[MUD-Dev] Re: An Introduction
Caliban Tiresias Darklock
caliban at darklock.com
Thu Jul 23 18:05:30 CEST 1998
On 05:09 PM 7/23/98 -0700, I personally witnessed J C Lawrence jumping up
to say:
>On Wed, 08 Jul 1998 06:44:01 +0200
>Asmodai <Jeroen> wrote:
>
>> Also, which books are recommended to start with network programming,
>> preferably TCP/IP based and for C/C++.
>
>I like the Steven's TCP/IP books, but also hear good things about
>Comer's (which I haven't looked at in detail).
In the defense industry, many of my colleagues believed you could spot the
real professional network programmers by which of these series they had on
the shelf. The competent professional had Comer. The expert professional
had Stevens. (The impostor had Stevens on the shelf and Comer in a drawer;
you could spot an impostor by opening his Stevens volume to see if the
binding cracked. Impostors are also known to subscribe to technical
magazines they don't read just so they can leave them lying around on their
desk.) In addition, the more worn and tattered the book was, the harder the
developer worked. Ideally, you would hire a developer who had two copies of
Stevens; one falling apart and unreadable, the other well-worn. That
generally indicated someone who not only worked his tail off, but was
emotionally attached to his first copy of Stevens, which marked him as a
true network hacker.
Use of the male pronoun above is not to be construed as restrictive, and I
make no claims as to the validity of the generalisation detailed above.
>From my own perspective, Stevens' books are based strongly on theory and
real understanding of the protocols and architecture, while Comer's are
more pegged toward real application and development. I would venture to say
that if you want to be writing code fast, go with Comer, but if you want to
really understand what you're doing (and you can handle the deep magic) go
with Stevens.
Definitely get Stevens at some point, because while you'll eventually
outgrow Comer... no one outgrows Stevens. "TCP/IP Illustrated" is one of
those series everyone OUGHT to have, but not enough people do. I consider
him about as essential as Knuth.
>On the C/C++ side it really depends on where you are coming from. I
>start with K&R for C and went straight to Strousup for C++ (not a
>generally recommended route), so can't comment terribly well there --
>tho I do like Pohl's C++ books.
I've been highly impressed with Jesse Liberty ("Teach Yourself C++ in 21
Days", Sams press). He's got a knack for taking complex ideas and
presenting them in easy-to-understand ways, and if you read his more
advanced books it becomes very evident that this guy isn't just some
halfwit programming wannabe that thought he'd write a book (which is the
sort of impression I get from Kris Jamsa, for example)... he's a serious
developer. "Clouds to Code" was particularly inspirational, and virtually
peerless in terms of subject matter.
More information about the mud-dev-archive
mailing list