asm-mode patch to allow per-file comment character setting from file locals

asm-mode patch to allow per-file comment character setting from file locals

[Alastair Houghton]

Hi there,

Here's a short patch to asm-mode.el that I use so that I can easily  
vary the comment style used for assembly language files on a per-file  
basis.  This is particularly useful on Mac OS X, where the Intel and  
PowerPC versions of the system assembler use different comment  
characters (I believe the PowerPC version actually accepts both ';'  
and '#', but the Intel one certainly requires '#').

An example of its use:

   ###
   ### Local Variables:
   ### asm-comment-char: ?\#
   ### End:
   ###

Here's the patch:

--- asm-mode.el.prev    2006-06-09 12:26:09.000000000 +0100
+++ asm-mode.el 2006-06-09 12:54:01.000000000 +0100
@@ -59,6 +59,14 @@
    "*The comment-start character assumed by Asm mode."
    :type 'character
    :group 'asm)
+(make-variable-buffer-local 'asm-comment-char)
+(put 'asm-comment-char 'safe-local-variable t)
+
+(defvar asm-current-comment-char nil
+  "Holds the current comment-start character in use in this buffer.")
+(make-variable-buffer-local 'asm-current-comment-char)
+
+(add-hook 'find-file-hook (lambda () (asm-set-comment-char asm- 
comment-char)))

(defvar asm-mode-syntax-table
    (let ((st (make-syntax-table)))
@@ -134,12 +142,9 @@
    ;; Make our own local child of asm-mode-map
    ;; so we can define our own comment character.
    (use-local-map (nconc (make-sparse-keymap) asm-mode-map))
-  (local-set-key (vector asm-comment-char) 'asm-comment)
+  (asm-set-comment-char asm-comment-char)
    (set-syntax-table (make-syntax-table asm-mode-syntax-table))
-  (modify-syntax-entry asm-comment-char "<")

-  (make-local-variable 'comment-start)
-  (setq comment-start (string asm-comment-char))
    (make-local-variable 'comment-add)
    (setq comment-add 1)
    (make-local-variable 'comment-start-skip)
@@ -151,6 +156,22 @@
    (setq fill-prefix "\t")
    (run-mode-hooks 'asm-mode-hook))

+(defun asm-set-comment-char (newch)
+  "Set the comment character for the current buffer"
+  (interactive "c")
+  (when asm-current-comment-char
+    (local-unset-key (vector asm-current-comment-char))
+    (modify-syntax-entry asm-current-comment-char
+                        (with-syntax-table asm-mode-syntax-table
+                          (string (char-syntax asm-current-comment- 
char)))))
+  (setq asm-comment-char newch)
+  (setq asm-current-comment-char newch)
+  (local-set-key (vector asm-comment-char) 'asm-comment)
+  (modify-syntax-entry asm-comment-char "<")
+  (setq comment-start (string asm-comment-char))
+  (when font-lock-mode
+    (font-lock-fontify-buffer)))
+
(defun asm-indent-line ()
    "Auto-indent the current line."
    (interactive)

Kind regards,

Alastair.

--
http://www.alastairs-place.net

Re: asm-mode patch to allow per-file comment character setting from file locals

[Stefan Monnier]

I don't have time to look in detail at the patch, but here are some comments:

> +(make-variable-buffer-local 'asm-comment-char)

Please don't do that.  USe make-local-variable where necessary instead.

> +(put 'asm-comment-char 'safe-local-variable t)

Use an actual function instead rather than just t.


        Stefan

Re: asm-mode patch to allow per-file comment character setting from file locals

[Alastair Houghton]

On 11 Jun 2006, at 22:20, Stefan Monnier wrote:

> I don't have time to look in detail at the patch, but here are some  
> comments:
>
>> +(make-variable-buffer-local 'asm-comment-char)
>
> Please don't do that.  USe make-local-variable where necessary  
> instead.

The reason I chose to use make-variable-buffer-local is that I want  
to be able to set it from a file local variables block (either the - 
*- version at the top, or the version at the bottom of my assembler  
files).

Is there a way to ensure that it's buffer local before that point  
without using make-variable-buffer-local?

>> +(put 'asm-comment-char 'safe-local-variable t)
>
> Use an actual function instead rather than just t.

I'm guessing you think it should check the type for int or (in XEmacs  
case) char.  Fair comment.

Kind regards,

Alastair.

--
http://www.alastairs-place.net

Re: asm-mode patch to allow per-file comment character setting from file locals

[David Kastrup]

Alastair Houghton <alastair@alastairs-place.net> writes:

> On 11 Jun 2006, at 22:20, Stefan Monnier wrote:
>
>> I don't have time to look in detail at the patch, but here are some
>> comments:
>>
>>> +(make-variable-buffer-local 'asm-comment-char)
>>
>> Please don't do that.  USe make-local-variable where necessary
>> instead.
>
> The reason I chose to use make-variable-buffer-local is that I want
> to be able to set it from a file local variables block (either the - 
> *- version at the top, or the version at the bottom of my assembler
> files).

Files set from a file local variables block are made buffer-local,
anyway.  Perhaps reading the documentation will help.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: asm-mode patch to allow per-file comment character setting from file locals

[Alastair Houghton]

On 12 Jun 2006, at 14:12, David Kastrup wrote:

> Files set from a file local variables block are made buffer-local,
> anyway.

Which isn't stated in the Elisp info documentation (which I *did* read),
though I do see now that it says that in the Emacs info file (which I  
only
skimmed), and it certainly makes sense that it should work that way.

> Perhaps reading the documentation will help.

I'm happy to admit that I missed that part of the documentation,  
though I'm
don't appreciate the undertone of cheekiness in your last remark,  
particularly
not from someone with a gnu.org e-mail address.  Maybe it wasn't  
intended that
way, but that's how it came across.

Kind regards,

Alastair.

--
http://www.alastairs-place.net

Re: asm-mode patch to allow per-file comment character setting from file locals

[David Kastrup]

Alastair Houghton <alastair@alastairs-place.net> writes:

> On 12 Jun 2006, at 14:12, David Kastrup wrote:
>
>> Files set from a file local variables block are made buffer-local,
>> anyway.
>
> Which isn't stated in the Elisp info documentation (which I *did*
> read),

File local variable blocks are not documented in the Elisp manual at
all.  They are a user-level feature.

> though I do see now that it says that in the Emacs info file (which
> I only skimmed), and it certainly makes sense that it should work
> that way.
>
>> Perhaps reading the documentation will help.
>
> I'm happy to admit that I missed that part of the documentation,
> though I'm don't appreciate the undertone of cheekiness in your last
> remark, particularly not from someone with a gnu.org e-mail address.

I'll leave replying to you to others in future.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: asm-mode patch to allow per-file comment character setting from file locals

[Alastair Houghton]

On 12 Jun 2006, at 23:58, David Kastrup wrote:

> Alastair Houghton <alastair@alastairs-place.net> writes:
>
>> On 12 Jun 2006, at 14:12, David Kastrup wrote:
>>
>>> Files set from a file local variables block are made buffer-local,
>>> anyway.
>>
>> Which isn't stated in the Elisp info documentation (which I *did*
>> read),
>
> File local variable blocks are not documented in the Elisp manual at
> all.  They are a user-level feature.

*My* copy of the Elisp manual has a node entitled "File Local  
Variables".
It seems to me that that should be the documentation for file local  
variables
from an Elisp perspective.

> I'll leave replying to you to others in future.

That's your prerogative, of course.

On 12 Jun 2006, at 23:58, Thien-Thi Nguyen wrote:
> Alastair Houghton <alastair@alastairs-place.net> writes:
>
>> I'm happy to admit that I missed that part of the documentation,
>> though I'm don't appreciate the undertone of cheekiness in your
>> last remark, particularly not from someone with a gnu.org e-mail
>> address.  Maybe it wasn't intended that way, but that's how it
>> came across.
>
> what do you expect from people w/ a gnu.org e-mail address?  what
> does that expectation have to do w/ not reading the docs properly?

What I expect is that people with gnu.org addresses (who represent  
the FSF,
right?) should be being polite to people who are trying to contribute  
to an FSF
project.  If I'm wrong or have misunderstood something, or have  
missed something
in the docs, I don't mind being told that, but I also don't see why I  
should
put up with remarks I perceive as cheeky given that I'm giving my  
time (and/or
my employer's time) for free.  I don't think that's an unreasonable  
expectation.

If it was not a gnu.org address, that's different because you don't  
have any
control over them and they don't represent the FSF.

(BTW, I'm not certain that David's remark was intended to be cheeky,  
but equally
I'm not altogether certain that it wasn't.  I wasn't really planning  
on starting
a discussion about it either.)

> if the docs are unclear to you in some way, please explain how so.

I think the Elisp manual's node (the larger of the two nodes that  
discuss
file locals) should explicitly state that they are automatically made  
buffer
local.  If it had done, I'd probably have seen it.  The node in the  
Emacs manual
is very short and looks like a menu node with an introduction, rather  
than actual
documentation per se.

Don't misunderstand; the fault here is mine, I should have read it  
more carefully.

Kind regards,

Alastair.

--
http://www.alastairs-place.net

Re: asm-mode patch to allow per-file comment character setting from file locals

[Eli Zaretskii]

> From: Alastair Houghton <alastair@alastairs-place.net>
> Date: Tue, 13 Jun 2006 11:36:29 +0100
> Cc: Stefan Monnier <monnier@iro.umontreal.ca>, emacs-devel@gnu.org
> 
> >>> Files set from a file local variables block are made buffer-local,
> >>> anyway.
> >>
> >> Which isn't stated in the Elisp info documentation (which I *did*
> >> read),
> >
> > File local variable blocks are not documented in the Elisp manual at
> > all.  They are a user-level feature.
> 
> *My* copy of the Elisp manual has a node entitled "File Local Variables".

Which says, as its very first paragraph:

    This section describes the functions and variables that affect
    processing of file local variables.  *Note (emacs)File variables, for
    basic information about file local variables.

So this section describes the _functions_and_variables_ related to
local vars _processing_, not the user-level information for which you
are referred to the user's manual.  And the user manual says, in the
very first paragraph of the node referenced above:

    Visiting the file checks for local variable specifications; it
    automatically makes these variables local to the buffer, and sets
    them to the values specified in the file.

Do we agree now that the information you were looking for is in the
docs?  If you think something in these texts is unclear or missing,
please tell what that is, thanks.

> If I'm wrong or have misunderstood something, or have missed
> something in the docs, I don't mind being told that, but I also
> don't see why I should put up with remarks I perceive as cheeky
> given that I'm giving my time (and/or my employer's time) for free.
> I don't think that's an unreasonable expectation.

I think you should assume a possible misunderstanding, before you
actually decide that a remark that sounds cheeky to you is indeed
cheeky.  Please don't forget that for most of the people here
(including David and myself) English is not their first language.
Fine nuances of English are not always clear to us.

And you are not the only one who gives their time for free.  All the
rest of us, including, but not limited to, those with gnu.org
addresses, are also volunteers, answering questions and helping
maintain Emacs on their own free time.  So the politeness and good
manners are equally reasonably expected on both sides.

> If it was not a gnu.org address, that's different because you don't
> have any control over them and they don't represent the FSF.

People with gnu.org address don't represent the FSF in any way, they
just happen to have accounts on gnu.org machines because they do work
for the GNU project.

> I think the Elisp manual's node (the larger of the two nodes that
> discuss file locals) should explicitly state that they are
> automatically made buffer local.  If it had done, I'd probably have
> seen it.

I'll leave it to Richard to decide whether this information should be
actually added to the ELisp manual, but please understand that we
cannot repeat in the ELisp manual everything that the user manual
says, as it will inflate the ELisp manual too much.  That's what
cross-references are for--to say things in one place, then point there
in several related ones.

> The node in the Emacs manual is very short and looks like a menu
> node with an introduction, rather than actual documentation per se.

See the citation above: what you were looking for is right there
before the menu, I think.

[OT] Netiquette (was: Re: asm-mode patch to allow per-file comment character setting from file locals)

[Alastair Houghton]

On 13 Jun 2006, at 18:56, Eli Zaretskii wrote:

> Do we agree now that the information you were looking for is in the
> docs?  If you think something in these texts is unclear or missing,
> please tell what that is, thanks.

There was never any dispute that the information was in the  
documentation.  Nor was there ever any dispute that the reason I  
didn't know about it was that I hadn't read it carefully enough.

> I think you should assume a possible misunderstanding, before you
> actually decide that a remark that sounds cheeky to you is indeed
> cheeky. Please don't forget that for most of the people here
> (including David and myself) English is not their first language.
> Fine nuances of English are not always clear to us.

If you read my original reply to David Kastrup, or the more recent  
reply to Thien-Thi Nguyen, you will find that I am well aware that it  
may not have been intentional.  I do have a sneaking suspicion that  
it *was* how it was intended to come across though (as I'm well aware  
that it's frustrating that people don't read documentation); but even  
if it wasn't, it would only have taken a second for David to reply  
that he hadn't meant it that way and that he was sorry if that was  
how I'd taken it.

> And you are not the only one who gives their time for free.

I'm well aware of that, and I'm equally well aware that there are  
others who donate much more of their spare time than I do.

> All the rest of us, including, but not limited to, those with gnu.org
> addresses, are also volunteers, answering questions and helping
> maintain Emacs on their own free time.  So the politeness and good
> manners are equally reasonably expected on both sides.

If you can find an instance where I haven't been polite, then I will  
happily apologise for it.

>> If it was not a gnu.org address, that's different because you don't
>> have any control over them and they don't represent the FSF.
>
> People with gnu.org address don't represent the FSF in any way, they
> just happen to have accounts on gnu.org machines because they do work
> for the GNU project.

People with gnu.org addresses represent the FSF every time they use  
that address, in the same sense that people wearing a uniform  
represent the organisation to which the uniform belongs.  That  
doesn't mean that they *speak for* the FSF, of course, in the same  
way that most individual members of (say) the Salvation Army do not  
speak for *their* organisation.  The fact remains, though, that they  
do *represent* that organisation to the wider world.

Anyway, to be clear, I'm really not interested in debating this, and  
I'm certain it's off-topic for the list as well.  All I wanted to do  
was submit my patch.  If you have a technical problem with it, then  
by all means comment (though please look at the most recent version,  
rather than the one in the first message in this thread).

Kind regards,

Alastair.

--
http://www.alastairs-place.net

Re: [OT] Netiquette (was: Re: asm-mode patch to allow per-file comment character setting from file locals)

[Thien-Thi Nguyen]

Alastair Houghton <alastair@alastairs-place.net> writes:

> I do have a sneaking suspicion that [...]

> If you can find an instance where I haven't been polite,
> then I will happily apologise for it.

there is no real need for politeness, nor apologies.
likewise, there is no real need for sneaking suspicion.

all of these can be misrepresented, misinterpreted and
mishandled.  so by all means, make the program precise,
but don't forget to leave slack for the programmers (at
least, until you get to know them personally :-).

for example, you may think i "represent" FSF saying these
things, but i don't mind considering that over time your
conception of the association between a thing and its
name(s) will mature.  in this way, i cut you slack,
personally (FWIW, YMMV, HAND, IMHO :-).

the real need is for programmers who contribute to
emacs to get the paperwork done, if the emacs
maintainer (RMS) thinks the contribution is good,
preferably in a manner that reduces administrative
work in the future.

to make evaluating the patch easier, please repost the
latest version using "diff -c".

thi

Re: [OT] Netiquette (was: Re: asm-mode patch to allow per-file comment character setting from file locals)

[Alastair Houghton]

On 13 Jun 2006, at 23:21, Thien-Thi Nguyen wrote:

> Alastair Houghton <alastair@alastairs-place.net> writes:
>
>> I do have a sneaking suspicion that [...]
>
>> If you can find an instance where I haven't been polite,
>> then I will happily apologise for it.
>
> there is no real need for politeness, nor apologies.
> likewise, there is no real need for sneaking suspicion.

Well I certainly think that people should be polite to one another.   
However, from the perspective of Emacs, I suppose you are right.

> for example, you may think i "represent" FSF saying these
> things,

Not quite in the sense you seem to have understood, but then you are  
not asking for an explanation or arguing about it, and I (like you,  
it seems) would rather get back to the issue of the patch.

> to make evaluating the patch easier, please repost the
> latest version using "diff -c".

Of course.  I've just done so.

Kind regards,

Alastair.

--
http://www.alastairs-place.net

Re: [OT] Netiquette (was: Re: asm-mode patch to allow per-file comment character setting from file locals)

[Eli Zaretskii]

> Cc: emacs-devel@gnu.org
> From: Alastair Houghton <alastair@alastairs-place.net>
> Date: Tue, 13 Jun 2006 20:27:27 +0100
> 
> > All the rest of us, including, but not limited to, those with gnu.org
> > addresses, are also volunteers, answering questions and helping
> > maintain Emacs on their own free time.  So the politeness and good
> > manners are equally reasonably expected on both sides.
> 
> If you can find an instance where I haven't been polite, then I will  
> happily apologise for it.

Your reaction to David's response strikes me as impolite, because it
interpreted a perfectly technical response as having a non-technical
agenda, and pounced on him.

> >> If it was not a gnu.org address, that's different because you don't
> >> have any control over them and they don't represent the FSF.
> >
> > People with gnu.org address don't represent the FSF in any way, they
> > just happen to have accounts on gnu.org machines because they do work
> > for the GNU project.
> 
> People with gnu.org addresses represent the FSF every time they use  
> that address, in the same sense that people wearing a uniform  
> represent the organisation to which the uniform belongs.

You are wrong, and your analogy is wrong.

> Anyway, to be clear, I'm really not interested in debating this

Well, you started it to begin with.

Re: [OT] Netiquette (was: Re: asm-mode patch to allow per-file comment character setting from file locals)

[Alastair Houghton]

On 14 Jun 2006, at 04:30, Eli Zaretskii wrote:

> Your reaction to David's response strikes me as impolite, because it
> interpreted a perfectly technical response as having a non-technical
> agenda, and pounced on him.

It wasn't a "perfectly technical response".  Only the first sentence  
was technical.

>> Anyway, to be clear, I'm really not interested in debating this
>
> Well, you started it to begin with.

No, I didn't.  Thien-Thi Nguyen didn't either; his question (about  
what I felt I expected from people with gnu.org addresses) was  
perfectly reasonable given what I'd written.  *You* "started it".

Alastair.

--
http://www.alastairs-place.net

Re: [OT] Netiquette

[David Kastrup]

Alastair Houghton <alastair@alastairs-place.net> writes:

> On 14 Jun 2006, at 04:30, Eli Zaretskii wrote:
>
>> Your reaction to David's response strikes me as impolite, because it
>> interpreted a perfectly technical response as having a non-technical
>> agenda, and pounced on him.
>
> It wasn't a "perfectly technical response".  Only the first sentence
> was technical.

The whole reply (excluding quotes from your post) was

    Files set from a file local variables block are made buffer-local,
    anyway.  Perhaps reading the documentation will help.

So it consisted of two sentences.  The first, longer sentence
explained your mistake.  The second, shorter one, pointed out that
this was supposed to be covered in the documentation.  Both sentences
were intended to help you deal with your problem.

>>> Anyway, to be clear, I'm really not interested in debating this
>>
>> Well, you started it to begin with.
>
> No, I didn't.

Yes, you did, and by now it is reflecting badly on the attitude of
people with an alastairs-place.net mail address.

I suggest that you drop your diatribe which nobody will miss and
instead try concentrating on the technical merits and issues of your
patch.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: [OT] Netiquette

[Alastair Houghton]

On 14 Jun 2006, at 13:04, David Kastrup wrote:

>>>> Anyway, to be clear, I'm really not interested in debating this
>>>
>>> Well, you started it to begin with.
>>
>> No, I didn't.
>
> Yes, you did,

I most certainly did not.  All *I* did was complain that I didn't  
like your remark; I didn't invite and don't care for a discussion on  
the subject.

> I suggest that you drop your diatribe which nobody will miss and
> instead try concentrating on the technical merits and issues of your
> patch.

Which is what I'd been saying from the beginning (though let's be  
clear: it isn't *my* diatribe; all I did was complain that I found  
your reply cheeky, and I even made clear that I realised it might not  
have been deliberate... all the subsequent pointless argument has  
been started and perpetuated by gnu.org people.)

At this point I'm thoroughly fed-up, so:

1. I won't be contributing to Emacs in future.  Instead, I'll either  
keep my changes to myself or I'll go back to XEmacs.

2. If I see *any* further email on this topic, I will withdraw my  
current patch and won't be signing any copyright assignment forms.  E- 
mail about the patch itself is, of course, fine.

Alastair.

--
http://www.alastairs-place.net

Re: [OT] Netiquette

[Sam Steingold]

> * Alastair Houghton <nynfgnve@nynfgnvef-cynpr.arg> [2006-06-14 16:31:03 +0100]:
>
> On 14 Jun 2006, at 13:04, David Kastrup wrote:
>
>>>>> Anyway, to be clear, I'm really not interested in debating this
>>>> Well, you started it to begin with.
>>> No, I didn't.
>> Yes, you did,
>
> I most certainly did not.  All *I* did was complain that I didn't like
> your remark; I didn't invite and don't care for a discussion on the
> subject.

this constitutes a start in the debate of netiquette, IMHO.

FWIW, I agree that whatever people do reflects on their uniform
(including e-mail domain), but I see nothing wrong in telling someone
that the answer to his question is covered in the manual, except that I
prefer to see a URL next to the note to RTFM.

> At this point I'm thoroughly fed-up, so:
>
> 1. I won't be contributing to Emacs in future.  Instead, I'll either
> keep my changes to myself or I'll go back to XEmacs.
>
> 2. If I see *any* further email on this topic, I will withdraw my
> current patch and won't be signing any copyright assignment forms.  E-
> mail about the patch itself is, of course, fine.

I doubt that these kinds of threats will win you any friends here or
elsewhere.

-- 
Sam Steingold (http://www.podval.org/~sds) on Fedora Core release 5 (Bordeaux)
http://jihadwatch.org http://iris.org.il http://mideasttruth.com
http://truepeace.org http://thereligionofpeace.com http://palestinefacts.org
Isn't "Microsoft Works" an advertisement lie?

Re: [OT] Netiquette

[Alastair Houghton]

On 14 Jun 2006, at 16:43, Sam Steingold wrote:

> I see nothing wrong in telling someone
> that the answer to his question is covered in the manual, except  
> that I
> prefer to see a URL next to the note to RTFM.

No, there's nothing wrong with telling someone that the answer to a  
question is covered in the manual.  But I didn't ask a question, and  
that isn't how David worded his reply.

I also don't see anything wrong with apologising if you offend  
someone, but apparently the correct approach is to attack them for  
taking offence instead, particularly if at the time they're trying to  
help out with some FOSS package you're working on.

>> At this point I'm thoroughly fed-up, so:
>>
>> 1. I won't be contributing to Emacs in future.  Instead, I'll either
>> keep my changes to myself or I'll go back to XEmacs.
>>
>> 2. If I see *any* further email on this topic, I will withdraw my
>> current patch and won't be signing any copyright assignment  
>> forms.  E-
>> mail about the patch itself is, of course, fine.
>
> I doubt that these kinds of threats will win you any friends here or
> elsewhere.

I wasn't after "winning friends", I was after putting an end once and  
for all to this pointless e-mail conversation.  But apparently *you*  
consider it more important to continue it than you do that someone  
makes a useful contribution to Emacs.

Maybe you thought I was bluffing, but I'm afraid I'm not, so---thanks  
to you---I'm withdrawing my patch and I won't be signing any  
copyright forms sent by the FSF.

Alastair.

--
http://www.alastairs-place.net

Re: [OT] Netiquette

[Sam Steingold]

Speaking of netiquette, please fix your newsreader to respect the
         Mail-Copies-To: never
header.  I do not appreciate receiving 2 copies of your e-mails.

-- 
Sam Steingold (http://www.podval.org/~sds) on Fedora Core release 5 (Bordeaux)
http://iris.org.il http://memri.org http://openvotingconsortium.org
http://jihadwatch.org http://camera.org http://thereligionofpeace.com
Linux: Telling Microsoft where to go since 1991.

Re: [OT] Netiquette

[Thien-Thi Nguyen]

Alastair Houghton <alastair@alastairs-place.net> writes:

> Maybe you thought I was bluffing, but I'm afraid I'm not,
> so---thanks to you---I'm withdrawing my patch and I won't
> be signing any copyright forms sent by the FSF.

the choice is yours.  thanks for posting the patch, anyway.
perhaps someone will get ideas from reading it and make
useful changes to asm-mode.el in the future.

some observations for the next patch, should you have a
change of heart:

 - the function asm-set-comment-char allows for very
   dynamic comment char.  i wonder how useful that would
   be in practice -- do assembler language files often
   contain different regions w/ varying comment chars?
   (i.e., can we get same results w/ asm-mode-hook?)

 - docstrings conventionally have the grammar style:
   "Return foo..." instead of "Returns foo...".  also,
   they should mention the function's arguments.

 - error messages conventionally begin w/ a capital
   letter.

 - the need to refontify strikes me as unclean.
   perhaps this is related to the first point above.

see info nodes `(elisp) Tips' and `(elisp) Major Mode'
for more information.  if you have difficulty w/ that
documentation, please start another thread.

(i will bow out of this thread now.)

thi

Re: [OT] Netiquette

[Chong Yidong]

> At this point I'm thoroughly fed-up, so:
>
> 1. I won't be contributing to Emacs in future.  Instead, I'll either
> keep my changes to myself or I'll go back to XEmacs.
>
> 2. If I see *any* further email on this topic, I will withdraw my
> current patch and won't be signing any copyright assignment forms.  E- 
> mail about the patch itself is, of course, fine.

I find this remark cheeky, though this may not be the intended effect.

(My opinions do not, however, reflect those of stupidchicken.com)

Re: [OT] Netiquette

[David Kastrup]

Alastair Houghton <alastair@alastairs-place.net> writes:

> On 14 Jun 2006, at 13:04, David Kastrup wrote:
>
>>>>> Anyway, to be clear, I'm really not interested in debating this
>>>>
>>>> Well, you started it to begin with.
>>>
>>> No, I didn't.
>>
>> Yes, you did,
>
> I most certainly did not.  All *I* did was complain that I didn't
> like your remark; I didn't invite and don't care for a discussion on
> the subject.

So if others don't like your remarks and complain, they start a
discussion, whereas you do no such thing when doing the same?

In particular when lecturing about the moral obligations of people
with a gnu.org address, when quite a few of those are hanging out on
the list?  It does not appear to you that you are asking people to
behave differently from what you do yourself?

This is a developer list.  Developers are not necessarily perfect
diplomats, and a lot of technical and design controversies are fought
out on the list.  I think it unlikely that you'll find a single list
participant who will state that he finds all others on the list
perfectly pleasant (and I certainly don't do much to improve the
record).  But the ultimate purpose of this list is to improve Emacs,
not improve the other developers.

My pointer to the documentation was not very helpful, but I was tired
when I posted it, and I remembered that I had read this in the
documentation.  I did not remember actually where, and I was not in
the mood to dig around.

If you want to have the last word on this matter, that's fine with me
and I'd ask others to leave it to you.  It is obvious enough that all
that could have been worth saying has been said.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: [OT] Netiquette

[Alastair Houghton]

On 14 Jun 2006, at 17:26, David Kastrup wrote:

> So if others don't like your remarks and complain, they start a
> discussion, whereas you do no such thing when doing the same?

I don't recall asking to discuss netiquette.  What I do recall is  
that once upon a time I was interested in submitting a patch.  I am  
not, now.

It shouldn't be surprising that if someone complains about something  
you have said, and then you and others proceed to attack them for  
complaining, they won't want to work with you.

Alastair.

--
http://www.alastairs-place.net

Re: [OT] Netiquette

[Sam Steingold]

> * Alastair Houghton <nynfgnve@nynfgnvef-cynpr.arg> [2006-06-14 17:40:34 +0100]:
>
> It shouldn't be surprising that if someone complains about something
> you have said, and then you and others proceed to attack them for
> complaining, they won't want to work with you.

1. I do not think your complaint was justified.

2. I do not think that you have been "attacked" any more than you
   "attacked" whoever you were unhappy about.

It is unfortunate that the conversation has degenerated into this.

-- 
Sam Steingold (http://www.podval.org/~sds) on Fedora Core release 5 (Bordeaux)
http://openvotingconsortium.org http://pmw.org.il http://jihadwatch.org
http://truepeace.org http://iris.org.il http://dhimmi.com http://ffii.org
Why use Windows, when there are Doors?

Re: [OT] Netiquette

[Richard Stallman]

    > It shouldn't be surprising that if someone complains about something
    > you have said, and then you and others proceed to attack them for
    > complaining, they won't want to work with you.

David Kastrup's remark was somewhat harsh, harsh enough that he ought
not to have said it.  Your complaint was justified, as far as that
goes.

However, what he said wasn't really an attack on you, just somewhat
harsh.  So I think that, even though you had grounds for a complaint,
it would have been wiser for you to let the matter slide.

It would likewise have been better, and wiser, if other people had not
responded to your complaint by arguing with you.  They were right in a
limited sense, that you had made a mistaken assumption and thus
reached a conclusion that was a little too strong.  But that was a
side issue, and they should not have argued about it.

Some people, after seeing that your feelings were hurt, responded in
an aggressive tone.  That was a bad thing to do, but I see why they
did it.  Your statement came across as a demand: "Everyone shut up
instantly, or I will refuse to contribute my code."  So their response
really meant, "I'll prove you can't order me to shut up!"  It was
foolish to respond that way, but it was almost inevitable that someone
would.

It seems that both you and others passed up opportunities to let the
quarrel end (by not responding).  If only people had used those
opportunities, we would have had a good outcome.

I hope that those who continue contributing to Emacs will learn the
lesson not to continue arguments about side issues where someone's
feelings have been hurt.  Don't take the risk of hurting the person
more.  Let it drop!

Re: [OT] Netiquette

[Alastair Houghton]

On 15 Jun 2006, at 09:30, Richard Stallman wrote:

>> It shouldn't be surprising that if someone complains about something
>> you have said, and then you and others proceed to attack them for
>> complaining, they won't want to work with you.
>
> David Kastrup's remark was somewhat harsh, harsh enough that he ought
> not to have said it.  Your complaint was justified, as far as that
> goes.
>
> However, what he said wasn't really an attack on you, just somewhat
> harsh.  So I think that, even though you had grounds for a complaint,
> it would have been wiser for you to let the matter slide.

Yes, I think I'd agree with that in hindsight, but it's much easier  
to say that now.

> It seems that both you and others passed up opportunities to let the
> quarrel end (by not responding).  If only people had used those
> opportunities, we would have had a good outcome.

Yes, both sides share the guilt to some degree and I for my part am  
sorry that it happened.  It's wasted a lot of time that people could  
better have spent either working on their job (which is my priority  
right now) or contributing to Emacs.

In light of your reply, I will be happy to sign any copyright  
assignment form that I do receive.  I still feel, sadly, that I'm now  
much less likely to contribute in future, but who knows; maybe I will  
change my mind at some point.

Kind regards,

Alastair.

--
http://www.alastairs-place.net

Re: [OT] Netiquette

[Richard Stallman]

    In light of your reply, I will be happy to sign any copyright  
    assignment form that I do receive.

Thank you.  Did you email that form as I asked?
If so, you should receive the papers in a week or so.

I think asm-set-comment-char needs to handle nil as an argument
meaning no comment character.  Interactively, it should provide a way
to say "none".

Re: asm-mode comment char patch (was: [OT] Netiquette)

[Alastair Houghton]

On 16 Jun 2006, at 07:01, Richard Stallman wrote:

>     In light of your reply, I will be happy to sign any copyright
>     assignment form that I do receive.
>
> Thank you.  Did you email that form as I asked?

Yes.

> If so, you should receive the papers in a week or so.

OK.  I'll get them sent back as soon as possible after they arrive.

> I think asm-set-comment-char needs to handle nil as an argument
> meaning no comment character.  Interactively, it should provide a way
> to say "none".

I'm probably going to be pretty busy until Tuesday of next week with  
work, so most likely I won't be able to add that feature until then  
(though I agree, it makes sense that you should be able to specify  
none as an option).  When I do, I'll post a new version of the  
patch.  I had a couple of other suggestions for improvements from  
Thien-Thi Nguyen that I'll try to incorporate as well.

Kind regards,

Alastair.

--
http://www.alastairs-place.net

Re: asm-mode comment char patch (was: [OT] Netiquette)

[Richard Stallman]

Happy hacking.

Re: [OT] Netiquette

[Miles Bader]

Alastair Houghton <alastair@alastairs-place.net> writes:
> Which is what I'd been saying from the beginning (though let's be  
> clear: it isn't *my* diatribe

It seems to have become so though:

> 1. I won't be contributing to Emacs in future.  Instead, I'll either  
> keep my changes to myself or I'll go back to XEmacs.
>
> 2. If I see *any* further email on this topic, I will withdraw my  
> current patch and won't be signing any copyright assignment forms.  E- 

Yeah, _that's_ a mature response...

-Miles
-- 
Would you like fries with that?

Re: asm-mode patch to allow per-file comment character setting from file locals

[Miles Bader]

Alastair Houghton <alastair@alastairs-place.net> writes:
> What I expect is that people with gnu.org addresses (who represent the
> FSF, right?)

Not really, no.

You seem to be thinking of GNU as some sort of business -- businesses
often maintain fairly close control over who has access to their systems
and for what purposes they use them, and usually try to present a
unified public front.  GNU is not like that.

Rather maybe think of it like a FOO.edu address.

-Miles

-- 
"An atheist doesn't have to be someone who thinks he has a proof that there
can't be a god.  He only has to be someone who believes that the evidence
on the God question is at a similar level to the evidence on the werewolf
question."  [John McCarthy]

[OT] Whether those with gnu.org addresses in any way represent FSF (was: Re: asm-mode patch to allow per-file comment character setting from file locals)

[Alastair Houghton]

On 14 Jun 2006, at 06:11, Miles Bader wrote:

> Alastair Houghton <alastair@alastairs-place.net> writes:
>> What I expect is that people with gnu.org addresses (who represent  
>> the
>> FSF, right?)
>
> Not really, no.
>
> You seem to be thinking of GNU as some sort of business -- businesses
> often maintain fairly close control over who has access to their  
> systems
> and for what purposes they use them, and usually try to present a
> unified public front.  GNU is not like that.
>
> Rather maybe think of it like a FOO.edu address.

Actually, that's *exactly* how I think of it.  The behaviour of a  
student with a FOO.edu address reflects on the institution of which  
they are a part, and it's in that sense that people with gnu.org  
addresses represent the FSF.

Kind regards,

Alastair.

--
http://www.alastairs-place.net

Re: [OT] Whether those with gnu.org addresses in any way represent FSF

[Miles Bader]

Alastair Houghton <alastair@alastairs-place.net> writes:
>> Rather maybe think of it like a FOO.edu address.
>
> Actually, that's *exactly* how I think of it.  The behaviour of a
> student with a FOO.edu address reflects on the institution of which
> they are a part, and it's in that sense that people with gnu.org
> addresses represent the FSF.

"Reflects on", sure, but rather weakly.  Nobody's going to lose much
sleep if a Harvard student flames somebody on a mailing list somewhere,
and it won't have much effect on Harvard's reputation, because it's
generally understood that there are lots of silly people at universities,
and that no attempt is made to control their behavior.

-Miles
-- 
"Nah, there's no bigger atheist than me.  Well, I take that back.
I'm a cancer screening away from going agnostic and a biopsy away
from full-fledged Christian."  [Adam Carolla]

Re: [OT] Whether those with gnu.org addresses in any way represent FSF

[Alastair Houghton]

On 14 Jun 2006, at 13:29, Miles Bader wrote:

> Alastair Houghton <alastair@alastairs-place.net> writes:
>>> Rather maybe think of it like a FOO.edu address.
>>
>> Actually, that's *exactly* how I think of it.  The behaviour of a
>> student with a FOO.edu address reflects on the institution of which
>> they are a part, and it's in that sense that people with gnu.org
>> addresses represent the FSF.
>
> "Reflects on", sure, but rather weakly.  Nobody's going to lose much
> sleep if a Harvard student flames somebody on a mailing list  
> somewhere,
> and it won't have much effect on Harvard's reputation, because it's
> generally understood that there are lots of silly people at  
> universities,
> and that no attempt is made to control their behavior.

OK, but there aren't lots of silly people with gnu.org addresses.   
Very few, if any, I suspect.

Kind regards,

Alastair.

--
http://www.alastairs-place.net

Re: [OT] Whether those with gnu.org addresses in any way represent FSF

[nferrier]

Alastair Houghton <alastair@alastairs-place.net> writes:

> On 14 Jun 2006, at 13:29, Miles Bader wrote:
>
>> Alastair Houghton <alastair@alastairs-place.net> writes:
>>>> Rather maybe think of it like a FOO.edu address.
>>>
>>> Actually, that's *exactly* how I think of it.  The behaviour of a
>>> student with a FOO.edu address reflects on the institution of which
>>> they are a part, and it's in that sense that people with gnu.org
>>> addresses represent the FSF.
>>
>> "Reflects on", sure, but rather weakly.  Nobody's going to lose much
>> sleep if a Harvard student flames somebody on a mailing list  
>> somewhere,
>> and it won't have much effect on Harvard's reputation, because it's
>> generally understood that there are lots of silly people at  
>> universities,
>> and that no attempt is made to control their behavior.
>
> OK, but there aren't lots of silly people with gnu.org addresses.   
> Very few, if any, I suspect.

I am extreemly silly.

-- 
Nic Ferrier
http://www.tapsellferrier.co.uk   for all your tapsell ferrier needs

Re: asm-mode patch to allow per-file comment character setting from file locals

[Thien-Thi Nguyen]

Alastair Houghton <alastair@alastairs-place.net> writes:

> I'm happy to admit that I missed that part of the documentation,
> though I'm don't appreciate the undertone of cheekiness in your
> last remark, particularly not from someone with a gnu.org e-mail
> address.  Maybe it wasn't intended that way, but that's how it
> came across.

what do you expect from people w/ a gnu.org e-mail address?  what
does that expectation have to do w/ not reading the docs properly?

if the docs are unclear to you in some way, please explain how so.

thi

Re: asm-mode patch to allow per-file comment character setting from file locals

[Masatake YAMATO]

> Hi there,
> 
> Here's a short patch to asm-mode.el that I use so that I can easily  
> vary the comment style used for assembly language files on a per-file  
> basis.  This is particularly useful on Mac OS X, where the Intel and  
> PowerPC versions of the system assembler use different comment  
> characters (I believe the PowerPC version actually accepts both ';'  
> and '#', but the Intel one certainly requires '#').
> 
> An example of its use:
> 
>    ###
>    ### Local Variables:
>    ### asm-comment-char: ?\#
>    ### End:
>    ###
> 
> Here's the patch:

Interesting.

I have met with the same situation when I read linux/arch/*/*.S files.
Font lock rules of asm-mode are not enough for the situation.
So I tried to write modes derived from asm-mode for each architecture/asm syntax.
I refer the info file of gas to know the asm syntaxes, so I called the derived
modes gas-*-mode, here * is a name of architecture.

You can do like: 

    M-x gas-i386-mode
,
    M-x gas-ppc-mode
or
    M-x gas-arm-mode if you want to hack iPod:-P

If your asm file is at /foo/bar/i386/baz.s or /foo/bar/ppc/baz.s,
gas-i386-mode or gas-ppc-mode is automatically selected.

Masatake YAMATO

(require 'asm-mode)
(require 'assoc)

(defgroup gas nil
  "Architecture specific mode for editing assembler code."
  :group 'asm)

(defcustom gas-mode-hook nil
  "*Hook that gets run after the gas mode ."
  :type  'hook
  :group 'asm)

(defvar gas-mode-architecture-history nil)
(defun gas-mode (&optional architecture)
  "Wrapper for gas-*-mode."
  (interactive (list 
		(let ((md (gas-choose-mode-automatically)))
		  (if current-prefix-arg
		      (completing-read "Architecture: "
				       gas-machine-dependents
				       nil
				       t
				       md
				       'gas-mode-architecture-history
				       md)
		    (gas-choose-mode-automatically)))))
  (unless (interactive-p)
    (setq architecture (gas-choose-mode-automatically)))
  (if architecture
      (let* ((machine-dependent (aget gas-machine-dependents architecture))
	     (mode-func         (nth 0 machine-dependent))
	     (asm-comment-chars (car (nth 1 machine-dependent))))
	(call-interactively mode-func))
    (asm-mode))
  (run-mode-hooks 'gas-mode-hook))

(defvar gas-machine-dependents nil)
(defun gas-register-machine-dependent (name
				       comment-chars
				       register-prefix-chars
				       immediate-prefix-chars
				       extra-symbol-chars
				       statement-separator-char
				       auto-mode)
  (let ((mode-func (intern (format "gas-%s-mode" name))))
    (aput 'gas-machine-dependents name 
	  (list mode-func comment-chars auto-mode))
    (eval
     `(define-derived-mode ,mode-func asm-mode ,(format "Gas/%s" name)
	,(format "%s specific asm mode" name)
	:syntax-table (gas-make-syntax-table 
		       ',comment-chars
		       ',register-prefix-chars
		       ',immediate-prefix-chars
		       ',extra-symbol-chars 
		       ',statement-separator-char)
	(make-local-variable 'font-lock-defaults)
	(setq font-lock-defaults (list 
				  (gas-make-font-lock-keywords ',comment-chars
							       ',register-prefix-chars
							       ',immediate-prefix-chars
							       ',extra-symbol-chars 
							       ',statement-separator-char)))))))

(defun gas-choose-mode-automatically ()
  (catch 'found
    (dolist (md gas-machine-dependents)
      (let ((auto-mode (nth 3 md)))
	(when (eq auto-mode t)
	  (setq auto-mode (car md)))
	(cond 
	 ((and (stringp auto-mode)
	       (buffer-file-name))
	  (when (string-match auto-mode (buffer-file-name))
	    (throw 'found (car md))))
	 ((functionp auto-mode)
	  (when (funcall 'auto-mode)
	    (throw 'found (car md)))))))))

(defun gas-make-syntax-table (comment-chars
			      register-prefix-chars
			      immediate-prefix-chars
			      extra-symbol-chars
			      statement-separator-char)
  (let ((st (copy-syntax-table asm-mode-syntax-table)))
    (mapc (lambda (c) (modify-syntax-entry c "< b" st))
	  comment-chars)
    (mapc (lambda (c) (modify-syntax-entry c "'" st))
	  register-prefix-chars)
    (mapc (lambda (c) (modify-syntax-entry c "'" st))
	  immediate-prefix-chars)
    (mapc (lambda (c) (modify-syntax-entry c "w" st)) ; "_"?
	  extra-symbol-chars)
    (mapc (lambda (c) (modify-syntax-entry c " " st))
	  statement-separator-char)
    st))

(defun gas-make-font-lock-keywords (comment-chars
				    register-prefix-chars
				    immediate-prefix-chars
				    extra-symbol-chars
				    statement-separator-chars)
  (let ((rp (when register-prefix-chars
	      (cons (concat 
		     "[" (mapconcat 'char-to-string  register-prefix-chars "") 
		     "]" 
		     "\\sw+") font-lock-variable-name-face)))
	(ip (when immediate-prefix-chars
	      (cons (concat
		     "[" (mapconcat 'char-to-string immediate-prefix-chars "") 
		     "]" 
		     "\\(?:\\s_\\|\\sw\\)+") font-lock-constant-face)))
	(pattern))
    (when rp (push rp pattern))
    (when ip (push ip pattern))
    (append
     pattern
     asm-font-lock-keywords)))

;; c:comment-char 
;; r:register-prefix-chars 
;; i:immediate-prefix-chars 
;; sym:extra-symbol-chars 
;; sep:statement-separator-char
;; am: regexp for auto mode selection.
;;                              name       c           r      i        sym        sep   am
(gas-register-machine-dependent "alpha"   '(?#)       '(?$)   nil      nil       '(?\;) t)
(gas-register-machine-dependent "arm"     '(?@)        nil   '(?# ?$) '(?_ ?\\)  '(?\;) t)
(gas-register-machine-dependent "cris"    '(?\; ?#)   '(?$)   nil      nil       '(?\@) t)

;; frv???
(gas-register-machine-dependent "frv"     '(?\; ?#)    nil    nil      nil        nil   t)

(gas-register-machine-dependent "h8300"   '(?\;)      '(?@)   '(?#)   '(?_)      '(?\$) t)
;; #include doesn't work well.
(gas-register-machine-dependent "i386"    '(?#)       '(?%)   '(?$)    nil       '(?\;) "i386\\|x86_64")
;; comment is broken
(gas-register-machine-dependent "ia64"     nil         nil     nil     nil       '(?\;) t)
(gas-register-machine-dependent "m32r"    '(?\;)      '(?@)   '(?#)   '(?+)       nil   t)
(gas-register-machine-dependent "m68k"    '(?\|)      '(?%)   '(?#)   '(?@)       nil   "68k")
(gas-register-machine-dependent "mips"    '(?#)       '(?$)    nil    '(?@)       nil   t)
(gas-register-machine-dependent "parisc"  '(?\;)      '(?%)    nil    '(?$)      '(?!)  t)
(gas-register-machine-dependent "ppc"      nil         nil     nil     nil       '(?\;) "powerpc\\|ppc")
(gas-register-machine-dependent "s390"    '(?#)       '(?%)    nil     nil        nil   t)
(gas-register-machine-dependent "sh"      '(?!)       '(?@)   '(?#)   '(?$ ?-)   '(?\;) t)
(gas-register-machine-dependent "sparc"   '(?!)       '(?%)    nil     nil       '(?\;) t)
(gas-register-machine-dependent "v850"    '(?#)        nil     nil     nil       '(?\;) t)
(gas-register-machine-dependent "xtensa"  '(?#)       '(?$)    nil     nil       '(?\;) t)

(add-to-list 'auto-mode-alist '("\\.[sS]\\'" . gas-mode))
(provide 'gas-mode)
;; gas-mode.el ends here

Re: asm-mode patch to allow per-file comment character setting from file locals

[Richard Stallman]

When you have a construct such as gas-register-machine-dependent
that generates a defun, it should be a macro, not a function.
The macro name should start with `def'.

And the name of the function it defines should not be constructed
like this

  (let ((mode-func (intern (format "gas-%s-mode" name))))

Instead, the name of the function should appear as the first argument
of the macro.

Those are general principles for Lisp programming which ought to be
applied here.


Comparing the usefulness for editing of these two proposed extensions
to asm mode is a different question.  It would be useful for those
who are interested in such things to compare them and state their
opinions.