*id attributes on xml and html spec

[timbrisc]

Issue migrated from trac ticket # 72

component: help | priority: minor | keywords: xml rdf sdmx

2011-03-11 11:00:59: yjaques@gmail.com created the issue

---

First off, thanks for all the work, truly amazing. I'm putting together a standards doc for a UN agency, and I would love to be able to point to actual IDs for the units in the spec - that way I can assign URIs to each code item in my document making the link to UCUM standards extremely clear. Could you guys do that? This way I would be able to put in a URI like:

http://aurora.regenstrief.org/~ucum/ucum-essence.xml#kcal

I think this is actually of major priority for anyone who is attempting to use XML and/or RDF to point to your standards. Ditto for those of us working with SDMX.

[timbrisc]

2011-03-13 20:05:27: werner.keil@gmx.net commented

---

Interesting. SDMX-ML seems more alive than Units-ML, at least telling from the history of some of the documents. I tried to get in touch with people related to Units-ML earlier during a review phase, but no response, so I consider that format not so relevant or barely used at the moment. The research especially with Java standards like JAXB or XMLBeans was however not lost and I'd be happy to apply some of that to your XML-schema and related files. Markup-support is part of the UOMo XML project, but what you propose also seems to require changes on the UCUM side. Please let us know here, if especially the typesafe Units of Measurement support by Unit-API or implementations like UOMo should be of interest for any of the Java tools for SDMX?

[timbrisc]

2011-03-16 17:04:54: yjaques@gmail.com commented

---

Thanks for your detailed reply. Yes SDMX is very active as it has the strong backing of large organizations such as World Bank, IMF, OECD, BIS, ECB, UNSD, etc. In addition, EuroSTAT is requiring that shortly all member countries report in this format. Though I am not on the implementation side I work closely with the implementors and I can tell you there is a strong likelihood Java will be the chosen technology for the system that is being fed requirements by the work my group is doing. In this case I assume your libraries would be useful and I will be sure the implementors are aware of them. The challenge I have found in defining corporate standards for classifications in general is the relative lack of stable, unambiguous URIs that can be referenced. To give you a clear example of what I'm talking about: http://aims.fao.org/aos/agrovoc/c_2724 If you point to this with an RDF browser you get back RDF/XML that describes using standard RDF elements what the definition means. My dearest hope is to be able to do the same for things like measurement units, and UCUM seems to me like the group that has the technical know how and the credibility to do it. BTW, I adore the precise syntax and grammar section of your spec.

[timbrisc]

2011-07-26 19:18:40: gschadow@pragmaticdata.com commented

---

Enumerating a list of all UCUM units is not possible, because the space of combinations of unit symbols is infinite. UCUM is a set of atomic symbols and rules to combine them. For example "kcal" is a combination of "k" (kilo) and "cal" (calorie).

However, there is no requirement for URIs to be enumerated, so it is in principle possible to take any UCUM unit and mangle the string in such a way that the character set restriction allowed for URIs are met. This may be an issue for our various parentheses, braces, brackets, apostrophes.

A specification for such URI scheme could be hosted here, but it could also be brought forth on other appropriate organizations with a normative reference to UCUM.

However, that all being said, I see little actionable activity that UCUM itself should do.

Comments welcome.

[timbrisc]

2013-10-22 03:29:59: gschadow@pragmaticdata.com changed priority from major to minor

[timbrisc]

2013-10-22 03:29:59: gschadow@pragmaticdata.com edited the issue description

[timbrisc]

2013-10-22 03:29:59: gschadow@pragmaticdata.com

[timbrisc]

2013-10-22 03:29:59: gschadow@pragmaticdata.com commented

---

not actionable and not required for R1.9

[timbrisc]

2021-01-15 21:03:21: mitchbre@regenstrief.org changed status from new to assigned

[timbrisc]

2021-01-15 21:03:21: mitchbre@regenstrief.org set owner to Simon Cox

[timbrisc]

2021-01-15 21:03:21: mitchbre@regenstrief.org set component to help

[timbrisc]

2021-02-16 05:34:51: simon.cox@csiro.au commented

---

I believe this ticket actually hides a genuine requirement, but I would phrase it differently.

The requirement is to have a web-identifier (URI) for every unit-of-measurement, and for the web-identifier to be based on or encapsulate the UCUM code. An associated expectation would be that the URI actually resolve to something useful, though that may be beyond the scope of UCUM as currently formulated.

There are some existing systems that provide resolvable identifiers for units, such as QUDT. However, the identifiers in the other systems have their own construction rules ands do not include UCUM syntax visibly - e.g.

http://qudt.org/vocab/unit/A-PER-MilliM2

Note that the definition denoted by this URI is annotated with UCUM codes, so it is possible to query this system using a UCUM code as an argument, though this will only get a result for the UCUM codes that are statically present in the dataset.

[timbrisc]

2021-11-02 20:25:07: mitchbre@regenstrief.org changed title from id attributes on xml and html spec to *id attributes on xml and html spec

[timbrisc]

2021-11-11 22:42:17: simon.cox@csiro.au commented

---

The service at https://units-of-measurement.org/ provides URIs in the namespace https://w3id.org/uom/ for referencing any UCUM code - e.g. https://w3id.org/uom/A-1.cg-3.s-3.T-3

Should we request that the URIs denoting the basic UCUM tokens also link to the matching point in the UCUM spec? Or to an XML fragment somewhere?

[timbrisc]

2021-11-11 22:42:17: simon.cox@csiro.au

[timbrisc]

2021-11-11 22:42:17: simon.cox@csiro.au