One of the things I hate most about HN is the fact that these undocumented systems exist, even when known about their exact mechanics aren't, and nobody really questions that or the motivations of...
One of the things I hate most about HN is the fact that these undocumented systems exist, even when known about their exact mechanics aren't, and nobody really questions that or the motivations of the venture capitalists behind the curtains there. Often times weird shit and weird behavior is extremely noticeable there and there is almost never any explanation available as to why.
E.g. When Deimos submitted the Tildes announcement blog post to HN it was doing insanely well and rapidly climbing the front page... then one of the mods changed the title to exactly what it was on the blog post (which removed much of the context from the submission) and suddenly within a matter of minutes the submission was off the front page and sitting way down amongst ones with 1/10th the number of votes and comments. It wasn't the worst outcome in the world, since at that point Tildes didn't want to draw too much attention to itself... but it was rather surprising and annoying to see happen with no explanation as to why.
That's one of the reasons I really love the fact that Tildes is opensource and Deimos pretty open about discussing the mechanics of the site. If people see "weird" behavior here (e.g. comments with more votes sorted under those with less, which is a result of the comment label system), they can just look in the sourcecode, ask in ~tildes and/or ideally just read the docs, to figure out why and how exactly the systems work.
Trust should go both ways IMO, but on HN one side (the users) are largely left blind.
This link was submitted to HN yesterday too, and there were a bunch of people posting comments requesting other features to be added. I made a comment saying that I wasn't sure if HN was even...
This link was submitted to HN yesterday too, and there were a bunch of people posting comments requesting other features to be added. I made a comment saying that I wasn't sure if HN was even under active development any more, because the last update announcement I could find was from over two and a half years ago (July 2016). One of the HN employees (as mentioned in the link, I think there are only two) replied to me and seemed disappointed that I would say that, since "we work so hard on the code".
I wasn't intentionally putting them down, but (at least as far as I could find) they haven't openly told the users anything they've been doing for over 2 years now! It's a bit of a weird cultural disconnect, because I would say that overall, the users of HN highly value things like open-source, transparency, community feedback, and just openness in general, but the site itself doesn't really share those principles.
HN remains a decently well preserved community despite having scaled up quite a bit. Common complaint about the site I agree with: I'm often frustrated at how many people are absurdly stuck up the...
HN remains a decently well preserved community despite having scaled up quite a bit.
Common complaint about the site I agree with: I'm often frustrated at how many people are absurdly stuck up the SV bubble, but that's the nature of the audience and says more about SV than it does about HN.
HN in my opinion remains an absurdly high quality link aggregator if you're interested in tech in any way. I define quality of link aggregators as their SNR, and HN is extremely-high signal. Comment quality has decreased over time but also remains very high signal. (As opposed to lobste.rs which is /r/programming-lite with very few comments anyway).
I've gotten into several disagreements with the mods myself (almost got banned several times) but I can't really fault them when what they're doing is working.
In theory, I like shadowbans as a means of "sandboxing" disruptive users because if they don't know they're banned they can't circumvent it by creating a new account and continuing on a usual. In...
In theory, I like shadowbans as a means of "sandboxing" disruptive users because if they don't know they're banned they can't circumvent it by creating a new account and continuing on a usual. In practice, the big problem is who gets to decide what constitutes a disruptive user. And of course there's the possibility of extreme abuse to stifle dissent, exact personal retaliation, etc. etc.
I think a separation of powers model might help here. If the guys who do the moderating aren't the guys who do the banning, you can avoid some of that bias and abuse. When it's ban-time the teams...
I think a separation of powers model might help here.
If the guys who do the moderating aren't the guys who do the banning, you can avoid some of that bias and abuse. When it's ban-time the teams kick the issue over to some other group that's there just to handle high level disciplinary activity.
Oh shoot I didn't even think about privacy! Oh shoot, I'll have to create a new account and re-do things. Luckily I didn't divulge anything that isn't already obvious about me, given all the other...
Oh shoot I didn't even think about privacy! Oh shoot, I'll have to create a new account and re-do things. Luckily I didn't divulge anything that isn't already obvious about me, given all the other info available online with my username.
I moderated a site in the past, and I gave one shadowban in total. It was to a user that had previously been told to improve, banned temporarily, and then banned multiple times permanently as they...
I moderated a site in the past, and I gave one shadowban in total. It was to a user that had previously been told to improve, banned temporarily, and then banned multiple times permanently as they evaded the bans repeatedly over weeks in order to cause more trouble. When they got shadowbanned, I watched their posts, and they continued to try to start shit. They also whined that nobody was paying attention to them. I would have found this depressing if I hadn't seen them do that a few days before and then immediately flame the people who replied to them. Within a day of nobody taking any of their now-hidden bait, the user got bored and then never returned. The shadowban was much more successful than I predicted. I had quietly implemented the shadowban feature into the site software specifically for that user, and I de-implemented it afterwards.
I consider it the nuclear option. I think it should only be used on people who have been told that they're in the wrong and have had the chance to adapt.
I once was playing a social web game with a friend where chat messages showed up above your character's head. Some players would put icon emojis in their chat, which made it look like an icon was just floating over you. A few people were putting three hearts over their head, which within the game UI and font made it look like a Zelda-style healthbar was over their head, which I thought was a pretty funny effect. I copied them into my clipboard, and about every 15 seconds for a few minutes, I quickly pressed enter, ctrl-v, enter to keep the chat message above my head. Shortly after, my friend and I could no longer find each other in the game. We assumed it was buggy so we spent 45 minutes trying to find each other in the game, and I looked at the network requests and code in Chrome devtools to try to figure out the issue. I did end up finding a few bugs. I emailed the developer about the bugs and the glitch that I couldn't find my friend in the game. ... It turns out that I had tripped some kind of chat rate-limiting and had gotten automatically shadowbanned. (... Why did the game have chat rate-limiting like that? The chat messages weren't disruptive. Your new message just replaced your current one on the screen, and didn't block other users. ... and why make the chat-rate limiting go straight to a shadowban?) I was pretty pissed. I didn't say so, though I just didn't touch the game again. I don't think shadowbans should be given to people that weren't warned in any way, and probably not by automated systems.
This kill - revive - kill again forever mechanic is kinda interesting. I've been wondering how to work in some kind of appeal angle to the moderation. It seems like we've discussed and rejected...
If a user has 31 Karma, they can also vouch for a [dead] submission/comment. A vouched submission/comment has its rank restored (and potentially improved as the vouch can counteract the effects of flags), but it can be [dead] again at which point it can't be re-vouched.
This kill - revive - kill again forever mechanic is kinda interesting. I've been wondering how to work in some kind of appeal angle to the moderation.
It seems like we've discussed and rejected most of the ideas on this HN list as either sub-optimal or just plain bad (like shadowbans). Where HN is using Karma, we'll be using trust as the metric to unlock things, and trust will be more sophisticated than simple karma.
Posting in ~tildes because I consider HackerNews a good ~comp-equivalent, and its idiosyncrasies and novelties always interest me. Please note I'm not endorsing or rejecting any specific features...
Posting in ~tildes because I consider HackerNews a good ~comp-equivalent, and its idiosyncrasies and novelties always interest me. Please note I'm not endorsing or rejecting any specific features mentioned in this document, I'm only posting it because I'm sure there's some interesting opinions which could come out of this :)
I find it interesting that HN has various tiers of status, but for the most part keeps those internal (sans the green "noob" username). This is in contrast to Reddit, which has cakedays, flair,...
I find it interesting that HN has various tiers of status, but for the most part keeps those internal (sans the green "noob" username). This is in contrast to Reddit, which has cakedays, flair, mod/admin colors, and of course user karma.
I was talking to my roommate about how Tildes' stated goal of being donation-supported seems unsustainable at scale, and it got me wondering if there was a way to promote users who have donated in a non-obnoxious way. Perhaps bolding a username or making it a different color if they've donated in the past month/year/etc. Reddit shows you people who have been guilded, but not the guilders, and I think that makes posting more of a competition rather than a communal discussion.
Outside of that thought, I do like the de-emphasis of downvoting on HN, and I wonder if that is one of the main reasons for higher quality discussion, such as on Tildes.
I really think we should avoid adding anything visible on a username in regular threads. No matter how subtle and well-meaning the gesture is it's going to have more and more weight the older this...
it got me wondering if there was a way to promote users who have donated in a non-obnoxious way. Perhaps bolding a username or making it a different color if they've donated in the past month/year/etc.
I really think we should avoid adding anything visible on a username in regular threads. No matter how subtle and well-meaning the gesture is it's going to have more and more weight the older this site gets and it's going to affect how you persieve that user from the first time you see it or interact with him/her. Now, that might not inherently be a bad thing but I'm a firm believer that everyone should come from the same standpoint no matter how old or frequent a user you are. Power users are automatically going to get noticed just by their username anyways. I would suggest that you can get a "donation badge" (or something similar) on your user page so it's not immitiately visible in threads but if you go to a users page anyways to learn more about this person you're going to see it.
Agreed, and it creates a barrier to entry for people who might otherwise be willing to donate, but can't afford it. They may have high trust and be well-known around the site, but they'd be...
Agreed, and it creates a barrier to entry for people who might otherwise be willing to donate, but can't afford it. They may have high trust and be well-known around the site, but they'd be singled out simply because they can't afford a monthly or one-time donation. I've seen it happen on other forums.
I think showing donator status is good, since it doesn't affect anyone at all and just shows that hey, this person has donated. I abhor reddit's gilded system.
I think showing donator status is good, since it doesn't affect anyone at all and just shows that hey, this person has donated. I abhor reddit's gilded system.
I disagree about moving it... IMO this submission is relevant to tildes and appropriate in ~tildes because it discusses the hidden features of a very similar site (albeit with different scope and...
I disagree about moving it... IMO this submission is relevant to tildes and appropriate in ~tildes because it discusses the hidden features of a very similar site (albeit with different scope and subject focus); It's a good jumping off point for discussing said features and whether they would be appropriate for use on here or not.
I think a jumping-off point for discussion about Tildes would include a mention of how some of these features might be useful for Tildes, rather than just "Here's a list of HackerNews features....
I think a jumping-off point for discussion about Tildes would include a mention of how some of these features might be useful for Tildes, rather than just "Here's a list of HackerNews features. Discuss."
Using this logic, we can post dumps of all features of all social media, content aggregators, and message boards here, with a note saying "Discuss."
The vast majority of posts on this platform do not have an accompanying textual content associated with them; I don't see how when the discussion suddenly becomes about Tildes, that needs to...
I think a jumping-off point for discussion about Tildes would include a mention of how some of these features might be useful for Tildes, rather than just "Here's a list of HackerNews features. Discuss."
The vast majority of posts on this platform do not have an accompanying textual content associated with them; I don't see how when the discussion suddenly becomes about Tildes, that needs to change.
Using this logic, we can post dumps of all features of all social media, content aggregators, and message boards here, with a note saying "Discuss."
This already happens on Tildes already, with every other link post. While the OP is certainly free to scope/descope/add context to a post they've made, it's neither required nor necessarily needed. We're all smart cookies with lots of ways to improve this site, we don't need a source of conversational entropy to speak up.
Sometimes less is more.
I don't feel the need to say more on this subject so I'll leave it at that.
Yeah, it's not worth arguing about. I edited a short prompt on to the end of the title to make it more obvious why it's in ~tildes (and hopefully I didn't misstate your purpose).
Yeah, it's not worth arguing about. I edited a short prompt on to the end of the title to make it more obvious why it's in ~tildes (and hopefully I didn't misstate your purpose).
Yep. The history of putting punctuation marks inside or outside quotation marks goes back to the early days of movable-type printing in England 400 years ago. After some variations, splits, and...
Had no idea that putting them on the outside was an American thing, apparently.
Yep. The history of putting punctuation marks inside or outside quotation marks goes back to the early days of movable-type printing in England 400 years ago. After some variations, splits, and changes, the end result is that Americans are the only English users who put periods that don't belong to a quotation inside the quotation marks ("printers punctuation"), while everyone else puts full stops that don't belong to a quotation outside the quotation marks ("logical punctuation").
As for putting an "apostrophe-s" on a word that already ends in an "s", that generally comes down to pronunciation: if you say the second "s", you write the second "s"; if you don't say the second "s", you don't write the second "s".
The intent was to get some discussion going about the mechanics of conversation board style sites that have similar goals to Tildes—and community feedback on whether we like/dislike certain...
The intent was to get some discussion going about the mechanics of conversation board style sites that have similar goals to Tildes—and community feedback on whether we like/dislike certain aspects, and how they'd fit on here; because previously when we've discussed Tildes in ~tech or ~comp, a lot of us end up apologising for going off topic or "meta" for discussing the current Tildes mechanics.
But if you feel it's best in another board, by all means, it's your prerogative to move it. I don't particularly mind, I've made my intent clear.
One of the things I hate most about HN is the fact that these undocumented systems exist, even when known about their exact mechanics aren't, and nobody really questions that or the motivations of the venture capitalists behind the curtains there. Often times weird shit and weird behavior is extremely noticeable there and there is almost never any explanation available as to why.
E.g. When Deimos submitted the Tildes announcement blog post to HN it was doing insanely well and rapidly climbing the front page... then one of the mods changed the title to exactly what it was on the blog post (which removed much of the context from the submission) and suddenly within a matter of minutes the submission was off the front page and sitting way down amongst ones with 1/10th the number of votes and comments. It wasn't the worst outcome in the world, since at that point Tildes didn't want to draw too much attention to itself... but it was rather surprising and annoying to see happen with no explanation as to why.
That's one of the reasons I really love the fact that Tildes is opensource and Deimos pretty open about discussing the mechanics of the site. If people see "weird" behavior here (e.g. comments with more votes sorted under those with less, which is a result of the comment label system), they can just look in the sourcecode, ask in ~tildes and/or ideally just read the docs, to figure out why and how exactly the systems work.
Trust should go both ways IMO, but on HN one side (the users) are largely left blind.
This link was submitted to HN yesterday too, and there were a bunch of people posting comments requesting other features to be added. I made a comment saying that I wasn't sure if HN was even under active development any more, because the last update announcement I could find was from over two and a half years ago (July 2016). One of the HN employees (as mentioned in the link, I think there are only two) replied to me and seemed disappointed that I would say that, since "we work so hard on the code".
I wasn't intentionally putting them down, but (at least as far as I could find) they haven't openly told the users anything they've been doing for over 2 years now! It's a bit of a weird cultural disconnect, because I would say that overall, the users of HN highly value things like open-source, transparency, community feedback, and just openness in general, but the site itself doesn't really share those principles.
I've definitely seen a couple of Tell HNs with some changelog style news posted by dang before. But it may have been two years ago admittedly...
Yeah, his list of submissions is here, that's where I found the July 2016 post (it's #14 on there): https://news.ycombinator.com/submitted?id=dang
HN remains a decently well preserved community despite having scaled up quite a bit.
Common complaint about the site I agree with: I'm often frustrated at how many people are absurdly stuck up the SV bubble, but that's the nature of the audience and says more about SV than it does about HN.
HN in my opinion remains an absurdly high quality link aggregator if you're interested in tech in any way. I define quality of link aggregators as their SNR, and HN is extremely-high signal. Comment quality has decreased over time but also remains very high signal. (As opposed to lobste.rs which is /r/programming-lite with very few comments anyway).
I've gotten into several disagreements with the mods myself (almost got banned several times) but I can't really fault them when what they're doing is working.
In theory, I like shadowbans as a means of "sandboxing" disruptive users because if they don't know they're banned they can't circumvent it by creating a new account and continuing on a usual. In practice, the big problem is who gets to decide what constitutes a disruptive user. And of course there's the possibility of extreme abuse to stifle dissent, exact personal retaliation, etc. etc.
I think a separation of powers model might help here.
If the guys who do the moderating aren't the guys who do the banning, you can avoid some of that bias and abuse. When it's ban-time the teams kick the issue over to some other group that's there just to handle high level disciplinary activity.
I'm about to go check how to start contributing to OSM for my city, but do you have any tips for starting out?
Oh shoot I didn't even think about privacy! Oh shoot, I'll have to create a new account and re-do things. Luckily I didn't divulge anything that isn't already obvious about me, given all the other info available online with my username.
P.S. Thank you for such a great guide!
This is great! I think I've found a new hobby :)
I moderated a site in the past, and I gave one shadowban in total. It was to a user that had previously been told to improve, banned temporarily, and then banned multiple times permanently as they evaded the bans repeatedly over weeks in order to cause more trouble. When they got shadowbanned, I watched their posts, and they continued to try to start shit. They also whined that nobody was paying attention to them. I would have found this depressing if I hadn't seen them do that a few days before and then immediately flame the people who replied to them. Within a day of nobody taking any of their now-hidden bait, the user got bored and then never returned. The shadowban was much more successful than I predicted. I had quietly implemented the shadowban feature into the site software specifically for that user, and I de-implemented it afterwards.
I consider it the nuclear option. I think it should only be used on people who have been told that they're in the wrong and have had the chance to adapt.
I once was playing a social web game with a friend where chat messages showed up above your character's head. Some players would put icon emojis in their chat, which made it look like an icon was just floating over you. A few people were putting three hearts over their head, which within the game UI and font made it look like a Zelda-style healthbar was over their head, which I thought was a pretty funny effect. I copied them into my clipboard, and about every 15 seconds for a few minutes, I quickly pressed enter, ctrl-v, enter to keep the chat message above my head. Shortly after, my friend and I could no longer find each other in the game. We assumed it was buggy so we spent 45 minutes trying to find each other in the game, and I looked at the network requests and code in Chrome devtools to try to figure out the issue. I did end up finding a few bugs. I emailed the developer about the bugs and the glitch that I couldn't find my friend in the game. ... It turns out that I had tripped some kind of chat rate-limiting and had gotten automatically shadowbanned. (... Why did the game have chat rate-limiting like that? The chat messages weren't disruptive. Your new message just replaced your current one on the screen, and didn't block other users. ... and why make the chat-rate limiting go straight to a shadowban?) I was pretty pissed. I didn't say so, though I just didn't touch the game again. I don't think shadowbans should be given to people that weren't warned in any way, and probably not by automated systems.
This kill - revive - kill again forever mechanic is kinda interesting. I've been wondering how to work in some kind of appeal angle to the moderation.
It seems like we've discussed and rejected most of the ideas on this HN list as either sub-optimal or just plain bad (like shadowbans). Where HN is using Karma, we'll be using trust as the metric to unlock things, and trust will be more sophisticated than simple karma.
Posting in ~tildes because I consider HackerNews a good ~comp-equivalent, and its idiosyncrasies and novelties always interest me. Please note I'm not endorsing or rejecting any specific features mentioned in this document, I'm only posting it because I'm sure there's some interesting opinions which could come out of this :)
I find it interesting that HN has various tiers of status, but for the most part keeps those internal (sans the green "noob" username). This is in contrast to Reddit, which has cakedays, flair, mod/admin colors, and of course user karma.
I was talking to my roommate about how Tildes' stated goal of being donation-supported seems unsustainable at scale, and it got me wondering if there was a way to promote users who have donated in a non-obnoxious way. Perhaps bolding a username or making it a different color if they've donated in the past month/year/etc. Reddit shows you people who have been guilded, but not the guilders, and I think that makes posting more of a competition rather than a communal discussion.
Outside of that thought, I do like the de-emphasis of downvoting on HN, and I wonder if that is one of the main reasons for higher quality discussion, such as on Tildes.
I really think we should avoid adding anything visible on a username in regular threads. No matter how subtle and well-meaning the gesture is it's going to have more and more weight the older this site gets and it's going to affect how you persieve that user from the first time you see it or interact with him/her. Now, that might not inherently be a bad thing but I'm a firm believer that everyone should come from the same standpoint no matter how old or frequent a user you are. Power users are automatically going to get noticed just by their username anyways. I would suggest that you can get a "donation badge" (or something similar) on your user page so it's not immitiately visible in threads but if you go to a users page anyways to learn more about this person you're going to see it.
Agreed, and it creates a barrier to entry for people who might otherwise be willing to donate, but can't afford it. They may have high trust and be well-known around the site, but they'd be singled out simply because they can't afford a monthly or one-time donation. I've seen it happen on other forums.
I think showing donator status is good, since it doesn't affect anyone at all and just shows that hey, this person has donated. I abhor reddit's gilded system.
But how is this post about Tildes? It's all about HackerNews. I'm inclined to move it to ~tech with all the other posts about social media.
I disagree about moving it... IMO this submission is relevant to tildes and appropriate in ~tildes because it discusses the hidden features of a very similar site (albeit with different scope and subject focus); It's a good jumping off point for discussing said features and whether they would be appropriate for use on here or not.
I think a jumping-off point for discussion about Tildes would include a mention of how some of these features might be useful for Tildes, rather than just "Here's a list of HackerNews features. Discuss."
Using this logic, we can post dumps of all features of all social media, content aggregators, and message boards here, with a note saying "Discuss."
But I'll leave it alone.
The vast majority of posts on this platform do not have an accompanying textual content associated with them; I don't see how when the discussion suddenly becomes about Tildes, that needs to change.
This already happens on Tildes already, with every other link post. While the OP is certainly free to scope/descope/add context to a post they've made, it's neither required nor necessarily needed. We're all smart cookies with lots of ways to improve this site, we don't need a source of conversational entropy to speak up.
Sometimes less is more.
I don't feel the need to say more on this subject so I'll leave it at that.
Yeah, it's not worth arguing about. I edited a short prompt on to the end of the title to make it more obvious why it's in ~tildes (and hopefully I didn't misstate your purpose).
Fixed both. Thanks for the heads up.
Yep. The history of putting punctuation marks inside or outside quotation marks goes back to the early days of movable-type printing in England 400 years ago. After some variations, splits, and changes, the end result is that Americans are the only English users who put periods that don't belong to a quotation inside the quotation marks ("printers punctuation"), while everyone else puts full stops that don't belong to a quotation outside the quotation marks ("logical punctuation").
As for putting an "apostrophe-s" on a word that already ends in an "s", that generally comes down to pronunciation: if you say the second "s", you write the second "s"; if you don't say the second "s", you don't write the second "s".
The intent was to get some discussion going about the mechanics of conversation board style sites that have similar goals to Tildes—and community feedback on whether we like/dislike certain aspects, and how they'd fit on here; because previously when we've discussed Tildes in ~tech or ~comp, a lot of us end up apologising for going off topic or "meta" for discussing the current Tildes mechanics.
But if you feel it's best in another board, by all means, it's your prerogative to move it. I don't particularly mind, I've made my intent clear.