Re: GETF Speedup?

IIRC, back in 1975, I was on a stipend for $3K per year, or about $250/month. So even $100 / week would have been a huge raise. Heh!



> On Oct 13, 2025, at 12:11, David McClain (as dbm at refined-audiometrics dot com) <[email protected]> wrote:
> 
> Wow! Thanks for that reference.
> 
> Zetalisp !! 
> 
> Holy Cow, I remember using Zeta Lisp back around 1975 (??). Just before Gerald Ford told NYC to “go to hell!” regarding its finances.
> 
> I was working on Madison Ave. around 70’th street or so. I think it was the old IBM Building. Just a meager grad student.
> 
> Those guys were running a SuperTanker Leasing company, funded by the famous Pirelli of Italy. They came to interview me at CCNY where I was studying Physics, and they asked me how much I wanted to work 20 hours/week for them? I was running the computer center software support activities in our Physics dept, on a DEC PDP-20 with 256 KB of storage !!
> 
> I remember telling them I would do it for $100 per week. And they chuckled and choked, and said, “How about $400 / week?”. Heh!!
> 
> I was so poor that I only had one pair of bluejeans, one shirt, and an old Army jacket that my brother gave me from his tour in S. Korea. 
> 
> The guys at the leasing agency have me $200 and the name of a tailor down the street, and told me to go get a haircut and buy a suit, to become presentable to their clients. Heh!! I had shoulder length hair at the time.
> 
> But alas, I was there to study Physics, and my Comprehensives were coming up in a month or so, 4 days of 4 hours each. I didn’t have a Bachelor’s Degree at the time, and was working toward a PhD on a provisional basis. They said everyone gets 2 tries at the Comprehensives, but since I didn’t have my Bachelor’s degree, I would only get 1 chance at them. And that year, nobody, out of 50, passed the exams.
> 
> So, I dropped kicked the lavish pay after a few weeks and went back to studying my buns off. When I took the exams, only two of us, out of 50, passed the exams. Heh! 
> 
> I don’t know, looking back, whether that was a smart or a dumb move… I think smart.
> 
> 
>> On Oct 13, 2025, at 11:14, Bradford Miller <[email protected]> wrote:
>> 
>> This discussion reminds me of the Symbolics “Table Management Facility” (q.v. https://www.chai.uni-hamburg.de/~moeller/symbolics-info/documentation/Symbolics-Common-Lisp-Language-Concepts.pdf ). 
>> 
>> The basic idea was you used the common lisp hash table interface in your code, and the machine would organize it for efficiency based on size (you could predeclare a size so you didn’t pay a runtime penalty for reorganization). 
>> 
>>> On Oct 13, 2025, at 2:06 PM, Yuri Davidovsky (as work at disclosure dot ie) <[email protected]> wrote:
>>> 
>>> 
>>> 
>>>> On 13 Oct 2025, at 17:28, David McClain <[email protected]> wrote:
>>>> 
>>>> And now the count is down to 10 associations!!
>>> 
>>> Pretty much what Bradshaw’s tests showed. 10 key/value pairs should still be sufficient for small time dictionary storage, especially if you place frequently used items at the top of the list. 
>>> 
>>> _______________________________________________
>>> Lisp Hug - the mailing list for LispWorks users
>>> [email protected]
>>> http://www.lispworks.com/support/lisp-hug.html
>> 
>