The Signpost

Discussion report

The deprecation of Persondata; RfA – A broken process; Complaints from users on Swedish Wikipedia

Contribute  —  
Share this
By Jonatan Svensson Glad

The deprecation of Persondata

Wikidata's logotype
Wikidata's logotype

Since the dawn of Wikipedia, or at least since 22 December 2005, the template named Persondata has existed. At first, it was an early attempt to incorporate structured metadata. Later it served to provide automatic extraction of key information about people, such as birthnames and dates. However, since the creation and growth of Wikidata the need for this template has become non-existent, since all the functions which it provides, and more, are now offered by Wikidata.

Wikidata allows users to add statements to Wikidata items (which are connected to Wikipedia articles) such as "birthdays", "occupations", "family members" etc. These statements rely on other Wikidata items and are therefore completely internationalized. There is no need to translate a person's occupation into multiple languages, since they are connected already to articles on other language Wikipedias. The template Persondata, however, is not internationalized and only provides services to the English Wikipedia. Therefore, having all this information on Wikidata makes it accessible in multiple languages and it can be used in articles on different wikis in the remote future.

All |SHORT DESCRIPTION=s had been copied over to Wikidata by the bot PLbot. With that in mind, plus that all other information in the Persondata template had been exported to statements on Wikidata, there is no real need for the template any more. This is at least what Msmarmalade was trying to communicate to the community with a proposal to deprecate the template.

The template had been discussed over many months and now an RfC had been created to settle once and for all the question of whether or not anyone was actually using it, or will use it in the future.

During this RfC many viewpoints and opinions were brought up and discussed.

The big problem with WikiData is that its "impossible" to edit and watch changes to articles (from enwiki). […] [H]ow do we suppose editors to figure out how to change any wrong data?

To this, users responded that it is possible to see all related Wikidata changes on your watchlist if you enable that gadget in your preferences.

Another issue was if usage of Persondata should be replaced with Wikidata, and to that Pigsonthewing responded:

This would only be true if we were talking about using [use of content from Wikidata on this Wikipedia] to replace [use of content from Persondata on this Wikipedia]. Since the set of articles in which the latter applies appears to be zero (do feel free to correct me), it's irrelevant to the discussion at hand.

There were also questions over how reliable statements moved by bots are, since Wikidata does not support free-text in their data fields, some dates become malformated. To this Technical 13 noted:

Wikipedia:Village pump (technical)/Archive 136#Wikidata date errors indicates that persondata is likely still more reliable than Wikidata. So, still going to have to go with "Wikidata is not stable and ready yet".

While this link to the technical villlage pump did not have anything to do with humans, it raised an interesting point that all date formats can not be handled on Wikidata.

Users noted that since these data can not be used in infoboxes due to their potential inaccuracy, Persondata should not be deprecated just yet, but to this Alakzi had a witty while still accurate point:

The reliability (accuracy) of the data and reliability (availability) of the service are two completely different things. (For God's sake, why are people using Module:Wikidata in mainspace? Whoever's said that it's ready for primetime?) Persondata cannot even be used in this manner, so what point are you trying to make?

— Alakzi

Persondata was not in use anywhere and was therefore deprecated with a 32-to-8 !vote. After this consensus had been reached, Magioladitis did not waste any time and requested for approval of his bot Yobot's 24th task: to "Remove {{Persondata}}". This has been met with some resistance. GoingBatty asked two questions; if this should wait until AWB (which Magioladitis is a developer of) had been programmed to stop adding the template, and if the policy WP:COSMETICBOT would apply, since it was an edit which wouldn't be visible or noticeable to the readers. The bot Rjwilmsi had once received permission to add Persondata as its only task and therefore there is precedence of fixing Persondata, despite the policy in question.

We already have consensus for the removal of Persondata. If the addition of Persondata by automated tools hasn't been a breach of COSMETICBOT, then neither should be its removal.

The user Dirtlawyer1 was also a bit sceptical to the removal of the template, despite the consensus, stating:

[…] This immediate removal without transfer of new input persondata to Wikidata violates the conditions upon which the RfC was approved. […]

If a persondata template with a short description has been added after PLbot transferred all the current once at the time, this might cause some problems with losing some information added by someone. However:

The outcome of the RfC is "deprecate and remove", not "deprecate and remove with caveats".

Author's note: The request for permission is currently being discussed while writing this report, hence no conclusion can be written, since it has yet to happen.

RfA – A broken process

After years of granting fewer and fewer users adminship and desysoping more and more of our current admins, you can naturally calculate that we are starting to reach a shortage of admins. At least that's what the current discussion on the Administrators' noticeboard is saying. As a result of this, backlogs on noticeboards such as Wikipedia:Administrator intervention against vandalism are bound to happen from time to time.

Users are reporting to feel discouraged to run for adminship, for two major reasons. One being that RfAs are like hell, with the scrutiny so intense you lose interest and want to go and hide somewhere. The other reason being that it is a broken process:

I know exactly how to become an admin. Stop getting involved in discussions at AN, ANI, RSN, etc., stop mediating at DRN, pick a poor-quality, uncontroversial article that nobody seems to be editing or watching and create high-quality content, withdrawing and moving on if anyone disagrees with me in any way, and repeat that pattern for at least a year. In other words, avoid anything that in any way resembles what an administrator is asked to do. Again, I do want to help but the price is too high.

— Guy Macon

In briefs

Style and naming

Policies and guidelines

WikiProjects and collaborations

+ Add a comment

Discuss this story

These comments are automatically transcluded from this article's talk page. To follow comments, add the page to your watchlist. If your comment has not appeared here, you can try purging the cache.

NPA?

"The users does not seem to be willing to come up with any suggestions or solutions other than to disrupt and make a fuss." That comes uncomfortably close to a personal attack. One does not always need to have a "solution" for a complaint to be legitimate.  --Lambiam 10:51, 4 June 2015 (UTC)[reply]

Indeed. Fel free to formulate how you wish to still be true to the situation. (tJosve05a (c) 12:20, 4 June 2015 (UTC)[reply]

Unclear

Not sure what this means "American Wikimedias make up around 80% of all money the Foundation makes" Does this mean donations from United States make up 80% of funding? Also it is best to use Canadians, people from the United States and Mexicans. While we from Canada and Mexico are technically Americans many do not use the term in that way. People often use it to mean just people from the United States. Doc James (talk · contribs · email) 19:34, 6 June 2015 (UTC)[reply]

Per here 20% of editors from US [1] unfortuantely no ref or when this was.
Pere here most donations from US in 2012.
In 2014 things were lumped together by continent [2] and yes America (all three) were about 60% Doc James (talk · contribs · email) 19:45, 6 June 2015 (UTC)[reply]
May be a bad quote on my part of this pipermail. However, this is a report of discussion, and I just reported what others stated. (tJosve05a (c) 23:16, 6 June 2015 (UTC)[reply]
The US is a relatively wealthy part of the world, and it is where the servers are hosted, I think if you go back a few years you will find there was a time when 80% of our money was raised in the US. However that day has long passed, especially if you look at all movement revenue including money raised directly by chapters such as Germany. ϢereSpielChequers 06:02, 8 June 2015 (UTC)[reply]

RfA

I believe these kinds of argument are wearing rather thin nowadays. Whether we at WP:RFA2011 actually launched any formal proposals for change or not it certainly sent the right message to those who were determined to disrupt RfA and/or turn it into a drama fest; we're also more active now at telling the trolls where to go with their votes. I think it's more a case that some would-be nominators are afraid of losing face if their nominee fails to get the mop. However, with very few exceptions RfA does what it says on the tin.

Most people commenting on the state of RfA fail to take into account the vast amount of data that was extracted by the participants at WP:RFA2011 which even if updated would still return the same factual conclusions today. The research generally put the blame squarely at the feet of those who were determined to disrupt the election process and give it a bad name, while the actual ‘facts’ revealed that the vast majority of voters are one-off, drive-by, fans, and vengeance seekers, with a few trolls and banned socks thrown in for good measure.

WT:RFA is the backbone of discussion on RfA, candidates, and adminship in general, and people would be well advised to at least watch that page more regularly before commenting anywhere else. However, it's also interesting to note - something that WereSpielChequers, who maintains a lot of the stats, might not yet have noticed - is that participation in discussions at WT:RfA has dropped in direct proportion year on year to the drop in promotions. I don’t personally know how to interpret that, but it is clear evidence that RfA being a ‘horrible and broken process’ does not quite ring entirely true today. --Kudpung กุดผึ้ง (talk) 01:02, 7 June 2015 (UTC)[reply]

  • As usual, Guy Macon wields his ginormous hammer of truth and hits the kotowaza nail of independent thought on the head, just as the Wikipedia petitfoggers who worship bureaucracy as their evil machine god do each time they appeal to deru kugi wa utareru in the form of an RfA. Its demise cannot come soon enough. Anyone who wants to be an admin simply doesn't get it. Viriditas (talk) 02:05, 8 June 2015 (UTC)[reply]
Well, Kudpung กุดผึ้ง, there was a pretty good turnout for NeilN's RfA. Your conclusions are interesting to me about the impact of who comes to vote in RfAs...I typically see very familiar names on the Support/Oppose/Neutral lists so I was unaware that there were a lot of drive-by editors participating. I'm not sure how quickly new editors would even come across an RfA as they are exploring Wikipedia. Liz Read! Talk! 01:48, 8 June 2015 (UTC)[reply]
I had noticed that WT:RFA had gone quiet, not sure why that happened, but when people try to change or improve RFA they don't necessarily do so there. I'd also noticed that while total participation at RFA has fallen, it hasn't fallen as fast as the number of RFAs, hence the increase in activity per individual RFAs, over a hundred !votes and a stack of questions now being the norm. Hopefully this means if we could reform RFA to the point where a dozen new admins were being appointed per month we would still have sufficient participation for the RFAs. As for how new editors come across RFA, in my case I noticed a list of current ones on someone's userpage. ϢereSpielChequers 11:51, 8 June 2015 (UTC)[reply]
  • @Liz:, WereSpielChequersTurnout at RfA has never been so good and a 100+ support is nowadays nothing to jump up and down about and hasn't been since my RfA nearly 5 years ago. However, in case anyone missed the conotation, if one lays the two bar graphs of the number of promotions and the activity at WT:RfA on top of each other, they are near idetical. Which leads to the logical assumption that interest in adminship in general has waned. That is different from editors of the right calbre being scared of what is in fact no longer so much a week of hell. --Kudpung กุดผึ้ง (talk) 18:03, 8 June 2015 (UTC)[reply]
Well, I think that RfAs that might have been highly contested are now being withdrawn at an earlier stage (like the same day they were posted) but I don't have any data to support that impression. I think the perception of the RfA process is that it will be a grueling 7 day experience where people look for skeletons in your closet and inquire about unlikely hypothetical situations and if an editor sees signs that it won't be an easy pass, they are much more likely to end the process early rather than stick it out for the entire period. And who can blame them? Liz Read! Talk! 18:31, 8 June 2015 (UTC)[reply]
Liz, I think you'll find that the vast majority of early closes are from candidates who didn't stand the slightest chance of ever getting through the first few minutes of transclusion and were either attempts to test our patience/vigilance etc or from other misguided souls who just couldn't be plain bothered to read one single advice page or even take notice of the last-minute 'in-your-face' warning banner about wasting their (and our) time. Such RfA are also a lot rarer than they used to be. So the experiment still stands: Anyone who is really interested should at least take the time to investigate by reading up on a lot of past failed and successful RfA, and note that interest in general in RfA/Adminship has waned significantly, otherwise what is reported is just conjecture.--Kudpung กุดผึ้ง (talk) 01:14, 9 June 2015 (UTC)[reply]
Indeed once again we have people needlessly whining about RfA being too hard when nine times out of ten those who are complaining that it is too hard didn't deserve the position anyways. So, to make themselves look better, they shout "Bureaucracy!" at everyone, as if we were the modern-day U.S. government.
Forget your own personal views and which administrators you like and do not like for a moment: what if we got more admins (not naming names) that tried to push their own ideological views onto others because they realise that Wikipedia is a platform that is very influential these days. So, with ulterior motives in hand, they go to RfA. With the current process, they don't get very far, because people smell fishiness early on.
If we loosen up the process, all that will happen is that more people who are not qualified to become administrators will get through. And, yes, the snowball may well head down a slippery slope into internal corruption (read: in such a case real, actual, bona fide bureaucracy would begin to appear). Tharthan (talk) 01:12, 16 June 2015 (UTC)[reply]

Persondata

  • The wholesale removal of {{Person data}} would destroy carefully hand-crafted information about alternative names, and short descriptions which have been added since PLbot did the transfer. We have had this sort of vandalistic implementation of consensus before, where a more careful phased approach would preserve the value of the deprecated component while still meeting the requirement to remove it. All the best: Rich Farmbrough, 12:44, 7 June 2015 (UTC).[reply]
    • Richard has been around here long enough to know that it is Wikipedia policy that "any good-faith effort to improve the encyclopedia is not vandalism... Mislabelling good-faith edits as vandalism can be considered harmful." The agreed removal of persondata has been discussed at length, both in the RfC and on Wikidata. Sadly, the remaining data held in {{Persondata}} and which has not already been transferred to Wikidata is not in a form which allows for its automated transfer (these problems with persondata have been known for some time). It's all well and good to say that the data is there, but if no-one is using it, it is unsuitable for transfer and it is impractical to transfer it manually, what's the point of keeping it? It just clutters article code and confuses new editors. Besides, it will still be in article histories for anyone needing to find it. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 12:14, 8 June 2015 (UTC)[reply]

Wikidata gadget

As usual, links to village pump discussions break very quickly. So where is this Wikidata gadget that was referred to in that discussion? I can't find it.--greenrd (talk) 15:21, 7 June 2015 (UTC)[reply]

It's not on the Gadgets tab of Preferences, where I had expected it. It's two separate settings, one on the Recent changes tab ("Show Wikidata edits by default in recent changes and watchlist (does not work yet with enhanced changes)") and one on the Watchlist tab ("Show Wikidata edits in your watchlist"). Stevie is the man! TalkWork 15:15, 9 June 2015 (UTC)[reply]



       

The Signpost · written by many · served by Sinepost V0.9 · 🄯 CC-BY-SA 4.0