Scott Williams 🐧Sep 14, 2023
Without looking it up, which one of these four was *not* one of the original four nodes of the internet/arpanet? #internet #computerhistory #networking
UC Berkeley
8.8%
UCLA
33.7%
UC Santa Barbara
26.5%
University of Utah
30.9%
Poll ended at .
Show threadJohnMasheySep 14, 2023
@vwbusguy
I didn';t have to look it up, since I've seen this map a few times at teh Computer History Museum:
https://www.computerhistory.org/collections/catalog/102658020
Also,l next to an early IMP, there's a display that cycles through the growth of the Arpanet.
Arpa Network four node map | 102658020 | Computer History Museum

Show threadJohnMasheySep 14, 2023
@vwbusguy
And as amusing bit of history, ~1971, 8 of the 15-site ARPANET computers were DEC PDP-10s, favored by leading CMPSC departments. Bell Labs had quit Multics, Ken Thompson had created Unix on PDP-7, and Computign Science Research Center wanted to buy a PDP-10 , but couldn't get the budget, so had to settle for (relatively tiny) PDP-11/20...
one of the coincidences that turned out to have far-reaching effects on computing:
https://techviser.com/wp-content/uploads/2022/06/Mashey.IEEE_.Micro_.2022.pdf
Show threadMark GardnerSep 14, 2023
@JohnMashey @darkuncle @vwbusguy And if you want an account on a system that uses the 36-bit road not traveled, head on over to @SDF’s https://twenex.org
SDF Public Access TWENEX

Show threadJohnMasheySep 14, 2023

@mjgardner @darkuncle @vwbusguy @SDF
From Ken Thompson’s Turing lecture:
“I thank the ACM for this award. I can't help but feel that I am receiving this honor for timing and serendipity as much as technical merit. UNIX swept into popularity with an industry-wide change from central mainframes to autonomous minis. I suspect that Daniel Bobrow would be here instead of me if he could not afford a PDP-10 and had had to "settle" for a PDP-11.”

Last 36-bitters I used were IBM 7090 and Univac 1108.

Show threadScott Williams 🐧Sep 14, 2023
@JohnMashey @mjgardner @darkuncle @SDF That's awesome! I've heard stories from John Maddog Hall about machines that used non 8-bit bytes and slicing memory in interesting ways as a sort of early cycle defragmentation technique before things all got standardized around 8-bit bytes. Whenever he gets into story mode, I always end up learning a few things along the way.
Show threadJohnMasheySep 14, 2023
@vwbusguy @mjgardner @darkuncle @SDF
Univac 1108: (via Wikipedia):
“ Whole word – 36-bit (ones' complement)
Half word – two 18-bit fields per word (unsigned or ones' complement)
Third word – three 12-bit fields per word (ones' complement)
Quarter word – four 9-bit fields per word (unsigned)
Sixth word – six 6-bit fields per word (unsigned)”
So C char could plausibly be 9 or 6-bit…
but byte-within word was encoded in instruction … which meant that C str* functions were pretty ugly.
Show threadMark GardnerSep 14, 2023
@JohnMashey @vwbusguy @darkuncle @SDF a/k/a VAXocentrism: http://catb.org/jargon/html/V/vaxocentrism.html
vaxocentrism

Show threadJohnMasheySep 14, 2023
@mjgardner @vwbusguy @darkuncle @SDF
Yes, although by mid-1980s, good software-engineering groups did portability well. Of course, CPUs other than 8-bit byte-addressed, were going extinct. (Of course, DEC's move to that for PDP-11/VAX was forced by IBM S/360.) For a while , both AMD 29K and Dec Alpha had byte addressing, but only word loads/stores, and had to fix that. Had we thought ahead more in ~1975, the size issues of early 1990s/64-bit might have been ameliorated:
https://queue.acm.org/detail.cfm?id=1165766
The Long Road to 64 Bits - ACM Queue

Show threadJohnMasheySep 14, 2023
@mjgardner @vwbusguy @darkuncle @SDF
Also, inside Bell Labs in 1970s, people (many of whom I knew) did C compilers for the plethora of computers we used, both Big and Litle-Endian, with stacks going up and down, word-addressed machines of various kinds and word-lengths, ones with unusual pointers.
Most of these never got outside.
At least by mid-1980s, the better third-party software companies would have an include file that specified the machine parameters.
Show threadMark GardnerSep 14, 2023
@JohnMashey @vwbusguy @darkuncle @SDF Now we have x86-centrism banging into Arm (https://learn.microsoft.com/en-us/windows/arm/apps-on-arm-troubleshooting-x86), GNU/Linuxisms running into BSDs and others (how the tables have turned!), and bashisms (https://mywiki.wooledge.org/Bashism)
Troubleshooting x86 desktop apps

Learn how to troubleshoot and fix common issues with an x86 desktop app running on Arm64 including information about drivers, shell extensions, and debugging.

Show threadScott Williams 🐧Sep 14, 2023

@mjgardner @JohnMashey @darkuncle @SDF You don't have to worry about bash-isms with my shell! It's time for CP/M-centrism! 😅

I even added CHDIR for PDP-10 backward compatibility!

https://qbsh.org

GitHub - vwbusguy/qbsh: Quick Basic Shell

Quick Basic Shell. Contribute to vwbusguy/qbsh development by creating an account on GitHub.

GitHub
Show threadMark Gardner
@vwbusguy @JohnMashey @darkuncle @SDF Hello PIP old friend. (PIP started on the #DEC #PDP6)
Sep 14, 2023 at 10:43pmWeb
Show threadScott Williams 🐧Sep 14, 2023
@mjgardner @JohnMashey @darkuncle @SDF Yup. I have an Osborne One sitting on the desk next to my as I type this, which is where I'm used to it. Gotta be careful to remember DEST=SRC - easy to get it backward when you're used to cp (like I still do half the time with ln..).
Show threadMark GardnerSep 15, 2023
@vwbusguy @JohnMashey @darkuncle @SDF I like to think the #Unix `ln` command continues that cheerful insanity. Which is source and which is destination? Depends on your perspective!
Show threadJohnMasheySep 16, 2023
@mjgardner @vwbusguy @darkuncle @SDF
Unix ln(1) was surely inspired by Multics lk.
http://www.bitsavers.org/pdf/honeywell/large_systems/multics/AG92-03A_multicsCmds_Feb80.pdf#page=393
That does pairs of source-newlinks, but I don't know if that feature was in Multics circa 1968 or added later.