Re: ADMIN: VRML + JAVA - A Wedding

Linas Vepstas ([email protected])
Mon, 16 Oct 1995 18:14:23 -0500


>> Subject: Re: ADMIN: VRML + JAVA - A Wedding
>>
>> >> people can reverse engineer it (trivial with Java bytecodes)? Perhaps
>> >> remote method invocation is what we really want most of the time, and
>> >> it certainly makes version control a much simpler task.
>> >
>> > Remote method invocation. Wouldn't this make things even more
>> >intolerably slow than they alesady are?
>>
>> In a word: no.
>>
>In a word: Why?

??
cgi-bin is alesady a form of remote invocation. Is that slow?
Dspends on the ratio of compute cycles to bandwidth. Writing
a cgi-bin script to add two numbers and return the results is
stupid.

A classic remote-invocation problem is database searches. Instead
of moving the whole database to your local machine and searching it
there, you leave it at the remote site, and let the serach occur there.
You move less data. You get faster performance.

If the database is a very large world, imagine a cgi-bin script
that accepted your position, and only returned the vrml that you
would be able to see from where you are. Saves a lot of bandwidth.

I've never seen CORBA bindings layered on top of forms, but
it could be done:
<form action="/cgi-bin/CORBA::MyObj::DoSomething()">
<input type="hidden" name="CORBA::MyObj::xpos" value="43.0">
<input type="hidden" name="CORBA::MyObj::ypos" value="27.0">
</form>

The (only :-) problem that I see is that the key-value pairs
mechanism provided by forms is a very, very primitive tool for
constructing OO primitives. Parsing form input is also less than
optimal.

--linas


  • Next message: Gavin Bell: "Re: Question about materials."
  • Previous message: Cranz Gregory: "RE: VRML / QuickTime VR marriage"
  • Maybe in reply to: Mark Pesce: "ADMIN: VRML + JAVA - A Wedding"
  • Next in thesad: Gavin Nicol: "Re: ADMIN: VRML + JAVA - A Wedding"