[Info-vax] Unthinking Existing VMS Networking, Moving Forward was Re: Rethinking DECNET ?

[David Froble]

JF Mezei wrote:
> On 14-08-31 09:07, Stephen Hoffman wrote:
> 
>> Restarting work with Phase IV means re-implementing a sizable hunk of 
>> what V already offers.  Which probably means ending up with something 
>> about the size of Phase V.
> 
> 
> I am not concerned about the transport itself. So decnet 4 or 5 by
> themselves are not relevant to me.
> 
> What I am concerned about is the functionality that is embedded within
> VMS and how to translate/maintain that functionality in a more modern
> environment.
> 
> the file system for instance understands that node::filename  gets sent
> over to FAL. So an updated FAL could then make use of IP instead of
> DECNET.  Same for task to task communications etc.
> 
> So in essence, keep the same DECNET functionality at the user level
> while ditching DECNET at the transport level.

FAL is an application ...
FAL is an application ...
FAL is an application ...

Get the idea yet?

What FAL does could be implemented using TCP/IP.

DECnet is strictly a task to task communication product.  Anything else 
is an application, perhaps bundled in with the communication package, 
but still an application.

Yeah, now go ahead and shoot down that argument by saying DECnet is also 
an application.  From some perspective, VMS is an application.