Mantis Experiment a Failure? Are we ready yet to move
on?


Hi Frank,

Thanks for your support.

Hi Goran,

Thanks for your time and reply.  I knew my comments
might be taken as disparaging Gjallar. I have not
looked into Gjallar and don't have an opinion on it. 
I do have an opinion about the wisdom of integrating
it into squeak here and now. Which is what my comments
refered to. 

In the future it might be worth trying. But as you say
it wasn't written for squeak and there would be
efforts needed to integrate it. And they would have to
be your efforts. Which right now I hope to see applied
to elsewhere so we can get beyond a difficult
maintainence system for squeak.



***
>goran at krampe.se goran at krampe.se 
>Wed Nov 7 09:34:57 UTC 2007 replies:
> 
>> I don't want to eat more squeak project dog food. 
At
>> least not here. Not now.

Dog food is being forced to use as tools, projects
that are still in development. This gives feedback to
the development project at the cost of time and focus
for the work at hand. And puts pressure on the
developers of those tools to fix the problems
immediately. The best time to do this is when there is
"time to spare" and space for play.

>
>Gjallar is not built for the Squeak community. It
just happens to be an
>issue tracker, so it is not dog food "for its own
sake".
>
> 
>> Each problem in squeak seems to cause some in the
>> community to try to solve it with a squeak tool
that
>> hasn't been invented yet.
>
>Again, if you mean Gjallar (which you may not mean)
then it was not
>created for the Squeak community - it is a system
built for the needs of
>a customer.

I was talking generally and as you say Gjallar was not
made to solve a squeak problem. Which is to its favor.

The current problems and solutions to those problems
are both following this pattern. MC and its
repositories and browsers are such tools and so is
Delta Streams.

An annoyance followed by code to fix the annoyance is
squeaks method of evolution.

The problem right now is that we have a very
inapporpriate way of maintaining the current image
which has superceded the older known ways of doing it.
Until the next tool proves itself we will have a very
tortured way of producing releases.

>> All experience has shown
>> that integrating a new tool into squeak comes with
>> risks and problems and diverts efforts from finding
>> current bugs into finding the newly introduced
bugs.
>> 
>> I use mantis all the time it works fine for my
>> purposes, which is reporting, analysing and fixing
>> problems. 
>> The release teams use mantis and it works well for
the
>> purpose of finding fixes in a harvestable state.
>
>Mantis is quite fine for most stuff if you ask me.
But it fails for some
>things IMHO - mainly lack of email integration and/or
doesn't fit our
>rather distributed package world. 

I don't know if it lacks hooks for email integration
or if we just don't know how to access them yet.
Mantis has all sorts of capabilities we have not tried
to learn how to tap. 


><...> Some of this we (me and Matthew) want to
>fix with DeltaStreams in fact.

Yes. I am praying for and counting on your success.
> 
>> The meta-problem is not that mantis is not used by
>> many of the community because they have not caught
on
>> to is merits and usefulness.
>> The problem is simply that they have not caught on
to
>> its merits and its usefulness.
>
>Ehm... lost me there.
>
Um. Maybe a little too zen? In English: The problem is
we have to train them to use it. And they have to
learn. The mailing list is a good place to discuss a
problem but Mantis is a good place to solve it and to
archive the solution.

>> The other meta-problem is communication to and
>> training of the community.
>> 
>> Squeak-dev has scarce resources. Mantis is
maintained
>> by a large resourceful group of folks outside of
the
>> squeak community.
>> They can provide better support for a bug tracker
that
>> we can even if our development tools are better.
>> 
>> Maybe we are not using the mantis communication
>> resources we have in the best way?
>> 
>> Mantis not only allows accumulating information on
a
>> single topic. It can also write letters to those
who
>> should know about them.
>> But nobody is maintaining the list of reporters to
see
>> if we have current emails. Or live reporters for
that
>> matter. 
>> 
>> (This requires the same thing we do with mailling
>> lists send out occasional reminders and are you
still
>> there mail).
>
>Would just like to also mention that Gjallar has far
more advanced email
>capabilities than Mantis has.

Which people would not use any more than they use
mantis's. The problem with mantis's mail abilities is
the quality and maintainence of the list as much as
anything else.  Also, I would like to find a good nuts
and bolts users documentation for mantis. Right now
all I see seems aimed toward administrators rather
than reports, updaters and developers. As Gjallar is a
commercial project I wonder if it has overcome the
user's guide problem?

>> You could also sign up mailing lists as reporters
so
>> reminders (I E. bug reports) could be sent to those
>> lists. This would need to be done cautiously lest a
>> list get swamped with mantis spam. But it could be
>> done and it would increase communication and
>> awareness.
>> 
>> The current urgent problem in developing squeak
images
>> is the mess that the MC decision in 3dot9 made of
>> image maintainence.
>> That is where I would hope to see the effort of the
>> communities best and brightest go. 
>
>I try to pull my share of that particular load in the
DeltaStreams
>effort.

Please do. I am pinning my hopes on Delta streams.  


>
>> Please use mantis now. It serves its purpose well.
>
>I will never advocate Gjallar unless there are
willing souls helping out
>with such an endeavour btw. And also, it needs some
testing and
>adaptation etc.

That was my point. Right now I prefer a tracker that
leverages the resources of another community for its
maintainence issues.
>
>But Gjallar continues completely independent on the
Squeak community's
>need for issue tracking - it is driven mainly by
business opportunities.
>
A source for funding one's efforts always helps the
quality.

Again thanks for your response.

Yours in curiosity and service, --Jerome Peace


__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com