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

[Phillip Helbig undress to reply]

In article <37452140-7970-40f4-9eca-02a20db0da3bo at googlegroups.com>,
"clair.grant at vmssoftware.com" <clairgrant71 at gmail.com> writes: 

> I tried my best to ditch DECnet on x86 but there are just too many
> customers using it (we asked, it was very disappointing). 

The customer is always right.  :-)

Seriously, some are using DECnet because it is better at what it does 
than TCPIP in many cases, and also does more things (such as running a 
procedure locally but accessing both local and over-the-network disks).
Modern routers don't even handle it.  As such, it is probably being used 
within secure, internal networks, so the fact that it is non-encrypted 
seems rather overblown (and a DECnet proxy is no worse than having an 
SSH key in place).

Suggesting a change to TCPIP is not a good idea until x86 TCPIP is 
stable.  Since the DEC days, TCPIP on VMS hasn't had the best 
reputation, and customers have had to deal with things such as 
functionality stopping to work in a new release without it even being 
deprecated with a note to get used to the new syntax in a previous 
release, and on top of that wonder why the syntax was changed at all.  
Even the name has changed (from UCX to TCPIP).  :-)

Many people use Dave Jones's excellent "OSU" web server, which calls 
itself across the network, within a node.  Is the lack of encryption in 
DECnet really a reason to stop using it?

The OSU server is now at SourceForge.  When can we expect a 
plug-compatible version which uses TCPIP instead of DECnet?