search

cessda / cessda.metadata.profiles

Contains DDI Profiles that are used by the CESSDA Metadata Validator Tool
Other
0 stars 0 forks source link

Mandatory language on Person name feels unnecessary #18

Closed cessda-bitbucket-importer closed 3 years ago

cessda-bitbucket-importer commented 3 years ago

Original report on BitBucket by Alexander Mühlbauer.

Feedback by Johan Marberg (SND)

Required: Mandatory if 'AuthEnty' element is present ElementType: Attribute Usage: Language of Person or Organisation name. ISO 639-1 codes must be used. Mandatory language on Person name feels unnecessary.

‌

cessda-bitbucket-importer commented 3 years ago

Original comment by Taina Jääskeläinen.

AuthEnty may be organisation in which case it may have different names in different languages. Affiliation attribute may also refer to the person’s organisation which again may be multilingual information.

In DDI 3.2 there are separate elements for organisations and persons but in DDI 2.5 both are put in AuthEnty element.

cessda-bitbucket-importer commented 3 years ago

Original comment by Taina Jääskeläinen.

@darrenbell2 Rephrase Usage information. Something to the effect: Language of the organisation name, which may be AuthEnty or the person’s affiliation.

‌

cessda-bitbucket-importer commented 3 years ago

Original comment by Darren Bell (GitHub: darrenbell2).

From: Bell, Darren S
Sent: 26 November 2020 18:18
To: email address removed' <[email address removed](mailto:email address removed)>
Subject: RE: (CESSDA Metadata Office) Review of DDI profiles for CDC

 Hi Benjamin. Thanks for the feedback.

  1. Regarding usage of AuthEnty\@‌xml:lang, this has been raised by others as well, and logged at

https://github.com/cessda/cessda.metadata.profiles/issues/18
The usage of @‌xml:lang in this context will be changed to “Recommended” in the next version of the profile (v1.0.3 to be released in the next two weeks) and the documentation updated to reflect the fact that you are strongly encouraged to provide a language if it is an Organization name, but not for a Person.

  1. Regarding the use of@vocab, there have many, many discussions.  The latest profile (

see v1.0.2) recommends the use of “CESSDA Topic Classification” and “ELSST” for and respectively but usage of other vocabularies is permitted.  However, those non-standard topic classifications and keywords will not be reflected in the CESSDA Data Catalogue.

 

Everyone, to help provide clarity when feeding back on the profiles (and feedback is very much appreciated), can you please ensure that you mention the version number of the profile you are reviewing. 
This appears in the annotations at the top of the profile document e.g. ‘Version: v1.0.2’

 Best wishes, Darren

‌

From: Benjamin Peuch (Basecamp) [[mailto:email address removed](mailto:email address removed)]
Sent: 26 November 2020 15:21
To: Bell, Darren S <[email address removed](mailto:email address removed)>
Subject: Re: (CESSDA Metadata Office) Review of DDI profiles for CDC

 

Hello everybody,

I have a few questions and suggestions.

@‌xml:lang

Like Johan , I'm not
sure why the xml:lang attribute should be mandatory for a person's name in the

element? It certainly makes sense to indicate the name of an organization because names in other languages than English are to be expected. For example, the names of several Belgian universities should not be translated because they mean the same thing: * Université libre de Bruxelles \(FR\) =/= Vrije Universiteit Brussel \(NL\) — Both literally mean "Free University of Brussels," hence the English name is to be avoided. * Université catholique de Louvain \(FR\) =/= Katholieke Universiteit Leuven \(NL\) — Likewise, "Catholic University of Louvain/Leuven" is ambiguous. Because fields such as are likely to receive names of either people or organizations, maybe the xml:lang attribute should be optional here? **@‌vocab** This attribute is mandatory for all and elements. Does this mean that CESSDA has taken an official stance against folksonomies / custom keywords? I am not saying this is necessarily a bad thing, and I know controlled vocabularies are extremely useful for \(meta\)data management and data discovery. For instance, we recommend the use of the _CESSDA Topic Classification_ by prepopulating the "Topic Classification" field in our Dataverse installation thanks to dataset templates. But I wonder if there have been discussions about this that led to an overarching policy decision, considering this seems to be a steadfast rule ‌
cessda-bitbucket-importer commented 3 years ago

Original comment by Darren Bell (GitHub: darrenbell2).

Changed @‌xml:lang to be Recommended for v1.0.3 of the 2.5 and 1.2.2 profile - see updated content below:

Required: Recommended ElementType: Attribute Usage: Language of the Organisation name, which may be the contents of the AuthEnty element or the @‌affiliation attribute of AuthEnty specifies a person. ISO 639-1 codes codes are strongly encouraged to be used. \]\]>
cessda-bitbucket-importer commented 3 years ago

Original comment by Darren Bell (GitHub: darrenbell2).

Change implemented in drafted in v1.0.3 Needs final approval before closure of ticket and publishing v1.0.3

cessda-bitbucket-importer commented 3 years ago

Original comment by Taina Jääskeläinen.

…or the @‌affiliation attribute of AuthEnty specifies a person.

Maybe rephrase or the @‌affiliation attribute when AuthEnty is a person.