Title about says it. I figure the American holiday stuff meant less time for server restarting.
screwtape
Creator of
Recent community posts
Okay! In my role as editor (of population:you) I took your good code cutout of my example and turned it into an asdf system (well, a :class :package-inferred-system) here: https://codeberg.org/screwlisp/mdh-guess
The gist is that instead of how it used to work it is now like:
(asdf:load-system :mdh-guess)
(mdh-guess/mdhzine:run-zine)
I guess the git either needs to be in ~/common-lisp/ or ~/.local/share/common-lisp/source/ for this to work. Can use nicknames. I nicknamed :mdh-guess/marklib to just be :marklib so your convention works.
It is as portable as asdf3+ is.
The big material change for you is probably that instead of using the local directory as the system directory, it now uses (asdf:system-source-directory :mdh-zine) as the source directory.
Now clicking, if the click is somewhere insensitive, moves the cursor hexagonally.
Actually, it is / is meant to be a drag-and-drop-translator, but since the :feedback tells the frame to call a command, the frame thinks that the drag-and-drop has ended after a single step.
Still, now we can both grab stuff by clicking, and also move by clicking so it is sort of a win.
Now when the frame opens, it renders the entire world (stretching out with scrollbars). However painting the entire world naturally places the screen at the top left corner of the entire world. Then, redisplays only happen where the current screen is looking (i.e. there is no meaning to redisplaying something you are not looking at). I haven’t quite made it to using :incremental-redisplay idiomatically, but we will get there.
So if you need to redisplay the entire world (to fix scrollbars?), reopen the application frame, otherwise just let it redisplay locally.
This thread is going to be my uploaded-a-new-nicclim.lisp notes.
Further commentary from Kent https://climatejustice.social/@kentpitman/116716609588549299
( https://en.wikipedia.org/wiki/Kent_Pitman )
N.b. I added a bunch of spot changes

From my perspective, the jam was 20 entries, from the forum post:
1/4
Lua with the flexibility of a lisp syntax and macro system.
1/4 clojure
1/4 The host guile scheme
1/4 chicken scheme + misc e.g. kawa scheme
Which I guess is a founder-effected version of wikipedia’s explanation of lisp given the small number of entries:
Today, the best-known general-purpose Lisp dialects are Common Lisp, Scheme, Racket, and Clojure.
I guess having a sense of incommensurability is on brand for lisp.
You and twenty other people! I did have the thought that McCLIM ( https://en.wikipedia.org/wiki/McCLIM implementation of https://en.wikipedia.org/wiki/Common_Lisp_Interface_Manager ) really wants for BSD ports / linux packages, because my
basically npm install but lisp
gives the wrong impression about what it and lisp are I think.
Yeah though I froze development for the jam rating season, you are right that click to place / modify is a must.
The deep learning only operates on Xs and Os; the word thing builds a phrase from symbols where the cell has an X. My plan for the DL is to run a predicate everywhere, which marks X or O according to the predicate, then run Dl infer, then run an update function everywhere (presumable changes).
I really think you should be able to run it in either sbcl or ecl common lisps and just use quicklisp for mcclim.
Here, I made it a deep-learning-evolving field of flowers as another example.
https://lispy-gopher-show.itch.io/dl-roc-lisp/devlog/1536402/deep-in-chrysanthemums

Edit: Scroll down to /post/16170658 to just see it working directly with sbcl in basically exactly the same way. The change to be portable was trivial.
I was doubling down on ECL viz sbcl’s lack of anti-llm policy. I was wondering if updating quicklisp might help
(ql:update-client)
(ql:update-all-dists)
(ql:quickload :mcclim)
. How do you install things on mac? I am a bit in the dark, but aren’t there basically BSD ports for mac named macports?
Anyway the only nonportable thing I think I have done is
(compile-file #P"some/file.lisp" :load t)
which portably would need to be
(compile-file #p"some/file.lisp")
(load #p"some/file.fas")
I think. I will try it on sbcl and change the useage notes to be compatible with sbcl.
Yeah, watching your game unfurl has been the real feature of mastodon this week.
I drastically improved my hextille: https://lispy-gopher-show.itch.io/dl-roc-lisp/devlog/1531334/dl-roc-hex-improved-ez-eg-hextille-gfx-game-lisp (just about installing and putting pictures on hexagons/edges, none of my deep learning theorizing).
Quick aside about “seeming slow”. That is because I do not cache images, and uncompress and make a new pattern out of the same image every single time it is used in every single frame. My justification was that I did not want to “try and outsmart the garbage collector” even if it produces the bizarrely slow loading experience every time the display updates. The DL itself at this scale is arbitrarily quick.
Alright, I decided to basically call the jam attempt on my dl-roc-hex.lisp.
~ https://lispy-gopher-show.itch.io/dl-roc-lisp/devlog/1529575/deep-learning-roc-lisp-gui-tool-super-demo-jam-sub . Do I have to do anything else to submit?
See https://lispy-gopher-show.itch.io/dl-roc-lisp/devlog/1529575/deep-learning-roc-lisp-gui-tool-super-demo-jam-sub for useage instructions please.
You have no idea (well, maybe) how difficult the way I decided to do this was.

I guess I am using and further developing my original deep learning infrastructure for this jam, as my excuse. WIP
Edit: For clarity, when I say deep learning, I mean like my own original lisp program and formulation of it, not some slop crap. Yᵩᵢ = F( Χᵩᵢ + ξᵩᵢ ) − F( Χᵩᵢ − ξᵩᵢ ) …
To make a personal comment, this use of my leonardo system (what this itchio is about) https://codeberg.org/screwlisp/pawn-75
to write an ontology, and from that ontology classically generating lisp snippets /is/ what is meant to happen.
Admittedly, this example was “programming the well known unix ls using conditions and restarts as generated by the leonardo system from this ontology” but this is a microcosm of the leonardo system, and its incidence with ansi cl’s condition system.
https://screwlisp.small-web.org/lispgames/nicclim.lisp I spot fixed the PATH thing, I didn’t figure out uiop:chdir not working yet though. Will try tomorrow.
















