Timeline for Questions about the new minimum age for diamond moderators
Current License: CC BY-SA 4.0
16 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Dec 5, 2020 at 21:30 | comment | added | Robert Columbia | Good point about ability to contract in various jurisdictions. This can get quite complex. I used to know an adult who had major developmental disabilities. He was under guardianship (could not sign a contract without the guardian's permission). He could also barely read anything more than his own name and certainly couldn't be expected to read and understand the terms of service of the websites he visited. Yet, despite all of those things, he loved to go online and watch videos and play games, and nobody seemed to make a big fuss over calling the guardian every time he wanted to click. | |
| Dec 3, 2020 at 23:13 | comment | added | Shog9 | It would, but I wanna emphasize that's not the only problem, @gnat. When I had access to this stuff, I built tools for identifying connections between accounts based on IPs without ever actually looking at those IPs. Just knowing (roughly!) when someone appeared from a given IP is enough to figure out an awful lot about them, if you have unfettered access to the data. Heck... As you might be aware, others have done a pretty decent job using only public data! It is hard to effectively obfuscate this stuff - arguably the effective utility of it would be entirely destroyed long before then. | |
| Dec 3, 2020 at 23:10 | comment | added | gnat | I see, thanks. This is exactly what I wanted to learn: looks like straightforward simplifying / obscuring of stuff that currently exposes PII would make rather substantial damage to moderation efficiency | |
| Dec 3, 2020 at 23:06 | comment | added | Mad Scientist | @gnat the CM team couldn't handle these issues alone right now, and I don't think they could with the planned expansion either. The amount of issues that require PII access varies a lot across sites. On smaller sites it can be almost irrelevant most of the time, on larger sites it can take a lot of work. Right now, you can't handle vote fraud, sock puppets or persistent trolls without these specific tools that reveal PII. | |
| Dec 3, 2020 at 23:06 | comment | added | Shog9 | That varies quite a bit, @gnat. Most mods on most sites don't use this at all. Many use it only occasionally, when under attack or faced with an unusual problem. Some use it daily, because they're on a site that's always under attack in one sense or another and they've made it their specialty. There's 4 CMs, and I doubt any of them are sitting on their hands; realistically, transferring work from mods -> CMs means work that takes longer or doesn't get done at all. | |
| Dec 3, 2020 at 23:01 | comment | added | gnat | @Shog9 I understand that. My question is, how much of this knowledge diamond mods use directly and how much loss would it be for them to be deprived of being able to immediately use it (with most likely increase of CMs load to handle this part based on mod escalations) | |
| Dec 3, 2020 at 23:00 | comment | added | Shog9 | The way to do this isn't to try to replace PII. It's to build tools for the (by now fairly well-understood) moderation tasks that don't need to expose it at all except in exceptional circumstances. BIG risk difference if mods on SO have to look at PII once or twice a day vs. hundreds of times a day. Right now, the tools are crude - barely better than raw log access, and in some cases worse. Stuff takes too much time and effort, is too error-prone, and creates too much risk for abuse. If SO, Inc ever gets serious about this, tool-building is where they'll start @gnat. | |
| Dec 3, 2020 at 22:57 | comment | added | Shog9 | And honestly... IPs are one of the weaker forms of PII involved, because generally you only get personal information from IPs by correlating them with other things - which means a hash or UID is not necessarily any less PII, if it can still be correlated with other things! | |
| Dec 3, 2020 at 22:55 | comment | added | Shog9 | The value of seeing IPs isn't just in being able to match up two users from the same IP, @gnat. It's also in being able to say, "this user connected from Texas Comcast IPs for 5 years, then today from a network in Australia at which point they started trolling." It's realizing that all the spam today is coming from AWS IPs, or a series of proxies associated with a service that was used to spam last week. It's recognizing that two cross-voting accounts are both from Madison, WI networks, one cable one mobile... | |
| Dec 3, 2020 at 22:51 | comment | added | gnat | @Shog9 I wonder how technically complicated would be to change UI to show moderators system-generated UIDs instead of user emails and post IPs. How much impact would it make on moderation efficiency. | |
| Dec 2, 2020 at 23:09 | comment | added | Mad Scientist | @Shog9 that is what my "lawyers covering their asses" line is meant to convey. This is one checkbox you can tick off that fulfills a formal requirements, it doesn't improve the situation and in my non-lawyer opinion probably doesn't even decrease liability for SE all that much. | |
| Dec 2, 2020 at 23:05 | comment | added | Shog9 | The tools necessary are a serious undertaking. Discussion regarding them - along with some prototyping and spec work - dates back half a decade; the resources to implement them were not available for the past five years, and likely will not be soon. This is nothing but a fig-leaf, it means nothing in terms of protection for anyone, beyond creating the illusion that such protection might exist. Do not waste your energy in assuming further. | |
| Dec 2, 2020 at 23:03 | history | edited | Mad Scientist | CC BY-SA 4.0 |
added 7 characters in body
|
| Dec 2, 2020 at 22:31 | history | edited | Mad Scientist | CC BY-SA 4.0 |
added 8 characters in body
|
| Dec 2, 2020 at 22:28 | comment | added | Sonic the Anonymous Hedgehog | Wikipedia does the same thing you pointed out in your last paragraph: normal admins only have access to wiki moderation tools, and access to PII requires a separate CheckUser bit that is heavily regulated (requires ID verification). However, they were only able to set it up that way because their architecture was developed specifically to allow for this (having user rights be separate from user groups). SE isn't developed this way (moderator rights are strictly tied to the moderator user group), so they had to go this way. (I know you know this; pointing it out for future readers.) | |
| Dec 2, 2020 at 22:10 | history | answered | Mad Scientist | CC BY-SA 4.0 |