NT vs. Unix DNS

Martin G. McCormick martin at dc.cis.okstate.edu
Thu Apr 20 14:47:32 UTC 2000


	One of the best tests of the robustness of any technology is
to answer the question, "How adaptable is it to unusual conditions?"

	As a computer user and experimenter who is blind, I deal with
this problem all the time.  UNIX was not written for blind people, but
it was designed to allow for standard input and output so we can
string things together in ways that the original developers never even
dreamed of.  Without going in to long detail, suffice it to say that
there are many UNIX users who are blind all over the world and they
don't use it terribly differently than anybody else because it is so
well designed that the fixes are rather easy and cheap to implement.
There are even people working on making X windows work, but this is a
tall order.  The reasons for this, however, are purely technical and I
suspect that there will eventually be fixes that we can at least live
with.

	On the other hand, the commercial OS which runs the huge
majority of personal computers in the world has no sacrosanct
mechanism for I/O.  Representatives of organizations dealing with
blindness matters began to cry foul because it was hard to write
patches or adaptive engines that could give speech or Braille
functionality to this environment.

	The company wouldn't lead, follow, or get out of the way.
They couldn't be bothered and were afraid to share information unless
one paid lots of money or signed all kinds of NDA's so what we now
have is buggy, expensive third-party commercial programs that give
access to some applications but do not work on a blanket basis.

	The problem is not graphics as such, but a design based on a
mistaken belief that there is only one way to do things, now, and
that's that.

	Rather than catching their collective breath the way most
humans do when confronted with a grave error, this organization just
put the tractor in full gear and continued on the same stupid path.

	Does one want to trust something as vital as a DNS to a
company with an attitude like that?

	This business of the blind access does not concern most people
at all.  Tomorrow, there may be some other design fault discovered
in which you find yourself on the short end of the stick.

	It just seems that the open-source material has far less truly
fatal defects in it and those that do show up are apt to be either
fixed or well-documented quickly.

Martin McCormick



More information about the bind-users mailing list