2026-09-10 19:08 UTC
@simon_brooke@mastodon.scot @sigue@universeodon.com @mdhughes@appdot.net @Ardubal@mastodon.xyz @vnikolov@ieji.de @screwlisp@gamerplus.org @kentpitman@climatejustice.social
I vaguely recalled error messages from lucid lisp when you tried to change T and NIL, so I asked my partner who remembered better. He says:
As I recall the errors were "Veritas Aeterna -- can't setq T" and "Nihil ex nihil -- can't setq NIL". The latin in the latter is wrong; it should be "Nihil ex nihilo", but the above is what I remember. This may date back to Maclisp. In Common Lisp there are constants that can't be set, so you don't really need special messages for T and NIL.
Replies (1)
-
@kentpitman@climatejustice.social 2026-09-10 19:53
@nosrednayduj@hachyderm.io @simon_brooke@mastodon.scot @sigue@universeodon.com @Ardubal@mastodon.xyz @vnikolov@ieji.de @screwlisp@gamerplus.org They do date back to Maclisp, as does the grammar complaint. Blame Quux[1] (Guy Steele), I suspect. There is also a third that happens when you assign another 'pure' (what we call readonly now) quantity. I was afraid I would not remember correctly (though, oddly, I did), so I looked it up in "COMMON;LINS >" on an ITS emulator: VERITAS AETERNA -- DON'T SETQ T. PURITAS NECESSE EST -- DON'T DO RANDOM BINDINGS. NIHIL EX NIHIL -- DON'T SETQ NIL. [1] https://speechcode.com/jargon/jargon.info.Quux.html