gnu vs. xemacs

gnu vs. xemacs

[H.]

I was just curious why people in this community prefer gnu over xemacs,
or if that is the case, and, whatever your preference is, why you feel
that way. I'm neutral myself, being just a beginner. All opinions
welcome. For instance, perhaps these difference forks are good for
different things...?

Re: gnu vs. xemacs

[Lennart Borgman (gmail)]

H. wrote:
> I was just curious why people in this community prefer gnu over xemacs,
> or if that is the case, and, whatever your preference is, why you feel
> that way. I'm neutral myself, being just a beginner. All opinions
> welcome. For instance, perhaps these difference forks are good for
> different things...?


See http://www.emacswiki.org/cgi-bin/wiki/EmacsAndXEmacs

Re: gnu vs. xemacs

[Peter Dyballa]

Am 27.12.2006 um 12:46 schrieb H.:

> I was just curious why people in this community prefer gnu over  
> xemacs,
> or if that is the case, and, whatever your preference is, why you feel
> that way. I'm neutral myself, being just a beginner. All opinions
> welcome. For instance, perhaps these difference forks are good for
> different things...?

GNU Emacs uses the key bindings I am used to.

It can be compiled as Carbon Emacs, X11 client in so many tastes,  
COCOA/OPENSTEP application, and as non-windowing programme for  
Terminal or xterm on Mac OS X.

--
Greetings

   Pete

“I hope to die before I *have* to use Microsoft Word.”
      - Donald E. Knuth, 2001-10-02 in Tübingen.

Re: gnu vs. xemacs

[Sven Bretfeld]

When I started using Emacs one year ago I knew almost nothing. I was
thinking XEmacs is Emacs for X-Window systems (like many people seem
to do). So I installed it, mainly because I use LaTeX since many
years and I was looking for a Linux editor as powerful as WinEdt which
I had been using under MS-Windows. 

To my disappointment XEmacs was not able to handle all the
utf8-characters I needed. It might be a matter of my too small
knowledge at that time, but I couldn't get the diacritics needed for
Sanskrit transcriptions to work. Somewhere I read that XEmacs doesn't
fully support unicode representation. However, it worked immediately
with GnuEmacs. Namely I need characters like underdotted t or n which
represent certain Indian consonants spoken with the tongue bent
backwards.

Best wishes,
Sven

Re: gnu vs. xemacs

[David Kastrup]

"H." <hbe123@gmail.com> writes:

> I was just curious why people in this community prefer gnu over
> xemacs, or if that is the case, and, whatever your preference is,
> why you feel that way. I'm neutral myself, being just a
> beginner. All opinions welcome. For instance, perhaps these
> difference forks are good for different things...?

Except for some features of its own, development of XEmacs trails that
of Emacs for about 5 years.  For those features where XEmacs is
leading, the documentation is still trailing for about 5 years, so one
can't actually make educated use of them.  Development for XEmacs is
not user-driven: problem reports and feature and documentation
requests from non-core programmers tend to fall by the wayside or get
argued away.

It is my impression that GNU Emacs has a much higher standard of
documentation and coherence in API design and user interface.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: gnu vs. xemacs

[Ramprasad]

On Wed, 2006-12-27 at 19:29, Sven Bretfeld wrote:

> years and I was looking for a Linux editor as powerful as WinEdt which

It's GNU/Linux system.

Please read 

* http://www.gnu.org/gnu/gnu-users-never-heard-of-gnu.html

* http://www.gnu.org/gnu/why-gnu-linux.html

-
Ramprasad B

Re: gnu vs. xemacs

[Tim X]

Ramprasad <ramprasad@gnu.org> writes:

> On Wed, 2006-12-27 at 19:29, Sven Bretfeld wrote:
>
>> years and I was looking for a Linux editor as powerful as WinEdt which
>
> It's GNU/Linux system.
>
> Please read 
>
> * http://www.gnu.org/gnu/gnu-users-never-heard-of-gnu.html
>
> * http://www.gnu.org/gnu/why-gnu-linux.html
>

You don't think your being a bit pedantic do you? After all, Linux is
what GNU/Linux is commonly called. You also have lots of other
references out there which don't include the "GNU" part, such as The
Linux Journal, the Linux Gazette, Linux Weekly news,
linuxquestions.org, linuxkernel.org, etc etc.

At least if your going to take the time to post a response, maybe add
something for the OP in addition to correcting his colloquial use of
Linux rather than the technically correct one. 

To the OP, my preferred editor is GNU Emacs rather than Xemacs because
not all the packages I want to use are supported by Xemacs. In
particular, there are some differences I've come across in the way
that font locking is handled and some of the packages I use regularly
don't work under Xemacs. I've also written a few of my own simple
elisp packages and really couldn't be bothered adding a heap of code
which checks to see which version you are running under and then take
appropriate action to handle diffeences in functions or the arguments
they take etc. I've also found the default settings in GNU Emacs to be
more in line with what I expect/want. 

Tim

-- 
tcross (at) rapttech dot com dot au

Re: gnu vs. xemacs

[insert name]

H. wrote:
> I was just curious why people in this community prefer gnu over xemacs,
> or if that is the case, and, whatever your preference is, why you feel
> that way. I'm neutral myself, being just a beginner. All opinions
> welcome. For instance, perhaps these difference forks are good for
> different things...?
> 


Emacs has much more documentation both on the web and in hard copy.

Re: gnu vs. xemacs

[Matthew Flaschen]

[-- Attachment #1.1: Type: text/plain, Size: 781 bytes --]

Tim X wrote:
> You don't think your being a bit pedantic do you? After all, Linux is
> what GNU/Linux is commonly called. You also have lots of other
> references out there which don't include the "GNU" part, such as The
> Linux Journal, the Linux Gazette, Linux Weekly news,
> linuxquestions.org, linuxkernel.org, etc etc.

> At least if your going to take the time to post a response, maybe add
> something for the OP in addition to correcting his colloquial use of
> Linux rather than the technically correct one.

I don't think he's being pedantic, and the fact that the error is common
does not mean it's insignificant.  It's not only technically correct,
it's ethically correct.

As for the OP, I can't help since I've never used XEmacs.

Matt Flaschen



[-- Attachment #1.2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 252 bytes --]

[-- Attachment #2: Type: text/plain, Size: 152 bytes --]

_______________________________________________
help-gnu-emacs mailing list
help-gnu-emacs@gnu.org
http://lists.gnu.org/mailman/listinfo/help-gnu-emacs

Re: gnu vs. xemacs

[David Kastrup]

insert name <nospam@nospam.org> writes:

> H. wrote:
>> I was just curious why people in this community prefer gnu over
>> xemacs, or if that is the case, and, whatever your preference is,
>> why you feel that way. I'm neutral myself, being just a
>> beginner. All opinions welcome. For instance, perhaps these
>> difference forks are good for different things...?
>
> Emacs has much more documentation both on the web and in hard copy.

Actually, given the amount of partly outdated, partly wrong junk you
can find for anything on the web and in hard copy, this is not
particularly persuasive.  However, Emacs comes with _integrated_
documentation (which you can also print out if necessary) that is, in
my opinion, more comprehensible, complete and up to date than the
respective integrated documentation of XEmacs.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: gnu vs. xemacs

[Tim X]

Matthew Flaschen <matthew.flaschen@gatech.edu> writes:

> Tim X wrote:
>> You don't think your being a bit pedantic do you? After all, Linux is
>> what GNU/Linux is commonly called. You also have lots of other
>> references out there which don't include the "GNU" part, such as The
>> Linux Journal, the Linux Gazette, Linux Weekly news,
>> linuxquestions.org, linuxkernel.org, etc etc.
>
>> At least if your going to take the time to post a response, maybe add
>> something for the OP in addition to correcting his colloquial use of
>> Linux rather than the technically correct one.
>
> I don't think he's being pedantic, and the fact that the error is common
> does not mean it's insignificant.  It's not only technically correct,
> it's ethically correct.
>
> As for the OP, I can't help since I've never used XEmacs.
>

Perhaps your right - I guess what got to me was the fact the
respondent was able to take the time to correct a common error which
nearly everyone is guilty of, but failed to actually address the OPs
question in any way. I mean, is he going to correct every post that
refers to just Linux rather than GNU/Linux?

Although I understand why we should encourage people to not drop the
GNU part, I suspect that basic human tendency to shorten descriptions
will see Linux being used far more often than GNU/Linux. We have
far greater ethical problems relating to how people use GNU software
and misunderstanding regarding free software and open source than
whether people use the "formal" title or not. IMO if we want users to
adopt a specific terminology, we need to provide positive
reinforcement to do so in addition to criticising them when they fail
to. Actually assisting someone together with correcting their
understandable error in terms would assist in this. Just criticism
without any assistance is more likely to have the opposite affect. 

Tim


-- 
tcross (at) rapttech dot com dot au

Re: gnu vs. xemacs

[Leo]

* Tim X. (2006-12-29 10:53 +1100) said:
  ^^^^^^
[...]

> Perhaps your right - I guess what got to me was the fact the
> respondent was able to take the time to correct a common error which
> nearly everyone is guilty of, but failed to actually address the OPs
> question in any way. I mean, is he going to correct every post that
> refers to just Linux rather than GNU/Linux?

In every emacs-* mailing lists, it has been extremely consistent to
use Linux and GNU/Linux i.e. Linux means the kernel and GNU/Linux is
the operating system that contains a Linux kernel.

Making effort to preserve this consistence should be appreciated.

-- 
Leo <sdl.web AT gmail.com>                         (GPG Key: 9283AA3F)

"MIT/GNU/Linux" (was: gnu vs. xemacs)

[B. Smith-Mannschott]

On Dec 29, 2006, at 02:12, Leo wrote:
> In every emacs-* mailing lists, it has been extremely consistent to
> use Linux and GNU/Linux i.e. Linux means the kernel and GNU/Linux is
> the operating system that contains a Linux kernel.

So lemme see...
(1) "Linux" is just kernel
(2) GNU/Linux is the operating system formed by running the GNU tools  
on top of that kernel

=>

Actually, we should use "MIT/GNU/Linux", at least when including X11  
on top of GNU/Linux. I'm sure I wouldn't feel at all tempted to  
shorten a handy phrase like "MIT/GNU/Linux". It just rolls of the  
tongue.

I'm inclined to agree with the point made by the pedants of "GNU/ 
Linux", but I do wonder: where does it end?

*ducks*


// ben

RE: "MIT/GNU/Linux" (was: gnu vs. xemacs)

[Micha Feigin]

 

> -----Original Message-----
> From: help-gnu-emacs-bounces+michf=post.tau.ac.il@gnu.org 
> [mailto:help-gnu-emacs-bounces+michf=post.tau.ac.il@gnu.org] 
> On Behalf Of B. Smith-Mannschott
> Sent: Friday, December 29, 2006 2:46 PM
> To: help-gnu-emacs@gnu.org
> Subject: "MIT/GNU/Linux" (was: gnu vs. xemacs)
> 
> 
> On Dec 29, 2006, at 02:12, Leo wrote:
> > In every emacs-* mailing lists, it has been extremely consistent to 
> > use Linux and GNU/Linux i.e. Linux means the kernel and 
> GNU/Linux is 
> > the operating system that contains a Linux kernel.
> 

The operating system is the kernel. GNU is one of the more popular option (but
not the only one) for the tools running over the operating system. I think
GNU/Linux is just RMS being overly pushy and and egomaniac.

> So lemme see...
> (1) "Linux" is just kernel

Linux is the operating system

> (2) GNU/Linux is the operating system formed by running the 
> GNU tools on top of that kernel
> 
> =>
> 
> Actually, we should use "MIT/GNU/Linux", at least when 

What about qt etc ?

> including X11 on top of GNU/Linux. I'm sure I wouldn't feel 

What if you replace GNU tools? (there are alternative options to most if not
all)

> at all tempted to shorten a handy phrase like 
> "MIT/GNU/Linux". It just rolls of the tongue.
> 
> I'm inclined to agree with the point made by the pedants of 
> "GNU/ Linux", but I do wonder: where does it end?
> 
> *ducks*
> 
> 
> // ben
> 
> 
> 
> 
> 
> _______________________________________________
> help-gnu-emacs mailing list
> help-gnu-emacs@gnu.org
> http://lists.gnu.org/mailman/listinfo/help-gnu-emacs
> 
> 
> 
> +++++++++++++++++++++++++++++++++++++++++++
>  This Mail Was Scanned By Mail-seCure System  at the Tel-Aviv 
> University CC.
> 
> 

RE: "MIT/GNU/Linux" (was: gnu vs. xemacs)

[Gian Uberto Lauri]

>>>>> "Micha" == Micha Feigin <michf@post.tau.ac.il> writes:

>> So lemme see...  (1) "Linux" is just kernel

Micha> Linux is the operating system

Nope, it's  just the kernel,  and take away  the GNU software  you are
left with just the kernel source. 

Micha> What if you replace GNU tools? (there are alternative options
Micha> to most if not all)

With what ?

-- 
 /\           ___
/___/\_|_|\_|__|___Gian Uberto Lauri_____
  //--\| | \|  |   Integralista GNUslamico
\/                 e coltivatore diretto di Software

Re: "MIT/GNU/Linux"

[Giorgos Keramidas]

On Fri, 29 Dec 2006 15:59:39 +0200,
"Micha Feigin" <michf@post.tau.ac.il> wrote:
>B. Smith-Mannschott wrote:
>>On Dec 29, 2006, at 02:12, Leo wrote:
>>> In every emacs-* mailing lists, it has been extremely consistent to
>>> use Linux and GNU/Linux i.e. Linux means the kernel and GNU/Linux
>>> is the operating system that contains a Linux kernel.
>
> The operating system is the kernel.

I have to respectfully disagree.  A kernel without the userland tools to
make good use of it is pretty much useless.

> GNU is one of the more popular option (but not the only one) for the
> tools running over the operating system. I think GNU/Linux is just RMS
> being overly pushy and and egomaniac.

No, RMS is right.

First of all, there are other examples where "parts" have a different
name from the "whole".  Consider for example the fine difference between
"SunOS" and "The Solaris Operating Environment" :)

Secondly, it may be cool to say that "Linux is only the kernel", but
this is a relatively strange way of looking at things.  You can't really
"boot" a standalong Linux kernel without the support of the myriad of
tools people have grafted on top of it.  A functional, and *complete*
operating environment -- at least from the perspective of the end user
-- includes both the Linux kernel *and* the surrounding tools.

Giorgos

Re: "MIT/GNU/Linux"

[Alexey Pustyntsev]

"Micha Feigin" <michf@post.tau.ac.il> writes:

> The operating system is the kernel. GNU is one of the more popular option (but
> not the only one) for the tools running over the operating system. I think
> GNU/Linux is just RMS being overly pushy and and egomaniac.
>

RMS is neither pushy nor egomaniac. 'GNU/Linux' is absolutely correct
and fair term as the GNU tools and the kernel are practically
indispensable to get the OS running.

-- 
Rgds
Alexey

Re: gnu vs. xemacs

[Hadron Quark]

Leo <sdl.web@gmail.com> writes:

> * Tim X. (2006-12-29 10:53 +1100) said:
>   ^^^^^^
> [...]
>
>> Perhaps your right - I guess what got to me was the fact the
>> respondent was able to take the time to correct a common error which
>> nearly everyone is guilty of, but failed to actually address the OPs
>> question in any way. I mean, is he going to correct every post that
>> refers to just Linux rather than GNU/Linux?
>
> In every emacs-* mailing lists, it has been extremely consistent to
> use Linux and GNU/Linux i.e. Linux means the kernel and GNU/Linux is
> the operating system that contains a Linux kernel.
>
> Making effort to preserve this consistence should be appreciated.

This is simply not true.

In nearly all but the kernel groups, "Linux" is known or accepted as Gnu/Linux.

Re: gnu vs. xemacs

[Ralf Angeli]

* Hadron Quark (2006-12-29) writes:

> Leo <sdl.web@gmail.com> writes:
>
>> In every emacs-* mailing lists, it has been extremely consistent to
>> use Linux and GNU/Linux i.e. Linux means the kernel and GNU/Linux is
>> the operating system that contains a Linux kernel.
>>
>> Making effort to preserve this consistence should be appreciated.
>
> This is simply not true.
>
> In nearly all but the kernel groups, "Linux" is known or accepted as Gnu/Linux.

Leo talked specifically about "emacs-* mailing lists" and those are
typically GNU lists.  And on such lists "Linux" is often _not_
accepted as "GNU/Linux".

-- 
Ralf

Re: "MIT/GNU/Linux"

[Kim F. Storm]

nospam@dev.null (Alexey Pustyntsev) writes:

> "Micha Feigin" <michf@post.tau.ac.il> writes:
>
>> The operating system is the kernel. GNU is one of the more popular option (but
>> not the only one) for the tools running over the operating system. I think
>> GNU/Linux is just RMS being overly pushy and and egomaniac.
>>
>
> RMS is neither pushy nor egomaniac. 'GNU/Linux' is absolutely correct
> and fair term as the GNU tools and the kernel are practically
> indispensable to get the OS running.

And in particular the GNU C compiler and libraries!

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

Re: "MIT/GNU/Linux"

[Jay Belanger]

"Micha Feigin" <michf@post.tau.ac.il> writes:
...
> What if you replace GNU tools? (there are alternative options to most if not
> all)

Then it wouldn't be GNU/Linux anymore, would it?  As far as this
thread goes, I'm not sure I see the point of this question.  Replacing
the GNU toolchain has been suggested before, but as far as I know
never done.  I suspect this would be difficult.

Jay

Re: "MIT/GNU/Linux"

[Tim X]

Giorgos Keramidas <keramida@ceid.upatras.gr> writes:

>
> First of all, there are other examples where "parts" have a different
> name from the "whole".  Consider for example the fine difference between
> "SunOS" and "The Solaris Operating Environment" :)
>

Can you expand on this point? I'm asking as this seems to contradict
what I was told by Sun and other sys admins and recall reading some
years ago. My understanding is that sunOS was what Sun called the
operating system they had prior to Solaris. When they brought out
Solaris, they faced a bit of industry resistance and released SunOS (I
can't remember, but think it might have been v4.5 or v5.4 or something
like that), which was essentially the same as solaris (v2.3?). If
there is a more precise difference, I'd be intrested in knowing it
and/or being corrected.

With respect to comments re kernel == OS, I don't agree. The kernel
and the operating system are two different things, but somewhat
dependent on each other. For example, you could run hurd instead of
the Linux kernel. 

Now for some final clarification - at what point does GNU/Linux become
something other than GNU/Linux? For example, I would expect a
distribution like Debian is certainly of the GNU/Linux variety.
However, what about Red Hat and SuSe? Possibly even more unclear, what
about the distribution Oracle is planning to release as the supported
platform for their Oracle databases, can you also call this GNU/Linux? 

If there are some distributions which are GNU/Linux and some which are
not, what is the generic name used to refer to all of them
collectively (i.e. in the sense of Unix or *nix)? In the old DOS days,
you referred to DOS when talking about the generic OS, and DR DOS, MS
DOS, PC MOS etc when referring to specific flavors. If we want to say
that something runs on all "Linux" based systems, what is the correct
terminology?

I also gather from listening to RMS and from some reading that we
should also avoid referring to GNU software as open source, but
instead as "Free Software". 

Tim

-- 
tcross (at) rapttech dot com dot au

Re: "MIT/GNU/Linux"

[Matthew Flaschen]

[-- Attachment #1.1: Type: text/plain, Size: 1249 bytes --]

Tim X wrote:
> 
> Now for some final clarification - at what point does GNU/Linux become
> something other than GNU/Linux? For example, I would expect a
> distribution like Debian is certainly of the GNU/Linux variety.
> However, what about Red Hat and SuSe? 

Of course.  They are using the same basic tools and software as Debian.

Possibly even more unclear, what
> about the distribution Oracle is planning to release as the supported
> platform for their Oracle databases, can you also call this GNU/Linux? 

Even more obvious.  Oracle's product is nothing but Red Hat Enterprise
Linux repackaged.

> If there are some distributions which are GNU/Linux and some which are
> not, what is the generic name used to refer to all of them
> collectively (i.e. in the sense of Unix or *nix)? 

Linux distributions, but you will not find a major distribution that
doesn't rely on GNU.  Only a few embedded systems use the kernel alone.

> I also gather from listening to RMS and from some reading that we
> should also avoid referring to GNU software as open source, but
> instead as "Free Software".

RMS started the movement and the GNU Project and that is the term he
chose to associate with both.

Matthew Flaschen


[-- Attachment #1.2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 252 bytes --]

[-- Attachment #2: Type: text/plain, Size: 152 bytes --]

_______________________________________________
help-gnu-emacs mailing list
help-gnu-emacs@gnu.org
http://lists.gnu.org/mailman/listinfo/help-gnu-emacs

Re: "MIT/GNU/Linux"

[Tim X]

Matthew Flaschen <matthew.flaschen@gatech.edu> writes:

> Tim X wrote:
>> 
>> Now for some final clarification - at what point does GNU/Linux become
>> something other than GNU/Linux? For example, I would expect a
>> distribution like Debian is certainly of the GNU/Linux variety.
>> However, what about Red Hat and SuSe? 
>
> Of course.  They are using the same basic tools and software as Debian.
>
> Possibly even more unclear, what
>> about the distribution Oracle is planning to release as the supported
>> platform for their Oracle databases, can you also call this GNU/Linux? 
>
> Even more obvious.  Oracle's product is nothing but Red Hat Enterprise
> Linux repackaged.
>

So then, should that be Red Hat Enterprise GNU Linux?

Also, when Larry initially announced that Oracle would do their own
distribution, while he indicated it would likely be based on RH
Enterprise, he was not prepared to commit to that. I was not aware he
(or Oracle) had yet made such a commitment.

Obviously, I wasn't clear enough. The point I wanted clarification on
is whether GNU Linux refers to Linux distributions which comprise of
only free software or whether it refers to all distributions which use
both the Linux kernel and GNU utilities and other free software
regardless of what other non-free software the distribution contains.

>> If there are some distributions which are GNU/Linux and some which are
>> not, what is the generic name used to refer to all of them
>> collectively (i.e. in the sense of Unix or *nix)? 
>
> Linux distributions, but you will not find a major distribution that
> doesn't rely on GNU.  Only a few embedded systems use the kernel alone.
>
>> I also gather from listening to RMS and from some reading that we
>> should also avoid referring to GNU software as open source, but
>> instead as "Free Software".
>
> RMS started the movement and the GNU Project and that is the term he
> chose to associate with both.
>

I think you missed the point. RMS does not like the use of the term
open source, but prefers free software. The problem is that open
source does not necessarily mean free (as in liberty) as you can have
software in which the sources are open, but the licensing is
restrictive and non-free. I personally agree with this distinction
unless I have misunderstood his arguement (which is possible and why I
mentioned it.). 

You might believe that this is all self evident, but I have been using
systems based on the Linux kernel since the first release of Slackware
and the terminology and how it is applied has not been a static thing.
The move away from the general term "Linux" to GNU Linux, while
positive in reducing confusion between the OS and kernel references,
was not emphasised initially as much as it seems to be now. Open
source was considered as synonymous with "free software" until Eric
Raymond and the OSI blurred things and an incresing number of
companies attempted to jump on the band wagon by releasing their
sources, but maintaining restrictive licenses etc. IIRC, even the
Linux kernel was not initially released under the GPL. 

Combine this with the fact few of the major distributions use the term
GNU Linux, preferring instead to either avoid explicit references to
Linux in the branding or more often, emphasising their company name,
the confusion in the press, deliberate FUD from SCO and MS etc, it
should not be any surprise that some find it unclear exactly when GNU
Linux is appropriate. 

Tim

-- 
tcross (at) rapttech dot com dot au

RE: "MIT/GNU/Linux"

[Micha Feigin]

 

> -----Original Message-----
> From: help-gnu-emacs-bounces+michf=post.tau.ac.il@gnu.org 
> [mailto:help-gnu-emacs-bounces+michf=post.tau.ac.il@gnu.org] 
> On Behalf Of Jay Belanger
> Sent: Saturday, December 30, 2006 1:13 AM
> To: help-gnu-emacs@gnu.org
> Cc: belanger@truman.edu
> Subject: Re: "MIT/GNU/Linux"
> 
> 
> "Micha Feigin" <michf@post.tau.ac.il> writes:
> ...
> > What if you replace GNU tools? (there are alternative 
> options to most 
> > if not
> > all)
> 
> Then it wouldn't be GNU/Linux anymore, would it?  As far as 
> this thread goes, I'm not sure I see the point of this 
> question.  Replacing the GNU toolchain has been suggested 
> before, but as far as I know never done.  I suspect this 
> would be difficult.
> 

Most if not all embeded systems do this ...

> Jay
> _______________________________________________
> help-gnu-emacs mailing list
> help-gnu-emacs@gnu.org
> http://lists.gnu.org/mailman/listinfo/help-gnu-emacs
> 
> 
> 
> +++++++++++++++++++++++++++++++++++++++++++
>  This Mail Was Scanned By Mail-seCure System  at the Tel-Aviv 
> University CC.
> 
> 

Re: "MIT/GNU/Linux"

[Matthew Flaschen]

[-- Attachment #1.1: Type: text/plain, Size: 3818 bytes --]

Tim X wrote:
> So then, should that be Red Hat Enterprise GNU Linux?

Yes, that would be more accurate (and fair).

> Also, when Larry initially announced that Oracle would do their own
> distribution, while he indicated it would likely be based on RH
> Enterprise, he was not prepared to commit to that. I was not aware he
> (or Oracle) had yet made such a commitment.

It is my understanding from reviews (like
http://ultramookie.com/wayback/2006/10/26/uncompatible-linux/) that it
is based on CentOS, which is in turn recompiled RHEL.
> 
> Obviously, I wasn't clear enough. The point I wanted clarification on
> is whether GNU Linux refers to Linux distributions which comprise of
> only free software or whether it refers to all distributions which use
> both the Linux kernel and GNU utilities and other free software
> regardless of what other non-free software the distribution contains.

Correct, any distribution that is based on both GNU and the Linux kernel
should be called GNU/Linux.  RMS would prefer GNU/Linux distributions
were completely free, but that is separate from the question of naming.

> I think you missed the point. RMS does not like the use of the term
> open source, but prefers free software. 

I'm well aware of that.

The problem is that open source does not necessarily mean free (as in
liberty) as you can have
> software in which the sources are open, but the licensing is
> restrictive and non-free. I personally agree with this distinction
> unless I have misunderstood his arguement (which is possible and why I
> mentioned it.).

This is mostly a misconception.  The reason that Stallman doesn't like
the open source movement is that it is based solely on practical
expediency, while free software is about morality.  I agree with him
here, and prefer "free software".  However, in practice, almost all open
source (as defined by the Open Source Initiative) licenses are also
free.  The Open Source Definition
(http://opensource.org/docs/definition.php) has detailed requirements,
including free redistribution (modified or unmodified) for free or for a
fee, and access to source.

> You might believe that this is all self evident, but I have been using
> systems based on the Linux kernel since the first release of Slackware
> and the terminology and how it is applied has not been a static thing.
> The move away from the general term "Linux" to GNU Linux, while
> positive in reducing confusion between the OS and kernel references,
> was not emphasised initially as much as it seems to be now.

Yes, the FSF says (http://www.gnu.org/gnu/gnu-linux-faq.html#always) "It
took a few years for us to realize what a problem this was and ask
people to correct the practice. By that time, the confusion had a big
head start."

> Open source was considered as synonymous with "free software" until Eric
> Raymond and the OSI blurred things

If I understand right, "open source" was not much used at all until the
OSI (and their founders) popularized it.

 and an incresing number of
> companies attempted to jump on the band wagon by releasing their
> sources, but maintaining restrictive licenses etc.

Again, almost every OSI-approved license is also free (as defined by the
FSF).  These restrictive licenses (e.g Microsoft Shared Source) aren't
OSI-approved open source either.

 IIRC, even the
> Linux kernel was not initially released under the GPL. 

No, it was originally under a non-commercial only license, which
wouldn't be OSI-approved open source OR Free.

> it should not be any surprise that some find it unclear exactly when GNU
> Linux is appropriate.

Again, all popular "Linux distributions" are actually GNU/Linux.

You may want to reply off-list.

Matthew Flaschen






[-- Attachment #1.2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 252 bytes --]

[-- Attachment #2: Type: text/plain, Size: 152 bytes --]

_______________________________________________
help-gnu-emacs mailing list
help-gnu-emacs@gnu.org
http://lists.gnu.org/mailman/listinfo/help-gnu-emacs

Re: "MIT/GNU/Linux"

[Robert Thorpe]

Micha Feigin wrote:
> > -----Original Message-----
> > From: help-gnu-emacs-bounces+michf=post.tau.ac.il@gnu.org
> > [mailto:help-gnu-emacs-bounces+michf=post.tau.ac.il@gnu.org]
> > On Behalf Of Jay Belanger
> > Sent: Saturday, December 30, 2006 1:13 AM
> > To: help-gnu-emacs@gnu.org
> > Cc: belanger@truman.edu
> > Subject: Re: "MIT/GNU/Linux"
> >
> >
> > "Micha Feigin" <michf@post.tau.ac.il> writes:
> > ...
> > > What if you replace GNU tools? (there are alternative
> > options to most
> > > if not
> > > all)
> >
> > Then it wouldn't be GNU/Linux anymore, would it?  As far as
> > this thread goes, I'm not sure I see the point of this
> > question.  Replacing the GNU toolchain has been suggested
> > before, but as far as I know never done.  I suspect this
> > would be difficult.
> >
>
> Most if not all embeded systems do this ...

Normally they don't replace it, they just do without most of it,
resulting in a system very unlike a *nix, but one that certainly
qualifies as "linux".

Re: "MIT/GNU/Linux" (was: gnu vs. xemacs)

[Robert Thorpe]

Micha Feigin wrote:
> > -----Original Message-----
> > From: help-gnu-emacs-bounces+michf=post.tau.ac.il@gnu.org
> > [mailto:help-gnu-emacs-bounces+michf=post.tau.ac.il@gnu.org]
> > On Behalf Of B. Smith-Mannschott
> > Sent: Friday, December 29, 2006 2:46 PM
> > To: help-gnu-emacs@gnu.org
> > Subject: "MIT/GNU/Linux" (was: gnu vs. xemacs)
> >
> >
> > On Dec 29, 2006, at 02:12, Leo wrote:
> > > In every emacs-* mailing lists, it has been extremely consistent to
> > > use Linux and GNU/Linux i.e. Linux means the kernel and
> > GNU/Linux is
> > > the operating system that contains a Linux kernel.
> >
>
> The operating system is the kernel.

No it's not.  This is an insidious idea started by Microsoft in the 80s
who wanted to ship DOS with virtually no utilities.  They tried to call
it an OS even though it was only a part of one.  Historically an OS has
meant  "the stuff that is needed to make a computer usable", not just
the kernel.  Even they have moved away from doing this, far away.

> > Actually, we should use "MIT/GNU/Linux", at least when
>
> What about qt etc ?
...
> > at all tempted to shorten a handy phrase like
> > "MIT/GNU/Linux". It just rolls of the tongue.
> >
> > I'm inclined to agree with the point made by the pedants of
> > "GNU/ Linux", but I do wonder: where does it end?

This is a problem.  The systems that are used today have been built
from software from many sources, large voluntary projects, private
individual hobbiests, and corporations.  The largest things in a distro
used to be, in terms of lines of code, Linux, GCC, X, Glibc and Emacs.
Now there are things like OO.org and Mozilla.  There is no really good
name anymore.

Perhaps these systems should be described as "Free/Open software
operating systems".

Re: "MIT/GNU/Linux" (was: gnu vs. xemacs)

[Matthew Flaschen]

[-- Attachment #1.1: Type: text/plain, Size: 891 bytes --]

Robert Thorpe wrote:
> Micha Feigin wrote:
> This is a problem.  The systems that are used today have been built
> from software from many sources, large voluntary projects, private
> individual hobbiests, and corporations.  The largest things in a distro
> used to be, in terms of lines of code, Linux, GCC, X, Glibc and Emacs.
> Now there are things like OO.org and Mozilla.  There is no really good
> name anymore.

Yes, but the center of the system is still GNU and Linux.  Moreover, the
 GNU Project is the impetus beyond the system's existence.

> Perhaps these systems should be described as "Free/Open software
> operating systems".

That description is weak, to say the least.  Half a dozen different
operating systems (GNU/Linux, NetBSD, OpenBSD, OpenSolaris, Darwin,
etc.) could be described like that.  GNU/Linux is much more descriptive.

Matthew Flaschen


[-- Attachment #1.2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 252 bytes --]

[-- Attachment #2: Type: text/plain, Size: 152 bytes --]

_______________________________________________
help-gnu-emacs mailing list
help-gnu-emacs@gnu.org
http://lists.gnu.org/mailman/listinfo/help-gnu-emacs

Re: "MIT/GNU/Linux"

[Tim X]

Matthew Flaschen <matthew.flaschen@gatech.edu> writes:

>
>> I think you missed the point. RMS does not like the use of the term
>> open source, but prefers free software. 
>
> I'm well aware of that.
>
> The problem is that open source does not necessarily mean free (as in
> liberty) as you can have
>> software in which the sources are open, but the licensing is
>> restrictive and non-free. I personally agree with this distinction
>> unless I have misunderstood his arguement (which is possible and why I
>> mentioned it.).
>
> This is mostly a misconception.  The reason that Stallman doesn't like
> the open source movement is that it is based solely on practical
> expediency, while free software is about morality.  I agree with him
> here, and prefer "free software".  However, in practice, almost all open
> source (as defined by the Open Source Initiative) licenses are also
> free.  The Open Source Definition
> (http://opensource.org/docs/definition.php) has detailed requirements,
> including free redistribution (modified or unmodified) for free or for a
> fee, and access to source.
>
Maybe we are getting down to "hair splitting" a little too much.
However, in an interview I heard with RMS on a podcast, he was very
critical of OSI and the problem with the use of the term open source.
Your quite correct regarding is arguments concerning expediency and
the danger this can represent for general freedom, especially in the
long term. However, he was very very clear about not using the term
open source and I have had e-mails from him correcting me for
referring to GPL's software as open source (this was software released
under the GPL, but not GNU software). AFter a few exchanges and after
listening to that interview, I think I understand exactly what his
concerns are and I have to say I agree. In the interview, the
arguement centred around provision of MS media codecs - one side
argued that it should be easier for people to obtain and install these
codecs as this would increase the number of people switching to GNU
Linux. On the other side, RMS argued that the cost to freedom this
would result in was too high and he would rather see a truely free
system used by less people than pseudo free OS used by a lot. Being a
bit of an old idealist and readily admitting to finding Marxist
theory, while flawed, better than any other, I could appreciate this
point.

>> You might believe that this is all self evident, but I have been using
>> systems based on the Linux kernel since the first release of Slackware
>> and the terminology and how it is applied has not been a static thing.
>> The move away from the general term "Linux" to GNU Linux, while
>> positive in reducing confusion between the OS and kernel references,
>> was not emphasised initially as much as it seems to be now.
>
> Yes, the FSF says (http://www.gnu.org/gnu/gnu-linux-faq.html#always) "It
> took a few years for us to realize what a problem this was and ask
> people to correct the practice. By that time, the confusion had a big
> head start."
>
>> Open source was considered as synonymous with "free software" until Eric
>> Raymond and the OSI blurred things
>
> If I understand right, "open source" was not much used at all until the
> OSI (and their founders) popularized it.

Hmm, not sure. The OSI wold seem a bit of a late arrival in my
chronology. The CS department at the University I worked at in '94 was
already starting to make a bit of a 'big deal' about the fact they
were moving to open source then and even started using the term in
their publicity in '94/95. All PCs in the CS labs were running
slackware at that time (but to be honest, a fair part of this was
because MS Win 31 and DOS were just crippled pretend OSs at that time
and even early, limited and buggy GNU Linux was better). 

Tim

-- 
tcross (at) rapttech dot com dot au

Re: "MIT/GNU/Linux"

[Giorgos Keramidas]

On Sat, 30 Dec 2006 11:28:50 +1100, Tim X <timx@nospam.dev.null> wrote:
>Giorgos Keramidas <keramida@ceid.upatras.gr> writes:
>> First of all, there are other examples where "parts" have a different
>> name from the "whole".  Consider for example the fine difference
>> between "SunOS" and "The Solaris Operating Environment" :)
>
> Can you expand on this point?

"Solaris" is considered to be the "SunOS" operating system, plus a
graphical user environment, and a few other components.

> I'm asking as this seems to contradict what I was told by Sun and
> other sys admins and recall reading some years ago. My understanding
> is that sunOS was what Sun called the operating system they had prior
> to Solaris.

"SunOS" 4.X was BSD-based.  Sun replaced the BSD-based core of the
system with a System V derivative, creating SunOS release 5.0.  At the
same time, a new marketing name was introduced for SunOS 5.0 and it
accompanying set of components.  This name was "Solaris 2".

> When they brought out Solaris, they faced a bit of industry resistance
> and released SunOS (I can't remember, but think it might have been
> v4.5 or v5.4 or something like that), which was essentially the same
> as solaris (v2.3?).

There is a nice table of Solaris vs. SunOS versions here:

    http://en.wikipedia.org/wiki/Solaris_Operating_System

I can't verify the correctness of *all* these release versions, but it
may help a bit.

The *real* point, however, is that there is ample precedent for calling
parts of a system with a codename that is different from the whole.

> With respect to comments re kernel == OS, I don't agree. The kernel
> and the operating system are two different things, but somewhat
> dependent on each other. For example, you could run hurd instead of
> the Linux kernel.

Exactly :)

> I also gather from listening to RMS and from some reading that we
> should also avoid referring to GNU software as open source, but
> instead as "Free Software".

Also true.

Re: "MIT/GNU/Linux"

[Giorgos Keramidas]

On Sun, 31 Dec 2006 13:36:52 +1100, Tim X <timx@nospam.dev.null> wrote:
> The CS department at the University I worked at in '94 was already
> starting to make a bit of a 'big deal' about the fact they were moving
> to open source then and even started using the term in their publicity
> in '94/95. All PCs in the CS labs were running slackware at that time
> (but to be honest, a fair part of this was because MS Win 31 and DOS
> were just crippled pretend OSs at that time and even early, limited
> and buggy GNU Linux was better).

Heheheh!  I know the feeling :)

Re: "MIT/GNU/Linux"

[Tim X]

Giorgos Keramidas <keramida@ceid.upatras.gr> writes:

> On Sat, 30 Dec 2006 11:28:50 +1100, Tim X <timx@nospam.dev.null> wrote:
>>Giorgos Keramidas <keramida@ceid.upatras.gr> writes:
>>> First of all, there are other examples where "parts" have a different
>>> name from the "whole".  Consider for example the fine difference
>>> between "SunOS" and "The Solaris Operating Environment" :)
>>
>> Can you expand on this point?
>
> "Solaris" is considered to be the "SunOS" operating system, plus a
> graphical user environment, and a few other components.
>
>> I'm asking as this seems to contradict what I was told by Sun and
>> other sys admins and recall reading some years ago. My understanding
>> is that sunOS was what Sun called the operating system they had prior
>> to Solaris.
>
> "SunOS" 4.X was BSD-based.  Sun replaced the BSD-based core of the
> system with a System V derivative, creating SunOS release 5.0.  At the
> same time, a new marketing name was introduced for SunOS 5.0 and it
> accompanying set of components.  This name was "Solaris 2".
>
>> When they brought out Solaris, they faced a bit of industry resistance
>> and released SunOS (I can't remember, but think it might have been
>> v4.5 or v5.4 or something like that), which was essentially the same
>> as solaris (v2.3?).
>
> There is a nice table of Solaris vs. SunOS versions here:
>
>     http://en.wikipedia.org/wiki/Solaris_Operating_System
>
> I can't verify the correctness of *all* these release versions, but it
> may help a bit.
>

Thanks. Its always worth getting clarification on this sort of stuff.
I didn't have to administer a solaris system for very long and I
didn't like it very much at the time (I was more experienced with
OSF/Digital Unix). This was solaris 2.4 I think and it had a few
issues - the sendmail bundled with it was buggy, there were problems
with defunct processes hanging around forever and a few other
irritating but not critical problems. 

What was really interesting was I replaced solaris with one of the
early Linux sparc ports and was very amazed at the performance/speed
improvement we got. I had thought the sparc hardware was over priced
and slow, but once we had GNU linux on it, it was merely over priced.
It actually gave us a really solid and reliable server for quite some
time. The people I was working for were really amazed at the uptime we
got and the performance. I was happy because I was able to install
emacs on it (I had problems getting emacs installed under solaris, I
can't remember what the exact problem was, but I had to tweak the
build to get it to compile and then it segfaulted quite often). I
think it did have xemacs installed, but one of the key elisp packages
I needed wouldn't run under xemacs, so it was no good to me. 

This is why Ive stuck with GNU Emacs rather than Xemacs - some of the
packages I use the most just won't run under Xemacs because of font
lock differences.

Tim

-- 
tcross (at) rapttech dot com dot au

Re: "MIT/GNU/Linux"

[Tim X]

Giorgos Keramidas <keramida@ceid.upatras.gr> writes:

> On Sun, 31 Dec 2006 13:36:52 +1100, Tim X <timx@nospam.dev.null> wrote:
>> The CS department at the University I worked at in '94 was already
>> starting to make a bit of a 'big deal' about the fact they were moving
>> to open source then and even started using the term in their publicity
>> in '94/95. All PCs in the CS labs were running slackware at that time
>> (but to be honest, a fair part of this was because MS Win 31 and DOS
>> were just crippled pretend OSs at that time and even early, limited
>> and buggy GNU Linux was better).
>
> Heheheh!  I know the feeling :)
>

The really sad thing is that the department, due to mainly economic
pressure, has moved from core comp. sci. courses to more of what I
would call IT courses. All the academics are now windows users and
from talking with the sys admin who is still there, instead of getting
really interesting questions or projects, such as setting up a beowulf
cluster or helping to improve new signal processing algorithms or
hardware designs etc, the questions/problems are now basic end-user
support problems, like how do I configure outlook or why won't my
anti-virus software run etc. Where once upon a time there was a number
of what would be called emacs "power users" who would often be
swapping bits of elisp and writing custom elisp packages to use in the
department, its now just office documents, abstract misguided
ramblings about DRM, ebusiness and enterprise management!

Tim

-- 
tcross (at) rapttech dot com dot au

Re: "MIT/GNU/Linux"

[Stefan Monnier]

> So lemme see...
> (1) "Linux" is just kernel
> (2) GNU/Linux is the operating system formed by running the GNU tools  on
> top of that kernel

I believe RMS's use of "GNU" in "GNU/Linux" does not really refer to the
so-called GNU tools, but the GNU project which endeavored, long before it
seemed even remotely possible, to build a 100% Free Software operating
system.  To this end, it built tools like gcc, it designed licenses such as
the GPL, it encouraged other people to release their code under a Free
license, and it kind of brought some sense and long term vision to all the
"free hacking" that was going on piecemeal around the world.


        Stefan

RE: "MIT/GNU/Linux"

[Drew Adams]

> I believe RMS's use of "GNU" in "GNU/Linux" does not really refer to the
> so-called GNU tools, but the GNU project which endeavored, long before it
> seemed even remotely possible, to build a 100% Free Software operating
> system.  To this end, it built tools like gcc, it designed
> licenses such as
> the GPL, it encouraged other people to release their code under a Free
> license, and it kind of brought some sense and long term vision to all the
> "free hacking" that was going on piecemeal around the world.

Well put. The "GNU = surrounding tools or environment" characterization is
too often too prevalent in these discussions. You put it right, IMO.

That said, getting people to refer to the OS as "GNU/Linux" might be a
losing, though worthwhile, battle, if only because people do take shortcuts
in speech. If the OS name were only one or two syllables, the battle would
be easier. If "GNU" didn't mean more than just an OS, one might try to push
to use just "GNU" instead of "GNU/Linux" - easier to say. If it were used
alone as the name for the OS, however, "GNU" would be more general than
"GNU/Linux", since, as someone pointed out, a different GNU kernel (e.g.
HURD) could be used. GNU/Linux is no doubt the right term for a GNU OS with
a Linux kernel, but it doesn't exactly roll off the tongue.

(I really wanted just to say "attaboy" for what you wrote.)

Re: "MIT/GNU/Linux"

[Matthew Flaschen]

[-- Attachment #1.1: Type: text/plain, Size: 429 bytes --]

Stefan Monnier wrote:
> I believe RMS's use of "GNU" in "GNU/Linux" does not really refer to the
> so-called GNU tools, but the GNU project which endeavored, long before it
> seemed even remotely possible, to build a 100% Free Software operating
> system.

That's true.  However, you should also consider the vastness of how much
 code the GNU Project has produced.  Everything from chown to GNOME.

Matthew Flaschen


[-- Attachment #1.2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 252 bytes --]

[-- Attachment #2: Type: text/plain, Size: 152 bytes --]

_______________________________________________
help-gnu-emacs mailing list
help-gnu-emacs@gnu.org
http://lists.gnu.org/mailman/listinfo/help-gnu-emacs

Re: "MIT/GNU/Linux"

[Stefan Monnier]

>> I believe RMS's use of "GNU" in "GNU/Linux" does not really refer to the
>> so-called GNU tools, but the GNU project which endeavored, long before it
>> seemed even remotely possible, to build a 100% Free Software operating
>> system.

> That's true.  However, you should also consider the vastness of how much
> code the GNU Project has produced.  Everything from chown to GNOME.

What I meant above is that a vast amount of non-GNU code is Free Software
thanks to the GNU project, so even if "the GNU project"[1] did not
contribute any code to it, its existence as a Free Software is partly due to
the GNU project.

In any case, to a large extent, the only difference between "GNU programs"
and "non-GNU programs" is the extent to which the maintainers agreed to
associate their project's name with the GNU moniker, which is mostly
a political decision.


        Stefan


[1] "The GNU project" is much more a concept than a bunch of people.

Re: "MIT/GNU/Linux"

[Matthew Flaschen]

[-- Attachment #1.1: Type: text/plain, Size: 681 bytes --]

Stefan Monnier wrote:
> What I meant above is that a vast amount of non-GNU code is Free Software
> thanks to the GNU project, so even if "the GNU project"[1] did not
> contribute any code to it, its existence as a Free Software is partly due to
> the GNU project.

Also true.  If nothing else, they created the licenses that is used most
commonly for both GNU/Linux and for free software generally.

> [1] "The GNU project" is much more a concept than a bunch of people.

That's true.  However, it's that concept (and those who implemented it,
however connected) that created the largest part of the OS.

I think we essentially agree here. :)

Matthew Flaschen


[-- Attachment #1.2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 252 bytes --]

[-- Attachment #2: Type: text/plain, Size: 152 bytes --]

_______________________________________________
help-gnu-emacs mailing list
help-gnu-emacs@gnu.org
http://lists.gnu.org/mailman/listinfo/help-gnu-emacs