[E-Lang] VLS for everyone

[Mark S. Miller]

At 01:49 AM Friday 3/23/01, Tyler Close wrote:
>Steve wrote:
>> Yes, I certainly have expressed an interest in reviving the VLS.
>> [...]  Regardless, it needs to happen and
>> there's no reason that I can't do it.
>
>Great. I think we have enough interested parties that getting
>something done should be feasible.

I wholeheartedly agree.  It's amazing not only how many capabiliity systems 
we have represented on this list, but how many distributed cryptographic 
capability systems -- E, Droplets, MojoNation, E-Speak2.2, Joule/Indra -- 
all of which did or could use such a service.  In the spirit of the 
unbundling Tyler advocates, I notice I don't need to be central to this 
effort.  To the degree it makes sense to talk about roles in a loose 
open-source community such as ours, the following roles make sense to me:

Steve Jenson, as the one doing the overall work, you'd also be in charge of 
defining what it is, through a process of open discussion on the e-lang 
list, just as we do with E itself.

Bill Frantz, representing the perspective and requirements of VatTP, the 
component of E, written by Bill, that currently both contains the VLS and 
"makes use" of it.  (I put "makes use" in scare quotes, because I've had 
that feature turned off, and allowed the VLS code to get stale.)  Bill, 
after the upcoming 0.8.10 release, I'd like you to formally take charge of 
VatTP in any case.  It's plausible to me that VatTP may also be able to be 
unbundled from E.  Would you like to use http://www.vattp.net for this purpose?

Tyler Close, representing the perspective and requirements of the Waterken 
suite.  (Is it appropriate to say "Droplets" to refer to the suite as a 
whole?).  Tyler, you should also represent our active memory of the httpy 
discussions.

Zooko (not Ppooko!), representing the perspective and requirements of 
MojoNation.

Steve, to the extent it's appropriate, could you also represent the 
perspective from openCOLA, and the requirements that would be needed for 
them to evolve from the current clerver protocol to this universal VLS?

Alan, does it make sense for you to represent the perspective from 
E-Speak2.2?  Would this also be relevant to E-Speak3?

Dean, if the perspective from Joule/Indra differs in any way from the 
perspective from E/VatTP, please speak up.

Steve, as the one doing the work, if you can't satisfy the needs of all 
these systems, which is likely, it's up to you to decide which to drop.  If 
it were me, I'd consider the high priority customers to be the active ones 
most likely to adopt the system if it met their needs (E, Waterken, Mojo?, 
openCOLA?).

I need to do a brain dump of my current thinking about the VLS, after which 
I can take a back seat.  (This won't happen until after the 0.8.10 release.) 
All the rest of us including myself should kibitz, contribute, criticize, 
argue, encourage, etc, as usual of course.

Does this sound good to everyone?  If I've offended anyone by volunteering 
them without first asking privately, please let me know, privately if you'd 
like.


>What exactly are the issues with firewalls that are related to the VLS
>service? I really can't think of any. [...]

You answer your question below:

>The place where firewalls do become an issue is when the client tries
>to use the fetched address information to contact the target host. 

That's the issue.

>I think we can and should ignore this step in the definition and
>implementation of the VLS service.

I admit to failing memory of the old httpy discussions, but I am mystified 
by this statement.  The purpose of the VLS is to bring about Granovetter 
introductions.  If the VLS reports to VatB an address for VatC that VatB is 
unable to use to contact VatC, what good has the VLS done anyone?  It would 
have failed at its one purpose.


>Sometime this weekend, I'll try to put a summary of my HTTPY proposal
>on the waterken.com site.

Great!  Looking forward to it.


        Cheers,
        --MarkM