bug#20702: 25.0.50; (elisp) `ImageMagick Images': say how to tell if Emacs has IM support, etc.

bug#20702: 25.0.50; (elisp) `ImageMagick Images': say how to tell if Emacs has IM support, etc.

[Drew Adams]

Someone using the Elisp manual might not have built the Emacs build
being used.  S?he might not even know what build a user might use.

This node should tell an Emacs-Lisp programmer how, using Lisp, to tell
whether the current build has ImageMagick support.

IOW, the opening phrase of this node, "If you build Emacs with
ImageMagick support" should be replaced by "If an Emacs build has
ImageMagick support", and this intro should be followed by telling users
how to check this using Lisp.

My guess (from trying) would be that (fboundp 'imagemagick-types) is the
test that tells you this - in my build it returns nil.

In any case, the doc should also say explicitly that this function is
not defined if the Emacs build has no ImageMagick support.  In
particular, the way `imagemagick-types' is described now, one could
reasonably (and mistakenly) expect that in an Emacs build that has no
ImageMagick support `imagemagick-types' would be defined but would
return () - no ImageMagick image types are supported.  (And why isn't
this the case?)

In sum, the intro to this node needs to be written for Elisp programmers
in general.


In GNU Emacs 25.0.50.1 (i686-pc-mingw32)
 of 2014-10-20 on LEG570
Bzr revision: 118168 rgm@gnu.org-20141020195941-icp42t8ttcnud09g
Windowing system distributor `Microsoft Corp.', version 6.1.7601
Configured using:
 `configure --enable-checking=yes,glyphs CPPFLAGS=-DGLYPH_DEBUG=1'

bug#20702: 25.0.50; (elisp) `ImageMagick Images': say how to tell if Emacs has IM support, etc.

[Eli Zaretskii]

> Date: Sun, 31 May 2015 08:08:03 -0700 (PDT)
> From: Drew Adams <drew.adams@oracle.com>
> 
> My guess (from trying) would be that (fboundp 'imagemagick-types) is the
> test that tells you this - in my build it returns nil.

Actually,

     M-: (image-type-available-p 'imagemagick) RET
and
     C-h v system-configuration-features RET

are better ways of getting at that information.  The latter should IMO
be documented in both manuals.

bug#20702: 25.0.50; (elisp) `ImageMagick Images': say how to tell if Emacs has IM support, etc.

[Glenn Morris]

Eli Zaretskii wrote:

>      C-h v system-configuration-features RET
>
> are better ways of getting at that information.  The latter should IMO
> be documented in both manuals.

As NEWS says, s-c-f is mainly (entirely) intended for Emacs bug reports.
I don't think that checking for elements in there is the right way to
see if Emacs supports a given feature. That's why it is intentionally
not mentioned in the manuals.

bug#20702: 25.0.50; (elisp) `ImageMagick Images': say how to tell if Emacs has IM support, etc.

[Eli Zaretskii]

> From: Glenn Morris <rgm@gnu.org>
> Cc: 20702@debbugs.gnu.org
> Date: Sun, 31 May 2015 14:06:50 -0400
> 
> Eli Zaretskii wrote:
> 
> >      C-h v system-configuration-features RET
> >
> > are better ways of getting at that information.  The latter should IMO
> > be documented in both manuals.
> 
> As NEWS says, s-c-f is mainly (entirely) intended for Emacs bug reports.
> I don't think that checking for elements in there is the right way to
> see if Emacs supports a given feature. That's why it is intentionally
> not mentioned in the manuals.

I cannot see any sense in this decision.  The value of the variable is
very clear to a human reader, so I don't understand why we should hold
it secret.

bug#20702: 25.0.50; (elisp) `ImageMagick Images': say how to tell if Emacs has IM support, etc.

[Eli Zaretskii]

> Date: Sun, 31 May 2015 21:35:36 +0300
> From: Eli Zaretskii <eliz@gnu.org>
> Cc: 20702@debbugs.gnu.org
> 
> > From: Glenn Morris <rgm@gnu.org>
> > Cc: 20702@debbugs.gnu.org
> > Date: Sun, 31 May 2015 14:06:50 -0400
> > 
> > I don't think that checking for elements in there is the right way to
> > see if Emacs supports a given feature. That's why it is intentionally
> > not mentioned in the manuals.
> 
> I cannot see any sense in this decision.  The value of the variable is
> very clear to a human reader, so I don't understand why we should hold
> it secret.

Maybe you meant the value should not be examined by Lisp programs.  In
which case I might agree: I was talking about humans examining the
value, not programs.

bug#20702: 25.0.50; (elisp) `ImageMagick Images': say how to tell if Emacs has IM support, etc.

[Glenn Morris]

Eli Zaretskii wrote:

> Maybe you meant the value should not be examined by Lisp programs.  In
> which case I might agree: I was talking about humans examining the
> value, not programs.

Yeah, I meant Lisp programs.

bug#20702: 25.0.50; (elisp) `ImageMagick Images': say how to tell if Emacs has IM support, etc.

[Lars Ingebrigtsen]

Drew Adams <drew.adams@oracle.com> writes:

> IOW, the opening phrase of this node, "If you build Emacs with
> ImageMagick support" should be replaced by "If an Emacs build has
> ImageMagick support", and this intro should be followed by telling users
> how to check this using Lisp.

I've now pushed a fix.

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

bug#20702: 25.0.50; (elisp) `ImageMagick Images': say how to tell if Emacs has IM support, etc.

[Drew Adams]

> > My guess (from trying) would be that (fboundp 'imagemagick-types)
> > is the test that tells you this - in my build it returns nil.
> 
> Actually, M-: (image-type-available-p 'imagemagick) RET and C-h v
> system-configuration-features RET are better ways of getting at that
> information.  The latter should IMO be documented in both manuals.

Very good.  I agree.