[Info-vax] Reimplementing VMS, was: Re: HP adds OpenVMS Mature Product Support beyond the end of Standard Support

[Paul Sture]

Craig A. Berry wrote:

> The user application has only two choices: use the unpublished status
> value that actually works or fail mysteriously. Yes, a reckoning could
> come under changing conditions, but when faced with getting code working
> or not, one does what one has to do. I think you greatly underestimate
> how very, very common such scenarios are, especially given the 35-year
> history of people reading the source listings to figure out how things
> work. I doubt anything really significant, say Rdb, could be made to
> work on a reimplemented VMS without considerable effort.

Rdb could be the killer indeed, paticularly given Oracle ownership.

Back in the late 90s we came across an Rdb bug which turned out to be a
bit of VMS itself only used by Rdb.  This was probably a leftover from
the time when the Rdb belonged to DEC and had its runtime distributed as
part of VMS, but it's a reminder of what could be lurking under the
surface.

-- 
If you make it well enough, you give it immortality.” -- Ernest Hemingway