We finally decided to bite the bullet and create a plugin for WordPress to allow knowledge building discussions, such as progressive inquiry. While Fle3 is a very good tool for doing KB, it’s quite challenging to install. The impulse came from Hans, when he (jokingly) said that he needed knowledge building on WordPress by “next Tuesday”.
Well it turned out that implementing KB discussion onto WordPress blogs wasn’t such a difficult task after all. I’ve now written a functioning plugin that does this, and it even uses the knowledge typesets from Fle3. Meet: Fle4 – Knowledge Building for the rest of us. It’s still very much in beta. There’s a live demo so you can test out Progressive Inquiry and Six Hat Thinking, but easy plugin installation is still a few days in the future (waiting for acceptance into the WordPress Plugin Directory).
So… Take a look and send us comments.
We today got together to organize work around preparation and writing of research papers. Our new research assistant Katrina is taking the greatest responsibility to keep us analyzing and reporting the research results from the last three or so years of work in the research group. We are working on the following papers (working titles):
Software as hypothesis – developing human-centered design research method through three design cases
Audio wiki for mobile communities – information systems for the rest of us
Design of LeMill – web community for finding, authoring and sharing learning resources
The first article is a formulation and conclusion of the design research method and practice we have been developing in the research group.
The second article is presenting new design direction for the MobilED project proposing that the audio wiki is probably more useful as an informal community information system rather than as an application for formal teaching and learning (for what it was originally designed for).
The third article is the first proper description of LeMill with explanation of principles behind it, design rationale and the process of designing it.
We are aiming to get these out in six coming months. We are targeting to some journals and some conferences. Can’t wait to get these out!
I just realized today that we have a good change to release three new major “products” still before the summer. I also have a student who is interested in to develop the Papanek idea machine. This would make it four.
So, what are these “products” and how they should be considered in the “academic context”?
Our products are actually our hypothesis. They represent and carry with them definitions of challenges related to some human activity system. They are also representing out understanding on how the challenges could be solved. As hypothesis they are also prototypes: they are something we can test if they really solve any of the challenges they are designed to solve. The product is not only a product, but also our definition of “better way of doing things”.
In the academic circles the “product” sounds very commercial. However, our products are avant-garde products. They are experimental and novel. This makes them naturally to be “products” coming from an University, rather than from any other place.
Are we going to have a major release party some day later in spring?
If, we’ll get all four products released before then end of May, I’ll promise to arrange a huge party. Hans will be the dj.
The rewrite of Fle3 on top of the Kala framework is progressing to the user interface part. Probably the largest difference to the earlier version is the move from DTML to ZPT in the page templates. Another new feature are human-readable URLs. The old Fle3 encodes its URLs in a very confusing way. The rewrite uses Zope’s folder hierarchy and the human-readable URLs it maps to as much as possible. This mostly works, although Zope expects applications to be structured in a tree hierarchy, and some tricks are needed before the structure of Fle fits into this.
For example in the old Fle, a jamming session in a course has the url http://localhost:8080/FLE/courses/1/jamming/js11? state_url=9,1course_id16,1inline19,12cc_order123, 9,10,2,157,10um_sortfirst_name14,1full_thread1_21, while in the new Fle the URL is http://localhost:8080/NewFle/courses/my_course/ jamming/jamming_instance/jam1, which can still be improved once we manage to convince Zope that the superfluous jamming_instance element can be dropped.
With the new Fle, it is possible to choose which modules you want to install. An installation with only knowledge building, only jamming or neither is possible. This is possible thanks to the Kala framework, which allows choosing the set of modules to install.
In the new modules, we’ve also started using Zope’s Python Script files to connect HTML requests to the Fle Python product, instead of parsing requests directly in the product API. This new convention makes the product API easier to understand and can help making lightweight customizations to installed Fle instances.
I want to document the discussion we were having the other day about a name for a new version of Fle3.
The new Fle3 is totally new code. So, in that way it would make sense to call it Fle4. The number has been changing always when the software has been written from the scratch. The Wikipedia article about Fle3 explains the history of FLE.
When asked I have said that we will never do Fle4. Our new developer Risto’s fast reply to this was that we should then call the new version Fle5. Hans has been promoting (as a joke) the name FLE 2.0, which would relate the new version to Web 2.0 (and everything else with the 2.0 addition). I think that this is would be a bit unfair because FLE has been web 2.0 since 1998.
From this we got the formula: Fle3 + 2.0 = Fle5.
It is possible that the new Fle3 will be called Fle5, but no decisions have been made yet.
The ITK 2007 conference takes place in April in Hämeenlinna. We should present some of our results in there. LeMill and Calibrate project should get some visibility. Hopeully the new FLE can be shown in there, too.
We started the complete rewrite of Fle3 on the 25th of September. Now, after two months, our competent new Python guy Risto Saarelma has completed the KB and Webtop modules (without UI). During the next two weeks we should have the Jamming module done, after which we’ll do the UI to the entire thing before Christmas, and we should have a demo up and running before the year is over, and a release candidate early next year.
As the project manager of sorts, I’m happy that we’re moving from dtml to zpt, and that the product is now properly covered in automated unit tests.
If you’re interested in more details about the development, head over to the Fle3 development site. We’re using Trac which we’ve found an extremely useful and simple solution to our dispersed software project which is going on concurrently.
After a few years of silence, we’re resuming our development of Fle3. After the development of ImaNote, we now have a tool that allows users to freely position and move objects in a 2D space using just their browser. This has originally been a key feature in Fle3, but back in 2001 the browsers just weren’t up for it. Now they are.
And we’re not just adding new stuff, we’re rewriting the whole thing to be better. The oldish Zope technologies like DTML that we’re using are just too clumsy and are being deprecated as time goes by. So we’re upgrading to ZPT and cleaning the whole code base as well. As Fle3 was the first Zope product that our team had worked on, it really shows in the source code, unfortunately. So now that we understand how acquisition works and all Zope’s special features are well in hand, we’ll turn out Fle3 that’s leaner and meaner than before.
We’re hiring a new Python to work exclusively on Fle3 development, so we’ll see how thing start rolling. I’m happy if we’re near a release during this year.