The only way I can think of is if there's a DLNA/UPnP command we could send to a renderer (if it were detected that is) that would result in a predictable reply that we could use to determine that traffic was being blocked if it didn't arrive. What do you guys It's not easy to check for incoming access to the computer you're actually on. I don't know if the install processes on the other OS'es would allow something like that, Linux is hopeless in that there's no standard for anything, and OS X is just foreign to me. We should also offer to open the necessary ports at the very least on Windows firewall, from the installer. I don't know if there are any such possibilities in DLNA or uPnP though. What would be really nice is if it would be possible to contact a renderer and request for it to contact UMS back (instead of replying in the estabilished channel) and in that way be able to test that it could reach UMS. It shouldn't be to hard for us to figure out what was needed though (isn't UDP port 5000 involved?), and we should probably show this information in the GUI dynamicly so that it reflects the current configuration. HTTP wise that's pretty simple, just two TCP ports (one for the DLNA server and one for the web interface)? On the uPnP/UDP side I really don't know what is needed, uPnP is a messy standard network wise with it's broadcasting and whatnot. The very least we should do is to inform the users of what incoming ports need to be open for UMS to work. I do think that more could be done though. Most OS'es these days also comes with firewalls enabled by default, and not-so-technical users will be scared to turn them off even if they are protected by a router and NAT. It can be quite tricky to deal with, as there are a number of different software firewalls available and it's hard to test incoming access without cooperating with an external host. Seems to me like UMS isn't taking care of firewalls and simply assumes that everybody will turn them off.
0 Comments
Leave a Reply. |