showing posts flaired Question — clear

0

Struggling with picking an instance

preview

I’m a lemmy.zip user and I moderate two communities. One of them is new and another has about 350 subscribers. A recent issue related to .zip weighs too heavily on me to be able to stay on the instance.

I’ve had PieFed on my radar for a while now, as evidenced by the registration date of this account, but because moving community subscribers is hard and requires regular reminders, I postponed their migration. To this end, it’s worth noting that PieFed currently allows to pull some old posts if you migrate, making it superior to Lemmy for moving a community.

Being fresh off of what I’m convinced are instances of brigading of my community, I also appreciate Rimu’s focus on voting and activity patterns, as well as what looks like attempts to offer mod tools that aid with those and other tasks.

So the PieFed software feels like an obvious choice. What I’m struggling with is picking an instance. Aside from fedinsfw, .social is the only instance that has enough monthly active users, 1,500, to compete with some of the bigger Lemmy instances. Excluding these two, the others get around 300 monthly at best. The latest version of PieFed, 1.8, is from a month ago, but many are running on outdated versions. This is a problem because I like PieFed’s active pace of development. Even my own bug report and a wording suggestion from just recently were quickly addressed. Outdated versions wouldn’t be able to make use of these improvements.

Going instance by instance, here’s my thought process:

  • Looking at the numbers, piefed.social is the way to go. However, given that my previous two instances involved serious disagreements with the staff, I don’t want to run into a situation where I’m at odds with Rimu, as it could also close the door to my contributions to PieFed as a whole, either by my choice or by Rimu’s.

  • piefed.zip and piefed.world are not acceptable, as I’ve already been through those.

  • piefed.ca is the next best based on its number of users, but it’s on PieFed 1.7 and I haven’t been able to reach the admins over the past two days. One criterion I’m looking for after being brigaded is being able to urgently get in touch with an admin, so responsiveness matters.

  • feddit.online is up to date, but I can’t quite understand it. It’s positioned as being separate from .social, but its main admin is a generic name, there are two main admins from .social, and one other admin with zero public activity.

  • piefed.blahaj.zone runs on an older version of PieFed. One moderator of a flagship community on lemmy.blahaj.zone also moderates a community with a long history of brigading and harassment, so I’m not sure it’s a great choice.

And that’s it for the instances with at least 100 monthly users. Given that everything is federated, you’d think the per-instance user counts don’t matter, but my limited understanding of how federation works is that posts may not even be pulled if there aren’t enough users subscribing to the communities they originate from. Moreover, if an instance is not attracting enough users, I feel it is more likely to be closed down.

I could go over some of the others, but there’s some seemingly petty drama involving the next on the list, so it may be best not to.

Edit: Corrected one fact about feddit.online.

I’m a lemmy.zip user and I moderate two communities. One of them is new and another has about 350 subscribers. A recent issue related to .zip weighs too heavily on me to be able to stay on the instance.

I’ve had PieFed on my radar for a while now, as evidenced by the registration date of this account, but because moving community subscribers is hard and requires regular reminders, I postponed their migration. To this end, it’s worth noting that PieFed currently allows to pull some old posts if you migrate, making it superior to Lemmy for moving a community.

Being fresh off of what I’m convinced are instances of brigading of my community, I also appreciate Rimu’s focus on voting and activity patterns, as well as what looks like attempts to offer mod tools that aid with those and other tasks.

So the PieFed software feels like an obvious choice. What I’m struggling with is picking an instance. Aside from fedinsfw, .social is the only instance that has enough monthly active users, 1,500, to compete with some of the bigger Lemmy instances. Excluding these two, the others get around 300 monthly at best. The latest version of PieFed, 1.8, is from a month ago, but many are running on outdated versions. This is a problem because I like PieFed’s active pace of development. Even my own bug report and a wording suggestion from just recently were quickly addressed. Outdated versions wouldn’t be able to make use of these improvements.

Going instance by instance, here’s my thought process:

  • Looking at the numbers, piefed.social is the way to go. However, given that my previous two instances involved serious disagreements with the staff, I don’t want to run into a situation where I’m at odds with Rimu, as it could also close the door to my contributions to PieFed as a whole, either by my choice or by Rimu’s.

  • piefed.zip and piefed.world are not acceptable, as I’ve already been through those.

  • piefed.ca is the next best based on its number of users, but it’s on PieFed 1.7 and I haven’t been able to reach the admins over the past two days. One criterion I’m looking for after being brigaded is being able to urgently get in touch with an admin, so responsiveness matters.

  • feddit.online is up to date, but I can’t quite understand it. It’s positioned as being separate from .social, but its main admin is a generic name, there are two main admins from .social, and one other admin with zero public activity.

  • piefed.blahaj.zone runs on an older version of PieFed. One moderator of a flagship community on lemmy.blahaj.zone also moderates a community with a long history of brigading and harassment, so I’m not sure it’s a great choice.

And that’s it for the instances with at least 100 monthly users. Given that everything is federated, you’d think the per-instance user counts don’t matter, but my limited understanding of how federation works is that posts may not even be pulled if there aren’t enough users subscribing to the communities they originate from. Moreover, if an instance is not attracting enough users, I feel it is more likely to be closed down.

I could go over some of the others, but there’s some seemingly petty drama involving the next on the list, so it may be best not to.

Edit: Corrected one fact about feddit.online.

0

Feedback : Piefed UI in other language (or in english)

preview

Hi everyone ! I assume most of us use Piefed in english and it render well. :)

In other language, i don’t know if the UI is still good or not. For example, in the french version, the reply button “répondre” overlaps with the threads and the arrow.

So, let’s chase together any UI bug you may have encountered. :)

To help us, you can :

  • join a screenshot and its location (eg : found in settings > blocklist…)
  • tell us the theme used
  • piefed version

If you use another front-end, you can also share it here, i will do my best to reach the related dev team.

Thank a lot for helping us :3

NB : if Piefed dev team has any suggestion for title or body text, you can PM me the change :3

Hi everyone ! I assume most of us use Piefed in english and it render well. :)

In other language, i don’t know if the UI is still good or not. For example, in the french version, the reply button “répondre” overlaps with the threads and the arrow.

So, let’s chase together any UI bug you may have encountered. :)

To help us, you can :

  • join a screenshot and its location (eg : found in settings > blocklist…)
  • tell us the theme used
  • piefed version

If you use another front-end, you can also share it here, i will do my best to reach the related dev team.

Thank a lot for helping us :3

NB : if Piefed dev team has any suggestion for title or body text, you can PM me the change :3

0

Question about new user onboarding, and thanks for the option to Proceed without choosing any topics!

preview

Logged onto piefed.social via my laptop instead of my phone for the first time. Got hit with a sequence of what looks like new user onboarding: a form giving me options on what I want to do with each of NSFW, NSFL, reposts, AI, and certain political stuff. Filtered, hidden, blurred, shown regularly, and maybe one other option. Also got hit with the posted screenshot asking me to pick topics that interest me (probably to populate my feed).

I am curious if this is a new feature that went out to PieFed users, or just PieFed.social. Or if it was always there and I just got hit by it on the first time when I switched from phone to laptop. Also curious who it went to; since I am subscribed to a few communities and have used this account to post and comment before; it’s not totally fresh. Did it go out to everyone, or just new enough accounts?

And I wanted to thank you so much for not forcing me to pick topics (and thus, I assume, sub me to pre-selected communities) and letting me just hit Proceed without choosing any topics. This way I can pick by myself without having to go to the trouble of unsubbing from forced-on-me subs. It is a small convenience, it would take maybe a minute max on my phone to undo the forced sub if unsubbing went really inconveniently, but I really appreciate it. You’re offering a familiar onboarding experience for social media that probably works for a lot of people (great for onboarding new folks who might need a little more help, people who I assume are outside of the typical Fediverse demographic and who I’d obviously like to be able to join us), while still offering choice to the user to curate totally for themselves (which I imagine the current Fediverse demographic would rather take). Helpful for folks like me not new to the Fediverse, just instance-hopping (I’m originally from programming.dev, lured over by the filtering features PieFed has). And I am sure it’s helpful for others too. It probably went to other people who also already set up an account and don’t want to be forced to add new communities to their setup.

Logged onto piefed.social via my laptop instead of my phone for the first time. Got hit with a sequence of what looks like new user onboarding: a form giving me options on what I want to do with each of NSFW, NSFL, reposts, AI, and certain political stuff. Filtered, hidden, blurred, shown regularly, and maybe one other option. Also got hit with the posted screenshot asking me to pick topics that interest me (probably to populate my feed).

I am curious if this is a new feature that went out to PieFed users, or just PieFed.social. Or if it was always there and I just got hit by it on the first time when I switched from phone to laptop. Also curious who it went to; since I am subscribed to a few communities and have used this account to post and comment before; it’s not totally fresh. Did it go out to everyone, or just new enough accounts?

And I wanted to thank you so much for not forcing me to pick topics (and thus, I assume, sub me to pre-selected communities) and letting me just hit Proceed without choosing any topics. This way I can pick by myself without having to go to the trouble of unsubbing from forced-on-me subs. It is a small convenience, it would take maybe a minute max on my phone to undo the forced sub if unsubbing went really inconveniently, but I really appreciate it. You’re offering a familiar onboarding experience for social media that probably works for a lot of people (great for onboarding new folks who might need a little more help, people who I assume are outside of the typical Fediverse demographic and who I’d obviously like to be able to join us), while still offering choice to the user to curate totally for themselves (which I imagine the current Fediverse demographic would rather take). Helpful for folks like me not new to the Fediverse, just instance-hopping (I’m originally from programming.dev, lured over by the filtering features PieFed has). And I am sure it’s helpful for others too. It probably went to other people who also already set up an account and don’t want to be forced to add new communities to their setup.

0

Suspect Registrations

preview

PieFed 1.8.7

There have been a few registrations which appear related to each other. When I deny the registration, a new one is made fairly quickly.

I adapted by not responding to the registration with either Approve or Deny, which in turn stopped the cycle.

For registrations I imagine there’s a feedback for when the denial is made. Can there be an option for Admin to select no feedback to resist automated bots filling out forms as soon as the denial is made?

Alternatively, can there also be an option to make no selection and to put the registration on some other no-action list to organize from the fresher and hopefully more legitimate registrations?

PieFed 1.8.7

There have been a few registrations which appear related to each other. When I deny the registration, a new one is made fairly quickly.

I adapted by not responding to the registration with either Approve or Deny, which in turn stopped the cycle.

For registrations I imagine there’s a feedback for when the denial is made. Can there be an option for Admin to select no feedback to resist automated bots filling out forms as soon as the denial is made?

Alternatively, can there also be an option to make no selection and to put the registration on some other no-action list to organize from the fresher and hopefully more legitimate registrations?

0

Titles of posts by one user not showing

preview

In the community !casualuk@feddit.uk (https://piefed.social/c/casualuk@feddit.uk), I’ve noticed that posts by one user in particular (ladybutterfly@piefed.blahaj.zone) don’t have post titles visible, when viewing from piefed.social (but other instances don’t have this issue). It doesn’t seem to be an issue with my account, because the issue still appears when logged out.

For example, these two posts appear to have no title:

But if you view the same posts on piefed.world, they do have a title:

Does anyone know if there’s a way to fix this issue, so I can see all post titles on piefed.social?

Thanks if anyone has any help.

In the community !casualuk@feddit.uk (https://piefed.social/c/casualuk@feddit.uk), I’ve noticed that posts by one user in particular (ladybutterfly@piefed.blahaj.zone) don’t have post titles visible, when viewing from piefed.social (but other instances don’t have this issue). It doesn’t seem to be an issue with my account, because the issue still appears when logged out.

For example, these two posts appear to have no title:

But if you view the same posts on piefed.world, they do have a title:

Does anyone know if there’s a way to fix this issue, so I can see all post titles on piefed.social?

Thanks if anyone has any help.

0

Flair being added to every post regardless of community

preview

I wasn’t paying attention and accidentally posted an article to !nsl@lemmy.ca instead of !nwsl@lemmy.world last night. I used the “move post” feature to move it between the two communities, but now my “Vancouver Rise” flair from the NSL community is being added to every post I make, regardless of which community I post in. I don’t think this is a client error, because I see the flair when I go to look at comments like this one in my browser while logged out. Is this a bug? Is there something I did wrong?

I wasn’t paying attention and accidentally posted an article to !nsl@lemmy.ca instead of !nwsl@lemmy.world last night. I used the “move post” feature to move it between the two communities, but now my “Vancouver Rise” flair from the NSL community is being added to every post I make, regardless of which community I post in. I don’t think this is a client error, because I see the flair when I go to look at comments like this one in my browser while logged out. Is this a bug? Is there something I did wrong?

0

Is there a way to remove specific followers?

preview

I don’t want to manually approve every follower but is there a way to remove specific followers and if there isn’t can this feature be added? And maybe a setting to hide the list of followers?

I don’t want to manually approve every follower but is there a way to remove specific followers and if there isn’t can this feature be added? And maybe a setting to hide the list of followers?

0

Are PieFed posts/comments vissable on Mastodon in a way?

preview

How do Tags works in a post? Does it make it visible to Mastodon in a way?

Mastodon has ~1.5 Million active users, so better integration between PieFed and Mastodon could see massive growth in users and activity.

My thinking is, it would be great that, if a user makes a post on c/SouthAfrica that it would automatically be tagged with “South Africa” and be visible as a post on Mastodon.

Comments through Mastodon would then also show up on the post in PieFed, and visa versa.

How do Tags works in a post? Does it make it visible to Mastodon in a way?

Mastodon has ~1.5 Million active users, so better integration between PieFed and Mastodon could see massive growth in users and activity.

My thinking is, it would be great that, if a user makes a post on c/SouthAfrica that it would automatically be tagged with “South Africa” and be visible as a post on Mastodon.

Comments through Mastodon would then also show up on the post in PieFed, and visa versa.

0

Another one about Lemmy and Piefed seeing/parsing things differently.

preview

Apparently Lemmy turns my backslashes (used to escape round brackets inside a link) into forward slashes, leading to Lemmy users complaining:

Lemmy does not require the round bracket to be escaped in this case. Why it turns backslashes into forward slashes I have no idea, but it does so even when I try the same thing straight from a Lemmy instance.

Is escaping things with backslashes not common practice in Markdown?

I know there’s so many implementations of Markdown it’s hard to even speak of a Spec, but backslashes serve this purpose almost universally. On computers that is.

I sense an overarching narrative of Lemmy vs Piefed developing here. Or maybe a need for communication between all client developers, on how to implement things so they work across the fediverse.

Hmm, it does feel more like a Lemmy problem in this case.
Where could I raise awareness on the Lemmy side, short of opening an issue on their git repos?

Apparently Lemmy turns my backslashes (used to escape round brackets inside a link) into forward slashes, leading to Lemmy users complaining:

Lemmy does not require the round bracket to be escaped in this case. Why it turns backslashes into forward slashes I have no idea, but it does so even when I try the same thing straight from a Lemmy instance.

Is escaping things with backslashes not common practice in Markdown?

I know there’s so many implementations of Markdown it’s hard to even speak of a Spec, but backslashes serve this purpose almost universally. On computers that is.

I sense an overarching narrative of Lemmy vs Piefed developing here. Or maybe a need for communication between all client developers, on how to implement things so they work across the fediverse.

Hmm, it does feel more like a Lemmy problem in this case.
Where could I raise awareness on the Lemmy side, short of opening an issue on their git repos?

0

Automatic Blurb Extraction

preview

Meaning that a link post automatically extracts some meta info - every time it’s clicked, not only when it’s created. Because it appears that Lemmy doesn’t automatically do it once at posting time.

Examples:

https://piefed.social/c/mediatheque@feddit.org/p/2446966/la-la-land
https://piefed.zip/c/netzkultur@feddit.org/p/1874208/we-noticed-that-you-are-blocking-our-ads

I only see a link. When I view the original post on a Lemmy-based instance, I see an automatically, real-time extracted blurb.

Is pyfedi’s stance something like “it should be enough to do that once when you create the post”? But apparently that’s not common practice?

Meaning that a link post automatically extracts some meta info - every time it’s clicked, not only when it’s created. Because it appears that Lemmy doesn’t automatically do it once at posting time.

Examples:

https://piefed.social/c/mediatheque@feddit.org/p/2446966/la-la-land
https://piefed.zip/c/netzkultur@feddit.org/p/1874208/we-noticed-that-you-are-blocking-our-ads

I only see a link. When I view the original post on a Lemmy-based instance, I see an automatically, real-time extracted blurb.

Is pyfedi’s stance something like “it should be enough to do that once when you create the post”? But apparently that’s not common practice?

0

Failed attempt to share post on Mastodon

preview

Attempted to share a post via Mastodon like this:

image

But I got an invalid link and a wrong text:

image

Clarification: The post I wanted to share is not about mental health and the link is broken.

Attempted to share a post via Mastodon like this:

image

But I got an invalid link and a wrong text:

image

Clarification: The post I wanted to share is not about mental health and the link is broken.

0

How should PieFed handle accounts that mix human activity with automated reposting?

preview

EDIT: Thanks for your thoughts everyone. Lots of support for simply banning or flagging as a bot, and I’m going to add functionality for flagging as a reposter, then letting people filter on it. Instance admins can apply whatever response they prefer.


Recently there have started to be a few accounts that seem to be used by a person for writing comments while also being used by scripts to automatically repost content from Reddit and possibly elsewhere. This doesn’t cleanly fit into our existing tools for classifying accounts and posts - a conventional bot might post RSS articles or other automated content, and users can choose to filter based on that bot setting. These new types of accounts are different from that because they’re partly human and partly automated.

This causes a few problems:

  • there are cases where the reposted content is written in the first person. Like for example, a Reddit post might say here’s a photo of my cat, but when that post is automatically reposted into the Threadiverse by a reposter script, it can appear to us that the PyFed account owner is claiming the cat as their own. Yet that’s obviously not the case, which gives an inauthentic feel to the environment.

  • sometimes some of the material being reposted is content that violates the rules and norms of Threadiverse. And so we see a lot of strange culture war stuff and racist content coming through sometimes.

  • a lot of people left reddit to get away from all the astroturfing and bots and LLMs running amok so letting that happen here too will just drive those people away.

Despite this, their content is often quite popular and many people upvote it and comment on it. Many people seem totally ok with it.

Before I change anything, it’d be really good to understand what the community wants to happen in response to this. I’m considering several possible approaches, including doing nothing, or introducing a new automated / reposted indicator for accounts (with a feed filter, separate from the bot indicator and filter), or continuing to apply the existing bot label onto those accounts, or restricting those kind of accounts completely. But I’m not going to propose a particular outcome yet. I’d like to hear what, if any, problems people are experiencing from these accounts and what information or controls would be useful.

EDIT: Thanks for your thoughts everyone. Lots of support for simply banning or flagging as a bot, and I’m going to add functionality for flagging as a reposter, then letting people filter on it. Instance admins can apply whatever response they prefer.


Recently there have started to be a few accounts that seem to be used by a person for writing comments while also being used by scripts to automatically repost content from Reddit and possibly elsewhere. This doesn’t cleanly fit into our existing tools for classifying accounts and posts - a conventional bot might post RSS articles or other automated content, and users can choose to filter based on that bot setting. These new types of accounts are different from that because they’re partly human and partly automated.

This causes a few problems:

  • there are cases where the reposted content is written in the first person. Like for example, a Reddit post might say here’s a photo of my cat, but when that post is automatically reposted into the Threadiverse by a reposter script, it can appear to us that the PyFed account owner is claiming the cat as their own. Yet that’s obviously not the case, which gives an inauthentic feel to the environment.

  • sometimes some of the material being reposted is content that violates the rules and norms of Threadiverse. And so we see a lot of strange culture war stuff and racist content coming through sometimes.

  • a lot of people left reddit to get away from all the astroturfing and bots and LLMs running amok so letting that happen here too will just drive those people away.

Despite this, their content is often quite popular and many people upvote it and comment on it. Many people seem totally ok with it.

Before I change anything, it’d be really good to understand what the community wants to happen in response to this. I’m considering several possible approaches, including doing nothing, or introducing a new automated / reposted indicator for accounts (with a feed filter, separate from the bot indicator and filter), or continuing to apply the existing bot label onto those accounts, or restricting those kind of accounts completely. But I’m not going to propose a particular outcome yet. I’d like to hear what, if any, problems people are experiencing from these accounts and what information or controls would be useful.

0

On AI-generated articles

preview

I thought of sharing some issues that preoccupy me here, because of how active piefed development is. Btw thank you all for your hard work!

I have noticed that more and more articles are created with LLMs without disclosing it. So, it seems to me that if someone wants to avoid posting this sort of content, one needs to at least:

  • check how many articles the author posts per day in the specific site,
  • then if the author really exist and
  • finally copy-paste part of the text in a couple of ai-detector sites.

Initially, I thought of making a post for a feature request like the one that detects AI generated images, but for text. But I can’t because if I got this right, the ai-detectors may flag an article as ai generated when the author is not a native english speaker [1].

Of course the feature that allows us to label ai-generated content ourselves before posting it is very important. In a way my questions are related to something else: what to do before that. As briefly as I can, here they are:

  • does the process mentioned above seem adequate?
  • if yes, what else can someone do to check an article before posting, and are there any alternatives/variations to this process?
  • if no, what would you suggest?

I thought of sharing some issues that preoccupy me here, because of how active piefed development is. Btw thank you all for your hard work!

I have noticed that more and more articles are created with LLMs without disclosing it. So, it seems to me that if someone wants to avoid posting this sort of content, one needs to at least:

  • check how many articles the author posts per day in the specific site,
  • then if the author really exist and
  • finally copy-paste part of the text in a couple of ai-detector sites.

Initially, I thought of making a post for a feature request like the one that detects AI generated images, but for text. But I can’t because if I got this right, the ai-detectors may flag an article as ai generated when the author is not a native english speaker [1].

Of course the feature that allows us to label ai-generated content ourselves before posting it is very important. In a way my questions are related to something else: what to do before that. As briefly as I can, here they are:

  • does the process mentioned above seem adequate?
  • if yes, what else can someone do to check an article before posting, and are there any alternatives/variations to this process?
  • if no, what would you suggest?
0

Post ranking

preview

How piefed currently ranks

I took a look into post.post_ranking which is an implementation of what has been described by Amir Salihefendic in 2015.

if post_date is None:  
    post_date = datetime.utcnow()  
if score is None:  
    score = 1  
order = math.log(max(abs(score), 1), 10)  
sign = 1 if score > 0 else -1 if score < 0 else 0  
seconds = self.epoch_seconds(post_date) - 1685766018  
return round(sign * order + seconds / 45000, 7)  

It looks esotheric to me.

  1. The log(… score) is quickly explained by the recursion of higher visibility through higher scores. The higher the score, the more views and more votes. That’s fine with me. A lot depends on the first votes, I can’t argue with that. But why they picked the base of 10 is left unexplained.
  2. The value 1685766018 refers to the unix date 2023-01-03T14:07:42. It looks like a magic/random date to me. However that date seems to be unimportant later on.
  3. The divisor of 45000 seconds is equivalent to 12.5 hours, 1 half-day. Through dividing the post age by 12.5 hours the age is basically converted from seconds to half-days. Again, unexplained why 12.5 h.
  4. In the end, we add up log(score, 10) + age in half days. I don’t get why they add them up.

Say, we choose a different log base like 2 instead of 10. How would the age measurement need to be adapted in order to get the same post_ranking results? All I know is it is unimportant. We just have to adapt the 45000 s to be 13546 s.

If we’d change the magic date we would need to adapt the arbitrary 45000 seconds again. (Wouldn’t change the rank, I suspect.)


Different approaches / statements

Here, something slightly different is claimed to be the reddit formula: Here, the sign decides if time is good or bad, not the log score.

score = log_10(|score|) + sign(score) * seconds / 45000  

Lobsters has this: (I removed a hotness-bonus, a comment score and several plus-ones.)

score = sign(score) * log(|score| + seconds / 81000)  

They call the 81000 s (22.5 hours) the “hotness window”. That article features an interesting animation of three kinds of posts: viral hit, sleeper hit, steady performer.

Evan Miller explored the reddit formula and boiled it down to:

ln( score ) + factor * seconds  

Seeing similarites to Bayesian beta distribution in a poisson process, Miller concludes:

I realize that proposing any change to how Reddit works is one of the Internet’s most dangerous games, so I hesitate to beat the drum in favor of MillerSort™. But I believe that expected-utility theory and a simple random-reload model can help explain why the Reddit formula has been so effective in the past, and shine a light on aspects that might be improved. In particular, the Reddit formula should probably take into account the percent of votes that are positive, rather than just taking the difference between positive and negative votes.

By the way, the original seems to be over 12 years old: https://github.com/reddit-archive/reddit/commit/50d35de04b928836b7ee955c8a26f197e24ab01e That commit explains the diversity of where to put the sign function.


Say, I’d suggest a different, more physical ranking function. How would we test it? Does somebody have statistics?


Edit: Correction concerning the choice of the log base. Precision on point 4. Guess about having to change 45000 for a different magic date.
Edit 2: Moved my comment beneath a new headline in my OP.
Edit3: Added more different approaches / statements. Consistent naming.
Edit4: Fixed an error in Lobsters’ function.

How piefed currently ranks

I took a look into post.post_ranking which is an implementation of what has been described by Amir Salihefendic in 2015.

if post_date is None:  
    post_date = datetime.utcnow()  
if score is None:  
    score = 1  
order = math.log(max(abs(score), 1), 10)  
sign = 1 if score > 0 else -1 if score < 0 else 0  
seconds = self.epoch_seconds(post_date) - 1685766018  
return round(sign * order + seconds / 45000, 7)  

It looks esotheric to me.

  1. The log(… score) is quickly explained by the recursion of higher visibility through higher scores. The higher the score, the more views and more votes. That’s fine with me. A lot depends on the first votes, I can’t argue with that. But why they picked the base of 10 is left unexplained.
  2. The value 1685766018 refers to the unix date 2023-01-03T14:07:42. It looks like a magic/random date to me. However that date seems to be unimportant later on.
  3. The divisor of 45000 seconds is equivalent to 12.5 hours, 1 half-day. Through dividing the post age by 12.5 hours the age is basically converted from seconds to half-days. Again, unexplained why 12.5 h.
  4. In the end, we add up log(score, 10) + age in half days. I don’t get why they add them up.

Say, we choose a different log base like 2 instead of 10. How would the age measurement need to be adapted in order to get the same post_ranking results? All I know is it is unimportant. We just have to adapt the 45000 s to be 13546 s.

If we’d change the magic date we would need to adapt the arbitrary 45000 seconds again. (Wouldn’t change the rank, I suspect.)


Different approaches / statements

Here, something slightly different is claimed to be the reddit formula: Here, the sign decides if time is good or bad, not the log score.

score = log_10(|score|) + sign(score) * seconds / 45000  

Lobsters has this: (I removed a hotness-bonus, a comment score and several plus-ones.)

score = sign(score) * log(|score| + seconds / 81000)  

They call the 81000 s (22.5 hours) the “hotness window”. That article features an interesting animation of three kinds of posts: viral hit, sleeper hit, steady performer.

Evan Miller explored the reddit formula and boiled it down to:

ln( score ) + factor * seconds  

Seeing similarites to Bayesian beta distribution in a poisson process, Miller concludes:

I realize that proposing any change to how Reddit works is one of the Internet’s most dangerous games, so I hesitate to beat the drum in favor of MillerSort™. But I believe that expected-utility theory and a simple random-reload model can help explain why the Reddit formula has been so effective in the past, and shine a light on aspects that might be improved. In particular, the Reddit formula should probably take into account the percent of votes that are positive, rather than just taking the difference between positive and negative votes.

By the way, the original seems to be over 12 years old: https://github.com/reddit-archive/reddit/commit/50d35de04b928836b7ee955c8a26f197e24ab01e That commit explains the diversity of where to put the sign function.


Say, I’d suggest a different, more physical ranking function. How would we test it? Does somebody have statistics?


Edit: Correction concerning the choice of the log base. Precision on point 4. Guess about having to change 45000 for a different magic date.
Edit 2: Moved my comment beneath a new headline in my OP.
Edit3: Added more different approaches / statements. Consistent naming.
Edit4: Fixed an error in Lobsters’ function.

0

Missing thumbnail for YouTube videos

preview

When I post a YouTube link, no thumbnail is generated with the post.

I see this issue on Codeberg as #269. It was created in 2024, and closed the same year. But was anything ever changed relating to video thumbnails in Piefed?

When I post a YouTube link, no thumbnail is generated with the post.

I see this issue on Codeberg as #269. It was created in 2024, and closed the same year. But was anything ever changed relating to video thumbnails in Piefed?