Ieee Double -> Ibm 370 Vb Source

Hello —

I'm interested in finding an algorithm written in VB to convert IEEE

format doubles to IBM 370 format. I've found a VB alogrithm on MSDN

that does the reverse, and C and FORTRAN alogrithms that do what I'm

looking for, but nothing that does IEEE to IBM 370 in VB.

While it is certainly possible to translate the C or FORTRAN

algorithms into VB or use a C/FORTRAN DLL, I'd like to avoid the

headaches if possible. If anyone knows where I could find such an

algorithm, I'd very much appreciate it if you could point me in that


Thanks in advance.

Request for Question Clarification bymorris-ga


"Welcome to, the web site for a DLL/Type Library that

gives Visual Basic many extra string handling functions, for those of

us that find this type of 'system utility' coding laborious or time


These functions include, EBCDIC/ASCII conversions,"

There is a free download (7 day trial) at

Unfortunately, I have no way of testing the program to see if it works

properly with floating point, though it would be strange if they left

that out.Please let me know if this does it.

Clarification of Question bydbright-ga

Hi — thanks for your response.

While there are several products that do this conversion, it would

require me to call the conversion function from a DLL or static

library which is what I'm trying to avoid. I know that as a last

resort I can use a freely available algorithm written in C or FORTRAN

or 3rd party product to do this conversion, but I'd like to avoid the

overhead of using a DLL.

Request for Question Clarification byxemion-ga

Would you mind posting the link to the MSDN article so we can review

that? Thank you.


Clarification of Question bydbright-ga

The link dextron provided in the comments is the one I was referring to:;EN-US;q235856

Clarification of Question bydbright-ga

Hi fireband, I am converting 8 byte IEEE to 8 byte IBM 370. I agree

that it shouldn't be too difficult to knock out in VB, it's just that

I have more pressing issues at the moment, so I'm gladly willing to

spend $20 to outsource this piece.

I do have the ability to spot check the code if you were to write it.

That is, I can do a IEEE -> IBM 370 -> IEEE round trip conversion and

compare the results (allowing for precision munging during the


I'm dealing mainly with real life data, so and underflow or overflow

condition will be unlikely. But in the event is happens, representing

an underflow as 0 and an overflow as the largest representable number

should be fine. Thanks.

Clarification of Question bydbright-ga

I, of course, meant "firebrand" 🙂

Leave a Reply

Your email address will not be published. Required fields are marked *