No statistical value and a privacy nightmare.
- 03:13, 13 July 2020 user Blackteahamburger proposed the head section of ArchMap for removal, citing:
ArchMap/List should be deleted, which means this page should also be deleted.
This discussion presumes that they both just used the wrong templates and actually intended for the entire project to be deleted (Template:Remove states that
To propose the removal of entire articles, use Template:Redirect or Template:Archive.).
Since somewhat taking over this project in early 2014 I've been well aware of the privacy implications of storing hundreds of users locations on the wiki, however it would be hard to believe that anyone was unaware of what they were doing when they added their location, it's not a trivial process. Because of that, I think it's safe to say that anyone adding their location did it willingly, and if they were uncomfortable using their exact location it would be absolutely fine to use a 'fake' location nearby. That may play into the point of the data not holding any statistical value, but with 1092 locations (at the time of writing) I don't see the argument that it contains none whatsoever.
I think that something like warnings being put in place on the articles notifying people of the privacy implications would be a good idea and believe that education about things like this is very important.
I would argue though that the pseudonymised data does hold value. It could be used before planning events, to make sure that localisation efforts are warranted or need to be increased, to enable a sense of community, and many more. Whatever the reasoning, over a thousand people must have held one of these or had their own reasons and I don't think that that should be ignored.
Note: the original deletion proposal was a month and a half ago and there has been no comments either way, however one more user has added their location.
- Privacy warnings will not help, because some infamous legislation gives users the right to withdraw their consent to share personal data. Because of the way MediaWiki stores content, the personal data of one user are not stored in just one page revision, but all subsequent revisions, which makes it extremely hard to selectively remove such information. Deleting the page is the easiest solution.
- As for the statistical value, this list has none because you don't know many Arch users did or did not record themselves in the list, which nationalities tend to record themselves more than other, or even if the users who recorded themselves 5 years ago are still using Arch.
- -- Lahwaacz (talk) 11:48, 13 July 2020 (UTC)
- Just want to preface this by saying I don't mean to sound rude and am sorry if I come across as being defensive. There's definitely a part of me just not wanting to give up on my first real software project (guyfawcus/ArchMap).
- I really don't envy your position, I've had to deal with managing GDPR myself and know how much of a nightmare it is. This is such a tricky subject and I'm not sure there are any real answers so I understand wanting to just get rid of any potential problem.
- That being said, I think the main reason I don't see it as being much of an issue is how do you know anyone put down anything near their actual location, not no mention their actual username. Because of that. wouldn't it come under the same category as any other user generated text in the wiki?
- For example: "I live in the UK". If I then asked you to remove my data, would you then just have to find whatever page I mentioned my location on and delete them, or would you have to go through every single page I'd made a contribution to?
- On the subject of statistics. I'm curious, would you say that Pkgstats is irrelevent because you don't know how many Arch users do or do not run the service, don't know where they actually come from or if someone just set up a load of VM's to skew the results? I agree that this data isn't incredibly significant but I still think it's useful for some things, even if it's just saying "1092 people have contributed to the list" 😜.
- Again, please don't take this is me being confrontational. I totally understand your position and am maybe playing a little devils advocate. This is definitely a conversation to be had.
- Of course we're open to all dicsussions, we're always happy to read counterarguments, and don't think anyone's confrontational just by objecting to a proposal.
- That said, I support removing the list for the points made above, as it would be one fewer place where to look for personal data that we may be asked to remove (there have been several cases of users who removed their entries from this list in the past).
- About pkgstats, we don't host the collected data on the wiki, so I'd argue that's not relevant in this discussion; moreover afaik the pkgstats data is stored without any reference to the client that submitted it, so it just gives some more guidance to developers to understand which packages may be more frequently installed or not, better than nothing, but of course it wouldn't be data that you could reference on a scientific paper.
- I don't necessarily support removing the main ArchMap page instead, would @Alux consider moving the locations list to a file in the GitHub repo?
- -- Kynikos (talk) 23:45, 13 July 2020 (UTC)