Can you say elaborate on your thinking here? Especially this line:
I think with the current shape of the feature it should be the other way around, posting Shortforms should be gated by the same kinds of things that remove rate limits or allow stronger voting.
If we already grant the premise that broken windows would make Shortform a low-quality place, and that it has less ability to defend itself and recover than other places, then it seems to follow that directing people who might have a greater than usual window-breaking disposition there is the opposite of what you would want to do. In this hypothetical it’s appropriate to keep the defense of this feature in mind when making new interventions, like choosing where to direct new users.
(This conclusion would change if the shape of the feature changes, making it more capable of defending itself, or if the premise that makes it out to be unusually defenseless is seen to be wrong. But I don’t understand how the conclusion wouldn’t follow from the premise.)
I want to distinguish “send people to shortform to write their own shortform posts” from “send people there to comment on other people’s shortform posts.” I meant the former, it sounds like your comment is expecting the latter, but I’m not sure if I’m missing something.
(Separately, to address one of the points here: I’m imagining also changing shortform to allow options like “only a whitelist of friends can comment on your shortform posts”, while making it easier to subscribe to shortform from people you like, which is how FB/Twitter actually work which is what shortform was largely inspired by)
I meant the former, it sounds like your comment is expecting the latter
No, I also mean “their own shortform posts”, with the expectation that it might make the list of all shortform posts (top level comments) made on some day a less interesting thing to browse. With how little voting there is currently on shortform posts, it’s not as easy to ignore most of them if you don’t want to go through all of them, trusting other users to have done a sensible job of evaluating them.
A lot of the shortform posts are just sitting at +1 to +4. A regular post would need to be doing something very wrong to retain such a near-default rating, but good shortform posts remain there, I assume because for most people All Posts is not the bookmarked page and so the shortform posts don’t get much more visibility than random comments. Not paying specific attention, I’m not even sure there is any difference between shortform posts and random comments on old posts, that influences their visibility on the main page.
(Btw, the limit=100 part of the All Posts URL that uncollapses lists of posts is a key usability feature for that page for me. I don’t remember how I learned about it and I don’t see it in the UI. And I don’t see a corresponding thing that works for shortforms. So with my obscure custom starting page URL that makes lists of regular posts more convenient than the default, shortforms become by comparison less convenient than lists of regular posts. Not having to go through the inconvenience of clicking on that Load More is important, the default alternative is ignoring everything behind it. Having to click it for regular posts and then separately and additionally for shortform posts makes it so much less likely for the unlucky shortform posts that are behind it to be seen.)
making it easier to subscribe to shortform from people you like
That’s the sort of thing I was gesturing at with “shortforms should be gated” (as an example intervention that makes some sense, though not necessarily that good overall, intended as a contrast to the intervention discussed in this thread). But using site-wide metrics, instead of having everyone do the work on their own. The latter is painful, never good enough to make it comfortable to ignore things that don’t make it to the list, creating a pressure to do the chore of keeping the list updated. Lots of costs.
An update this made me have: shortform comments should maybe get a self-upvote based on your strongvote strength rather than weakvote, so that it’s a bit easier to scan the new content for stuff that’s likely to be interesting (and this seems fairly principled since we already do this for posts, which shortform are more like than comments)
(although Ruby noted “maybe we want to build Medium votes (which we’ve been thinking about anyway) and make shortform use them”)
Huh? How does strong-upvote-by-default make it any easier to distinguish the “likely to be interesting”? Oh, with the model that old-timers are more interesting always, I guess I get it, but I don’t actually agree.
I’d drop strong voting entirely, perhaps replacing it with medium-voting, but definitely keep it intentional and optional. I really hate seeing comments (even mine) with karma in the teens or twenties and just a few votes.
No, I also mean “their own shortform posts”, with the expectation that it might make the list of all shortform posts (top level comments) made on some day a less interesting thing to browse.
Ah. My impression was that ship had kinda already sailed.
Ah. My impression was that ship had kinda already sailed.
You can always make the problem worse.
My feeling is that yes, you could probably find ways to make low quality users posting to Shortform not dilute better Shortforms from better users, but (1) that’s a bunch of work we haven’t yet done and has opportunity cost, (2) once I’m doing that work, I might as well do it for posts? To the extent low quality users want to write posts (and I think they do), it’s better for author and readers if they’re a post (e.g. you get ToC, etc)
Without necessarily disagreeing with the rest of the comment, I’m not sure I buy this? The mechanism by which I see shortform is either by seeing it in recent discussion (in which case posts, comments or shortforms contribute to about the same rate of dilution, there’s nothing special about shortform), or I subscribe to the shortform of a particular author I like (which I can find via clicking on the user profiles of authors with otherwise highly upvoted comments)
I’m not saying there’s anything special about it, but just as I don’t want to make the SNR of the others worse, I don’t want that for Shortform either. I think discoverability for Shortform is a problem to be solved. Question is how do you know which authors have shortforms they regularly post to? Seems like not the best flow for the only way to find them is via manually checking profiles. (Not that this isn’t solvable, but it’s a thing to do, and if you give up on Shortform SNR, you lose option value and things that maybe could be made to work, etc.
Oh, that disconnect has caused odd decisions before. I never look at the homepage, and just have a bookmark to allPosts. I’ve missed out on Petrov Day announcements (and impact), and am often confused about discussions on “discoverability” which assume homepage starting point for everyone.
Can you say elaborate on your thinking here? Especially this line:
If we already grant the premise that broken windows would make Shortform a low-quality place, and that it has less ability to defend itself and recover than other places, then it seems to follow that directing people who might have a greater than usual window-breaking disposition there is the opposite of what you would want to do. In this hypothetical it’s appropriate to keep the defense of this feature in mind when making new interventions, like choosing where to direct new users.
(This conclusion would change if the shape of the feature changes, making it more capable of defending itself, or if the premise that makes it out to be unusually defenseless is seen to be wrong. But I don’t understand how the conclusion wouldn’t follow from the premise.)
I want to distinguish “send people to shortform to write their own shortform posts” from “send people there to comment on other people’s shortform posts.” I meant the former, it sounds like your comment is expecting the latter, but I’m not sure if I’m missing something.
(Separately, to address one of the points here: I’m imagining also changing shortform to allow options like “only a whitelist of friends can comment on your shortform posts”, while making it easier to subscribe to shortform from people you like, which is how FB/Twitter actually work which is what shortform was largely inspired by)
No, I also mean “their own shortform posts”, with the expectation that it might make the list of all shortform posts (top level comments) made on some day a less interesting thing to browse. With how little voting there is currently on shortform posts, it’s not as easy to ignore most of them if you don’t want to go through all of them, trusting other users to have done a sensible job of evaluating them.
A lot of the shortform posts are just sitting at +1 to +4. A regular post would need to be doing something very wrong to retain such a near-default rating, but good shortform posts remain there, I assume because for most people All Posts is not the bookmarked page and so the shortform posts don’t get much more visibility than random comments. Not paying specific attention, I’m not even sure there is any difference between shortform posts and random comments on old posts, that influences their visibility on the main page.
(Btw, the limit=100 part of the All Posts URL that uncollapses lists of posts is a key usability feature for that page for me. I don’t remember how I learned about it and I don’t see it in the UI. And I don’t see a corresponding thing that works for shortforms. So with my obscure custom starting page URL that makes lists of regular posts more convenient than the default, shortforms become by comparison less convenient than lists of regular posts. Not having to go through the inconvenience of clicking on that Load More is important, the default alternative is ignoring everything behind it. Having to click it for regular posts and then separately and additionally for shortform posts makes it so much less likely for the unlucky shortform posts that are behind it to be seen.)
That’s the sort of thing I was gesturing at with “shortforms should be gated” (as an example intervention that makes some sense, though not necessarily that good overall, intended as a contrast to the intervention discussed in this thread). But using site-wide metrics, instead of having everyone do the work on their own. The latter is painful, never good enough to make it comfortable to ignore things that don’t make it to the list, creating a pressure to do the chore of keeping the list updated. Lots of costs.
An update this made me have: shortform comments should maybe get a self-upvote based on your strongvote strength rather than weakvote, so that it’s a bit easier to scan the new content for stuff that’s likely to be interesting (and this seems fairly principled since we already do this for posts, which shortform are more like than comments)
(although Ruby noted “maybe we want to build Medium votes (which we’ve been thinking about anyway) and make shortform use them”)
Huh? How does strong-upvote-by-default make it any easier to distinguish the “likely to be interesting”? Oh, with the model that old-timers are more interesting always, I guess I get it, but I don’t actually agree.
I’d drop strong voting entirely, perhaps replacing it with medium-voting, but definitely keep it intentional and optional. I really hate seeing comments (even mine) with karma in the teens or twenties and just a few votes.
Ah. My impression was that ship had kinda already sailed.
You can always make the problem worse.
My feeling is that yes, you could probably find ways to make low quality users posting to Shortform not dilute better Shortforms from better users, but (1) that’s a bunch of work we haven’t yet done and has opportunity cost, (2) once I’m doing that work, I might as well do it for posts? To the extent low quality users want to write posts (and I think they do), it’s better for author and readers if they’re a post (e.g. you get ToC, etc)
Without necessarily disagreeing with the rest of the comment, I’m not sure I buy this? The mechanism by which I see shortform is either by seeing it in recent discussion (in which case posts, comments or shortforms contribute to about the same rate of dilution, there’s nothing special about shortform), or I subscribe to the shortform of a particular author I like (which I can find via clicking on the user profiles of authors with otherwise highly upvoted comments)
I’m not saying there’s anything special about it, but just as I don’t want to make the SNR of the others worse, I don’t want that for Shortform either. I think discoverability for Shortform is a problem to be solved. Question is how do you know which authors have shortforms they regularly post to? Seems like not the best flow for the only way to find them is via manually checking profiles. (Not that this isn’t solvable, but it’s a thing to do, and if you give up on Shortform SNR, you lose option value and things that maybe could be made to work, etc.
I’m just one datapoint, but just an hour ago I found an interesting shortform via allPosts and left a comment on it.
Yeah I guess I’m just not modeling allPosts page, which is maybe more on the borderline of usable
Oh, that disconnect has caused odd decisions before. I never look at the homepage, and just have a bookmark to allPosts. I’ve missed out on Petrov Day announcements (and impact), and am often confused about discussions on “discoverability” which assume homepage starting point for everyone.