1. 10

Every so often I check the Invitation Queue. But really there is no way to verify that the person is the one they claim.

I myself got invited here by posting a link to my then HN profile. Nowadays I see a LOT of links to random github profiles/blogs.

I could send these people an invite, but really, if i were to spam lobste.rs I’d search up random blogs and register profiles with a similar name than the ones uses in random tech blogs.

Essentially there is no way to verify that the people submitting invitation requests are really the people behind the github profiles/blogs.

Should there maybe be a way to at least confirm that the blog/github/whatever you submit in your invitation request is under your control by giving people a random string to post there and then link to that?

As it is right now there is no way to check if the URL they linked is under their control, so I stopped inviting people I don’t personally know.

What are the current guidelines to vet invitees?

Invite everyone with a link to something that seems programming related and whose usernames seem to match the URL they submitted? I’d be fine with that, I’m just not sure how much it helps against spam. I’d appreciate some official guidelines here.

  1.  

  2. 12

    if i were to spam lobste.rs I’d search up random blogs and register profiles with a similar name than the ones uses in random tech blogs.

    I don’t think anyone’s done this in all the time the invite queue has been around, but such blatant spam would be really easy to spot and those accounts would quickly be banned. The bigger problem is legitimate accounts posting corporate blog spam and marketing for their startups, which account verification would do nothing to prevent.

    1. [Comment removed by author]

      1. [Comment removed by author]

        1. 6

          I think is OK to err on the side of not inviting. I like to think I’m like most people. I rarely submit. I’m mostly here to read.

          I guess the question is what are we trying to optimize for? I suspectour answers will fall in place once we answer that question.

          1. 5

            “I guess the question is what are we trying to optimize for? “

            This is what I use when describing the site:

            https://lobste.rs/s/oackyq/lobsters_community_standards/comments/sybvqw#c_sybvqw

            It got a lot of smart people interested who would’ve done good discussions. They didn’t join when I actually sent the invitation and I didn’t bother them about why. My guess has always been front page was full of fluff/noise during that time that contradicted what’s in the link above. It stabilized and we got back to normal. So, the Is and Isn’t parts of that comment still seem like a good standard to use describing it to select others.

      2. 3

        @lawl In general it seems to work ok. If you look on the user tree there are not really that many deleted / inactive accounts so there seems to not be a problem that needs solving.

        I’ve been looking at the invitation queue a bit recently and inviting people I like the look of. Plenty of people post URL’s that are not particularly impressive - I don’t think it’s being abused in any significant way.

        1. 2

          I will admit that I don’t usually recognize the people I invite, but I also don’t invite people very often. I might be violating the guidelines a tiny bit. I usually select for interesting writing at whichever link is posted. Just a github profile is rarely instructive in that regard.

          1. 2
             I myself got invited here by posting a link to my then HN profile. Nowadays I see a LOT of links to random github profiles/blogs.
            

            Thats how I got invited years ago, not seeing the issue with linking to github to be honest.

            1. 1

              Me either–frankly, I’d like to see more people who aren’t active on HN.

              (Personally I question the value of the invitation system over open signups. As our userbase has grown, it’s not clear to me that the invitation system has prevented any of the problems that accompany similar growth on open sites. But it’s not my site, so….)

              1. 1

                As our userbase has grown, it’s not clear to me that the invitation system has prevented any of the problems that accompany similar growth on open sites.

                It assumes people will be careful about who they invite. I was looking into their technical skills, quality of discussions, and filtering out anyone who gets in flamewars a lot over trivial stuff. Such people should have interesting comments, esp if I get 2-4 each in diverse sub-fields. Many of the invitee’s during the mass invites were clearly not assessed to that degree. That’s why the front page looked like almost pure noise. It might have been a problem with incentives as some people were competing over number of new members to “win” acclaim. I’m not sure, though, since I didn’t look to see which user trees the noisy ones came out of.

                The original concept was good, though. It’s basically F2F model. Only as strong as the judgment of its weakest members if design is share-all (eg forum).

                1. 3

                  Only issue with friend to friend is if you’re in the out group. If I had to go through the hoops described above I’d just have ignored this site entirely.

                  I don’t think thats conducive to good discussion, just promotes insular friend groups in the end. Though I acknowledge it normally works fine.

            2. 2

              Should there maybe be a way to at least confirm that the blog/github/whatever you submit in your invitation request is under your control by giving people a random string to post there and then link to that?

              It reminds me of Keybase that attempts to solve such identification problem.

              1. 1

                FWIW, I got my invite by simply proving my more-than-two-indentities keybase account to someone.

              2. 1

                I think that the purpose of the invitation is not to (for the lack of a better word) be “paranoid” about inviting people, but to not have a lot of people spamming and trolling, as well as keeping the overall quality up.

                For this the system seems to work. I’d assume however that if you start with a few people posting high quality content then it will still drop if invites remain at a consistent level, simply because the pool of people to invite is diverse and even if you’d manage to only get the highest quality poster eventually one would end at them bottom. Of course that would take a while, but I also think that people aren’t likely to “attack” lobste.rs as it’s probably not such an incredible effort to get an invite. Also being able to post here isn’t something that will make you rich or something like that.

                1. 1

                  I got my invite yesterday after weeks of trying to get to people on twitter without any response.

                  I dont think there’s anything wrong with the invite system. The public request page could use some sort of change. It’s probably not too hard to cook up a github page with commit history that goes back to years. And that’s for people who have some sort of online presence. I don’t have any online presence on any major site.

                  I’d suggest vetting people on irc’s like a lot of private torrent trackers do.

                  Thats my 2 cents.

                  1. 1

                    Should there maybe be a way to at least confirm that the blog/github/whatever you submit in your invitation request is under your control by giving people a random string to post there and then link to that?

                    No need for a random string, just use Oauth2 to do that work for you.