This page is not created by, affiliated with, or supported by Slack Technologies, Inc.
- # aleph (1)
- # announcements (2)
- # aws (3)
- # bangalore-clj (1)
- # beginners (136)
- # boot (3)
- # calva (89)
- # cider (44)
- # cljdoc (1)
- # cljs-dev (31)
- # clojure (101)
- # clojure-europe (3)
- # clojure-italy (52)
- # clojure-nl (7)
- # clojure-spec (12)
- # clojure-uk (34)
- # clojurescript (62)
- # community-development (46)
- # cursive (25)
- # datomic (6)
- # duct (26)
- # events (6)
- # figwheel-main (4)
- # fulcro (25)
- # graphql (2)
- # jackdaw (4)
- # jukebox (3)
- # kaocha (57)
- # leiningen (31)
- # off-topic (3)
- # onyx (4)
- # other-languages (22)
- # pathom (18)
- # re-frame (9)
- # reitit (3)
- # shadow-cljs (60)
- # spacemacs (9)
- # speculative (22)
- # sql (39)
- # tools-deps (45)
- # vim (37)
Some very annoying gif (animation, not content) at #off-topic from an account which is now deactivated. Can we remove at least the inline preview?
Removed. I didn't know the spammer had posted there too. I was mostly busy last night removing all the porn spam that the other rogue user posted everywhere :face_with_symbols_on_mouth:
They weren't leveraging the Slack exploit so this was just drive-by spam/porn because we have an auto-inviter. The Admin Team are discussing possible options to curtail future spammers but we don't want to just close off the community to new members. We're looking at email filtering (the spammers, conveniently, mostly use obvious domains when they sign up -- but they might soon get wise to that). We're also contemplating having the "invite" page just send candidates to an admin mailing list and we would manually approve them (a fair bit of work, given the number of new members most days). Or we could add some simple, Clojure-related questions to the sign up page (should deter spammers -- and some of the one-off random promoters we get too -- but, again, easy enough to discover the answers out there and bypass that sort of security).
Slack is designed for closed, trusted teams so it really doesn't have any sort of pre-moderation (requiring Admins approve your first message would also likely stop this -- but could be very frustrating for genuine new Clojurians).
I know it’s a really tough sell, but we could try encouraging people to move over to https://clojurians.zulipchat.com Zulip is designed to accommodate open chat rooms like what we are trying to use slack for. Any exploits that may appear in zulip like the one we saw the other day should be taken more seriously by the zulip team.
Folks are welcome to try Zulip. We've mentioned it a few times in several channels. Getting 15k people to move en mass is... challenging tho'... And a lot of people like Slack (sometimes because they use it at work). I suspect any large open community is going to attract spammers.
I think the only way to get people on zulip is to first have a weekly reminder, on all channels, to please migrate to zulip. And eventually move to a once daily notification.
from the Clojure survey I can give you some concrete data about what people are actually using
slack is the highest with 63%, reddit 54%, mailing list 37% (big drop), ClojureVerse 22%, Zulip 5%
the idea of replacing one web browser chat with another web browser chat, where I still have to have the first web browser chat around for work is very uninspiring
For a lot of folks who use Slack at work, the idea of running yet another chat client is... unappealing...
I'm on slack at work too. But that's quite a pill to swallow for other community members that aren't, that some of us are
You can't make everyone happy 🙂 That's why there are so many little chat communities already. Discord, Matrix, Zulip, Braid... IRC 🙂
Could this be automated. Something that looks for x amount of links or gifs in a certain frequency then blocks that user and\or alerts an admin?
The Admins would be happy for community volunteers to write a bot that helped auto-moderate stuff 🙂
What about a bot that worked something like this: Spammer logs in, begins posting crap. Regular users add a reaction of and/or :abuse:, bot recognizes the reactions, checks that reactions are coming from users who have been members longer than X weeks (to prevent DOS spamming), and blocks spammer, with alerts to Admins for reviewing.
RE Slack vs Zulip: hopefully the searchable slack-archive (https://clojurians.zulipchat.com/#narrow/stream/180378-slack-archive) will help attract users looking for an alternative
That is a definitely an idea. I take it there would be a threshold on # of those reactions before the bot kicks in?
Definitely would cover the use case for when someone spams a bunch of unwanted links and one random one in a channel. May want to combine that with a count of how many links a user sent between all channels within the past 30 min or something though so it can’t easily be used against innocent users.
Yeah, min number of people marking as spam/abuse, and/or a whitelist of trusted users who can trigger the bot by adding a reaction
If in the event no trusted users are in the list, should it alert an admin if a message exceeds that minimum?
This may be out there, but could it be created as a subscription product for community slacks and profits invested back into Clojure?
Say for example, threshold is 5 reactions, but each reaction from a trusted user counts as 3.
Feel free to bikeshed the implementation details in issues on a GitHub repo for this bot if you folks want to build it 🙂
It doesn’t need to be under my ownership, someone can start a diff one and I’ll delete it or I can transfer ownership.
> I think the only way to get people on zulip is to first have a weekly reminder, on all channels Is there an option to set this to remind people of the searchable archive on Zulip?
https://github.com/eccentric-j/cjl-slack-auto-mod/issues/1 added more detail and structure to the proposal.