• 1 Post
  • 21 Comments
Joined 2 months ago
cake
Cake day: October 12th, 2024

help-circle



  • Thank you for responding. Here are my plans

    1. After some discussion with another fediverse developer, he recommended we move to sublinks library. I posted our tech plans here https://lemmy.world/comment/12922172. This will achieve a number of things - move db to postgres, deployment to docker/k8s, enable lemmy clients, make some security changes so our passwords are not exposed, this in turn will enable open sourcing and self hosting. This seems the best path forward.
    2. We almost completed the move when we found out that sublinks library itself does not have federation implemented. I was told it will be picked up in 2025 but it is also being developer by volunteers, so the timeline is not certain. Since we almost finished move to sublinks, as soon as they have federation, we should be able to move very quickly since work on our side is mostly done.

    There is almost no traffic today, users are not missing out on any content. Since the timelines are not in my hands, my ask is for admins to give me benefit of doubt and be patient until I wait for sublinks federation implementation (or if clubsall have traffic in which case, users will be missing out on content. In that case, I will have to think of something else)

    1. In the long run, idea is to have an simplified fediverse frontend that can realistically be a real open alternative to reddit.

    Feel free to ask me anything else.



  • Thank you for considering and responding. While I cannot change your mind, I can explain some of the strikes

    1. This is not monetized platform, there is no way to pay me, no advertisements, no premium plans etc. I am guessing you are saying since it is not open source yet, it is monetized. That will change, we started working on changing architecture to use sublinks, but got stuck as sublinks itself is under development.
    2. As soon as sublinks federation works, we will be able to adjust quickly. Without this I have only two choices - spend a lot of money or shut it down. I would like to avoid both if possible. Since there is no traffic, I am requesting the community to give us benefit of doubt.
    3. The developer used various subdomains while trying to implement federation. As far as I know, the final subdomain was api.clubsall.com. My guess is that admins are blocking clubsall.com and not api.clubsall.com. If another subdomain is still active, please let me know.



  • I just checked, the homepage of Clubsall shows content from !196@lemmy.blahaj.zone, while lemmy.blahaj.zone have you defederated: https://lemmy.blahaj.zone/instances, as “clubsall.com” is in the blocked instances. But if we look at the linked instances, there is now “api.clubsall.com” and “clubsall-api4.renchesterjramos.workers.dev”

    This is because federation is fairly complex and we are new to this and do not fully understand. I really do not know why we are getting content from lemmy.blahaj.zone when they have blocked us. The content is coming from federation, so how is it being pushed to clubsall after blocking?

    You are channeling the traffic from 44k monthly active users to your website without giving credit to the instances where the content is created. In our last discussion, you said “your site is small, so people just ignore it. Should it become more active, then users are probably going to call their admins to defederate.” ClubsAll has not grown. This time you are saying we are not attributing. It seems even you change mind on to what is fair.

    My request stays the same, give us some breathing room until some traffic threshold. Is that fair?


  • Thanks for this. My worry is that other admins have different set of rules. Someone said clubsall should be open sourced, someone said clubsall should be self hostable, another person said, it should attribute posts and have canonical links. Unfortunately the admins do not seem to have common agreed set of rules that are fair. So no matter what we do, it seems we will be blocked. Can we have a viable alternative?

    There was a discussion about having a common set of rules which had proposal which is different than yours. So I am requesting some breathing room exception based on traffic, so we can survive until we can develop more or there is a common set of rules agreed to. Would that be fair?








  • Hello Blaze,

    One request, whenever you post about clubsall, can you please tag me so I can also participate? Thank you.

    As for this discussion, when we had a discussion last time on reddit, I thought we agreed that as long as there is not meaningful traffic, it should be ok. I guess you were not fully onboard with this.

    My request with all admins is - can we agree on some rules. If we block all new sites on hunch or rules as we make up, it will kill anyone trying to do something new.

    There was another thread (that I cannot find now) where some rules were proposed. (If someone can find, please reply to this).

    There are 2 things being discussed

    1. federating out
    2. open sourcing clubsall

    I can give an update on both of these. As I mentioned previously, someone was helping me in doing some audit, implementing federation and then we could also open source. We were going to use “sublinks”, which would do a number of upgrades. So that seems the right choice moving forward. It will allow self hosting, federation, move to postgres, move to docker etc.

    However during implementation we found that sublinks itself does not implement federation out. So we were blocked by that. I checked with developer just few days ago and I was told that since all developers are volunteers and they got busy, federation was delayed. The work will pickup again in new year.

    I am not sure when they will be able to finish federation. I can quickly move there once done. However that seems months away.

    In the meantime, can we have a discussion about a set of rules for clarification and fairness?


  • Update on this request: A developer approached me and is not only helping me with review and fix security issues. We found quite concerning security issues, so I think the decision to have another person look at this was right one.

    We discussed and found that we need to do following work

    1. Redo backend/api so it is more robust, while doing that it will also become Lemmy API compatible
    2. Fix client so it adapt to any API changes
    3. Move from cloudflare workers to docker, so it can be self hosted
    4. Move from D1 to postgres (D1 has 10GB limit, ClubsAll db is already 5GB), so it is scalable
    5. Move production to a VM or k8s cluster so we can host our own DB, backend and frontend instead of CF workers

    We have some work to do but will have a good product at the end of it. We will update further once we get this work done. Thanks to everyone to your thoughts and offers to help.




  • Do you have any idea of the timeframe when you’ll be able to open source the project?

    I replied in the other thread, copy pasting here: Open sourcing is not time dependent. I just need

    1. security review by someone experienced to make sure we do not instantly get hacked as soon as we open
    2. and some commitment to fix critical bugs and hacks that will kill ClubsAll or steal resources

    what are the languages used to develop ClubsAll Ah, good question. It is typescript, next, React, Cloudflare