On Mon, Jul 21, 2025 at 11:00 AM Larry McVoy <lm(a)mcvoy.com> wrote:
Please don't take this as an insult, Clem,
No worries.
I'm wondering if maybe us "less smart"
people, just don't have the extra cycles
it takes to love emacs?
Maybe, but I've known a lot of smart people who stayed with ex/vi over
Emacs.
I never found Emacs any more intuitive than ex/vi - just a different
learning curve.
Since I came to UNIX from TSS and TOPS-10, I found that UNIX presented a
steep learning curve. I was accustomed to having a programmable command
system, editors, and specialized tools/programs to do specific things.
Yes, the 10's had things like SUDS and ISPS, which we did not (yet) have in
Unix land. And clearly, BLISS generated better code than C, so why would I
give all that up?
I was getting paid to program UNIX, so I learned to use it. The original
value was that I was getting paid. But the idea of things like standard
I/O, the pipeline, and >>small<< tools that I could reuse was intuitive.
Yup, the command had strange names, but I learned to >>love<< find scripts
and grep. I quickly discovered that, even on a small 11/40, I could
accomplish a lot. I quickly became more effective than I was on the PDP-10
- the 10 forced every program to be its subsystem, and while what was there
was excellent, I could not make it up -- i.e., I liked the tools approach
way more than the give a man a fish approach.
That said, given the capabilities I had in PDP-10 land, I still had an
account on them to run SUDS when I needed to create a schematic as an
example (CMU did not obtain UCDS for EE-CAD until my Senior year). FWIW:
Ted Kowalski's PhD was in developing ISPS for UNIX, although by the time he
finished, I had long since left. So the "special" tools that the PDP-10s
had became less and less critical, but the >>value<< I received for
learning UNIX was immense.
I suspect that if I had been forced to use Emacs and found its value (like
using m-lisp), like my UNIX investment, it might have been worth it.
Indeed, many intelligent people I know share this view.
But the key is I looked at what people did with Emacs and never thought, I
want/need that feature. As I said, I was not a LISP person, so things like
m-lisp were not valued. They liked it, but I could do everything they
could that >>I desired<< to do. The mode/non-mode argument never bothered
me before, so again, its value was not seen as necessary.
Is it possible that if we had cycles to spare,
we'd like emacs too?
Maybe - but (I think) it's a value proposition. If the cost is high and
you don't see a return on investment. If it had been available, I might
have even used it. We toyed with bringing RT-11 TECO from the Harvard
tape up on our V6++ systems, but never did, as we were fluent with ed(1)
and by then, fred(1) had shown up from Cornell [funny we did bring up the
ddt as the early debuggers for V5 and V6v were weak and adb did not yet
exist].
On the last team I built, all of the core of that
team were crazy smart,
way, way ahead of me.
Amen, brother - I always want people smarter than me on the team, and I can
say that I have been thrilled to have that occur with most of the major
projects I have been a part of.
Clem