Piero Bosio Social Web Site Personale

@cwebber

All hail Emacs. The only real editor.

Iä iä ph'nglui mglw'nafh Emacs R'lyeh wgah'nagl fhtagn!

@thaodan @m3tti @cwebber

Or this: https://github.com/tmccombs/mexpr

That's the nice thing working with a mature language (that has skinnable syntax): You just know something should be feasible, and lo-and-behold: Somebody already has done it for you.

@thaodan @m3tti @cwebber

Indeed: https://github.com/vcerovski/binfix/

@thaodan @m3tti @cwebber

All procedure calls are actually polish notation ...

If one wants to develop longer math expressions (rare, actually, in most software engineering), e.g. for sceintific purposes, then there should be a macro "compiling" infix to polish.

@larsbrinkhoff @rnd @cwebber

Too late ...

I should have known and didn't pay attention.

@larsbrinkhoff @rnd @cwebber

And yet, everything compiles to binary and is executed as binary. Big reveal: processor doesn't understand Lisp, nor Python.

So what?

@chris_evelyn @cwebber @vindarel

Well, that's likely a search engine / AI problem. They have gotten much worse the last couple of years. That's why I am now paying for a search engine... My lifetime is to to valuable (to me) than to waste it on Google or Bing search results.

@m3tti @cwebber

Also refactoring what you wrote easily because everything is an expression. It's almost algebra, especially as long as you stay in functional style.

@eniko @cwebber

Just fast. My understanding is, that in CL tail calls are not guaranteed and anyway I think tail calls are not this relevant for performance --- more relevant are likely how much the compiler can optimize and the "fatness" of the underlying compiler / VM model.

@rnd @cwebber

SBCL is pretty fast. Faster than Python at first glance.

The thing that riled people up against Lisp where the parenthesis. A really stupid thing to base your language selection on.

@chris_evelyn @cwebber

Portacle.

Otherwise just emacs and then load sly or slime. You need to have those, since otherwise the indentation for macros will be off.

Here => https://lispcookbook.github.io/cl-cookbook/getting-started.html and here => https://lispcookbook.github.io/cl-cookbook/emacs-ide.html is some more info.

And maybe @vindarel 's Common Lisp course has something on initial setup.

@davidaugust

Don't let @scottjenson catch you disseminating defeatist news on AI.

It's utterly your fault that we have this bad reputation on the Fedi with respect to AI.

@xdydx

@scottjenson @pmdj @dalias

> My point is that there is a pattern here: there are topics this community actively hates

My problem here: I cannot just stop advocating against the current strain of AI only because an overwhelming majority here also dislikes it.

I suppose others are in the same position and --- bam! --- you suddenly got a community that dislikes AI.

Looking for root causes, I'd guess that is, because such a lot of creatives (in a wide sense) are here.

@frog_reborn @Weirdaholic @fae2535

This is what I also believe.

@Weirdaholic @fae2535 @Gargron

Since my banking app refuses to run on LineageOS, because some Google security or attestation framework shit is not there, I very much suspect it will be the same for any digital Wallet. So I think we are already there and not 1 business decision away.

@Drums_in_the_night @Gargron @EUCommission

Count me in. I also want to see some consequences.

@_RyekDarkener_ @Gargron

I hope so. Can we do something about this? Like, e.g. burn their[1] sorry asses?

[1] This is pretty large plural. Indeed I want to see everybody who thought this was Ok and participated in such a design decision to be "freed" up for tasks for which they are more suited. Disinfecting phones, for example. And if everything else fails there is still an arch to Golgafrincham we could could offer them a place on.

@mttaggart

"Tim Rogers, principal product manager for Copilot at GitHub, took to Hacker News on Monday to say that giving Copilot the ability to add "tips" to PRs was intended "to help developers learn new ways to use the agent in their workflow."

Hearing feedback from the community following Manson's post and the kerfuffle it generated, Rogers said, has helped him realize that "on reflection," letting Copilot make changes to PRs written by a human without their knowledge "was the wrong judgement call." "

What a dishonest take. "help developers". LOL.

"Was the wrong judgement call".

No. It was an eklatant breach of trust. From which there is no painless recovery. Full stop.

You (remaining clients of github) should just refuse to do business with Github until Tim Rogers has been fired ignominiously (without a golden parachute).

@mttaggart

... until you loose. Frankly, this always follows the same pattern: Provocative change, backing down and after this gauging of the public opinion slowly boiling frog release of the very same feature over a couple of months or years. While our attention is elsewhere.

We only win, if the perpetrators cannot commit future deeds. Going back to the status quo is no win, because the status quo can be attacked again.

@Noortjevee @CStamp @jakebrake @stux

They could just have said "predicting the future doesn't work, suck it up". Unfortunately fantasy on magic is rarely courageous enough to handle the "prophesy of the one who will be our salvation" properly (ie. this way).