absurd (was: Xemacs 21.5.32 & org-8.03 almost)

absurd (was: Xemacs 21.5.32 & org-8.03 almost)

[Uwe Brauer]

[-- Attachment #1: Type: text/plain, Size: 2013 bytes --]


Now the make process is getting absurd. 
For example I put in org-agenda.el

(if (featurep 'xemacs)
	(define-obsolete-variable-alias 'org-agenda-menu-two-column 'org-agenda-menu-two-columns)
  (define-obsolete-variable-alias 'org-agenda-menu-two-column 'org-agenda-menu-two-columns "24.3"))

Now when I run make, with the local.mk for Xemacs, 

oldorg:	# do what the old Makefile did by default.

,----
[+]
| ##----------------------------------------------------------------------
| ##  CHECK AND ADAPT THE FOLLOWING DEFINITIONS
| ##----------------------------------------------------------------------
| .PHONY: xemacs
| xemacs: up0 doc uncompiled
| EMACS   = /usr/local/bin/xemacs
| prefix  = /home/oub/xemacs/site-lisp/packages
| lispdir = $(prefix)/org
| datadir = $(prefix)/etc/org
| infodir = $(prefix)/info
| 
| BTEST = /bin/true
| BATCH = $(EMACS) -batch -q -vanilla # XEmacs
| # How to byte-compile the whole source directory
| ELCDIR  = $(BATCH) \
|                 --eval '(add-to-list '"'"'load-path ".")' \
|                 --eval '(byte-recompile-directory "." 0)'
`----




I obtain

,----
[+]
| While compiling toplevel forms in file
| /home/oub/ALLES/xemacs/site-lisp/packages/org-8.0.3/lisp/org-agenda.el:
|   !! Wrong number of arguments ((define-obsolete-function-alias 2))
|   !! Wrong number of arguments ((define-obsolete-function-alias 2))
| 
| xemacs exiting.
`----


But  according to the docstring:

,----
[+]
| 
| `define-obsolete-function-alias' is a built-in compiled Lisp function
|   -- loaded from
| "/home/oub/ALLES/Add-Import/xemacs-21.5.32-xsymbol/lisp/obsolete.elc"
| (define-obsolete-function-alias OLDFUN NEWFUN)
| 
| Documentation:
| Define OLDFUN as an obsolete alias for function NEWFUN.
| This makes calling OLDFUN equivalent to calling NEWFUN and marks OLDFUN
| as obsolete.
`----

When I byte compile org-agenda.el individually no such problem occurs. 

What is wrong?

thanks

Uwe Brauer 


[-- Attachment #2: smime.p7s --]
[-- Type: application/pkcs7-signature, Size: 5556 bytes --]

Re: absurd

[Achim Gratz]

Uwe Brauer writes:
> Now the make process is getting absurd. 

What does that problem have to do with make?

> For example I put in org-agenda.el
>
> (if (featurep 'xemacs)
> 	(define-obsolete-variable-alias 'org-agenda-menu-two-column 'org-agenda-menu-two-columns)
>   (define-obsolete-variable-alias 'org-agenda-menu-two-column 'org-agenda-menu-two-columns "24.3"))

That cannot work since the byte-compiler necessarily sees both branches
of the if clause.


Regards,
Achim.
-- 
+<[Q+ Matrix-12 WAVE#46+305 Neuron microQkb Andromeda XTk Blofeld]>+

Waldorf MIDI Implementation & additional documentation:
http://Synth.Stromeko.net/Downloads.html#WaldorfDocs

Re: absurd

[Achim Gratz]

Achim Gratz writes:
> That cannot work since the byte-compiler necessarily sees both branches
> of the if clause.

I've changed the advices to actually remove any arguments past the
second from the argument list.  That should allow you to compile and
load Org (at least it works with my copy of XEmacs).  Expect further
incompatibilities when trying advanced Org features (for instance Babel
doesn't seem to work, or at least ob-perl doesn't).


Regards,
Achim.
-- 
+<[Q+ Matrix-12 WAVE#46+305 Neuron microQkb Andromeda XTk Blofeld]>+

SD adaptations for Waldorf Q V3.00R3 and Q+ V3.54R2:
http://Synth.Stromeko.net/Downloads.html#WaldorfSDada

Re: absurd

[Uwe Brauer]

[-- Attachment #1: Type: text/plain, Size: 1154 bytes --]

>> "Achim" == Achim Gratz <Stromeko@nexgo.de> writes:

   > Uwe Brauer writes:
   >> Now the make process is getting absurd. 

   > What does that problem have to do with make?

Because I can byte compile it manually, in dired via Shift-B!
That I consider as bizarre, byte compiling works, but via make it does
not.


   >> For example I put in org-agenda.el
   >> 
   >> (if (featurep 'xemacs)
   >> (define-obsolete-variable-alias 'org-agenda-menu-two-column
   >> 'org-agenda-menu-two-columns)
   >> (define-obsolete-variable-alias 'org-agenda-menu-two-column
   >> 'org-agenda-menu-two-columns "24.3"))

   > That cannot work since the byte-compiler necessarily sees both branches
   > of the if clause.
I don't understand:

(defun my-test-if ()
  (interactive)
  (if (featurep 'xemacs)
      (message "Xemacs")
    (message "GNUS")))


when executed in Xemacs returns Xemacs, when executed in GNU emacs returns
GNUS.

In any case I removed the ifs leaving 

(define-obsolete-variable-alias 'org-agenda-menu-two-column
 'org-agenda-menu-two-columns)

And again make failed with the same error message!

Uwe 

[-- Attachment #2: smime.p7s --]
[-- Type: application/pkcs7-signature, Size: 5556 bytes --]

Re: absurd

[Uwe Brauer]

[-- Attachment #1: Type: text/plain, Size: 722 bytes --]

>> "Achim" == Achim Gratz <Stromeko@nexgo.de> writes:

   > Achim Gratz writes:
   >> That cannot work since the byte-compiler necessarily sees both branches
   >> of the if clause.

   > I've changed the advices to actually remove any arguments past the
   > second from the argument list.  That should allow you to compile and
   > load Org (at least it works with my copy of XEmacs).  Expect further
   > incompatibilities when trying advanced Org features (for instance Babel
   > doesn't seem to work, or at least ob-perl doesn't).

Oops I just saw you second message. This new version I can obtain via 
git clone git://orgmode.org/org-mode.git 
I presume.

Lets see whats happening
thanks

Uwe  

[-- Attachment #2: smime.p7s --]
[-- Type: application/pkcs7-signature, Size: 5556 bytes --]

Re: absurd

[Uwe Brauer]

[-- Attachment #1: Type: text/plain, Size: 918 bytes --]

>> "Achim" == Achim Gratz <Stromeko@nexgo.de> writes:

   > Achim Gratz writes:
   >> That cannot work since the byte-compiler necessarily sees both branches
   >> of the if clause.

   > I've changed the advices to actually remove any arguments past the
   > second from the argument list.  That should allow you to compile and
   > load Org (at least it works with my copy of XEmacs).  Expect further
   > incompatibilities when trying advanced Org features (for instance Babel
   > doesn't seem to work, or at least ob-perl doesn't).

Hm no the version I obtain via 

git clone git://orgmode.org/org-mode.git
Still contains 

(define-obsolete-variable-alias 'org-agenda-menu-two-column
'org-agenda-menu-two-columns "24.3") 

But according to the webpage this is the most recent version, the one in
which you just performed the changes you indicated.

I don't understand

regards


Uwe 

[-- Attachment #2: smime.p7s --]
[-- Type: application/pkcs7-signature, Size: 5556 bytes --]

Re: absurd

[Achim Gratz]

Uwe Brauer writes:
> Oops I just saw you second message. This new version I can obtain via 
> git clone git://orgmode.org/org-mode.git 
> I presume.

If your current version was also obtained via git, then simply

git checkout master
git pull

will update your work tree.  You don't need to copy all the history
again since you already have it.


Regards,
Achim.
-- 
+<[Q+ Matrix-12 WAVE#46+305 Neuron microQkb Andromeda XTk Blofeld]>+

SD adaptation for Waldorf microQ V2.22R2:
http://Synth.Stromeko.net/Downloads.html#WaldorfSDada

Re: absurd

[Achim Gratz]

Uwe Brauer writes:
> I don't understand:
>
> (defun my-test-if ()
>   (interactive)
>   (if (featurep 'xemacs)
>       (message "Xemacs")
>     (message "GNUS")))
>
>
> when executed in Xemacs returns Xemacs, when executed in GNU emacs returns
> GNUS.

Yes, but the byte-compiler has to compile both branches, since the if is
actually resolved at run-time.  You'd need to use a compatibility macro
to make that decision at compile-time and only use the XEmacs branch.
That was once in Org, but later reverted by Bastien.


Regards,
Achim.
-- 
+<[Q+ Matrix-12 WAVE#46+305 Neuron microQkb Andromeda XTk Blofeld]>+

Wavetables for the Terratec KOMPLEXER:
http://Synth.Stromeko.net/Downloads.html#KomplexerWaves

Re: absurd

[Uwe Brauer]

[-- Attachment #1: Type: text/plain, Size: 662 bytes --]

>> "Achim" == Achim Gratz <Stromeko@nexgo.de> writes:

   > Uwe Brauer writes:
   >> Oops I just saw you second message. This new version I can obtain via 
   >> git clone git://orgmode.org/org-mode.git 
   >> I presume.

   > If your current version was also obtained via git, then simply

   > git checkout master
   > git pull

   > will update your work tree.  You don't need to copy all the history
   > again since you already have it.

Well I already downloaded it again, but as I said, the change you
describe is not in, and returning to my previous post, I still don't
understand why I can byte compile it manually but make *fails*.


[-- Attachment #2: smime.p7s --]
[-- Type: application/pkcs7-signature, Size: 5556 bytes --]

Re: absurd

[Uwe Brauer]

[-- Attachment #1: Type: text/plain, Size: 707 bytes --]

>> "Achim" == Achim Gratz <Stromeko@nexgo.de> writes:

   >> 
   >> when executed in Xemacs returns Xemacs, when executed in GNU emacs
   >> returns GNUS.

   > Yes, but the byte-compiler has to compile both branches, since the
   > if is actually resolved at run-time.  You'd need to use a
   > compatibility macro to make that decision at compile-time and only
   > use the XEmacs branch.  That was once in Org, but later reverted by
   > Bastien.

Ok, maybe this is off topic, but what would be a correct construction, I
am asking since my changes to icalendar.el suffer from the same mistake.

BTW I still do not understand why manual byte-compiling works, while
make does not.

Uwe 

[-- Attachment #2: smime.p7s --]
[-- Type: application/pkcs7-signature, Size: 5556 bytes --]