On Wed, 8 Jul 2020 17:28:17 -0700, Charles Mills wrote:
>Regarding 2: *if* it was a "round trip" translate table and *if* one could
>get a copy of the table then the IEBCOPY data could be reconstructed
>programmatically.
>
>Even if not, I suspect that if one defined the problem not as "do a 100% job
>of recovering *any* IEBCOPY unload that has been translated to ASCII" but
>rather as "do a 95% job of recovering FB/80 source code" that one might have
>a manageable task. Not trying to "undo" the translate but just looping
>through the unload data and pulling out the actual source records.
>
One of the worst possibilities is that they used the "dd" utility:
https://pubs.opengroup.org/onlinepubs/9699919799/utilities/dd.html#tag_20_31_18
... which converts between no two identifiable code pages.
Why is a utility targeted for IBM mainframes (other than Linux for z)
translated into "ASCII"? That makes about as much sense as translating
the Zohar into Sanskrit.
It appears that the "ASCII" translation was never validated, and that
for a long interval there was no interest in having or maintaining
the source.
>-----Original Message-----
>From: Seymour J Metz
>Sent: Wednesday, July 8, 2020 4:55 PM
>
> 1. I would be willing to bet that if Gerhard gave up on it then it was well
>and truly hosed.
>
> 2. The web site mentions IEBCOPY unload. If they translated that to ASCII then
> you're in for interesting times.
-- gil
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
