by ryan » Wed Jun 18, 2014 12:22 am
Hi akramer,
As you can see from the remote access dialog, the links provided are http. The remote access dialog also explicitly states that Basic auth is used. There is no attempt to obscure this, a packet scan isn't necessary to find this out.
This is no different for users who would normally forward their IP cameras out of their network. I don't know of any cameras which are https enabled by default.
There are any number of scenarios someone might use remote access for which require any number of different options and configurations. Some of those are not addressed by the current implementation, but many are. Again, there is no attempt to hide the current behavior, in fact the app explicitly calls it out.
We would certainly like to add other options to accommodate other usage scenarios in the future. These are decidedly non-trivial, and must be addressed one by one. HTTPS for example, requires a certificate as you note. This certificate must be securely signed and protected for it to be trustworthy, which means it *cannot* be distributed with the application, so we cannot pre-package a cert. If we did, it could trivially be extracted, exploited, and subsequently revoked. Could we make a “fake” self-signed cert to include? Yes, but browsers would warn (or block) you about the unknown authority unless you subsequently added the certificate to each browser and device you access from. It would be possible to leave this as an option for users to do, as a few cameras are beginning to do, but the process is highly technical and full of hiccups (getting on to all devices, having the apps recognize this as a legitimate certificate, etc..) and would not be suitable for the majority of our users.
The more traditional use of certificates and why we think of https when we think of web traffic is that the server hosting the website is typically singular (not every users machine) with only the owning company having access to the cert. This would be possible for the mobile apps, with the caveat that *all* of your traffic would be flowing through a Sighthound server, which is something many of our users would be opposed to on principal, they only want their video on their machine(s).
Another option, not much worth mentioning, would be rolling our own security layer, but you should probably be generally skeptical when anyone claims to do something like that.
Hopefully that helps clarify why the application behaves as it does now. Remote access is a very young part of Sighthound Video. It may not be optimal for everyone in the current incarnation - you might need https, others might require alternate layouts, someone else might need rule manipulation - but it certainly has not stopped evolving. We have an extensive list of features we would like to add, spanning every spectrum - functionality, optimizations, access methods, etc...
Best,
- ryan