> From: Rob Pike > I still marvel at the productivity and precision of his generatio We noticed the same thing happening in the IETF, as the number of people working on networking went up. The explanation is really quite simple, once you think about it a bit. If you have a very small group, it is quite possible to have a very high level. (Not if it's selected randomly, of course; there has to be some sorting function.) However, as the group gets much larger, it is _necessarily_ much more 'average' in the skill/etc level of its members. This rule applies to any group - which includes the members of TUHS, of course. Noel
On Tue, Aug 09, 2022 at 01:42:32PM -0400, Noel Chiappa wrote:
> > From: Rob Pike
>
> > I still marvel at the productivity and precision of his generatio
>
> We noticed the same thing happening in the IETF, as the number of people
> working on networking went up. The explanation is really quite simple, once
> you think about it a bit.
>
> If you have a very small group, it is quite possible to have a very high
> level. (Not if it's selected randomly, of course; there has to be some
> sorting function.) However, as the group gets much larger, it is
> _necessarily_ much more 'average' in the skill/etc level of its members.
I used to complain about this at Sun and was dryly told "We get it,
Larry, you are yeast. You need flour to make bread."
And as time went on, I found that the smart people tended to find each
other. So it was fine.
It is more fun when it is a highly curated group of smart people. Made
me work hard to keep up.
On 8/9/22 2:49 PM, Larry McVoy wrote:
> On Tue, Aug 09, 2022 at 01:42:32PM -0400, Noel Chiappa wrote:
>> > From: Rob Pike
>>
>> > I still marvel at the productivity and precision of his generatio
>>
>> We noticed the same thing happening in the IETF, as the number of people
>> working on networking went up. The explanation is really quite simple, once
>> you think about it a bit.
>>
>> If you have a very small group, it is quite possible to have a very high
>> level. (Not if it's selected randomly, of course; there has to be some
>> sorting function.) However, as the group gets much larger, it is
>> _necessarily_ much more 'average' in the skill/etc level of its members.
> I used to complain about this at Sun and was dryly told "We get it,
> Larry, you are yeast. You need flour to make bread."
>
> And as time went on, I found that the smart people tended to find each
> other. So it was fine.
>
> It is more fun when it is a highly curated group of smart people. Made
> me work hard to keep up.
Put another way, "If you're always the smartest person in the room,
you're spending your time in the wrong rooms."
On Tue, Aug 09, 2022 at 02:54:08PM -0400, Tom Teixeira wrote: > On 8/9/22 2:49 PM, Larry McVoy wrote: > >On Tue, Aug 09, 2022 at 01:42:32PM -0400, Noel Chiappa wrote: > >> > From: Rob Pike > >> > >> > I still marvel at the productivity and precision of his generatio > >> > >>We noticed the same thing happening in the IETF, as the number of people > >>working on networking went up. The explanation is really quite simple, once > >>you think about it a bit. > >> > >>If you have a very small group, it is quite possible to have a very high > >>level. (Not if it's selected randomly, of course; there has to be some > >>sorting function.) However, as the group gets much larger, it is > >>_necessarily_ much more 'average' in the skill/etc level of its members. > >I used to complain about this at Sun and was dryly told "We get it, > >Larry, you are yeast. You need flour to make bread." > > > >And as time went on, I found that the smart people tended to find each > >other. So it was fine. > > > >It is more fun when it is a highly curated group of smart people. Made > >me work hard to keep up. > > Put another way, "If you're always the smartest person in the room, you're > spending your time in the wrong rooms." I was usually the dumbest one in the room, I found the right rooms :-) I personally like being "dumb", the other people just make you want to work harder to reach their level. Back when I used to play pool pretty seriously, I always tried to play people better than me. You get lazy if you are the best. -- --- Larry McVoy Retired to fishing http://www.mcvoy.com/lm/boat
> I've always believed that pic was so well designed because it took a day to get the print out (back then), so you had to have a language where you could see what it was doing. > I'll confess: I was never very good at bench checking batch programs, but only had at most a handful of assignments in college: generally cycles were cheap on time-sharing systems and I quickly adapted to interactive debugging. Along these lines, if I'm understanding correctly, my hunch would be that part of the precision being discussed was born out of necessity. When you can't debug interactively, you're forced to be precise with your changes, influencing how you think. On the flip side, when interactive development is an option, there's an easy route to take - and so that's what ends up informing those developmer's thought patterns. I think it's possible that if you were to force a new generation to only be able to iterate once a day, you may end up with a new generation with that precision. Perhaps material for a fun experiment for the teachers on the list. Cheers, Marshall On Tue, Aug 9, 2022 at 3:01 PM Larry McVoy <lm@mcvoy.com> wrote: > > On Tue, Aug 09, 2022 at 02:54:08PM -0400, Tom Teixeira wrote: > > On 8/9/22 2:49 PM, Larry McVoy wrote: > > >On Tue, Aug 09, 2022 at 01:42:32PM -0400, Noel Chiappa wrote: > > >> > From: Rob Pike > > >> > > >> > I still marvel at the productivity and precision of his generatio > > >> > > >>We noticed the same thing happening in the IETF, as the number of people > > >>working on networking went up. The explanation is really quite simple, once > > >>you think about it a bit. > > >> > > >>If you have a very small group, it is quite possible to have a very high > > >>level. (Not if it's selected randomly, of course; there has to be some > > >>sorting function.) However, as the group gets much larger, it is > > >>_necessarily_ much more 'average' in the skill/etc level of its members. > > >I used to complain about this at Sun and was dryly told "We get it, > > >Larry, you are yeast. You need flour to make bread." > > > > > >And as time went on, I found that the smart people tended to find each > > >other. So it was fine. > > > > > >It is more fun when it is a highly curated group of smart people. Made > > >me work hard to keep up. > > > > Put another way, "If you're always the smartest person in the room, you're > > spending your time in the wrong rooms." > > I was usually the dumbest one in the room, I found the right rooms :-) > > I personally like being "dumb", the other people just make you want to > work harder to reach their level. Back when I used to play pool pretty > seriously, I always tried to play people better than me. You get lazy > if you are the best. > > -- > --- > Larry McVoy Retired to fishing http://www.mcvoy.com/lm/boat
[-- Attachment #1: Type: text/plain, Size: 4669 bytes --] On Tue, Aug 9, 2022 at 1:23 PM Marshall Conover <marzhall.o@gmail.com> wrote: > > I've always believed that pic was so well designed > because it took a day to get the print out (back then), so you had to > have a language where you could see what it was doing. > > > I'll confess: I was never very good at bench checking batch programs, > but only had at most a handful of assignments in college: generally cycles > were cheap on time-sharing systems and I quickly adapted to interactive > debugging. > > Along these lines, if I'm understanding correctly, my hunch would be > that part of the precision being discussed was born out of necessity. > When you can't debug interactively, you're forced to be precise with > your changes, influencing how you think. On the flip side, when > interactive development is an option, there's an easy route to take - > and so that's what ends up informing those developmer's thought > patterns. > > I think it's possible that if you were to force a new generation to > only be able to iterate once a day, you may end up with a new > generation with that precision. Perhaps material for a fun experiment > for the teachers on the list. > I think it was a confluence of many things. Programs had to be smaller (bigger ones didn't fit). Interactive terminals were non-existant or extremely limited (80x24). Printing out listings and 'desk checking' the output was something you had plenty of time to do. Computing budgets were tiny: You had only so many $$$ for your runs and if you made too many, you'd run out of $$$ before you were done (more applicable as a student than as a professional post school though). Consequently your time was plentiful and computer time was scarce. Plus people from that generation tended to think globally and didn't compartmentalize as much as is done today (where people are told that everything below you in the stack can be considered hardware don't worry about how it works). The systems were also simpler to program, since all the 'go fast' caveats you have to cope with in todays system didn't exists, which also encourage global thinking. Plus, computer programmers tended to be the best and the brightest because they were the only ones that could (a) afford to undertake their study and (b) the only ones that didn't wash out of very demanding university programs. Plus companies tended to only trust their super expensive machines to the best and the brightest, further enhancing their skills (which we now know are built with repetition) while the less bright tended to be relegated to other machines with fewer opportunities. (yes, I know the previous paragraph way over-generalizes a very complex and subtle dynamic that was at play, hence 'tendency' rather than some other more definite word). Warner > Cheers, > > Marshall > > On Tue, Aug 9, 2022 at 3:01 PM Larry McVoy <lm@mcvoy.com> wrote: > > > > On Tue, Aug 09, 2022 at 02:54:08PM -0400, Tom Teixeira wrote: > > > On 8/9/22 2:49 PM, Larry McVoy wrote: > > > >On Tue, Aug 09, 2022 at 01:42:32PM -0400, Noel Chiappa wrote: > > > >> > From: Rob Pike > > > >> > > > >> > I still marvel at the productivity and precision of his > generatio > > > >> > > > >>We noticed the same thing happening in the IETF, as the number of > people > > > >>working on networking went up. The explanation is really quite > simple, once > > > >>you think about it a bit. > > > >> > > > >>If you have a very small group, it is quite possible to have a very > high > > > >>level. (Not if it's selected randomly, of course; there has to be > some > > > >>sorting function.) However, as the group gets much larger, it is > > > >>_necessarily_ much more 'average' in the skill/etc level of its > members. > > > >I used to complain about this at Sun and was dryly told "We get it, > > > >Larry, you are yeast. You need flour to make bread." > > > > > > > >And as time went on, I found that the smart people tended to find each > > > >other. So it was fine. > > > > > > > >It is more fun when it is a highly curated group of smart people. > Made > > > >me work hard to keep up. > > > > > > Put another way, "If you're always the smartest person in the room, > you're > > > spending your time in the wrong rooms." > > > > I was usually the dumbest one in the room, I found the right rooms :-) > > > > I personally like being "dumb", the other people just make you want to > > work harder to reach their level. Back when I used to play pool pretty > > seriously, I always tried to play people better than me. You get lazy > > if you are the best. > > > > -- > > --- > > Larry McVoy Retired to fishing > http://www.mcvoy.com/lm/boat > [-- Attachment #2: Type: text/html, Size: 6156 bytes --]
Early on in my career at IBM Yorktown, ca. 1976, I was submitting many long running simulation jobs to the 360/91 there. At one point, the head of computer systems (I.T. if you will) wrote to the head of computer sciences (my department) complaining that I had just spent $50K over some short period, asking if this was justified. My management shrugged it off, encouraged me to continue what I was doing. I might still have the letter somewhere. A couple of years later, while on the faculty at U.T. Austin, one of the main budgetary items in research grant proposals was purchase of mini-computers, assuming those were a more efficient use of funds than paying for time at the campus computing center (then using CDC 6600 and successors). COFF? Charlie On 8/9/2022 3:19 PM, Warner Losh wrote: > Computing budgets were tiny: You had only so many $$$ for your runs and > if you made > too many, you'd run out of $$$ before you were done (more applicable as > a student than > as a professional post school though). Consequently your time was > plentiful and > computer time was scarce. -- voice: +1.512.784.7526 e-mail: sauer@technologists.com fax: +1.512.346.5240 Web: https://technologists.com/sauer/ Facebook/Google/Twitter: CharlesHSauer
[-- Attachment #1: Type: text/plain, Size: 2731 bytes --] LOL I joined IBM Research in Yorktown in 1978. I was an electrical engineer and one of the first problems I was given was modeling a novel concept for an X-Y touch panel. I realized that the model is basically solving Laplace's equation in the plane. I was not a programmer at the time, so I asked what was the recommended thing for that. I was told APL, so I grabbed a manual and got to work. Within a day or two I had a nice solver working and was getting useful results. (Of course, solving Laplace in the plane by relaxation is the slowest possible way to get to the answer, but I didn't know much about numerical methods back in those days.) The next week I got a visit from the same IT weenies who had bothered you. They told me that in my first week on the job I had managed to be the biggest consumer of CPU cycles on the 370/168 and that I had to learn to program in PL/I because compiled was better than interpreted. It took me several weeks to get it working, since PL/I was such a pain in the neck and I had to learn all sorts of stuff about how numbers were represented in the hardware. Obviously my time was worth less than the computer's. Bleh. ===== nygeek.net mindthegapdialogs.com/home <https://www.mindthegapdialogs.com/home> On Tue, Aug 9, 2022 at 4:43 PM Charles H Sauer (he/him) < sauer@technologists.com> wrote: > Early on in my career at IBM Yorktown, ca. 1976, I was submitting many > long running simulation jobs to the 360/91 there. At one point, the head > of computer systems (I.T. if you will) wrote to the head of computer > sciences (my department) complaining that I had just spent $50K over > some short period, asking if this was justified. My management shrugged > it off, encouraged me to continue what I was doing. I might still have > the letter somewhere. > > A couple of years later, while on the faculty at U.T. Austin, one of the > main budgetary items in research grant proposals was purchase of > mini-computers, assuming those were a more efficient use of funds than > paying for time at the campus computing center (then using CDC 6600 and > successors). > > COFF? > > Charlie > > On 8/9/2022 3:19 PM, Warner Losh wrote: > > > Computing budgets were tiny: You had only so many $$$ for your runs and > > if you made > > too many, you'd run out of $$$ before you were done (more applicable as > > a student than > > as a professional post school though). Consequently your time was > > plentiful and > > computer time was scarce. > > -- > voice: +1.512.784.7526 e-mail: sauer@technologists.com > fax: +1.512.346.5240 Web: https://technologists.com/sauer/ > Facebook/Google/Twitter > <https://technologists.com/sauer/Facebook/Google/Twitter>: CharlesHSauer > [-- Attachment #2: Type: text/html, Size: 4691 bytes --]
And, of course, it should be noted that $50K was significantly more than typical annual salaries for researchers back then. On 8/9/2022 3:58 PM, Marc Donner wrote: > LOL > > I joined IBM Research in Yorktown in 1978. I was an electrical engineer > and one of the first problems I was given was modeling a novel concept > for an X-Y touch panel. I realized that the model is basically solving > Laplace's equation in the plane. I was not a programmer at the time, so > I asked what was the recommended thing for that. I was told APL, so I > grabbed a manual and got to work. > > Within a day or two I had a nice solver working and was getting useful > results. > > (Of course, solving Laplace in the plane by relaxation is the slowest > possible way to get to the answer, but I didn't know much about > numerical methods back in those days.) > > The next week I got a visit from the same IT weenies who had bothered > you. They told me that in my first week on the job I had managed to be > the biggest consumer of CPU cycles on the 370/168 and that I had to > learn to program in PL/I because compiled was better than interpreted. > It took me several weeks to get it working, since PL/I was such a pain > in the neck and I had to learn all sorts of stuff about how numbers were > represented in the hardware. > > Obviously my time was worth less than the computer's. > > Bleh. > ===== > nygeek.net <http://nygeek.net> > mindthegapdialogs.com/home <https://www.mindthegapdialogs.com/home> > > > On Tue, Aug 9, 2022 at 4:43 PM Charles H Sauer (he/him) > <sauer@technologists.com <mailto:sauer@technologists.com>> wrote: > > Early on in my career at IBM Yorktown, ca. 1976, I was submitting many > long running simulation jobs to the 360/91 there. At one point, the > head > of computer systems (I.T. if you will) wrote to the head of computer > sciences (my department) complaining that I had just spent $50K over > some short period, asking if this was justified. My management shrugged > it off, encouraged me to continue what I was doing. I might still have > the letter somewhere. > > A couple of years later, while on the faculty at U.T. Austin, one of > the > main budgetary items in research grant proposals was purchase of > mini-computers, assuming those were a more efficient use of funds than > paying for time at the campus computing center (then using CDC 6600 and > successors). > > COFF? > > Charlie > > On 8/9/2022 3:19 PM, Warner Losh wrote: > > > Computing budgets were tiny: You had only so many $$$ for your > runs and > > if you made > > too many, you'd run out of $$$ before you were done (more > applicable as > > a student than > > as a professional post school though). Consequently your time was > > plentiful and > > computer time was scarce. > > -- > voice: +1.512.784.7526 e-mail: sauer@technologists.com > <mailto:sauer@technologists.com> > fax: +1.512.346.5240 Web: https://technologists.com/sauer/ > Facebook/Google/Twitter > <https://technologists.com/sauer/Facebook/Google/Twitter>: CharlesHSauer > -- voice: +1.512.784.7526 e-mail: sauer@technologists.com fax: +1.512.346.5240 Web: https://technologists.com/sauer/ Facebook/Google/Twitter: CharlesHSauer
“The trouble with folk songs is thst they are written by the people.” -Tom Lehrer.
Ches
> On Aug 9, 2022, at 13:42, jnc@mercury.lcs.mit.edu wrote:
>
> However, as the group gets much larger, it is
> _necessarily_ much more 'average' in the skill/etc level of its members.
[-- Attachment #1: Type: text/plain, Size: 448 bytes --] On Wed, Aug 10, 2022 at 5:13 PM Bill Cheswick <ches@cheswick.com> wrote: > “The trouble with folk songs is thst they are written by the people.” -Tom > Lehrer. Lehrer's songs, especially "The Periodic Table", have now become subject to the folk process themselves. "Goldilocks and the Three Bears" was originally written by the Romantic poet Robert Southey, although in his version the bears' antagonist was a "wicked old woman". [-- Attachment #2: Type: text/html, Size: 1087 bytes --]
[-- Attachment #1: Type: text/plain, Size: 714 bytes --] Here's your blackmail video: https://www.youtube.com/watch?v=02N_UOmT8Kk&t=2s ===== nygeek.net mindthegapdialogs.com/home <https://www.mindthegapdialogs.com/home> On Wed, Aug 10, 2022 at 5:32 PM John Cowan <cowan@ccil.org> wrote: > > > On Wed, Aug 10, 2022 at 5:13 PM Bill Cheswick <ches@cheswick.com> wrote: > >> “The trouble with folk songs is thst they are written by the people.” >> -Tom Lehrer. > > > Lehrer's songs, especially "The Periodic Table", have now become subject > to the folk process themselves. "Goldilocks and the Three Bears" was > originally written by the Romantic poet Robert Southey, although in his > version the bears' antagonist was a "wicked old woman". > [-- Attachment #2: Type: text/html, Size: 2002 bytes --]