Frank,
	Thanks for your interest and thoughtful reply.

>Peace Jerome <peace_the_dreamer at yahoo.com> wrote:
>> 
>> Are what the janitors are doing, what needs to be
>> done?
>> 
>> The janitors are doing work that is useful and is
that
>> the work that needs to get done to fix bugs and
have a
>> better squeak?
><snip>
>> When I first saw the janitor project I thought  it
>> would be one I could join and add my bug fixing
skill
>> to. But as I looked at the early stuff from the
list I
>> realized this wasn't what the janitors intended to
be.

Frank replied:
>
>Janitors make sure that the people who need to hear
about bugs, well,
>hear about the bugs. In other words, if you I find a
bug in, say,
>Scamper, and send a bug report to the list, one of
the janitors looks
>at the bug, tries to reproduce it, contacts me if
there's a problem
>trying to reproduce the bug. If it turns out the
bug's not a bug at
>all, the janitor closes the report (making sure I,
the reporter, know
>it's closed). If the janitor can reproduce the bug,
he or she adds it
>to Mantis and contacts the maintainer/s of the
affected package (in
>our example, the Scamper maintainer/s).
>
>I (as a Janitor bystander) had made the same mistake,
thinking of
>Janitors as sweeping up the first & whatnot. Ken
corrected me (I hope
>you don't mind me quoting you, Ken?), saying this:
>
>"I don't agree with this.  My goal for this team has
nothing to do with
>fixing bugs or reviewing or any of that stuff.  It is
simply to get bug
>reports and fixes from those that have them to those
that want them and
>can make use of them.  A sorting function with some
filtering included."
>
>In the longer term, the Janitors aim to make the
process of Harvesting
>as easy as possible. That's not the same as actually
harvesting!

My point exactly. The Janitors task is like the guy
who lost his keys in the dark of the parking lot who
chooses to look for them under the lamp post because
the light is better. Its easier but will it get the
job done?

>> Quite frankly, My perception is the janitors will
>> generate a lot of noise and result in very little
of
>> the actual problems being solved. It will obscure
the
>> fact that what is needed is problem solvers and bug
>> fixers and not just text processors.
>
>Too many bug reports languish in obscurity amongst
the ~70 posts a day
>this list usually generates. We have a central bug
database, Mantis,
>but right now you can't just hit "mail out a bug
report" off your
>Debugger, say, and have it appear in Mantis. How do
these bugs
>actually end up in Mantis?

Well why not let that be the responsibility of the
reporter?

>The Janitors. Who makes sure that bugs
>aren't re-reported, that they're reproducible (i.e.,
actual bugs and
>not the result of someone toasting their image), etc?
The Janitors.

Anyone has the ability to get a bug reporting account
on mantis. So the janitors are not really able to
filter out the unreproducible, toasted image bugs from
the real ones. They can only act as filters if the bug
is not reported to mantis in the first place.

>
>Yes, they do produce a number of mails per bug report
to the list (at
>least two: the [caught] and the [closed] mails).
That's temporary,
>until we figure out the proper way for people to
report bugs easily.
>
>> Right now most of my bug reports sit on mantis
>> unperused and unacknowledged. Often the resolution
is
>> to leave them open. Sometimes if the fix is small
>> enough (cause the problem is small) somebody will
>> grasp that this can be added to squeak and it is.
But
>> the bigger problem is making a better squeak.
>
>I think (being somewhat in this position) that if
your fix is easy,
>small, etc., then mail someone you think is in
charge. 
><snip>
What I was trying to say was that the small fixes
stand a chance of being attended to and getting fixed.

The larger fixes. The ones I’ve spent a reasonable
amount of time understanding, explaining and fixing
(e.g. mantis #503 fixing slider re-targeting ) can’t
seem to find a champion or anyone beside myself
interested. And I have written Ned about it. 

The larger the fix the less likely someone will be
willing to adapt it. Instead the bug remains and
assists in the writing of more bugged code. I chose
mantis #503 deliberately because the broken code from
sliders was coped verbatim and became broken code in
button re-targeting as well. I can fix both but there
is no sense in putting time into extending the fix
when the real problem is to get the commitment on the
part of the code maintainers to care about including
the fixes in the image. 



IMHO the goal is to solve the problem of getting
things 
	1) fixed and 
	2) included in the image with 
	3) the least amount of effort on everyones part.

>
>> There is a need for recruiting the people who will
>> take the bug reports and turn them in to fixes that
>> can be committed to squeak. I am not in a position
>> from where I sit to see how that will be done.
>
>I agree. And once we have those fixes, we need a
group of people who
>can take those fixes and push them to the update
stream. (Again, the
>external package fixes are easier: nag the maintainer
of the
>package). Right now, I don't even know who can update
the stream, for
>instance. I'm sure these issues have been
exhaustively discussed
>before on the list.

Yeah, but I think since then the November revolution
has happened twice I think.

Thanks again for your interest and thoughtful reply.

Yours in service, --Jerome Peace




		
__________________________________ 
Do you Yahoo!? 
Yahoo! Small Business - Try our new resources site!
http://smallbusiness.yahoo.com/resources/