[Info-vax] What would you miss if DECnet got the chop? Was: "bad select 38" (OpenSSL on VMS)

[Johnny Billquist]

On 2016-10-02 15:48, Dirk Munk wrote:
> David Froble wrote:
>> Dirk Munk wrote:
>>> Anything is possible, but let's stay with the reality.
>>>
>>> Multinet is offering a kind of DECnet Phase IV over IP interface,
>>> using IP port 700. It is not covered by any RFC. I don't know if it
>>> can be used with IPv6.
>>>
>>> The DEC/HP way of DECnet over IP is not only offering DECnet over IP,
>>> but also OSI over IP (you can look at DECnet as just another OSI
>>> application). It is covered by three RFC's, RFC1006, RFC1859, and
>>> RFC2126, the latter is for IPv6.
>>>
>>> Both versions of DECnet over IP are incompatible.
>>>
>>> Now my simple question is, what should VSI offer, two incompatible
>>> versions of DECnet over IP?

Because of two incompatible versions of DECnet?
Let me bounce the question back: Why should VSI drop support for Phase 
IV? And why should VSI remove existing, working, and used functionality?

Note that I'm just talking about what exist today for DECnet phase IV.

And, if they retain the Multinet DECnet-over-IP for Phase IV, in which 
way is that bad for Phase V? Essentially, we're talking about 
functionality that have nothing to do with Phase V.

They are two different protocol suites, with disjoint domains, functions 
and implementations. Why do you try to bundle the two things together?

>> If you look up thread at Michael Moroney's post, you'll see the
>> reality.  He got DECnet built, but that's all the time VSI is going to
>> put into DECnet.
>>
>> What you see today is all that you're ever going to see, with the
>> understanding of "never say never".  Your question(s) are already
>> answered.  Not that you're going to like the answer(s).
>
> Let's start again shall we?

Sure.

> VSI is going to use the IP stack of Multinet as a base for their new IP
> services for VMS version 10.5.

Correct.

> The Multinet IP stack has no provisions to support DECnet Phase V / OSI
> over IP, you will not find the RFC's that I mentioned in their SPD.

I suspect that is correct today.

> Multinet has a different non-compatible version of DECnet Phase IV over IP.

Right. But even so, they can not be compatible, since phase IV and phase 
V are not the same thing, so why do you try to make it so?

> Mr. Billquist stated the following:
>
> "At most, it could make sense to provide the ability that Multinet
> already have, of using TCP/IP as a transport for DECnet circuits, which
> can be done for Phase IV. I suspect that could actually be of some use
> at a few places. And it has already been implemented.
> But anything beyond that, just would not make sense."
>
> That would break DECnet Phase V OSI, since it would no longer have an IP
> stack, and I told him so.

That is FUD on your part.
Mutlinet have a DECnet-over-IP for Phase IV today. I'm pointing out that 
I don't see any gains in just deleting this functionality out of spite. 
It is there, it works. Why not keep it?

This has *nothing* to do with phase V.

Now, if you want phase V, then I guess you should request that VSI 
implement those bits for their new TCP/IP stack.

This would be a different thing than the Multinet phase IV tunnels, and 
there is no conflict between them. You can certainly have both.

However, the phase V issue is that nothing exists today for the new 
stack. So here you need some new development, which is totally separate 
from anything that already exists, and is in no way in conflict with it.

> He replied:
>
> "Who cares? You have two Phase IV nodes, they can connect using IP. All
> else is unchanged. Phase IV couldn't care less about OSI applications
> anyway. The same goes for RFCs. You do not have to have an RFC to use a
> protocol. We are talking about DECnet here, remember? The fact that it
> can be carried over IP just means that you have your phase IV DECnet,
> nothing changed there. All that happened is that you can connect two
> DECnet Phase IV machines who only have connectivity through IP
> otherwise. A simple, obvious win, without any downsides at all (except
> in your head)".
>
> So he doesn't care that his suggestion would break DECnet Phase V OSI.

My suggestion in no way breaks phase V. You are utterly mad if you are 
trying to sell that story.

The fact is, that, if anything, the mere switch to this new stack, is 
what will break phase V, since no ability to do DECnet phase V over IP 
exists in the new stack.

And that have nothing to do with anything I've written. It's merely a 
fact. y suggestion of keeping the current Multinet DECnet-over-IP for 
phase IV in no way breaks phase V. It has, in fact, nothing at all to do 
with phase V.

Now, deleting the current DECnet-over-IP in Multinet can be done. Will 
that make phase V over IP work any better? No. It will not make any 
difference.
So, what about keeping DECnet-over-IP in Multinet. Will that make phase 
V over IP work any better? No. It will not make any difference.

So, in short. Who cares? The DECnet-over-IP in Multinet today have no 
bearing on phase V DECnet over IP.

If you want to fight for DECnet phase V over IP, feel free. Nothing will 
prevent you. And note that whatever currently exists in the Multinet 
stack will neither hinder, nor improve the ability to do DECnet phase V 
over IP.

But you need to convince VSI that they should work on actually implement 
that thing for you. There in lies your problem.

Stop taking that out on another piece of software that have nothing to 
do with you or your issues.

> So I told him that he apparently doesn't care that customers using OSI
> would not be happy about that.

I personally do not care, but more to the point, my suggestion does not 
make any difference for the customers using OSI, so in which way would 
they not be happy with my suggestion?

> Then you accused me of not being able to read, because Mr. Billquist
> didn't say anything about Phase V / OSI, but in fact he did.

You are, in fact, unable to read what I wrote.

> Hence my question to you, if you think VSI should support both
> incompatible versions of DECnet over IP.

Your question is formulated wrongly, and biased.

What you should ask, are the following two questions:
1) Should VSI delete the current DECnet-over-IP for phase IV, which have 
no impact or relevance for phase V. (Essentially we're talking about 
doing work to delete something that exists, which in no way is hurting you).
2) Should VSI implement DECnet-over-IP for phase V, which in no way 
relates to the phase IV tunnels. (Essentially we're talking about doing 
new work for DECnet).

These two questions are independent of each other. It would seem that 
you are arguing for yes to both of them, which is the maximal amount of 
work for VSI.

My, I am arguing for no to question 1, and I couldn't care less for 
question 2, so I'm not voting on that one. I'll leave it to those who 
actually have some staked interest in phase V.

See, I'm being nice. I actually am not trying to destroy anything for 
you. But you are trying to destroy something for me.

Go figure...

	Johnny