directory-discuss
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: linphone, still free?


From: Narcis Garcia
Subject: Re: linphone, still free?
Date: Wed, 18 Dec 2024 09:49:53 +0100

(sorry for my bad english)

I think e-mail client example fits in the debate, and two SIP endpoints (softphones) when communicate do not become a software entity to be evaluated as a single part of freedom matter.

Same as two computers with different OS when they share data in same LAN.


El 17/12/24 a les 23:33, Lorenzo L. Ancora ha escrit:
I would actually be grateful if someone who has more time checked how free and 
nonfree versions of Linphone talk to each other.

Yes, some experimentation would be appropriate.
Meanwhile, for common knowledge, this is how apps like Linphone work, very 
roughly:

0. The interlocutor is found in a remote SIP directory.
1. A direct SIP connection is negotiated and established.
2. Your voice is digitalized with VoIP and stored in memory.
3. RTP is used to safely rely the stored data back and forth.

The VoIP interlocutor will be able to capture your IP and version during points 
0 and 1.
As a consequence, in most circumstances it would also be possible to implement 
a warning about communication with non-libre endpoints before point 2 starts.


โ” Would you consider the project responsible if the user of the libre version 
unknowingly includes a non-free client in a conversation?
โ” What could be the ethical and safety implications of not warning the free 
user that this is happening? (so the free user can at least prevent or close 
the conversation)
I am open to different outcomes, but it just occurred to me that we don't see 
an issue with free e-mail clients being able to send/receive e-mails from 
nonfree e-mail clients, do we?

...but the web email archives kept by GNU will still store and publicly display 
the name of the e-mail client used by each thread participant. Amongst the many 
email headers, keeping and showing this one in particular. Perhaps, while as 
you say us three don't see an issue (at least in asynchronous protocols), other 
F/LOSS lovers may not be so open minded. ๐Ÿคจ

Privacy-side (which I'm sure won't be an issue here, but it is still worth 
noting) the nature and sensitivity of the information exchanged via email is 
much different.
With emails, all users are well-aware of the privacy risks, because the 
presence of at least two servers involved in each asynchronous communication is 
made obvious by the clients and all emails are saved by default, so the 
information is known to be permanent. The same cannot be said for calls and 
videocalls, which are naturally considered ephemeral, fully private and 
personal. Now we have AI and we all know its dangers: exfiltrating company or 
family meetings, affairs, negotiations and other sensitive interactions could 
be very bad, especially because anything can happen during synchronous 
real-time communications, even unwanted things that should remain OTR.

On the side of ethics, a person could legitimately decide that they have the desire or even 
necessity (depending on the walk of life they're currently in) to mainly or only interact with apps 
which are free and hold no surprises in their codebases. The word "free/libre" can 
provide a sense of assurance and trust which should be well-founded. If their interlocutor says 
"I'm using Linphone", the logical expectation is that a libre client is executed on both 
sides, no strings attatched.๐Ÿ˜‰

~Lorenzo








--
Narcis Garcia



reply via email to

[Prev in Thread] Current Thread [Next in Thread]