Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 30 Dec 2011 12:20:47 -0500 (EST)
From:      doug@safeport.com
To:        Polytropon <freebsd@edvax.de>
Cc:        David Jackson <djackson452@gmail.com>, freebsd-questions@freebsd.org, Robert Bonomi <bonomi@mail.r-bonomi.com>
Subject:   Re: FreeBSD Kernel Internals Documentation
Message-ID:  <alpine.BSF.2.00.1112301204560.15394@fledge.watson.org>
In-Reply-To: <20111230170339.6d6af931.freebsd@edvax.de>
References:  <CAGy-%2Bi-NN_SOYrrE6WgHyCBa5VzFexwT_C9UYhO3GyjvfsxpAA@mail.gmail.com> <201112300604.pBU64dqB069626@mail.r-bonomi.com> <CAGy-%2Bi-m7hUfrUPKaG6FAiv8cor6%2BWXUW-OwWA=uOAt0yDToag@mail.gmail.com> <20111230170339.6d6af931.freebsd@edvax.de>

next in thread | previous in thread | raw e-mail | index | archive | help
On Fri, 30 Dec 2011, Polytropon wrote:

> On Fri, 30 Dec 2011 10:22:31 -0500, David Jackson wrote:
>> Of course, those best able to document are those who wrote it in
>> the first place, since they already know how it works.
>
> A fact seems to be: "Modern" programmers don't bother
> with documenting, or coding guidelines, or style or
> other things that "slow down" development. This
> attitude isn't new in general, as it has been done
> that way even in IT dinosaur times: There are COBOL
> programs still running, and nobody knows _why_ they
> are running and _how_. If someone had written usable
> documentation at the time the program was created
> and maintained, skilled COBOL h4x0rs wouldn't be
> able to write the desired salary on the contract
> as _they_ wish. :-)

As a [former] mainframer, I might take issue limiting the above to modern. I 
learned MVT and then MVS from the microfish and crashing rather larger 
(physically) and expensive computers. Not much else was available [outside of 
IBM].

> But keep in mind: Writing code and writing documentation
> are two different things. There are people who are
> excellent coders, but bad writers. In some teams,
> you'll find code writers and doc writers separated,
> but working together. This approach isn't free of
> problems, but also seems to work.
>
In this thread if anyone mentioned Robert Watson, kernel source cross-reference, 
I missed it. Also every so often Kirk McKusick teaches a course on BSD 
internals. Kind of expensive but very thorough.



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?alpine.BSF.2.00.1112301204560.15394>