[Info-vax] OpenVMS V9.0-C Released July 29th

[Terry Kennedy]

On Monday, August 3, 2020 at 1:01:55 PM UTC-4, Stephen Hoffman wrote:
> On 2020-08-03 15:25:02 +0000, clair.grant at vmssoftware.com said:
> 
> > I tried my best to ditch DECnet on x86 but there are just too many 
> > customers using it (we asked, it was very disappointing).
>
> From hard lessons years ago, polls accrue incremental info at best and 
> most of which the product admin should already know or be able to 
> extrapolate. Polls seldom garner info on where the product should be 
> headed. If ever.

As the post you quoted from Clair said, VSI asked their current paying
customers and Phase IV was in demand.

If you want to suggest that NETCONFIG.COM say "Are you really, REALLY
sure you want to enable this feature because <blah blah>", that's one
thing. But leaving it out when your customers told you they are using
it and telling them "adapt or die" is a recipe for dying yourself.

Way back when the possibility of VMS on x86 was just being announced,
I said that it was going to be a hard decision between porting it
as-is, so customers could simply recompile (or VEST, if they didn't
have the source code for their applications) and go, lowering the
barrier to entry for existing VMS customers, or if there should be new
compilers that accepted modern syntax. It appears from what I've read
here that VSI chose the former, with select new compilers as needed.
I assume they did enough research to decide that that was the best way
to succeed.