News for hackers. Stuff that matters.
Re:Paul Graham: Lisp in Web-Based Applications (2001) by helium on Wednesday May 13, 2009 @04:26AM (#606638)
Counterpoint: Use any language you want, as long as you keep in mind that you might have to re-write libraries that are available to you in more mainstream languages or use inferior or incomplete alternatives.

On many projects this isn't as much of a problem, but I think it's dangerous to assume that one can easily use 'any language' when developing web-based software.

Re:Paul Graham: Lisp in Web-Based Applications (2001) by dchest on Wednesday May 13, 2009 @05:59AM (#606685)
This is not a counterpoint. Actually, I think it's a point of the article: use any language that better suits your needs. If your web app needs a library that's not available in language A, use language B. Or mix languages -- write feature X in language A, and feature Y in language B. The point is -- you can do whatever you want, as you control the environment, unlike with desktop apps.

See, for example, my comment in this thread: http://news.ycombinator.com/item?id=595513

Re:Paul Graham: Lisp in Web-Based Applications (2001) by Retric on Wednesday May 13, 2009 @08:36AM (#606790)
Note the 2001. I think his point had more to do with a lack of library's making any powerful language useful. Ruby on rails vs Lisp is not really in favor of lisp, but there is a reason nobody has good C web frameworks.

PS: Which is something to think about, if you are spending as much money on HW as Google it might be a good idea to use C or other fast but hard to write language for some of your core functions.

Re:Paul Graham: Lisp in Web-Based Applications (2001) by wheels on Wednesday May 13, 2009 @10:34AM (#606941)
Well, in 2001 the hot web languages were Perl and PHP and CPAN was already riding high.

I think this argument in some ways hinges on how far from the tree you're falling. If you're doing essentially a variation on a common theme, you'll be the fastest by picking a language where the bulk of the work is already done for you in a variety of reusable modules. It'd have to be a really poor language for its badness to outweigh having most of the work done for you up front.

If, however, you're starting from scratch on something relatively novel, e.g. where Viaweb was in 1996 (and not so much in 2001), then it's true, you should write in the language that fits the bill rather than following the course of momentum.

One proviso I would add there though is that Lisp is in the position of being fairly mature and non-traditional, unlike many flavor-of-the-weeks. There are many pains to be found in using tools which haven't stabilized yet, even if they do have a gee-whiz-bang factor.

As to your postscript, Google does also extensively use C++; I think that's still fairly common in high-performance stuff at the lower levels.

Re:Paul Graham: Lisp in Web-Based Applications (2001) by helium on Wednesday May 13, 2009 @06:10AM (#606694)
The point of the article is that you can use more powerful, but less used languages when developing web based apps. That hasn't always been feasible when writing software targeting the desktop. Which of course is true, but could have implications of it's own.
Re:Paul Graham: Lisp in Web-Based Applications (2001) by plinkplonk on Wednesday May 13, 2009 @04:42AM (#606647)
"you might have to re-write libraries that are available to you in more mainstream languages or use inferior or incomplete alternatives."

Use (something like) Clojure (or a dot net equivalent)and you can have the best of both - a lisp and lots of libraries. The (minor) price is some grunginess at the interface of the lisp and the libraries, but this is paid only once.

Re:Paul Graham: Lisp in Web-Based Applications (2001) by yummyfajitas on Wednesday May 13, 2009 @11:03AM (#606994)
Try clojure. I'm only getting started with it, but I'm amazed at how clean the interface between lisp and the JVM is.
Re:Paul Graham: Lisp in Web-Based Applications (2001) by jimbokun on Wednesday May 13, 2009 @04:39PM (#607591)
I am commonly using clojure now to interactively test methods in my Java code.
Re:Paul Graham: Lisp in Web-Based Applications (2001) by axod on Wednesday May 13, 2009 @04:44AM (#606650)
Rewriting libraries can work out really well. As well as a great way to learn, you may often end up with something better than the existing libraries - giving you another edge.

I'm not saying you should write everything yourself, just the bits that matter :)

Re:Paul Graham: Lisp in Web-Based Applications (2001) by helium on Wednesday May 13, 2009 @04:56AM (#606657)
Startups often don't have the time or resources available to them to re-write everything. Many open source libraries have been developed by dozens of very smart people over many years, do you really think your solution will be better?
Re:Paul Graham: Lisp in Web-Based Applications (2001) by wglb on Wednesday May 13, 2009 @11:09AM (#607014)
I like the Franz story about NFS. The found it was easier to reimplement NFS from scratch rather than to figure out how to configure the commercial product. Bleeping NFS!

Sometimes it is quicker to write the exact thing you need rather than struggling with something that wants you to think it is the solution to your problem.

Another case in point was the Aha! moment someone described when realized that the Lisp Postgress interface was not going against the library but against the wire protocol itself. Just as easy.

Re:Paul Graham: Lisp in Web-Based Applications (2001) by subbu on Wednesday May 13, 2009 @06:45AM (#606714)
What axod is suggesting is programmers can always write something of their own if there isn't anything they want. They don't have to write everything in the first go. But incrementally. This is how Django and Rails were born. Smart people will join later if they find the base interesting.
Re:Paul Graham: Lisp in Web-Based Applications (2001) by axod on Wednesday May 13, 2009 @05:09AM (#606661)
In specialized spaces that matter 100% to my startup, yes.
Re:Paul Graham: Lisp in Web-Based Applications (2001) by davidw on Wednesday May 13, 2009 @05:16AM (#606665)
That's sort of the point, though, isn't it? You should be concentrating on that stuff, not replicating (or writing from scratch!) N other libraries that you do need, but that aren't the core of your business.
Re:Paul Graham: Lisp in Web-Based Applications (2001) by helium on Wednesday May 13, 2009 @05:27AM (#606670)
Well sometimes it's a case of gluing together the right components and quickly getting it out there. Not everyone has to be the next Google.
Re:Paul Graham: Lisp in Web-Based Applications (2001) by axod on Wednesday May 13, 2009 @09:31AM (#606845)
OTOH though, a lot of things aren't rocket science. For example, if you want to roll your own webserver/framework for good reasons, definitely do it.

But I agree. It's a matter of deciding when it matters, and when it doesn't.

Re:Paul Graham: Lisp in Web-Based Applications (2001) by helium on Wednesday May 13, 2009 @05:19AM (#606666)
Sure. On some projects your advantage is on a technical level. But on others it's time to market. Or an unique idea that's never been tried before.

In these cases it's to your advantage to use the mainstream tools because there are so many resources available to you.

Re:Paul Graham: Lisp in Web-Based Applications (2001) by briancooley on Wednesday May 13, 2009 @09:42AM (#606861)
The question to ask isn't whether the solution will be better, but rather whether the differentiation that gives your product value is derived from rewriting libraries.

Libraries tend to be generic, so you can often do better for your specific problem if you are willing to pay the opportunity cost.

Re:Paul Graham: Lisp in Web-Based Applications (2001) by tsally on Wednesday May 13, 2009 @05:27AM (#606668)
Remember that libraries licensed under the GPL don't count, as far as startups go. You'd be surprised at how many of the libraries you use in hacked up code on a daily basis are ones you can't release with a closed source product.
Re:Paul Graham: Lisp in Web-Based Applications (2001) by plinkplonk on Wednesday May 13, 2009 @05:34AM (#606673)
"You'd be surprised at how many of the libraries you use in hacked up code on a daily basis are ones you can't release with a closed source product"

I thought GPL libs could be used n a web based project? PG's article says (emphasis mine)

"One of the reasons to use Lisp in writing Web-based applications .. "

Re:Paul Graham: Lisp in Web-Based Applications (2001) by jrockway on Wednesday May 13, 2009 @06:14AM (#606696)
The GPL will also kill your entire family when you're not looking. You should use the much better "jrockway is awesome license", which gives you the freedom to give me all your money. I like that one a lot more.

(Also, what is with all the anti-GPL FUD on this site? Have any of you guys read the license, or do you just like to bash it because you don't like Free Software?)

Re:Paul Graham: Lisp in Web-Based Applications (2001) by tsally on Wednesday May 13, 2009 @06:22AM (#606701)
I'm actually very much for the GPL. Before you criticize people for not reading, you might want to do a bit of reading yourself? My comment history clearly places me in the pro-GPL camp, and there's nothing about my parent comment that suggests that I don't like the GPL.
Re:Paul Graham: Lisp in Web-Based Applications (2001) by weavejester on Wednesday May 13, 2009 @08:08AM (#606758)
Well, the second sentence of the talk excerpt reads:

"When you're writing software that is only going to run on your own servers, you can use whatever language you want."

If you're not distributing software then the GPL doesn't apply. A web-based start up can use as many GPL-based libraries as they want.

Re:Paul Graham: Lisp in Web-Based Applications (2001) by gaius on Wednesday May 13, 2009 @05:50AM (#606680)
"Release" being the operative word. You can do anything you like with GPL code you run on your own kit. SAAS is a huge end-run around the GPL.
Re:Paul Graham: Lisp in Web-Based Applications (2001) by randallsquared on Wednesday May 13, 2009 @08:59AM (#606808)
I like the GPL, and I like SAAS. It's not an "end-run around the GPL" unless you believe that the object of the GPL is to stop people from making money from programming. If, instead, you believe the rightful purpose of the GPL is encouraging people to control the software on their own hardware, then SAAS fits in with that goal quite well. Letting people use my computer with my software on it for some task does not make my computer (or my software) theirs, and it shouldn't.
Re:Paul Graham: Lisp in Web-Based Applications (2001) by bodhi on Wednesday May 13, 2009 @08:27AM (#606783)
Isn't the Affero GPL supposed to address the end-run problem? I guess I should check myself really: [Seems it is](http://en.wikipedia.org/wiki/Affero_General_Public_License).
Re:Paul Graham: Lisp in Web-Based Applications (2001) by jcromartie on Wednesday May 13, 2009 @10:21AM (#606919)
I think the best approach is (paraphrasing something I read once): don't reinvent the wheel unless you want to learn more about wheels.
Re:Paul Graham: Lisp in Web-Based Applications (2001) by ken on Wednesday May 13, 2009 @01:35PM (#607237)
Does it often happen that higher-level languages have "inferior or incomplete" versions of libraries? My experience has been the exact opposite: in Common Lisp, for example, when there's a native library, it's invariably more flexible and powerful than in any other language I've used.