It was thus said that the Great Liam Proven once stated:
When I was a baby geek, I remember reading about this seminal text, a
distillation from a long-active mailing list called UNIX-HATERS. Now,
it's available for free:
Darn. And to think I bought the dead tree version in the mid 90s (I still
have the barf bag that came with it).
It's a good & enjoyable read. I'm nearly
at the end of it now.
It's interesting to look back at this 1991 (-ish) book from the
perspective of 2008. How many of the criticisms levelled against Unix
were stuff that users of then-older OSs thought was deranged.
Today, the same sort of rivalry exists between Unix and Windows
people; the stuff before them is nearly forgotten now. I mean, I've
been in this business for some 20y (and another 5-10y before that as a
hobbyist) and I've never seen TOPS or MULTICS or ITS or anything like
them.
I've certainly heard of them myself, and some of the concepts they came up
with have made me want to learn more about each one, but I've yet to
actually use any of them.
What I'm wondering is, how many of the criticisms
levelled against
Unix (and thus, by association, Linux) in this book from 17y ago are
still current or valid today. I've been using Linux for 11-12y now,
but I still regard myself as something of a beginner, whereas I've
known Windows since it was 2 and can make it jump backwards through
flaming hoops.
Chapter 1---not so much any more. This was a time (94) when the ANSI-C
standard had only been out a few years so there was still quite a bit of K&R
C floating around. This was also a time when there were still plenty of
Unix versions floating around, instead of the what? Five we have left now?
(Linux, Open|Free|NetBSD, Solaris).
Chapter 2---mostly user expectations here. And yes, it's still somewhat
valid---VMS was a lot more consistent in the command line than Unix ever
was.
Chpater 3---Eh. They go on about how horrible man pages are, but they're
infinitely better than the whole info crap you get nowadays with GNU (god, I
love man pages that say crap like "The full documentation for cut is
maintained as a Texinfo manual. If the info and cut programs are properly
installed at your site, the command info coreutils cut should give you
access to the complete manual"). Nowadays you can probably paste any
command or error message into Google and find out the problem (but then
again, you might find plenty of other people with said problem but no
definitive answer).
Chapter 4---mail. Or more specifically sendmail. Still relevent, and
sadly, the situations presented there are infinitely easier to deal with
than the crap I have to deal with today.
Chapter 5---Usenet. Strictly speaking, not specifically Unix, and
frankly, nowadays, not really relevent all that much (which means, it's
probably usuable these days). Replace USENET with
Slashdot|Reddit|Digg|etc.
Chapter 6---terminals. Not so much anymore. Pretty much vt100/xterm and
you're good to go (alhough I still have issues with BS vs. DEL on Linux).
Chapter 7---Pretty much spot on, although the alternatives to X Windows
are worse (that is, if you want remote capabilities). And sadly, I do miss
NeWS (used it in college; was sad when SGI dropped it).
Chapter 8---who uses csh anymore? I think we all use bash nowadays, but
this still holds up pretty well (man, I can't make heads or tails of the
startup scripts on Linux, but then again, I never did learn to really read
shell scripts).
Chapter 9---Not so much anymore, as more and more people use languages
other than C and make (and there are a few alternative to make nowadays).
Chapter 10---Dead on. (I don't like C++ 8-P
Chapter 11---If anything it's gotten better and worse in my opinion. If
you set up Unix correctly, it can run smoothly for years without problems.
The major problem comes when you have other users on the system, or trying
to get a modern Linux distribution set up correctly (when did "which" become
optional? Or "traceroute"? Don't even get me started on so-called
"package
managers").
Chapter 12---security. Sadly, this book was written before the Internet
became as popular as it did and Microsoft showed us just how bad security
could get. Or PHP showed us just how bad security could get (really! You
can embed PHP in gif files and PHP will gladly execute it! How ...
Microsoftian of it ... ).
Chapter 13---file systems. They're gotten much better over the years, and
even Linux is slowly learning about removable media [1].
Chapter 14---I don't know anyone using NFS anymore (I think the last time
I saw NFS in a commercial setting was the late 90s, and even at home, I
don't use NFS all that much). But replace NFS with Samba, and it's spot on
(more or less).
-spc (Been using Unix since 1989 ... )
[1] My own personal theory about Unix: If it wasn't in Unix V7, then
it's never been supported all that well under Unix. That explains
the crappy support for removable devices, graphical systems and
threads.