--- "K.S. Bhaskar" <[EMAIL PROTECTED]> wrote:

> An open, royalty-free, formal standard would allow anyone to
> implement
> it unencumbered by concerns as to whether the definition of a
> specific
> feature or function violates a copyright or patent....

I think I agree with you up to this point

>  But for standards to be meaningful, the process must be
> open, multiple conforming implementations must exist, users must use
> standard features in preference to non-standard alternatives.  

But here I do not agree. Standards do not exist for the purpose of
*preventing* users from using implementation features that go beyond
the standard. Rather, they exist to *ensure* that applications
following the standard will be portable across conforming
implementations.

> ...
 
> However, let me throw in something to think about.  Let's
> conceptually
> separate the MUMPS database from the MUMPS language.  It seems to me
> that if we define a callable API for the database (one that includes
> not
> just global variables but also M transaction processing and M locks),
> allow access to the database via this API, and allow code in other
> languages to freely call back and forth to and from M code, then an
> application like VistA can use other languages as well as M (not just
> C,
> but also languages like Perl that have much of the expressive power
> of M
> but with more modern features like information hiding). 

Indeed, I would welcome such a development, and I'd be surprised if
there was not considerable interest in making use of it. It is not as
simple as it may appear on the surface, though, because there needs to
be an efficient mechanism for synchronizing processes written in many
languages (both interpeted and compiled). Off-hand, a kernel extension
(and yes, folks they have them in OS X, too) would be the most natural
approach, but a lock manager running in application space is also a
possibility.

> New modules
> can
> be written in the most appropriate language, but can freely
> interoperate
> with modules in other languages.

Exactly. And that would even make MUMPS more attractive to people who
might want to use it as it is no longer an all or nothing commitment.
You can write code in MUMPS today, knowing you are not forever tied to
the language. On the other hand, if you prefer C, Java or Scheme, you
are free to use that language from the start.
> 
> ...
 
> Back to M standards: Fidelity cannot commit to implementing an
> unspecified standard that doesn't exist yet. 

Which is only reasonable, of course. 

> But if there is a
> standards development effort, I will participate enthusiastically if
> there is broad involvement and a community commitment to it.

That's the key. 


===
Gregory Woodhouse  <[EMAIL PROTECTED]>
"Interaction is the mind-body problem of computing."
--Philip L. Wadler


-------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc. Do you grep through log files
for problems?  Stop!  Download the new AJAX search engine that makes
searching your log files as easy as surfing the  web.  DOWNLOAD SPLUNK!
http://ads.osdn.com/?ad_id=7637&alloc_id=16865&op=click
_______________________________________________
Hardhats-members mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/hardhats-members

Reply via email to