Add www redirect

[redrun45]

As two separate commits, in case there's trouble getting a valid cert for www.

[notartom]

Not 100% sure, but I suspect the same can be achieved by just adding ServerAlias www.librivox.org [1] to the existing config file for librivox.org. Are you able to try it in your localdev env? We still need the other stuff though.

[1] https://httpd.apache.org/docs/2.4/mod/core.html#serveralias

[redrun45]

For *:80, we certainly could just do a ServerAlias. I tested that, it worked just fine.
For *:443, though, the web site does not behave properly when reached as www.librivox.org, so I think it does need its own VirtualHost.
I figured it would be neater to have all things www as their own file, since they wouldn't cleanly fit into (the existing) one. Does that make sense?

[notartom]

For *:443, though, the web site does not behave properly when reached as www.librivox.org, so I think it does need its own VirtualHost.

#30 seems to work fine on staging? How does it "not behave properly" for you? (Not trying to steal your thunder, just proving stuff out for myself).

[redrun45]

(Not trying to steal your thunder, just proving stuff out for myself).

Not at all! If it works, it works. I saw a SQL error styled in a way I haven't seen before, and figured I'd better keep things cleanly separate for the PR. Now I'm curious myself, I'll see if I can trigger that again!

[redrun45]

Well, if I alter my hosts file to the staging IP, then open https://www.librivox.org/ in Chromium, it... appears to work alright. Click through a few links, and you'll pretty soon be on the regular librivox.org, with no funny business I can see.

On Firefox, I have separate containers for live, dev, and staging. I pulled some proxy shenanigans to get that, and while it's worked well since I set it up, I wonder if that's what's got me seeing funny things. Being a permanent redirect, it's hard to tell for sure, but I think #30 will work. If it causes unexpected trouble, all it takes is a redirect. 😉

Edit to add: And while I haven't seen that exact error again, I'm pretty sure it was WordPress. Which, given that staging and live both have a WordPress database, it makes sense we would get an error only on localdev.

[notartom]

Replaced with #30