I can vouch that the feeling goes back more than 40 years.. (I know, 
no help there :)  One interesting one involved a colleague at IBM who 
had worked out an interesting alternative to a Fast Fourier Transform 
- it was done and worked very nicely in APL. But the patent attorney 
required rewriting it in FORTRAN...

Another was a brief letter to a geology journal by the late Donald 
McIntyre. He submitted a letter that included a brief APL function 
that modeled tectonic plate movements. The journal rejected it saying 
they were anxious to print it, but required that the program be in 
FORTRAN. When Donald said he would be happy to do that, but they 
should expect a 15 page long document to replace the brief APL, they 
reluctantly agreed to publish the APL. A funny side effect of that 
was that some people at the USGS wanted to run the model, and so they 
gave the short APL function to their computation group who said that 
it would take weeks of work to re-implement... At that point, the 
USGS users decided the best solution was to fly 500 miles so they 
could try their parameters in Donald's program on the machine in his 
office, an IBM 5100....

I've had some more recent ones (only 10 years ago...) that were j 
based. My favorite was one that the "Official Group" upon hearing 
that the project was implemented in j said, "Oh, we can easily port 
that to our system, you do mean  Java, yes?"

Your story and frustration is one that has been repeated many times. 
I suspect every "old timer" in these forums can tell several such 
stories. Some of the stories involve huge efforts, to replace a 
system that the users where quite happy with, that ended in abject 
failure...

So, cheer up! Thanks for adding to the litany.



At 10:14  +0800 2010/05/05, Alex Rufon wrote:
>Ok. First off, I'm just venting so if you have better things to do 
>... just ignore this email (this is why this is in Chat).
>
>I'm a bit frustrated with some programmers attitude regarding 
>development. Unfortunately, English is my 3rd language so bear with 
>me on this one.
>
>One of the Project Managers have a problem. They need to import an 
>EDI 850-4010-X12 into their system which only handles XML document. 
>His problem is that his Technical Manager is quoting 22 man days to 
>implement support for it. Her boss referred him to me and I said I 
>can read the EDI document and export it to XML that their system 
>currently handles. As a favour to the guy, my wife helped out and 
>spent 2 days mapping the EDI document and it took me another 2 days 
>to implement my wife's work and export the file to XML. Side note: 
>This is why I really like J, 4 days work (2 days programming in J) 
>as compared to 22 days in C#.
>
>Now here lies the rub. When we were implementing, I am now getting 
>grief from his Technical Manager because the importer was not done 
>in C# and is insisting on redoing the whole thing again in his 
>language. I could have understood where he stand if the thing 
>doesn't work ... but it does as what was in the specification. 
>ARRRRGGGGHHHHH (this is my shorthand to the heated arguments).
>
>What surprise me though is that for the last decade, I've been 
>encountering this attitude, and I still get affected by it. Just so 
>frustrated right now.
>
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm

Reply via email to