Re: IDL subroutine improvements

[<asowter(at)synopticsga.freeserve.co.uk>]

Jeez, this is a real bummer.  As a commercial IDL user, I've been sold
partially on it's flexibility to upgrade.  It's also a personal bummer as
I'm a regular user of the correlation and CONGRID functions.......!

Andy

Craig Markwardt wrote in message ...
>
>
>wbiagiot@suffolk.lib.ny.us writes:
>>
>> To all,
>>
>> This is a really small issue.  I'm just wondering if anyone else has
>> submitted an improvement to an existing IDL subroutine to RSI and seen it
>> incorporated into a subsequent version of IDL? A while back I submitted
>> (what I considered to be) a significant speed enhancement to the cross
>> correlate and auto correlate functions with only minor modifications.  My
>> benchmarks were showing me about a 60%+ speed improvement (which is
important
>> if your code is constantly banging on these functions, like mine was). I
had
>> to convince the rep over a couple of emails what the
advantage/improvement
>> was.
>>
>
>I was in a similar position.  I found an inconsistency in CONGRID --
>which still exists today, by the way.  I reported it for version 4 of
>IDL, and I convinced RSI tech support people that it truly was an
>inconsistency.  Unfortunately it was never corrected.
>
>Craig
>