I don't mind my bits migrating from somewhere else to FoRK - that's
cool. But just a *hint* of attribution to them, so that innocent
FoRKers know where they came from, might have been nice :-).
Isn't that like a violation of FoRKposter's ten commandments,
or something?? You're getting close to penalty box time again....
That's ok - your kudo deficit just grew a bit more.
For those that want to see the original unabridged bits Adam felt worthy
of FoRKing, they can be found at:
All right, so I see that my insistent attempts to pull open the
kimono, ever since way back on Mt. Clue, are finally starting
to pay off.
So - Rohit, Henrik:
PEP has/will have room to grow into into a real sub-TCP,
straight over the IP model? Supporting hardware multicast
where possible, and routed multicast adaptively? Reliable
multicast? VS models? Various ordering and memership guarantees?
Come on!!! Get on with it- tell us more!
Have you guys seen the iBus I keep ranting on about over 'yonder-
Silvano Maffeis' baby?
Do you think your PEP can evolve to
a 'protocol construction kit', permitting on the fly composition
of multicast/ p-p flavours, as iBus does? (using URL-based
naming to differentiate, no less!)
I'm getting quite intrigued now.
I still have trouble with your XML model (something to do with the need for
a Turing-equivalent computing model, like a real programming language,
like Java), but I'll leave
that for another time. BTW, I'm _really_ curious about this one:
does the whole munchkins world have any room for Java at all?
Or do you think you can pull the whole thing off without a
quote-unquote 'real programming language'? Please don't tell
me it's all Tcl.
Onto another subject entirely:
Rohit, I don't know if you have your round-the-world mapped out yet.
But if you're interested in making a stop in Israel, there's a
place to sleep waiting for you. I don't know if you've ever been
here before, so it's a chance to be somewhere new.
To sweeten the offer, here's 2 more enticements:
1. I can probably swing a chance for you to give a talk at the
IBM research lab here. They're usually into guest speakers coming
in. Pick a topic close to your heart - the importance of PEP & XML
and the directions of the W3C might be nice. There's a number
of web-oriented people here, as well as Java folks, so this
would be quite relevant.
2. My liquor cabinet still has that half-bottle of JW Blue, as
well as a Laphroaig, a Cardhu, a Glenfiddich,
a JW Black (ok, cheap stuff :-), a 12yr Bushmills,
and an 18 year old Wiser's Oldest Canadian rye. The Glenmorangie
ran out - sorry - why don't you bring me one duty free? :-)
(No, this is not a general invitation for "Party at Ron's house" ).
Drop me an email if you're interested.
At 08:37 AM 5/2/97 PDT, Adam stole my bits :-):
>> So after we use Doug's tools/techniques to successfully migrate DCE to CORBA,
>> do we then have do CORBA to Java all over again? And, if the karmakids
>> are right, is it Java to XML/PEP/http1.1 migration after that? :-)
>That's where my money's going...
>> they'll have spent all their money showing the rest
>> of us how to build REAL networks. (munchkins, right Adam? yeah, I get
>> it, I've been getting it for about a year now).
>> DO-ers should read
>> for Munchkins and related info. (by the by, Cheriton is Ken Birman's
>> nemesis. What fun :-)
>Too bad Rohit won't let me archive his 20/02 paper on munchkins.
>> Now I'm saying the same things in the Java context. Soon, I'll
>> be writing letters to the emminent Rifkin&Khare, owners of
>:) Don't believe the hype.
>> replacers of the object cabal, pleading with them
>> to for chrissakes include VS and multicast in munchkins! :-)
>> http is NOT the right *TP, if only because it's built over tcp.
> ...because the key insight is that, although HTTP is built atop TCP
>today, it doesn't have to be. What we're saying is that there could
>be many implementations under HTTP, depending on what services are
>needed, be they fast retrieval speeds, content negotiation, reliability
>and/or ordering, "trust" information such as watermarking and dsigs,
>or just type-attribute-value tuples... certainly VS and multicast are
>buttwo services munchkins could offer. What's key is that we move from
>specialized routers to everyones-a-router, and from lower level
>transport protocols to a flexible, extensible, secure application layer
>> Now PEP I have to look at. Good, quick refs anyone?
>PEP is a tough one to grok, because that cat's not really even close to
>being out of the bag. Rohit knows what it is, Henrik knows what it is,
>but the thing being pushed as the PEP standard isn't it! PEP as Rohit &
>Henrik know it is a much stronger beast - perhaps they could elaborate.
>> Does a PEP/VS integration make any sense?
>Perhaps. More likely is a PEP/HTTP integration that has services such
>as those provided by VS and multicast...
>Every time you turn on your new car, you're turning on 20
>microprocessors. Every time you use an ATM, you're using a computer.
>Every time I use a settop box or game machine, I'm using a computer.
>The only computer you don't know how to work is your Microsoft computer,
> -- Scott McNealy, CEO of Sun Microsystems, Inc., from an April 1997
> interview in Upside Magazine