social-discuss
[Top][All Lists]
Advanced

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

Re: [Social-discuss] PHP-Based GNU Social structure


From: Ted Smith
Subject: Re: [Social-discuss] PHP-Based GNU Social structure
Date: Wed, 31 Mar 2010 17:54:55 -0400

On Tue, 2010-03-30 at 09:52 +0200, Carlo von Loesch wrote:
> Ted Smith typeth:
> | ...is because we seem to have a consensus on what UIs are necessary - a
> | PHP-based web UI would be nice for users who don't want to install
> | things, and a desktop client would be nice for users who do.
> 
> It is I who has started doubting the validity of the commodity
> webhosting approach, not because I don't like cheap server space
> but because of your remark about scrapping private keys off virtual
> machine memory. It sharpened my doubts about VM technology into
> realizing how this can be just as dangerous as giving your data
> to Facebook. ISPs can easily be subpoenad into monitoring the
> virtual machine of any activist group or whoever is to run GNU
> Social, even if we design the system to use plenty of HTTP-based
> cryptography. I was further alerted by RMS's words on SaaS.
> Most of them apply to our constellation, even if he thinks social
> computing should be exempted.
> 
This is a more compelling reason to have a clear core/UI distinction -
even if the core is running on commodity webhosting, all (end-to-end)
crypto could be done in the UI, and then the private keys scraped from
VM memory would just be session keys for talking to other nodes, rather
than persistent keys for talking to and identifying other users.

> I am not fundamentally against web-based implementations, in fact
> I started out in this discussion thinking that a Facebook-like UI
> would be pretty cute to have, but I feel it should be combined
> with a license that goes beyond the Affero GPL. It should be
> forbidden to run this software in virtual machines as the privacy
> of the users can no longer be secured. If you run the software on
> your own hardware server, even if it's in somebody else's rack,
> it is much harder to harvest. Maybe I'm just some five years early
> thinking about the dangers of VM computing, but that's traditionally
> my role hurrying ahead of time.

That would not be a free license, so I think that's impossible here.

> 
> Another point of critique against server-based social networking
> is the power of the user's private key: User-bound end-to-end
> and one-to-many encryption are a whole new dimension of privacy
> within social exchange and it's a pity if we don't leverage that
> just because we have a requirement to run on commodity web hosting.
> So in that sense having to be able to provide a web frontend
> seriously limits the possibilities where this technology could head
> off to.
> 

We can do crypto in javascript; it'll be a pain, but a web UI can be
done securely.

Attachment: signature.asc
Description: This is a digitally signed message part


reply via email to

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