Is Signal a non-free software?

"Is Signal a non-free software or still being one?

https://directory.fsf.org/wiki/Signal"

It would be nice if the FSF gave some clarification on their reasoning. From what I can gather the clients are under the GPLv3 license and the servers is under the AGPLv3 license so from a license standpoint they seem to be good. They have their code up on Github. I can't see a reason I would classify them as non free.

https://github.com/signalapp/Signal-Android

In the end what the FSF says may be true but I would be asking what non free dependencies Signal requires.

What the FSF actually says is this: "Signal, a messaging app for mobile phones, has earned widespread deserved praise as a free software encrypted voice and text communication client, but currently has a proprietary dependency on Google libraries. Helping the project remove this dependency and operate the necessary infrastructure is a high priority."

Also: "Developers are needed to work on projects that are creating free software real-time voice and video chat programs. There are several such programs, such as Ekiga, Jitsi Meet (and Jitsi Desktop), and Jami, and various projects based on WebRTC."

https://www.fsf.org/campaigns/priority-projects/voicevideochat (published four years ago on Jan 13, 2017 03:01 PM)

The fact that Signal by default requires some Google dependencies doesn't mean that the app is not forkable.

In fact, exist a FOSS edition of Signal >> https://github.com/tw-hx/Signal-Android/

More info about this fork >> https://www.twinhelix.com/apps/signal-foss/

In the worse case scenario, you can use Jami, which works in a similar way than Signal.

https://freedombone.net/faq.html#org96b7e85

I think I read somewhere that Signal does not require Google Play services anymore to work.

Anyway its license may be good, but Signal is not decentralized, which I think is very bad, as bad as this talk:

https://yewtu.be/watch?v=Nj3YFprqAr8

I like the SFS thing.

> Signal is not decentralized, which I think is very bad

That is the reason why I am not using Signal myself. This is a very deliberate choice by the Signal dev, which is based on their vision of networking technologies and control. They are totally transparent about it (cf. the talk you linked to). I discovered its other limitations only because of various discussions in other threads on this forum and elsewhere. They do not seem insuperable but the centralized structure will remain.

At the moment, Jami/Ring also requires several central servers to work properly, especially with the mobile clients, but the goal here has always been to go as fully decentralized as possible and anyone can decide to set their own servers and modify their client settings accordingly. Also, contrary to Signal, they have not chosen to focus on the mobile usage case, which might explain the often reported glitches. Jami is still work in progress but there have been noticeable improvements recently, especially when chatting with Android-based users. Splitting resources between so many sub-projects has its limitations (there are currently seven clients under active development in the Jami project). Maybe a Jami fundraiser would be in order to help it pass the crucial stability line.

Another messaging option is RetroShare, which almost exclusively focuses on desktop usage. The tool can be used for private, authenticated communication (which is what I think people are looking for when they pick a messaging app) but also for massive, anonymous file sharing (which has arguably a completely different appeal), and the intricate settings reflect this contradiction. Some people might be looking for both but I do not think having them served by the same tool is a good idea. Anyway, it can easily be used for private networking and only lacks a full external code review to rank top of the list for privacy and security - for old-school people like me who do not mind staying away from mobile networks, that is.

All in all, we are getting the usual picture of any libre software ecosystem: many projects with specific goals partially overlapping and a tendency for the less uncompromising ones to gain traction. Looking at the business model (or the absence thereof) behind any given project can also be helpful. For instance, who is funding Signal? Some answers here: https://www.wired.com/story/signal-foundation-whatsapp-brian-acton.

it's not free software, altough it's released with a free software license.

"#This project was abandoned because Moxie Marlinspike said he was not OK with LibreSignal using the Open Whisper Systems servers and the name "Signal"."
https://github.com/LibreSignal/LibreSignal

"The safest and easiest way to install Signal for Android is through the Google Play Store."
https://signal.org/android/apk/

^ non-free javascript, pushing google play smh.

app was silly when i tried it, like only a pin for passwords or whatever.

also, rms has said it's non-free in some videos

I'd send a link of signal body-guarding skype on twitter, but twitter is forcing non-free JavaScript now, but definitely saw it on chatsecure's twitter.

"The difference is huge; one we have control over, the other we don't. I understand that federation and defined protocols that third parties can develop clients for are great and important ideas, but unfortunately they no longer have a place in the modern world. Even less of a place for an organization the size of ours. Everyone outside the FOSS community seems to know it"
https://github.com/LibreSignal/LibreSignal/issues/37#issuecomment-217211165

^ weird

> Even less of a place for an organization the size of ours.

That sums it up. The "modern world" is a fanciful concept used to cover "what suits us better, as an organization".

I have recently taken to use smoke signals instead anyway. In order to reduce pollution, I am exclusively using steam, which I obtain through immersing myself in frozen ponds. This hugely helps keeping the amount of time lost in frivolous communication minimal.