Comment

Comment on alien.top is a new level of Reddit crossposting spam

The database of “1M comments” is taking less than 10GB of disk space. Looking at the last backup, the whole database uncompressed is 18GB. It’s running on commodity hardware. Even with the mirrors making copies of the images to object storage, my object storage bill this month was a whooping $0.66.

I guess if you just link the images from Reddit it’s not that computationally intensive. I very much doubt that Reddit is going to let this slide if Lemmy ever gets that big though.

Why shouldn’t at least try to do it?

Because there are things to lose, and this isn’t a risk-free process. I expanded more on my reasoning in my last paragraph:

If this bridging was an opt-in system, I’d be fine with it. But because it’s currently an opt-out system, and an opt-out system where you have to block hundreds of accounts, I really don’t like it. Perhaps a system to make these opt-in, like a menu in the settings to select which bridges you want enabled could be added to Lemmy, and I’d be fine with these mirror/bridge bots then. This is sort of like how it works on Matrix, and I like the bridging there. But with the current circumstances on Lemmy, I don’t like the mirror/bridge bots.

source

Sort:hotnew top

rglullis@communick.news ⁨1⁩ ⁨year⁩ ago
I guess if you just link the images from Reddit it’s not that computationally intensive

The images are actually copied to the mirrored server.

Perhaps a system to make these opt-in, like a menu in the settings to select which bridges you want enabled could be added to Lemmy,

It’s not that simple to do that per user. You’d need:

An actual Reddit client per user

A Lemmy client with OAuth support so that the bridges don’t need to hold the user’s password.

An “official” map of reddit-to-lemmy communities, so that we know where to point all those bridges for posts. I’m working on such map, but I really don’t want to call it official unless it gets significant community support.

Is the opt-out solution aggressive? Yes, no doubt. But I thought that this “aggression” was pointed to Reddit and therefore justifiable. The whole reason that this approach forces its hand to be able to get the data is because Reddit API changes was a clear sign that they want to treat the data from the users as their own. The protests were not effective against this, and showed to Reddit that they can win any conflict against dissenting mods. If Reddit tracked back on their policies and showed to be a good steward of one of the most vast amount of user data, I wouldn’t be putting so much effort in this project.

If you can think of any other approach to make this work and is aligned with the clear goal of the project (make it easy for people to migrate away from Reddit, in a way that those that come here can already find their niche communities) I’m all for trying it.
source
- Jumuta@sh.itjust.works ⁨1⁩ ⁨year⁩ ago
  
  The images are actually copied to the mirrored server.
  
  That’s really interesting, but why do you do that? Surely having the clients fetch the data from Reddit’s servers themselves would be easier?
  
  But I thought that this “aggression” was pointed to Reddit and therefore justifiable.
  
  I hate Reddit as a platform too, but I very much disagree with this philosophically. I don’t break the rules against the enemy because then the enemy would be allowed to break the rules against me. If we want to grow as a platform, we have to stay civilized. The one that fails to do that dies.
  
  If you can think of any other approach to make this work and is aligned with the clear goal of the project (make it easy for people to migrate away from Reddit, in a way that those that come here can already find their niche communities) I’m all for trying it.
  
  I think you misunderstood my idea about opt-out bridges. I meant that there should be a toggle for Lemmy users on Lemmy which mirrored/bridged content should be shown to them. These should be off by default, but easily changeable.
  
  source
  - rglullis@communick.news ⁨1⁩ ⁨year⁩ ago
    
    Surely having the clients fetch the data from Reddit’s servers themselves would be easier?
    
    Easier? Yes. Reasonable? Not at all. Reddit wants to control all the data, the whole API fiasco started because they started to abuse their power, do you think they can trusted of stewards of social media data?
    
    I don’t break the rules against the enemy because then the enemy would be allowed to break the rules against me.
    
    This is a fight, not a game. There are no rules. Do you think they care about rules when they started forcing moderators out of the protesting subs? Or lying about what Christian was asking during pricing negotiations? Or when they get mods working for them to do their bidding?
    
    Let’s not be naive. They will leverage anything they have to get the upper hand. We are not going to win anything by pretending there is a higher moral ground to stand on.
    
    source
    Jumuta@sh.itjust.works ⁨1⁩ ⁨year⁩ ago
    
    We are not going to win anything by pretending there is a higher moral ground to stand on.
    
    The moral high ground is the ONLY thing we have. Lemmy as a platform exists to be a non-evil counterpart to Reddit. It would have no purpose to exist were it not for our better ethics.
    
    source
    -> View More Comments