| Re: What does an optimal scientific programming language/environment need? [message #36502] |
Mon, 22 September 2003 11:44 |
Richard Maine
Messages: 1
Registered: September 2003
|
Junior Member |
|
|
Duane Bozarth <dp_bozarth@swko.dot.net> writes:
> Well, since F77 there is little that has actually been removed and a
> major consideration (as is evidenced in converstions in c.l.f) is
> maintaining compatability w/existing code. In practice, virtually
> nothing is ever removed from a commercial compiler although most have
> switches to allow specific standard level violations to be flagged...
Note that the "since f77" applies to the whole paragraph. Whether
you intended it to or not, I'm unsure; but it needs to. Several
f66 features were dropped from compilers. For example, I'm not
sure whether you can still find any compilers that support the
extended range of a DO loop; certainly most compilers don't.
There is actually quite a list of obscure f66 behaviors that f77
was incompatible with.
Though the most widely cited example of f66 features, which some
commercial compilers still do support, was never actually an f66
standard feature at all. It was a nonstandard and nonportable
practice used by some codes and supported by some compilers. I'm
referring, of course, to the behavior of zero-trip DO loops.
--
Richard Maine | Good judgment comes from experience;
email: my first.last at org.domain | experience comes from bad judgment.
org: nasa, domain: gov | -- Mark Twain
|
|
|
|
| Re: What does an optimal scientific programming language/environment need? [message #36504 is a reply to message #36502] |
Mon, 22 September 2003 08:07  |
Duane Bozarth
Messages: 4
Registered: September 2003
|
Junior Member |
|
|
phil chastney wrote:
>
> "grunes" <grunes@yahoo.com> wrote in message
> news:2c0d6c85.0309191029.3efe3a99@posting.google.com...
>>
>> 5. 100% upwards compatibility with earlier versions of the same
>> language/environment, and between platforms.
>
> I don't see this as either necessary or desirable -- Fortran has survived so
> well precisely because of its ability to move the language definition sideways,
> without being shackled to earlier imperfections
Well, since F77 there is little that has actually been removed and a
major consideration (as is evidenced in converstions in c.l.f) is
maintaining compatability w/existing code. In practice, virtually
nothing is ever removed from a commercial compiler although most have
switches to allow specific standard level violations to be flagged...
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
