dynamic-completion-mode doc

dynamic-completion-mode doc

[Madhu]

[written up for gnu-emacs-help but posted here on second thoughts]

I'm trying to find out what (dynamic-completion-mode) does.

C-h f dynamic-completion-mode
C-h v dynamic-completion-mode

Only show the boilerplate documentation.  But I cannot understand what
the impact of the command is, or what changes when minor-mode
dynamic-completion-mode is turned on or off.

[I've used what is called "dynamic completions" before:

1. use M-/ dabbrev-exapand to complete the word at point

2. Use completion-table-dynamic to support completion-at-point for lisp
modes like sly or slime

3. I used comint-dynamic-complete-as-filename for a long time until its
functionality was handicapped and then obsoleted then removed, and I've
had to locally revert the changes to "comint-filename-completion"
"comint-dynamic-complete-filename" to get the functionality back.

This makes me think I've misunderstood what "dynamic completion" means,
so perhaps I should start with what dynamic-completion-mode is supposed
to do.

Re: dynamic-completion-mode doc

[Visuwesh]

[வியாழன் செப்டம்பர் 07, 2023] Madhu wrote:

> [written up for gnu-emacs-help but posted here on second thoughts]
>
> I'm trying to find out what (dynamic-completion-mode) does.
>
> C-h f dynamic-completion-mode
> C-h v dynamic-completion-mode
>
> Only show the boilerplate documentation.  But I cannot understand what
> the impact of the command is, or what changes when minor-mode
> dynamic-completion-mode is turned on or off.

Does not the package commentary of completion.el answer this question?

Re: dynamic-completion-mode doc

[Eli Zaretskii]

> From: Madhu <enometh@meer.net>
> Date: Thu, 07 Sep 2023 10:10:14 +0530
> 
> [written up for gnu-emacs-help but posted here on second thoughts]

It should have been submitted as a bug report instead.

> I'm trying to find out what (dynamic-completion-mode) does.
> 
> C-h f dynamic-completion-mode
> C-h v dynamic-completion-mode
> 
> Only show the boilerplate documentation.

Well, it did say

  Toggle dynamic word-completion on or off.

And that "word-completion" part is not really boilerplate.  But yes,
the doc string was clearly insufficient; I've fixed that now on the
emacs-29 release branch.

You can find the full documentation of this mode in the commentary at
the beginning of completion.el.

> But I cannot understand what the impact of the command is, or what
> changes when minor-mode dynamic-completion-mode is turned on or off.

After turning the mode ON, type at least 3 characters of some symbol
and press M-RET, perhaps several times.  You will see the mode in
action.

RE: [External] : dynamic-completion-mode doc

[Drew Adams]

> [written up for gnu-emacs-help but posted here on second thoughts]
> 
> I'm trying to find out what (dynamic-completion-mode) does.
> 
> C-h f dynamic-completion-mode
> C-h v dynamic-completion-mode
> 
> Only show the boilerplate documentation.  But I cannot understand what
> the impact of the command is, or what changes when minor-mode
> dynamic-completion-mode is turned on or off.
> 
> [I've used what is called "dynamic completions" before:
> 
> 1. use M-/ dabbrev-exapand to complete the word at point
> 
> 2. Use completion-table-dynamic to support completion-at-point for lisp
> modes like sly or slime
> 
> 3. I used comint-dynamic-complete-as-filename for a long time until its
> functionality was handicapped and then obsoleted then removed, and I've
> had to locally revert the changes to "comint-filename-completion"
> "comint-dynamic-complete-filename" to get the functionality back.
> 
> This makes me think I've misunderstood what "dynamic completion" means,
> so perhaps I should start with what dynamic-completion-mode is supposed
> to do.

Others have pointed you to the Commentary in the
Lisp file, `completion.el'.  That's really all
there is, I'm afraid.  It's relatively complete,
but it, and the library itself, was written many
moon ago, and the presentation is not as good as
one might expect nowadays.

That said, it's a good, interesting, and still
useful library.  But it's not well known, to put
it mildly.  It could really use some love and
understanding - some "modernization", taking
advantage of the last 30 years of Emacs dev.

I have hopes someone might get interested in
that, e.g., leveraging its approach/features to
either improve it or improve some other Emacs
library/feature.  Its description might give the
impression that it's the same or similar to some
other existing approaches, but I don't think so.
It might at least serve as food for thought...

It's really quite simple to use, but yeah, you
need to read the Commentary and give it a try.

Re: [External] : dynamic-completion-mode doc

[Madhu]

*  Drew Adams <drew.adams@oracle.com>
 <SJ0PR10MB5488D1CB450A1EBF0C3913C4F3EEA @SJ0PR10MB5488.namprd10.prod.outlook.com>
Wrote on Thu, 7 Sep 2023 13:58:52 +0000
> That said, it's a good, interesting, and still useful library.  But
> it's not well known, to put it mildly.  It could really use some
> love and understanding - some "modernization", taking advantage of
> the last 30 years of Emacs dev.
>
> I have hopes someone might get interested in that, e.g., leveraging
> its approach/features to either improve it or improve some other
> Emacs library/feature.  Its description might give the impression
> that it's the same or similar to some other existing approaches, but
> I don't think so.  It might at least serve as food for thought...
>
> It's really quite simple to use, but yeah, you need to read the
> Commentary and give it a try.

Thanks, yes, I have used this package in the 90s but not in the last 2
decades and didn't spot any doc on it (which is probably fixed now,
thanks Eli)