Comments (5)
  • Add a Comment
  • Edit
  • More Actions v
  • Quarantine this Entry

1 localhost commented Trackback

Long, long ago, when dinosaurs roamed the earth and I was a young(er) USAF developer, I wrote FORTRAN and assembler on VAX/VMS machines for the Over-the-Horizon radar system. I was "embedded" at GE and wrote the code for the hardware/software interfaces at the Receive and Transmit sites.<div>&nbsp;</div> It was also pretty much the last time I did a substantial amount of coding, so I have fond memories of FORTRAN and assembler.<div>&nbsp;</div> Thanks for the trip down memory lane. :-)

2 localhost commented Permalink

Kelly has a nice quote from John Backus here:<div>&nbsp;</div>

3 localhost commented Permalink

In a previous life when I was a geophysicist I wrote some CODE in FORTRAN 77, it was my first professional code (apart BASIC) and yes it was not so nice. FFT and some other IBM libraries were in daily usage...and with vector facility we was able to do a lot of hard work. Ok it wasn' elegant but he did its work in scientific computing. Now you've a babel of languages mainly for WEB sites, the most of them rubbish. <div>&nbsp;</div> But it was when MF was THE COMPUTER. Now it's even better but these good old days never will be back again.<div>&nbsp;</div> Cheers<div>&nbsp;</div> Max<div>&nbsp;</div>

4 localhost commented Permalink

Thanks Max for stopping by. We seem to have a theme encapsulated in this thread of "not pretty but got the job done". Hopefully we're in an era where the mainframe IS pretty again. :-)<div>&nbsp;</div> Personally I'd like to see us more vocal about what we can do in the modern (web) space, regardless of whether some of the software stack is "rubbish" or not.

5 localhost commented Permalink

"Now, does anyone remember using FORTRAN (or even Assembler) on the 3090 Vector Facility?"<div>&nbsp;</div> Yes I used both, back around 1985/1986 on a 3090 being used for seismic data processing. The IBM Fortran compiler was very efficient in generating vectorized code, it was almost impossible to do better by using hand-coded assembler. I managed to beat it only in a situation where I knew the vector length was rather short, and the compiler had no way of knowing the actual vector lengths so it always generated the same type of code.<div>&nbsp;</div> Fortran was a good language and got the job done. We also used PL/I which was a much more powerful language but programmers tended to use too much memory and other resources. I suppose neither language is used much these days.<div>&nbsp;</div> As I recall, James Cooley of IBM was involved in the FFT routines and maybe some others that IBM produced for the Vector Facility. He was a genius, one of the inventors of the FFT.<div>&nbsp;</div> I no longer work in that area so I always wondered what eventually happened to the Vector Facility.