Yeah Jumbofiles is pretty Good for free users. I will really recomend it too.

Take a look at this then, good for free users.

http://jumbofiles.com/

Funnily enough, they seem to be erroring out nowadays...
Urgh, okay, I take your point. I don't really want to be using too much more bandwidth but I'll see if I should get rid of some other service (like UploadMirrors).
Any votes on a single file host?

Having a MF folder actually would be the preferred way for people using software like jDownloader since it allows you to download all the files automatically by copying only one link instead of multiple.
I haven't suggested that before only because I thought it's not possible with anonymous uploads which does reduce the risk of having all the files deleted at once. Maybe if the backend could automatically create a new account e.g. every day, that could solve the problem but it does seem like too much work.

When it works, uploadmirrors is good. But looking at the front page not many of the latest, uploadmirrors have not yet mirrored many of the additions even after some hours.

That's probably only a temporary thing with MirrorStack being broken these days.
It's already been disabled, but what about Bayfiles vs UploadMirrors?

Thanks for your hard work.
But I have to say in defense of Bayfiles. Looking at the homepage, the only mirror that every file has seems to be embedupload. The others are mirroring very late. When Bayfiles was around, it used to mirror pretty fast. So we had an alternative to embedupload.

Just the awesomeness of the awesomeness! You made it simple and easy for us, the selfish, ignorant and fast consuming internet's users. But no one knows in real how much hard working is behind all of this!
Thank you for everything!

For point 1, I'm a bit apprehensive about modifying original filenames - my preference is to keep it the way that the original uploader wanted it. Though you did mention "show", so maybe I could make the CRC a little more visible... (if you were already unaware, you can click on a file to see its CRC, as well as some other hashes)

I've generally avoided archives for convenience sake, but I guess mangled filenames are a bit problematic too, so that's not a bad idea. There's a bit of technical difficulty with trying to stream the archive process in the way I'm doing it, but I'll consider it.

Thanks for the suggestions :)

It's nothing more than a filename match ;) Similar concept to how NZBs are matched.
(so it's possible the files don't actually match, but I presume that it usually should)

It would be nice if you could show / add the CRC32-checksum to files where it isn't included into the filename. Using ZIP/RAR for filehosters which cripple the original filename (for example '_' instead of '[') would also be nice to have (and maybe adding a password).

Anyways: thanks a lot for creating this site here and keep on being awesome =3