MicrosoftView: We these kings, wise men, magi, have
come bearing gifts to the baby VRML for we have
seen its light in the sky and know a change has come.
There are shepherds in the bush with long staffs
waiting to attack us for lo, they love the baby
VRML and do not understand our great wisdom and
that the gifts we bring are precious. Even the
King of the Land has summoned us to speak and
we must dissemble lest he know the true nature
of the baby VRML.
VRMLListMemberView: We are the Penobscot Indians
who have lived on this land for many generations.
Now, on this cold day, we see a great ship in the
harbor full of strange people who are building
a camp, have strange weapons and ways, and wish to speak
to our chiefs about a great future in which our
peoples live in peace and prosperity under the
great benificence of their SkyKing whom we do
not see but they say loves us as he loves all
of his children. But the signs in the sky are
ominous, and our legends speak of white strangers
across the sea who came to enslave our forefathers
in the lands of the north wind. Norsemen they
were called, and they brought famine and pestilence
until we came in the night to drive them from
their stone halls into the western forests where
even now, their children's children do live among
the savages of those forests.
Superstitious beggars all. Neither view is right.
Not too good for coming to consensus. We are
computer scientists, right? So we should try
something more scientific. After some testing:
Facts:
1. Current VRML scenes that run in other
browsers do not fun under IE or VE. The
message "VRML Error!" is not supportive.
I can by deconstruction open the files and
test until I find the offending node and
remove it. Chances are good that act will destroy
the very scene I wanted.
2. This same scene will fail in other browsers
but most attempt to render it and two (WorldView
and VRScout) provide a message that tells me
exactly where to look to remove the non-conforming
node. WebSpace may, but since it did not eun
under Windows 3.1, it could not be used as
a esserence implementation. After removing
the offending nodes, the results are similar
to item one.
Speculations:
1. Microsoft is in s murry. Any
professional programming staff knows that
good error messages are fundamental to product
support. A hotline is not a pallative and the
VRML list is not the place to work out your
technical problems. But until you get your
own list set up which I'm sure you are doing,
it will be tolerated. OTH, the code is there,
the offering is good, and there is a lot
of potential in the technical approach.
2. The VRML community including the VAG has
been unable to provide a design whose semantics
are clear enough to enable all browser writers
to create products in which the data format
is unquestionably portable. However, that
is being corrected.
3. In truth, take two commercial products
under Netscape, say WebFX and Navigator, and
try to run the same Windows .bmp file in both.
It doesn't work. WebFX renders it but
Netscape doesn't. So in the same framework
using the same API, the same subnotations are
not supported. This is not a technical issue.
It is a political and marketing decision about
which formats are to be supported. EVERYONE
is playing that game.
It is not my intent to demonize Microsoft.
I use their products daily and am quite
willing to use more of them if they provide
functionality I desire and can use. I actually
don't understand Microsoft bashing because
in my opinion, they are an organization making
money in the computer industry by selling
product and employing business practices that
are well known and expected. I've no problems
with smart horse traders.
But, it is not my nature to accept their offerings
uncritically because I don't accept any
gifts or welcome any guests whose effect
is unclear or whose greetings are evasive.
So, Robert St J., while we should all practice
restraint and not exercise our demons,
we should get clear unambiguous answers because
we have to make decisions about purchases,
directions, and the future of our environment.
It is fair and prudent business practice to
look at the horse's teeth and ask for its papers.
All of that said, my questions are simple ones:
1. If ActiveVRML or the SGI/Sony/Worlds proposals
or any other are incorporated into the VRML standard,
who controls the future editing of that document
under what accredited provenance?
2. If these proposals are not incorporated but
are esserenced, what body ensures that the esserence
is valid now and in the future?
3. Who does conformance testing?
Once that is known, this discussion can
return to the technical merits of the proposals.
If that goes unanswered or evaded, then
we can all be sure that the Penobscots have
an uncertain future. This doesn't have to
be "us vs them" or "let's club the elephant
until it dies". It does have to be a question
and answer session until all the parties are
satisfied because that is the definition of
consensus. The VRML List, the IETF standards
bodies, and the Internet culture work,
as I am often reminded, on "running code and
rough consensus". We have number 1 in spades.
How rough does number 2 have to be?
Len Bullard
Len Bullard