bug#50775: ispell help message times out before most people can read it

bug#50775: ispell help message times out before most people can read it

[積丹尼 Dan Jacobson]

I typed
 <escape> $		 ;; ispell-word
then I typed
 ? ? ? ?
in fact I need to type even more of them.

That's because the three line message that "?"
shows times out in one second.

But some people need more than one second to read it.

So I have a better idea.

Make it into a four line message, so that way there would be no
need to switch back to the original prompt, because the original prompt
would now be at the bottom, just like before.

emacs-version "27.1"

bug#50775: ispell help message times out before most people can read it

[Eli Zaretskii]

> From: 積丹尼 Dan Jacobson
>  <jidanni@jidanni.org>
> Date: Fri, 24 Sep 2021 21:13:18 +0800
> 
> I typed
>  <escape> $		 ;; ispell-word
> then I typed
>  ? ? ? ?
> in fact I need to type even more of them.
> 
> That's because the three line message that "?"
> shows times out in one second.

For me, it times out in 5 sec, not one.

bug#50775: ispell help message times out before most people can read it

[積丹尼 Dan Jacobson]

EZ> For me, it times out in 5 sec, not one.

In fact you're right!

But it is way to short for many of us.

Indeed it violates web accessibility.

bug#50775: ispell help message times out before most people can read it

[Eli Zaretskii]

> From: 積丹尼 Dan Jacobson <jidanni@jidanni.org>
> Cc: 50775@debbugs.gnu.org
> Date: Fri, 24 Sep 2021 21:52:34 +0800
> 
> EZ> For me, it times out in 5 sec, not one.
> 
> In fact you're right!
> 
> But it is way to short for many of us.

It's a simple matter of increasing the timeout, since when the user
touches any key, the help goes away.

bug#50775: ispell help message times out before most people can read it

[積丹尼 Dan Jacobson]

EZ> It's a simple matter of increasing the timeout,

Maybe just use no timeout:

EZ> since when the user touches any key, the help goes away.

Anyway, slower readers would probably have a hard time figuring out how
to increase the timeout.

bug#50775: ispell help message times out before most people can read it

[Juri Linkov]

>> EZ> For me, it times out in 5 sec, not one.
>>
>> In fact you're right!
>>
>> But it is way to short for many of us.
>
> It's a simple matter of increasing the timeout, since when the user
> touches any key, the help goes away.

I have another problem with ispell timeouts: oftentimes text contains
characters that mismatch the charsets used by ispell.  Then ispell
waits 5 sec for every such character to display a useless error message.
I used such change to minimize the time spent looking at these messages:

diff --git a/lisp/textmodes/ispell.el b/lisp/textmodes/ispell.el
index b650ab3871..f940dad460 100644
--- a/lisp/textmodes/ispell.el
+++ b/lisp/textmodes/ispell.el
@@ -2723,7 +2723,7 @@ ispell-parse-output
    ((equal 0 (string-match "[\ra-zA-Z]" output))
     (ding)				; error message from ispell!
     (message "Ispell error: %s" output)
-    (sit-for 5)
+    (sit-for 1)
     nil)
    (t					; need to process &, ?, and #'s
     (let ((type (aref output 0))	; &, ?, or #
-- 

bug#50775: ispell help message times out before most people can read it

[Eli Zaretskii]

> From: Juri Linkov <juri@linkov.net>
> Cc: 積丹尼 Dan Jacobson <jidanni@jidanni.org>,
>   50775@debbugs.gnu.org
> Date: Fri, 24 Sep 2021 18:39:03 +0300
> 
> >> EZ> For me, it times out in 5 sec, not one.
> >>
> >> In fact you're right!
> >>
> >> But it is way to short for many of us.
> >
> > It's a simple matter of increasing the timeout, since when the user
> > touches any key, the help goes away.
> 
> I have another problem with ispell timeouts: oftentimes text contains
> characters that mismatch the charsets used by ispell.  Then ispell
> waits 5 sec for every such character to display a useless error message.
> I used such change to minimize the time spent looking at these messages:

Are you still using Ispell the program as the spell-checking back-end?
AFAIK, this only happens with that old speller, but not with Hunspell
or Aspell.

That said, I have no problem with the proposed patch, since I don't
think this message should happen in any reasonable use nowadays.

bug#50775: ispell help message times out before most people can read it

[Juri Linkov]

>> I have another problem with ispell timeouts: oftentimes text contains
>> characters that mismatch the charsets used by ispell.  Then ispell
>> waits 5 sec for every such character to display a useless error message.
>> I used such change to minimize the time spent looking at these messages:
>
> Are you still using Ispell the program as the spell-checking back-end?
> AFAIK, this only happens with that old speller, but not with Hunspell
> or Aspell.
>
> That said, I have no problem with the proposed patch, since I don't
> think this message should happen in any reasonable use nowadays.

I'm seeing these errors with hunspell.  Maybe it's possible to configure it
to avoid errors about wrong encodings, I don't know.

bug#50775: ispell help message times out before most people can read it

[Eli Zaretskii]

> From: Juri Linkov <juri@linkov.net>
> Cc: jidanni@jidanni.org,  50775@debbugs.gnu.org
> Date: Sat, 25 Sep 2021 21:57:44 +0300
> 
> > That said, I have no problem with the proposed patch, since I don't
> > think this message should happen in any reasonable use nowadays.
> 
> I'm seeing these errors with hunspell.  Maybe it's possible to configure it
> to avoid errors about wrong encodings, I don't know.

You mean, your dictionary isn't in UTF-8?

Or maybe you have a buggy Hunspell, which counts bytes instead of
characters?

bug#50775: ispell help message times out before most people can read it

[Lars Ingebrigtsen]

積丹尼 Dan Jacobson <jidanni@jidanni.org> writes:

> That's because the three line message that "?"
> shows times out in one second.
>
> But some people need more than one second to read it.

I've now added a new user option 'ispell-help-timeout' so that you can
control this.

> So I have a better idea.
>
> Make it into a four line message, so that way there would be no
> need to switch back to the original prompt, because the original prompt
> would now be at the bottom, just like before.

Hm...  That's a good idea, I think.  But how does that affect people
that have a lower number on the max height of the echo area?

-- 
(domestic pets only, the antidote for overdose, milk.)
   bloggy blog: http://lars.ingebrigtsen.no

bug#50775: ispell help message times out before most people can read it

[Stefan Kangas]

Lars Ingebrigtsen <larsi@gnus.org> writes:

> 積丹尼 Dan Jacobson <jidanni@jidanni.org> writes:
>
>> That's because the three line message that "?"
>> shows times out in one second.
>>
>> But some people need more than one second to read it.
>
> I've now added a new user option 'ispell-help-timeout' so that you can
> control this.

Do we even need a timeout here?

If yes, could we double the default to 10?  Or nil to disable it, even?

>> So I have a better idea.
>>
>> Make it into a four line message, so that way there would be no
>> need to switch back to the original prompt, because the original prompt
>> would now be at the bottom, just like before.
>
> Hm...  That's a good idea, I think.  But how does that affect people
> that have a lower number on the max height of the echo area?

That would be even better, IMO.

bug#50775: ispell help message times out before most people can read it

[Eli Zaretskii]

> From: Stefan Kangas <stefan@marxist.se>
> Date: Sat, 25 Sep 2021 11:17:27 -0700
> Cc: 50775@debbugs.gnu.org,
>  積丹尼 Dan Jacobson <jidanni@jidanni.org>
> 
> > I've now added a new user option 'ispell-help-timeout' so that you can
> > control this.
> 
> Do we even need a timeout here?

We've been having that for eons, so changing the default now should
only be done if we have a very good reason.  You have a customizable
variable now, so why not change the value for your convenience?

Let's not let the tail wag the dog: it's a minor issue, no need to
turn the world around each time we add some minor convenience.

bug#50775: ispell help message times out before most people can read it

[Stefan Kangas]

Eli Zaretskii <eliz@gnu.org> writes:

> We've been having that for eons, so changing the default now should
> only be done if we have a very good reason.  You have a customizable
> variable now, so why not change the value for your convenience?
>
> Let's not let the tail wag the dog: it's a minor issue, no need to
> turn the world around each time we add some minor convenience.

[ Removing a timeout for a message does not exactly amount to "turning
  the world around"... ]

We don't avoid fixing minor issues just because they are old, do we?
That's not something I have ever heard proposed before.

For all we know, users have been frustrated with this for ages, but no
one told us about it until now.

That said, the other alternative I proposed was increasing the timeout
to 10 seconds to give the user enough time to read it.

bug#50775: ispell help message times out before most people can read it

[Juri Linkov]

> That said, the other alternative I proposed was increasing the timeout
> to 10 seconds to give the user enough time to read it.

Or type another key to hide it.

bug#50775: ispell help message times out before most people can read it

[Eli Zaretskii]

> From: Stefan Kangas <stefan@marxist.se>
> Date: Sat, 25 Sep 2021 12:38:10 -0700
> Cc: larsi@gnus.org, 50775@debbugs.gnu.org, jidanni@jidanni.org
> 
> Eli Zaretskii <eliz@gnu.org> writes:
> 
> > We've been having that for eons, so changing the default now should
> > only be done if we have a very good reason.  You have a customizable
> > variable now, so why not change the value for your convenience?
> >
> > Let's not let the tail wag the dog: it's a minor issue, no need to
> > turn the world around each time we add some minor convenience.
> 
> [ Removing a timeout for a message does not exactly amount to "turning
>   the world around"... ]

It does, for those who are accustomed to the message go away.

> We don't avoid fixing minor issues just because they are old, do we?

We don't.  But there's no issue here.

> That said, the other alternative I proposed was increasing the timeout
> to 10 seconds to give the user enough time to read it.

You are free to increase it for your own use.  And so is jidanni.  But
I see no reason to do that for everyone.

bug#50775: ispell help message times out before most people can read it

[積丹尼 Dan Jacobson]

Slower thinkers need longer timeouts.
Faster thinkers know how to customize defaults.
So longer default timeouts win!

bug#50775: ispell help message times out before most people can read it

[Eli Zaretskii]

> From: 積丹尼 Dan Jacobson <jidanni@jidanni.org>
> Cc: Eli Zaretskii <eliz@gnu.org>,  larsi@gnus.org,  50775@debbugs.gnu.org
> Date: Sun, 26 Sep 2021 14:30:52 +0800
> 
> Slower thinkers need longer timeouts.
> Faster thinkers know how to customize defaults.
> So longer default timeouts win!

Not if slower thinkers also know how to customize.

bug#50775: ispell help message times out before most people can read it

[Stefan Kangas]

tags 50775 fixed
close 50775 28.1
thanks

Eli Zaretskii <eliz@gnu.org> writes:

>> Slower thinkers need longer timeouts.
>> Faster thinkers know how to customize defaults.
>> So longer default timeouts win!
>
> Not if slower thinkers also know how to customize.

;-)

I think we are satisfied with the outcome here, and I'm consequently
closing this bug.