Aquamacs 2.0: slow start, Japanese fonts, Kotoeri, visual line

[Jose Figueroa-O'Farrill]

Hi,

I have started using Aquamacs 2.0 and I find the following main problems.  All this is without customizations.

1) it takes an inordinate amount of time to start up.  If I simply launch it from the Dock then no windows appear (at least for 1 minute) until I click on the dock icon  again (but not before waiting for about 30 seconds) and then I get a number of windows: the scratch buffer and N windows where N is the number of times I click on the dock icon.  [This is on a relatively new Unibody MacBook Pro 17" with 4Gb of RAM running 10.5.8.  I will be upgrading to 10.6 as soon as my wife returns from Edinburgh with the DVD :) ]

I have isolated the problem to my Preferences.el file and I suppose I could try to narrow it down further by taking chunks of it at a time.  I'll do that later if I have the time.  However, if I hide the Preferences.el file so that it is not found at start up then it starts up reasonably quickly; although not as quickly as Aquamacs 1.8.  If I then load the hidden Preference.el file, it loads immediately and without any appreciable delay.  So I don't know what the problem is.

2) There is a problem with Japanese fonts -- this has been remarked by Jean-Christophe previously.  The kana characters are nice and antialiased, but the kanji characters are scraggly and almost unreadable.  How was this problem sorted (if it was) in the end?

3) Aquamacs seems to trap the standard Kotoeri keyboard shortcuts.  I switch input system between latin and japanese by Control-Shift-J and Control-Shift-; but these no longer work because Aquamacs does not distinguish between Ctrl-Shift-J and Ctrl-J.  Neither does Ctrl-Shift-; work with Aquamacs claiming that "^: is undefined."

4) C-p and C-n are now bound to aquamacs-*-nonvisual-line.  Is this intentional?  It seems that it's the arrow keys which now behave visually.  However this is not consistent with the bindings for C-a and C-e which just like Option-arrows behave visually.
Anyway, I would think that if is a lot more ergonomic to to set the C-a,C-e,C-p,C-n according to the value of visual-line-mode.  If someone bothers to set visual-line-mode to 't' is because one wants to work with visual lines, in which case the the usual emacs keys for operating on lines should behave visually as well, right?  Anyway, that's easy enough to take care of via the Preferences/Customisations.  I just point it out in case it was not intentional.    And at any rate, I think that the default keybindings are inconsistent as they stand.

More later,

José


--
Professor José Miguel Figueroa-O'Farrill
Institute for the Physics and Mathematics of the Universe
University of Tokyo, Kashiwa, Chiba, Japan
Phone: [phone] and Skype: josemiguelfigueroa
Web: http://www.maths.ed.ac.uk/~jmf
Blog: http://empg.maths.ed.ac.uk/blog
_______________________________________________
Aquamacs-bugs mailing list
Aquamacs-bugs@...
http://lists.aquamacs.org/mailman/listinfo.cgi/aquamacs-bugs

[Jose Figueroa-O'Farrill-2]

Hi again,

To follow up on my own post, I narrowed down the cause of the slow
start:

Jose> 1) it takes an inordinate amount of time to start up.

It was due to the weblogger package, which was prompting for the
username of a blog (this was in customizations.el, but I was testing
without loading it).   So Aquamacs was waiting for input, but had not
yet displayed any windows, so I was none the wiser.

That is now resolved (although it still feels slower starting up than
Aquamacs 1.8) but it has highlighted another problem.

I was setting up a weblog via M-x weblogger-setup-weblog.  This
function prompts you in sequence for a URL, a username, a password and (if
successful) for a label for the blog.  Aquamacs seems to hang after
the first prompt: I type the URL, hit return and nothing happens: in
particular, it does not go on to prompt me for the username.  Well, in
fact, it has but the minibuffer doesn't reflect it.  I have to
recenter (i.e., C-l) for the username prompt to appear.

Cheers, José


--
Professor José Miguel Figueroa-O'Farrill
Institute for the Physics and Mathematics of the Universe
University of Tokyo, Kashiwa, Chiba, Japan
Phone: [phone] and Skype: josemiguelfigueroa
Web: http://www.maths.ed.ac.uk/~jmf
Blog: http://empg.maths.ed.ac.uk/blog
_______________________________________________
Aquamacs-bugs mailing list
Aquamacs-bugs@...
http://lists.aquamacs.org/mailman/listinfo.cgi/aquamacs-bugs

[davidswelt]

José, JC,

On Sep 5, 2009, at 4:11 PM, Jose Figueroa-O'Farrill wrote:
> 2) There is a problem with Japanese fonts -- this has been remarked  
> by Jean-Christophe previously.  The kana characters are nice and  
> antialiased, but the kanji characters are scraggly and almost  
> unreadable.  How was this problem sorted (if it was) in the end?

JC, can you fill us in on this one?

> 3) Aquamacs seems to trap the standard Kotoeri keyboard shortcuts.  
> I switch input system between latin and japanese by Control-Shift-J  
> and Control-Shift-; but these no longer work because Aquamacs does  
> not distinguish between Ctrl-Shift-J and Ctrl-J.  Neither does Ctrl-
> Shift-; work with Aquamacs claiming that "^: is undefined."

It looks like Emacs is intercepting these in the NS port.  Perhaps the  
old mechanism to pass control combinations to the system doesn't work  
any longer.

> 4) C-p and C-n are now bound to aquamacs-*-nonvisual-line.  Is this  
> intentional?  It seems that it's the arrow keys which now behave  
> visually.  However this is not consistent with the bindings for C-a  
> and C-e which just like Option-arrows behave visually.

Yes, this is intentional.  Won't rebind keys as a side-effect.  The  
arrow keys have always behaved visually; the binding mechanism was  
different.

I agree about the inconsistency criticism.

> Anyway, I would think that if is a lot more ergonomic to to set the  
> C-a,C-e,C-p,C-n according to the value of visual-line-mode.

The mode is on by default, and that is good.  People will need buffer-
line based navigation on occasion, so having C-p/n for that is a good  
thing.

Do other people have an opinion regarding C-a/e ?

- David

--
http://aquamacs.org -- Aquamacs: Emacs on Mac OS X
http://aquamacs.org/donate -- Could we help you? Return the favor and  
support the Aquamacs Project!

_______________________________________________
Aquamacs-bugs mailing list
Aquamacs-bugs@...
http://lists.aquamacs.org/mailman/listinfo.cgi/aquamacs-bugs

[Jean-Christophe Helary]

Sorry for the late reply.

On 21 sept. 2009, at 01:48, David Reitter wrote:

> José, JC,
>
> On Sep 5, 2009, at 4:11 PM, Jose Figueroa-O'Farrill wrote:
>> 2) There is a problem with Japanese fonts -- this has been remarked  
>> by Jean-Christophe previously.  The kana characters are nice and  
>> antialiased, but the kanji characters are scraggly and almost  
>> unreadable.  How was this problem sorted (if it was) in the end?
>
> JC, can you fill us in on this one?

Well, I just made a movie with the new Quicktime, showing Japanese  
input in both 1.9 and today's nightly build and the only thing I  
notice is the difference in font. Overall system reaction is  
reasonably similar. I can send the movie offlist if necessary (about  
5mb).

In both cases I have started the apps without customizations.




Jean-Christophe Helary

_______________________________________________
Aquamacs-bugs mailing list
Aquamacs-bugs@...
http://lists.aquamacs.org/mailman/listinfo.cgi/aquamacs-bugs

[Jean-Christophe Helary]

On 21 sept. 2009, at 01:48, David Reitter wrote:

> José, JC,
>
> On Sep 5, 2009, at 4:11 PM, Jose Figueroa-O'Farrill wrote:
>> 2) There is a problem with Japanese fonts -- this has been remarked  
>> by Jean-Christophe previously.  The kana characters are nice and  
>> antialiased, but the kanji characters are scraggly and almost  
>> unreadable.  How was this problem sorted (if it was) in the end?
>
> JC, can you fill us in on this one?

I just noticed a problem with the nightly build, besides for the  
difference of font I had noticed a few days ago:

When I type a sequence of Japanese characters the output blinks at  
each new character. When I  validate the string, it is redrawn one  
character at a time with a visible delay between each character.

In 1.9, this problem does not occur, but when I start 2.0 without  
configuration files it does not occur either.

I can send you a movie to show you what is wrong.

JC
_______________________________________________
Aquamacs-bugs mailing list
Aquamacs-bugs@...
http://lists.aquamacs.org/mailman/listinfo.cgi/aquamacs-bugs

[davidswelt]

On Oct 14, 2009, at 8:00 PM, Jean-Christophe Helary wrote:
> When I type a sequence of Japanese characters the output blinks at  
> each new character. When I  validate the string, it is redrawn one  
> character at a time with a visible delay between each character.
>
> In 1.9, this problem does not occur, but when I start 2.0 without  
> configuration files it does not occur either.

Sorry, you're confusing me.  What do you refer to by "2.0"?
2.0 isn't out yet.  We have several 2.0 previews, and the nightly  
build.  Which is it that doesn't show the problem?

If this is related to your customizations, can you find out which one  
is causing this?

The code handling input is completely different between 1.9 (Carbon,  
Emacs 22) and 2.x (Cocoa, Emacs 23), so we don't need to look further  
into 1.9.

- David
_______________________________________________
Aquamacs-bugs mailing list
Aquamacs-bugs@...
http://lists.aquamacs.org/mailman/listinfo.cgi/aquamacs-bugs