* customize-apropos-options takes forever...
@ 2006-07-05 15:14 Drew Adams
2006-07-05 15:41 ` Chong Yidong
` (2 more replies)
0 siblings, 3 replies; 9+ messages in thread
From: Drew Adams @ 2006-07-05 15:14 UTC (permalink / raw)
M-x customize-apropos-options RET RET
On Emacs 22, this takes quite a while. On Emacs 20, it finishes very
quickly.
Emacs 22 has a lot more options, of course, and more faces and buttons and
such, but it still feels like things are less efficient now. In emacs -Q,
there are 4698 lines in the buffer in Emacs 22; in Emacs 20, there are 2333
lines (about half). But Emacs 22 takes *far* longer than twice as long.
[Why would anyone want to customize all options? I don't know, but you might
want to search among option values or something. If you just want
descriptions, `apropos-variable' is of course better (too bad the name
doesn't say "option"). `edit-options' will give you values too, but Emacs
spits a nasty obsoletion message at you (forgetting to put `options' in
quotes, BTW).]
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: customize-apropos-options takes forever...
2006-07-05 15:14 customize-apropos-options takes forever Drew Adams
@ 2006-07-05 15:41 ` Chong Yidong
2006-07-06 13:32 ` Richard Stallman
2006-07-06 16:43 ` Chong Yidong
2 siblings, 0 replies; 9+ messages in thread
From: Chong Yidong @ 2006-07-05 15:41 UTC (permalink / raw)
Cc: Emacs-Devel
> M-x customize-apropos-options RET RET
>
> [Why would anyone want to customize all options? I don't know, but you might
> want to search among option values or something.
In that case, you simply do customize-apropos RET regexp RET.
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: customize-apropos-options takes forever...
2006-07-05 15:14 customize-apropos-options takes forever Drew Adams
2006-07-05 15:41 ` Chong Yidong
@ 2006-07-06 13:32 ` Richard Stallman
2006-07-06 15:52 ` Drew Adams
2006-07-06 16:43 ` Chong Yidong
2 siblings, 1 reply; 9+ messages in thread
From: Richard Stallman @ 2006-07-06 13:32 UTC (permalink / raw)
Cc: emacs-devel
Emacs 22 has a lot more options, of course, and more faces and buttons and
such, but it still feels like things are less efficient now. In emacs -Q,
there are 4698 lines in the buffer in Emacs 22; in Emacs 20, there are 2333
lines (about half). But Emacs 22 takes *far* longer than twice as long.
Probably something really is wrong. The question is what.
Would you like to try running Emacs 22 under a debugger
and stopping it a few times uring this operation?
If you notice a place that shows up frequently in the backtrace,
that might be the place that is using the time.
^ permalink raw reply [flat|nested] 9+ messages in thread
* RE: customize-apropos-options takes forever...
2006-07-06 13:32 ` Richard Stallman
@ 2006-07-06 15:52 ` Drew Adams
2006-07-07 19:30 ` Richard Stallman
0 siblings, 1 reply; 9+ messages in thread
From: Drew Adams @ 2006-07-06 15:52 UTC (permalink / raw)
Emacs 22 has a lot more options, of course, and more faces
and buttons and such, but it still feels like things are
less efficient now. In emacs -Q, there are 4698 lines in the
buffer in Emacs 22; in Emacs 20, there are 2333
lines (about half). But Emacs 22 takes *far* longer than
twice as long.
Probably something really is wrong. The question is what.
Would you like to try running Emacs 22 under a debugger
and stopping it a few times uring this operation?
If you notice a place that shows up frequently in the backtrace,
that might be the place that is using the time.
I feel comfortable with the lisp debugger, but not with gdb etc.
Also, I just don't have the time right now. If this can be checked with the
lisp debugger (debug-on-entry...), then I will try to check it later.
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: customize-apropos-options takes forever...
2006-07-05 15:14 customize-apropos-options takes forever Drew Adams
2006-07-05 15:41 ` Chong Yidong
2006-07-06 13:32 ` Richard Stallman
@ 2006-07-06 16:43 ` Chong Yidong
2006-07-06 17:06 ` Drew Adams
2006-07-07 19:31 ` Richard Stallman
2 siblings, 2 replies; 9+ messages in thread
From: Chong Yidong @ 2006-07-06 16:43 UTC (permalink / raw)
Cc: Emacs-Devel
"Drew Adams" <drew.adams@oracle.com> writes:
> M-x customize-apropos-options RET RET
>
> On Emacs 22, this takes quite a while. On Emacs 20, it finishes very
> quickly.
>
> Emacs 22 has a lot more options, of course, and more faces and buttons and
> such, but it still feels like things are less efficient now. In emacs -Q,
> there are 4698 lines in the buffer in Emacs 22; in Emacs 20, there are 2333
> lines (about half). But Emacs 22 takes *far* longer than twice as long.
I took some time to look into this.
For M-x customize-apropos RET RET:
Emacs 22 takes 140.04 seconds to make 1904 entries, i.e. 13.6 / second.
Emacs 21 takes 76.38 seconds to make 1048 entries, i.e. 13.7 / second.
Conclusion: >>>> This Is Not A Bug <<<<
I think the reason you thought Emacs 22 is slower is that the echo
area tells you how much % of the job is done. Obviously, 1% of 1904
entries takes longer to do than 1% of 1048 entries, giving the
illusion that Emacs 22 is progressing more slowly.
In the reality, of course, it's silly to do M-x customize-apropos with
no regexp in the first place.
^ permalink raw reply [flat|nested] 9+ messages in thread
* RE: customize-apropos-options takes forever...
2006-07-06 16:43 ` Chong Yidong
@ 2006-07-06 17:06 ` Drew Adams
2006-07-06 22:55 ` Chong Yidong
2006-07-07 19:31 ` Richard Stallman
1 sibling, 1 reply; 9+ messages in thread
From: Drew Adams @ 2006-07-06 17:06 UTC (permalink / raw)
> M-x customize-apropos-options RET RET
>
> On Emacs 22, this takes quite a while. On Emacs 20, it finishes very
> quickly.
>
> Emacs 22 has a lot more options, of course, and more faces
> and buttons and such, but it still feels like things are less
> efficient now. In emacs -Q, there are 4698 lines in the buffer
> in Emacs 22; in Emacs 20, there are 2333 lines (about half).
> But Emacs 22 takes *far* longer than twice as long.
I took some time to look into this.
Thanks Chong.
For M-x customize-apropos RET RET:
Emacs 22 takes 140.04 seconds to make 1904 entries, i.e. 13.6 / second.
Emacs 21 takes 76.38 seconds to make 1048 entries, i.e. 13.7 / second.
Conclusion: >>>> This Is Not A Bug <<<<
I was comparing Emacs 20, not Emacs 21. Perhaps the same difference in
performance exists between 20 and 21. Believe me, there is a marked
difference between 20 and 22. Again, I know that Customize in 22 is a lot
fancier than in 20, and it does more behind the scene, so degraded
performance is not a full measure of usefulness - Customize in 22 is better
than in 20, and I don't want to go backward here.
I think the reason you thought Emacs 22 is slower is that the echo
area tells you how much % of the job is done. Obviously, 1% of 1904
entries takes longer to do than 1% of 1048 entries, giving the
illusion that Emacs 22 is progressing more slowly.
No, I was not paying attention to the echo area. I was simply waiting for
the buffer to appear (!) and for my keyboard and mouse to come back to me
from Never-Never Land.
In the reality, of course, it's silly to do M-x customize-apropos with
no regexp in the first place.
Agreed, but people do silly things sometimes. Searching among all options is
a reasonable thing to want to do, but I agree that there are other ways to
do that.
apropos-variable is good for getting descriptions of all user options (but
not faces) - but it doesn't show you the values. And edit-options is quick
and useful for getting both descriptions and values, but it is deprecated.
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: customize-apropos-options takes forever...
2006-07-06 17:06 ` Drew Adams
@ 2006-07-06 22:55 ` Chong Yidong
0 siblings, 0 replies; 9+ messages in thread
From: Chong Yidong @ 2006-07-06 22:55 UTC (permalink / raw)
Cc: Emacs-Devel
> No, I was not paying attention to the echo area. I was simply waiting for
> the buffer to appear (!) and for my keyboard and mouse to come back to me
> from Never-Never Land.
C-g works.
In the meantime, let's work on the release instead of trying to make
Emacs 22 Custom faster than Emacs 20 (!!) for an unrealistic usage
situation.
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: customize-apropos-options takes forever...
2006-07-06 15:52 ` Drew Adams
@ 2006-07-07 19:30 ` Richard Stallman
0 siblings, 0 replies; 9+ messages in thread
From: Richard Stallman @ 2006-07-07 19:30 UTC (permalink / raw)
Cc: emacs-devel
Would you like to try running Emacs 22 under a debugger
and stopping it a few times uring this operation?
If you notice a place that shows up frequently in the backtrace,
that might be the place that is using the time.
I feel comfortable with the lisp debugger, but not with gdb etc.
You can try using debug-on-quit and the Lisp debugger. If the
slowness is at Lisp level, you would find it that way. But I would
guess it is at the C level. If so, you could only see it with a
C level debugger.
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: customize-apropos-options takes forever...
2006-07-06 16:43 ` Chong Yidong
2006-07-06 17:06 ` Drew Adams
@ 2006-07-07 19:31 ` Richard Stallman
1 sibling, 0 replies; 9+ messages in thread
From: Richard Stallman @ 2006-07-07 19:31 UTC (permalink / raw)
Cc: drew.adams, emacs-devel
Thanks for checking this.
^ permalink raw reply [flat|nested] 9+ messages in thread
end of thread, other threads:[~2006-07-07 19:31 UTC | newest]
Thread overview: 9+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2006-07-05 15:14 customize-apropos-options takes forever Drew Adams
2006-07-05 15:41 ` Chong Yidong
2006-07-06 13:32 ` Richard Stallman
2006-07-06 15:52 ` Drew Adams
2006-07-07 19:30 ` Richard Stallman
2006-07-06 16:43 ` Chong Yidong
2006-07-06 17:06 ` Drew Adams
2006-07-06 22:55 ` Chong Yidong
2006-07-07 19:31 ` Richard Stallman
Code repositories for project(s) associated with this external index
https://git.savannah.gnu.org/cgit/emacs.git
https://git.savannah.gnu.org/cgit/emacs/org-mode.git
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.