Publicly Available Standards | Hacker News Reader

Publicly Available Standards

standards.iso.org

107 pointsTomte57 comments

Suggested title: Publicly Available Standards by ISO

ISO is somewhat unique in requiring you to pay money to even know what the standard says before you can ever think about being in compliance.

Most other standards organizations (i.e. NIST, IETF, W3C) do not have this weird financial burden, so merely "Publicly Available Standards" doesn't sell the significance of this.

Making their standards publicly available is a good move by ISO. I hope we see more of this.

This offering isn't, as far as I know, anything new.

AFAIK ISO is funded by their paywall and as an international treaty organisation is probably far harder to reform than industry organisations, so I don't have much hope for this to change.

ISO's attitude to this was recently expressed on Twitter: https://nitter.snopyta.org/MetaCatie/status/1368910356664483...

ISO even takes freely available standards from other organizations and paywalls them, even reneging on previous agreements to make a standard freely available, for example: https://www.theregister.com/2021/06/18/schematron_standard_p...

I'm not sure about the numbers but I would have thought that ISOs main funding comes from all the member nations who pay a bunch a money proportional to the countries GDP?

So they're funded by dozens of nations yet charge a fee to access their so-called Public Standards? Operational costs cannot be covered by the member countries? After all, public standards benefit everyone, and is the raison d'etre for the ISO to be constituted in the first place. A corporate paywall makes sense, not from an org acting for the public.

National standardization organisations participating in ISO usually make their budgets publicly available. Last time I looked I got an impression that the standards sales are contributing to their budgets (perhaps 1/3 or 1/4 of income). From the annual reports it was a bit unclear why the usual national standardization organization need to have wealth in $50M (instead of between 100'000 and a million) for being able to continue distributing standards.

> Making their standards publicly available is a good move by ISO. I hope we see more of this.

Afaik there were always publicly available standards in the ISO portfolio. Is this list new?

I wonder if there exists a Sci-Hub equivalent for ISO standard PDFs somewhere.

My pet fish found ISO standards PDFs on Sci-Hub a few times.

Clever girl

Some can be found on libgen.

I'm envious of people who can work with freely available standards.

In my industry (machine design, industrial automation), almost all of the standards cost money. UL, CSA, NSF, ANSI, ASME, BSi, IEC, ISO, DIN, etc. Our library contains many thousands of dollars worth of documents, and it costs many more to keep them up to date. If your company has a slow or complicated purchasing process, it's a real hinderance when you realize you need a standard and have to stop working so you can deal with acquiring it.

I think the only ones that are freely available are NFPA 70 (national electric code), NFPA 79 (electrical standard for industrial machinery), but if you want to download or print them, you still have to pay.

My library (in the same industry) contains a few hundred dollars worth of documents, because that's all I could get purchasing to pay for. We're just a small integrator, many to most in our field simply don't know what's in the documents because they don't buy any of them.

You can go a long way with cargo-cult designs, and a ton of companies get by with approximations, but there are a ton of mistakes made because ANSI/ISO/DIN/IEC refuse to make the documents available for free.

I feel your pain. Most projects I work on are medical software, often connected to hardware. Getting this stuff certified is impossible without paying thousands of euros for the standards and then you pay even more for a few people who judge if you did it like the standard says.

But maybe the next project will be outside of the medical field.

Sorry to intrude with an ignorant question, but I know next to nothing about this and it sparked my curiosity. These thousands of dollars in documents are the text descriptions/instructions of various ISO standards? Do they really cost that much or am I misinterpreting? Wouldn't their content have been pirated by now if they were so expensive? (assuming i'm correctly imagining what you said)

I challenge you to find a pirated copy of IEC 61362 "Guide to specification of hydraulic turbine governing systems" online. It does not exist. The company I used to work for, which specialized in turbine governing systems costing millions of dollars, would not pay the $300 for that document.

As pointed out in another comment and previously on HN, my new favourite source for specs are regional English translations from Estonia. Here’s one for IEC 61362 for 22 EUR: https://www.evs.ee/en/evs-en-61362-2012

> It does not exist

Sure does! First line from a random page (44) of 61362:2012: "...the runner blade opening is brought to a certain position to provide a given flow rate after load..."

They wouldn't pay $300 for it because the price is too high or because they don't need it? If their specialty makes them need the document, why would they not pay $300? I'm curious.

If you are in this type of business, you probably also get regular audits. Checking that you have access to the relevant standards is something that ends up checked from time to time. Pirated copies will do more harm then good when that time arrives.

Standards of questionable origin are fine when an engineer just needs to check something and the result is not critical. (Have you seen how many standards are called out in a simple specification document for something as mundane as a resistor? Don't tell me IEC 60757 is really worth owning!) Managers don't even have to know these standards exist.

Anything you claim compliance or certification with, though... that's your core business we're talking about! At least buy those from Estonia!

Actually, HN guidelines are to do the reverse:

> If the title includes the name of the site, please take it out, because the site name will be displayed after the link.

e.g. if this had been 'ISO announces publicly available standards' on iso.org itself, the submitter should (or if not and a popular/noticed submission, a mod would) change it to the 'Publicly available standards' that it is.

An ISO page with a list of publicly available IETF, NIST, and W3C standards would not be nearly as interesting.

Oh I see what you mean, I suppose 'Publicly available ISO standards' is more reasonable, but personally I assumed that when I saw 'iso.org'.

There's been an extensive debate on an external mailing list about standards, and why TCP won over OSI. The free and open nature of IETF had a hell of a lot to do with it.

Among all the other advantages noted here, it leads to a much shorter development cycle and a greater number of people participating -- people whose energy is invaluable.

Just as an FYI.

If you want to obtain ISO standards legally but for a price significantly lower than ISO, the answer is simple. Buy them from EVS, the Estonian Centre for Standardisation and Accreditation[1].

They have a nice English website, and you don't have to have any ties whatsoever to Eestonia, or Europe.

The customer service people speak perfect English too if you have any questions.

[1] https://www.evs.ee/en/

Wow, that’s big information. Thank you for sharing.

Its been a little while since I shopped there, IIRC their search engine is generally good, but if you see high prices come up, just make sure you're looking at the price for the EN version and not the ISO version.

EN just means the regionally adopted version of the standard. For all intensive purposes EN and ISO are typically identical[1][2]. EN is ISO with additional Z Annexes as applicable.

[1]https://northridgequality.com/difference-en-iso-14971-part1/ [2]https://www.wearwell.co.uk/whats-the-difference-between-an-i...

Also if the standard you’re looking for only has a high price in their catalog, it’s possible it’s freely available online. Not every standard is published as a regional version, generally only those that would cost more to obtain or are popular enough to require translation.

But be careful with the prefixes.

DIN EN 13850 was "Postal service - Quality of service - Measurement of the transit time of end-to-end services for single piece priority mail and first class mail", but it has been rescinded.

DIN EN ISO 13850 is "Safety of machinery - Emergency stop function - Principles for design"

> intensive purposes

"intents and purposes"?

wow they have IEC standards for 22 euro instead of 300!

ANSI used to sell many ISO standards for 18 dollars. C and C++ were available at that price. Unfortunately I didn't buy it then and now they have adjusted their prices.

If I can't read a standard to follow it without paying a lot to see it then it isn't a standard in my world. ISO is more than useless. It is actively doing harm.

These standards were made freely available because unusually enlightened committees, common sense, public interests and/or private interests occasionally prevailed over ISO traditions.

In contrast, for all other standards ISO works similarly to Elsevier.

People who write the ISO standards are required to transfer copyright to ISO, and lose control over their work. ISO says the paywall is to pay for development, but the authors don't get a cut of the sales.

https://twitter.com/isostandards/status/1367138676162105344

I don't know how these things usually get done, but the JPEG XL standard has been written for free by volunteers, and ISO paywalled it. The authors would like to change that: https://twitter.com/jonsneyers/status/1415696597585993732

The usual workaround is to publish the final preprint publicly, and give Elsevier/etc an identical paper just with the “[Preprint]” watermark or letterhead removed.

The Ada Rapporteur Group managed to have the standard be free.

ISO/ANSI still sell it, but it's identical and the free Language Reference Manual is accepted as canonical: http://www.ada-auth.org/standards/ada12.html

I think the trick was to do make sure the US Government holds copyrights, and then there is some rule that government documents need to be free.

Maybe someone can comment on the specifics, it's been years since I've heard that.

Indeed, and I encourage people to look for "preprints" and "drafts" for standards they are interested in. Sometimes you need to look for the WG pages/forums that developed it, and they are usually public, especially for the IT-related standards.

I've learned C intricacies by printing and going through an "ISO C9X" draft of the standard while in high school.

In a similar vein, it seems fairly unlikely that anyone's implementing the actual C++ specifications rather than the final drafts.

Microsoft has the ISO specs.

Microsoft was also so appalled by the ISO's modus operandi that they chose to publish their open specs with the ECMA that doesn't impose a paywall.

The Common Lisp Hyperspec is one good example of this principle.

http://www.lispworks.com/documentation/HyperSpec/Front/

By "Elsevier" you mean "research journal publishers". There's more than one, and many have the same policies, depending on the publication. Elsevier doesn't even publish the biggest journals.

What's weird to me is that nobody has to submit their standard or paper to ISO or a journal, they could just put them on a blog post somewhere. But people still complain that these organizations, who people voluntarily submitted their papers to, are actually charging people for what they said they would charge for, or take the rights they said they would take. It's not like somebody was tricked. This is like a manufacturer giving their jeans to Wal-Mart to sell and complaining that Wal-Mart is taking a cut. Yeah..... they're selling it. That's how that works. The manufacturer could always sell them out of the back of a truck... but they don't want to.

Of course you can just put your specification somewhere in a blog post, but chances are it will be taken more seriously if it goes through a process of review and ratification by national and international standardization organizations.

The point of standardization is that you reach an international consensus on how to do something. If everyone can just create their favorite variant of the spec and put it on their blog, then that would be quite fun, but it wouldn't exactly be the best way to achieve interoperability.

Here's an opinion: the world would be a better place if copyrights (on creative works and similar, including texts) lasted for several years and exclusive rights would not be transferrable.

I've thought that for a while. It gets tricky with works for hire. Perhaps that distinction needs to go away?

There are also some machine interpretable bits of ISO standards available [1].

[1] https://standards.iso.org/

Not me thinking "all standards should be freely available". Guess I'm a just filthy communist...

Or just sensible.

I wish there were more open access standard organizations.

At least members could force ISO to publish standards once a predefined amount of development costs (let's say 100k per standard is recouped).

ISO doesn't fund the development costs.

For many standards they do host the committees and do print formatting, redactions, end of life notices, etc.

Sound like a job for AI and some simple workflows

Some languages, like Ada, Java, and C# have some parts of their standards available. Notably missing are any for C++ (14882) or C (9899).

Cool, ISO 10646 was there, but sadly ISO 9995-7 is not.

Nor is ISO 10303-21 for CAD data (STEP).

That isn't very useful without 10303-11 and at least one of the model definitions that you want to exchange.

Ok, nothing under 10303 is there for free.

Reply on news.ycombinator.com