pr25039 Frontend problem?

[Richard E Maine]

On Nov 28, 2005, at 10:26 AM, Steve Kargl wrote:

> We have numerous PR's that contain a variation of "g77 accepted this
> code, why is gfortran broken".  After awhile, it gets tiresome to
> tell people to fix their nonconforming code; a position you surely
> understand.

Yes, I well understand. I've seen the same kinds of complaints about 
other compilers in other contexts. I've even seen the horror of 
acceptance testing that required a compiler to get demonstrably 
incorrect results to match the buggy results of a baseline test case. 
Not an extension, mind you, but results that were demonstrably 
incorrect for the physical problem being solved. On finding one of 
those, I seriously suggested that all analysis done with the baseline 
code for the last several years really ought to have been rechecked for 
validity. :-(

My understanding of the presssures doesn't keep me from injecting 
occasional input on the opposing side. I don't actually expect everyone 
to agree with my personal judgments on everything. This is one of many 
matters where I recognize valid arguments on both sides (or even more 
than two sides) and thus I understand that perfectly reasonable and 
intelligent people can come to different conclusions. So I understand 
your position and I respect it as a reasonable one. I'll still 
occasionally speak up on the other side, but I try not to do so in a 
derogatory manner. (If I occasionally fail in my intent to keep my 
disagreement civil, that is a fault of mine, for which I apologize.)

-- 
Richard Maine                |  Good judgment comes from experience;
Richard.Maine@nasa.gov       |  experience comes from bad judgment.
                             |        -- Mark Twain