Are decentralized services unable to innovate?
| Description | On Signal, XMPP and the future of decentralized services. |
|---|---|
| Website(s) | https://www.int21.de/slides/33c3-decentralized/#/ |
| Type | Talk |
| Kids session | No |
| Keyword(s) | social, software |
| Tags | federation, decentralized services, xmpp, email, signal, innovation |
| Processing assembly | Free Software Foundation Europe |
| Person organizing | Eal |
| Language | en - English |
| Other sessions... | |
| Starts at | 2016/12/29 20:30 |
|---|---|
| Ends at | 2016/12/29 21:30 |
| Duration | 60 minutes |
| Location | Hall 13-14 |
The Signal messenger is generally considered as the most secure encrypted messaging application available today. Signal's code is licensed under the GPL 3 and thus free software. But in many ways Signal goes against the values of the free software community. It depends on proprietary Google services, it requires Google libraries to run and it depends centralized servers which should not be used by alternative implementations.
In justification of these decisions Signal lead developer has written a provocative blog post in which he basically claims that federated services are unable to deliver innovations.
For Signal's critics there is one problem: Moxie is right. Everyone with an unbiased look at XMPP or E-Mail or any other federated service should clearly be able to see this. While Signal provides secure end-to-end encryption for all its users XMPP has 5 different end-to-end encryption methods - and none of them is enabled by default.
What can decentralized services do to break out of this cycle? And what other problems are there around abuse handling (xmpp is seeing a major spam flood these days)? And does the ability to self-host mean that everyone should host their own infrastructure?
