

Aloha,

> I'm not sure. It seems like the only way a bug can be verified is by 
> being closed, that takes it out of the regression cycle.

Aha, maybe I see where we're missing each other...what do you mean when 
you say 'verified'?

> How about:
> (bug,resolved) -> (bug,verified) -> (bug,closed)

I would say 'bug fix verified' in the middle of the above.
Is that what you mean?

> Where (bug,verified) means its been checked and is OK,

Checked how?  That is, what's OK,the bug, or the fix?

If the fix worked, the bug will no longer be reproducible - will no 
longer exist - and will then be closed.

If the fix didn't work and the bug can still be reproduced - still 
exists - then it will go back to 'open' and be assigned back to the 
developer to try again.

> but we do not 
> want to close it as we wish to reopen it in the next release.

If the bug is fixed - gone - then this would not make sense.

If the bug is not fixed - not gone - then this is exactly what I am 
saying would happen.  It would go back to open and back to the developer 
to try again.

> On the JEP project I'm involved with in the forum (often used for bug 
> reporting) there seems to be about equal numbers of posts from logged in 
> users/anonymos users. 90% of bugs come through the forum, 1% or less 
> actually use the source forge issue tracker.

So are you saying you'd rather just have some kind of blog/forum for bug 
reporting?

SF issue tracker is pretty lame/limited IMO so perhaps that's why most 
don't use it.

> Because thats the way people work. The more steps you need to
>  go through to participate the less people will. Take all the "free 
> registration required" sites such a NYT, you have two options
>
> 1) spend 5 mins registering, try to remember yet another password, worry 
>  about privicy.

> 2) forget about it
>
> In case 1 we get the bug report, case 2 we don't :-(

What are the quality levels of bugs entered in case 1) vs. case 2)??

My experience is that even a lot of paid "professional" testers write 
pretty crappy bugs, even using a system that really tries to constrain 
them to write at least decent bugs.

My experience is also that crappy bugs are often more trouble than no 
bugs at all.  Red herrings, non-reproducible stuff, lack of basic info 
needed to reproduce, etc.

In short, I think we disagree significantly on some testing/bug 
reporting/tracking issues.

Everyone please chime in on this issue.  Most basic question IMO is:

Do you think our user community will be willing to register and use the 
issue tracker or not?

If not, what do you propose?

thx,
John S.


--

John Schinnerer - MA, Whole Systems Design
------------------------------------------
- Eco-Living -
Whole Systems Design Services
People - Place - Learning - Integration
john@eco-living.net
http://eco-living.net