What custom is and does

[Alex Schroeder <alex@emacswiki.org>]

"Robert J. Chassell" <bob@rattlesnake.com> writes:

> Please explain in more detail.  I cannot infer your implications.

I guess the problem is that do not understand what your problem with
custom is.  Therefore, what Per says makes perfect sense to me but I
cannot explain it to you -- it seems that you two are talking about
two different problems.  Per wrote what he did in response to
something Miles wrote:

Miles Bader <miles@gnu.org> wrote:

> Agreed.  Customize should be a friendly face for the existing
> infrastructure, not a distant relative...

Perhaps the two of you understood two different things, there?  I
spent quite a bit of time explaining my mental model of custom, and
its relation to setq statements.  Perhaps in your question, you could
take my answer and tell me what questions this leaves unanswered, or
what implication you do not follow.

The way I saw it, I said that custom-set-variables was only remotely
connected to setq.  You can use custom-set-variables as a complex and
unreadable variant of hand-written setq statements.  I explained that
this was a very simplified mental model and that it would not be good
enough to explain more powerful features.

Per's position is that custom is a very powerful library to build a
user interface for the customization of Emacs.  He also said (more or
less) that to look at custom as just another way to write elisp into
your .emacs and to expect it to have the very same qualities
hand-written elisp such as readability, modifyability, etc. was to
constrain the powers of custom.  He seems to be saying that working on
this part of custom was worth less than working on extending custom in
new directions (such as keymaps, wizards/saints, etc.).  His point is
that custom can be used to build a customization infrastructure for
Emacs that has very little to do with the underlying lisp variables
and functions.  This is why he does not mind two variables instead of
one -- because custom can be used for greater things and by more
newbies that know nothing of lisp.  Per might argue that to use the
underlying lisp code as an argument in discussions about custom was an
indication of the lack of understand people have for the potential of
custom.

To this you responded along the lines that you consider this lack
readability, modifyability, of the saved data as a bug and it seemed
that you wanted people to work on this with rather high priorities.

Perhaps you need to explain your priorities, or perhaps you need to
restate the problem.  Or perhaps we need to have a discussion about
what custom is and does and what custom could and should be.  I have
read lots of posts by Per so I basically know where he is coming
from...  :)  I just don't know where you stand in all of this.

Alex.