Standard interfaces

[Andrew J Bromage]

G'day all.

On Tue, Nov 09, 1999 at 05:41:42PM -0500, Alex Nicolaou wrote:

> I think you missed my point. I'm not saying standards are good or bad,
> or that de facto standards are right or wrong. I'm saying that the fact
> that people defend standards generated by competition in the market
> place is evidence that some people accept the idea of de facto standards
> and reject official standards.

If you don't mind the comment, there's quite a difference between the
way that the US looks at things and Europe looks at things here.

Look, for example, at digital mobile phone systems.  When the US wanted
to implement digital mobiles, basically every company got to design and
implement their own.  The result is that every digital mobile in the US
has to be an AMPS analogue phone too, otherwise you can't roam across
the country, into cells owned by a phone company which didn't implement
your standard.  (Well, that was the situation about five years ago.  It
may have changed a bit now.)

Europe, on the other hand, got all the manufacturers around a table and
got them to work out a standard together.  They came out with GSM, which
is what was almost universally adopted in the world outside North
America, including all of Europe and Australia and most of Africa and
Asia.

Now there are fairly sound economic reasons why the USA didn't adopt
GSM (it would have meant effectively scrapping the AMPS system and
starting over again, as happened in Australia, which in the USA would
not have been economically feasable).  But it still highlights a big
difference in culture: On the whole, Europe prefers official standards,
and the USA prefers de facto standards.  It may have something to do
with Europe being a loose federation, and not wanting to promote one
country's standard at the expense of some other country.  (The rest of
the world, of course, waits for Europe and the USA to come out with
their standards and picks whichever looks best. <grin>)

> The bottom line is that standards are an obstacle to freedom.

I could easily argue that standards promote freedom, too.

You forcing me to use a certain standard certainly does stomp on my
freedom, but picking the wrong standard reduces the freedom of my
customer base.  If, for example, I write software which uses only the
Win32 API, I reduce my customers to those who have Win32.  Those who
only have POSIX (which, when Win2K comes out, will be pretty much
everyone) I cover more people.  Therefore I should use POSIX because
it gives more people the freedom to use my software.

Now if I were to come up with my own interface to my own you-beaut
operating system, I'm in an even worse position.

> I think that there are many
> examples in the GNU software world where the software has been
> "improved" rather than conforming to the standard. To this day I hate
> operating "info" and curse the person who decided that man pages were
> not good enough; but there you have it: a better way was chosen over the
> standard way, because standards are an obstacle to progress.

Well, I can understand why info was used in preference to man pages.
Have you ever tried to find anything in the gcc man page?

	% zcat /usr/share/man/man1/gcc.1.gz | wc -l
	   4191

Ouch.

However, if the GNU project had picked (and extended if appropriate) an
existing standard (SGML comes to mind) rather than come up with their
own, imagine how much better off we'd all be now...

That, IMO, is one of the reasons for the adoption of the World Wide Web
standards as they were originally formulated: they were based on
EXISTING standards, such as SGML and MIME, and they would interoperate
with other standards like ftp and gopher if you wanted them to.

It's also one of the reasons for the popularity of C++, despite it
being a programming language theorist's worst nightmare: The new
features behaves just like you would expect C to behave under the same
circumstances.

Cheers,
Andrew Bromage