01:23:30
##forth
<LispBot860>
Chargé 7 commande(s)
01:31:37
##forth
<LispBot860>
Chargé 7 commande(s)
09:36:33
##forth
<tpnix>
good video,very interesting imho
13:34:52
##forth
<KipIngram>
Hans has quite a few interesting videos.
13:35:01
##forth
<KipIngram>
I binge watched his channel one weekend a few months ago.
13:36:44
##forth
<KipIngram>
I particularly got a laugh from that bit where the young couple are walking along in one direction and pass anotoher pretty girl walking the other way, who the young man is ogling over his shoulder, quite upsetting his companion. The young man is labeled "Forth Novice." His companion is labeled "The Stack." And the other young lady - the "ogle-ee" - is labeled "Local Variables." :-)
13:37:54
##forth
<KipIngram>
I think I posted a screen shot of it here, actually.
13:40:17
##forth
<KipIngram>
In one video Hans is pretty harsh on a paper written by someone named Jose Betancourt. I actually had a professional colleague by that name some years back and couldn't help wondering if the paper had been his work. I sent him a message on LinkedIn and asked him, but never heard anything back. I could easily have pictured him being interested in Forth, and particularly being interested in the
13:40:20
##forth
<KipIngram>
sort of things Hans was bearing down on.
13:43:56
##forth
<DKordic>
KipIngram: But isn't "swap", for example, yet another Variable (not meant to be updated in-place)? "": an-example Value a-local ... ;""?! "": an-example : local-data ... ; ... ; is even more general from FP PoV!
13:46:04
##forth
<DKordic>
BTW De Bruijn Index is Normal Form of Identifiers.
13:47:27
##forth
<DKordic>
Shifting (said) Indices _is_ _explicit renaming_.
13:55:22
##forth
<DKordic>
[[smlckz]]: "[ dfn ( swap : D.1 D.2 ; ) ( : D.2 D.1 ; ) ]" or "[ dfn ( swap < D.1 D.2 > ) ( < D.2 D.1 > ) ]" (identical values)?
16:09:32
##forth
<KipIngram>
DKordic: I'm regarding variables as allocated RAM - data on the stack is an entirely different thing, because that space is used dynamically and released when you're done with it.
16:12:21
##forth
<DKordic>
I think such decisions require clarification, and beyond that formalization. So far it is formalized, and therefore clarified, only in AMD64 Machine Code.
16:19:41
##forth
<DKordic>
Key point is Inherent Complexity. It can be balanced with Interfaces. Example: You could theoretically do all the (UNIX) feti.sh to BlinkLED, as long as it's only connection to reality is MMIO Interface it's Inherent Complexity will be limited. And possibly optimzed into just a few bytes of FLASH.
16:31:36
##forth
<DKordic>
Cheat Code is W/Partial_evaluation and Futamura Projections.
16:43:05
##forth
<ghodawalaaman_>
Hello o/
16:50:39
##forth
<DKordic>
Welcome ghodawalaaman_.
16:57:04
##forth
<ghodawalaaman_>
:)
21:30:04
##forth
<[[smlckz]]>
What was that about? Oh, you're going into the concatenative languages' territory..
22:36:13
##forth
<KipIngram>
DKordic: I try to use variables in Forth only when the data I'm saving in them is actually "long term data"; I avoid using them merely to reduce stack activity. I invented an alternative mechanism for that (which I also try not to use unless absolutely necessary, but at least it still IS stack-based and therefore only makes transient use of storage).
22:37:07
##forth
<KipIngram>
It would be legitimate to call those stack cells local variables, but they're extremely cheap and efficient so I find it an acceptable approach.
22:37:31
##forth
<KipIngram>
I waste no energy trying to apply names to them.
22:37:54
##forth
<KipIngram>
Though if you wanted to complicate the compiler and extract names from stack comments you could.
22:39:23
##forth
<KipIngram>
And that would at least have the advantage of adding additional work only at compile time - not runtime.