Equally, the page may usefully serve to inform contributors to Wikidata about legitimate concerns from other projects that have arisen out of test integrations, that there is a need to do more to address.
-- James.
On 23/09/2017 22:18, john cummings wrote:
Hi all
I'm putting aside time next week to write up an information page on Wikidata for contributors to other Wikimedia projects who want to know more about/may have concerns about reusing Wikidata on other projects. I hope this will help people having the same discussions over and over and allay many of the concerns of users from other projects.
I'm starting off with a list of common arguments for not using data from Wikidata and working my way back from there, please do take a look and brain dump
https://www.wikidata.org/wiki/User:John_Cummings/Wikidata_in_Wikimedia_proje...
Thanks very much
John
On 23 September 2017 at 14:34, James Heald j.heald@ucl.ac.uk wrote:
It's not just other wikis where cryptic template invocations can be an issue.
I sometimes think that on Wikidata itself, with templates {{P|...}} and {{Q|...}}, we could use a bot to add the label of the property or item in the default language of the page as an extra parameter to the template.
(If I remember correctly, both the P and Q templates permit the presence of such a extra, undisplayed parameter).
For one thing, this would make discussions significantly easier to interpret for anyone who is following the diffs as raw wikitext.
It also might help with people arguing at cross-purposes, basing their arguments on the label of a property or item in their own language, which is what is visible to them because it is their language labels that the {{P|...}} and {{Q|...}} templates show them -- but may be different to what the {{P|...}} and {{Q|...}} templates show to other participants who have a different mother tongue. Often both sides think their arguments are right and obvious, based on the different native labels that the P and Q templates are showing them. If there was a label added in a single language, even if displayed only in the wikitext of the page, they might sometimes realise this sooner.
So, for both of these reasons, I think there can be a case for human-meaningful "explanatory" or "identificatory" parameter-slots in templates, even if they are never displayed in actual page-output.
A bot-added Harvardesque-ref courtesy field in {{Cite_Q}} could be exactly another such example.
-- James.
On 23/09/2017 05:50, LeadSongDog wrote:
My point, Andy, was that some parameters can be required, such as CS1 requiring the parameter Title. Further, the Ref parameter can be automated, as with ref=harv.
On Sep 22, 2017, at 5:09 PM, Andy Mabbett andy@pigsonthewing.org.uk
wrote:
On 22 September 2017 at 01:45, LeadSongDog leadsong@webname.com wrote:
Not "enforcing", but it's certainly possible to show an error message for missing parameters. Many other cite templates do so.
The subject under discussion was "a legible refname"; that's not a parameter of the template and no cite templates currently warn if a refname is missing, let alone not "legible".
--- This email has been checked for viruses by AVG. http://www.avg.com