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

Re: changes and standards documents



Hi,
>>"Raul" == Raul Miller <rdm@test.legislate.com> writes:

 Raul> Manoj Srivastava <srivasta@datasync.com> wrote:
 Marcus> a) Without documentation, you can't use the software.
 >> 
 >> Does not apply to a standard. You use the standard by reading
 >> it -- nothing has to be modified. A standard is not documentation for
 >> a program. 

 Raul> "This program implements the ___ standard."

	So? The program should still come with usage and configuration
 documentation. Even then, the standard does not describe th e program
 -- if the program does not do what the standard says, the program is
 buggy, and should be changed -- not the standard. There is a big
 difference between documentation for a program that may need to be
 modified to match the behaviour, and a standard that should not be
 changed. 	

 Raul> So, I don't like the way the program behaves... is it buggy or have
 Raul> I misunderstood what the program is supposed to do?

	If you do not merely like how the program behaves, that does
 not make it buggy. I don't like how ask wqorks. Big deal. 

	If the program does not follow what the stadard says, the
 program is buggy. The standard is always authoritative (unless some
 one has monkeyed with the standard, in which case it is no loger a
 standard). 

	The exception is when the program is a reference
 implementation. Thereference implementation should probably come with
 its own documentation talking about usage and configuration, if at
 all applicable.

	If you modify a reference implementation, it is no longer a
 reference implementation. And, if your modified copy behaves any
 different fom the reference, your program is buggy as far as the
 standard is concerned.

 Raul> For example, consider /bin/sh: as it stands right now we're not
 Raul> even free to distribute the standard, and we're certainly not free
 Raul> to derive /bin/sh's documentation from a standard document.

	I fail to see what this has to do with whether standards
 should be mutable in order to get into main. 


	manoj
-- 
 Always verify your witchcraft. --anonymous
Manoj Srivastava  <srivasta@acm.org> <http://www.datasync.com/%7Esrivasta/>
Key C7261095 fingerprint = CB D9 F4 12 68 07 E4 05  CC 2D 27 12 1D F5 E8 6E


Reply to: