[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: [Stds-754] Language-oriented vs primitive-oriented



COBOL did add "common exception handling" in its 2002 Standard.  I believe
(but am not certain) that one implementor has STARTED providing it - but in
general the "customers" could care less.

One issue with how COBOL did it, was that it (like the MUCH older PL/I
version) caused a "transfer of control" to an exception routine (based on
the exception detected).  In *theory* one could "get back" to where one
started, but not easily and probably NOT being able to "repair" the original
problem.

If anyone would like more information on the existing COBOL model, let me
know (on or off-list) 

-----Original Message-----
From: stds-754-bounces@xxxxxxxxxxxx 
[mailto:stds-754-bounces@xxxxxxxxxxxx] On Behalf Of Keith Bierman
Sent: Friday, August 04, 2006 11:32 AM
To: Michel Hack (1-914-784-7648)
Cc: (backup); stds-754
Subject: Re: [Stds-754] Language-oriented vs primitive-oriented

Michel Hack (1-914-784-7648) wrote:
My interpretation of the lack of proper language support 
for 754 (1985)
is inertia, and the general unease of some languages to deal with
exception handling -- perhaps because older languages like PL/I went
overboard and left a bad taste...
  
Try attending some Language standardization meetings and work on 
proposals. Fortran spent a decade, and the effort finally was 
given up 
in favor of programmatic testing of flags and such (in 
spirit, similar 
to the existing practice on many Unix systems). This didn't happen 
because any of the participants believed it was *better*, it happened 
because despite years of trying, getting detailed agreement on how a 
refined exception handling facility ought to be specified in 
the Fortran 
context. It's not as easy as it looks.
I fully agree that the new standard should strive to 
prevent this from
happening again, and should therefor give explicit guidance and even
mandatory requirements to the providers of programming environments.
  
As far as I know IEEE lacks the standing to provide *mandatory* 
requirements to ISO. Calling them such will probably just annoy the 
people who have to actually do the work, and may just choose 
to ignore 
your Standard entirely if "mandatory" requirements make it 
impractical 
or just plain hard. LIA has pretty much that status, even 
though it IS 
an ISO group which nominally has the authority to place mandatory 
requirements on other ISO bodies.
 fickleness of voting 
A problem with the Language committee's have, often worse than the 
floating point committee.
m

-- 
Keith H. Bierman    keith.bierman@xxxxxxx|
Sun Microsystems PAE                     | khbkhb@xxxxxxxxx
500 Eldorado Blvd UBRM05-169             | 650-352-4432 voice+fax
Broomfield, CO 80021                     | sun internal 68207
http://blogs.sun.com/khb
<speaking for myself, not Sun*> Copyright 2006
------------------------------------------------------------
N
_______________________________________________
Stds-754 mailing list
Stds-754@xxxxxxxxxxxx
http://mailman.oakapple.net/mailman/listinfo/stds-754


754 | revision | FAQ | references | list archive