Re: Done

[Manoj Srivastava <srivasta@debian.org>]

On Sun, 14 Sep 2003 14:53:53 -0500, Steve Langasek <vorlon@netexpress.net> said: 

> If you are able to recognize that a given package description is
> inadequate, you are also capable of discerning *what* is wrong with
> it, even if you don't know what the package does; and you are

	Umm. 

Package: blacs-lam-test
Description: Basic Linear Algebra Communications Subprograms
 This package provides programs to test your BLACS libraries.

	Let us see. Whats is a Basic Linear Algebra Communications
 Subprogram? What kind of tests are provided? Why should I want to
 install this package?

	I recognize this desciption is inadequate, mostly because I
 have no idea what the package in question does.  What are Linear
 Algebra Communications?  Are there communications programs, as
 opposed to subprograms? Heck, I don't even know enough to ask
 intelligent questions. 

> therefore very much qualified to contribute comments to individual
> package description bugs in the BTS, which is likely the most time
> consuming part of the exercise of getting these bugs addressed, and

	Hmm. I do not think I have the information to create a
 description, sorry.  Of course, I can reasearch the program, and
 probably do the maintainers job for him, but I do not have time to do
 this kind of research and work for a package that supposedlyu has a
 full time maintainer devoted to the package. 

> which does not require familiarity with the package.  Contrary to
> Manoj's sneering, there is very good reason to believe that those
> most qualified to maintain a given package in other respects are
> also least likely to be able to identify those traits that render
> their package descriptions problematic for users unfamiliar with the
> package.  So if you want good package descriptions, don't complain
> -- help.

	You thought I was merely sneering? Lord, I must have been the
 model of restraint.

	This is not about people not turning out perfect description
 for people unfamiliar with the package; I understand how one can be
 too close to a package to figure out exactly what areas need further
 explanation. But if you can't expand the *LONG* description beyond a
 measly 80ncharacters, you are not even trying.

	All this red herring of an excuse that "we are too close to a
 package to ber the best person to document it" is just that -- a red
 herring.  The issue at hand is nto a the optimal description -- it is
 a a description that tries to explain what a package does, why it
 could be useful, and how to distinguich it from any competition, and
 whether one need install it directly -- all in under 80 chars.

	The complain is not asking the developoer to turn out a
 description that satisfies all kinds of novice users -- it is to try
 and provide a Long description that even attempts to address some of
 these questions.

	No matter how close I am to a package (and I have been ribbed
 at length for the ucf man page) -- I should be able to provide a
 multi sentence paragraph describing my package.

	And, BTW, you did not see me whining that people who
 complained about the ucf man page provide me with a working patch to
 improve it -- or worse, yelled at them for daring to complain without
 providing clues, helps, suggestions, and patches. I took a poll, and
 fixed the man page ti the best of my ability.

	manoj
-- 
A good memory does not equal pale ink.
Manoj Srivastava   <srivasta@debian.org>  <http://www.debian.org/%7Esrivasta/>
1024R/C7261095 print CB D9 F4 12 68 07 E4 05  CC 2D 27 12 1D F5 E8 6E
1024D/BF24424C print 4966 F272 D093 B493 410B  924B 21BA DABB BF24 424C