Re: info assembly line, "flits" (long)

["Josh Sled" <jsled@cory.EECS.Berkeley.EDU>]

On Sun, 23 Jun 96 17:27:25 -0700, Vladimir Z. Nuri wrote:

>>think about the the number of atoms that are necessary 
>>to hold the information for a single page, let alone an entire book...  
>
>yep. not saying what I am talking about is feasible this moment.
>"moore's law"

But I don't think that it's feasible at ANY point in the near future...
in fact, I think a better question is WHY?  What would we use this
newfound ability for?  

>>in a virtual reality "cyberspace", this would be an insurmountable data 
>>storage... on the small-scale.
>
>not insurmountable. quite practical and sensible in say 10 years.

Ok... maybe practical to store the data... but I don't think sensible. 
Think: if we had this ability with atoms, what would we use it for? 
Maybe tracking crimes... or political prisoners (where have the atoms
in this anonymous leaflet spelling out steps for rebellion come from?)
or just curiousity... sure, there are justifable uses, but does the
"savings outweigh the cost", so to speak?

>to move a pencil I only pick it up and set it down. to move a 
>document through cyberspace, 
>the process is infinitely more complex, requiring an
>immensity of thoughts and coordinated actions. when we create
>a system that matches the real-world difficulty, then we will
>be approaching the limit. we are very, very far from that limit
>even though we have climbed the ladder a long ways as you note.

I think we're too far from the limit to even begin to guess about how
information would be stored in such a medium.    But that's just my
thought.  For the sake of argument: why would we want to have this
level of complexity in a simulated world.  There is so much that can be
assumed that is repetative and wasteful in the real world, why would we
want a computer-generated system that must keep track of all this
waste?

>>I think the thing that's most important in this sentence is _"move"_ ... 
>>this is the main problem for computers... it's SO easy to DUPLICATE 
>>information... but near impossible to make sure that you've MOVED it... 
>
>indeed, the concept of "moving" information implies TIME-- at one
>time, it is at one place, at another time, it is in a new place.
>but it is the SAME INFO. today, the disconnected idea of a "bit"
>does not give you this *continuity*. I make a copy somewhere else
>that is not tied to the original document.

That's why it's so easy to make a copy...  that's why we can have
backups and the cypherpunks list and almost everything on the Net.  If
we had to keep track of flits and update them constantly, things would
slow down a lot... but you're compensating for that by having
arbritarbily advanced computer technology.

>imagine the same thing on a totally universal cyberspatial level,
>not merely within a single company or university. I agree, we have
>rudiments of what I'm talking about in place. but my point is
>mainly that they are rudiments compared to what is possible. the
>web is a very good sample framework for the kind of seamlessness
>I'm talking about. like I say, the future information assembly
>line will be built on top of it. it has a long ways to go too.

Ok... in a far advanced level of technology, this and more would be
possible... but, the question is: do we need it?

>in a sense, I think the flit concept is a magic bridge between bits
>and atoms. bits are too abstract. atoms are too real. flits are
>a nice compromise. we have to get our bits to behave more like
>atoms: persistence, etc.  there are a whole lot of very nice
>"properties" of atoms that are staring us in the face that we
>would benefit from immensely implementing in cyberspace.

For some reason, distributed objects come to mind... though I'm not
knowledgeable enough on the subject to know why...

>>It means that you'd have an INSANELY large ammount of storage for a 
>>single small document.
>
>early stages would not be much different than RCS systems already in
>use in companies.

But I think that this is the best system available for today's or the
future's needs... if we extended this concept to document fragments
(ie, if you use a quote from a document, that section of your document
has the same revision history, changed as necessary), then we have a
system that could be in use for a long time.

>you could have mechanisms that don't keep the entire history of the
>flit. I agree, a flit as a 0 or 1 is very unlikely in the near future.
>but at a document level, i.e. a document as a flit, we already have
>it in RCS systems that companies are struggling to implement well
>as we speak.

Again, I think a document system is the best suited for information
storage... the flit concept seems to be a great overkill.

>I'm saying that in the information assembly line of the future,
>they *must* have context. they must be tied together. you only
>have disconnected chaos otherwise.

Hmm... the only problem is that the "context" would be completely
machine enforced... you can't actually MOVE or COPY the bits off a
computer onto another computer... each side just would have to
intrepret the bits as being moved or copied... and unless you propose
storing flits in something other than bits, the disconnected chaos you
refer to is nothing other than software understanding bit patterns...
same as ASCII or a word processor file or an OpenDoc document.

>again, this requires the human to interpret the changes. what if
>there was an actual "link" between the old and the new document
>that is "stuck" to the new document? and furthermore, software
>could traverse these links? that's more what I have in mind.

And if the old document moved?  How would the new document continue to
know where the old document is?

>data doesn't have to be disconnected. I told you this was a radical
>paradigm shift that I was proposing. you obviously have the previous
>concepts down quite well. I'm not arguing that what you are saying
>is the conventional system. I pointed out exactly that.

*nod*  You're right... I don't understand your shift in thinking... it
doesn't make sense to me.  I think that conventional systems are very
well suited to information creation, retrieval and transport, and that
virtual reality, though still in development, wouldn't benefit from the
flit concept.

>>But cyberspace is NOT real space...
>
>it will evolve to become more and more like a real space, a point
>of my essay.

I sure hope not... why do you think so many people escape to muds...
real space is bad... virtual space has the potential to become so much
more.

>>VERY VERY VERY far...
>
>10 years or so, I would say, before people implement things that
>sound like they came right out of what I was talking about.

Perhaps... I hope not, because I don't think that it's a reasonable
concept for working with data in an on-line or even mostly digital
environment, but perhaps...

>right. I think it is much more realistic to speculate that we will
>be virtually unlimited than limited, and to ask the question, 
>what would we implement if we truly were unlimited? to build something
>limited when you are unlimited shows an impoverished imagination.

well... I'll continue to think about the present, and you can continue
to dream about the future....

Take care,
Josh

PS I wonder: do you envision flits being able to be created out of thin
(virtual) air?