Frederic wrote:

Guten Tag Jeff V. Merkey,

am Mittwoch, 2. August 2006 um 23:28 schrieben Sie:

JVM> Julien Lemoine wrote:

...

...

Hello,

I have wrote a "google suggest" like service for wikipedia under GPL licence [1]. By the way, I am not sure if this project will interest you, I am open to all comments from your community.

I wrote a simple web site for english/french on http://suggest.speedblue.org, all software are available in the download section.

Best Regards. Julien Lemoine

[1] If you want me to change the licence to something other than GPL, please let me know.

---

---

foundation-l mailing list foundation-l@wikimedia.org http://mail.wikipedia.org/mailman/listinfo/foundation-l

JVM> It's a start -- we also need a specialized file system for a google JVM> replacement.

JVM> Jeff JVM> _______________________________________________ JVM> foundation-l mailing list JVM> foundation-l@wikimedia.org JVM> http://mail.wikipedia.org/mailman/listinfo/foundation-l

I don't understand what the discussion is all about. Plse Explain.

Was hast du gemacht! :-)

The search engine wiki is a great idea. I am willing to develop an enhnaced replica of Googles File System archtiecture to run beneath it for this project. We will need developers on the php side to define a set of interaction rules to speed searches.

I would like to work out some set of agreements on the underlying technology to ensure it is controlled by the community. Open Sourcing it is not out of the question, but I would like to get some sort of consensus on how we develop it and keep it under the control of the Foundation and Wikipedia Community initially. I would propose we develop it as a project jointly owned by the Foundation and the contributors to the project on a concensus basis on who gets to use it and how as part of the licensing terms. No such public license has ever been created, but we should attempt to do so. This will ensure we have control over who uses it and for what purpose based on community concessus. i.e. the work should be jointly owned by the creators as a "shared copyright". Using it should be free -- owning it should be a community based position.

We need the following:

Clustered file system for massive search engine capabilities distributed content crawlers to harvest the internet by scraping (we can start by scraping google and yahoo -- Wales already figured a lot of this out with DMOZ. He would be a big help here. Distributed Search capability.

I'm in for the file system and underlying hardware platforms and hardware support, including SAN based and distributed models. We should not open source the code, but develop a community licensing scheme for the technology. This way we can build a revenue model around hardware and platform sales and certain classes of subscription services (not for the content itself but for services like "I want to link our educational organizations cluster of services to your central hub for automated image and content updates, translation services, and service and support and would like to compensate the foundation for these added services which are above and beyond simply public access to the information itself, which is free). I can see us creating all sort of jobs for folks helping with content management, translation, publishing textbooks and selling hard copy books for education, etc.) Stuff that keeps the information free, but lets the Foundation have a viable revenue model to pay for everyones expenses and help give folks a stake to build something that will endure 100 years from now. Sort of like combining the power of Google and Amazon together.

That's what I'm thinking.

Jus!

Jeff

On 03/08/06, Julien Lemoine speedblue@happycoders.org wrote:

Hello,

...

Very nice! How often is the database updated?

For the moment, I do not update the database, I analyze a grabbed wikipedia.

I developed this program as a proof of concept. For the update of the database, it will depend of the possibility of integration in wikipedia. I do not have enough details at this step.

Have you played with WikiWax? It's a bit temperamental (and uses an old dump), but much the same idea - see wikiwax.com (or [[en:WikiWax]])

Hello Andrew,

Andrew Gray wrote:

Have you played with WikiWax? It's a bit temperamental (and uses an old dump), but much the same idea - see wikiwax.com (or [[en:WikiWax]])

Yes I tried it, but it gives a full list of matching articles without ranking. I did not find it really useful and very fast, I preferred to keep a small list of ranked results.

Best Regards. Julien Lemoine