Re: What is going on with Lahey?

Al,

You may have been using a Fortran 2003 construct that Lahey (windows)
compiles correctly but does not detect uninitialized variables for. Example,
Derived Type Integer (or Real) array. For more complete lists see:
Table III at: http://ftp.aset.psu.edu/pub/ger/fortran/test/results.txt
and Polyhedron results at:

http://www.polyhedron.com/pb05/win32/diagnose.html

Skip

On 31 Jan 2007 10:19:21 -0800, awgreynolds@xxxxxxxxxxxxx wrote:

-|On Jan 31, 7:41 am, Herman D. Knoble <SkipKnobleL...@xxxxxxxxxxxxxxx>
-|wrote:
-|> Al: Please post a short run-time example that IVF catches but LF95 v5.7 or 7.1
-|> does not catch.
-|>
-|> Skip Knoble
-|
-|Unfortunately, its been several weeks since the last time IVF 9.1
-|found a uninitialized variable at runtime that Lahey 7.1 did not. All
-|I remember is that it was in a module procedure.
-|
-|For years, I have depended on Lahey's runtime checking although I was
-|aware of some of its quirks. For example, they "set" uninitialized
-|data to a specific bit pattern (a different one for each data type)
-|and then check for that pattern at runtime. The problem I ran into is
-|that their integer(1) pattern was valid data for me!
-|
-|I have been using Lahey compilers since f77l (only a ~100kB
-|executable!) in the 1980's. I will miss having an up-to-date version
-|of one of their Windows compilers. The aforementioned contact with
-|one of their representatives was for a bug that was known but will
-|sadly never get fixed.
-|
-|Al

.