Because not enough of the people here have their Lips on my weenie.
-
I'm glad to see you took a neutral position on this.
Actually....hmmn, OK, I guess that was a bit harsher than I meant it to be. It's just that AFAIK, Graham has done very little with the language since the 2008 release other than maintain the Y-Combinator web site, most of which he'd already written when he released the language. The reference implementation is very minimally documented, and is tightly bound to Dr Scheme - not Racket, but to the version before it was changed, and it was only a few months later that Racket changed enough that the tarball on his web site no longer worked. The only libraries he's released so far were the ones for the website, and a Bayesian mailfilter, last I checked. It's a start, but not a stellar one, and with him busy on other issues I don't see him coming back to it any time soon.
I'd love to see Arc catch on, even if it is meant for the same niche as my own language, but... well, it just doesn't seem to be going anywhere. Am I wrong? I don't see anyone talking about it anymore, not even Graham himself, but maybe I just not looking hard enough.
I'll admit that my own ideas are very different from his. I don't see that as a conflict; there's enough room for more Lisp languages, provided that the ones that arise have enough going for them. I'm not sure if Arc does or not. Its got potential, so far that potential isn't tapped. I dunno.
breezed in: -abarker
-
Arc, grounded out, neutral... Whoosh, dude. Not that your other points aren't valid.
-
OK, in that case I deserve it because... za?
-
Breezy.
-
Not sure what you mean by that...
If you're implementing on an x86 (or related) machine, the base “language” must be x86 machine code. That's defined as a sequence of bytes, not as a collection of S-expressions. There therefore must be a level beneath the Lisp, and so the system cannot be fully homoiconic when running on an x86. The same argument applies to all other current widely-used platforms. The only platform I'm aware of that really had the full homoiconicity was the LispMachine, which really did run Lisp natively, but who uses one of those now?
The moral of this is don't get too hung up on claiming that Lisp is super-special. It's not really, other than by virtue of being a pioneer (and getting a number of things really horribly wrong too; pioneering isn't the easy path).
-
-
Paging @ScholRLEA...
-
An interesting page that. Arguably there are a whole bunch of C-variants (especially C++, Objective-C, Java and C#, in order of initial creation) but the main adherents mostly agree to disagree. They also definitely agree that delivering actual useful code beats out having flamewars over which is best. It's this last point that seems to have escaped so many Lisp programmers.
Back when I started programming professionally, the only fast-to-start Lisp I had available was in Emacs. The others were large environments that you started up in the morning before going and having a long slow coffee, while not being stable enough to leave running overnight. They were also not usefully documented; there was plenty of docs, but no way to find anything relevant. So I abandoned ship for a “toy” language that happened to be far faster and delivered far more value to everyone. Who lost out? (Holy moly. LispWorks is still a supported product. I wonder who was buying it back then in sufficient numbers to justify keeping it alive and doing future releases…)
-
Arguably there are a whole bunch of C-variants (especially C++, Objective-C, Java and C#, in order of initial creation) but the main adherents mostly agree to disagree.
Also, no one refers to them all generically as "C". Scheme, Clojure and Common Lisp are at least as far apart as the languages you listed, but it's common for people to refer to them generically as "Lisp".
Back when I started programming professionally, the only fast-to-start Lisp I had available was in Emacs. The others were large environments that you started up in the morning before going and having a long slow coffee, while not being stable enough to leave running overnight. They were also not usefully documented; there was plenty of docs, but no way to find anything relevant.
How long ago was that?
-
Also, no one refers to them all generically as "C".
True that. But they're definitely evolutions of it, and will merrily admit the fact.
How long ago was that?
I'm trying to forget that.
-
Also, no one refers to them all generically as "C".
C# doesn't claim to be C or a C-variant. The most it says is it adopts C conventions. From Wikipedia, referring to C#'s ECMA standard doc (the only place it refers to C or C++):
Portability is very important for source code and programmers, especially those already familiar with C and C++.
Closure on the other hand, in the second paragraph on its homepage says:
Clojure is a dialect of Lisp, and shares with Lisp the code-as-data philosophy and a powerful macro system.
So anyway. I guess the reason people don't call C# a C is because C# itself doesn't claim to be a C. Clojure, however, does claim to be a LISP.
-
The most it says is it adopts C conventions.
It was C based originally, then it evolved. It's still definitely C-family; the fingerprints are all over the syntax and much of the semantics.
-
I never said it wasn't based on C, I said it doesn't claim to be based on C.
So good job responding to those invisible aliens on your shoulder, but I don't understand why you thought it was worth sending me a notification for.
-
I don't understand why you thought it was worth sending me a notification for.
I turn all that shit off as much as I can because Discourse by email is far more annoying than Discourse by web. (It could be worse I guess: Discourse by email on iOS, with networking supplied by MilwaukeePC…) Would I suggest that others do this too? Perhaps…
The whole point of my post was that “claim to be” vs “claim to be derived from” is a very murky area. Also, much of the scholarship in this area hardly meets the name. Something about the history of programming languages makes those who study these things lose all sight of what is and what isn't the case. It also makes them myopic as hell. Probably because if they knew what they were doing, they wouldn't be studying that topic and would be actually earning some real money.
-
So anyway. I guess the reason people don't call C# a C
That's not what I said.
Also, no one refers to them all generically as "C".
I'm not talking about people who say that Scheme, for example is a Lisp. I'm talking about people who say that Scheme is Lisp, and ignore the other Lisps, assuming they knew about them at all.
-
Lead developer of EMACs resigns, cites "that fucking LISP bullshit" as the reason.
-
cites "that fucking LISP bullshit" as the reason.
But I'll leave these decisions to someone else, because I also take this
opportunity to step down as head maintainer.It's time for me to move on, and it's time for new blood to take
the lead. I'm not about to disappear, but I won't be reading
emacs-devel (nor bug-gnu-emacs) at least for a while, so if you need
something from me, put me explicitly in the Cc.Thank you all for bearing with me as head maintainer, it's been a great
ride, I hope you enjoyed it as much as I did.
?BTW, Emacs Lisp does suck, and everyone knows it, but there would be a shitton of code that would have to be rewritten if they switched to another Lisp.
-
?Yeah, well, Blakey seems to be listen to - what is he always calls them? - shoulder aliens on this. Nowhere in that thread does Monnier or anyone else say anything about Lisp, good, bad, or indifferent.
BTW, Emacs Lisp does suck, and everyone knows it, but there would be a shitton of code that would have to be rewritten if they switched to another Lisp.
True. Saying ELisp sucks is about as controversial as saying Turbo C++ sucks, and programmers familiar with other Lisps are the ones saying it the most. In 1985, ELisp seemed, well, acceptable at any rate; in 2015, it is archaic. The Emacs developers have been looking to switch to another dialect - preferably one with lexical scoping - for at least twenty-five years, but the sheer volume of code that would have to be re-written is daunting. Even if they could expect to use automated re-writing tools to fix 90% of it the remaining 10% is huge and often long out of date.
-
(use-package 'cl)is usually the first line in any modern elisp package, just to get to sometinhg closer to common-lisp.I must confess that I have only ever worked with elisp, so I don't know any better, but I think it is rather decent. But maybe that's just lisp for you...
-
Well, like with all religious and ideological fanatics, programmers in general are more likely to go after the heretic or deviationist next door than the heathen down the street. Lispers complaints about ELisp being a bad Lisp often involve details that don't seem much of a much to non-Lispers, but the same is true about the details of most forks or schisms to anyone not caught up in it.
I mean, does it really matter that much if you are saying 'similar in form to God' (ὁμοιούσιος) or 'is of the same form as God' (ὁμοούσιος)? Yet a lot of early Christians killed each other over that one letter, in a debate most non-Christians would have found incomprehensible.
-
Lispers complaints about ELisp being a bad Lisp often involve details that don't seem much of a much to non-Lispers, but the same is true about the details of most forks or schisms to anyone not caught up in it.
Which is why the surest way to piss off a Common Lisp user is to take a complaint that would be perfectly valid against Scheme and apply it to Common Lisp. Bonus points for referring to LISP instead; after all, Scheme is a Lisp, Common Lisp has Lisp in the name, how different could they be?
