Date: Mon, 16 Jan 2006 00:48:35 -0500
To: WSFA members <WSFAlist at KeithLynch.net>,
        WSFA members <WSFAlist at KeithLynch.net>
From: "Mike B." <omni at omniphile.com>
Subject: [WSFA] Re: Eats, Shoots & Leaves....
Reply-To: WSFA members <WSFAlist at KeithLynch.net>

At 1/15/2006 03:17 PM, Keith F. Lynch wrote:

>Throughout computer history, numerous people tried to enhance
>communications and standardization:
>
>* The inventors of the ASCII code, which replaced dozens of mutually
>   incompatible character codes.

This works if you are North American anyway...or British, or one of the
colonies or others who use English.  ASCII doesn't work so well if you use
Spanish, Norwegian, Chinese, Arabic, etc. where the character set isn't our
usual 26, or where it is, but you need things like accent marks, umlauts,
inverted question marks, etc. to go with them.

The main advantage of ASCII these days is that pretty much every computer
made can use it...even IBMs which might prefer EBCDIC locally.

>* The inventors of the Internet, which used TCP/IP to allow file
>   transfer and email between any two computer systems whatsoever.

Well, any two *unix* computers, or unix-like systems.  Other OSs support
TCP/IP, but there are aspects to some of the application protocols that use
it, particularly the oldest ones, that are extremely awkward to work with
if you aren't using some variant or clone of unix.

LPR (Line Print Remote) for instance assumes that you can easily determine
how many bytes are in a print job's file, and has you send that information
before you send the data...so the receiving system will know when it's
gotten it all.  On OpenVMS, for example, this information is *not* easily
determined ahead of time (if you want to let users use the standard
printing system the OS provides), so you have to spool the file, counting
the bytes as you go, then send it (VMS formats as it prints...).  For large
print jobs this can take a considerable amount of time (several
minutes).  It also doesn't specify that the control file be sent first,
which implies that the receiver will be able to spool all the files that
make up the job until that gets sent...which is generally not true of
network-aware printers (in its defense, LPR was never intended as a
printing protocol...it's a print job transfer protocol...a way to move a
print job from a print queue on one system to a print queue on another
system.  It is commonly used for printing these days though).

Another example is NFS (Network File System), which is just a network
extension of the Unix file system.  Any other OS that wants to use this
protocol must figure out some way to map its own file system into the unix
model...which often causes problems due to the characteristics of the unix
file system model (case sensitive long file names, only three classes of
accessors (owner, root and group), etc.).  Some OSs support more features,
and have to deal with them not being available on NFS, other OSs support
fewer, and have to figure out how to deal with the extra features (like
Windows not having OS-supported file locking for instance).  NFS wasn't an
attempt to come up with a way for a variety of OSs to share files over the
network...it was just a way to extend the Unix file system for network access.

>* The inventors of the World Wide Web, which used the HTTP protocol
>   to make any web page visible to any browser on any operating system.

Well, it lets any web page be transferred, with information describing it
in a general way (the operation being done, the character set assumed, the
length of the data, etc.).  Whether the page can be displayed by the
browser is another issue.  Memory limits, display device limitations,
locally available character sets, security policies, etc. might prevent
viewing in some cases.  HTTP is remarkably non-OS-specific though...so long
as it supports UTF-8 (a variation of ASCII, basically).

>For a while, things looked dim for the Discordians.  But in recent
>years, the advocates of chaos, confusion, and mutual incomprehension
>have made remarkable strides.  Many web pages now proudly bear the
>banner "Best viewed with ..." or "Requires <foo> version <N> or
>better."  Soon it will be impossible to read any web page more than
>a year or two old, which will cut down on the obsolete information
>that currently clutters the Web.

Reading old ones isn't generally the problem.  The problem is HTML and
various other protocols and programs often used by web designers, and newer
pages are more likely to have limitations in audience than old ones that
don't use all the newer, and more restricted, technologies.

HTML is up to about version 4 now.  The things that have been added since
HTML 1.0 are mostly there to make the web designer's job easier, or to
allow making the displayed page look more like it was intended to
look.  With HTML 1.0 the designer just suggested what sorts of data were
being included, and the browser could figure out how it wanted to show
it.  Designer would say, "This is a list" or "this is a medium-sized
headline", and the browser would show it the way it wanted to.  This made
for some really ugly web pages at times, when the viewer's browser had very
different ideas about things than the one used by the designer.  In HTML
4.0 the designer can specify relative locations and sizes of things down to
the pixel if they want to...which is both good and bad, depending on how
good the designer is at their job.

A lot of the audience size limitations for many web pages are due to the
designers being unaware of, or dismissive of, the fact that not everyone
uses Windows and IE.  They use proprietary technologies, such as ActiveX,
which are not available on most computers...or on any used on the net by
those who know what ActiveX really is...

MicroSoft isn't the only producer of proprietary web
technologies.  MacroMedia does too, with their Flash stuff.  They make
players for things other than Windows, but not for everything...and they
prohibit anyone else making a player without explicit written permission in
their licenses.

There's nothing stopping anyone from writing their web page using only HTML
1.0 and none of these proprietary technologies.  If all you need to do is
display some text, and an occasional picture, and you don't care a whole
lot exactly what it looks like, that will work fine.

>If current trends continue, it's only a matter of time until we bring
>back the pre-Enlightenment golden age of a few centuries ago, when
>every village had not just a different system of weights, measures,
>and currency, but even a different spoken language.

The way spoken language is going these days, I'm guessing that most of
these will consist of grunts, howls and yelps.

-- Mike B.
--
There are 10 kinds of people in the world: those who understand binary, and
those who don't.