I note the expense of hosting and wonder why tixati does not offer downloads of new releases via torrents

Sounds like a good idea. I would also suggest a better and smoother update system so we can update the client without having to go to the website.

It would set a bad precedent to have people download their releases via P2P methods.

Next thing you know, you've got jokers in the channels passing around magnet links to "new releases" that aren't what they claim to be, and some users _will_ be caught off-guard since it is no longer an unusual practice to get Tixati via a torrent.  Yes, of course most users are savvy and will probably not be fooled, but some definitely would.

And since bandwidth these days is dirt-cheap, there is very little benefit in exchange for the significant security risk to such a scheme.  It's not the bandwidth that makes running a web site expensive any more, it's the hardware, the time, and the expertise to setup and run the whole thing smoothly and securely.

As for in-app updates, this is always a majorly bad idea outside of operating-system package managers.  If it isn't implemented perfectly, with a very solid signature scheme, it's a pipeline into the user's machine for running an attacker's executable code.  Bespoke in-app schemes should generally not be trusted, I have seen many insecure implementations (Napster had a trivial one that was exploited), and have only seen just a few trustworthy ones, all of which came from major operators like Mozilla, goog, MS.

Why not make it into a built-in feature for current users? Hear me out. You could even make it completely automatic (and optional of course). In the end, it could benefit the users through increased ease of use and the developers through decreased bandwidth costs, while de-centralizing further the design of Tixati. You know, like this:

1. Devs release a new version and seed it through DHT, even possibly by Tixati itself. (While also releasing it on the Tixati website for regular download of course.)
2. End-user Tixatis detect a new version has been released via whatever current method for such detection is already in use. (great feature by the way)
3. Tixati downloads the hash tag for the new version from tixati.com (or through the update detection itself) and initiates DHT download/download-upload (see below) either automatically or by prompting the user.
4. With the download complete, Tixati goes through an automated process of shutting down, installing and re-initializing.
5. Tixati seeds the new version for other users via DHT. (see below)

Additional ideas (important):
1. Users would be able to set whether they actually seed the new version for other users and the maximum seed ratio (By default this would be set to zero of course, with no upload activity - many people have limited upload bandwidth, would be unwilling to seed anyway, etc. Seeding the new version would be a new way to help the devs, for users who would be willing to and capable of doing so. My guess is that there would be enough such users to ensure good download speeds. Tixati is really light anyway, and user updates could be queued.)
2. End-user seeding could be smartly implemented, with upload activity only occuring during low bandwidth use.
3. To make it very secure, new versions could have encrypted confirmation keys for the confirmation of authenticity of (future) updates. This would mean that hackers would need to gain control of Tixati servers and also the physical confirmation keys to distribute unauthentic, malicious code through the automatic update system, which is unlikely in any case.
4. This is a bit off-topic but a Tixati free of the re-booting process would be amazing, especially if coupled with the automated update system I (attempted to) have described. Totally hassle-free, lightning fast, de-centralized updating. Probably hard to implement, if not impossible. Just thinking.

As one poster above mentioned, bandwidth these days is cheap with costs falling, nevertheless I believe this feature could mean progressive thinking in software design and increase the ease of use of Tixati, eliminating the need for manual updates. However cost-effective anyway this all would be to implement is completely beyond my knowledge, I am just a janitor in real life, heh.

I would toss it in my seedbox for sure

Hi
wow amazing ideas, I doubt the implementation,
but this would be a great Evolution of the update system,
unique system in the genre, I love it,
this kind of ideas gives the best of world.

with respect.

This would work perfectly as long as the tixati website giving the infohash isn't compromised. The only downside I can think of is this will put tixati much higher on the radar with the copyright trolls, with there being an infohash that can be used to track tixati users/clients that share and download and use the collected IP's for further correlation. In essence a updated version downloaded while the VPN is off by accident could end up giving enough data to the peer tracking trolls to correlate user IP's to VPN IP's. This method makes it much too easy for the trolls to monitor exactly who's using tixati, and if client/nodeID, etc. matched across 2 IP's, etc. you see where this goes. In time they could even send fake/troll 'takedown' requests to the tixati website for distributing 'piracy software' (in future), etc.

Hi,
(Personally:
1: automatic changing of the node ID every day.
2: I use a VPN pass through (including kill switch protection) (SHA-256, AES256, RSA2048))

I do not think there is truly more risk for me to download it via eg: the Tixati channel group compared to the direct download on the website, It will ask an improvement to notifications of channels to be savvy, download it, and set seed in conventional manner.
a decentrally distribution, no saturation of the servers, a seductive idea...
from france with love.
Mohadib.