Are these files precise copies of files in the ISO standard

[velle]

Most of us don't have access to the official ISO standard, e.g. the latest ISO/IEC 19757-3:2025.

This repo is labeled as: Unofficial copy of schema(s) for Schematron

Does that mean that these are exact copies of files that are part of the iso standard?

Or are they adapted in some way, maybe by fixing "bugs"? Or maybe more limited/simplified, or alternative implementations?

[velle]

I found this snippet in all four schema files:

# "THIS SCHEMA HAS BEEN MODIFIED FROM THE SCHEMA DEFINED IN ISO/IEC 19757 3, 
# AND SHOULD NOT BE INTERPRETED AS COMPLYING WITH THAT STANDARD". 

So that actually answers the most essential part of my original question.

However, it seems that these disclaimers could have been added many years ago (header suggests 2017 and 2019). Do you know if its still the case that they are different from the ISO files, now for the 2025 version? And if so, could you please share an example of such a difference?

[AndrewSales]

Hi @velle and thanks for your good questions.

Does that mean that these are exact copies of files that are part of the iso standard?

They should be, in principle (errors and omissions excepted).

Or are they adapted in some way, maybe by fixing "bugs"? Or maybe more limited/simplified, or alternative implementations?

The history around this and original rationale for placing them in this repo is that, thanks to the copyright notice, they are the only portion of the standard that can be reproduced without violating ISO copyright.

Since then, I have been updating them while editing the standard to give at least some visibility of the directions being taken and it's also useful if we need to refer to proposed schema changes in the other repos here, such as (and especially) the enhancement proposals.

However, it seems that these disclaimers could have been added many years ago (header suggests 2017 and 2019)

The text you refer to relates to the immediately preceding paragraph, "any modified copy of the Schema shall include the following notice:", in the event that someone makes their own copy of the schema. Does that help, or have I misunderstood?

[velle]

Thanks :)

The text you refer to relates to the immediately preceding paragraph, "any modified copy of the Schema shall include the following notice:",

If I understand correctly: its mostly if someone actually makes a modified version that this disclaimer becomes relevant. But since there might be minor corrections / omissions, it is probably still a tiny bit relevant/required to to add this disclaimer directly in the https://github.com/Schematron/schema/ repo.

But any such differences from the original ISO version are so insignificant that I should not worry about that.

Let me know if I misunderstood.

[rjelliffe]

Yes, the intent is that it is fine to have a customized version of the Standard schemas, as long as you note they are customized. And if you use a customized version to validate, you note this. I think it is usual practise, and IIRC correctly I did it after feedback from the ISO working group. But there should be no problem in hosting your own versions, distributing them, etc. If you extend Schematron or restrict it, then that is fine as long as you are upfront about it. (Extend it by using a different namespace, otherwise it is rude and disfunctional, of course. But if your idea works out well in implementation, suggest it to editor Andrew and the Schematron folk or to the ISO WG perhaps through your working group representative at ISO if your country's national body has a representative.) Ideally, ISO should publish the schemas free on their site, of course. But unless national bodies fund ISO so that it doesn't have to rely on sales (or price things in ways that small business, students and private individuals can afford), then that is not reliable. (Of course, that relies on national bodies having their own funding that doesnt rely on sales, which comes down to governments: government's really botch the productivity strategy by not funding national and consortium standards bodies adequately: I think it is part of the strategy of some nations which want to support their businesses by trying to suppress technologies that come from outside their country, rather than allowing a level-playing field. ) Rick

…

On Thu, May 21, 2026 at 4:46 AM velle ***@***.***> wrote: *velle* left a comment (Schematron/schema#11) <#11 (comment)> Thanks :) The text you refer to relates to the immediately preceding paragraph, "any modified copy of the Schema shall include the following notice:", If I understand correctly: its *mostly* if someone actually makes a modified version that this disclaimer becomes relevant. But since there might be minor corrections / omissions, it is probably still a tiny bit relevant/required to to add this disclaimer directly in the https://github.com/Schematron/schema/ repo. But any such differences from the original ISO version are so insignificant that I should not worry about that. Let me know if I misunderstood. — Reply to this email directly, view it on GitHub <#11 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AF65KKOVNXNXM736ZU4MHXT43X4PVAVCNFSM6AAAAACZCNAAYKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHM2DKMBRGU3TGMJSGY> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>. You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

[rjelliffe]

I wonder if the schemas for Schematron could have more of the standard reflected into them? To expose more? Rick On Thu, May 21, 2026 at 12:47 PM Rick Jelliffe ***@***.***> wrote:

…

Yes, the intent is that it is fine to have a customized version of the Standard schemas, as long as you note they are customized. And if you use a customized version to validate, you note this. I think it is usual practise, and IIRC correctly I did it after feedback from the ISO working group. But there should be no problem in hosting your own versions, distributing them, etc. If you extend Schematron or restrict it, then that is fine as long as you are upfront about it. (Extend it by using a different namespace, otherwise it is rude and disfunctional, of course. But if your idea works out well in implementation, suggest it to editor Andrew and the Schematron folk or to the ISO WG perhaps through your working group representative at ISO if your country's national body has a representative.) Ideally, ISO should publish the schemas free on their site, of course. But unless national bodies fund ISO so that it doesn't have to rely on sales (or price things in ways that small business, students and private individuals can afford), then that is not reliable. (Of course, that relies on national bodies having their own funding that doesnt rely on sales, which comes down to governments: government's really botch the productivity strategy by not funding national and consortium standards bodies adequately: I think it is part of the strategy of some nations which want to support their businesses by trying to suppress technologies that come from outside their country, rather than allowing a level-playing field. ) Rick On Thu, May 21, 2026 at 4:46 AM velle ***@***.***> wrote: > *velle* left a comment (Schematron/schema#11) > <#11 (comment)> > > Thanks :) > > The text you refer to relates to the immediately preceding paragraph, > "any modified copy of the Schema shall include the following notice:", > > If I understand correctly: its *mostly* if someone actually makes a > modified version that this disclaimer becomes relevant. But since there > might be minor corrections / omissions, it is probably still a tiny bit > relevant/required to to add this disclaimer directly in the > https://github.com/Schematron/schema/ repo. > > But any such differences from the original ISO version are so > insignificant that I should not worry about that. > > Let me know if I misunderstood. > > — > Reply to this email directly, view it on GitHub > <#11 (comment)>, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/AF65KKOVNXNXM736ZU4MHXT43X4PVAVCNFSM6AAAAACZCNAAYKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHM2DKMBRGU3TGMJSGY> > . > Triage notifications on the go with GitHub Mobile for iOS > <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> > or Android > <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>. > > You are receiving this because you are subscribed to this thread.Message > ID: ***@***.***> >

[rjelliffe]

I mean such as ``` < sch:title>Elements</title> < sch:group id="patterns"> < sch:title>Patterns< /sch:title> < sch:p>A pattern element holds rules (which hold assertions) in first-match priority, like a switch statement. In contrast, a group element holds rules (which hold assertions) so that each rule is tested. It is not an error if no rule matches. < /sch:p> < sch:p>An abstract pattern is one that is parameterized to allow systematic variations of a common base.< /sch:p> < sch:rule ***@***.***"> < sch:assert test="key('abstract-pattern', @is-a)" id="abstract-patterns"> The is-a attribute of a pattern element shall match the id attribute of an abstract pattern. < /sch:assert> < sch:assert test="every $name in ***@***.*** satisfies $name eq key('abstract-pattern', @***@***.***" id="param-names-match"> The name attribute of a pattern param element shall match the name attribute of an abstract pattern param element. < /sch:assert> < /sch:rule> < sch:rule ***@***.***/sch:param"> <sch:assert ***@***.*** or key('abstract-pattern', ..***@***.******@***.***" id="missing-param-value">A pattern parameter value shall be specified.< /sch:assert> < /sch:rule> < sch:group> ``` I wouldn't see the need to reflect the content model in, from the rnc, but it would be possible. ``` < sch:assert test="count(sch:*) = count(sch:let) + count( sch:title) + count(sch:p) + count(sch:rule) + count(sch:param)"> A pattern element holds rules and let elements. It also may contain documentary p and title elements. Abstract patterns and concrete patterns may also have param elements. < /sch:assert> ``` On Thu, May 21, 2026 at 1:03 PM Rick Jelliffe ***@***.***> wrote:

…

I wonder if the schemas for Schematron could have more of the standard reflected into them? To expose more? Rick On Thu, May 21, 2026 at 12:47 PM Rick Jelliffe ***@***.***> wrote: > Yes, the intent is that it is fine to have a customized version of the > Standard schemas, as long as you note they are customized. And if you use a > customized version to validate, you note this. I think it is usual > practise, and IIRC correctly I did it after feedback from the ISO working > group. But there should be no problem in hosting your own versions, > distributing them, etc. If you extend Schematron or restrict it, then that > is fine as long as you are upfront about it. (Extend it by using a > different namespace, otherwise it is rude and disfunctional, of course. But > if your idea works out well in implementation, suggest it to editor Andrew > and the Schematron folk or to the ISO WG perhaps through your working group > representative at ISO if your country's national body has a representative.) > > Ideally, ISO should publish the schemas free on their site, of course. > But unless national bodies fund ISO so that it doesn't have to rely on > sales (or price things in ways that small business, students and private > individuals can afford), then that is not reliable. (Of course, that > relies on national bodies having their own funding that doesnt rely on > sales, which comes down to governments: government's really botch the > productivity strategy by not funding national and consortium standards > bodies adequately: I think it is part of the strategy of some nations which > want to support their businesses by trying to suppress technologies that > come from outside their country, rather than allowing a level-playing > field. ) > > Rick > > > On Thu, May 21, 2026 at 4:46 AM velle ***@***.***> wrote: > >> *velle* left a comment (Schematron/schema#11) >> <#11 (comment)> >> >> Thanks :) >> >> The text you refer to relates to the immediately preceding paragraph, >> "any modified copy of the Schema shall include the following notice:", >> >> If I understand correctly: its *mostly* if someone actually makes a >> modified version that this disclaimer becomes relevant. But since there >> might be minor corrections / omissions, it is probably still a tiny bit >> relevant/required to to add this disclaimer directly in the >> https://github.com/Schematron/schema/ repo. >> >> But any such differences from the original ISO version are so >> insignificant that I should not worry about that. >> >> Let me know if I misunderstood. >> >> — >> Reply to this email directly, view it on GitHub >> <#11 (comment)>, >> or unsubscribe >> <https://github.com/notifications/unsubscribe-auth/AF65KKOVNXNXM736ZU4MHXT43X4PVAVCNFSM6AAAAACZCNAAYKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHM2DKMBRGU3TGMJSGY> >> . >> Triage notifications on the go with GitHub Mobile for iOS >> <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> >> or Android >> <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>. >> >> You are receiving this because you are subscribed to this thread.Message >> ID: ***@***.***> >> >

[velle]

Thanks for the very elaborate answers :)

If you extend Schematron or restrict it, then that is fine as long as you are upfront about it.

Got it. I have no such intentions though, but I still appreciate the explanation.

Ideally, ISO should publish the schemas free on their site, of course.

I think everybody would agree. So its appreciated that you guys try share what you can here.

I don't quite understand the idea (link) you are contemplating. But when I hopefully get to a level where I understand and have any feedback, I will return to give my thoughts.

[Hibou57]

Do someone know if this book contains the same information as the ones in the ISO document? : https://xmlpress.net/2026/05/31/schematron-2025-a-language-for-validating-xml/ . I wonder, because it is far more affordable than the ISO.

[AndrewSales]

Do someone know if this book contains the same information as the ones in the ISO document? : https://xmlpress.net/2026/05/31/schematron-2025-a-language-for-validating-xml/ . I wonder, because it is far more affordable than the ISO.

@xatapult can you comment here (as the author!)?

[xatapult]

@Hibou57 My book about Schematron 2025 is based on the ISO document and, unless I've overlooked something, contains the same information. It is of course very differently organized: standards are written for engine implementors, my book is for users. I have tried to organize and describe the material in such a way that it is easier to understand. There is a learning curve in the book: it starts with the basics and then describes the more complex topics.

As a bonus the book contains "technology starter" appendices that describe XPath and Namespaces.

Although I am obviously biased, I would not recommend spending your money on the standard. That is a formal and terse document (and it should be!) but, IMHO, doesn't add much value for mere users. Either buy my book 😉 or look on the web for information and examples.