EmacsW32 invocation options

EmacsW32 invocation options

[Will]

Hi, there!

I'd like to use a shortcut on my desktop to start GNU Emacs with W32 
bundle (patched, V1.46 - V1.50) so that my personal init-file is 
evaluated at start-up, which is _not_ in my home directory. The 
following commands however do not work (cf. Shortcut > Target):

"C:\Program Files\Emacs\emacs\bin\emacsclientw.exe" -n  -e 
"Z:\00_Preferences\Emacs\.emacs"

"C:\Program Files\Emacs\emacs\bin\emacsclientw.exe" -n  --eval 
"Z:\00_Preferences\Emacs\.emacs"

What am I doing wrong?

Thanks for help in advance,

Will

Re: EmacsW32 invocation options

[Eli Zaretskii]

> From: Will <schimpanski@gmx.de>
> Date: Tue, 01 May 2007 15:15:24 +0200
> 
> I'd like to use a shortcut on my desktop to start GNU Emacs with W32 
> bundle (patched, V1.46 - V1.50) so that my personal init-file is 
> evaluated at start-up, which is _not_ in my home directory. The 
> following commands however do not work (cf. Shortcut > Target):
> 
> "C:\Program Files\Emacs\emacs\bin\emacsclientw.exe" -n  -e 
> "Z:\00_Preferences\Emacs\.emacs"
> 
> "C:\Program Files\Emacs\emacs\bin\emacsclientw.exe" -n  --eval 
> "Z:\00_Preferences\Emacs\.emacs"
> 
> What am I doing wrong?

Why do you make this so complicated?  Why not

   "C:\Program Files\Emacs\emacs\bin\runemacs.exe" --eval 
         "Z:\00_Preferences\Emacs\.emacs"

? why use emacsclientw?

Btw, please note that eval'ing .emacs like this will work in subtly
different ways that what Emacs does with .emacs found in your home
directory.  Some customizations, especially those related to setup of
the Emacs display, will only work correctly when Emacs finds .emacs in
your home directory, because then it evaluates .emacs before the
display is set, whereas --eval is done after that, when it's too late
to influence how Emacs initial frames and windows are set.

Re: EmacsW32 invocation options

[Lennart Borgman (gmail)]

Eli Zaretskii wrote:
>> From: Will <schimpanski@gmx.de>
>> Date: Tue, 01 May 2007 15:15:24 +0200
>>
>> I'd like to use a shortcut on my desktop to start GNU Emacs with W32 
>> bundle (patched, V1.46 - V1.50) so that my personal init-file is 
>> evaluated at start-up, which is _not_ in my home directory. The 
>> following commands however do not work (cf. Shortcut > Target):
>>
>> "C:\Program Files\Emacs\emacs\bin\emacsclientw.exe" -n  -e 
>> "Z:\00_Preferences\Emacs\.emacs"
>>
>> "C:\Program Files\Emacs\emacs\bin\emacsclientw.exe" -n  --eval 
>> "Z:\00_Preferences\Emacs\.emacs"
>>
>> What am I doing wrong?
> 
> Why do you make this so complicated?  Why not
> 
>    "C:\Program Files\Emacs\emacs\bin\runemacs.exe" --eval 
>          "Z:\00_Preferences\Emacs\.emacs"
> 
> ? why use emacsclientw?


That is good advice for the unpatched version of Emacs, but for the 
patched version that comes with Emacs+EmacsW32 it should be handled 
differently.

The reason is that with the patched version you do not have to start 
Emacs server in your .emacs, it is started automatically by the first 
invocation of emacsclient. To still use that feature you can do 
something like this:

   emacsclientw.exe -n -e "load-file \"path-to-your-file\""

BTW there is a misunderstanding above. When you use -e the parameter is 
evaled, not the contents in the file. So you are trying to use the file 
name as an emacs lisp expression. (Fortunately that fails.)

Re: EmacsW32 invocation options

[Lennart Borgman (gmail)]

Lennart Borgman (gmail) wrote:

>   emacsclientw.exe -n -e "load-file \"path-to-your-file\""

Eh, and that should be

   emacsclientw.exe -n -e "(load-file \"c:/path/to/your/file\")"

and you should use FORWARD slashes in the file names.

Re: EmacsW32 invocation options

[Eli Zaretskii]

> Date: Tue, 01 May 2007 22:31:18 +0200
> From: "Lennart Borgman (gmail)" <lennart.borgman@gmail.com>
> CC: Eli Zaretskii <eliz@gnu.org>,  help-gnu-emacs@gnu.org
> 
> >> "C:\Program Files\Emacs\emacs\bin\emacsclientw.exe" -n  --eval 
> >> "Z:\00_Preferences\Emacs\.emacs"
> >>
> >> What am I doing wrong?
> > 
> > Why do you make this so complicated?  Why not
> > 
> >    "C:\Program Files\Emacs\emacs\bin\runemacs.exe" --eval 
> >          "Z:\00_Preferences\Emacs\.emacs"
> > 
> > ? why use emacsclientw?
> 
> 
> That is good advice for the unpatched version of Emacs, but for the 
> patched version that comes with Emacs+EmacsW32 it should be handled 
> differently.
> 
> The reason is that with the patched version you do not have to start 
> Emacs server in your .emacs, it is started automatically by the first 
> invocation of emacsclient.

So not only is your Emacs patched and behaves differently, it also
teaches users habits that will prevent them from using the same .emacs
on other systems?  I find this disturbing.

Re: EmacsW32 invocation options

[Brian Elmegaard]

Eli Zaretskii <eliz@gnu.org> writes:

> So not only is your Emacs patched and behaves differently, it also
> teaches users habits that will prevent them from using the same .emacs
> on other systems?  I find this disturbing.

Why do users like to have to start emacs server themselves?

-- 
Brian (remove the sport for mail)
http://www.et.web.mek.dtu.dk/Staff/be/be.html
http://www.rugbyklubben-speed.dk

Re: EmacsW32 invocation options

[Lennart Borgman (gmail)]

Eli Zaretskii wrote:

>> The reason is that with the patched version you do not have to start 
>> Emacs server in your .emacs, it is started automatically by the first 
>> invocation of emacsclient.
> 
> So not only is your Emacs patched and behaves differently, it also
> teaches users habits that will prevent them from using the same .emacs
> on other systems?  I find this disturbing.


You know very well, Eli, that we have discussed if Emacs server should 
be started automatically when Emacs client is invoked. I have offered 
the code that does so, but you are one of those that did not want that 
code to be in Emacs at the moment.

I think it is one of the important features in the patched version that 
makes it quite a bit easier to get started using Emacs for new users. 
Not one single of the new users have complained about this.

Re: EmacsW32 invocation options

[Juanma Barranquero]

On 5/2/07, Lennart Borgman (gmail) <lennart.borgman@gmail.com> wrote:

> You know very well, Eli, that we have discussed if Emacs server should
> be started automatically when Emacs client is invoked. I have offered
> the code that does so, but you are one of those that did not want that
> code to be in Emacs at the moment.

I'm obviously not Eli, but I think there were good reasons for not
wanting the code in Emacs at the present. For one, althought I like
the functionality, I think it could benefit from some more discussion;
also, I don't like the implementation.

> I think it is one of the important features in the patched version that
> makes it quite a bit easier to get started using Emacs for new users.

I don't think "quite a bit easier" is a fair description. We're
talking of the difficulty of starting Emacs as a server before using
emacsclient. Hardly rocket science.

> Not one single of the new users have complained about this.

Users who enter Emacs through your distribution don't know it is
currently supposed to be different. Why should they complain, unless
it causes problems (which is not the same that being better, of
course)? And I seem to remember at least a few bug reports...

             Juanma

Re: EmacsW32 invocation options

[Hadron]

"Lennart Borgman (gmail)" <lennart.borgman@gmail.com> writes:

> Eli Zaretskii wrote:
>
>>> The reason is that with the patched version you do not have to
>>> start Emacs server in your .emacs, it is started automatically by
>>> the first invocation of emacsclient.
>>
>> So not only is your Emacs patched and behaves differently, it also
>> teaches users habits that will prevent them from using the same .emacs
>> on other systems?  I find this disturbing.
>
>
> You know very well, Eli, that we have discussed if Emacs server should
> be started automatically when Emacs client is invoked. I have offered

I would vote for this.

The current way it works seems to make little sense. If someone is
running the client then clearly they want the server. If there is no
server, start it. Makes perfect sense.

Re: EmacsW32 invocation options

[Hadron]

"Juanma Barranquero" <lekktu@gmail.com> writes:

> On 5/2/07, Lennart Borgman (gmail) <lennart.borgman@gmail.com> wrote:
>
>> You know very well, Eli, that we have discussed if Emacs server should
>> be started automatically when Emacs client is invoked. I have offered
>> the code that does so, but you are one of those that did not want that
>> code to be in Emacs at the moment.
>
> I'm obviously not Eli, but I think there were good reasons for not
> wanting the code in Emacs at the present. For one, althought I like
> the functionality, I think it could benefit from some more discussion;
> also, I don't like the implementation.
>
>> I think it is one of the important features in the patched version that
>> makes it quite a bit easier to get started using Emacs for new users.
>
> I don't think "quite a bit easier" is a fair description. We're
> talking of the difficulty of starting Emacs as a server before using
> emacsclient. Hardly rocket science.

Is there a solid reason NOT to start the server automatically if someone
starts a client? Clearly they want the server to be running. Why have
something "easy enough to do" when you can have it "nothing to do"?

Re: EmacsW32 invocation options

[Lennart Borgman (gmail)]

Juanma Barranquero wrote:

> I'm obviously not Eli, but I think there were good reasons for not
> wanting the code in Emacs at the present. For one, althought I like
> the functionality, I think it could benefit from some more discussion;
> also, I don't like the implementation.

Yes, more discussion would be good. I really tried to open a discussion 
about it, but I did not get much company.


>> Not one single of the new users have complained about this.
> 
> Users who enter Emacs through your distribution don't know it is
> currently supposed to be different.

Did you read the download page for Emacs+EmacsW32? Did you read the 
documentation on the web pages about the patches?


> And I seem to remember at least a few bug reports...

I am not aware of any bugs that have not been fixed. None of the 
problems where related to the changes to Emacs client.

Re: EmacsW32 invocation options

[Juanma Barranquero]

On 5/2/07, Brian Elmegaard <brian@rkspeed-rugby.dk> wrote:

> Why do users like to have to start emacs server themselves?

Because then they have full control over what's running, and how.
Irrelevant if you only run one instance of Emacs, or just one release,
or always with the same setup.

             Juanma

Re: EmacsW32 invocation options

[Juanma Barranquero]

On 5/2/07, Hadron <hadronquark@gmail.com> wrote:

> The current way it works seems to make little sense. If someone is
> running the client then clearly they want the server. If there is no
> server, start it. Makes perfect sense.

Which server? With which options? Local or remote?

             Juanma

Re: EmacsW32 invocation options

[Juanma Barranquero]

On 5/2/07, Hadron <hadronquark@gmail.com> wrote:

> Is there a solid reason NOT to start the server automatically if someone
> starts a client?

Yes. You can read the discussions in the emacs-devel list (those that
Lennart seems to think never happened). Basically, for the client to
be able to start the server, either you put too much "inteligence" in
the client, or you have a way to clearly specify how to pass that info
from the client to the server. I'm quite sure there's a clean way to
do that, and I'm also quite sure it's not the one in EmacsW32-*, at
least the last time I took a look at it.

Also, there's another good reason NOT to start the server
automatically: there was *not enough* time during the Emacs 22
freeze/pretest to properly develop that feature (hell, we were on the
verge of simply *not* supporting emacsclient on Windows at all, and I
know that for a fact). Lennart's rolled his own, which means that
either we're forced to follow his path, even if we don't like it, or
to do an incompatible change later.

             Juanma

Re: EmacsW32 invocation options

[Juanma Barranquero]

On 5/2/07, Lennart Borgman (gmail) <lennart.borgman@gmail.com> wrote:

> Yes, more discussion would be good. I really tried to open a discussion
> about it, but I did not get much company.

I'd say you're being unfair, but that would be far too nice. You seem
to have suddenly forgotten tens of messages (public and private) and
quite a few hours devoted to this and related issues. Once again I get
the feeling that you're very conservative with your time, and quite
liberal with other people's.

Also, the fact that your patches (for this issue) where not included
is not because "you did not get much company"; it is because they were
late, not convincing to some people (me included), and not deemed
urgent enough; and, to my knowledge, not agreeing with your assessment
of a problem's gravity/urgency is not a failure in the Emacs
development process. I'm sure you'll disagree.

> Did you read the download page for Emacs+EmacsW32? Did you read the
> documentation on the web pages about the patches?

Yes, Lennart, I did. Once. Fortunately, I'm not morally or
contractually obliged to memorize every single word you write.

> I am not aware of any bugs that have not been fixed.

I haven't said "not fixed".

> None of the
> problems where related to the changes to Emacs client.

I'm speaking from memory, but I remember at least one bug report that
explicitly mentioned trouble in the automatic start of the server
(though, if my memory doesn't fail, you fixed it in your patches). But
that's irrelevant; I've said that I seem to remember a few problems,
but that does not affect in any way whether I think that your patches
are a good or a bad idea (as long as you do the support for them,
which you do).

             Juanma

Re: EmacsW32 invocation options

[Lennart Borgman (gmail)]

Juanma Barranquero wrote:
> On 5/2/07, Hadron <hadronquark@gmail.com> wrote:
> 
>> Is there a solid reason NOT to start the server automatically if someone
>> starts a client?
> 
> Yes. You can read the discussions in the emacs-devel list (those that
> Lennart seems to think never happened). Basically, for the client to
> be able to start the server, either you put too much "inteligence" in
> the client, or you have a way to clearly specify how to pass that info
> from the client to the server. I'm quite sure there's a clean way to
> do that, and I'm also quite sure it's not the one in EmacsW32-*, at
> least the last time I took a look at it.

I am sure you know these things much better than me, I just wanted 
something that worked now. You did the heroes work when you got Emacs 
client to run on MS Windows. I only reorganized what you wrote and added 
simple functions for user feedback.

I also added a switch

   --server-file FILENAME

for which I wrote "Set filename of the TCP authentication file". I do 
not know if that is enough, since I did not get any comments on that and 
I have hardly used it myself.


> Also, there's another good reason NOT to start the server
> automatically: there was *not enough* time during the Emacs 22
> freeze/pretest to properly develop that feature (hell, we were on the
> verge of simply *not* supporting emacsclient on Windows at all, and I
> know that for a fact). Lennart's rolled his own, which means that
> either we're forced to follow his path, even if we don't like it, or
> to do an incompatible change later.

Maybe I am missing something, but I tried to avoid compatibility 
problems. My guess is that the default action will be to start the local 
server just as it is started in .emacs with (server-start). That is what 
is done by my patched version of Emacs client now.

As I have pointed out several times there are no changes to Emacs itself 
for this.


Another problem is if the code runs on other OS:es. It would be nice if 
someone tested that. If someone wants to do that, please contact me.

Re: EmacsW32 invocation options

[Hadron]

"Juanma Barranquero" <lekktu@gmail.com> writes:

> On 5/2/07, Brian Elmegaard <brian@rkspeed-rugby.dk> wrote:
>
>> Why do users like to have to start emacs server themselves?
>
> Because then they have full control over what's running, and how.
> Irrelevant if you only run one instance of Emacs, or just one release,
> or always with the same setup.
>

I don't see that as true at all. If you run emacsclient you obviously do
it with the intention of connecting to emacs-server.

Re: EmacsW32 invocation options

[Hadron]

"Juanma Barranquero" <lekktu@gmail.com> writes:

> On 5/2/07, Hadron <hadronquark@gmail.com> wrote:
>
>> The current way it works seems to make little sense. If someone is
>> running the client then clearly they want the server. If there is no
>> server, start it. Makes perfect sense.
>
> Which server? With which options? Local or remote?

The configured one I would assume...

Clearly this could be overwritten by emacsclient command line options.

Re: EmacsW32 invocation options

[Lennart Borgman (gmail)]

Juanma Barranquero wrote:
> On 5/2/07, Lennart Borgman (gmail) <lennart.borgman@gmail.com> wrote:
> 
>> Yes, more discussion would be good. I really tried to open a discussion
>> about it, but I did not get much company.
> 
> I'd say you're being unfair, but that would be far too nice. You seem
> to have suddenly forgotten tens of messages (public and private) and
> quite a few hours devoted to this and related issues.

I am sorry if I gave the impression that you and others did not take 
part at all. But when I finally gave Emacs client a (logical) structure 
that seemed relevant to me I did not get feedback on that. And I really 
tested different possibilities a lot.

> Once again I get
> the feeling that you're very conservative with your time, and quite
> liberal with other people's.

You must be misunderstanding something very much.

> I'm speaking from memory, but I remember at least one bug report that
> explicitly mentioned trouble in the automatic start of the server
> (though, if my memory doesn't fail, you fixed it in your patches).

There were a lot of such reports before people realized that they did 
not had to start the server in .emacs.

Re: EmacsW32 invocation options

[Juanma Barranquero]

On 5/2/07, Lennart Borgman (gmail) <lennart.borgman@gmail.com> wrote:

> I just wanted something that worked now.

Yes. I think that's the main problem. There was no need to rush that.
It is nice for emacsclient to be able to start the server, but it is
not a pressing need. The alternative is a few more clicks every now
and then (not each time emacsclient is run).

> I only reorganized what you wrote and added simple functions
> for user feedback.

"Reorganized" is not how I would describe it, of course ;-)

> I also added a switch
>
>    --server-file FILENAME
>
> for which I wrote "Set filename of the TCP authentication file".

You must be talking about some other thing. IIRC, I added that switch:

2006-10-31  Juanma Barranquero  <lekktu@gmail.com>

        * emacsclient.c: [...]
        (server_file): New global variable.
        (longopts): New option --server-file.
        (decode_options): Process new option --server-file and environment
        variable EMACS_SERVER_FILE.
        (print_help_and_exit): Document new option.

though I just did what Stefan asked for; it was him who designed that
feature (except the environment variable, which was useful to me).

> Maybe I am missing something, but I tried to avoid compatibility
> problems. My guess is that the default action will be to start the local
> server just as it is started in .emacs with (server-start).

I'm not keen on rehearsing the discussion; I think there's not one
single way to start the local server. A way to pass that setup info is
needed.

             Juanma

Re: EmacsW32 invocation options

[Juanma Barranquero]

On 5/2/07, Lennart Borgman (gmail) <lennart.borgman@gmail.com> wrote:

> I am sorry if I gave the impression that you and others did not take
> part at all. But when I finally gave Emacs client a (logical) structure
> that seemed relevant to me I did not get feedback on that.

You *did* get feedback. Some of that feedback said: "it is too late".
Some was "I don't agree with your (logical) structure, I think we
should discuss it a little more". Some of it said "the implementation
is confusing". There was also "it's not that big of an issue, and we
don't want to further delay the issue". Some feedback was silence (and
then Warnock's dilemma apply). Some feedback was positive. You seem to
talk of non-positive feedback as "no feedback".

> And I really
> tested different possibilities a lot.

I *know* you invested a lot of effort. Why should that mean that the
answer reached is optimal, or even good?

> You must be misunderstanding something very much.

Yes, I'm "misunderstanding" your comment: "I really tried to open a
discussion about it, but I did not get much company."

> There were a lot of such reports before people realized that they did
> not had to start the server in .emacs.

Then you agree with me: there were bug reports.

             Juanma

Re: EmacsW32 invocation options

[Lennart Borgman (gmail)]

Juanma Barranquero wrote:

> Yes. I think that's the main problem. There was no need to rush that.
> It is nice for emacsclient to be able to start the server, but it is
> not a pressing need. The alternative is a few more clicks every now
> and then (not each time emacsclient is run).

I was concerned about new users. Every difficulty when trying to start 
using Emacs is a big thing. And if you are on MS Windows it was very 
difficult to get started.

>> I also added a switch
>>
>>    --server-file FILENAME
>>
>> for which I wrote "Set filename of the TCP authentication file".
> 
> You must be talking about some other thing. IIRC, I added that switch:

Eh, yes you are right.

> I'm not keen on rehearsing the discussion; I think there's not one
> single way to start the local server. A way to pass that setup info is
> needed.


I am sure there is more to know about that than I do. However if you are 
talking about user specific elisp code that must be ran I believe it is 
possible to extend the solution in my patched Emacs client without 
difficulties.

Re: EmacsW32 invocation options

[Lennart Borgman (gmail)]

Juanma Barranquero wrote:
> On 5/2/07, Lennart Borgman (gmail) <lennart.borgman@gmail.com> wrote:
> 
>> I am sorry if I gave the impression that you and others did not take
>> part at all. But when I finally gave Emacs client a (logical) structure
>> that seemed relevant to me I did not get feedback on that.
> 
> You *did* get feedback. Some of that feedback said: "it is too late".
> Some was "I don't agree with your (logical) structure, I think we
> should discuss it a little more". Some of it said "the implementation
> is confusing". There was also "it's not that big of an issue, and we
> don't want to further delay the issue". Some feedback was silence (and
> then Warnock's dilemma apply). Some feedback was positive. You seem to
> talk of non-positive feedback as "no feedback".

It is more that the feedback at that time had to be on a more detailed 
level to be meaningful, at least for me.


>> And I really
>> tested different possibilities a lot.
> 
> I *know* you invested a lot of effort. Why should that mean that the
> answer reached is optimal, or even good?


No, of course not. What I meant was that I perhaps could have added 
something to your view if you had tried to be more concrete at that 
time. (I am not sure I remember the details any more.)

I understood it as both a lack of time on your side and that you thought 
a bit different than I did. That is ok, but it is not my responsibility.

Re: EmacsW32 invocation options

[Juanma Barranquero]

On 5/2/07, Lennart Borgman (gmail) <lennart.borgman@gmail.com> wrote:

> I was concerned about new users. Every difficulty when trying to start
> using Emacs is a big thing.

No way. Not being able to print from Windows would be big. Having to
click on the Emacs icon *once* before using the client is hardly big.
IMHO, etc. etc.

> And if you are on MS Windows it was very
> difficult to get started.

I've read you say that more than once, but I'm still unconvinced.

> However if you are
> talking about user specific elisp code that must be ran I believe it is
> possible to extend the solution in my patched Emacs client without
> difficulties.

Almost everything can be extended, often without difficulties. That
does not imply cleanliness of design.

             Juanma

Re: EmacsW32 invocation options

[Juanma Barranquero]

On 5/2/07, Lennart Borgman (gmail) <lennart.borgman@gmail.com> wrote:

> It is more that the feedback at that time had to be on a more detailed
> level to be meaningful, at least for me.

I don't know whether you're joking. I was simplifying. There was
detailed criticism as well as lighter one.

> What I meant was that I perhaps could have added
> something to your view if you had tried to be more concrete at that
> time. (I am not sure I remember the details any more.)

It could have added something to your view if you had tried to accept
that some of us thought there was no time to do a proper design (much
smaller features fell off the wayside because of the "freeze"). You
didn't seem to keen on waiting.

> That is ok, but it is not my responsibility.

No, but you *are* responsible of quite aggressively promoting a
distribution that takes pride in doing some things different from the
standard one (all these things that you affirm that "do not work" or
"are very difficult" in Windows and that you've "fixed" with your
ad-hoc solutions).

             Juanma

Re: EmacsW32 invocation options

[Lennart Borgman (gmail)]

Juanma Barranquero wrote:
> On 5/2/07, Lennart Borgman (gmail) <lennart.borgman@gmail.com> wrote:
> 
>> I was concerned about new users. Every difficulty when trying to start
>> using Emacs is a big thing.
> 
> No way. Not being able to print from Windows would be big. Having to
> click on the Emacs icon *once* before using the client is hardly big.
> IMHO, etc. etc.

That is not the issue. The issue is having to add something to .emacs 
before you even know what it is.

> Almost everything can be extended, often without difficulties. That
> does not imply cleanliness of design.

No, of course not. And views may differ completely about what a clean 
design is.

Re: EmacsW32 invocation options

[Lennart Borgman (gmail)]

Juanma Barranquero wrote:

> No, but you *are* responsible of quite aggressively promoting a
> distribution that takes pride in doing some things different from the
> standard one (all these things that you affirm that "do not work" or
> "are very difficult" in Windows and that you've "fixed" with your
> ad-hoc solutions).


I do not take pride in it. I am just glad it is working.

And some of those things that you call ad-hoc solutions are just that I 
follow the Window's API.

Re: EmacsW32 invocation options

[Brian Elmegaard]

"Juanma Barranquero" <lekktu@gmail.com> writes:

> Because then they have full control over what's running, and how.
> Irrelevant if you only run one instance of Emacs, or just one release,
> or always with the same setup.

I just started emacs on our unix system. I didn't have to start the
server before emacs.

-- 
Brian (remove the sport for mail)
http://www.et.web.mek.dtu.dk/Staff/be/be.html
http://www.rugbyklubben-speed.dk

Re: EmacsW32 invocation options

[Eli Zaretskii]

> Date: Wed, 02 May 2007 09:18:47 +0200
> From: "Lennart Borgman (gmail)" <lennart.borgman@gmail.com>
> CC:  help-gnu-emacs@gnu.org
> 
> I think it is one of the important features in the patched version that 
> makes it quite a bit easier to get started using Emacs for new users. 

This is not the issue I was talking about.  I was talking about the
fact that you are actively promoting a variant of Emacs that behaves
differently from what is described in the Emacs manual.  I think this
is wrong, because it makes harder for newbies to learn Emacs, and
makes it harder for us veterans to help those newbies when they ask
questions.

Re: EmacsW32 invocation options

[Eli Zaretskii]

> From: Brian Elmegaard <brian@rkspeed-rugby.dk>
> Date: Wed, 02 May 2007 07:01:44 +0100
> 
> Eli Zaretskii <eliz@gnu.org> writes:
> 
> > So not only is your Emacs patched and behaves differently, it also
> > teaches users habits that will prevent them from using the same .emacs
> > on other systems?  I find this disturbing.
> 
> Why do users like to have to start emacs server themselves?

I didn't say they like this.  In fact, I didn't talk about server at
all.  I was talking about how to start _Emacs_, not the server, and I
was talking about making it harder for users to share the same .emacs
file between different OSes.

In particular, a .emacs that starts the server will do the wrong thing
with Lennart's binaries.  And .emacs that does not start the server
will give you a server on Windows if you use Lennart's binaries, but
on other systems the server will not start.

Re: EmacsW32 invocation options

[Eli Zaretskii]

> From: Hadron <hadronquark@gmail.com>
> Date: Wed, 02 May 2007 13:54:57 +0200
> 
> The current way it works seems to make little sense. If someone is
> running the client then clearly they want the server. If there is no
> server, start it. Makes perfect sense.

It makes sense only if Emacs behaves like that on all supported
platforms.  Uniform behavior is IMO more important than any random
feature, especially since it's so easy to turn it on in your .emacs.

Re: EmacsW32 invocation options

[Eli Zaretskii]

> Date: Wed, 2 May 2007 17:01:01 +0200
> From: "Juanma Barranquero" <lekktu@gmail.com>
> Cc: "Eli Zaretskii" <eliz@gnu.org>, help-gnu-emacs@gnu.org
> 
> Yes, I'm "misunderstanding" your comment: "I really tried to open a
> discussion about it, but I did not get much company."

Let me translate: "I wanted you to accept my views and change Emacs,
but you didn't."

Re: EmacsW32 invocation options

[Lennart Borgman (gmail)]

Eli Zaretskii wrote:
>> Date: Wed, 2 May 2007 17:01:01 +0200
>> From: "Juanma Barranquero" <lekktu@gmail.com>
>> Cc: "Eli Zaretskii" <eliz@gnu.org>, help-gnu-emacs@gnu.org
>>
>> Yes, I'm "misunderstanding" your comment: "I really tried to open a
>> discussion about it, but I did not get much company."
> 
> Let me translate: "I wanted you to accept my views and change Emacs,
> but you didn't."


I think that is a good translation of your view of the discussion. But 
it is not my view of it.

Re: EmacsW32 invocation options

[Lennart Borgman (gmail)]

Eli Zaretskii wrote:

> In particular, a .emacs that starts the server will do the wrong thing
> with Lennart's binaries.  And .emacs that does not start the server
> will give you a server on Windows if you use Lennart's binaries, but
> on other systems the server will not start.


I do not think that it necessarily does that. The automatic server 
startup does

   (server-mode 1)

and if you do that again in your .emacs I do not think that it creates 
any major problems. The server will be restarted, but that then it seems 
to work without problems.

Are you really sure that it creates problem if there is a (server-mode 
1) in .emacs when using my patched version? I might well be missing 
something, but if I do I will be glad to correct it.

Re: EmacsW32 invocation options

[Juanma Barranquero]

On 5/2/07, Lennart Borgman (gmail) <lennart.borgman@gmail.com> wrote:

> That is not the issue. The issue is having to add something to .emacs
> before you even know what it is.

Not different of a miriad other configuration options, assuming (which
I don't) that you're right when saying "before you even know what it
is". It is not necessary to run emacsclient/server to successfully use
Emacs (on Windows or elsewhere). There's absolutely no trouble using
Emacs without emacsclient and only adding it to your .emacs after
reading the documentation.

> And views may differ completely about what a clean
> design is.

Yes, but not all views are created equal. Programming experience and
knowledge of the target environment are a plus.

             Juanma

Re: EmacsW32 invocation options

[Lennart Borgman (gmail)]

Eli Zaretskii wrote:
>> Date: Wed, 02 May 2007 09:18:47 +0200
>> From: "Lennart Borgman (gmail)" <lennart.borgman@gmail.com>
>> CC:  help-gnu-emacs@gnu.org
>>
>> I think it is one of the important features in the patched version that 
>> makes it quite a bit easier to get started using Emacs for new users. 
> 
> This is not the issue I was talking about.  I was talking about the
> fact that you are actively promoting a variant of Emacs that behaves
> differently from what is described in the Emacs manual.  I think this
> is wrong, because it makes harder for newbies to learn Emacs, and
> makes it harder for us veterans to help those newbies when they ask
> questions.


I agree that it makes it a bit harder for you as a veteran to support 
the newbies. I have however tried to be very careful pointing out the 
(in fact small) differences.

When you say that it makes it harder for newbies to learn Emacs I 
disagree. The differences are not of that kind. I try to get users 
started quickly, that is the most important part of my distro.

I am actively promoting making it easy to get started with Emacs. I am 
not promoting my package as such and would be glad if it was that easy 
to get started with the standard distro.

And I have told that many times, in different words. Still you are 
continuing saying that I am "actively promoting a variant of Emacs". In 
my opinion that is a bit destructive.

Re: EmacsW32 invocation options

[Juanma Barranquero]

On 5/2/07, Lennart Borgman (gmail) <lennart.borgman@gmail.com> wrote:

> I do not take pride in it. I am just glad it is working.

The feeling I get reading your webpage is that you take pride in
detecting (presumed) shortcomings of the standard distribution and
offering (presumed) fixes for them. I dislike it because I don't agree
with many of the supposed shortcomings you point out, and because I
think you're *taking* options from the user (at least, from the novice
user). Pre-packaged releases with lots of goodies stink of patronizing
to me.

> And some of those things that you call ad-hoc solutions are just that I
> follow the Window's API.

That means nothing. There are a gazillion ways to use them.

             Juanma

Re: EmacsW32 invocation options

[Juanma Barranquero]

On 5/2/07, Lennart Borgman (gmail) <lennart.borgman@gmail.com> wrote:

> I think that is a good translation of your view of the discussion.

FWIW, it is also a very accurate translation of my view of the discussion.

             Juanma

Re: EmacsW32 invocation options

[Juanma Barranquero]

On 5/2/07, Lennart Borgman (gmail) <lennart.borgman@gmail.com> wrote:

> I am actively promoting making it easy to get started with Emacs.

You're actively promoting the view that getting started with Emacs on
Windows is difficult without your changes.

> Still you are
> continuing saying that I am "actively promoting a variant of Emacs".

It's you who says "my distro", isn't?

> In my opinion that is a bit destructive.

As I explained once before, your "distro" is potentially destructive
too. It's not hard to imagine how it could lead to a fork of Emacs on
Windows (not that it's *going* to happen, but it paves the way for it
to happen).

             Juanma

Re: EmacsW32 invocation options

[Lennart Borgman (gmail)]

Juanma Barranquero wrote:
> On 5/2/07, Lennart Borgman (gmail) <lennart.borgman@gmail.com> wrote:
> 
>> I do not take pride in it. I am just glad it is working.
> 
> The feeling I get reading your webpage is that you take pride in
> detecting (presumed) shortcomings of the standard distribution and
> offering (presumed) fixes for them. I dislike it because I don't agree
> with many of the supposed shortcomings you point out, and because I
> think you're *taking* options from the user (at least, from the novice
> user). Pre-packaged releases with lots of goodies stink of patronizing
> to me.

What are you actually trying to do, Juanma? Why did you write this?


>> And some of those things that you call ad-hoc solutions are just that I
>> follow the Window's API.
> 
> That means nothing. There are a gazillion ways to use them.

Now you have just left logical reasoning. I see no reason to respond to 
that.

Re: EmacsW32 invocation options

[Lennart Borgman (gmail)]

Juanma Barranquero wrote:

> Yes, but not all views are created equal. Programming experience and
> knowledge of the target environment are a plus.


And so is sound reasoning without personal attachs.

Re: EmacsW32 invocation options

[Juanma Barranquero]

On 5/2/07, Lennart Borgman (gmail) <lennart.borgman@gmail.com> wrote:

> What are you actually trying to do, Juanma? Why did you write this?

What do you think I'm trying to do? Why do you think I wrote it?

> Now you have just left logical reasoning.

Yes, of course pointing out that it is possible to err even using
Windows API calls is "leaving logical reasoning".

> I see no reason to respond to that.

Your call.

             Juanma

Re: EmacsW32 invocation options

[Hadron]

"Juanma Barranquero" <lekktu@gmail.com> writes:

> On 5/2/07, Lennart Borgman (gmail) <lennart.borgman@gmail.com> wrote:
>
>> I am actively promoting making it easy to get started with Emacs.
>
> You're actively promoting the view that getting started with Emacs on
> Windows is difficult without your changes.

A brief browse of forums and the emacs irc channel suggests that most/a
good % of people do indeed find it daunting to get emacs working under
windows. I am surprised you would disagree. What is simple to you is not
necessarily simple for complete newbies - most of whom I would guess run
Windows.

Having said that I must agree with you that patching over many of the
"must learns" at too early a stage can also leave a nOOb stranded even
further out to sea at a later date. By far the most important thing to
instill in any  emacs apprentice is to learn how to use the built in
help.

C-h i
C-h m
C-h a
C-h k

>From there the world is their Oyster. Oh and add the arrow keys into the
tutorial :-;

>
>> Still you are
>> continuing saying that I am "actively promoting a variant of Emacs".
>
> It's you who says "my distro", isn't?
>
>> In my opinion that is a bit destructive.
>
> As I explained once before, your "distro" is potentially destructive
> too. It's not hard to imagine how it could lead to a fork of Emacs on
> Windows (not that it's *going* to happen, but it paves the way for it
> to happen).
>
>             Juanma
>
>

-- 

Re: EmacsW32 invocation options

[Juanma Barranquero]

On 5/2/07, Lennart Borgman (gmail) <lennart.borgman@gmail.com> wrote:
> Juanma Barranquero wrote:

> And so is sound reasoning without personal attachs.

Rest assured I haven't attacked you.

You think that explaining the feeling I get from your webpage is an
attack. It's not. I cannot read your mind, but I have no trouble
whatsoever reading mine.

I think that saying that the Emacs developers did not take the time to
discuss rationally or in deep what you wanted to do, or asking again
and again to people who disagrees with you whether they have read your
webpages, or messages in a discussion thread, instead of assuming that
they *have* read it and just *do* disagree... that seems like a
personal attack to me, yes.

             Juanma

Re: EmacsW32 invocation options

[Hadron]

"Juanma Barranquero" <lekktu@gmail.com> writes:

> On 5/2/07, Lennart Borgman (gmail) <lennart.borgman@gmail.com> wrote:
>
>> What are you actually trying to do, Juanma? Why did you write this?
>
> What do you think I'm trying to do? Why do you think I wrote it?
>
>> Now you have just left logical reasoning.
>
> Yes, of course pointing out that it is possible to err even using
> Windows API calls is "leaving logical reasoning".

It is possible to "err" tying up your shoelaces. Or have you discovered
some magical "perfection" dust?
>
>> I see no reason to respond to that.
>
> Your call.

You appear to be unnecessarily hostile and protective of your toys.

>             Juanma

Re: EmacsW32 invocation options

[Lennart Borgman (gmail)]

Juanma Barranquero wrote:
> On 5/2/07, Lennart Borgman (gmail) <lennart.borgman@gmail.com> wrote:
>> Juanma Barranquero wrote:
> 
>> And so is sound reasoning without personal attachs.
> 
> Rest assured I haven't attacked you.
> 
> You think that explaining the feeling I get from your webpage is an
> attack. It's not. I cannot read your mind, but I have no trouble
> whatsoever reading mine.

Reading ones own mind is also about understanding where the feelings 
come from. And about acting/reacting/contemplating.

Maybe you think I am saying to much above but I really have found some 
of your comments strange.


> I think that saying that the Emacs developers did not take the time to
> discuss rationally or in deep what you wanted to do, or asking again
> and again to people who disagrees with you whether they have read your
> webpages, or messages in a discussion thread, instead of assuming that
> they *have* read it and just *do* disagree... that seems like a
> personal attack to me, yes.

I am not the one who brings up this discussion. I would prefer to be 
silent if I did not find the comments destructive.

Re: EmacsW32 invocation options

[Juanma Barranquero]

On 5/2/07, Hadron <hadronquark@gmail.com> wrote:

> I don't see that as true at all. If you run emacsclient you obviously do
> it with the intention of connecting to emacs-server.

Yes. But there's no single "right" way to run the Emacs server.

There has to be a way for emacsclient to detect that the Emacs server
is not running, start it, pass it the info to start the server (if you
want a non-default server setup), wait till Emacs *and* the server are
operative, and then pass it the original arguments.

That's a lot of details, both of implementation and design. Lennart
and I disagree about the right way to do both. But design worries me
more; that and divergence from the Emacs interface in other OSes.

             Juanma

Re: EmacsW32 invocation options

[Juanma Barranquero]

On 5/2/07, Hadron <hadronquark@gmail.com> wrote:

> The configured one I would assume...

Which one is "the configured one"?

> Clearly this could be overwritten by emacsclient command line options.

Yes. That's part of the problem.

             Juanma

Re: EmacsW32 invocation options

[Juanma Barranquero]

On 5/2/07, Brian Elmegaard <brian@rkspeed-rugby.dk> wrote:

> I just started emacs on our unix system. I didn't have to start the
> server before emacs.

Perhaps you have ALTERNATE_EDITOR set.

             Juanma

Re: EmacsW32 invocation options

[Juanma Barranquero]

On 5/2/07, Hadron <hadronquark@gmail.com> wrote:

> A brief browse of forums and the emacs irc channel suggests that most/a
> good % of people do indeed find it daunting to get emacs working under
> windows.

Do you mean, more people find it daunting to get Emacs working on
Windows than on GNU/Linux?

> I am surprised you would disagree. What is simple to you is not
> necessarily simple for complete newbies - most of whom I would guess run
> Windows.

The kind of things that make Emacs difficult are not going to
disappear by making emacsclient start Emacs. A complete newbie must be
ready to delve into a (very good, but extensive) documentation, and
perhaps even do some elisp programming, to take full advantage of
Emacs.

> By far the most important thing to
> instill in any  emacs apprentice is to learn how to use the built in
> help.

Exactly. I did three attempts over the years before I finally warmed
to Emacs. The definitive push was taking the time to follow the
tutorial and read at least some of the docs.

             Juanma

Re: EmacsW32 invocation options

[Juanma Barranquero]

On 5/2/07, Hadron <hadronquark@gmail.com> wrote:

> It is possible to "err" tying up your shoelaces. Or have you discovered
> some magical "perfection" dust?

I think you meant that for Lennart, didn't you? He's the one talking
about "leaving logical reasoning".

> You appear to be unnecessarily hostile and protective of your toys.

On one hand, I'm not sure what do you think are "my toys". Emacs
certainly isn't. On the other hand, "unnecesarily" is quite difficult
to define in an objective way, let's not even mention "hostile". Allow
me to disagree on both accounts. On the third hand, I'm prompted to
quote Erik Naggum: "People who seek agreement do not seek
understanding."

             Juanma

Re: EmacsW32 invocation options

[Hadron]

"Juanma Barranquero" <lekktu@gmail.com> writes:

> On 5/2/07, Hadron <hadronquark@gmail.com> wrote:
>
>> A brief browse of forums and the emacs irc channel suggests that most/a
>> good % of people do indeed find it daunting to get emacs working under
>> windows.
>
> Do you mean, more people find it daunting to get Emacs working on
> Windows than on GNU/Linux?

I remember years ago I certainly did. But having abandoned Windows now I
can only report from my perception/experience.

>
>> I am surprised you would disagree. What is simple to you is not
>> necessarily simple for complete newbies - most of whom I would guess run
>> Windows.
>
> The kind of things that make Emacs difficult are not going to
> disappear by making emacsclient start Emacs. A complete newbie must be
> ready to delve into a (very good, but extensive) documentation, and
> perhaps even do some elisp programming, to take full advantage of
> Emacs.

There are many things in emacs which can and should be simplified. The
server thing is one such. I see no reason whatsoever that emacsclient
cant do exactly the same as "server-start" - certainly on a single
machine. or maybe default to using the system clipboard for cut &
paste. Things like that turn new users away. Saying "its because thats
the way it is" is inflexible and non productive.

ps if anyone with any shout reads this - is there any change
update-alternatives could change emacsclient when it changes emacs? I
constantly fall foul of that.

>
>> By far the most important thing to
>> instill in any  emacs apprentice is to learn how to use the built in
>> help.
>
> Exactly. I did three attempts over the years before I finally warmed
> to Emacs. The definitive push was taking the time to follow the
> tutorial and read at least some of the docs.

I can't disagree with that. Its very easy to cry for help but its there
under your nose all the time. I have made the mistake more than once in
emacs and probably will do again.


>
>             Juanma
>
>

-- 

Re: EmacsW32 invocation options

[Eli Zaretskii]

> Date: Wed, 02 May 2007 21:49:22 +0200
> From: "Lennart Borgman (gmail)" <lennart.borgman@gmail.com>
> CC:  help-gnu-emacs@gnu.org
> 
> Eli Zaretskii wrote:
> 
> > In particular, a .emacs that starts the server will do the wrong thing
> > with Lennart's binaries.  And .emacs that does not start the server
> > will give you a server on Windows if you use Lennart's binaries, but
> > on other systems the server will not start.
> 
> 
> I do not think that it necessarily does that. The automatic server 
> startup does
> 
>    (server-mode 1)

That is not what the Emacs manual says one should do.  Here's the
relevant excerpt:

       First, the preparations.  Within Emacs, call the function
    `server-start'.  (Your `.emacs' init file can do this automatically if
    you add the expression `(server-start)' to it, see *Note Init File::.)

So it says to add `(server-start)', not `(server-start 1)'.

Re: EmacsW32 invocation options

[Eli Zaretskii]

> Date: Wed, 02 May 2007 22:01:07 +0200
> From: "Lennart Borgman (gmail)" <lennart.borgman@gmail.com>
> CC:  help-gnu-emacs@gnu.org
> 
> When you say that it makes it harder for newbies to learn Emacs I 
> disagree.

It makes it harder because they use Emacs that behaves not as
documented in subtle ways.

> And I have told that many times, in different words. Still you are 
> continuing saying that I am "actively promoting a variant of Emacs". In 
> my opinion that is a bit destructive.

My experience of talking to you indicates that anything that you
disagree with is deemed ``destructive'' by you.

Re: EmacsW32 invocation options

[Brian Elmegaard]

Eli Zaretskii <eliz@gnu.org> writes:

> with Lennart's binaries.  And .emacs that does not start the server
> will give you a server on Windows if you use Lennart's binaries, but
> on other systems the server will not start.

I see the point, and if the only problem is that the solution was not
found before releasing 22 I understand the issue and hope it will be
fixed. However, users of emacsW32 and particularly those who choose
the patched version are informed the whole way through. 

Many windows users will not even consider running another
system. Those experts who need to run on several systems may be able
to do something clever with system-type in their .emacs wouldn't they?

-- 
Brian (remove the sport for mail)
http://www.et.web.mek.dtu.dk/Staff/be/be.html
http://www.rugbyklubben-speed.dk

Re: EmacsW32 invocation options

[Brian Elmegaard]

Eli Zaretskii <eliz@gnu.org> writes:

> It makes it harder because they use Emacs that behaves not as
> documented in subtle ways.

>From a short google on groups it seems that the emacs developers find
emacsW32 ("EmacsW32 is not Emacs for MS Windows. Instead it is an
add-on to Emacs for MS Windows.") worse than aquamacs ("Aquamacs
features extensive customization: it will feel and behave mostly like
an Aqua program"). 

Why is that?

-- 
Brian (remove the sport for mail)
http://www.et.web.mek.dtu.dk/Staff/be/be.html
http://www.rugbyklubben-speed.dk

Re: EmacsW32 invocation options

[Brian Elmegaard]

"Juanma Barranquero" <lekktu@gmail.com> writes:

> Perhaps you have ALTERNATE_EDITOR set.

I don't know. I just use a shortcut in the menu that pops up when I
click the lower left corner of the screen. 

I don't think that is bad for me.

-- 
Brian (remove the sport for mail)
http://www.et.web.mek.dtu.dk/Staff/be/be.html
http://www.rugbyklubben-speed.dk

Re: EmacsW32 invocation options

[Brian Elmegaard]

Eli Zaretskii <eliz@gnu.org> writes:

>> From: Brian Elmegaard <brian@rkspeed-rugby.dk>
>> Date: Wed, 02 May 2007 07:01:44 +0100
>> 
>> Eli Zaretskii <eliz@gnu.org> writes:
>> 
>> > So not only is your Emacs patched and behaves differently, it also
>> > teaches users habits that will prevent them from using the same .emacs
>> > on other systems?  I find this disturbing.
>> 
>> Why do users like to have to start emacs server themselves?
>
> I didn't say they like this.  

I see.

So I took a look at the manual. The manual for some reason is focused
on unix:    
"As an alternative to using `emacsclient', the file `etc/emacs.bash'
defines a Bash command `edit' which will communicate with a running
Emacs session, or start one if none exist."

Afaics this is similar to what Lennart does for w32, but I may be
wrong. 
To me this indicates that there might be a reason for making something
easier for windows people, too. 

In my experience emacsW32 is a very nice customization for my use of
emacs and it is easier for me to make our students get going with
emacs. 

-- 
Brian (remove the sport for mail)
http://www.et.web.mek.dtu.dk/Staff/be/be.html
http://www.rugbyklubben-speed.dk

Re: EmacsW32 invocation options

[Juanma Barranquero]

On 5/3/07, Brian Elmegaard <brian@rkspeed-rugby.dk> wrote:

>From a short google on groups it seems that the emacs developers find
> emacsW32 ("EmacsW32 is not Emacs for MS Windows. Instead it is an
> add-on to Emacs for MS Windows.") worse than aquamacs ("Aquamacs
> features extensive customization: it will feel and behave mostly like
> an Aqua program").

I know nothing about Aquamacs, so I cannot say. The point, anyway, is
not whether the program has features which help it to blend with its
environment. I'm arguing that Emacs on Windows is not harder to use or
learn that Emacs on GNU/Linux or other environments, so changes to
"help the user" seem like gratuitous divergences to me unless there's
a *specific* need (window manager integration, keyboard handling,
etc.).

             Juanma

Re: EmacsW32 invocation options

[Juanma Barranquero]

On 5/3/07, Brian Elmegaard <brian@rkspeed-rugby.dk> wrote:

> I don't know. I just use a shortcut in the menu that pops up when I
> click the lower left corner of the screen.
>
> I don't think that is bad for me.

No. My point was that you can do that already with the stock
emacsclient on Windows (using --alternate-editor). The end result is
different of what we're discussing here (because Emacs started as
alternate editor does not receive its arguments via the
emacsclient/server connection).

             Juanma

Re: EmacsW32 invocation options

[Lennart Borgman (gmail)]

Eli Zaretskii wrote:
>> Date: Wed, 02 May 2007 21:49:22 +0200
>> From: "Lennart Borgman (gmail)" <lennart.borgman@gmail.com>
>> CC:  help-gnu-emacs@gnu.org
>>
>> Eli Zaretskii wrote:
>>
>>> In particular, a .emacs that starts the server will do the wrong thing
>>> with Lennart's binaries.  And .emacs that does not start the server
>>> will give you a server on Windows if you use Lennart's binaries, but
>>> on other systems the server will not start.
>>
>> I do not think that it necessarily does that. The automatic server 
>> startup does
>>
>>    (server-mode 1)
> 
> That is not what the Emacs manual says one should do.  Here's the
> relevant excerpt:
> 
>        First, the preparations.  Within Emacs, call the function
>     `server-start'.  (Your `.emacs' init file can do this automatically if
>     you add the expression `(server-start)' to it, see *Note Init File::.)
> 
> So it says to add `(server-start)', not `(server-start 1)'.

It looks to me like server-mode was added afterwards to make things more 
convenient. Maybe the manaul should be updated to reflect that?

Re: EmacsW32 invocation options

[Kim F. Storm]

"Juanma Barranquero" <lekktu@gmail.com> writes:

> As I explained once before, your "distro" is potentially destructive
> too. It's not hard to imagine how it could lead to a fork of Emacs on
> Windows (not that it's *going* to happen, but it paves the way for it
> to happen).

That's the "risk" of free software ... somebody taking what you've written,
modify it (as the license permits), and release it as a modified version,
offering you (and everybody else) the patches to make your version "comply".

If you don't like those patches, well, that's your freedom.

I don't see how someone using the freedom of GPL should be accused of
being destructive, just because he does something you don't approve of.

I can fully understand why Lennart would prefer his changes to be adopted
into the core (it's no fun to maintain your "own distro"), but if you don't
want to do that (I haven't considered the actual changes), then IMO it is
fine -- and fully legal -- for him to make his own distro.

BTW, IMO, Lennart is doing a fine job of helping getting the 22.1
release out by providing (unpatched) W32 binaries for all the pretests.

-- 
Kim F. Storm  http://www.cua.dk

Re: EmacsW32 invocation options

[Hadron]

Brian Elmegaard <brian@rkspeed-rugby.dk> writes:

> Eli Zaretskii <eliz@gnu.org> writes:
>
>>> From: Brian Elmegaard <brian@rkspeed-rugby.dk>
>>> Date: Wed, 02 May 2007 07:01:44 +0100
>>> 
>>> Eli Zaretskii <eliz@gnu.org> writes:
>>> 
>>> > So not only is your Emacs patched and behaves differently, it also
>>> > teaches users habits that will prevent them from using the same .emacs
>>> > on other systems?  I find this disturbing.
>>> 
>>> Why do users like to have to start emacs server themselves?
>>
>> I didn't say they like this.  
>
> I see.
>
> So I took a look at the manual. The manual for some reason is focused
> on unix:    
> "As an alternative to using `emacsclient', the file `etc/emacs.bash'
> defines a Bash command `edit' which will communicate with a running
> Emacs session, or start one if none exist."

As a relative convert back to emacs (I used back in the early 90s) I
agree with you. I remember back then when I moved to PCs I found it
incredibly frustrating to get emacs working under window. Whether its
because of missing paths, lack of bash, lack of documentation I don't
know/cant remember. I think any move to facilitate emacs use under
windows for nOObs can only be a good thing so long as that the basic
platform independent manual is not made redundant. Lennart's "additions"
should probably be totally done in a single line autoload in the basic
.emacs so that it can be simply unplugged.  Maybe it is.

>
> Afaics this is similar to what Lennart does for w32, but I may be
> wrong. 
> To me this indicates that there might be a reason for making something
> easier for windows people, too. 
>
> In my experience emacsW32 is a very nice customization for my use of
> emacs and it is easier for me to make our students get going with
> emacs. 

In SW there are always those who prefer not to do rather than do because
then there is no risk of the apple cart being upset. This attitude can
pay dividends in the long run. "Half assed" solutions to "partial
problems" often come and bite one in the ass at a later date. I suspect
people are worried that W32Emacs might be just such a venture. There is
already too much confusion in the emacs world - try looking for a
php-mode a cscope interface - it's a version nightmare with everyone
maintaining their own versions. One of the few the drawbacks of open source
IMO.

As far as the issue which sparked it, my partial ignorance on the
internal workings of emacs might well bar me from commenting. However,
casting a dis partial eye of the issue it seems fairly obvious to me for
the "single machine" issue. If (server-start) is able to start a server,
then I see no reason why emacsclient cant glean the same "parameters"
and start it itself if it is not running. I assume emacs server can also
run on a remote machine?  If so then the start parameters would, I
guess, be different. I don't know.

I think the bottom line is that any improvements should be cautiously
welcomed.

Certainly Lennart's xhtml stuff is way and above better than the
incomplete and buggy bits and pieces that were lying around in various
incarnations for editing mixed mode php/css/html/xhtml files. It got to
a stage where I wondered if anyone in the whole world was actually using
emacs to edit such files. Nearly ever readme I found indicated a
different "branch" of php-mode or html-helper and then went on to say
things like "may not work with x", "not sure if it will work for
you". So, I for one am grateful for Lennart's work on this issue
anyway. Is it perfect? No way. Is it better than went before? Yes. Is
Lennart open to suggestions for improvement from the people who actually
use it? Yes.

Re: EmacsW32 invocation options

[Juanma Barranquero]

On 5/3/07, Kim F. Storm <no-spam@cua.dk> wrote:

> That's the "risk" of free software ... somebody taking what you've written,
> modify it (as the license permits), and release it as a modified version,
> offering you (and everybody else) the patches to make your version "comply".

Let me see... Hmm... I think I already knew that :)

> I don't see how someone using the freedom of GPL should be accused of
> being destructive, just because he does something you don't approve of.

I think it is destructive. I don't think it is destructive "just
because" I don't approve of it; I think it is destructive because
developer resources are scarce. I'm sure the XEmacs folk are a bunch
of very nice guys, but I see the GNU Emacs/XEmacs split as destructive
all the same.

And BTW, perhaps you skipped the following bit from Lennart to Eli, or
you do not consider it relevant:

> And I have told that many times, in different words. Still you
> are continuing saying that I am "actively promoting a variant
> of Emacs". In my opinion that is a bit destructive.

Are you going to call him to task, too, for describing Eli's opinion
as destructive...? :)

> I can fully understand why Lennart would prefer his changes to be adopted
> into the core (it's no fun to maintain your "own distro")

Of course. I, personally, have committed changes *by* Lennart *to* emacsclient:

2006-11-22  Lennart Borgman  <lennart.borgman.073@student.lu.se>

        * emacsclient.c: Include <stdarg.h>.
        [WINDOWSNT]: Include <windows.h>.
        (w32_check_console_app): New function.
        (message): New function.
        (decode_options, print_help_and_exit, fail, main)
        (initialize_sockets, get_server_config, set_tcp_socket)
        (set_local_socket, set_socket): Use message().

so it's not like I'm opposed to either Lennart or Lennart's ideas. I'm
just disagreeing on some of them (and his timing).

> then IMO it is
> fine -- and fully legal -- for him to make his own distro.

It is of course fully legal (at least once he made the source tarballs
for some utilities more readily accessible). "Fine", well, that's your
opinion. Mine is different.

> BTW, IMO, Lennart is doing a fine job of helping getting the 22.1
> release out by providing (unpatched) W32 binaries for all the pretests.

Yes. I don't remember ever saying otherwise. "Unpatched" being the
operative word in your comment above.

             Juanma

Re: EmacsW32 invocation options

[Juanma Barranquero]

On 5/3/07, Hadron <hadronquark@gmail.com> wrote:

> In SW there are always those who prefer not to do rather than do because
> then there is no risk of the apple cart being upset. This attitude can
> pay dividends in the long run. "Half assed" solutions to "partial
> problems" often come and bite one in the ass at a later date. I suspect
> people are worried that W32Emacs might be just such a venture.

And in this case, is not that some people (me included) preferred "not
to do". More like "not now, because it could benefit from some
additional thought".

> So, I for one am grateful for Lennart's work on this issue
> anyway. Is it perfect? No way. Is it better than went before?
> Yes.

Of course.

> Is
> Lennart open to suggestions for improvement from the people who actually
> use it?

That's a good question.

             Juanma

Re: EmacsW32 invocation options

[Hadron]

"Juanma Barranquero" <lekktu@gmail.com> writes:

> On 5/3/07, Hadron <hadronquark@gmail.com> wrote:
>
>> In SW there are always those who prefer not to do rather than do because
>> then there is no risk of the apple cart being upset. This attitude can
>> pay dividends in the long run. "Half assed" solutions to "partial
>> problems" often come and bite one in the ass at a later date. I suspect
>> people are worried that W32Emacs might be just such a venture.
>
> And in this case, is not that some people (me included) preferred "not
> to do". More like "not now, because it could benefit from some
> additional thought".

I have often found in the development world that sometimes "additional
thought" is a euphemism for "not today, not tomorrow, maybe in a few
years when I have nothing better to do". There are times when it can
benefit a development to let the young guns go blazing and see what
happens. Sometimes :-;

>
>> So, I for one am grateful for Lennart's work on this issue
>> anyway. Is it perfect? No way. Is it better than went before?
>> Yes.
>
> Of course.
>
>> Is
>> Lennart open to suggestions for improvement from the people who actually
>> use it?
>
> That's a good question.
>
>             Juanma

And one fairly easy to answer - yes. But note the important bit - "from
those who actually use it".

Re: EmacsW32 invocation options

[Juanma Barranquero]

On 5/3/07, Hadron <hadronquark@gmail.com> wrote:

> I have often found in the development world that sometimes "additional
> thought" is a euphemism for "not today, not tomorrow, maybe in a few
> years when I have nothing better to do". There are times when it can
> benefit a development to let the young guns go blazing and see what
> happens. Sometimes :-;

Yeah, well, I don't think that was the case right now.

> And one fairly easy to answer - yes.

I was suggesting that there are other good questions.

> But note the important bit - "from
> those who actually use it".

Irrelevant, I think. Everybody involved in the emacsclient discussion
were also emacsclient users.

             Juanma

Re: EmacsW32 invocation options

[David Kastrup]

"Juanma Barranquero" <lekktu@gmail.com> writes:

> On 5/3/07, Kim F. Storm <no-spam@cua.dk> wrote:
>
>> That's the "risk" of free software ... somebody taking what you've written,
>> modify it (as the license permits), and release it as a modified version,
>> offering you (and everybody else) the patches to make your version "comply".
>
> Let me see... Hmm... I think I already knew that :)
>
>> I don't see how someone using the freedom of GPL should be accused of
>> being destructive, just because he does something you don't approve of.
>
> I think it is destructive. I don't think it is destructive "just
> because" I don't approve of it; I think it is destructive because
> developer resources are scarce. I'm sure the XEmacs folk are a bunch
> of very nice guys, but I see the GNU Emacs/XEmacs split as
> destructive all the same.

It is divisive.  I don't see it as destructive since it does not make
anything cease to be that would otherwise continue to exist.

People speaking different languages is not generally considered
"destructive" either, even though many might derive from a common
ancestor.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: EmacsW32 invocation options

[Brian Elmegaard]

"Juanma Barranquero" <lekktu@gmail.com> writes:

> environment. I'm arguing that Emacs on Windows is not harder to use or
> learn that Emacs on GNU/Linux or other environments, so changes to

hat may be true, but my own experience and my impression of watching
students required to learn emacs, is that emacs+EmacsW32 is easier to
learn emacs.

-- 
Brian (remove the sport for mail)
http://www.et.web.mek.dtu.dk/Staff/be/be.html
http://www.rugbyklubben-speed.dk

Re: EmacsW32 invocation options

[Juanma Barranquero]

On 5/3/07, David Kastrup <dak@gnu.org> wrote:

> I don't see it as destructive since it does not make
> anything cease to be that would otherwise continue to exist.

No, but perhaps it makes something not to be that could've been, were
the resources more abundant.

> People speaking different languages is not generally considered
> "destructive" either, even though many might derive from a common
> ancestor.

Yeah, I'm sure the native speakers of latin do fully agree with you ;-)

             Juanma

Re: EmacsW32 invocation options

[David Kastrup]

"Juanma Barranquero" <lekktu@gmail.com> writes:

> On 5/3/07, David Kastrup <dak@gnu.org> wrote:
>
>> I don't see it as destructive since it does not make
>> anything cease to be that would otherwise continue to exist.
>
> No, but perhaps it makes something not to be that could've been, were
> the resources more abundant.
>
>> People speaking different languages is not generally considered
>> "destructive" either, even though many might derive from a common
>> ancestor.
>
> Yeah, I'm sure the native speakers of latin do fully agree with you ;-)

Like the active users of Emacs 18 will fully agree with you about your
sentiments concerning the Emacs/XEmacs fork.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: EmacsW32 invocation options

[Juanma Barranquero]

On 5/4/07, David Kastrup <dak@gnu.org> wrote:

> Like the active users of Emacs 18 will fully agree with you about your
> sentiments concerning the Emacs/XEmacs fork.

That's a bit of a cheat. I'm sure there are active users of Emacs who
were once users of Emacs 18. The Emacs/XEmacs fork is not some barely
half-remembered thing of the past.

             Juanma

Re: EmacsW32 invocation options

[Eli Zaretskii]

> From: Brian Elmegaard <brian@rkspeed-rugby.dk>
> Date: Thu, 03 May 2007 07:08:12 +0100
> 
> Many windows users will not even consider running another system.

True, but many others do (or have to).  And they are not necessarily
experts.

For example, on my daytime job, there's a rumor going around that I
have something to do with Emacs.  You wouldn't believe how many such
users come to me for help, and they specifically want to have similar
environments on both platforms.  This is because we have many projects
that target multiple platforms, including Windows and GNU/Linux.

> Those experts who need to run on several systems may be able
> to do something clever with system-type in their .emacs wouldn't they?

Sure, they can.  But system-type is never mentioned in the Emacs user
manual, so it will take them quite some time to find out about it.
And it's a nuisance anyway.

Re: EmacsW32 invocation options

[Eli Zaretskii]

> From: Brian Elmegaard <brian@rkspeed-rugby.dk>
> Date: Thu, 03 May 2007 07:24:02 +0100
> 
> Eli Zaretskii <eliz@gnu.org> writes:
> 
> > It makes it harder because they use Emacs that behaves not as
> > documented in subtle ways.
> 
> >From a short google on groups it seems that the emacs developers find
> emacsW32 ("EmacsW32 is not Emacs for MS Windows. Instead it is an
> add-on to Emacs for MS Windows.") worse than aquamacs ("Aquamacs
> features extensive customization: it will feel and behave mostly like
> an Aqua program"). 
> 
> Why is that?

Maybe because most Emacs developers don't know anything about
Aquamacs.  That includes myself.

Re: EmacsW32 invocation options

[Eli Zaretskii]

> Date: Thu, 03 May 2007 10:26:09 +0200
> From: "Lennart Borgman (gmail)" <lennart.borgman@gmail.com>
> CC:  help-gnu-emacs@gnu.org
> 
> >        First, the preparations.  Within Emacs, call the function
> >     `server-start'.  (Your `.emacs' init file can do this automatically if
> >     you add the expression `(server-start)' to it, see *Note Init File::.)
> > 
> > So it says to add `(server-start)', not `(server-start 1)'.
> 
> It looks to me like server-mode was added afterwards to make things more 
> convenient. Maybe the manaul should be updated to reflect that?

Maybe it should, but _right_now_ users who adhere to the manual will
be surprised by the deviant behavior on Windows if they move to
another platform.

Re: EmacsW32 invocation options

[Eli Zaretskii]

> From: no-spam@cua.dk (Kim F. Storm)
> Date: Thu, 03 May 2007 11:43:25 +0200
> Cc: help-gnu-emacs@gnu.org
> 
> "Juanma Barranquero" <lekktu@gmail.com> writes:
> 
> > As I explained once before, your "distro" is potentially destructive
> > too. It's not hard to imagine how it could lead to a fork of Emacs on
> > Windows (not that it's *going* to happen, but it paves the way for it
> > to happen).
> 
> That's the "risk" of free software ... somebody taking what you've written,
> modify it (as the license permits), and release it as a modified version,
> offering you (and everybody else) the patches to make your version "comply".
> 
> If you don't like those patches, well, that's your freedom.
> 
> I don't see how someone using the freedom of GPL should be accused of
> being destructive, just because he does something you don't approve of.

Kim, I hope you've read this thread before responding, because you've
taken one phrase out of context and jumped the gun on it.

The issue is not freedom, and Juanma hardly needs a lecture about it.
(He might be better off choosing a less radical word than
``destructive'', but then he's not a native English speaker, so we all
might cut him some slack.)  The issue is how to use this freedom in
the best interests of the community at large.

In this case, Lennart's attitude to put into his distribution every
patch he finds useful, but cannot convince the Emacs developers to
accept, is IMO not the best use of that freedom.  It is confusing to
users who expect Emacs to behave as documented and to do that
consistently on all platforms.  And it's an attitude that smells of
schism and forking, which many of us will not approve, because it
diminishes the already small group of active developers by spreading
their insufficient resources between several competing projects.

> I can fully understand why Lennart would prefer his changes to be adopted
> into the core (it's no fun to maintain your "own distro"), but if you don't
> want to do that (I haven't considered the actual changes), then IMO it is
> fine -- and fully legal -- for him to make his own distro.

Of course, it's legal!  No one suggested that Lennart should be
arrested for illegal trafficking in Emacs Lisp or something.

> BTW, IMO, Lennart is doing a fine job of helping getting the 22.1
> release out by providing (unpatched) W32 binaries for all the pretests.

Sure, he does, and he deserves a lot of credit and thanks for that.
As does Juanma, btw, who made emacslient work on Windows to begin
with.

Re: EmacsW32 invocation options

[Eli Zaretskii]

> From: Hadron <hadronquark@gmail.com>
> Date: Thu, 03 May 2007 13:36:59 +0200
> 
> I have often found in the development world that sometimes "additional
> thought" is a euphemism for "not today, not tomorrow, maybe in a few
> years when I have nothing better to do".

It is the task of whoever suggests a change to convince the other
developers of its virtue.  Failure to convince them should give you a
more objective estimate of its _real_ virtue, because the other
developers normally provide reasoning for why they dislike the
suggestion.  If you (or anyone else who is reading this) want to learn
about those reasons, I invite you to read the relevant discussions on
emacs-devel and emacs-pretest-bug.

As things frequently are with changes suggested by Lennart that other
developers don't like, Lennart will continue arguing and will not
accept differing opinions or modify his suggestions according to
comments he receives, no matter what.  When such endless and futile
arguments go out of control, it is quite possible that someone would
say "additional thought" meaning "not now, and I don't care when".
But you can hardly accuse people who don't want their scarce free time
devoted to work on Emacs be monopolized by a long heated discussion
leading nowhere, can you?  When cornered, people would do anything to
cut their losses.

> There are times when it can benefit a development to let the young
> guns go blazing and see what happens. Sometimes :-;

Sometimes, yes.  When development stagnates, for example, or if major
usability features are rejected by a dictatorship-like clique for no
good reason.  If you find evidence of these in the discussions related
to the change suggested by Lennart that is the issue of this thread,
please do point that out.

> >> Is
> >> Lennart open to suggestions for improvement from the people who actually
> >> use it?
> >
> > That's a good question.
> >
> >             Juanma
> 
> And one fairly easy to answer - yes. But note the important bit - "from
> those who actually use it".

I think a more accurate answer is ``yes, if he agrees with the
suggestion''.

And I don't see why one needs to actually use an Emacs where server is
started by default to be eligible for an opinion on that.  It's not
that understanding the implications of this requires actual
experience.

Re: EmacsW32 invocation options

[Eli Zaretskii]

> From: Brian Elmegaard <brian@rkspeed-rugby.dk>
> Date: Thu, 03 May 2007 23:13:14 +0100
> 
> "Juanma Barranquero" <lekktu@gmail.com> writes:
> 
> > environment. I'm arguing that Emacs on Windows is not harder to use or
> > learn that Emacs on GNU/Linux or other environments, so changes to
> 
> hat may be true, but my own experience and my impression of watching
> students required to learn emacs, is that emacs+EmacsW32 is easier to
> learn emacs.

We are not talking in general here, we are talking specifically about
starting server automatically and about the way the Emacs icon is set
up to start Emacs.  Can you please explain how the changes in these
two aspects of the default Emacs behavior make it easier to learn
Emacs on Windows?  Because I cannot imagine how these two minor
subtleties can change anything in the generally steep curve of
learning Emacs from scratch.  The major difficulties are elsewhere,
and if we really care about that, we should divert our energy to
overcoming those major difficulties, instead of wasting it on whether
server should be up or down by default.

Re: EmacsW32 invocation options

[Juanma Barranquero]

On 5/4/07, Eli Zaretskii <eliz@gnu.org> wrote:

> (He might be better off choosing a less radical word than
> ``destructive'', but then he's not a native English speaker, so we all
> might cut him some slack.)

I should've said "not productive", but I was mimicking the use of
"destructive" in one of Lennart's answers to you.

> In this case, Lennart's attitude to put into his distribution every
> patch he finds useful, but cannot convince the Emacs developers to
> accept, is IMO not the best use of that freedom.  It is confusing to
> users who expect Emacs to behave as documented and to do that
> consistently on all platforms.  And it's an attitude that smells of
> schism and forking, which many of us will not approve, because it
> diminishes the already small group of active developers by spreading
> their insufficient resources between several competing projects.

You've nicely summarized the main points of the discussion. Thanks.

             Juanma

Re: EmacsW32 invocation options

[Juanma Barranquero]

On 5/4/07, Eli Zaretskii <eliz@gnu.org> wrote:

> As things frequently are with changes suggested by Lennart that other
> developers don't like, Lennart will continue arguing and will not
> accept differing opinions or modify his suggestions according to
> comments he receives, no matter what.

And finally he will incorporate these changes into his distro and
claim that without them Emacs is too complicate or lacking essential
functionality. Which I don't think it's true.

> When such endless and futile
> arguments go out of control, it is quite possible that someone would
> say "additional thought" meaning "not now, and I don't care when".

Funnily enough, that wasn't even the case in that discussion. IIRC,
most people commenting in the original thread *wanted* the
functionality. No one was suggesting it was bad, just underdesigned
and/or not cleanly implemented *for the moment being*.

             Juanma

Re: EmacsW32 invocation options

[Brian Elmegaard]

Eli Zaretskii <eliz@gnu.org> writes:

> up to start Emacs.  Can you please explain how the changes in these
> two aspects of the default Emacs behavior make it easier to learn
> Emacs on Windows?  

No, except that I have experienced all kind of trouble with getting it
to work in the past and I haven't got that with the latest installers
from Lennart.

> and if we really care about that, we should divert our energy to
> overcoming those major difficulties, instead of wasting it on whether
> server should be up or down by default.

I see.

-- 
Brian (remove the sport for mail)
http://www.et.web.mek.dtu.dk/Staff/be/be.html
http://www.rugbyklubben-speed.dk

Re: EmacsW32 invocation options

[Hadron]

Eli Zaretskii <eliz@gnu.org> writes:

>> From: Brian Elmegaard <brian@rkspeed-rugby.dk>
>> Date: Thu, 03 May 2007 23:13:14 +0100
>> 
>> "Juanma Barranquero" <lekktu@gmail.com> writes:
>> 
>> > environment. I'm arguing that Emacs on Windows is not harder to use or
>> > learn that Emacs on GNU/Linux or other environments, so changes to
>> 
>> hat may be true, but my own experience and my impression of watching
>> students required to learn emacs, is that emacs+EmacsW32 is easier to
>> learn emacs.
>
> We are not talking in general here, we are talking specifically about
> starting server automatically and about the way the Emacs icon is set
> up to start Emacs.  Can you please explain how the changes in these
> two aspects of the default Emacs behavior make it easier to learn
> Emacs on Windows?  Because I cannot imagine how these two minor
> subtleties can change anything in the generally steep curve of
> learning Emacs from scratch.  The major difficulties are elsewhere,
> and if we really care about that, we should divert our energy to
> overcoming those major difficulties, instead of wasting it on whether
> server should be up or down by default.

I'd say you have wasted more energy arguing about the "waste of energy"
of this little "nice to have" than you would have implementing it.

Re: EmacsW32 invocation options

[Eli Zaretskii]

> From: Brian Elmegaard <brian@rkspeed-rugby.dk>
> Date: Fri, 04 May 2007 12:21:18 +0100
> 
> Eli Zaretskii <eliz@gnu.org> writes:
> 
> > up to start Emacs.  Can you please explain how the changes in these
> > two aspects of the default Emacs behavior make it easier to learn
> > Emacs on Windows?  
> 
> No, except that I have experienced all kind of trouble with getting it
> to work in the past and I haven't got that with the latest installers
> from Lennart.

The installer is a different issue.  No one objects to that, and
personally I think it's a good idea.  But the installer is not the
subject of this thread.

Re: EmacsW32 invocation options

[Juanma Barranquero]

On 5/4/07, Hadron <hadronquark@gmail.com> wrote:

> I'd say you have wasted more energy arguing about the "waste of energy"
> of this little "nice to have" than you would have implementing it.

That's beyond unfair.

Eli has spent prodigious amounts of energy in developing Emacs in
general, and Emacs on DOS and Windows in particular, not to mention
his attention and dedication to improving the documentation. There are
at least 1313 ChangeLog entries bearing his name, and the AUTHORS file
says:

Eli Zaretskii: wrote codepage.el rxvt.el tty-colors.el
and changed msdos.c Makefile.in makefile.w32-in files.el info.el fileio.c
  startup.el mainmake.v2 config.bat menu-bar.el pc-win.el rmail.el
  simple.el msdos.h internal.el xfaces.c emacs.c frame.c dosfns.c
  faces.el frame.el and 531 other files

I'd say he's quite knowledgeable about spending energy in the
betterment of Emacs...

             Juanma

Re: EmacsW32 invocation options

[Hadron]

"Juanma Barranquero" <lekktu@gmail.com> writes:

> On 5/4/07, Hadron <hadronquark@gmail.com> wrote:
>
>> I'd say you have wasted more energy arguing about the "waste of energy"
>> of this little "nice to have" than you would have implementing it.
>
> That's beyond unfair.

No it isn't. I made absolutely zero inference on the great things Eli
has done in the past. In *this* instance it strikes me that the old
guard are being unfairly hostile to Lennart and his efforts. Possibly
it's just the way it comes across.

If it seemed I was in anyway degrading or disputing Eli's efforts for
Emacs then I sincerely apologise. That was not my intention.

>
> Eli has spent prodigious amounts of energy in developing Emacs in
> general, and Emacs on DOS and Windows in particular, not to mention
> his attention and dedication to improving the documentation. There are
> at least 1313 ChangeLog entries bearing his name, and the AUTHORS file
> says:
>
> Eli Zaretskii: wrote codepage.el rxvt.el tty-colors.el
> and changed msdos.c Makefile.in makefile.w32-in files.el info.el fileio.c
>  startup.el mainmake.v2 config.bat menu-bar.el pc-win.el rmail.el
>  simple.el msdos.h internal.el xfaces.c emacs.c frame.c dosfns.c
>  faces.el frame.el and 531 other files
>
> I'd say he's quite knowledgeable about spending energy in the
> betterment of Emacs...
>
>             Juanma
>

-- 

Re: EmacsW32 invocation options

[Juanma Barranquero]

On 5/4/07, Hadron <hadronquark@googlemail.com> wrote:

> No it isn't. I made absolutely zero inference on the great things Eli
> has done in the past.

Have you at least checked the original discussion of the emacsclient issue?

             Juanma

Re: EmacsW32 invocation options

[Lennart Borgman (gmail)]

Juanma Barranquero wrote:
> On 5/4/07, Eli Zaretskii <eliz@gnu.org> wrote:
> 
>> (He might be better off choosing a less radical word than
>> ``destructive'', but then he's not a native English speaker, so we all
>> might cut him some slack.)
> 
> I should've said "not productive", but I was mimicking the use of
> "destructive" in one of Lennart's answers to you.
> 
>> In this case, Lennart's attitude to put into his distribution every
>> patch he finds useful, but cannot convince the Emacs developers to
>> accept, is IMO not the best use of that freedom.  It is confusing to
>> users who expect Emacs to behave as documented and to do that
>> consistently on all platforms.  And it's an attitude that smells of
>> schism and forking, which many of us will not approve, because it
>> diminishes the already small group of active developers by spreading
>> their insufficient resources between several competing projects.
> 
> You've nicely summarized the main points of the discussion. Thanks.


For those interested here is a list of what patches there are in the 
patched version of Emacs+EmacsW32:

   file:///C:/EmacsW32/etc/EmacsW32.html#patchcont

Re: EmacsW32 invocation options

[Lennart Borgman (gmail)]

Lennart Borgman (gmail) wrote:
> Juanma Barranquero wrote:
>> On 5/4/07, Eli Zaretskii <eliz@gnu.org> wrote:
>>
>>> (He might be better off choosing a less radical word than
>>> ``destructive'', but then he's not a native English speaker, so we all
>>> might cut him some slack.)
>>
>> I should've said "not productive", but I was mimicking the use of
>> "destructive" in one of Lennart's answers to you.
>>
>>> In this case, Lennart's attitude to put into his distribution every
>>> patch he finds useful, but cannot convince the Emacs developers to
>>> accept, is IMO not the best use of that freedom.  It is confusing to
>>> users who expect Emacs to behave as documented and to do that
>>> consistently on all platforms.  And it's an attitude that smells of
>>> schism and forking, which many of us will not approve, because it
>>> diminishes the already small group of active developers by spreading
>>> their insufficient resources between several competing projects.
>>
>> You've nicely summarized the main points of the discussion. Thanks.
> 
> 
> For those interested here is a list of what patches there are in the 
> patched version of Emacs+EmacsW32:
> 
>   file:///C:/EmacsW32/etc/EmacsW32.html#patchcont


Damn, wrong window ;-)

  http://ourcomments.org/Emacs/EmacsW32.html#unpatched

Re: EmacsW32 invocation options

[Eli Zaretskii]

> Date: Fri, 04 May 2007 16:54:18 +0200
> From: Hadron <hadronquark@googlemail.com>
> Cc: help-gnu-emacs@gnu.org, Hadron <hadronquark@gmail.com>
> 
> In *this* instance it strikes me that the old guard are being
> unfairly hostile to Lennart and his efforts.

I challenge you to find and point out any hostility whatsoever in
anything I wrote here.  Disagreement, yes; but hostility, no.

Re: EmacsW32 invocation options

[Kim F. Storm]

Eli Zaretskii <eliz@gnu.org> writes:

>> I don't see how someone using the freedom of GPL should be accused of
>> being destructive, just because he does something you don't approve of.
>
> Kim, I hope you've read this thread before responding, because you've
> taken one phrase out of context and jumped the gun on it.

I did try to follow the arguments - but the discussion was going around
in circles, and was becoming quite personal.  Accusing Lennart of being
destructive was unjustified IMHO.

> (He might be better off choosing a less radical word than
> ``destructive'', but then he's not a native English speaker, so we all
> might cut him some slack.)  

Neither am I :-)

>                             The issue is how to use this freedom in
> the best interests of the community at large.

Lennart obviously believes that he is helping _part of_ the community
to a better Emacs experience.

> In this case, Lennart's attitude to put into his distribution every
> patch he finds useful, but cannot convince the Emacs developers to
> accept, is IMO not the best use of that freedom.  It is confusing to
> users who expect Emacs to behave as documented and to do that
> consistently on all platforms.  

IMO, we have to balance this "cross-platform consistency" with the
user's expectation of "Emacs working seamlessly on _my_ platform".

>                                 And it's an attitude that smells of
> schism and forking, which many of us will not approve, because it
> diminishes the already small group of active developers by spreading
> their insufficient resources between several competing projects.

So why don't we try to find ways to accomodate people like Lennart who
really wants to contribute, but whose ideas are too radical to be the
default behaviour?

E.g. by supporting two installation options

a) maximum cross-platform consistency
b) maximum adaptation to current platform APIs.


> Sure, he does, and he deserves a lot of credit and thanks for that.
> As does Juanma, btw, who made emacslient work on Windows to begin
> with.

Sure.  I never said or meant otherwise.


-- 
Kim F. Storm  http://www.cua.dk

Re: EmacsW32 invocation options

[Peter Dyballa]

Am 05.05.2007 um 00:09 schrieb Kim F. Storm:

>>                                 And it's an attitude that smells of
>> schism and forking, which many of us will not approve, because it
>> diminishes the already small group of active developers by spreading
>> their insufficient resources between several competing projects.
>
> So why don't we try to find ways to accomodate people like Lennart who
> really wants to contribute, but whose ideas are too radical to be the
> default behaviour?
>
> E.g. by supporting two installation options
>
> a) maximum cross-platform consistency
> b) maximum adaptation to current platform APIs.

For Mac OS X we have the X clients of GNU Emacsen 21.x (with or  
without GTK), 22.x (with or without GTK), and 23.0.0; we have also a  
few flavours of "Carbon Emacsen" – and we have the Cocoa or GNUstep/ 
OPENSTEP version of Emacs.app based on version 23.0.0.

I don't think it's a bad situation. Their developers have particular  
hobbies or preferences. If they were not allowed to change things  
they might be like me: not developing anything. But they do write and  
improve patches to add transparency to Emacs, or to open every buffer  
in its own frame, or to overcome the inability of GNU Emacs (and I  
mean the "official" and "stable" version here) to deal with Unicode.  
Or to enable Emacs to print non-US-ASCII and non-ISO 8859-1 characters.

IMO it would be fine if the "official" GNU Emacs site would point to  
sites providing "non-official" flavours of GNU Emacs, and these sites  
would document their changes and additions.

--
Greetings

   Pete

Upgraded:  Didn't work the first time.

Re: EmacsW32 invocation options

[David Kastrup]

Eli Zaretskii <eliz@gnu.org> writes:

>> From: Brian Elmegaard <brian@rkspeed-rugby.dk>
>> Date: Thu, 03 May 2007 07:24:02 +0100
>> 
>> Eli Zaretskii <eliz@gnu.org> writes:
>> 
>> > It makes it harder because they use Emacs that behaves not as
>> > documented in subtle ways.
>> 
>> >From a short google on groups it seems that the emacs developers find
>> emacsW32 ("EmacsW32 is not Emacs for MS Windows. Instead it is an
>> add-on to Emacs for MS Windows.") worse than aquamacs ("Aquamacs
>> features extensive customization: it will feel and behave mostly like
>> an Aqua program"). 
>> 
>> Why is that?
>
> Maybe because most Emacs developers don't know anything about
> Aquamacs.  That includes myself.

Well, I had the doubtful pleasure of getting to try showing someone on
a conference how to work with it.  What a piece of crock.  C-x 3 and
your window splits vertically, showing the same buffer in both
windows.  So far, so good.  C-x b and a separate frame opens.  C-x 4 b
and another frame opens.  It is impossible to actually switch buffers
in one of the vertically aligned windows: instead new frames open
every time.

After a while, I give up on it and tell the person to get and install
Carbonemacs which is supposed to be less "Maclike", a euphemism for
"braindead" that more likely than not is unjust to Apple.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: EmacsW32 invocation options

[David Kastrup]

no-spam@cua.dk (Kim F. Storm) writes:

> So why don't we try to find ways to accomodate people like Lennart
> who really wants to contribute, but whose ideas are too radical to
> be the default behaviour?
>
> E.g. by supporting two installation options
>
> a) maximum cross-platform consistency
> b) maximum adaptation to current platform APIs.

I consider choice a bad idea here.  The correct thing to do is to have
both.  This should be done using customization themes.  If a person
wants to have his Emacs behave like on Windows when on Windows,
chances are that he wants it to behave like on Windows also when he is
on the Mac.  One point of Emacs is that it works the same everywhere.
It is fine to let a user have it work like a Windows application at
the drop of a hat under Windows, but it does not make sense to not let
the user have the application work identically on a Mac.

The point of Emacs is feeling at home everywhere.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: EmacsW32 invocation options

[Juanma Barranquero]

On 5/5/07, Kim F. Storm <no-spam@cua.dk> wrote:

> but the discussion was going around
> in circles

Yes. In a previous post Eli has described the process quite accurately.

> and was becoming quite personal.

Personal?

> Accusing Lennart of being
> destructive was unjustified IMHO.

And accusing Eli of the same wasn't? I'm sitting here waiting for you
to say to Lennart the same you said to me...

> Neither am I :-)

Your command of English is obviously much better than mine (I'm
self-taught, after all), but as I've said, I choose to use
"destructive" for a specific reason; that was no slip.

> IMO, we have to balance this "cross-platform consistency" with the
> user's expectation of "Emacs working seamlessly on _my_ platform".

There are specific limits to that balance. For example, if I'm not
mistaken, RMS forbids changes that would give Emacs an advantage on
Windows over GNU/Linux or other free operating systems.

> So why don't we try to find ways to accomodate people like Lennart who
> really wants to contribute, but whose ideas are too radical to be the
> default behaviour?

Perhaps you're right, but here we aren't discussing a radical idea,
just an half-cooked one (IMO, etc.).

             Juanma

Re: EmacsW32 invocation options

[David Kastrup]

"Juanma Barranquero" <lekktu@gmail.com> writes:

> There are specific limits to that balance. For example, if I'm not
> mistaken, RMS forbids changes that would give Emacs an advantage on
> Windows over GNU/Linux or other free operating systems.

No, he forbids changes that would give Windows an advantage beneath
Emacs over GNU/Linux or other free systems.

"I would not be using Windows if it were not for Emacs" is no outcome
he is interested in.  I can follow him on that.

I also find it disconcerting if programmers enjoying the benefits of
free software don't bother helping anybody else reaping the benefits
of benefitting from their adaptations.  It always has been one goal of
free software to be available to as many people as possible, and the
"I cater for myself, let everybody else cater for themselves" stance
is not particularly helpful when 90% of potential users are actually
using non-free systems.

At least I don't care for a "Emacs - best used on Windows" situation.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: EmacsW32 invocation options

[Eli Zaretskii]

> Cc: help-gnu-emacs@gnu.org
> From: no-spam@cua.dk (Kim F. Storm)
> Date: Sat, 05 May 2007 00:09:14 +0200
> 
> IMO, we have to balance this "cross-platform consistency" with the
> user's expectation of "Emacs working seamlessly on _my_ platform".

I think we already do.  However, in this particular case, I cannot
believe that Windows users somehow expect the server to be on, while
users on other platforms do not.  There's no similar feature in any
popular Windows program, AFAIK, so I don't expect any expectations
about it.

> So why don't we try to find ways to accomodate people like Lennart who
> really wants to contribute, but whose ideas are too radical to be the
> default behaviour?

How do you do that when the radical idea in the case in point is
_precisely_ to change the default behavior?

> E.g. by supporting two installation options
> 
> a) maximum cross-platform consistency
> b) maximum adaptation to current platform APIs.

These a and b are not of the same category: the issue we are talking
about is user expectations, not programmer expectations.  API is
something users are not aware of and are not concerned with.

But other than that, I think Emacs already balances these two to a
very good measure, and the developers are well aware of user
expectations and the need to cater to them.  For example, the support
for different GUI toolkits (including the Windows one) is the result
of this.

Re: EmacsW32 invocation options

[Eli Zaretskii]

> Cc: Eli Zaretskii <eliz@gnu.org>,
>  help-gnu-emacs@gnu.org
> From: Peter Dyballa <Peter_Dyballa@Web.DE>
> Date: Sat, 5 May 2007 01:03:23 +0200
> 
> I don't think it's a bad situation. Their developers have particular  
> hobbies or preferences. If they were not allowed to change things  
> they might be like me: not developing anything.

No one can disallow changing Free Software.  But no one can disallow
me from having an opinion about such projects, either.

Now, while I'm quite ignorant about the Mac and the user expectations
on that platform, I do know a lot about Windows (and work on that OS
myself many hours a day).  So, while my opinion about the situation on
the Mac does not matter (and therefore I won't comment on it), that is
not the case on Windows.  I don't think the Windows environment is so
different from the Unix environment to justify different defaults in
most aspects of Emacs behavior.  (We do have different defaults in
some aspects that do matter, such as the end-of-line convention, for
example.)

> IMO it would be fine if the "official" GNU Emacs site would point to  
> sites providing "non-official" flavours of GNU Emacs, and these sites  
> would document their changes and additions.

Maybe it is a good idea, but I still think that proliferation of
different default behaviors is an impediment for newbies.  Witness the
problems with helping Debian users because the packages are influenced
by site-init files that change many aspects of the default behavior.

Re: EmacsW32 invocation options

[Juanma Barranquero]

On 5/5/07, David Kastrup <dak@gnu.org> wrote:

> No, he forbids changes that would give Windows an advantage beneath
> Emacs over GNU/Linux or other free systems.
>
> "I would not be using Windows if it were not for Emacs" is no outcome
> he is interested in.  I can follow him on that.

He didn't want a patch to make Emacs act as a COM server. It's hard to
think that many people would say "I would not be using Windows if it
were not for Emacs being a COM server".

> I also find it disconcerting if programmers enjoying the benefits of
> free software don't bother helping anybody else reaping the benefits
> of benefitting from their adaptations.  It always has been one goal of
> free software to be available to as many people as possible, and the
> "I cater for myself, let everybody else cater for themselves" stance
> is not particularly helpful when 90% of potential users are actually
> using non-free systems.

Sorry, you lost me here. I don't see how this applies to the
discussion. Could you please elaborate?

             Juanma

Re: EmacsW32 invocation options

[David Kastrup]

"Juanma Barranquero" <lekktu@gmail.com> writes:

> On 5/5/07, David Kastrup <dak@gnu.org> wrote:
>
>> No, he forbids changes that would give Windows an advantage beneath
>> Emacs over GNU/Linux or other free systems.
>>
>> "I would not be using Windows if it were not for Emacs" is no outcome
>> he is interested in.  I can follow him on that.
>
> He didn't want a patch to make Emacs act as a COM server. It's hard
> to think that many people would say "I would not be using Windows if
> it were not for Emacs being a COM server".

"I would not be using Windows if it wasn't the only platform where I
could seamlessly integrate Emacs into my desktop" is quite likely.
After all, that's the whole point of making it a COM server.  So
instead of making it a Windows-only application server, I find it
legitimate to demand a solution that won't work Windows-only and might
very well _block_ a clean implementation of, say, CORBA because of
Windows-only design decisions.

>> I also find it disconcerting if programmers enjoying the benefits
>> of free software don't bother helping anybody else reaping the
>> benefits of benefitting from their adaptations.  It always has been
>> one goal of free software to be available to as many people as
>> possible, and the "I cater for myself, let everybody else cater for
>> themselves" stance is not particularly helpful when 90% of
>> potential users are actually using non-free systems.
>
> Sorry, you lost me here. I don't see how this applies to the
> discussion. Could you please elaborate?

If 90% of Emacs functionality worked only on Windows, it would be
pretty useless for promoting free software.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: EmacsW32 invocation options

[Brian Elmegaard]

Eli Zaretskii <eliz@gnu.org> writes:

> The installer is a different issue.  No one objects to that, and
> personally I think it's a good idea.  But the installer is not the
> subject of this thread.

No, I understand that now and I agree that portability between systems
is an important issue.

-- 
Brian (remove the sport for mail)
http://www.et.web.mek.dtu.dk/Staff/be/be.html
http://www.rugbyklubben-speed.dk

Re: EmacsW32 invocation options

[Brian Elmegaard]

David Kastrup <dak@gnu.org> writes:

> both.  This should be done using customization themes.  If a person
> wants to have his Emacs behave like on Windows when on Windows,
> chances are that he wants it to behave like on Windows also when he is
> on the Mac.  One point of Emacs is that it works the same everywhere.

I think you are right. And even though I am now again going OT, this
is what emacsW32 users get when they see the complete EmacsW32
customization buffer.

-- 
Brian (remove the sport for mail)
http://www.et.web.mek.dtu.dk/Staff/be/be.html
http://www.rugbyklubben-speed.dk

Re: EmacsW32 invocation options

[David Kastrup]

Brian Elmegaard <brian@rkspeed-rugby.dk> writes:

> David Kastrup <dak@gnu.org> writes:
>
>> both.  This should be done using customization themes.  If a person
>> wants to have his Emacs behave like on Windows when on Windows,
>> chances are that he wants it to behave like on Windows also when he is
>> on the Mac.  One point of Emacs is that it works the same everywhere.
>
> I think you are right. And even though I am now again going OT, this
> is what emacsW32 users get when they see the complete EmacsW32
> customization buffer.

And nobody else gets this, and the emacsW32 users don't get the option
to let their Emacs behave like Carbon Emacs or Aquamacs.

So either you are missing my point, or not replying to it.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: EmacsW32 invocation options

[Lennart Borgman (gmail)]

David Kastrup wrote:
> Brian Elmegaard <brian@rkspeed-rugby.dk> writes:
> 
>> David Kastrup <dak@gnu.org> writes:
>>
>>> both.  This should be done using customization themes.  If a person
>>> wants to have his Emacs behave like on Windows when on Windows,
>>> chances are that he wants it to behave like on Windows also when he is
>>> on the Mac.  One point of Emacs is that it works the same everywhere.
>> I think you are right. And even though I am now again going OT, this
>> is what emacsW32 users get when they see the complete EmacsW32
>> customization buffer.
> 
> And nobody else gets this, and the emacsW32 users don't get the option
> to let their Emacs behave like Carbon Emacs or Aquamacs.
> 
> So either you are missing my point, or not replying to it.

It is a good point. From the beginning EmacsW32 was written to make that 
possible (though there are now some things specific to w32, like looking 
up programs for use in inferior processes).

Re: EmacsW32 invocation options

[David Reitter]

On May 5, 1:58 am, David Kastrup <d...@gnu.org> wrote:

> So far, so good.  C-x b and a separate frame opens.  C-x 4 b
> and another frame opens.  It is impossible to actually switch buffers
> in one of the vertically aligned windows: instead new frames open
> every time.

M-x switch-to-buffer-here, or C-x B.
Or, intuitively: right mouse click into the window, (or C-mouse-1),
then select a buffer in the submenu "Switch Buffer".

Don't expect Aquamacs to behave like Emacs. Otherwise there'd be no
point in making it.

Re: EmacsW32 invocation options

[Juanma Barranquero]

On 5/5/07, David Kastrup <dak@gnu.org> wrote:

> "I would not be using Windows if it wasn't the only platform where I
> could seamlessly integrate Emacs into my desktop" is quite likely.

Well, I think that's also unlikely, but I suppose we'll have to agree
to disagree on this one.

> So
> instead of making it a Windows-only application server, I find it
> legitimate to demand a solution that won't work Windows-only and might
> very well _block_ a clean implementation of, say, CORBA because of
> Windows-only design decisions.

I find it legitimate, too. I wasn't complaining about Richard's decision.

> If 90% of Emacs functionality worked only on Windows, it would be
> pretty useless for promoting free software.

Aha. We're agreeing here, then. There's no point in ad-hoc,
system-specific "fixes" when a general one could be developed with
similar effort.

             Juanma

Re: EmacsW32 invocation options

[Peter Dyballa]

Am 05.05.2007 um 02:58 schrieb David Kastrup:

> Well, I had the doubtful pleasure of getting to try showing someone on
> a conference how to work with it.  What a piece of crock.  C-x 3 and
> your window splits vertically, showing the same buffer in both
> windows.  So far, so good.  C-x b and a separate frame opens.  C-x 4 b
> and another frame opens.  It is impossible to actually switch buffers
> in one of the vertically aligned windows: instead new frames open
> every time.
>
> After a while, I give up on it and tell the person to get and install
> Carbonemacs which is supposed to be less "Maclike", a euphemism for
> "braindead" that more likely than not is unjust to Apple.

Aquamacs Emacs seems indeed to be stuck in the farthest end of a dead- 
end street: recently it stopped working completely when my ~/.emacs  
file (for versions 20...23, running with or without windows) is  
present (of course it has its own preferences files hidden here and  
there and elsewhere). It just loops endlessly (but *I* don't let it  
that long!). Maybe this lack of backward compatibility is its  
particular Mac likeliness. And this might be in the diagonally away  
corner of Apple's HID (Human Interface Definition guide line). (At  
least there was an option to make it create new frames only when its  
user wishes this.)

--
Mit friedvollen Grüßen

   Pete

There are three types of people in this world: those who can count,  
and those who cannot.

Re: EmacsW32 invocation options

[David Kastrup]

David Reitter <david.reitter@gmail.com> writes:

> On May 5, 1:58 am, David Kastrup <d...@gnu.org> wrote:
>
>> So far, so good.  C-x b and a separate frame opens.  C-x 4 b and
>> another frame opens.  It is impossible to actually switch buffers
>> in one of the vertically aligned windows: instead new frames open
>> every time.
>
> M-x switch-to-buffer-here, or C-x B.
> Or, intuitively: right mouse click into the window, (or C-mouse-1),
> then select a buffer in the submenu "Switch Buffer".
>
> Don't expect Aquamacs to behave like Emacs. Otherwise there'd be no
> point in making it.

At least I would no longer bother with the subjunctive mode on that
last sentence.

Carbonemacs _does_ behave like Emacs, and there is a point in making
it, even if you don't believe it.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: EmacsW32 invocation options

[Brian Elmegaard]

David Kastrup <dak@gnu.org> writes:

> And nobody else gets this, and the emacsW32 users don't get the option
> to let their Emacs behave like Carbon Emacs or Aquamacs.

For the moment. It would be great to have the option.
-- 
Brian (remove the sport for mail)
http://www.et.web.mek.dtu.dk/Staff/be/be.html
http://www.rugbyklubben-speed.dk