Update Metadata Customization page to suggest how to implement booleans

[luddaniel]

What this PR does / why we need it:

Explains how to implement a kind of boolean fieldtype in metadata customization
Related to #7961 (comment)

Which issue(s) this PR closes:

• Closes Custom Metadata: support boolean values #7961

Is there a release notes update needed for this change?:

Included.

Preview doc changes at https://dataverse-guide--11064.org.readthedocs.build/en/11064/admin/metadatacustomization.html (search for "boolean").

[pdurbin]

@luddaniel this looks great to me but I'm going to request a review from @jggautier and @scolapasta who think about creating custom metadata blocks more than I do.

I did just make a pull request with some suggestions on wording. Please feel free to close and ignore if you don't them! Here:

• suggested rewording Recherche-Data-Gouv/dataverse#9

[jggautier]

Hey all. I'm commenting here since @pdurbin asked me to help review.

I agree with @pdurbin's edits I see at Recherche-Data-Gouv#9.

I read the points that @luddaniel very helpfully wrote at #7961 (comment). Thanks so much!

I'd like to learn more about the third and fourth points:

• Being able to internationalize "true" / "false" or third value depending on the input title
• Being able to export the "boolean"

And maybe it would be help to ask: From those two points, is it fair to say that using "Yes" / "No" makes it more difficult to internationalize values and makes it difficult to export the "boolean"? I think I don't have any experience to know this. But I'm thinking of cases where users are already using metadata fields with dropdown menus that show "Yes" and "No".

[jggautier]

Hmm, I'll limit my review to the text changes being made and not try to understand why we're recommending this, since others worked on this earlier.

I'll review the latest text changes today 👍🏽

[pdurbin]

@jggautier sounds good. Thanks!

[jggautier]

@luddaniel There's just a missing "be" between may and used. Could you add that?

[pdurbin]

@jggautier whoops, that was my bad, part of my rewrite. I just fixed it in 649e2af.

[jggautier]

Cool thanks. That was my only suggestion for the changes in the documentation. I'll remove myself from the Assignment list on this PR.

[jggautier]

Cool thanks. That was my only suggestion for the changes in the documentation. I'll remove myself from the Assignment list on this PR.

[pdurbin]

Looks good to me! Thanks, @luddaniel! Approved.

[luddaniel]

Hi @jggautier

I'd like to learn more about the third and fourth points:

• Being able to internationalize "true" / "false" or third value depending on the input title
• Being able to export the "boolean"

And maybe it would be help to ask: From those two points, is it fair to say that using "Yes" / "No" makes it more difficult to internationalize values and makes it difficult to export the "boolean"? I think I don't have any experience to know this. But I'm thinking of cases where users are already using metadata fields with dropdown menus that show "Yes" and "No".

In fact, exported metadata will not use any properties language, so exported values will match tsv values : True/False or Yes/No but Yes/No is likely unusable by a machine/code interpretor.
And it's not complicated to add internationalized value (GUI) .properties file to change True/False to Yes/No Oui/Non.

For users who are already using Yes/No in TSV : the damage is done haha! Nah it's just a recommendation, let's say to clarify our position on the boolean field type : use controlled values. I didn't tested but maybe you can use alternative value in TSV to change/add Yes/No to True/False.