Hi Marc,

On Mon, 2007-05-28 at 18:37 +0200, Marc Fargas wrote:
> Hi there,
> Since we got the triagers system patches and bugfixes are moving much
> faster than before but there are still somethings that get stuck,
> specially on the "Design Decission Needed" state.

One man's "stuck" is another man's "prioritised".

> In theory, if I got it right, triagers set this state when they think
> that the scope of a ticket should be brought to discussion on this list
> (django-developers). It's here, on this part of the diagram where it
> says "The community and/or core developers decide whether..." and the
> question arises: What if a ticket is brought to django-developers for
> discussion and there are no answers either positive or negative?

It doesn't happen too often for serious tickets. Unfortunately it does
happen sometimes and that is usually a sign that we are busy on other
things. The answer is "wait some more". Try again in another couple of
months if nothing has happened and try to pick a time when there isn't a
lot else going on.

This is the beauty of Open Source, you get to scratch your own itch and
so does everybody else and if your itch doesn't coincide with the core
developers' (and our "itches" implicitly include "helping others and
providing points of final decision making"), so be it; that happens
sometimes.

Not unsurprisingly, some of us do read every single post on the list. So
they don't go unnoticed. Personally, I will deliberately not comment on
a particular thread because I don't want to think about it and people
sometimes think that what I say matters. I really don't want to say "I
don't like this yet, but am not sure why", because then I'll be pressed
on the why and suddenly I have to spend some time backing up my
intuition and participating in a long thread about something that isn't
on my TODO list at the moment.

> Is the ticket left there to die? Or can triagers assume that if nobody
> has said anything in some period of time that no developer minds about
> the changes introduced by this ticket?

No. If it hasn't been changed it means that no decision has been made
yet.

> If you are thinking "What ticket is he talking about?" I'm talking about
> #1051 which has been around for 1 year and brought for discussion
> without success two times atleast.

Speaking only for myself, the reaosn I haven't done anything about that
ticket is that it is much lower priority than any of the other tickets
I'm thinking about or working on. I'm not unaware of it. It isn't a
trivial design decision, despite your various comments in the ticket
itself. For example, another ticket was closed as a duplicate of #1051
and the recent patches on that ticket don't address that duplicate's
issues at all (you seem to have dismissed them out of hand). There is
broader design thinking to be done there. How much support for schemas?
How much customisabilty is required/useful? What should the defaults be?
How does this fit in with other database backends? We try to commit
reasonably well thought-out solutions, not something that is going to be
in constant flux, so some of those decisions should be made up front and
that design work hasn't been done yet.

I didn't respond to your first email about this over the weekend,
because the reply I originally wrote wasn't particularly polite. At some
point, you have to accept that this ticket can wait a bit longer. It's
not like no action is happening anywhere in Django. You are in the queue
of things to be processed and being in the queue is better than not
being in the queue, at least.

Right at the moment, the lack of support for PostgreSQL schemas is not a
showstopper for the broader user base. It should even be a major
inconvenience for many cases. For the group of people who might need it
for policy reasons, they can apply the patch in the ticket and get on
with their lives. At some point one of the core developers will get
around to looking at the ticket. Staying in "design decision needed"
isn't a rejected state. It means "not yet".

There are many fish (bugs) in the sea here. Getting hung up on one
particular ticket isn't worth the stress.

> But, anyway, I think that the triaging process could be improved if the
> ones involved can define about this issue: What happens if nobody
> discusses a ticket when it's brought for that?
> 
> Cheers,
> Marc
> 
> PS: Note that "Design Decision" currently holds 290 tickets, not sure
> how many can be hanled by "accept/reject if nobody answers about that on
> django-developers" ;)

Many of them aren't really big issues, either. A lot of them are
enhancement requests and propsing ways to do things that are already
possible. We don't have a "one day/maybe" state (and before you say "so
we should have one", note that it's been suggested before -- you aren't
the only one who has threads with no resolution) and so we have to make
a yes/no decision early on, or leave it in design decision as a way of
saying "we haven't made a decision yet".

Be aware: looking at raw counts is a not uncommon trap to get stuck in,
particularly on Open Source projects, but it is a trap. Not all open
tickets are the same. When you encourage the broad userbase to file
tickets at will and when your userbase has a wide variety of skills an
experiences, you are going to get a lot of tickets that aren't problems
for everybody or are essentially enhancement requests making an
assumption that everybody works in a particular way or that are possibly
overlooking part of the bigger picture. These aren't bad bug reports and
we should be encouraging them because they sometimes point to bigger
things. But they also aren't of the same value/import as some other
tickets.

We try very hard to fix the high impact problems quickly and also to fix
everything correctly (robustly, good design, reusable... all those
motherhood and apple-pie features). Those two goals don't always align,
of course, so there are some high impact problems that have been open
for a while whilst we work on the "correctness" phase. But that's the
guiding principle that every decent programmer works towards implicitly
and if you look at the big picture, we really are achieving the same
here.

So whilst the open ticket counts always *could* be lower, it's not a
worry to me that they aren't. Note also, that they aren't increasing at
a great rate lately, despite a much larger number of trivial things and
that is where the triagers identifying that the trivial ones are
complete is helping a lot.

With best wishes,
Malcolm


--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups 
"Django developers" 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/django-developers?hl=en
-~----------~----~----~----~------~----~------~--~---

Reply via email to