W3C home > Mailing lists > Public > public-sweo-ig@w3.org > January 2007

Re: Hacker rally questionnaire draft, take 2

From: Kingsley Idehen <kidehen@openlinksw.com>
Date: Fri, 12 Jan 2007 12:13:45 -0500
Message-ID: <45A7C1C9.8080408@openlinksw.com>
CC: W3C SWEO IG <public-sweo-ig@w3.org>

Kjetil Kjernsmo wrote:
> On Thursday 11 January 2007 17:01, Kingsley Idehen wrote:
>   
>> Starting a conversation is just that. I don't see how the notion of
>> "hackers" and their feelings come into play here per se. We just want
>> to have a conversation about the relevance of the Semantic Web to Web
>> Developers. During this conversation we explore usage scenarios and
>> possible solutions or seek: "How would the Semantic Web address this
>> problem..." type commentary. This kind of thing happens every other
>> second across the blogosphere today. We just need to being SWEO into
>> this mix by seeding relevant conversations.
>>     
>
>   

Kjetil,


> Well, this depends on your perspective. My perspective is this: Outside 
> of the Semantic Web community I have seen little of that commentary, to 
> the contrary, to me it seems like 90% of our audience allready thinks 
> we have failed miserably. They think that semweb is only talk, but that 
> it is hardly possible to get things to run, because it is too complex. 
> They think it is a dream of a small number of reality-detached 
> academics. 
>   
If 90% of the people outside the semweb community believe we have 
failed, then don't you think that they possess a good chunk of the 
missing use cases and demo scenarios around which education & outreach 
oriented collateral should be constructed? You will also find that this 
90% is comprised of individuals who inextricably link RDF the Data Model 
and RDF/XML.

> What they can cite is that RSS was not successful before it abandoned 
> RDF (not entirely true of course, but still). Mozilla has had RDF 
> support for years, but to leading figures there, it has had no value at 
> all, even though they hardly understood it from the beginning. People 
> who have understood RDF well includes Aaron Swartz, who allready told 
> this list that "now SW has such a bad rap that I wonder if it's too 
> late."
>   

I watched the RSS & RDF imbroglio from the side lines from the onset. 
This was a case of RDF/XML obscuring the RDF Data Model above everything 
else.
> In this context, "How would the Semantic Web address this problem..." 
> commentary is meaningless, because it is only talk. There has been 
> enough talk, in fact, there has been far too much talk. People are sick 
> of it, and more talk will only underline the presumption that we have 
> failed. 
>   

I disagree.

Look, when all is said an done this is how the Web Breaks down:

Basic Hypertext Web - traditional Web Sites (as created by TimBL)
Blogosphere - centered around RSS (bootstrapped by Dave Winer)
Web Services - SOAP and REST (bootstrapped by mashups of Data exposed as 
RSS or basic XML)
Semantic Web - were the bootstrap is in progress (as a manifestation of 
TimBL's web enhancement manifesto)
> I believe it is not possible to reach out beyond the Semantic Web 
> community with conversation at this point. It is too late. The train 
> left the station. The rocket has been launched. There are too many 
> high-profile failed attempts to use SW technologies out there and 
> hardly any prominent success stories.  
>
>   

Here is a simple example.

If you can demonstrate to the Blogopshere  that you can build much more 
powerful mashups via RDF Data Sources as opposed to RSS 2.0 or Atom Data 
Sources, the dialog starts. When this is achieved, you can then go to 
next step of showing how Online Communities such as Blog clusters 
(Blogosphere), Wiki Collections (Wikisphere), 1st Gen. Social Networks 
(that expose FOAF for instance) are used to construct RDF Data Source 
that are easy to consume across a myriad of environments.

> There are a few things that can reverse this trend. One thing is 
> widespread adoption in at the enterprise level, something that is 
> clearly happening, such as the Norwegian oil industry, which will make 
> SW technologies a job requirement. Another thing is academic adoption, 
> which will expose a lot of students to it. It took only two such 
> students to start Yahoo and another two to start Google. These things 
> are also happening, as there are a few academic programs that include 
> it, and projects including the Mesur project at LANL and Virtual 
> Observatory for astronomy. A third possibility is that it spreads from 
> entertainment industry projects, such as the Venice Project.
> It could certainly spread from these angles, but I fear that all these 
> things will be backend-deployments, it will not really expose the data 
> to the web, and on the web, there are few programmers who actually use 
> it. That's not really the Semantic Web, is it?
>
> As a big part of the early web, beyond the High Energy Physics 
> community, was personal homepages with a lot of pretty interesting 
> stuff, I think it is equally important to have on board a large number 
> of individuals. I think it is extremely important to reach out to them, 
> but as they think that all they need to know is that the Semantic Web 
> has been tried and failed, there is only one thing that I think is 
> meaningful to do at this point, and that's to stop talking and prove 
> them wrong with code. Genuinely useful stuff. And we have to create it 
> ourselves.
>
> And that's why hacker's feelings are not only relevant, but extremely 
> important, because they are the only ones who can do it. I'm not even 
> trying to appeal to the bloggers outside of the SW community because I 
> think I allready know what they think ("bad rap"). I have such a great 
> intellectual investment in the Semantic Web, that I want to write the 
> code to make it work, but I know other hackers who doesn't feel quite 
> the same allthough they would actively participate if we don't give 
> them the impression that this is something that they must do, but 
> something that's interesting to them.
>   

Hackers write code, code manipulates data, no hacker worth his or her 
salt has a problem working with powerful data sources once they 
understand the essence of the Data Source. The only issue with all of 
this is that a majority of people simply don't grok the Data Model.

Look, a majority of Semantic Web communications collateral (Tutorials, 
demos, white papers, presentations) take the following sequence:

1. Define the Semantic Web Vision
2. Describe RDF (typically as RDF/XML)
3. Provide a simple Triple example using RDF/XML

Whereas the sequence should be:

1. Define the Semantic Web
2. Describe the RDF Data Model
3. Describe the basic Triple
4. Provide examples of basic Triple in: N3/Turtle and RDF/XML
5. Explain the purpose of the interchange formats
6. Describe RDF storage options
7. ....



The skeptics are not going to grok the Semantic Web based on what we 
say, they are going to do so based on what we demonstrate (Dave Winer 
demonstrated what became the Blogosphere via his Userland platform).
As stated earlier, Mashups drive Web 2.0, and RSS drives the 
Blogosphere. Thus, we should produce demos and tutorials that show how 
the Semantic Web enhances Mashups, Online Communities (Social-networks, 
Blogs, Wikis, Discussion Forums, shared Bookmarks etc).
> There is certainly value in engaging the average web developer in 
> conversations about SW, as we get a better overview of what they 
> actually think. It is probably not quite as categorical as I say above, 
> especially since you say people are talking about it. It could also, 
> with time, motivate some to do something practical, but I see that as 
> orthogonal (or at least 80 degrees in Euclidian geometry :-) ) to my 
> initiative here. My initiative here is to gather hackers allready 
> interested in the Semantic Web and produce a quick consensus around 
> what's interesting, and then do it with least possible talk involved. 
>   
We can establish rapport with the Web Development communities by  simply 
producing a demo that explore enhancing perspectives that are well 
understood in the Web domain (as per my comments above).

You can never educate anyone without identifying and establishing 
context as the basis of rapport. You can't establish said context 
without conversation with the target community. Note, conversations take 
many forms. A classic example is TimBL's statement about not seeking 
permission to create the Web, he simply did it. That's the kind of 
conversation that works with networked communities (i.e. Network Effects 
Exploitation: locate the conversation clusters, analyze the essence of 
the conversation, and then participate in a palatable manner).



Kingsley
> Cheers,
>
> Kjetil
>   


-- 


Regards,

Kingsley Idehen	      Weblog: http://www.openlinksw.com/blog/~kidehen
President & CEO 
OpenLink Software     Web: http://www.openlinksw.com
Received on Friday, 12 January 2007 17:13:55 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 8 January 2008 14:17:35 GMT