[mythtv-users] Mythweb finds no remote controllable frontends

[Michael T. Dean]

On 07/21/2011 08:12 AM, E. Westbrook wrote:
> On Wed, Jul 20, 2011 at 09:55, Michael T. Dean wrote:
>> This is why we generally try to use the in-database information to find
>> hosts, rather than just assume that every network has name resolution
>> properly configured.
> Many (I daresay most), don't have local resolution at all, so the need is
> certainly there.
>
> However, I hope we can be sure not to cause problems for those that do
> properly set up their networks, and for whom the infrastructure should be
> relied on and used as 100% authoritative in these matters.

No, MythTV database configuration is authoritative.  If you specify 
invalid IP addresses in your MythTV configuration, we will use them and 
we will fail until you correct your configuration.

> For example, by necessity, I use the "hostname" key to identify different
> groups of settings for different purposes.

And you would  only have problems if you specified invalid IP addresses 
for some of those LocalHostName overrides.

>    This always troubles me, but
> it's simply how it's done, apparently, to have named groups of settings in
> MythTV for different purposes.  I haven't encountered a place where that
> misnomer breaks anything, but I worry.  And it seems to me that this is
> where it could likely surface.
>

It's the same code we have scattered all around MythWeb.  Nothing would 
be different--this is just one location where such code was never 
added--likely because anyone who has ever used the MythWeb remote page 
until now has functional name resolution on their network.

If anyone is really that offended by the hack, you'll need to do the 
full refactoring of MythTV connection handling and tracking that Raymond 
mentioned so that you can remove all the existing MythWeb (and MythTV) 
code that uses the hack.

Mike