From: JD Smith <jdtsmith@gmail.com>
To: Eli Zaretskii <eliz@gnu.org>
Cc: 68236@debbugs.gnu.org, Stefan Kangas <stefankangas@gmail.com>
Subject: bug#68236: [PATCH] help.el: allow help-quick to use local commands/quick-sections
Date: Fri, 5 Jan 2024 12:00:32 -0500 [thread overview]
Message-ID: <F4B94E6F-8841-4AE6-B123-02B89EE22DAE@gmail.com> (raw)
In-Reply-To: <835y08w4v1.fsf@gnu.org>
[-- Attachment #1: Type: text/plain, Size: 3792 bytes --]
> On Jan 5, 2024, at 3:40 AM, Eli Zaretskii <eliz@gnu.org> wrote:
>
>> From: JD Smith <jdtsmith@gmail.com>
>> Date: Thu, 4 Jan 2024 20:28:27 -0500
>> Cc: 68236@debbugs.gnu.org
>>
>>> On Jan 4, 2024, at 8:57 AM, Eli Zaretskii <eliz@gnu.org> wrote:
>>>
>>> If we are going to expose help-quick-sections as a defcustom, then I
>>> don't understand why we need to change the code at all. Is the idea
>>> that sections will depend on the current buffer? If so, then we just
>>> need to add an element to the list members which will store the
>>> major-mode for which the member is relevant.
>>>
>>> Or what am I missing?
>>
>> Right now the code does
>>
>> (with-current-buffer (get-buffer-create "*Quick Help*")
>>
>> right away, then checks `where-is-internal' for each listed command in `help-quick-sections'. So only global bindings (and bindings available in help-mode) are accessible for display. My patch simply delays switching to *Quick Help* buffer, so that binding information can be gathered from the local buffer from which quick help was summoned. Note that help-quick omits any bindings that are nil, as well as any empty sections. So adding sections to the defcustom that do not apply (=have no bindings) in some buffer is not a problem.
>
> Your proposal has the disadvantage that the user must switch to a
> buffer under some major mode to see the entries for that mode in the
> quick-help window. It could be an annoyance; e.g., consider a user
> who wants to see this while in a *Help* buffer. And I don't think
> being in the buffer under the major mode is the only way of getting
> mode-specific bindings; for example, where-is-internal can accept a
> KEYMAP argument, which will be used to find key bindings.
The current disadvantage is related, but much worse than this: you currently cannot configure quick help to show any bindings other than global and help-mode bindings. It might be nice to see e.g. org-bindings from anywhere, but to me it’s an advantage to show “quick help for this mode”.
> Or maybe we should have a separate command for cheat sheets specific
> to a major mode. The window we pop up cannot be too large, so if the
> user only wants a quick help for the current mode, she might consider
> global bindings an annoying waste of screen estate.
I’d probably disable most of the global bindings, since I already know those. But some are useful (e.g. I often forget the project bindings).
> Moreover, the
> current quick help shows "popular commands", which are likely to be
> already known to some users, whereas when the user works in a major
> mode that is new to the user, one is likely to be in the need of the
> cheat sheet for that one mode. (Yes, we do already have "C-h b", but
> the output of that could be overwhelming: for example in an Org buffer
> I get almost 1400 lines in the *Help* buffer showing the Org-specific
> bindings.)
Agree, this is a real problem for discoverability. Also command names are not always informative enough to find what you are actually looking for (org is terrible for this).
> IOW, if we want to consider mode-specific quick help, we should
> perhaps discuss more about the goals before we consider code tricks to
> implement it.
I do like this idea. Perhaps mode authors could add a “starter pack” of this information themselves, if they are suitably admonished to keep this information brief, high level, and with clear few-word command descriptions. Perhaps adding a ‘help-quick property to the mode command symbol?
One guiding principle I suggest: users in the end should be able to decide what quick help they need; what’s memorable for some may be frequently forgotten for others.
[-- Attachment #2: Type: text/html, Size: 20149 bytes --]
next prev parent reply other threads:[~2024-01-05 17:00 UTC|newest]
Thread overview: 13+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-01-04 3:08 bug#68236: [PATCH] help.el: allow help-quick to use local commands/quick-sections JD Smith
2024-01-04 7:39 ` Eli Zaretskii
2024-01-04 13:45 ` JD Smith
2024-01-04 13:57 ` Eli Zaretskii
2024-01-05 1:28 ` JD Smith
2024-01-05 8:40 ` Eli Zaretskii
2024-01-05 17:00 ` JD Smith [this message]
2024-01-10 12:51 ` Stefan Kangas
2024-01-10 13:53 ` Eli Zaretskii
2024-01-10 15:46 ` JD Smith
2024-01-10 15:50 ` Stefan Kangas
2024-01-10 15:58 ` Eli Zaretskii
2024-01-10 22:49 ` JD Smith
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
List information: https://www.gnu.org/software/emacs/
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=F4B94E6F-8841-4AE6-B123-02B89EE22DAE@gmail.com \
--to=jdtsmith@gmail.com \
--cc=68236@debbugs.gnu.org \
--cc=eliz@gnu.org \
--cc=stefankangas@gmail.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
Code repositories for project(s) associated with this public inbox
https://git.savannah.gnu.org/cgit/emacs.git
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).