In the short time since this instance and bot launched, I’ve been seeing the same questions resurface multiple times. This is totally understandable, since the concept of a Fediverse is still new to most (myself included), and this server is not like the others.

Q: What is Lemmit?

A: Lemmit is a Lemmy instance specifically designed for archiving Reddit content. Users can request new subreddits to be included in the archiving process by posting in the [email protected] community. It is powered by an open source python bot, which periodically checks the request list, adds new requests to the queue, and continuously monitors the Hot feed of those subs for new posts to cross-post here.

Q: Does it synchronize comments?

A: No, that would be impossible. Considering there are thousands of posts already on Lemmit, many of them having at least several hundred comments on Reddit, often buried in deep layers, it simply wouldn’t be feasible to index those for more than a few posts, let alone keep them up to date.

Unfortunately, this means that archiving certain subreddits, such as Ask Historians/Men/Women/Hyperintelligentshadesofthecolourblue-type subs, is going to be rather pointless.

Q: Can it send comments back to Reddit?

A: No, it cannot. The purpose is to help bootstrap the Lemmy platform, not to serve as a bridge between the two networks. Also, see the answer about synchronizing comments.

Q: Can I request any subreddit?

A: Technically, yes. However, as the list of subs grows, the time it takes to update all of them will also increase. I do not have strict guidelines in place for this, so I’m relying on your common sense (hoooo boy). At some point, I will probably have to either stop accepting new requests or disable scraping for very low-traffic communities. Nope. Automatic requests for new subreddits are closed. You can send me a PM if you think a specific subreddit would be a valuable addition to the Lemmyverse, but you’ll have to make a strong case.

Q: Does this use the API? Will it keep working after July 1st?

A: Nope, it uses a combination of the public feed and scraping old.reddit.com. So, as long as those are still available, it will continue working. And even if they close those sources, there will probably be new ways to achieve the same effect. “Content, eh, finds a way.”

Q: This is spam, can you stop?

A: First of all, I apologise for the inconvenience. All you have to do is block @[email protected], and none of its posts will ever show up on your instance. If you you don’t want anyone else on your server to be exposed to this bot/instance, you should convince your admin to defederate from lemmit.online. Since there are no other users on here, there will be no harm done.

Obviously I could stop, because running this server and software is only ever going to cost me time and money. But for the reasons listed above, I still think this server is a useful addition to the lemmyverse at this time. But I’m looking forward to the day where I can turn the bot off because it’s no longer needed.

Q: What started this?

A: Okay, nobody asked this, but I’m going to tell you anyway. After Reddit made it clear that they are effectively killing third-party apps and implementing plenty of other anti-end user decisions, I realized that I would either have to accept not being able to access my time-wasting content or have to do so in a rather uncomfortable way (either through the official app or old.reddit.com for as long as they’ll allow it to exist).

Being a stubborn developer, naturally, I chose option C: Have my own Reddit. With blackjack, and hookers. This way, I would still be able to access my beloved content without being beholden to Reddit’s mood swings and abusive relationship tendencies.

Besides that, I also know that Content is King. So I’m order to counter the network effect (No users because no content, No content because no users), I figured it would be better to have some inorganic content to bootstrap the adoption of Lemmy.

Q: Are NSFW subreddits allowed?

A: Absolutely. Like I said: Blackjack and hookers.

Q: My request isn’t picked up by the bot!

A: That isn’t a question. But yeah, the process isn’t flawless yet. I’m trying to iron out all the bugs as I encounter them. In the meantime, feel free to re-request the subreddit by making a second post. No harm done. New subreddit requests are closed for the time being.

Q: No new posts are showing up at all on Lemmit

A: If no posts are appearing on the Lemmit Frontpage (sorted by NEW), it’s possible that the bot has crashed or is stuck on something. Since no software is flawless, this sometimes happens. I usually fix this as soon as I’m aware, and I’m happy to say that these kinds of fatal errors are becoming less and less frequent. However, they may still occur, and as a human with needs of sleep and other responsibilities, I’m not always able to fix them immediately.

Q: Posts aren’t showing up on my instance, what’s up?

A: Due to the spammy nature of the bot, some server admins choose to block this server, and that is completely understandable. So first of all, make sure to check the instances link in the footer of your home server. If Lemmit is the Blocked Instances list, you’re out of luck.

When you have verified that Lemmit is not blocked on your instance, try unsubscribing, waiting a little, and then re-subscribing. That tends to fix things.

  • @adminOPMA
    link
    fedilink
    English
    11 year ago

    Not sure for which subreddit this is. But if the corresponding posts also get removed, this could be alleviated by increasing the delay/upvote threshold before something gets cross posted. I’m not in favour of implementing a imgur ban, especially not on a per-community basis. And I have no interest on writing (and paying for) an imgur-rehosting service.

    • @[email protected]
      link
      fedilink
      English
      11 year ago

      I could see myself writing a backup service for that. What kind of storage per day and bandwidth usage could I expect for such a project?

      In terms of storage I could afford something like 10TB/year, so I’d guess archiving all imgur images (perhaps with the largest at a reduced quality) should be feasible, but I am restricted on upload bandwidth to about 5MB/s (not Mb/s) I can guarantee more or less reliably.
      I see a path to gaining reliable 1GB/s upload but that would require some work (and some cost) on my part.

      Given the bandwidth restrictions how about some system to detect removed imgur images that edits the posts to only then point to my mirrored version?