[Freedombox-discuss] Social Sync (was: Re: Is / Is Not)

[Josef Spillner]

Hello,

:: Jonas Smedegaard Donnerstag 09 Juni 2011
> Isn't Akonadi a KDE framework.  I don't follow - do you suggest to run
> KDE on the FreedomBox?!?

Akonadi is just an example which I'm slightly familiar with so I put it there 
as potential candidate. I'm happy to include other personal data storage 
frameworks or web resource caching frameworks in that diagram. It's supposed 
to somehow provide the big picture of how things currently work together or 
are supposed to work together some months from now when taking today's ongoing 
development efforts into account.
The lack of documentation is indeed an issue; today I've seen that there's 
also been a connector for some Google services which process personal data 
available and Debian-packaged for about two years.

(NB1. Akonadi is not dependent on KDE, and seems to run fine on smaller 
devices like smartphones.)

(NB2. If a desktop should be run on the FreedomBox probably depends on its 
nature. If it's the proverbial plug computer, then my understanding is that it 
might run some of the components which are currently contained in the desktops 
of every device with a desktop, such as the wallet containing the credentials 
to online services or the personal data storage, and makes them accessible to 
connecting devices. If however my notebook or smartphone serves as a 
FreedomBox by just installing or activating the right packages, then it will 
obviously run a desktop by itself.
Such differentiations are not the focus of my work though; I simply assume 
that there will be some integration with the desktop as the user's control 
interface to his or her devices and services.)

> I guess it is useful with a URL to some more info than the picture.

That's what my plan is, to write down some text once I've found enough 
information about what is out there and especially how well it works already.

> Agreed.  Any suggestions on where to gather?
> 
> Personally (since Perl is what I can code) I hang out with perlrdf[0]
> folks on IRC and also follow the foaf-protocols[1] mailinglist.
> 
> [0]: http://www.perlrdf.org/
> 
> [1]: http://lists.foaf-project.org/mailman/listinfo/foaf-protocols

Thanks for these two. It's certainly a good idea to start with a list of 
already existing communication spaces instead of opening yet another one 
without reason.

> Sounds great.  Could you list some concrete examples of ready-to-use
> tools in this area - or perhaps a pointer to some summary list.  It
> might very well be that I am missing lots of cool progress on this
> front.

Those not yet contained in the diagram (for the simple reason of me also being 
mostly uninformed about them) seem to be built on Folks or QtFolks and 
Telepathy. At least that's my impression from what I experienced when using 
Maemo and Meego devices where the whole setup of contacts seems to be more 
intuitive, direct and first-class when compared to traditional desktops.
However, this experience severely degrades when looking at events and 
especially content in social networks.
There are also a couple of additional libraries like libsocialweb and its -
plugins-extra extension, both of which seem to target using (or migrating off) 
proprietary services. Their use is certainly goverened by some ToS, though.

A specific question still to decide is if it should be a goal that all 
information ever pushed by a user to a remote service should be kept as a 
local replica just in case the site or network will become (permanently) 
unavailable. None of the current systems gives such a hard guarantee. Most 
users of today's sites don't invest the effort to keep manual copies; the 
question is therefore if they expect it to be magically there or if they just 
don't care if it won't be there.
If the guarantee shall be given, a much tighter integration with the desktop 
including its application (both conventional desktop applications and local 
web applications) would need to be implemented for enforcement, consistency 
and manageability.

Josef