emacs-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Patch proposal: display symbol source code in help buffers


From: Arthur Miller
Subject: Re: Patch proposal: display symbol source code in help buffers
Date: Thu, 23 Sep 2021 00:38:28 +0200
User-agent: Gnus/5.13 (Gnus v5.13) Emacs/28.0.50 (gnu/linux)

martin rudalics <rudalics@gmx.at> writes:

>> That sounds like useful approach. Something like:
>>
>> (defcustom help-buffer-default-height 25
>>    "The default height of a help-buffer window."
>>    :type 'fixnum
>>    :group 'help
>>    :version "28.1") <-- :-)
>
> Help-buffer windows are as a rule displayed via 'display-buffer' and are
> easily recognizable because their buffer is always called *Help*.  So
> you can simply add a 'window-height' action alist entry for them which
> is considerably more versatile than what you propose above.

Yes, I know; but think of new users comming from other editors and applications,
having no experience with Lisp and never heard of conses and alists.

I use display-buffer-alist to achieve something with *Help* buffer (and some
others) myself. I posted that snippet in the other thread I don't know if all
that stuff  existed back in time when original poster posted his bug repport
(2011), but at least nowadays, the bug should be marked as resolved, since it is
possible to tell Emacs to reuse Help buffer and window, instead of opening new
ones.

However, I don't think my solution to that problem, same as you suggest here,
is very convenient to new users. People are used to see some option they can
toggle on/off, some slider they can pull left right, some value they can pull
from a list or enter into some box. I think it is more helpful to see some
variable with clear and descriptive name, I think new users would appreciate
that. But that is just my opinion.

Anyway, I have read both threads you have refered too. One of them, the
Florians, (Bug#9054), is completely unrelated to what this patch does. Well it
is about help buffer, but there you are concerned with wath happens *before* 
user
enters the help buffer and in which way user will come to that buffer.

>        Where to pop up the location of the source and how to get rid of
> it is a question we currently discuss in Bug#9054

And that is exactly what I say above; this patch is completely not concerned
with where and how you will pop that window and source code file. I am concerned
with what happens in a help buffer, after user requested help, *not* how user 
come
to that point. I don't even pop a source file; just code for a function or var.

I wish to make help buffer more usable, because I wisth to go away from Helpful
and I see no reason why we should wait for something unrelated you guys had 10
years to get consensus on but didn't :). Sorry, that one was hard not to pick on
:).

I know I sound biased, but the suggested patch works above my personal
expectations. Suddenly the help buffer acts like a code browser. I really
suggest you to try and click a bit around. It wasn't my intention, since I never
do that in Helpful. I even don't know if it is possible there, but due to
back/forward functionality in built-in help buffer, it works really nice.

There is one bug I know off: it can't display code for *some* aliases, for
example 'find' which is alias for cl-find. Some other aliases works
fine. Besides that, I am very happy with it.

>                                                                   If and
> when we reach a consensus there, we can add an option to immediately
> display the source in the *Help* buffer or a window right below it

I would like to cite Eli here from the other thread, and say "life is too
short." :)

I am sure you guys will come up with something wonderful and amazing, but untill
you reach concensus, I'll go and work on some other options, because the
proposed patch is my proposal. It came up completely unrelated and independent
of that discussion, but it is touching the things you are discussing there. If
the code is bad you can improve on it and rework it, but as I understand my
proposal seems not be accepted, since I see you guys discuss options in the 
other
thread (Bug#36767), which I have already implemented and demoed. Nobody but
Tassilo has commented on, so I guess nobody has neither cared to try it nor see
the demo.

Anyway, thanks for your input in the other comment, it was helpful.




reply via email to

[Prev in Thread] Current Thread [Next in Thread]