www.digitalmars.com         C & C++   DMDScript  

digitalmars.D - Sun's "Fortress" language

reply Bill Baxter <dnewsgroup billbaxter.com> writes:
Many of you probably already saw this on Slashdot, but apparently Sun's 
working on some sort of "Fortran-killer" by the name of Fortress.
   http://research.sun.com/projects/plrg/
I know there are a few others here who are interested in numerical 
computing so I thought I'd repost it here.  Guy Steele, Jr. is the 
principal investigator behind the project.  Wasn't he a big Lisp guy?

Anyway, it looks kind of interesting.  This is supposedly actual 
Fortress source code:
    http://research.sun.com/projects/plrg/faq/NAS-CG.pdf

Looks pretty much like a LaTeX typeset algorithm.  So seems like you'll 
need a fancy editor to program in Fortress.  FAQ says it uses unicode 
symbols for variables and operators.

Looks like it's still in the very early stages.  The language is meant 
to be cross-platform, but their current implementation targets "the JVM".

--bb
Jan 15 2007
parent reply Steve Horne <stephenwantshornenospam100 aol.com> writes:
On Tue, 16 Jan 2007 10:21:07 +0900, Bill Baxter
<dnewsgroup billbaxter.com> wrote:

Anyway, it looks kind of interesting.  This is supposedly actual 
Fortress source code:
    http://research.sun.com/projects/plrg/faq/NAS-CG.pdf

Looks pretty much like a LaTeX typeset algorithm.  So seems like you'll 
need a fancy editor to program in Fortress.  FAQ says it uses unicode 
symbols for variables and operators.
Interesting. I've had the idea that programming languages should use more sophisticated source documents than plain text for some time. It's been a long time since the days when word processors used a plain text view with special notations for style changes, and we all know that WYSIWYG word processing works well, so why not get some similar benefits for programmers - and I don't just mean syntax highlighting. That said, I looked at this listing and it was pretty wierd. I suppose part of that is the fact that the listing looks more like math than code, though, which makes a lot of sense for a language taking on a similar role to Fortran WRT numerical work. It's not really a new idea, though. Mathematica has its own built in programming language, and IIRC that is presented on screen formatted like mathematical notation. The saved source files are plain text, but this isn't quite the source as input by the user in mathematica - there's a translation between the formatting-based syntax and plain text equivalents, which is a pretty good idea. Exaggerating the principle somewhat, it can be compared with OpenOffice which uses XML files (which can be considered plain text) rather than binary files to represent the on-screen/page formatting - the two representations of the document are not the same, but they do represent the same thing. Mathematica uses a much simpler translation to plain text than XML, which is much easier and much more practical to edit in a text editor, and that is the kind of thing I'd like to see in a programming language that uses formatting as part of its syntax. -- Remove 'wants' and 'nospam' from e-mail.
Jan 16 2007
parent "Ralf Schneider" <ralfs72_at_ gmx.net> writes:
 It's not really a new idea, though. Mathematica has its own built in
 programming language, and IIRC that is presented on screen formatted
 like mathematical notation. The saved source files are plain text, but
 this isn't quite the source as input by the user in mathematica -
 there's a translation between the formatting-based syntax and plain
 text equivalents, which is a pretty good idea.
If you look at the Fortress slides at: http://research.sun.com/projects/plrg/PLDITutorialSlides9Jun2006.pdf you will see there exists a pure ascii version of the same file. IMHO the mapping from ASCII to the typed document is were well done. But anyway, I can not imagine such a language would be used in wide areas. It looks to "Mathematic". And in addition I have doubts a more complex formating tool than a text editor is very productive. May be someday a real / useful graphical programming tool will be born, but it's probably a long way until there.
Jan 16 2007