Should we require @dtd-version?

[Klortho]

See http://jatspan.org/niso/publishing-1.1d3/#p=attr-dtd-version.

This attribute is defined as "FIXED", which means:

• The attribute is optional -- it doesn't have to be present in an instance document
• If it is present, its value must be the exact same as that specified in the DTD, e.g. "1.1d3".

It's a nice attribute, because it enables consumers (bots) to determine the version of the schema that the document conforms to, without having to read the doctype declaration or any processing instructions. OTOH, making this attribute required by JATS4R would put an extra burden on the content providers. Then again, it's best practice, and I'd guess that most of them are doing it anyway.

[jats-laura]

I'd guess that most of them are doing it anyway

I don't think they are, actually. PMC gets very little data with the attribute explicitly stated.

[hubgit]

Presumably the intention is that the value is read from the DTD rather than the XML file, so there's no reason to add it to the XML.

However, if we're encouraging re-use in environments that don't use the DTD, it might make sense to require it. It would have to be accurate, though…

FWIW, PeerJ's articles do have this attribute set.

[Klortho]

I think it's a good best practice. What if we made it a warning if the attribute is not present?

[Klortho]

It came up in the call just now, that we can use @dtd-version in the Schematron, to determine which license rules to apply (see #50). That would simplify things quite a bit for us. And using the test that "what's good for us is good for reuse", suggests that we should require it.

[ahamelers]

There seemed to be widespread support to require it on the call. Can anyone
think of any compelling reason not to require it?

On Thu, Oct 15, 2015 at 3:03 PM, Chris Maloney [email protected]
wrote:

It came up in the call just now, that we can use @dtd-version in the
Schematron, to determine which license rules to apply (see #50
#50). That would
simplify things quite a bit for us. And using the test that "what's good
for us is good for reuse", suggests that we should require it.

—
Reply to this email directly or view it on GitHub
#113 (comment)
.

[Klortho]

This is a warning if it is not there.