🚀 go-pugleaf

Phil Royse <proyse@peeras.demon.co.uk> writes:
|
|   I AM NOT WHINING....

I'm sorry.  I inadvertently grouped you with all the "Mommy! ISO won't
let me have their standards" folks.  I'm happy to see the exceptions.

|   But, Erik the fact that you and I can afford to pay ISO (& BSI's)
|   prices does not justify their decision to prevent them from being
|   available via a lower cost, easy-to-disseminate route such as FTPing
|   over the internet.

They don't need to justify their decision.  As long as ISO can make the
money they think they need to make by their present policies, and as
long as there are people who're willing to pay for the ISO specs, ISO
has no incentive to change their policies.  As I said last time this
thing erupted, it doesn't matter what we opine about ISO's policies, as
long as it's up to ISO to change them.  We should, IMHO, give reasonable
arguments for reaching the goals that the policies and decisions are
intended to make reachable today with _other_ policies and decisions.
We can't redefine ISO's goals, however.  If they need X dollars a year
to operate, we need to raise X + deltaX dollars a year with a new
scheme.  ISO will probably be most interested in this, and more so the
bigger the deltaX.

I think this discussion would benefit from regarding ISO as the
publisher it is, and refrain from taking the consumer side, that ISO's
in the business to benefit the readers of their publications.  No other
publisher has such ardent demands on them to make their books and other
publications available for FTP.  (Although, ever since I was delivering
newspapers in the biting cold of Oslo's winters, I have a Grand Plan to
let subscribers get their newspapers delivered electronically.  Today,
it's technically feasible.  Ten years from today, maybe it's reality.)

|   Precisely because that knowledge which I can gain from them will
|   be to the client's benefit, and indirectly to the whole comms community.
|   The more people who know more about OSI the better for OSI.

The latter point is true, but ISO probably figures that since you
benefit from their work, you should pay them for doing it.  I can't
disagree with this, or I'd have to give my own service away for free,
too.  Not that many experts don't already, in this present medium!

|   That's a fair point for you and I, but what about students and young
|   people starting off in a career in comms - surely it's good for them
|   to be able to have low cost acces to OSI docs & specs?

Yes, it's good for them.  That's why Universities should be more
involved in their work, and should subscribe to the working documents of
relevant committees, with unobstructed access for students and faculty.
(It was by accident that I discovered that this University keeps many
past working documents in my committees of interest.  They are not part
of the library, but are in an obscure, locked lab devoted to completely
different things.  Go figure.)

I think libraries should be the low-cost venue to OSI docs and specs,
not ISO.  Hey, it works with other publications.  I don't subscribe to
the Federal Register (equivalent), but I still have unlimited, free
access to it, provided I'm willing to travel downtown and visit the
government libraries.  (I'm assuming that people actually live in
relative proximity to the center of the world, as it were, not several
million miles out on the equivalent of a wayward space station.)

|   What about the small software authors - the ones who potentially
|   could write REALLY low cost implementations?  The easy dissemination
|   of OSI standards would encourage PD and shareware for OSI.  I see a
|   great benefit in (say) an MHS UA for DOS at $25.00, or a TP4/8473
|   stack for $50.00, not the large software fees that these products
|   currently command. (Eg. Unisys U6000 stack for TP4/8473 (over
|   8802/3, X.25 etc.) runs at 1200 UK pounds!

I see your point.  I just don't think it's ISO's job to cater to it.

Also, Look at the number of books about ANSI C.  Look at the cost of
these books, and how many your average reader will buy.  Compare with
the cost of ANSI X3.159.  Books cost more, are inaccurate, and are
written by people who are very liberal about interpreting the standard,
which they probably can't even read.  (I hold courses in ANSI C, and it
was absolute hell trying to find good textbooks among the trash.)

|   >Another thing to remember is that standards are often held by large
|   >public libraries, and specialized libraries.
|
|   Yes, I ought to start investigating this route...

I think this is the least known and most cost-effective way for a
student or other interested party to learn about things.  If you plan to
make a living, you need the specs yourself, but then it's a deductible
cost, too.  I think ISO knows this.  It would perhaps help if we could
compile a list of major Universities who have working document archives,
subscribe to national and international standards, etc?  (The Norwegian
member body to ISO gives handsome discounts to educational institutions
on national standards, but I haven't seen anything similar for ISO or
other standards.)

|   My DIS access to the internet costs ten UK pounds per month, no
|   usage charges and only local telephone costs.  The s/w is an
|   FTP/Telnet/TCP/IP stack running over PPP, in DOS.  It is available
|   as free copyright s/w, which underwrites my earlier point - when
|   will the equivalent OSI stack be so cheaply available?

This is impressive.  I wasn't aware of this level of availability.
Completely tangential to the issue at hand: is this ISDN at 64kbps or a
slow modem at, say, 19.2kbps?  Maybe I should dust off the plans to
erect such a network in and around Oslo, too?  Last time I tried, I lost
so much money I still haven't completely recovered.  That was five years
ago.  At GBP 10 a month, I can foresee more losses, though...  :-(

|   OK, but I don't need the whole text of most standards, just enopugh
|   to understand them.  For example, a few months ago I needed ISO
|   10589 in order to formulate a router strategy for a very large
|   public sector client.  I needed to know all about the concepts of
|   IS-IS routing, areas, subdomains etc. in order to plan ahead and
|   spec. our purchasing policy.  It was also necessary to plan for
|   routing policies in the initial development of the WAN.  I didn't
|   need all the detail stuff at the end of the standard on encoding
|   PDUs.  ( I suppose I need "technical overview" material, but that is
|   never avialable except in the body of the standard.)

This is a point I can relate to, and fully.  I've always wanted to have
what I would dub "a conceptual introduction to X for people who can read
the whole standard, but don't have time to".  Something about this idea
tells me that it would be foolish to try to write these for profit, as
they would easily become even more expensive than the standards
themselves, and maybe sell only twice as many copies.  Maybe it turns
out that ISO standards aren't so expensive, after all?  After all, they
only cover the costs of the organizational overhead, as opposed to the
development costs, which would easily run into several hundred million
dollars for even small standards.  These are paid for by vendors and
others who send experts to meetings, and by those who host meetings.
I'm sure glad I don't pay for the development costs...

I think the newsgroups could be a little more densely populated with
such conceptual introductions, however.  This is an excellent forum for
just the students and other interested parties you allude to.  Funny,
though, how the Internet protocol suite, and a handful other (ANSI, ISO,
IEEE) standards get wide coverage, but almost nothing about OSI
standards...  Something tells me that this is not ISO's fault at all.

|   Yes, that must work for some consultants, and I know many who are on
|   BSI, EWOS, OSI committees.  But not all consultants can be, that's
|   illogical.

Don't worry.  There are more committees than there are independent
consultants.  Well, not really, but we won't run out of committees to
absorb consultants in a long time.  Besides, committees reproduce faster
than independent consultants.

|   >OSI is spreading faster than Internet protocols did with the same
|   >distribution.  The Internet growth is approaching linear growth,
|   >while we're only in the early exponential phase of OSI protocols,
|   >still.  That means Internet protocols have a pure time advantage it
|   >will take many years to overtake, but with the commercial drive
|   >behind OSI, I would hedge my bets.

This paragraph has caused more comments than I anticipated.  If we're
looking at the user base for OSI products, and at the time from initial
deployment until now, and compare this with the user base at the same
time in the life of the Internet suite, or at the time it took the
Internet suite to reach the same user base, OSI is leading.  This could
_very_ well be because the Internet protocols pioneered the field and
blazed the trail.  In fact I think it did.  However, as far as I can
tell from the data I receive (from the Internet Monthly Report, and a
few other sources), the growth of the Internet is accelerating at a
slower pace, that is, if its growth matches (x^t), for some x, x is
getting smaller, indicating that we're approaching linear growth.

|   Where do you get your figures?  My understanding (without any
|   official stats) is that OSI is creeping slowly into the public
|   sector, but TCP/IP systems (and IP router) sales are booming in the
|   private sector, and keeping ahead in the public sector.

Sure, but a handful of years has an enormous effect on exponential
growth curves.  I'm assuming a growth curve which goes linear spread,
through exponential growth, to linear at the tangential, and matures at
a logarithmic curve (range (0,->)) approaching saturation.

(Note: if you didn't read all of this article, please mail me and
complain about its length.)

Best regards,
</Erik>
--
Erik Naggum             |  ISO  8879 SGML     |      +47 295 0313
                        |  ISO 10744 HyTime   |
<erik@naggum.no>        |  ISO 10646 UCS      |      Memento, terrigena.
<enag@ifi.uio.no>       |  ISO  9899 C        |      Memento, vita brevis.