[svlug] -who option

Rob Landley rob at landley.net
Sat Mar 2 16:34:35 PST 2019


On 3/2/19 5:31 PM, Mehma Sarja wrote:
>> And so on. (Computer history is a hobby of mine, although
>> https://landley.net/history/mirror is sadly out of date...)
> 
> That’s part of my point. The information is scattered. Although if
> the community does not passionately back the -who idea, it will become
> just another place people stash info creating more scatter.

That would be your best case outcome, yes.

The cat man page says:

AUTHOR
       Written by Torbjorn Granlund and Richard M. Stallman.

But you don't seem to be aware of that?

Gnu's cat was 833 lines of C code, probably a _lot_ of contributors over the
years. Busybox had a cat, which Erik Anderson wrote, because 833 lines was
ridiculous and it could be much smaller/simpler. Toybox has a cat, which I wrote
from scratch when I left busybox (which is its own story having to do with
https://lwn.net/Articles/202106/ and later https://lwn.net/Articles/478308/
which became https://lwn.net/Articles/629362/ which  means my version is
probably _the_ most widely deployed one in the world in a way you didn't know
and doesn't necessarily matter). BSD had its own (for legal reasons couldn't use
AT&T's), Coherent had its own (that was the first clone of Unix that Dennis
Ritchie personally confirmed contained no AT&T code and thus they couldn't stop
it from being sold.) Linux forked of of a project called Minix which Andrew
Tanenbaum wrote after Apple sued the makers of the speak-and-spell and thus
changed the Unix license to something he couldn't teach with anymore, so he came
up with his own clone from scratch so yes Andrew wrote yet another cat
implementation... And none of these were the one from Unix version 7, which is
not the same as the one in Unix System V. (Bell Labs vs AT&T corporate; the Bell
Labs guys went on up through Unix version 10 but it never got released outside
of the labs, and then they started over with Plan 9 which would have been a MUCH
bigger deal if AT&T's lawyers hadn't also kept that bottled up with stupid IP
claims for decades...)

Also, the industry used BSD en masse (because the IMP switchover in 1979
promulgated vax unix through the entire internet as the router of choice, ala
https://www.salon.com/2000/05/16/chapter_2_part_one/) and then AT&T went through
with lawyers and convinced them all to switch over from BSD-based systems to
System V based systems in the late 80's (hence SunOS->Solaris and IBM AOS->AIX
and so on, this was covered in Robert Young's book "under the radar" but there's
a bunch of other sources...) so the "System-V like" axis vs the "BSD-like" axis
were established then (which you need to understand to explain the references in
https://www.oreilly.com/openbook/opensources/book/appa.html or in Linus's book
"Just for Fun" or http://www.landley.net/history/mirror/unix/srcos.html or... oh
good grief WHY is https://en.wikipedia.org/wiki/Larry_McVoy linking to my mirror
of that page, did the original I linked from http://landley.net/history/mirror/
go down? Yes, it did. Sigh, this is why I mirror stuff...)

Anyway, which portion of this history should your --who cover?

I'm all for having better collections of this info out there. I've "wanted to
but not had time to do" a computer history book (or podcast series or something)
for years. But it's like trying to summarize a 50 year long soap opera.

Here, read this:

http://landley.net/history/mirror/cpm/history.html

Watch this (all 4? parts):

https://www.youtube.com/watch?v=sX5g0kidk3Y

Get a copy of "a quarter century of unix" by peter salus, "where wizards stay up
late" on the history of the internet, "crystal fire" on the history of the
transistor, "big blues" (the history of IBM up vaguely until y2k), Steven Levy's
book "hackers" (which is the other half of
http://landley.net/history/mirror/interviews/olsen.html), Linus's "just for fun"
book, and then read these four different <strike>views of mt fuji</strike>
reminiscences about the creation of the microprocessor:

http://landley.net/history/mirror/interviews/Moore.html
http://landley.net/history/mirror/intel/Hoff.html
http://landley.net/history/mirror/interviews/Faggin.html
http://landley.net/history/mirror/intel/shima.html

And the Long Strange Trip to Java:

http://landley.net/history/mirror/java/javaorigin.html

And if you can track down a copy of the HOPL II proceedings (the UT Library in
Austin has one) read Grace Hopper's keynote because it's lovely...

THEN worry about piecing together the rest of the history.

Rob

P.S. Howard Aiken importantly followed up on Charles Babbage's long-abandoned
work, leading directly to the creation of the Eniac and Univac and the IBM 701
and the modern computer industry, with a bit of help from the UK projects going
on at the same time. Grace Hopper, who invented the compiler, was Howard Aiken's
assistant, assigned to him by the Navy. Meanwhile some guy named Konrad Zuse
tinkered in a nazi basement and all his work was destroyed by allied air raids
and nobody ever followed up any of it or was even aware it had hapened until
decades later when his son started trying to get him credit for stuff, but the
industry would have turned out EXACTLY THE SAME if that man was never born and
every documentary that even mentions him can safely be stopped at that point and
discarded.)



More information about the svlug mailing list