[Info-vax] Encrypted TCP/IP network printserver spooled printing for OpenVMS (secure-IPP?)

[Stephen Hoffman]

On 2018-05-03 23:54:13 +0000, seasoned_geek said:

> There is no future in a version of OpenVMS riddled with OpenSource feces.

For a number of cases, open source is the only path forward, as VSI 
doesn't have the staff or the schedule or the budget to reimplement 
everything themselves, and doesn't (yet?) have the scale of customers 
to attract folks to bespoke APIs in enough volume to matter.

> There is no future in perpetuating the failed architecture of Linux 
> when it comes to IP. Period. If that is the plan they can end the port 
> right now because VMS died 3 hours before they (whoever they are) made 
> the decision.
> 
> We have had this discussion on the TCP/IP appliance. I "thought" you 
> understood at one point.

A bespoke approach won't sell.   IP and BSD sockets are what we have to 
work with now and for the foreseeable future.  It's what apps and 
developers expect.  TCP and SCTP and UDP and otherwise, and with TLS 
and DTLS, and with IPsec too, and with current compilers for common 
languages.   Those and other fundamental APIs and tools have to be 
available, and have to be current.  Again, VSI doesn't have the scale 
or the budget or the time to gather the skills to write bespoke 
implementations of common network services, either.  That means the use 
of Apache or nginx, and of libtls or OpenSSL, or of 
mqtt/mosquitto/rabbitmq or Erlang, otherwise.   Wrap and abstract some 
of the lower layers and APIs, certainly.   But those lower-level APIs 
still have to be available directly and far better integrated and 
better secured, as OpenVMS is not (yet?) in a position to be a bespoke 
platform.  Better integrated tools, easier to develop, better security, 
whatever other customers VSI is targeting.  But not having expected 
APIs and competitive security is not going to succeed in most any 
market I'm aware of.




-- 
Pure Personal Opinion | HoffmanLabs LLC