Talk:Q76525

date of death

@Pichpich: Thanks for noticing the bug in my import :) but I checked the original dataset, and it’s not my bug – the Bundestag dataset actually says that he died on that day. Since the reference is correct (even though the data obviously isn’t), I propose it makes more sense to keep it as a deprecated statement than to completely remove it. Do you agree? --Lucas Werkmeister (talk) 15:30, 8 May 2018 (UTC)

I strongly disagree. We know the source sucks and very few people will notice that the statement is deprecated. I wouldn't mind a deprecated statement that says his birth day is on the 5th of May when more reliable sources say he was born on the 6th, but the mistake here is so blatant it serves little purpose to keep the bad info. If I'm not mistaken, deprecated statements do show up in searches with tools like Petscan and queries at Wikidata Query so keeping the really poor info is not doing anybody any good. Pichpich (talk) 15:47, 8 May 2018 (UTC)

@Pichpich: deprecated statements do not show up in WDQS queries unless the query is specifically phrased to include them. (Try it out: wd:Q76525 wdt:P570 ?dateOfDeath. does not yield any results.) If any other other tool includes deprecated statements indiscriminately, then IMHO that’s a serious problem with the tool, and one which is going to affect a lot more items than this one. This case is, in my opinion, a prime example for why the deprecated rank exists and is useful: that’s also why the item error in referenced source or sources (Q29998666) exists.

More practically speaking, if the statement is removed, anyone who processes the Bundestag dump (including me – I’m not done yet) might add the same statement again, with normal rank. On the other hand, most import tools will not add another statement if a statement with the same value already exists, so if we leave the deprecated statement there it should stay that way regardless of who else imports it. --Lucas Werkmeister (talk) 18:26, 8 May 2018 (UTC)

@Masegand: did you see this discussion before you removed the statement again? I don’t want to get into an edit war, but so far I remain convinced that the statement should be kept. --Lucas Werkmeister (talk) 11:01, 11 May 2018 (UTC)

No i did not see it. I have repeatedly deleted such statements where dob and dod got confused as it looked like a trivial import bug instead of a really existing ambiguity. There is also no direct weblink to the referenced content of Q51850225 on this import. Of course this might happen again if the bad data is still there and no check for P569==P570. --Masegand (talk) 13:54, 11 May 2018 (UTC)

@Masegand: basic data about the members of the Bundestag (Q51850225) has both official website (P856) and full work available at URL (P953), so you can check for yourself that the statement is present in the import if you want to. --Lucas Werkmeister (talk) 10:42, 14 May 2018 (UTC)