I stand corrected. Apologies. Nonetheless, you're definitely a "easier to ask forgiveness rather than permission" kind of guy. There will always be a "WE CHANGED SOMETHING" announcement. There is almost never a "WE'RE THINKING OF CHANGING SOMETHING OPINIONS PLEASE" announcement. Even when there is, you'll still do your own thing regardless of concensus. Tags, for example.
BAM. cgod sent me the following text last night: RL is kicking my ass atm. The second day of a training I have to be a part of is about to start. I want to write something worthwhile, and I can't do it. TBH, I agree with a lot of what you've said. You and cgod are right in that one way to preserve what we have/had is to have the tools to do it. Share sorting might be a more valuable tool if you don't follow a narrow band of users. That was my approach with lists. It regards a function that people have been asking about for quite some time, the ability to categorically save and share content. Markup is on the shortlist, and forwardslash has designed a more robust implementation. He has a lot more chops than I do in that regard, but his chops have been most recently dedicated to hubski mobile which is another thing that people have been clamoring for. I have more to add to this conversation. I will add more.I miss the ability to see posts with no, one, or two votes badly. Spending way less time on hubski
There is almost never a "WE'RE THINKING OF CHANGING SOMETHING OPINIONS PLEASE" announcement.
So long as it's a two-sided conversation, it's a conversation worth having. Thanks, and good luck with "real life." My schedule at the moment includes eight Youtube episodes that were supposed to be trickled to me two per month starting in September… but which I got the first one of on Dec 4 and which are all due by the end of the year. Wish I could link to 'em - they're gonna be fuckin' AWESOME - but my name's on 'em.
If we really wanted to couldn't we just TinEye your photo, though? I'm saying this on a purely theoretical level, not a stalker-ish one. An #askhubski question I have considered, but not asked, for a while now, is "Do we share too much personal information on Hubski?" Frankly, as someone who's usually pretty tight on my internet identity - i.e., different usernames across websites, no personal identifying information - my use of Hubski just totally blows the lid off my life. It's a huge gaping security hole. I've been considering bombing my Reddit account because of Hubski.Wish I could link to 'em - they're gonna be fuckin' AWESOME - but my name's on 'em.
Thanks. I am going to look for those videos, but won't let on if I find them. I've been thinking on this. IMO one of the problems that Hubski doesn't yet address in a way that is significantly better or different from other places, including Reddit, is that comments are exposed to a 'tragedy of the commons' scenario. As you say, someone voices their valid yet uninteresting top level comment opinion, and upvotes pour in with little regard to its value as a conversational merit. There are a bunch of ways to address this one issue, some global, and some individual. I don't put too much stock in global solutions (for example, comments with long threads could tend to float up over orphan comments, as they are indicative of better conversation), but sometimes they can nudge things in a better direction. Sometimes they go wrong however, as we could just be featuring flame wars with the same mechanism. One global mechanism i have considered is comment length, or even some sort of language score, but there are other unintended consequences there as well. As you know, we've typically tried to make the solutions based on user choice. And yes, sometimes I drag my feet too long, and sometimes I push things out too willy-nilly. I've been thinking about a user-specific way to alter the comment space, and I think we could do an experimental run, and see if it helps, and if so, if it can be improved upon. In short, not everyone's comments need to sort equally. Here's what I am thinking (a rough idea): You can give any given user a rating: poor, neutral, or good. I'm thinking about three face icons on their profile that you see: frown, neutral, smile. Neutral is the default. If you are so motivated, you can rate that user 'poor' or 'good' instead. The user doesn't know how you have rated them, and the way that you rate them only concerns you. However, when you view a post, the way that you rate users provides extra bias to the comment sort. The users you particularly like will tend towards the top, and the users you don't will tend towards the bottom. Everything is still there, but things will tend to sort in a way that reflects your input. There are comments here that wouldn't induce me to ignore or mute, but I'm typically only going to read them if I am burning time. Also, there are people here that make comments that I almost never want to miss. It only addresses one issue, but I could see it going some distance to better insulate from dilution. I'd like to give it a run, or perhaps some variation of it, as an experiment. Nixing shares was a mistake. Actually, now that they are applied to the feed, I like them even more.
I've been thinking on it for too long myself. For the record: I recognize that I'm often shouting at you for things you do with your website on your time with your money for your entertainment, and that's a dick move. Sorry for that. And for the record, I do it because I've invested a lot of my time in the exact same place and I have nothing to show for it - and whenever something changes without my input, I'm reminded of that fact. Allow me to appeal to you as a scientist: One of the arguments I've made against Team Reddit is that they betatested once. Go check my user page - I got a badge out of it. It worked. They learned a lot. And then they never did it again. I don't know nearly enough about coding to even phrase the conversation in a useful way, but allow me to fumble around for a little: you need a beta interface. You need some code to sit on top of your code that allows users to tweak their own personal functionality. You need a EULA that says "here's what I'm tracking, here's what I'm not, here's why, and by strapping this thing on, you consent to let me watch how you use it." You need to be able to fuck with it regularly. You need to give me the option to turn it on and turn it off and see what it does to functionality. You need to say "we're rolling out this beta feature we've been messing with and we're rolling it out site-wide for a week or until we decide it was a horrible idea, whichever comes first." You do have god-like powers around here - might as well tweak things to see what happens. Perform some versioning, take some notes, run some experiments, see what you get out of it. Shit, write it up well enough and get someone to sign off on it and call it a legit experiment. Maybe get some grant money, who knows? You really do have a unique opportunity here in that you have a small, tight-knit community that doesn't torch'n'pitchfork much. You don't have a massive userbase to rabblerouse. You're a lot more agile than any other aggregator and you have the wherewithal to flight test. No matter what you roll out, you're going to get a global answer. Put the stuff you're thinking about messing with into several peoples' hands before you do it, and you can get a better handle on what you get out of it. Surveymonkey is your friend.
Thanks. No worries. I understand where you are coming from, and know you well enough. The one luxury that Reddit had that we don't is a lack of day jobs. We feel it. There are opportunity costs that we are paying, and many we don't even know about. I wanted to cry when akkartik stepped down. I was drunk at a buddy's birthday party when he sent a text from the Scottish highlands, telling me that he had to pursue a personal coding project that kept eating at him. I was drunk and gobsmacked, but I understood. Hubski is that project for me. forwardslash has been a huge win for Hubski, and the wheels are on tighter than they have ever been. However, we are rationing and it is not ideal. I doubt we can code such a layer, but I do recognize both the experimentation and the approach to it are critical. At the very least, when we do experiment, we can make it clear, and benefit from the discussion that occurs around it. Years ago, I used to be part of this weird blog/rpg thing running among several friends on Blogger. We all grew up gaming, and some sort of rpg/story/rap battle thing organically emerged. One of my pals was the GM and admin, and nothing was more infuriating than the time he started fucking with what people were writing. Sometimes, he deleted posts wholesale. It was supposed to be part of the experience, and the whole thing was a joke to begin with, but our reaction was viscerally negative. As much as I slave over this thing, every one here is here by choice, and I don't want to give the impression that I am not acutely aware of the time and trust people give to the site. There's no getting around that I am always going to be a blackbox of sorts, but I don't want to be arbitrary. Also, I won't pretend that the site's character is a product of my vision alone.