[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: It's Huntin' Season



On Wed, Feb 06, 2002 at 10:53:31PM -0600, Manoj Srivastava wrote:

> 	We are not waffling abpout this. We are talking about a very
>  distinct set of files: those that live in /etc/emacs*/site-start.d/
>  There is absolutely no ambiguity at all.
> 
>  Nick> Debian policy distinguishes (or tries to) between configuration
>  Nick> and code.
> 
> 	Where? 

By saying that configuration files go in /etc, and code doesn't, effectively.

>  Nick> Take it to the (ridiculous) limit and you put *everything* in
>  Nick> /etc, as "you might want to run a hex editor over libc"...
> 
> 	You are the one taking things to such extremes. Not the other
>  participants in this discussion so far.

I'm just pointing out that it is not cut-and-dried, and that the section
of policy that you quoted, if it is going to be taken absolutely literally
(specifically the wording "A file that affects the operation of a program"),
is ridiculous.

I conclude that it is not intended to be taken absolutley literally, and
therefore that things are not as cut-and-dried, and "clear", as you seem
to be implying.

> 	If it modifies program behaviour, it is a configuration file,
>  and goes in /etc. If it does not modify program behaviour, why is it
>  there in the first place?

No; that's the point - *lots* of files that quite definitely do "affect
the operation of a program" quite definitely *don't* live in /etc.

Since emacs-lisp is "just another interpreted language", there is no better
reason (if we're reading policy literally) why they should go in /etc than
why a bunch of, say, shell-script libraries should go in /etc. All may
affect the startup behaviour of a particular program.

> 	Rubbish. This is where we do development and design work. We
>  do not tell people to go away and not bother us with the hard
>  technical stuff,

I'm trying to say that the solution is that in this case it's up to the
emacs maintainers, and that having made that decision, the work of this
list would be done. Since you don't agree, I'm trying to persuade you that
I'm right (or get someone to point out how I'm not). That's the "hard"
bit.

> 	If you can't exercise a modicum of common sense, there is no
>  point in wasting the time of the list. Why the hell bring up
>  executable in the middle of a policy discussion of a specific set of
>  non binary text files that modify the behaviour of a program? 

Because there is no real difference. And once you accept that, it becomes
clear that it is the intention of the author and maintainers that should
determine whether particular files are config files, and hence whether or
not they should go in /etc.

> 	Are you just trying to be contrary, and enusre no work gets
>  done in this thread?

No, I'm trying to break what appears to me to be an assumption that you
are making to support your argument - that because a startup file makes
emacs do something, that makes it a configuration file.

Apologies if I come/came across confrontational; It's not meant that way.
-- 
Nick Phillips -- nwp@lemon-computing.com
Truth will out this morning.  (Which may really mess things up.)



Reply to: