As explained in the last post, I was trying to get s3 (R2) set up with kbin.earth, and now it's complete. Now storage costs should be significantly lower (with more capacity) and maybe images will load even faster now.
Also, think of this as an experiment; if it doesn't work out, I'll switch it back to file storage.
Sorry for the past 90 minutes of downtime; I've been trying to set up s3 storage for kbin.earth, but was unfortunately unsuccessful. I'll likely try again tomorrow (in case you notice it down again).
https://blog.rust-lang.org/2024/05/02/Rust-1.78.0.html
Empowering everyone to build reliable and efficient software.
kbin.earth is now running Mbin! Please let me know here if you run into any problems.
https://kbin.earth/m/kbinEarth/t/69127/The-future-of-kbin-earth
Hello kbin.earth users,...
Hello kbin.earth users,
I think, at this point, it's clear that Kbin is not being actively developed; the last commit by the lead developer was four months ago, and the latest commit, period, was two months ago. There are many bugs/issues present that make not only using the platform difficult but also hosting and administrating it (even worse, there was a recent issue that affected many non-Kbin servers in the fediverse).
Given the situation, I see three options for moving forward:
We could continue using Kbin as is, knowing there will be ongoing issues, and hope that maybe one day there will be an update.
Pros:
Cons:
Alternatively, we could migrate kbin.earth to Mbin, a fork of Kbin that is actually supported by its developers and are actively fixing issues and adding features. I was initially skeptical of Mbin when it first came out, but at this point, it's had many improvements that likely make it more suitable than the current Kbin system.
Pros:
Cons:
As a last resort, we could just shut down the server. kbin.earth was initially created to help develop the Interstellar app for kbin (since kbin.social didn't, and still doesn't, support the API). Still, at this point, the initial purpose has been served, and further updates to the app could be tested on an already existing Mbin server.
Pros:
Cons:
To conclude, we could either not change anything (and stay on Kbin), switch to Mbin, or shut down kbin.earth. For me personally, I think option 1 is ruled out; we've already been waiting on Kbin long enough, and I don't think much is going to change. We could potentially try option 2 (Mbin) for at least a little while and see how it goes; if people like it, we could keep kbin.earth for the time being (running Mbin); if people don't like it, we could just go to option 3 (shutdown). What do you all think? If someone has anything to add to the pros/cons, feel free to share.
https://go.dev/doc/devel/release#go1.22.minor
@jwr1
@kbin.earth