Changes to emacs/doc/emacs/building.texi,v

Changes to emacs/doc/emacs/building.texi,v

[Juri Linkov]

In the 2008-10-31 commit of building.texi I noticed the following change:

   @findex compile-goto-error
  +@vindex compilation-auto-jump-to-first-error
     You can visit the source for any particular error message by moving
   point in the @samp{*compilation*} buffer to that error message and
   typing @key{RET} (@code{compile-goto-error}).  Alternatively, you can
   click @kbd{Mouse-2} on the error message; you need not switch to the
  - <at> samp{*compilation*} buffer first.
  +@samp{*compilation*} buffer first.  If you set the variable
  +@code{compilation-auto-jump-to-first-error} to a non-@code{nil} value,
  +Emacs automatically jumps to the first error (if any exists) once
  +compilation finishes.

Actually it jumps to the first error immediately after it occurs.

Perhaps this is caused by the inaccuracy in the doc string of
`compilation-auto-jump-to-first-error' that says:

  If non-nil, automatically jump to the first error after `compile'.

A better variant would be:

  If non-nil, automatically jump to the first error during `compile'.

BTW, maybe a new option `first-error' of `compilation-scroll-output'
should be documented in the manual too?

-- 
Juri Linkov
http://www.jurta.org/emacs/

Re: Changes to emacs/doc/emacs/building.texi,v

[Chong Yidong]

Juri Linkov <juri@jurta.org> writes:

> Actually it jumps to the first error immediately after it occurs.

Thanks.  I've fixed the manual and the docstring.

> BTW, maybe a new option `first-error' of `compilation-scroll-output'
> should be documented in the manual too?

Yes.  I've fixed the manual.  It also needed a NEWS entry---I've added
that.

Re: Changes to emacs/doc/emacs/building.texi,v

[Stefan Monnier]

>   If non-nil, automatically jump to the first error after `compile'.
> A better variant would be:
>   If non-nil, automatically jump to the first error during `compile'.

I'm not convinced: `compile' runs the command asynchronously, so the
jump is definitely done *after* compile, strictly speaking.  Maybe it
should just not say `compile', especially since it's also used in
circumstances where `compile' is not involved (e.g. sml-compile which
passes the compilation request to an inferior process running in
a comint buffer).

So, I'd prefer "during compilation" rather than "during `compile'".


        Stefan

Re: Changes to emacs/doc/emacs/building.texi,v

[Juri Linkov]

>>   If non-nil, automatically jump to the first error after `compile'.
>> A better variant would be:
>>   If non-nil, automatically jump to the first error during `compile'.
>
> I'm not convinced: `compile' runs the command asynchronously, so the
> jump is definitely done *after* compile, strictly speaking.  Maybe it
> should just not say `compile', especially since it's also used in
> circumstances where `compile' is not involved (e.g. sml-compile which
> passes the compilation request to an inferior process running in
> a comint buffer).
>
> So, I'd prefer "during compilation" rather than "during `compile'".

Certainly, there is uncertainty in the phrase "after `compile'"
as it can mean both "after starting" and "after finishing".
I agree that the phrase "during compilation" has no such problems.

-- 
Juri Linkov
http://www.jurta.org/emacs/

Re: Changes to emacs/doc/emacs/building.texi,v

[martin rudalics]

 >> So, I'd prefer "during compilation" rather than "during `compile'".
 >
 > Certainly, there is uncertainty in the phrase "after `compile'"
 > as it can mean both "after starting" and "after finishing".
 > I agree that the phrase "during compilation" has no such problems.

Maybe we should explain in two or three sentences what really happens
instead of using such fuzzy terms.

martin