Thread:Crazy teitoku/@comment-1915363-20151024174001/@comment-26091666-20151025081159

Ar-cen-ciel wrote: Currently planned features are the option to specify which game assets to download (game CG, BGM, and other stuffs), though I'll have to find a way to dynamically generate links via Javascript for that. As you can see in the repo though, all the server HTML pages are virtually identical with different server IP addresses. It's gonna be a living hell to update them all as it is.

Another thing I might consider is to limit the max number of concurrent connections to 5 via controlled Javascript requests, but again, it needs dynamic links, not to mention I didn't have the know-how to do so.

Ah yeah for now manually updating the links would be somewhat a pain but most text editors should be able to search and replace strings of text from one to another. I know for *nix there's things like sed which can be used to somewhat partially automate the search and replacement of specific strings (it also does regexp).

That with maybe some script that will do it for each file with specified variable for each of the file, i.e. define the IP address for each server as a variable and then use something like sed to search and replace entries of a specified IP with the variable defined.

Either that, or maybe create a template and have the script generate each file dynamically. Not sure if that is a solution you would be wanting to look into however.

Ar-cen-ciel wrote: Might be possible in KC3Kai, but I couldn't think of any means to retrieve the referrer.

Oh I meant for that if it is possible to set referrers in the iframe then that is half the problem eliminated. The other is to maybe set the files up to have referrers automatically set in place which shouldn't be too hard to obtain imo with Chrome. It just means more work and may probably not be needed at this stage. Just a thought however.

Ar-cen-ciel wrote: These links you provided are rather transient assets actually. These files tend to get changed everytime a new update arrives, so the only time these files would be requested is when you start from a clean cache, and even then, it's better off to just explore the game around to recache them to get the latest version number. That being said, I already have a super-prototype version sitting in my hard drive already. It's a 200KB HTML page.

Ah I see, yeah that would potentially require it to be more adaptive which is yet another thing to consider about.