4 ms·

Lisp: The Golden Age Isn't Coming Back, Let's Welcome a Bright Future

16 points by pchristensen 19y ago

jimbokun 19y ago
I think this nails the challenge of getting started with Common Lisp: 'With Lisp, newbies say "How do I get started?" and instead of a URL, they get "Lisp is a family of languages, with the primary branches being Common Lisp and Scheme. But these are just language standards with a number of implementations. Which implementation is best for you depends on your platform and computing needs." Huh?' I think SBCL is closest to being the default CL implementation, but the functionality on Mac/Windows still isn't quite caught up to the Linux version. Threading seems to be the biggest sticking point. If you have a Mac, the best starting point I know of is ReadyLisp: http://www.newartisans.com/software/readylisp.html http://www.newartisans.com/software/readylisp.html
- pivo 19y ago
  OpenMCL (http://clozure.com/ http://clozure.com/) works great on the Mac and 64 bit linux too. Threads and all.
- cstejerean 19y ago
  A common misconception with SBCL is that it cannot handle threading on OSX. The feature has been there for quite some time and even though it's marked as experimental, don't let that stop you. It should work just fine, especially for development. If you're curious about how to turn on threading you can see a post I wrote on my blog a little while ago (or you can just read the INSTALL file SBCL comes with). http://blog.offbytwo.com/2007/12/03/installing-sbcl-with-threads-on-os-x-leopard/ http://blog.offbytwo.com/2007/12/03/installing-sbcl-with-thr...
mechanical_fish 19y ago
This is nice. Thanks for the various links and pointers; I dream of the day that I have time to sit back down with Lisp and finally grok the mysteries of the macro, and articles like yours will help. Language designers seem to agree that the secret to giving Lisp a bright future is to invent a brand new Language of the Future that just happens to encompass Lisp. This is not exactly a new argument around here, since this very site is written in a FutureLisp candidate. Other aspiring FutureLispers include Guy Steele and Matz (http://www.paulgraham.com/quotes.html http://www.paulgraham.com/quotes.html). The thing to recognize is that Lisp's storied history is not an advantage. It's actually a big problem for the language. People like new languages. They're not big yet. You can grasp them. Their core team is young and obscure enough that you have a fighting chance to join it. They haven't accumulated decades' worth of parasitic features. They have no installed base that must be accommodated. Their design flaws are not yet so old that they've been enshrined as advantages. They are not steered by a committee of middle-aged managers. They haven't spawned so many children that you can't decide between them all. The experts haven't yet had the chance to become intimidatingly skilled, let alone entrenched and opinionated, let alone bitter and burnt out. The good news is that each decade's new language is a bit more Lispy than the last. Ruby seems likely to tide me over nicely until it's finally time to jump on the Arc bandwagon.
- pchristensen 19y ago
  Since pg has been pretty mum on Arc and he has designed it more as a pure language than a stack, I'm interested in Qi from Lambda Associates (http://www.lambdassociates.org/lC21.htm http://www.lambdassociates.org/lC21.htm). That sounds like a balanced approach that solves the library problem and the implementation problem nicely. Hopefully pg will surprise us (this winter, he says! http://news.ycombinator.com/item?id=98330 http://news.ycombinator.com/item?id=98330 )
  - mechanical_fish 19y ago
    Well, this is good news. It would be nice to have FutureLisp ready for prime time when the Ruby programmers (weaned on closures and metaprogramming!) get frustrated with their stale, old, crufty language sometime around the year 2015.
- cstejerean 19y ago
  I think the key to success for any FutureLisp is to not call itself a Lisp dialect. Unfortunately the word itself has developed a negative connotation for most people.
  - pchristensen 19y ago
    Some cool name like ... Arc? I can't help but think that pg noticed that the name Lisp definitely has baggage.
Tichy 19y ago
I've restarted learning Scheme now, and found an Eclipse plugin: http://schemeway.sourceforge.net/ http://schemeway.sourceforge.net/ I don't know how it compares to other Scheme environments, but it seems to work well enough. It bundles Kawa and SISC, but can apparently also be used with external scheme interpreters like MzScheme.
- herdrick 19y ago
  That plugin was a little buggy when I used it this summer. If you do something with MzScheme you're much better off with DrScheme. DrScheme is pretty good, and getting better - they just added rudimentary code completion, for example. You can just use DrScheme with Kawa, too, but you lose the REPL (which is not acceptable in my opinion).
  - Tichy 19y ago
    I don't know, the editor just seemed ugly. But perhaps I'll give it another shot.
bobochan 19y ago
This was a good article and I enjoyed reading it, but I have to say I disagree with the central premise. I love the "family of languages" aspect of Lisp and the incredible amount of innovation that is going on with Lisps and functional programming. Lisp has become like Jedi training, you aren't a real Lisp geek until you have built your own light saber (aka rolled your own Lisp dialect). The story should be to tell new people that part of using Lisp embracing this richness. There are implementations that can get people started in a hurry and anyone with ten minutes and access to a search engine can find them. I used to post and even answer some questions on comp.lang.lisp and I honestly cannot understand where it got the reputation that it seems to have. I can't tell you how many incredibly helpful messages I received from people correcting some of my misconceptions about macros and even helping me track down a type in one of the books I was learning from. The regular posters there really just had a few simple rules that were not hard to live by, e.g. don't ask the group to do your homework, try and make some effort on a problem before asking for help, don't continually make inflammatory claims about why Language X or Feature Y are so much better when you have only been using the language for three days.
- pchristensen 19y ago
  Yeah, I agree with your points. In the article, I was trying to summarize the impression that is conveyed by reading what people have written about Lisp. I actually had the Jedi analogy in a draft but took it out for length - small in number, deceptively powerful, greater than several people armed with clumsier weapons, etc. As for c.l.l, I've read enough people express trepidation to go there. I think it's mostly the tongue-in-cheek fear that CS students have of compiler class. Also, I don't doubt that most posters are courteous and helpful, but the anecdote about the noob who asked a dumb question and got a script to format his hard drive traveled pretty far. The 1% of people with bad experiences are just a lot more vocal than the rest with acceptable or positive experiences. I think that if there was an easier setup and more thorough online documentation, fewer people would go to c.l.l for help.
  - gruseom 19y ago
    the anecdote about the noob who asked a dumb question and got a script to format his hard drive traveled pretty far The exact same complaints about the Lisp community were being repeated just as much before that incident as after it. I remember reading that hard-drive post the day it came out and thinking, this is really going to please a lot of people - now they'll have something canonical to point to.

16 points by pchristensen 19y ago

I think this nails the challenge of getting started with Common Lisp: 'With Lisp, newbies say "How do I get started?" and instead of a URL, they get "Lisp is a family of languages, with the primary branches being Common Lisp and Scheme. But these are just language standards with a number of implementations. Which implementation is best for you depends on your platform and computing needs." Huh?' I think SBCL is closest to being the default CL implementation, but the functionality on Mac/Windows still isn't quite caught up to the Linux version. Threading seems to be the biggest sticking point. If you have a Mac, the best starting point I know of is ReadyLisp: http://www.newartisans.com/software/readylisp.html http://www.newartisans.com/software/readylisp.html

pivo 19y ago
OpenMCL (http://clozure.com/ http://clozure.com/) works great on the Mac and 64 bit linux too. Threads and all.
cstejerean 19y ago
A common misconception with SBCL is that it cannot handle threading on OSX. The feature has been there for quite some time and even though it's marked as experimental, don't let that stop you. It should work just fine, especially for development. If you're curious about how to turn on threading you can see a post I wrote on my blog a little while ago (or you can just read the INSTALL file SBCL comes with). http://blog.offbytwo.com/2007/12/03/installing-sbcl-with-threads-on-os-x-leopard/ http://blog.offbytwo.com/2007/12/03/installing-sbcl-with-thr...

mechanical_fish 19y ago

This is nice. Thanks for the various links and pointers; I dream of the day that I have time to sit back down with Lisp and finally grok the mysteries of the macro, and articles like yours will help. Language designers seem to agree that the secret to giving Lisp a bright future is to invent a brand new Language of the Future that just happens to encompass Lisp. This is not exactly a new argument around here, since this very site is written in a FutureLisp candidate. Other aspiring FutureLispers include Guy Steele and Matz (http://www.paulgraham.com/quotes.html http://www.paulgraham.com/quotes.html). The thing to recognize is that Lisp's storied history is not an advantage. It's actually a big problem for the language. People like new languages. They're not big yet. You can grasp them. Their core team is young and obscure enough that you have a fighting chance to join it. They haven't accumulated decades' worth of parasitic features. They have no installed base that must be accommodated. Their design flaws are not yet so old that they've been enshrined as advantages. They are not steered by a committee of middle-aged managers. They haven't spawned so many children that you can't decide between them all. The experts haven't yet had the chance to become intimidatingly skilled, let alone entrenched and opinionated, let alone bitter and burnt out. The good news is that each decade's new language is a bit more Lispy than the last. Ruby seems likely to tide me over nicely until it's finally time to jump on the Arc bandwagon.

pchristensen 19y ago
Since pg has been pretty mum on Arc and he has designed it more as a pure language than a stack, I'm interested in Qi from Lambda Associates (http://www.lambdassociates.org/lC21.htm http://www.lambdassociates.org/lC21.htm). That sounds like a balanced approach that solves the library problem and the implementation problem nicely. Hopefully pg will surprise us (this winter, he says! http://news.ycombinator.com/item?id=98330 http://news.ycombinator.com/item?id=98330 )
- mechanical_fish 19y ago
  Well, this is good news. It would be nice to have FutureLisp ready for prime time when the Ruby programmers (weaned on closures and metaprogramming!) get frustrated with their stale, old, crufty language sometime around the year 2015.
cstejerean 19y ago
I think the key to success for any FutureLisp is to not call itself a Lisp dialect. Unfortunately the word itself has developed a negative connotation for most people.
- pchristensen 19y ago
  Some cool name like ... Arc? I can't help but think that pg noticed that the name Lisp definitely has baggage.

Tichy 19y ago

I've restarted learning Scheme now, and found an Eclipse plugin: http://schemeway.sourceforge.net/ http://schemeway.sourceforge.net/ I don't know how it compares to other Scheme environments, but it seems to work well enough. It bundles Kawa and SISC, but can apparently also be used with external scheme interpreters like MzScheme.

herdrick 19y ago
That plugin was a little buggy when I used it this summer. If you do something with MzScheme you're much better off with DrScheme. DrScheme is pretty good, and getting better - they just added rudimentary code completion, for example. You can just use DrScheme with Kawa, too, but you lose the REPL (which is not acceptable in my opinion).
- Tichy 19y ago
  I don't know, the editor just seemed ugly. But perhaps I'll give it another shot.

bobochan 19y ago

This was a good article and I enjoyed reading it, but I have to say I disagree with the central premise. I love the "family of languages" aspect of Lisp and the incredible amount of innovation that is going on with Lisps and functional programming. Lisp has become like Jedi training, you aren't a real Lisp geek until you have built your own light saber (aka rolled your own Lisp dialect). The story should be to tell new people that part of using Lisp embracing this richness. There are implementations that can get people started in a hurry and anyone with ten minutes and access to a search engine can find them. I used to post and even answer some questions on comp.lang.lisp and I honestly cannot understand where it got the reputation that it seems to have. I can't tell you how many incredibly helpful messages I received from people correcting some of my misconceptions about macros and even helping me track down a type in one of the books I was learning from. The regular posters there really just had a few simple rules that were not hard to live by, e.g. don't ask the group to do your homework, try and make some effort on a problem before asking for help, don't continually make inflammatory claims about why Language X or Feature Y are so much better when you have only been using the language for three days.

pchristensen 19y ago
Yeah, I agree with your points. In the article, I was trying to summarize the impression that is conveyed by reading what people have written about Lisp. I actually had the Jedi analogy in a draft but took it out for length - small in number, deceptively powerful, greater than several people armed with clumsier weapons, etc. As for c.l.l, I've read enough people express trepidation to go there. I think it's mostly the tongue-in-cheek fear that CS students have of compiler class. Also, I don't doubt that most posters are courteous and helpful, but the anecdote about the noob who asked a dumb question and got a script to format his hard drive traveled pretty far. The 1% of people with bad experiences are just a lot more vocal than the rest with acceptable or positive experiences. I think that if there was an easier setup and more thorough online documentation, fewer people would go to c.l.l for help.
- gruseom 19y ago
  the anecdote about the noob who asked a dumb question and got a script to format his hard drive traveled pretty far The exact same complaints about the Lisp community were being repeated just as much before that incident as after it. I remember reading that hard-drive post the day it came out and thinking, this is really going to please a lot of people - now they'll have something canonical to point to.