Which leads to a lot of passive-aggressive ironic tagging.
So just like Twitter, then.
The interface was borne of Hacker News but has been rewritten since.
HN and reddit are approximately the same general interface with a different theme, but okay.
following and filtering users, tags and domains gives you a much more complex and nuanced feed.
This is way too much work to get anywhere at the moment. I'm pretty well versed in how social media and content aggregators work - you may even recognize my username from reddit. I came here, and it took me several minutes to figure out how to get any sort of active customization, and while I've got a handful of tags... well, you've described the problem pretty clearly in that interest-algebra post, getting started on the Y axis is borderline impossible and as a new user I have no idea who to give a shit about on the X axis. Twitter and facebook, in their "social media" mode, bootstrap that problem by using peoples' real-world relationships to get things started.
Hubski is trying to straddle some sort of in between territory and it's doing a bad job of it at the moment, as far as I can make out. If tags could be nested somehow and content out of lower tags bubbled up into higher tags once they got popular enough, maybe... ( #nfl having 32 team tags under it, and when content for a particular team has enough interest it would "overflow" to the parent tag; you'd need multiple inheritance to also have teams linked to their local area - if the Blackhawks win the Stanley Cup it should show up directly from #chicago even if it's not tagged chicago directly.)
Of course, the issue may just be that I have little interest in the "X axis" of following individuals. The people I know in the real world I talk to on facebook or whatever. I'm not interested in following random people on the internet.
(3) comments and replies become a part of an active conversation that can go on for days weeks months or years. Hubski does not archive; it's possible to reactivate old content from weeks ago and have it appear on a user's page.
Yeah, my point there was to contrast it with twitter, wherein "replies" aren't really any different from other top-level tweets. (There's a bit of metadata now tracking them back to where they came from so it's easier to follow conversations after the fact, but that used to not be there.)
PMs are not necessarily person-to-person; they thread and concatenate just like comments.
How does the privacy on this work?
(Side note: this markup language blows, and the lack of preview isn't helpful at all. Apparently an open-parentheses before the tag markup breaks the tag markup.)