10:51 a.m.
It seems that a (require 'something) line in a file is executed during
byte-compilation. That seems wrong to me, since if that is really what you
want, you can 'eval-and-compile it.
Is it a bug or am I missing something ?
--
/ / _ _ Didier Verna http://www.inf.enst.fr/~verna/
- / / - / / /_/ / EPITA / LRDE mailto:didier＠lrde.epita.fr
/_/ / /_/ / /__ / 14-16 rue Voltaire Tel. +33 (1) 53 14 59 47
94276 Kremlin-Bicêtre cedex Fax. +33 (1) 44 08 01 99
1:45 p.m.
Didier Verna <didier(a)xemacs.org> writes:
It seems that a (require 'something) line in a file is executed
during
byte-compilation.
Yes, that's in order to get macros, variable declarations, function
definitions, etc., from the package that is being required.
That seems wrong to me, since if that is really what you want, you
can 'eval-and-compile it.
Maybe that's a matter of taste, but I rarely use eval-and-compile and
friends, and when I do, it's only for special hacks with the
byte-compiler, usually to achieve compatibility across Emacs versions.
`require', on the other hand, means that you require the presence of
the package to run and presumably to compile your own package. Having
to wrap each `require' in `eval-and-compile' would be annoying for me.
Finally, there's the matter of what Common Lisp does. I wouldn't be
surprised if it were the same.
2:18 p.m.
Hrvoje Niksic writes:
Didier Verna <didier(a)xemacs.org> writes:
> It seems that a (require 'something) line in a file is executed during
> byte-compilation.
Yes, that's in order to get macros, variable declarations, function
definitions, etc., from the package that is being required.
> That seems wrong to me, since if that is really what you want, you
> can 'eval-and-compile it.
Maybe that's a matter of taste, but I rarely use eval-and-compile and
friends, and when I do, it's only for special hacks with the
byte-compiler, usually to achieve compatibility across Emacs versions.
`require', on the other hand, means that you require the presence of
the package to run and presumably to compile your own package. Having
to wrap each `require' in `eval-and-compile' would be annoying for me.
Finally, there's the matter of what Common Lisp does. I wouldn't be
surprised if it were the same.
ANSI Common Lisp has deprecated the use of PROVIDE and REQUIRE.
Their behaviour was never completely specified and, consequently, they
were implemented differently by different vendors.
Craig
--
----------------------------------------------------------------------
Craig Lanning Principal Engineer
Logicon, Inc.
a Northrop Grumman company E-Mail: Lanning(a)scra.org
5300 International Blvd. Voice: 843-760-3246
North Charleston, SC 29418 FAX: 843-760-3780
USA
2:21 p.m.
Hrvoje Niksic wrote:
`require', on the other hand, means that you require the presence
of
the package to run and presumably to compile your own package.
The "presumably" is actually the problem here...
Having to wrap each `require' in `eval-and-compile' would be
annoying for
me.
I understand this argument, but there are problems to solve here:
1/ AFAICS, XEmacs 21.1 doesn't behave this way. This introduces some backward
incompatibility, notably because some packages can compile properly under
XEmacs 21.1 but not under XEmacs 21.2
2/ I don't know of a standard way to do avoid evaluation of 'require at
compile time, and I need it.
3/ What does GNU Emacs do ? Are we compatible ?
--
/ / _ _ Didier Verna http://www.inf.enst.fr/~verna/
- / / - / / /_/ / EPITA / LRDE mailto:didier＠lrde.epita.fr
/_/ / /_/ / /__ / 14-16 rue Voltaire Tel. +33 (1) 53 14 59 47
94276 Kremlin-Bicêtre cedex Fax. +33 (1) 44 08 01 99
2:33 p.m.
Didier Verna <didier(a)xemacs.org> writes:
I understand this argument, but there are problems to solve
here:
1/ AFAICS, XEmacs 21.1 doesn't behave this way.
I find that hard to believe. To be 100% sure, I've just created two
files, x.el and y.el, with the following contents:
;; x.el:
(require 'y)
;; y.el
(message "foo")
(provide 'y)
I started XEmacs 21.1 with -vanilla, pushed "." to the load-path, and
byte-compiled x.el. After the compilation, `C-h l' showed that `y'
was loaded ("foo" was in the message log).
That's with XEmacs 21.1.10. Maybe Martin's recent changes to the
byte-compiler broke something. In that case, the compiler should be
fixed.
2/ I don't know of a standard way to do avoid evaluation of
'require at
compile time, and I need it.
I guess you could wrap the `require' in an appropriate `eval-when'.
3/ What does GNU Emacs do ? Are we compatible ?
Also prints "foo"; that is, we're compatible.
3:23 p.m.
Hrvoje Niksic wrote:
Didier Verna <didier(a)xemacs.org> writes:
> I understand this argument, but there are problems to solve here:
>
> 1/ AFAICS, XEmacs 21.1 doesn't behave this way.
I find that hard to believe.
You're right. I was mistaken by some local configuration here.
Sorry.
--
/ / _ _ Didier Verna http://www.inf.enst.fr/~verna/
- / / - / / /_/ / EPITA / LRDE mailto:didier＠lrde.epita.fr
/_/ / /_/ / /__ / 14-16 rue Voltaire Tel. +33 (1) 53 14 59 47
94276 Kremlin-Bicêtre cedex Fax. +33 (1) 44 08 01 99
2:31 a.m.
>>>> "drv" == Didier Verna
<didier(a)xemacs.org> writes:
drv> It seems that a (require 'something) line in a file is executed
during
drv> byte-compilation. That seems wrong to me, since if that is really what you
drv> want, you can 'eval-and-compile it.
drv> Is it a bug or am I missing something ?
This is the way it has always been.
Lispref says:
The Compilation Functions
=========================
...
Normally, compiling a file does not evaluate the file's contents or
load the file. But it does execute any `require' calls at top level in
the file. One way to ensure that necessary macro definitions are
available during compilation is to `require' the file that defines them
(*note Named Features::). To avoid loading the macro definition files
when someone _runs_ the compiled program, write `eval-when-compile'
around the `require' calls (*note Eval During Compile::).
8717
days inactive
8718
days old
6 comments
4 participants
participants (4)
-
Craig Lanning -
Didier Verna -
Hrvoje Niksic -
Martin Buchholz