@interlisp 
The syntax of FORMAT and LOOP blends with Common Lisp in a way similar to how the CLISP notation augments Interlisp. I don't mind the odd syntaxes of these little languages as they don't force to step outside the language for expressing specialized tasks.
#lispyGopherClimate #LIVE #technology #podcast
https://communitymedia.video/w/vY4A9hxkZEQCe47UBASXFa
@kentpitman #climate #haiku and note on https://climate.us
@jns #gopher #phlog about #vcfmw
Mastodon #lisp happenings recently @mdhughes etc
#emacsconf proposal deadline is Friday!
Otherwise, I did lots of #commonLisp #programming this week and will talk about the #gof_ai & #DeepLearning #AI one, and maybe my spacetime-box #NicCLIM #gamedev.
https://screwlisp.small-web.org/fundamental/classical-ai-and-deep-learning-ai-difference-in-two-examples/
https://screwlisp.small-web.org/lispgames/spacetime-boxes-nicclim/
#classicalAI ( #gof_ai ) and #deeplearning #AI . A minimal #programming #example in #commonLisp for identifying the artist of famous pictures with either approach.
The code fits on the back of a cereal box. (Barely. Argh, numerically encoding data for DL..).
I think this is as clear an example as can be. Though I look forward to your feedback and criticisms.
I still care about you, #lisp newcomer: https://lispcookbook.github.io/cl-cookbook/variables.html
We see: defparameter vs defvar, let, let* and their lexical scope, unbound variables, style guidelines.
It just was missing on the Cookbook.
I wish this article pointed out explicitly that Lisp (if we need to be concrete, Common Lisp) has two levels of syntax:
the simple syntax of S-expressions
and
the idiosyncratic syntax of each special operator.
I wish this article considered XSLT, an ugly but homoiconic language.
I wish it mentioned that what is _usually_ called the Lisp reader _also_ does a scanner's job of constructing tokens.
Etc., which will remain for another time.
On Lisp Macros
My thesis is that they become easy
once you realize they extend the language
_without_ touching the language implementation,
no more and no less.
That is, macros extend the compiler _without_ modifying it.
It so _happens_ in Lisp that they are functions mapping lists to lists.
They are related to backquotes only because
they necessarily need to construct nested lists
and backquotes are one of the tools for that.
A Tiny Puzzle
What use might one have for
(defvar *foo*)
_without_ an initial value?
Which C programming language construct does this resemble a little?
On Naming Common Lisp
Erik Naggum once aptly pointed out
that it could have been named Garden Lisp.
In biology, the common variety and the garden variety (of a species) are synonym terms.
I was a gardener once.
It is a very worthy endeavor.
Spacetime boxes and #NicCLIM #gamedev : An ordeal #devlog #commonLisp
https://screwlisp.small-web.org/lispgames/spacetime-boxes-nicclim/
I made a gamedev-my-NicCLIM-oriented spacetime-box class to try and figure upon Olum's stuff. I got tied up making two spacetime boxes, then making one spacetime-box every neighbor of the other, which was a huge problem actually. It seemed to be working at the end. Skip quickly to the extensive and far-ranging if poorly-named conclusions.
All right.
If we suppose those years in the past are legitimate, I wouldn't want to lose those entries, so I would consider
(max 1900 year)
for the year.
(Leaving aside the matter that this is really a job for full-fledged date manipulation.)
Does `encode-universal-time' signal an error for an invalid year or does it return a bogus value?
If the former, I would wrap the call with `ignore-errors',
because if the year can be invalid,
then the types and values of all the other arguments are also suspect.
Soon, at 10:30 CDT / 15:30 UTC, I plan on streaming McCLIM app development (no backend work). This week I'm attempting to render 2D game tilesets using the recent work I've done with INDEXED-PATTERNs.
Has it occurred to you that an #LLM is a very good for generating boilerplate code? It's especially very good in frontend development.
But have you sometimes thought that the language shouldn't require you to write a lot of duplicated codes? And it should instead provide #metaprogramming capabilities like #macros so you could summarize multiple similar patterns?
I really miss macros in #Typescript and other frontend applications.
I mean using an LLM with like 400B parameters to autogenerate some frontend stuff is insane when you could simply fix the underlying tech instead of introducing new massive size tech to go around the problem, instead of solving it.
I think I really need to get back to #CommonLisp or maybe some other #Lisp languages. #Rust also has got macros and I could learn more advanced usages of them.
#lispyGopherClimate screwlisp reading Kent M Pitman's vignette remembering the #commonLisp ANSI standards process
#LispyGopherClimate #live
Missing guests! But joined by https://en.wikipedia.org/wiki/Kent_Pitman @dougmerritt @mdhughes
https://communitymedia.video/w/2YtDzQ55UsgSfuVqoexPAZ
The #ANSI #commonLisp committee standards process conversation between @vnikolov and @kentpitman , also #climate #climateCrisis haiku and things.
Some notes about the #knowledgeRepresentation handbook apropos #lisp #programming.
@jeremy_list haha, I probably should have warned you I wouldn't be putting haikuos on anything ;p.
#commonLIsp
Threads: https://bordeaux-threads.common-lisp.dev/ de factow standard, proposed by Barlow,
multiple event loops and the GUI: (honestly I would throw both of these in McCLIM which does both of these) https://mcclim.common-lisp.dev/
Surely it would still be Theseus' habitat though ;p.
For me I would quite like to hold onto the fact that your code is C++ in that magicaly ECL sffi way. C++ compatibility is good
Some of these "how deep is an atom in a #commonLisp form in a file" pictures are deeply satisfying to look at.
#gnuplot #clseries Sharpsign... Visualizing lisp expression leaf depth
https://screwlisp.small-web.org/momentary/alexandria-form-atom-heights/
Make your own in this thread plz. Will send help.
Requests, conceivably.
Revisits how to write
(let* ((current-dir (car (directory #p"./")))
(wild-dir
(make-pathname :directory
'(:relative :wild-inferiors))
)
(wild-here
(merge-pathnames wild-dir current-dir))
I usually sabotage my own learning process for #rust by not finding any good reason to learn it because of previous knowledge in #python, #nim and #golang.
So... Any good ideas to do in #rust that could be convincing enough to stick to it, without telling myself... I can do this in #python, #nim, #golang? (Other than saying writing a kernel or a bootloader)
Could even be a concept that is generally applicable (like I did for #commonlisp and #haskell).
@screwlisp wrote:
> I've had the idea that the loved-and-hated #commonLisp #formattedOutput is basically markdown
Again, I'd rather draw a (partial) similarity between Common Lisp's format language
and extended Backus-Naur (or Normal) form (EBNF).
Note also the common root of the two F-words.
It must be intentional; probably this has already been said.
Some format directives are analogous to lexical tokens
(as are almost all `printf' directives), of course.
