<?xml version="1.0" encoding="UTF-8" ?>
<?xml-stylesheet type="text/xsl" href="http://cs.jjbresearch.org/rss.xsl" media="screen"?><rss version="2.0" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:slash="http://purl.org/rss/1.0/modules/slash/" xmlns:wfw="http://wellformedweb.org/CommentAPI/"><channel><title>The Hub goes live</title><link>http://cs.jjbresearch.org/blogs/os/archive/2006/05/12/hubFSGoesLive.aspx</link><description>I have been so busy, I haven't had a chance to announce one of the initiatives that I have been working so hard to complete.&amp;nbsp; So here it is . . .
hubFS - THE place for F#
This is my first effort at an organized community site with focus on F#.&amp;nbsp;</description><dc:language>en-US</dc:language><generator>CommunityServer 2.0 (Build: 60217.2664)</generator><item><title>re: The Hub goes live</title><link>http://cs.jjbresearch.org/blogs/os/archive/2006/05/12/hubFSGoesLive.aspx#3</link><pubDate>Fri, 12 May 2006 13:38:30 GMT</pubDate><guid isPermaLink="false">c1975ec3-f1be-4eba-a03b-f29c6b5007dd:3</guid><dc:creator>Adam</dc:creator><description>Fortran? &amp;nbsp;Really? &amp;nbsp;That's so retro it's awesome.&lt;br&gt;</description></item><item><title>re: The Hub goes live</title><link>http://cs.jjbresearch.org/blogs/os/archive/2006/05/12/hubFSGoesLive.aspx#4</link><pubDate>Fri, 12 May 2006 18:30:29 GMT</pubDate><guid isPermaLink="false">c1975ec3-f1be-4eba-a03b-f29c6b5007dd:4</guid><dc:creator>optionsScalper</dc:creator><description>Adam,&lt;br&gt;&lt;br&gt;Actually, not Fortran. &amp;nbsp;But that too is what I first thought last year when I adopted F#.&lt;br&gt;&lt;br&gt;F# is based on ML/OCaml and is a research language from Microsoft Research Cambridge. &amp;nbsp;While it is not a production language as of yet, it is CLS-compliant, plays well with others in the .NET suite, etc. &amp;nbsp;The language is concise and allows for functional/imperative/OO constructs in any mixture.&lt;br&gt;&lt;br&gt;The reason that I'm interested is that this language will likely become THE math/scientific computing platform for .NET. &amp;nbsp;Because I can seamlessly integrate with all other .NET stuff, I can choose how to deploy F# in an application. &amp;nbsp;I can use F# within SQL Server 2005, interop with C#, do some stuff with it in BizTalk, etc. &amp;nbsp;I can also deploy F# in Web Services and WCF (Indigo). &amp;nbsp;So where F# provides advantage in my scientific computing environment, I can use it and still continue to use my other tools as needed.&lt;br&gt;&lt;br&gt;Having been a big MatLab, Mathematica, Maple, S-Plus, R and other math engine user, I am also glad to move to F#. &amp;nbsp;Data used for any research needs to be shuttled to the compute engine and if I have to cross a program boundary into MatLab or Mathematica for example, I'm beholden to their interface, it's robustness and performance. &amp;nbsp;With F#, I incur none of those penalties and am free to work out architecture and design issues for large scale models.&lt;br&gt;&lt;br&gt;Check out Chad's article that I linked above. &amp;nbsp;He just did a proof of concept to run F# on a Virtual Supercomputer, i.e. deployment of F# compute engines on a grid of machines.&lt;br&gt;&lt;br&gt;---O&lt;br&gt;&lt;br&gt;p.s. F# supports BigInts and BigRationals (arbitrary precision math) as first-class language types. &amp;nbsp;I know the F# team has just completed some work to further optimize those types and related operations. &amp;nbsp;Crypto work that requires large number use becomes easier with F#.&lt;br&gt;</description></item></channel></rss>