Re: [Tails-dev] GSoC 2014 - Linphone Implementation ... ?

Supprimer ce message

Répondre à ce message
Auteur: carol3
Date:  
À: intrigeri
CC: The Tails public development discussion list
Sujet: Re: [Tails-dev] GSoC 2014 - Linphone Implementation ... ?

> Note that Tails is not participating to GSoC this year, but perhaps
> someone at the Tor Project will be happy to mentor something close to
> what you're suggesting here.


ok ...

>
>> In this context I thought it could be useful to get one step further
>> with the voip topic on tails.
>
> This would certainly be most welcome!
>
>> My idea is to implement and extend linphone to get a possibilty of
>> anomynous voice calls (e.g. with a button to register oneself
>> automatically at ostel etc.).
>
> You probbly mean pseudonymous, then, not anonymous, right?


no i mean anonymous, by creating new random account on every call (in
this case we have to trust ostel.co - guardian project).

>
>> Also I am looking forward to test and fit the delays and codecs.
>
> Indeed, testing this kind of stuff in the context of Tor-ified
> connections seems very useful to me. Note that the Tor folks have
> recently tested Mumble, and documented their findings:
> https://trac.torproject.org/projects/tor/wiki/doc/TorifyHOWTO/Mumble
>
>> What do think about this idea?
>
> It's unclear to me that Linphone can work in Tails without using
> onioncat. Is your plan to use onioncat, or? More information can be
> found there:
>
> * https://labs.riseup.net/code/issues/5709
> * https://tails.boum.org/blueprint/VoIP_support/


perhaps i have a error in reasoning. why you need vpn for it? for e.g.
with linphone you can tunnel the traffic over tcp

>
> Also, I have to say the situation regarding "what's the best VoIP
> software for Tails?" is quite less clear now than it was a few years
> ago, as the landscape has changed a bit:
>
>   * the Tor folks have been experimenting with Mumble (which arguably
>     is not meant for P2P VoIP, but perhaps this can be fixed with
>     OnionCat)
>   * Jitsi entered Debian (it was previously discarded from our panel,
>     due to the difficulty of packaging it properly, but this was
>     apparently dealt with since then)
>   * much energy has been put in WebRTC, that likely won't work over
>     Tor any time soon

>
> Do you think Linphone is a better bet for Tails than Jitsi?
> If so, I'd be delighted to learn why :)


jitsi is not working without udp. and: i think zrtp is better supported
by linphone

>
>> What do you think about the implicit influence interference of
>> projects by google (also with GSoC)? For example the google partnership
>> with mozilla has the result that firefox phone depends more and more on
>> parts of android.
>
> The influence on... what exactly? :)


i don´t know gsoc ... but normally google don´t give anything without
taking influence on it

>
>> And: Why you open the bluetooth device on tails? What is the reason?
>
> Tails is a general purpose OS, and people happen to have bluetooth
> hardware they want to use.
>
>> I think it is a problem in a view of security. Similar to the
>> mac-address it can identify people (e.g. at least by fhs-packets in the
>> discoverable modus).
>
> Last time I checked, we did not ship the bluez stack. Knowing this,
> does the problem you're evoking here apply to Tails? E.g. is this
> identifying info broadcasted in current Tails?


current tails: yes i think so. but i don´t have bluetooth - so i can´t
test it (wireshark).

regards