[nepomuk-kde] Social Desktop in 4.3?
karlitschek at kde.org
Thu Apr 30 09:04:39 CEST 2009
I would love to work closer together with you.
I think there is a huge potential if we combine community data and
semantic desktop ideas.
And you are right we should support existing standards whenever
possible. In fact we support OpenID already and I´m looking
into OAuth at the moment if we could use it for API authentication. I
share your opinion about dataportability.org by the way.
A lot of big players discussion interesting ideas with zero results.
I think if KDE wants to be the market leader in this area we should
move forward and implement functionality and experiment with it.
This is what we try to do with OCS. The API is as simple as possible
so that different websites and application can use it. For example the
guys from forum.kde.org are interested in implementing the knowledge
base functionality. I don´t want to wait for Microsoft/Google/Oracle
and others to agree on a knowledge base specification standard. And
the idea is to improve the API in the future. For example by
integrating ideas from you and adapting the API to other standards if
they become available and it makes sense for KDE.
Don´t get me wrong. If there are standards for the innovative features
we are doing at the moment and plan to do later this year I´m totally
agree that we should use them. But at the moment I think that the
great frameworks of KDE 4 enables us to move forward and become the
leader in Semantic Desktop, Social Desktop and other key areas.
This reminds me of the corba -> dcop -> dbus story.
1. First we tried to implement the corba "official" standard designed
by a industry forum. But it was slow and bulky.
2. Than we decided that we can do better and faster by creating dcop.
We used dcop for years as an advantage over other Desktops and it
3. Other Desktops realized the using a light weight solution is the
way to go. And we all created dbus as a de-facto standard.
On 29.04.2009, at 22:37, Gunnar AAstrand Grimnes wrote:
> The standardisation efforts for social web things go far beyond the
> semantic web - although of course we should put semantic things into
> many things as possible :)
> For technologies, there is of course OpenID, now supported by almost
> everyone, for basic authentication, there is OAuth for the exchange of
> user-data, and there is Google's OpenSocial for social widgets (i.e.
> almost like an application framework for social pages, like Facebook
> Then there is dataportability.org, which is the main interest group,
> joined by google, microsoft, etc. - although in the end they seem to
> little more than joining and talking a lot.
> I know Dan Brickley, who is behind FOAF, also works hard to on the
> intersection of these technologies and semantic web things, he blogs
> about it, but his page http://danbri.org/ is unfortunately down at the
> I also talked briefly to Frank Karlitschek about this at Academy last
> year, but did not take the time to convince him that joining these
> people was the way to go :) (Disclaimer: I've not really followed the
> kde social desktop, for all I know has started using all of the above
> since then!)
> =- Gunnar
> Richard Dale wrote:
>> On Tue, Apr 28, 2009 at 8:53 PM, Sebastian Kügler <sebas at kde.org>
>>> On Tuesday 28 April 2009 17:49:48 Richard Dale wrote:
>>>> On Tue, Apr 28, 2009 at 2:12 PM, Sebastian Kügler <sebas at kde.org>
>>>>> As some of you know, I've spent some intense cycles on the
>>>>> social desktop
>>>>> support in KDE and specifically in Plasma. This functionality
>>>>> a social network, in this case openDesktop.org which powers
>>>>> sites such as
>>>>> kde- ook.org, kde-apps.org and opendesktop.org. The server-side
>>>>> parts are
>>>>> specified in the Open Collaboration Services API, to be seen on
>>>>> The grand vision behind all that is to closely integrate the
>>>>> social web
>>>>> into the desktop. What's there right now is the tip of the
>>>> What I don't understand about this stuff is why it completely
>>>> the Sematic Web. Why doesn't it use FOAF for instance, instead of
>>>> redoing the same thing in XML? Why do we have one set of
>>>> ontologies in
>>>> Nepomuk, and yet a completely different non-RDF bunch of stuff on
>>> I think that's mainly because it's all still very young
>>> technology, but I also
>>> have to admit that I don't know enough of the semantic stuff yet
>>> to make real
>>> sense of it.
>> Yes, the RESTful apis actually look very nice and easy to use. It's
>> just that I think
>> data should be the starting point, and not a specific api. Otherwise,
>> we'll just end
>> with another 'data silo' even if it has a public open api, because to
>> do anything with
>> the data well will need to convert it to other formats.
>>> Also, the KDE side of these things is not quite there yet. When I
>>> was for
>>> example looking for a generic concept of a Person (for this, but
>>> also for
>>> another project), my impression was that Nepomuk is not quite
>>> there yet.
>> Nepomuk uses different ontologies optimized for the semantic
>> desktop. But
>> if the data is in the form of triples it is much easier to transform
>> it and combine
>> it with other RDF based data. For KDE the semantic web ontologies
>> that seem
>> be the most suitable might be:
>> * FOAF - info about people
>> * SIOC - blogging info
>> * SCOT - tags that are connected to semantic web ontologies, rather
>> then just
>> being names
>> * DOAP - description of a project
>>> That's of course the desktop side of things. How this looks on the
>>> web, I
>>> don't know. Heck, I don't even know FOAF (OK, *now* I know, thanks
>>> Wikipedia ... ;-)).
>>> All this is very new, and we're in no way setting things in stone.
>>> But I guess
>>> it'll take some time until it all fleshes out, becomes clear how
>>> this should
>>> all be connected, and for the technologies and environment to
>>> mature enough so
>>> it's actually useful. It's also one of the reasons why I want to
>>> get it out
>>> there, so visibility increases and more people start working on
>>> it. Especially
>>> I hope more smarter people than me do. :-)
>>> I'm dying to combine all those subjects, but I don't see how to do
>>> it at this
>>> point in time, pointers are of course welcome.
>> I think the name to search for is 'Social Semantic Desktop'. Here
>> is a
>> slide show by
>> John Breslin (it mentions KDE4):
>> I'll cc this to the Nepomuk list as the people there know the most
>> about it, and I would be
>> interested to hear what they have to say about the Social Desktop
>> -- Richard
>> nepomuk-kde mailing list
>> nepomuk-kde at semanticdesktop.org
> Plasma-devel mailing list
> Plasma-devel at kde.org
More information about the Plasma-devel