Re: [Gimp-developer] Interaction design. Used to be: Re: gimp gradients

[Simon Budig <simon budig de>]

Hi Peter.

Let me state as the first thing that I at first did not really try and
later had troubles following the dispute since I did not follow the
development of the UTT, had not read the spec and just knew your mails
from the mailinglist, basically pushing away the undertones because I
failed to understand what you were angry about.

(For me I think that I now have some grasp on what was going on there
and will try to describe it in detail later, because it is not at all
clear from the mailinglist alone...)

peter sikking (peter mmiworks net) wrote:
> the outreach, endowment, the relationships, they need to
> be recovered and this time explicitly, in writing, here on
> this list, not in gone-tomorrow talk on irc.

At the libre graphics meeting we had the usual gimp meeting and also
discussed the problems that currently seems to make it impossible for you
to continue working on GIMP.

In the meeting there was the general consensus that the collaboration
with you was very successful and we'd like it to continue. I really hope
that the differences between you and - I think - mostly Mitch can be
resolved and everybody can return to the fun parts of GIMP development.

Anyway, before this can happen we need to rehash what happened, try to
understand the issues and try to come up with a plan to resolve this.

Now - I understand that this problem is not just about UTT but the
general expectations on how we collaborate on the interaction design.
However, since the UTT seems to be the elephant in the room I'll focus
on that one for now, we should try to resolve this first.

Here is my short summary of what I believe has happened regarding the
UTT.

On Feb. 22 there was some discussion about the goals and designs for the
UTT and the other transform tools. Later in the (european) evening you
and mitch started discussing the current state in IRC. Mitch was
enthusiastic on how good the interaction on the UTT works and starts
talking about problems in the spec he and mikatchu faced when
implementing it. There are some parts of the discussion I still don't
really understand (the precision vs. feel stuff).

Then you apparently saw one or two screenshots and suddenly the tone of
the discussion went downhill - you apparently were quite pissed and I am
not sure if you actually read what mitch and mikatchu tried to explain
before you left the channel. Later you rejoined, giving some statements
on how you think the collaboration has to work and left.

Subsequently there were mails on the mailinglist where among other
things you expressed your frustrations, I am not going to rehash these
in detail.

I at that time peripherally followed the IRC discussion (I recall not
understanding what went wrong there) and the mails left me more and more
confused.

At LGM I compared the current state of the UTT with your spec and it
suddenly became apparent to me how it is possible that you take that as
a slap into your face and a stab in your back: The visual appearance of
the tool has nothing to do with what you specified, the handles have
been substituted with basically lots of line noise.

I imagine this must feel like the tailor delivering a quality business
suit tailored for the client and then witnessing the client taking the
scissors, randomly cutting holes for better ventilation.

Yes, I'd probably get pissed too.

In the defense of Mikatchu and Mitch however I don't believe this is
what happened.

They had troubles with the spec, found holes in it and focussed on the
parts they understood and left the rest for clarification together with
you. They implemented stubs for the visual part, implementing just
enough stuff with the stock-handle-code in GIMP to test-drive the mouse
handling, getting the link to the transform core correct etc. pp.

By no means the current visual appearance is supposed to be a "better"
replacement for your design, nobody claims that this is an improvement
over your design and noone intended to show you any middle fingers.

It is just that the uncertainties in the spec (e.g. how do the handles
behave while shearing/perspective changes) don't encourage the
implementation, since without an overall understanding - including the
corner cases - it is likely to have to be rewritten multiple times.

Then there is finally the chance of catching up with you, mitch wants to
clarify stuff, is in a hurry, you are apparently feverish, and a simple
screenshot makes things go sour, communication breakdown in about 20
minutes. Grgh.

(I'll send the relevant parts of the irc log to you off list, please
read it and try to look at it from an outsiders point of view...)

We're now in the weird situation that you're pissed towards - at least -
M&M, because they supposedly fail to respect your work and Mitch is
likewise pissed because you supposedly are unwilling to respect the work
from Mikatchu and him that has gone into the invisible part of the tool.

Frankly, I believe this aspect is something only a real-world-meeting
between mitch and you can resolve, I'll gladly donate the beer.


I sincerely hope that we can get this specific issue worked out and then
broaden the discussion to the "overall" collaboration. I won't do this
now, because this mail already is too long, and loading more potentially
loaded issues won't help.

I really want to stress, that - contrary to your perception - there is a
lot of trust in you and your work, I think the defense of the save vs.
export design has proven that. In fact it is not only trust but also
interest, which on the other hand means that sometimes your designs get
challenged. Please don't take that as questioning your competence.

Thanks,
        Simon

-- 
              simon budig de              http://simon.budig.de/