Hi Folks,
I'm a newbie with Roundcube, but I'm a very experienced system administrator, so big-picture sorts of replies are great, but "the devil is in the details," of course!
The Goal: Provide existing users an alternative to accessing their email accounts via mail applications such as Thunderbird or (-gack!-) Outlook.
Existing Email Config: Pure port-based, always encrypted access through a firewall / gateway via port-forwarding to well protected "internal" servers.
Existing Web Config: Standard Apache on an "external" box on the internet that's also connected to an internal net. The occasional need to get to internal systems, for services such as email list services such as MailMan are performed via standard mechanisms like pipermail.
Existing Database Config: Postgres on an internal server, accessed via internal net using Postgres' access security mechanisms.
Questions:
I presume that this is not a big deal to configure by installing Roundcube on a web server box and then accessing the internal servers via the internal net. That is, I presume Roundcube doesn't want / require direct access to mail folders but can use established servers like Dovecot for this kind of access. If that's a problem, I'd like to know before I get deep into this!
Does Roundcube have knowledge of or "care" about the URL that gets the client there? That is, can I provide individualized "branding" for the various domains that are served by the environment? (This computing environment serves some 50 to 60 domains, variously .coms and .orgs, etc.) It would be nice if, at the very least, Roundcube doesn't require some "hard coded" configuration information that locks it to a domain, or appear to a clueless user as if they're "talking" with a server from "some other company," or otherwise distress them.
Future Goal: We want to start offering "virtual email accounts" sometime soon. Are these going to be an issue with Roundcube?