bug#7726: reveal-mode bypassed by compile-goto-error

bug#7726: reveal-mode bypassed by compile-goto-error

[Lennart Borgman]

If I use compile-goto-error from a grep-mode buffer to a org-mode
buffer where the destination line is in a hidden outline then the
outline will not be revealed until after next command.

GNU Emacs 24.0.50.1 (i386-mingw-nt5.1.2600) from trunk 2010-10-19 (patched)

bug#7726: reveal-mode bypassed by compile-goto-error

[Thierry Volpiatto]

Lennart Borgman <lennart.borgman@gmail.com> writes:

> If I use compile-goto-error from a grep-mode buffer to a org-mode
> buffer where the destination line is in a hidden outline then the
> outline will not be revealed until after next command.

Indeed,

also, goto-line is unable to reach line in an outline based buffer.
See in anything and ioccur, it is fixed for org-mode and
outline-minor-mode.

Is there other modes based on outline to handle?

-- 
A+ Thierry
Get my Gnupg key:
gpg --keyserver pgp.mit.edu --recv-keys 59F29997 

bug#7726: reveal-mode bypassed by compile-goto-error

[Chong Yidong]

Thierry Volpiatto <thierry.volpiatto@gmail.com> writes:

> Lennart Borgman <lennart.borgman@gmail.com> writes:
>
>> If I use compile-goto-error from a grep-mode buffer to a org-mode
>> buffer where the destination line is in a hidden outline then the
>> outline will not be revealed until after next command.
>
> Indeed,
>
> also, goto-line is unable to reach line in an outline based buffer.
> See in anything and ioccur, it is fixed for org-mode and
> outline-minor-mode.
>
> Is there other modes based on outline to handle?

Anyone have a test case?

bug#7726: reveal-mode bypassed by compile-goto-error

[Lennart Borgman]

On Sun, Jan 23, 2011 at 12:47 AM, Chong Yidong <cyd@stupidchicken.com> wrote:
> Thierry Volpiatto <thierry.volpiatto@gmail.com> writes:
>
>> Lennart Borgman <lennart.borgman@gmail.com> writes:
>>
>>> If I use compile-goto-error from a grep-mode buffer to a org-mode
>>> buffer where the destination line is in a hidden outline then the
>>> outline will not be revealed until after next command.
>>
>> Indeed,
>>
>> also, goto-line is unable to reach line in an outline based buffer.
>> See in anything and ioccur, it is fixed for org-mode and
>> outline-minor-mode.
>>
>> Is there other modes based on outline to handle?
>
> Anyone have a test case?

Yes, but why do you want one? Is not the recipes we have given enough?

bug#7726: reveal-mode bypassed by compile-goto-error

[Thierry Volpiatto]

Chong Yidong <cyd@stupidchicken.com> writes:

> Thierry Volpiatto <thierry.volpiatto@gmail.com> writes:
>
>> Lennart Borgman <lennart.borgman@gmail.com> writes:
>>
>>> If I use compile-goto-error from a grep-mode buffer to a org-mode
>>> buffer where the destination line is in a hidden outline then the
>>> outline will not be revealed until after next command.
>>
>> Indeed,
>>
>> also, goto-line is unable to reach line in an outline based buffer.
>> See in anything and ioccur, it is fixed for org-mode and
>> outline-minor-mode.
>>
>> Is there other modes based on outline to handle?
>
> Anyone have a test case?
Switch to any org buffer, close all, and:
M-g g 20 RET

-- 
A+ Thierry
Get my Gnupg key:
gpg --keyserver pgp.mit.edu --recv-keys 59F29997 

bug#7726: reveal-mode bypassed by compile-goto-error

[Stefan Monnier]

>>> If I use compile-goto-error from a grep-mode buffer to a org-mode
>>> buffer where the destination line is in a hidden outline then the
>>> outline will not be revealed until after next command.
>> also, goto-line is unable to reach line in an outline based buffer.
>> See in anything and ioccur, it is fixed for org-mode and
>> outline-minor-mode.
>> Is there other modes based on outline to handle?
> Anyone have a test case?

I think the problem is that reveal-mode runs from post-command-hook and
reveals what's in the buffer that's current when post-command-hook is
run, whereas in the above cases, the buffer that needs revealing is not
current at that point.
That's one of the cases where we could use something like
a "pre-redisplay-hook".


        Stefan

bug#7726: reveal-mode bypassed by compile-goto-error

[Lennart Borgman]

On Mon, Jan 24, 2011 at 1:50 AM, Stefan Monnier
<monnier@iro.umontreal.ca> wrote:
>>>> If I use compile-goto-error from a grep-mode buffer to a org-mode
>>>> buffer where the destination line is in a hidden outline then the
>>>> outline will not be revealed until after next command.
>>> also, goto-line is unable to reach line in an outline based buffer.
>>> See in anything and ioccur, it is fixed for org-mode and
>>> outline-minor-mode.
>>> Is there other modes based on outline to handle?
>> Anyone have a test case?
>
> I think the problem is that reveal-mode runs from post-command-hook and
> reveals what's in the buffer that's current when post-command-hook is
> run, whereas in the above cases, the buffer that needs revealing is not
> current at that point.
> That's one of the cases where we could use something like
> a "pre-redisplay-hook".

Isn't it more like post-post-command-hook?

(Didn't I take up such a problem with cua-mode long ago? But at the
moment I can't remember what it was.)

bug#7726: reveal-mode bypassed by compile-goto-error

[Stefan Monnier]

>>>>> If I use compile-goto-error from a grep-mode buffer to a org-mode
>>>>> buffer where the destination line is in a hidden outline then the
>>>>> outline will not be revealed until after next command.
>>>> also, goto-line is unable to reach line in an outline based buffer.
>>>> See in anything and ioccur, it is fixed for org-mode and
>>>> outline-minor-mode.
>>>> Is there other modes based on outline to handle?
>>> Anyone have a test case?
>> I think the problem is that reveal-mode runs from post-command-hook and
>> reveals what's in the buffer that's current when post-command-hook is
>> run, whereas in the above cases, the buffer that needs revealing is not
>> current at that point.
>> That's one of the cases where we could use something like
>> a "pre-redisplay-hook".
> Isn't it more like post-post-command-hook?

No I really meant pre-redisplay-hook.  I don't see why
post-post-command-hook would have anything to do with this problem.
E.g. I think a case that needs to work is when you're in the *compile*
buffer and select one entry to display the source but while staying in
the *compile* buffer: the current buffer before and after the command is
*compile*, so no amount of post/pre-command-hooks can tell you that some
other buffer needs revealing.


        Stefan

bug#7726: reveal-mode bypassed by compile-goto-error

[Lennart Borgman]

On Mon, Jan 24, 2011 at 5:12 PM, Stefan Monnier
<monnier@iro.umontreal.ca> wrote:
>>>>>> If I use compile-goto-error from a grep-mode buffer to a org-mode
>>>>>> buffer where the destination line is in a hidden outline then the
>>>>>> outline will not be revealed until after next command.
>>>>> also, goto-line is unable to reach line in an outline based buffer.
>>>>> See in anything and ioccur, it is fixed for org-mode and
>>>>> outline-minor-mode.
>>>>> Is there other modes based on outline to handle?
>>>> Anyone have a test case?
>>> I think the problem is that reveal-mode runs from post-command-hook and
>>> reveals what's in the buffer that's current when post-command-hook is
>>> run, whereas in the above cases, the buffer that needs revealing is not
>>> current at that point.
>>> That's one of the cases where we could use something like
>>> a "pre-redisplay-hook".
>> Isn't it more like post-post-command-hook?
>
> No I really meant pre-redisplay-hook.  I don't see why
> post-post-command-hook would have anything to do with this problem.
> E.g. I think a case that needs to work is when you're in the *compile*
> buffer and select one entry to display the source but while staying in
> the *compile* buffer: the current buffer before and after the command is
> *compile*, so no amount of post/pre-command-hooks can tell you that some
> other buffer needs revealing.

Oh, I see.

And haven't we been discussing before that such a hook maybe could be
useful for cases where we now are using
window-configuration-change-hook?