Fork me on GitHub
#clojure-uk
<
2021-06-03
>
dharrigan05:06:55

Good Morning!

3
djm06:06:56

https://lispcast.com/some-comments-from-a-numerical-analyst/ - this was much more interesting than I expected

djm06:06:05

“Many people who set out originally to solve some problem in mathematical physics found themselves temporarily deflected by the problems of computerology.” - I guess that means that yak shaving is the essence of programming (and “computerology” is a term that needs to make a comeback)

djm06:06:31

It fits somewhat with my own experience, as much of my early programming was around solving maths/physics problems, and I switched to CS after one year of a physics degree

djm06:06:52

No, I’m not familiar with that

Aron06:06:38

that video was a big moment in life, because who Kahan is and how he behaves and what he says. I recognized the pattern.

Aron06:06:00

I have not seen posits implemented in Clojure ( https://groups.google.com/g/numerical-clojure/c/QE6-4vE7fQ8 )

Aron06:06:55

Obviously, it will be best when it gets hardware support. There are some chips made, but I am not sure about their popularity.

Aron06:06:47

imho, if anything is keeping back physics, it's the bad assumptions of mathematics that they import so willy-nilly, rather than practical considerations of the hardware and software they have to use.

Conor08:06:46

What bad assumptions do you mean?

Aron08:06:01

Continuum.

Conor08:06:00

I don't think I ever saw that TV show

Aron09:06:03

Very funny. I don't watch tv at all, so anything related to that is not my fault. I am entirely serious. The way mathematicians deal with the nature/idea/whateveryoucallit continuum is adopted by physicists without any reservations, and it's causing problems historically.

Conor09:06:36

I don't really understand the problem you are referring to (speaking as someone who has a PhD in computational physics) - maybe you could expand a bit more?

Aron09:06:07

well, computational physics exactly avoids this problem, doesn't it?

Aron09:06:12

I would be very happy to discuss it at length, if it's ok to pm you. I would like to ask a bunch of questions.

Conor09:06:48

Sure, though I can't promise I will have any good answers 😄

Aron06:06:11

One more link about this subject, it's the 4th and last episode of a short series https://youtu.be/4DNlEq0ZrTo?list=PLIljB45xT85Bfc-S4WHvTIM7E-ir3nAOf

paulspencerwilliams08:06:54

@djm_uk I think that observation is somewhat correlated to various technical platforms. When I’ve attended general development conferences (QCon etc), and even some technical specific conferences (A Ruby one), speakers seem highly aligned to the technology - they might be Java evangelists, or Mongo experts etc. I’ve attended 3 Clojure{X/:re) conferences, and what struck me was the amount of speakers that were using technology to solve problems, be they musical, statistics, modelling or whatever. From my observations, this correlation is somewhat related to Clojure’s data orientation, and lisps’ expressiveness which lends to solutions that are very close representations of the underlying mathematical models.

djm09:06:13

It would be interesting to think through what the equivalents would have been at the time the paper was written

danm09:06:18

Morning from the new house!

🎉 3
alexlynham09:06:32

i think it might also be that lisps traditionally attract a lot of hacker/tinkerer type mindset people, hence potential broadness of solution space... but idk that i see people generally doing things in clj they aren't doing in other langs i think the reason you see talks like this in the clj space is 'cos the community is so small that a lot of people are using it for small side hacks, so a talk on generative art may be the only thing that it's being used for, or building keyboards in scad, or whatever in the js/typescript space or somewhere else they'd have 3 talks just on different option flags for the same library, cos so many people use the tools, all day, every day by comparison so i think that might be a reflection of the amount of day job professional work done in a given lang as well

alexlynham09:06:04

just a suspicion anyway, that it's something to do with the intersection of those two things - more hackers x smaller community

6
Aron09:06:51

the only typescript talk I ever cared about was where the inventor/creator of typescript described how the language came to be https://www.youtube.com/watch?v=ET4kT88JRXs then I hear this in the first 5 minutes: emphasis mine. "[talking about what missing features typescript adds for js]...but then when you compile your code you are just left again with plain old javascript. So there is really no notion of any of this having any of this having existed other than at development time." So I am like, wait, at runtime I can't use this information? No, this is about writing the same programs over and over again, not about writing new things based on old things.

Luis Thiam-Nye10:06:09

Good Morning™

djm11:06:29

Are you aware of #remote-jobs?

rickmoynihan11:06:31

Yes, but thanks for the reminder, I’ll cross post there too.

👍 3
rickmoynihan11:06:01

Isn’t every (software) job remote these days? 😩

djm11:06:37

Apparently not

dominicm19:06:41

Morning 🙂

3