Re: A bit of bomb throwing....

From: Zack Rosen <zack_at_civicspacelabs.org>
Date: Mon, 16 Jan 2006 21:59:36 -0800

> I think you are on the wrong email list - seriously. You basically
> are telling universities to develop implementations (and
> consequently business models) for a new technology.

Not the case. Im advocating that semantic web research be done
within established open-source communities meeting 'real-world' needs
of partners instead of in isolation as examples.

> AFAIK TBL developed (his part of) the WWW to exchange documents
> with his geeky friends not to develop the most successful,
> commercial, ubiquitous network ;o) As you pointed out the WWW took
> off because of Netscape, EBay and so on. They developed the
> applications for the masses and invested the money, they picked
> HTML and HTTP as their choice of technology. There was plenty of
> innovation around (e.g. connected mailbox servers etc.) - sure in
> hindsight it is easy to say 'Well it was supposed to happen this way'.
> Also comparing WWW and semantic web is not fair as the WWW in its
> basic form is simple (excluding basic technology like TCP/IP, DNS,
> etc.) - the semantic web utterly complicated (still excluding
> previously named technologies).

Yes, the semantic web is multitudes more complex which makes it all
the more difficult to gain external momentum.

> I think SIMILE is doing a good job and I wish more projects would
> be so well organized and oriented towards a useful outcome. But to
> bring it to the market the industry has to step in - the
> universities do not have the expertise nor resources to do this
> job. Cooperative approaches are as close as it gets to your utopia
> but that again is initiated by the industry as the provider of
> funds. I for my part work for SAP Research and here you see the gap
> being bridged from the other side meeting projects, like SIMILE,
> half way. Companies do research too and pick up where universities
> leave if it is interesting enough.

What is the path to industry adoption of semantic web technologies?
All I see are obstacles: threatened business models (such as SAPs')
stalling and early adopters (such as SalesForce) opting for the much
simpler and easier approach of REST API's.

I think the much more aligned and willing allies of the semantic web
are the tinkerer's: open-source developers and entrepreneurs. In
fact it is these people who are currently paving the way for
automated data-interchange on the web (RSS, microformats, etc).

> Your frustration seems to be born out of the high hopes the
> semantic web sparks in many but you should realize it will take
> time if it ever happens. Contemplating about it is necessary but if
> done excessively a waste of energy. As mentioned before you should
> decide to either help develop the technologies (from either side of
> the gap), use something else or if you think neither will work for
> your problem then you might face a challenge which is unsolvable
> today.

We are eight years down this path. It isn't working - at-least not
fast enough.

-Zack

>
> -----Original Message-----
> From: Zack Rosen [mailto:zack_at_civicspacelabs.org]
> Sent: Tuesday, 17 January 2006 2:41 PM
> To: general_at_simile.mit.edu
> Subject: Re: A bit of bomb throwing....
>
> On Jan 15, 2006, at 3:03 PM, Prokopp, Christian wrote:
>
>
>> _at_Zack:
>> You are making valid points but I think also contradicting
>> yourself. On
>> the one hand you ask the researches to come down from their ivory
>> tower
>> and meet the practical needs and on the other you complain that when
>> they do (like SIMILE) that it is not the 101% semantic web you would
>> like it to be. Some of your arguments also apply to any new
>> technology
>> (e.g. you could have made all your barrier points about the WWW too).
>> I was very pleasantly surprised when I cam across SIMILE - a
>> realistic
>> approach to bring semantic web technologies to the masses and on top
>> added some documentation and usability consideration. It also orients
>> itself at a real world use case so I really have trouble following
>> all
>> your points. Maybe you picked the wrong project for your argument ;o)
>>
>> No hard feelings but I strongly disagree although in general I could
>> agree with some points in your analysis but just not with SIMILE.
>>
>
> SIMILE definitely is a positive step forward, and you are right, I am
> being overly harsh and picking a fight with the wrong crowd. That
> said I think there is still an impossibly wide gap between semantic
> web research and the earliest adopters who wish to employ the
> technologies in 'real-world' applications. My goal in approaching
> this community is to asses what it will take to close that gap.
>
> My understanding is that the intent of SIMILE is to create user-
> facing applications that demonstrate semantic web technology. But
> 'user-facing' simply does not equate to 'real-world'. It is pretty
> easy to create a user-facing application that is of no real-world
> practical use (and indeed it seems SIMILE has created quite a few).
> This is fine in the context of academia and standard academic
> research, but the goals of the semantic web are extremely ambitious,
> and I am convinced this approach simply isn't going to produce
> results quick enough. As long as research is pursued under the
> assumptions that the semantic web will be created within a university
> a whole universe of potential contributers and adopters that could be
> leveraged towards their goals will be ignored. Consider me one of
> the more outspoken and crankier of these potential contributors.
>
> The WWW was born on the frontier of the internet and helped catalyzed
> it's tremendous inception. But it's a different landscape today.
> The semantic web has a trillion dollar legacy problem.
>
> 1990: When the WWW was created it offered critical and unique
> functionality in a new and creative environment. It was ingenious,
> relatively simple to implement, and the only game in town. Down the
> mountain the little snowball went - JPL puts up pictures of
> asteroids, Compuserve signs up some customers, Netscape IPO's, Ebay
> revolutionizes junk sales, Google monetizes search and voilà the web
> is born.
>
> 2006: The web works. Billions (if not trillions) have been invested
> in platform technologies such as LAMP, JSP, .NET and wells of
> innovation are being tapped more quickly than ever. Meanwhile up
> atop the hill Universities and far sighted corporate research
> vehicles have spent eight years perfecting their new snowball are
> apparently stuck waiting for others to push it down the hill for
> them. But it simply isn't happening yet. Structured data
> interchange technologies for consumers on the web (RSS) appear to
> steadily advance at speed at which Dave Winers beard grows (http://
> zacker.org/zacks-law) while simple-to-implement / simple-to-
> understand web-service data interchange techniques (REST, SOAP, etc.)
> solve the immediate 'enterprise' needs leaving no immediate critical
> functionality path for semantic web technologies to toboggan down.
>
> The semantic web has a well articulated vision, it has a proven
> leader, it has it's pick of incubating vehicle, but what it is
> desperately is lacking is a workable pathway for adoption. I firmly
> believe one exists.
>
> -Zack
>
>
>
Received on Tue Jan 17 2006 - 05:59:10 EST