[linux-elitists] advocacy ad hoc

Heather star@betelgeuse.starshine.org
Mon Sep 11 16:32:18 PDT 2000


> Actually, I think that's a bit of overgeneralizing. There are people,
> particularly those in larger organisations, where the formal process is
> more important than the work. And there's the rest of us.
 
Well, yes... in a sense, all analogies are either over-generalizing, 
applying one paradigm to an invalid context, or both.  Sometimes it works;
sometimes it merely inspires;  others, it does neither.

> I personally think a lot of people missed the point in some books where it
> says that the documentation is supposed to only be 1/n of the duration of
> the project. If it's a six-week project, two weeks is the *maximum* you
> can allocate to design time. 1 week is better, especially if in 2 stages.
> :)
 
And to take things to a greater sense of reality, if something was designed
for the use of one person or a family or dorm, with a total project horizon
of your described 6 weeks, then even this "good planning percentage" won't
be worth much when the code is expanded and re-purposed.

A good rate of introducing more design time, and the means of seeing that 
to be done, would probably make a good article somewhere, but I'm sure not 
the right person to do it.  

The open source world fakes this by getting together at conventions, for
beers and dinner runs, and on IRC.  So it can be awful hard to tell when 
we're goofing off, when we're designing, and then, there's designing cool
toys to goof off with.

> > Lots of play with these analogies is possible, including the questions
> > of mistranslation or misapplications of Good Book practices to result
> > in "holy" code but unusable products, and knowing that just because we
> > can get faster mutations to the better does not mean that all
> > "evolved" products are any good, either.
> 
> Ask me sometime about well-designed and documented vaporware.

Heh, they got the 1/N ratio wrong in the other direction, it's not supposed
to be 99% doco, right?

-* Heather Stern * I have yet to see any problem, however complicated, 
                   which, when you looked at it in the right way, did 
                   not become still more complicated.  -- Poul Anderson




More information about the linux-elitists mailing list