Re: ISO standards costs

From: Markus Kuhn (
Date: Sat Mar 04 2000 - 06:32:52 EST

"Robert A. Rosenberg" wrote on 2000-03-03 20:39 UTC:
> Forget about the Unicode Standard book's cost (wrong paradyne since you are
> comparing Paper vs Paper that way). ISO should realize and enter the 21st
> Century (which is already here [or will be in less than 10 months <g>]) and
> stop pretending that the cost of creating a CD is based on the number of
> pages of content it contains. Once the actual content is created, the cost
> of PRODUCING the CD is the same whether the media contains a 50Meg document
> or 650Megs of Documents. For a paper copy there actually is some
> justification for a per-page price since the larger it is, the more it
> costs to produce (more paper, ink, time to print the required number of
> total copies, etc.) unlike a CD where the CD is pressed and thus the amount
> of data on it is not an issue (ie: It takes the same time/effort and cost
> to press a 50Meg CD as a 650M one).

The problem with ISO is far more fundamental. They insist on using the
very same completely outdated production process for all of their paper
standards. Most of the over 12000 ISO standards are as obscure as

  ISO 4074-9:1996 Rubber condoms -- Part 9: Determination of tensile properties


  ISO 1208:1982 Spices and condiments -- Determination of filth

and only a few hundred copies are ever sold (most to places like patent
offices, national standards bodies, and big corporate libraries that
archive all of them systematically). Very few people ever read them.

I had an opportunity to visit ISO in 1995. ISO has its own very small
printshop in Geneva (I'd estimate fewer than 25 employees there), where
they produce all the standards and handbooks in editions of 500 copies a
time. The fixed production overheads are quite significant, because they
throw away the offset plates after printing only 500 issues and they use
very small sheet presses and a manual binding process. They have a huge
warehouse where they can store up to 500 copies of each of the 12000
standards plus the camera-ready copy (often some paper cut&paste
product) to produce more. [Whenever a fly shits onto the camera-ready
copy in ISO's warehouse, you run the risk of having a new decimal point
added to the authoritative master version of an ISO standard forever,
which is the true reason why ISO strictly prefers decimal commas ... ;-]

Their 1950s infrastructure and processes might once have been suitable
for low-volume obscure standards, but it is completely inappropriate for
standards such as UCS, C, POSIX, and others that ten thousands of
developers would love to have on their desk if they were available at a
normal school book price in every bookshop. ISO really should hand over
the production and shipping of at least the more popular JTC1 standards
to established international publishers of technical and academic
literature who know how to efficiently turn PDF into filled bookshop
shelf space (Addison-Wesley, Springer, and friends). For the low-volume
standards, ISO and its national members should buy a few of the larger
Xerox document printers that can now perform high quality binding fully
automatically directly after printing, and use it to do pure on-demand
printing from PDF master copies, which should also be available on a
small set of quarterly updated CD-ROMs. Get rid of the entire expensive
printshop and warehouse. Turn ISO into a small e-commerce business.
ETSI, ITU, ECMA, and all the others have already done this long ago.
ISO's and IEC's market flexibility can only be described as pathetic.

But then, ISO's management is perhaps not not the real problem here, but
the few influential national member bodies that fear change most as it
would also affect their profitable national monopoly business (DIN, to
name *the* prime culprit of international standards publication
inefficiency, closely followed by ANSI and BSI).


Markus G. Kuhn, Computer Laboratory, University of Cambridge, UK
Email: mkuhn at,  WWW: <>

This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:20:59 EDT