I'm putting together a structured wiki that documents what languages, frameworks, and libraries are used by various websites. One of the goals of the wiki is to highlight the diverse set of languages and other components that people can and do use to power their sites.
I'd like to add in some sites that use Lisp and various related frameworks or components like UnCommon Web (UCW), AllegroServe, Hunchentoot, mod_lisp, etc.
Does anyone here have a Lisp powered website? If so, would you mind sharing the URL and what other support components are you using (e.g, UCW, mod_lisp)?
dan <Daniel....@gmail.com> writes: > Does anyone here have a Lisp powered website? If so, would you mind > sharing the URL and what other support components are you using (e.g, > UCW, mod_lisp)?
> Does anyone here have a Lisp powered website? If so, would you mind > sharing the URL and what other support components are you using (e.g, > UCW, mod_lisp)?
Volkan YAZICI <volkan.yaz...@gmail.com> writes: > On Dec 2, 2:37 pm, Zach Beane <x...@xach.com> wrote: >> http://wigflip.com/is powered by TBNL. It's been running odd little >> Lisp-powered graphical amusements since 2005. The most popular >> application is <http://wigflip.com/signbot/>, which is powered by >> Skippy. The most recent is tinytags at <http://wigflip.com/tinytags/>, >> which I'm pleased to say topped delicious.com for a while (see >> <http://flickr.com/photos/xach/3059723046/>) a few days ago.
> Could you mention about the received traffic statistics in > delicious.com effect? How did Hunchentoot stand that load?
It resulted in about 10,000 page views. I don't use Hunchentoot for wigflip.com; TBNL didn't have any trouble with the traffic. The server never got much load as a result.
The movie charts at <http://xach.com/moviecharts/> reached digg.com and stumbleupon.com popularity, and the traffic effect was much, much, much bigger than delicious.com. Those files are generated statically by Lisp, though, and served by nginx, not TBNL.
dan <Daniel....@gmail.com> writes: > I'm putting together a structured wiki that documents what languages, > frameworks, and libraries are used by various websites. One of the > goals of the wiki is to highlight the diverse set of languages and > other components that people can and do use to power their sites.
> I'd like to add in some sites that use Lisp and various related > frameworks or components like UnCommon Web (UCW), AllegroServe, > Hunchentoot, mod_lisp, etc.
> Does anyone here have a Lisp powered website? If so, would you mind > sharing the URL and what other support components are you using (e.g, > UCW, mod_lisp)?
> Does anyone here have a Lisp powered website? If so, would you mind > sharing the URL and what other support components are you using (e.g, > UCW, mod_lisp)?
I wrote a few bioinformatics applications in Lisp:
They are all based on an HTTP server + web development framework that I wrote from scratch. I didn't use any of the existing ones, mainly because... they didn't exist yet when I started :)
> I'm putting together a structured wiki that documents what languages, > frameworks, and libraries are used by various websites. One of the > goals of the wiki is to highlight the diverse set of languages and > other components that people can and do use to power their sites.
> I'd like to add in some sites that use Lisp and various related > frameworks or components like UnCommon Web (UCW), AllegroServe, > Hunchentoot, mod_lisp, etc.
> Does anyone here have a Lisp powered website? If so, would you mind > sharing the URL and what other support components are you using (e.g, > UCW, mod_lisp)?
> Thanks, Dan
Thanks, to everybody who responded both to the group and over e-mail. As soon as I get a chance, I'll add all of the sites.
dan <Daniel....@gmail.com> writes: > Does anyone here have a Lisp powered website? If so, would you mind > sharing the URL and what other support components are you using (e.g, > UCW, mod_lisp)?
Portable Allegroserve/Webactions and CLSQL running CMUCL. All in Norwegian. I haven't had any time to develop it futher but I have had more than 6000 unique hits a month (mostly due to the real estate statistics at http://gratismegler.no/finn).
Petter
-- A: Because it messes up the order in which people normally read text. Q: Why is top-posting such a bad thing? A: Top-posting. Q: What is the most annoying thing on usenet and in e-mail?
http://toonlet.com runs on ucw_ajax on sbcl behind apache/mod_lisp. It uses plenty of libraries, including the usual suspects like clsql, alexandria, mel-base, and stefil.
Our biggest day saw over 20k pageviews. Everything held up pretty well, though i had to shorten the session expiry time for ucw to keep memory usage down on our VPS.
I'd be happy to give more specifics on- or off-list if desired.
> From: dan <Daniel....@gmail.com> > I'm putting together a structured wiki that documents what > languages, frameworks, and libraries are used by various websites. > One of the goals of the wiki is to highlight the diverse set of > languages and other components that people can and do use to power > their sites.
It's not clear whether you're talking about the master HTTP server itself, which in most cases is Apache, or the language used to write individual CGI or similar applications, whereby one user's "web site" might be all written in PHP and another's might be all written in Perl and another's might be all written in Common Lisp, but all three users are on a single ISP sharing a single Apache HTTP server. If you're asking about individual users' Web sites within the overall framework of an Apache HTTP server which all those users share even though the various users use different languages for their CGI applications, my personal answer is that I have (so-far) used CMUCL for all serious applications on my Web site, but have small demos of how to do simple things in several other languages. In the future I plan to switch to PHP for the first-step of getting into services, so that floods from dDOS botnets won't overload the system, and link PHP into CMUCL only after I'm sure a particular connection isn't from a dDOS botnet or other unauthorized automated bulk connection flood. One idea is to keep the URL of the actual CMUCL CGI script secret, and have a different alias to that script for each user online at a given time. The entry-point PHP script then verifies non-botnet, creates a new alias, and redirects to that alias, and all URLs given in subsequent Web pages/forms for that user will specify that same alias again, so that one user will stay with that one alias for the rest of the session. If any particular alias gets overloaded, meaning that the user has switched over to a bot, that alias is disabled, causing subsequent HTTP requests by that user's bot to get 404 NOT FOUND from the Apache server at very little CPU overhead, and then if the sysadmin sees infinite spew of additional HTTP requests to the no-longer-existing URL despite the 404 already sent, the sysadmin can block the relevant client IP numbers at the Apache level or at the router level.
> Does anyone here have a Lisp powered website?
(Ambiguous question, see two interpretations above.)
> If so, would you mind sharing the URL and what other support > components are you using (e.g, UCW, mod_lisp)?
Just regular CGI under Apache here. To see the overall HTTP server configuration as viewed by PHP: <http://www.rawbw.com/~rem/HelloPlus/phpinfo.php> = <? phpinfo(); ?> Let me know if that doesn't answer all your questions about HTTP server configuration here, if you know a way I can get the additional info you seek. In particular, there's mod_macro but no mention of mod_perl or mod_php or even mod_cgi in that report, and I don't know whether that's because phpinfo doesn't report that info, or phpinfo *does* report that kind of info but neither mod_perl nor mod_php is installed here and CGI is built into Apache without need of any mod_cgi. CGI is explicitly mentionned in the report, but perl isn't, so maybe mod_perl isn't installed here and perl is instead running as an ordinary CGI application with a brand-new start-up of the perl interpretor for each HTTP/CGI request that tries to run a script starting with #!/usr/bin/perl A few months or years ago /etc/motd used to mention something about mod_perl or mod_php, but that info has been removed for a long time and I don't remember the details of what it used to say.
By the way, timing evidence seems to indicate that re-starting the same Unix-level application (such as CMUCL) several times in a row is much faster than the first time that same application has been started after a long time of not using it. I'm guessing that's because most of the often-used pages of the core-image (and also any lisp or fasl files it loads after startup) are already swapped from disk into disk-cache and/or RAM if the application has been running recently. Thus mod_lisp isn't really necessary for efficient operation under medium load, plain old CGI is good enough. Still PHP ought to be even faster, both the first time and also repeat times.
When time permits I plan to write some distributed applications that pass public-key encrypted+signed data objects across the net from one component to another, typically between CGI or PHP applications running on this FreeBSD shell account and others running on a Ubuntu shell in the UK.
> From: Lars Rune =?ISO-8859-1?Q?N=F8stdal?= <larsnost...@gmail.com> > http://varefalne.no/ runs on SymbolicWeb > ..this one really only runs on FF and IE based browsers for now..
This Web site seems to really be based on JavaScript, not Lisp. When I go there using lynx, I see a blank screen with one non-blank line: JavaScript needs to be enabled. Please fix your Web site to use Lisp instead of JavaScript for the work, so that it will be usable from here. Then get back to me on the topic below:
> ..it is in Norwegian: > S=C3=B8k =3D Search > Etternavn =3D Last name > Fornvan =3D First name > Fritekst =3D Free text > =20 > Nytt s=C3=B8k =3D New search > Neste side =3D Next page > Forrige side =3D Previous page
(The Unix program 'more' generated the hexadecimal codes above from non-USASCII bytes in your newsgroup article. I have software to parse those 'more' codes to produce the correct 8-bit bytes you actually posted, then to diagnose whether that's UTF-8 or Latin-1 encoding, then convert either to UniCodePoints, then convert any that it knows to brace pictures and the rest to {u+xxxx} notation, but it's not handy at the moment.)
That's close enough to German/Dutch roots that it's actually not terribly difficult to learn. I currently have Web-based CAI to teach English Spanish and Mandarin (pinyan), and am planning to include Chinese characters also. Since several active/productive people on the net are Norwegian, and the composer of two of my three most-favorite classical works (Piano concerto, Sonata) is also Norwegian, maybe you could (after you figure out how to make your Web sites lynx-compatible) help me include Norwegian? Note that for Spanish I already have invented what I call "brace pictures" to represent accented Latin-1 and other characters within US-ASCII text, in order to make languages with accented characters (Spanish, French, German, and also now Norwegian perhaps) accessible to US-ASCII browsers such as I use here. See <http://www.rawbw.com/~rem/NewPub/es1.html> for my ask-for-help (for English/Spanish translation pairs) plea, which allows you to choose to get output in either UTF-8 or brace pictures, so that you can see what I'm talking about, what I mean by "brace pictures".
OT: I'm also trying to figure out a good system for entering Chinese characters if you know what a character looks like but you don't know how it is pronounced, such as might be useful if you see something in a Chinese newspaper (hardcopy) and wish to have it translated to English. Such an entry system would also be useful in a CAI program that requires you to draw the Chinese character that has a specified/given meaning, to prove you really know that character, rather than using multiple-choice where you might just guess which of the five characters is the correct one by eliminating the other four you already know have other meanings. The Chinese-character input system installed on MS-Windows requires you type the pinyan in order to generate the character, which is of no use if you don't know how the character is pronounced. A rough idea of one way to do it is sketched here, selecting strokes one by one, by multiple choice, to build a character: <http://www.rawbw.com/~rem/WAP/ChineseInput/start1.html> An alternative idea I haven't prototyped yet, because it requires a PHP or CGI application that is rather "smart", which requires more work than I have time for currently, would be for the user to check just the corners of a new stroke and have the program guess how to "connect the dots" from upper/left to bottom/right to draw the new stroke. On graphics browser with mouse, checking corners would be best for the user, but in lynx or on cellphone with only cursor-stepping and type-in, multiple-choice would be best. I wonder if anyone can think of an even better idea for either kind of browser platform?
VOT: Really stupid brainwork: Some church in Germany, shown on DW-TV just now, planted a tree upside down, but it grew, and now the church claims that's proof of life after death.
> > From: Lars Rune =?ISO-8859-1?Q?N=F8stdal?= <larsnost...@gmail.com> > > http://varefalne.no/ runs on SymbolicWeb > > ..this one really only runs on FF and IE based browsers for now..
> This Web site seems to really be based on JavaScript, not Lisp.
No. The application is written in 100% Lisp. There is not a single line of custom or inlined or whatever JS-code in that application. It is all done using manipulation of higher-level Lisp "widgets" (CLOS instances).
> When I go there using lynx, I see a blank screen with one non-blank line: > JavaScript needs to be enabled.
I'm not going to say I follow W3C and ECMA (JavaScript-standard) 100% in all places (yet..(#1)), but I'm close and I have 99% coverage (IE, FF, webkit, Opera) when I do not do URL-manipulation like I do in this particular application. URL-manipulation like done in this one only works in FF and IE for now(#2); that's around 92-95% coverage depending on where you look.
> Please..
No. You have already freely voted with your browser and you ended up at 0.1%; the application is explicitly informing you of this via the NOSCRIPT feature. Others have voted with their browsers (which _have_ proper W3C and ECMA support; good things wrt. competition (Flash/Silverlight) and what people actually want wrt. the Internet) and they ended up at 99% and they are also paying (voting, again) me money.
> ..use Lisp instead of JavaScript for the..
The application is written in 100% Lisp. The (interesting) JavaScript is generated by a Lisp library. I include a static jquery.js file (not interesting) to get rid of some of the browser-bugs; again, the rest, the actual application stuff, is generated dynamically 100% by Lisp.
> ..work, so that it will be usable from here.
No. _You_ get to work for me now! I want you to write an application for me which, in real-time, presents numeric data from sensors placed around in this facility we have here. I'd like to have a UI that lets one create alarms that trigger on user-defined thresholds etc. It should work in IE, FF and webkit-based browsers (Chrome/Safari). It must scale reasonably well. I want this by the end of this month.
Stuff like this is what Some Interesting People want from applications (not web-documents, but web-applications). They do not care about what _you_ want or prefer. ..but you do get to use Lisp.
#1: ..and W3C as a platform is evolving; I'm working "close to the edge" in some places. I know I'm sloppy with the already well-defined or implemented details wrt. standards in some places, those are bugs, not features; opposite of NOSCRIPT, which is a feature.
#2: Note that this is caused by browser bugs; it is supposed to work based on W3C/ECMA standards.
..i don't have time to respond to the rest of your post .. there are many ways to read mouse events and generate graphics based on this..
<seeWebInst...@teh.intarweb.org> wrote: >> From: Lars Rune =?ISO-8859-1?Q?N=F8stdal?= <larsnost...@gmail.com> >> http://varefalne.no/ runs on SymbolicWeb >> ..this one really only runs on FF and IE based browsers for now..
> This Web site seems to really be based on JavaScript, not Lisp.
You have to send Javascript to a browser, because that's what they understand as a scripting language. Javascript is needed to have some semi-intelligent behavior in form widgets and the like.
You could send Lisp in the HTTP response, but it wouldn't do much good, would it, now?
Javascript in the output of a website doesn't imply that the server is executing Javascript, any more than HTML in the output implies that the server is executing HTML.
If he removed the Javascript, would you believe that the site is HTML-based rather than Lisp-based?
OK, I clicked on that URL, which took me to a page that looks like what I've shown here: <http://www.rawbw.com/~rem/NewPub/BadWebSites/toonlet1.txt> I don't see any explanation of the purpose of the Web site, nor any link for "About this site" or any such, so I decided my best plan would be to click on this link: Linkname: take the tour! URL: http://toonlet.com/tour That takes me to a page that looks like what I've shown here: <http://www.rawbw.com/~rem/NewPub/BadWebSites/toonlet2.txt> It says "CLICK THE SINGLE-ARROW HERE, AND WE'LL GET STARTED." but I don't see anything that looks like an arrow, so I don't know how to get started with the "tour".
> Sasha
Are you the "queen bee" in an early episode of Smallville? How do you measure the covariance of a swarm of bees? Are you related to Tovar, my partner in the SpaceWar team tournament?
> OK, I clicked on that URL, which took me to a page that looks like > what I've shown here: > <http://www.rawbw.com/~rem/NewPub/BadWebSites/toonlet1.txt> > I don't see any explanation of the purpose of the Web site, nor any > link for "About this site" or any such, so I decided my best plan > would be to click on this link: > Linkname: take the tour! > URL: http://toonlet.com/tour > That takes me to a page that looks like what I've shown here: > <http://www.rawbw.com/~rem/NewPub/BadWebSites/toonlet2.txt> > It says "CLICK THE SINGLE-ARROW HERE, AND WE'LL GET STARTED." > but I don't see anything that looks like an arrow, so I don't know > how to get started with the "tour".
> > Sasha
> Are you the "queen bee" in an early episode of Smallville? > How do you measure the covariance of a swarm of bees? > Are you related to Tovar, my partner in the SpaceWar team tournament?
Here, here, Look at The Lispers. This is how they support their peers who actually use Lisp for something which is actually made use of in the real world.
Next, you'll be telling me/us how I/we do not understand (you):
* Lars Rune Nøstdal <1228775795.29169.352.ca...@blackbox.nostdal.org> : Wrote on Mon, 08 Dec 2008 23:36:35 +0100:
| | No. _You_ get to work for me now! I want you to write an application for | me which, in real-time, presents numeric data from sensors placed around | in this facility we have here. I'd like to have a UI that lets one | create alarms that trigger on user-defined thresholds etc. It should | work in IE, FF and webkit-based browsers (Chrome/Safari). It must scale | reasonably well. I want this by the end of this month. | | http://sw.nostdal.org/pg-888
I can't see this --- not going to enable JS to see it.
None of what youve stated seems to REQUIRE any of the Web2.0 technolgies you are using lisp for.
Is there any requirement other than "SHOULD USE WEB2.0 TECHNOLOGIES"?
|
