Re: [RFC] Restructuring of the Debian Project

[kai@khms.westfalen.de (Kai Henningsen)]

bruce@pixar.com (Bruce Perens)  wrote on 06.03.97 in <m0w2kQf-00MFgDC@golem.pixar.com>:

> > Incidentally, there's ECMA, who publishes their stuff for free. That may
> > be because ECMA is the European Computer Manufacturers Association - that
> > is, their members are used to paying for standards creation, not to
> > charging for standards distribution.
>
> Yes. Indeed they sent me _their_version_ of the ANSI terminal standard
> via international air-mail when I requested it. They spent 3 or 4 bucks
> to do so. They could have saved the money and put the document up for
> electronic retrieval, but I doubt their license from ANSI would allow
> them to do that.

Incidentally, they do that, for the standards they have an electronic form  
of. See http://www.ecma.ch/.

And I don't think they have a "license from ANSI"; I suspect they simply  
are an ISO member or some such. However, this ought to be somewhere on  
that web page. Interestingly enough, their paper standards don't seem to  
include *any* copyright reference.

Of course, everybody reading comp.std.international (or cs.unix or ...)  
knows that the copyright statements from ANSI, ISO etc. are on fairly  
shaky legal grounds - the authors (committees) don't seem to sign them  
over, they don't seem to sign anything at all. AFAIK, those copyrights  
have never been tested in court, though I'm no expert.

> Kai:
> > And I gather all the computer standards committees would much rather have
> > the standards available for free, too. Or at least for reasonable prices
> > (like other computer books, that is).
>
> The problem is they currently don't have any standards organization
> _other_than_IETF_ that works that way. I think IETF is one of the most

I just pointed out that ECMA does.

> successful standards organizations ever (otherwise the Internet would not
> work), just because it works with little bureaucracy.

Probably more to do with the "two cooperating implementations" rule.  
However, the fact that their committees - even if they meet at lush  
resorts - are, these days, essentially mailing lists (and the lists count  
more than the meetings) also contributes. This is how I can participate in  
the current renovation of the mail standards (RCFs 821/822 and friends,  
the DRUMS working group, tends to go up in flames because of a certain  
mailer author who also participated in some flames on debian-devel some  
time ago and whose software is involved in running these lists, and whom I  
don't name because that tends to guarantee his appearance).

Yes, it would be nice if every standards organization worked like the IETF  
or ECMA. It would also be nice if every program was available for free,  
and there was no hunger or war in the world.

Unfortunately, these goals are of approximtely similar probability.

Personally, I'll settle for reasonable prices for those standards.

(Incidentally, I gather that the FSF had input into both the ANSI/ISO C  
standard, and at least one of the POSIX standards. I'm not clear on how  
this worked; my copy of POSIX.1 has been vanished, and my copy of ANSI C  
doesn't (at first glance) list an obviously FSF name.)

> > This may, or may not work; however, I don't think it would be wise to
> > have  standards that conflict with what the rest of the industry uses.
>
> No, I would expect the free standards to be supersets rather than to
> conflict with things like POSIX. They could be designed while consulting
> the other standards, however we would have to write a new standard rather
> than plagiarize and old one word-for-word.

This is, of course, a *lot* of work.


MfG Kai