It's that time again...

[Koobie]

it all started with https://www.massassi.net/levels/files/2177.shtml
thank you so much Brian for creating the best modding community in the world

[Koobie]

Hey Brian
I rewrote Massassi in Python
Take that, Elon Musk

[Koobie]

I hacked you through your foot! lolkopter

[Koobie]

Мы покорим звёзды

[Nikumubeki]

In any case, whoever's gonna take over the site should finally overthrow JKE from the permanent LOTW position with Backup Base

[SMLiberator]

what do you mean it's always been backup base

[Brian]

From another thread re: jkhub:

So awesome!!!!!!!!!!! We should definitely figure out a way to get regular backups of this stuff. I'm working on a rewrite of Massassi (almost done) and I think I'm going to put all the file downloads, screenshots, programs, etc., up on S3 in a public-readable bucket. That way even if Massassi ever goes down at least people will have had a chance to mirror/archive it (regularly, if desired).

I enabled s3 as storage backend in my new django-based massassi app and the whole thing slows to a crawl It took over 13 hours to upload the ~8GB of files up to linode's s3-compatible object storage. And maybe it's because it's hosted in NJ and I'm closer to the west coast, but loading pages that have images loading from linode's object storage is too slow as well. I guess I could experiment with actual s3 but I've previously had good luck with linode. And amazon's interfaces for this stuff are so odd, disjointed, confusing, and needlessly complex.

Does anyone know if there's a way to make a public s3 bucket "browseable" via a web browser like old apache index pages? Right now the bucket is public but it just returns xml when you access it, you have to know the full resource name to get the actual files.

Maybe I need to consider linode's block storage instead.

[Nikumubeki]

You know, removing the TODOA TC downloads (and the unlinked original TODOA download lingering in the directory) would reduce that filesize