NetNews Usenet Archive 1992 #18/NN_1992_18.iso/spool/comp/protocol/iso/1018

Internet Message Format | 1992-08-19 | 8.5 KB

Path: sparky!uunet!kithrup!hoptoad!decwrl!deccrl!bloom-beacon!eru.mt.luth.se!lunic!sunic!ugle.unit.no!nuug!ifi.uio.no!enag
From: [e-mail] (Erik Naggum)
Newsgroups: comp.protocols.iso
Subject: Re: Archive site for docs?
Message-ID: <[e-mail]>
Date: 19 Aug 92 13:03:23 GMT
References: <[e-mail]> <[e-mail]> <[e-mail]>
Organization: Department of Informatics, University of Oslo, Norway
Lines: 134

James S. Vera <[e-mail]> writes:
|
|   While the above is certainly ISO's policy on the matter, it is an
|   unfortunately policy which is contributing to OSI's failure in the
|   marketplace.  The free exchange of standards documents provides a
|   springboard for the development of compatible products.  As for the
|   danger of tampering, this is nonsense.  What do you think would
|   happen to a producer of a TCP/IP product which didn't interoperate
|   with the net?  Claims that their spec document was defective aren't
|   going to buy them a thing.  If ISO provided the documents online,
|   the quality ftp cites would carry them and no problems would occur.

Please note that the International Organization for Standardization is
much, much more than Open Systems Interconnection, and that ISO has a
real concern for the integrity of their standards in a much wider
context than the connected networks.  The standards are copyright ISO
and sold through authorized outlets, so that you know you get a real ISO
standard when you buy one.  Since most end users of ISO standards and
ISO standard conforming products don't even know ISO exists, much less
want to read the specifications, it would be a simple matter for one
vendor to edit out the hardest parts to implement, increase an error
margin, or whatever, and to supply the customers with this copy.  This
would be much easier if they were electronically available, and ISO has
a real concern that ISO standards would be diluted if a tight control
over their integrity is not maintained.  Considering the many companies
who already claim their deficient software and systems are conforming, I
certainly understand this attitude.

Now, you can't compete with something which is freely available, such as
the RFC's.  RFC's also have built-in integrity checks in that anyone,
wherever they are on the connected Internet, can FTP to nic.ddn.mil and
grab the authoritative copy with no delta cost.  We all know that
getting on the connected Internet isn't exactly free, though.  In fact,
I can't go anywhere and "buy" Internet connectivity off the shelf.  Then
there's the Acceptable Use Policy, which means the customer can't use
the network for many of his needs, _if_ he gets connected in the first
place.  For _us_ on the inside, the RFC's are free, but for _them_ on
the outside, they are essentially not available, except by the favor of
a nearby University or direct-ordered from the NIC, ordering info being
available on the net.

You can say whatever you want about OSI, but ISO and CCITT aim at having
the networks operate in the commercial world with commercial up-time and
no silly restrictions on what you send on the wire.  They also think
it's fair that anyone can buy their standards from them or from national
member bodies, and not have to be connected before they can get things.

|   Not a small amount from what I recall.  Certainly not for a college
|   student seeking to explore.

This situation is new to ISO.  Never before have their standards been of
such interest to more than vendors and manufacturers.  Academia has been
actively hostile towards ISO and standards which "thwart" creativity and
are big and unimplementable by your favorite graduate student and would
ruin the University library if he should want access to all of them on
the same basis that he expects to have access to other documentation.

We should stop staring so intently at our navels and realize that we're
damn lucky and privileged to live in the electronic society where
information flows with no delta cost, and that we can't expect to see
this happen to the rest of the industrialized world for decades to come,
if ever.  When computer networks connections are as ubiquitous as the
telephone wall sockets (and we're not talking ISDN, here, mind you; it
isn't even enough for personal use) and the hub computers are big enough
to handle thousands of FTP sessions to retrieve multi-megabyte pictures
and other wasteful multi-media "infotainment" for the masses, you should
also be able to retrieve documentation and standards for "free", meaning
that your 50M standards documents is delivered to your computer in about
25 millisecond (AT&T plan to have 8Gbps pipes to all residences and
offices with an expected throughput of 2Gbps), at a cost so marginal
that you would perhaps have to do without a lump of sugar for your cup
of coffee to afford it.

We're living on the forefront of available technology, and we enjoy it
immensely, but we should never forget that this forefront costs a hell
of a lot of money which mostly governments have poured into "research"
so we can complain to each other that ISO documents aren't as free as
the T1 we have to the the T3 backbone, either of which we didn't pay
for ourselves.

We've got to understand that this is luxury, paid for by taxes and the
infrequent private donation to give a few people the knowledge and
vision to shape the future for the rest of the people.  We should not be
so damn snotty about it that we don't want to stoop to the level of mere
mortals and pay for things once in a while, no matter how much we "need"
it or "could use it" or for whatever beneficial academic cause we can
dream up.

Notwithstanding the disgusting company which used this slogan first, ISO
attends to the needs of the rest of the people, and the rest of the
people are going to pay for the standards that the manufacturers had to
buy for what seems like much money to a college student with a miniscule
fraction of the end product cost.  If the manufacturers had people on
the ISO committees, the cost of the standards themselves are blown out
of the water by costs three of four orders of magnitude larger for
sending experts all over the world to luxury hotels to twist a phrase
here and phrase there and perhaps reach a compromise about which people
in academia are going to spend megabytes of complaints and cries of
"mindless, OSIfied requirements" in comp.protocols.iso and elsewhere.

The procedure and the results can be understood and appreciated, but it
requires that your favorite academic institution and/or company is
willing to pay for the forest ISO kills every year in preparing
standards for, among other things, electronic document interchange.  (I
know, it's laughably tragic.)  I can do it, and I'm a lowly undergrad
(still) at the University of Oslo doing some consulting jobs whenever I
need money, and benefiting so much from the U of Oslo's Internet link
that I probably couldn't pay for comparable services if I worked 14
hours a day year 'round.  I consider spending about $5,000 on ISO and
CCITT standards over a three-year period to be affordable, especially
since I don't smoke or drink, and walk to the U and don't incur the
"normal" society's cost of living.  The U gets my expertise back for
free, and so does comp.text.sgml, where I write the most.

Maybe I'm just stupid, but I figured that if I wanted to learn about
this, spending a year doing so is more costly than the standards, and I
would use more money on textbooks than standards had I aimed at getting
credit points with the same level of effort I invest in standards.
Considering that I've become one of the world's handful of experts on
SGML and can reap high consulting fees now and hopefully a high salary
some years down the road, this is not such a bad deal.

Take a look at the price commercial enterprises pays for OSI courses.
If you buy the OSI standards set, and learn enough about them with the
normal academic quality standards (which are *way* higher than the
commercial quality standards), you should be able to make a killing
teaching other people, helping manufacturers and developers, etc, etc.
I manage to make a living working perhaps 12 weeks a year.  You can do
it, too, if you put your money where your mouth is and make a longer-
term cost-benefit analysis than this "Waaah, I can't get ISO 10744 with
FTP."

Best regards,
</Erik>
--
Erik Naggum             |  ISO  8879 SGML     |      [phone]
                        |  ISO 10744 HyTime   |
<[e-mail]>        |  ISO 10646 UCS      |      Memento, terrigena.
<[e-mail]>       |  ISO  9899 C        |      Memento, vita brevis.