1. what did you find (articulate the answers in your own words)

1.1. interesting about the article?

The first idea I found interesting was the notion of trade-offs in EUD languages. I would think that the more time one is willing to spend learning an EUD language would directly relate to the degree of scope it provides, but it seems there are ideals which provide such scope without excessive learning requirements. What kind of low-cost, high-scope tools are available? I am familiar with Excel macros (or macros in any large program), but what are Alice and Agentsheets, and how close to the EUD ideal are they? Presumable they are some kind of interactive graphical world?

I also found the presentation of the Open Source development process as a meta-design interesting. The open source methodology seems to exactly fit the idea of a meta-design. Typically an Open Source project starts when a developer has the need for a program and starts writing one. They have instantly created a meta-design, they will be a user of their own software, and will be participating in an ongoing design process. Other people may use the software and have no need to modify it, and will not participate in the meta-design. But some users will find a problem with the software or want to add some feature and will modify it, contributing in the meta-design process. The problem which Open Source projects often face is the learning cost of contributing to a project is so high, many people do not wish to invest the time nor energy to learn the system well enough to modify it. Imagine trying to fix something like an html rendering bug in a large Open Source project like Firefox, where would you even begin looking through the thousands of source files? You may be able to narrow the search down by looking at filenames, but once you find the region of code which handles html rendering, you will have to study and understand the design and flow of the code before you are capable of fixing the problem.

1.2. not interesting about the article?

I thought the whole paper was interesting.

2. what do you consider the main message of the article?

The article introduced the idea of EUD languages, and proposed meta-design as a way to help address the learning issues behind them.

3. is meta-design an

3.1. interesting concept? why?


Meta-design is an interesting concept because it leverages the user as a designer, as we have talked about in class before. It comes back to no single person being able to tackle a large design problem on their own. With something like an EUD, the original designer of the software does not need to envision and design every possible usage scenario, an extensible system is provided which the user can manipulate into solving their particular problem.

Another interesting concept is that meta-design requires that users are able to continuously extend the design. If one user sees a way for the system to work better, the system must provide them a way to modify it to their liking.

3.2. not interesting? why?

Meta-design, as presented does not really overcome the high-cost of learning associated with most EUD languages as presented early in the paper. The urban planning system is easy to use, but how easy is it to extend? Open Source, as described above, has its own set of problems relating to learning. Certainly meta-design is a step in the right direction, but I do not believe it adequately adresses the learning problem.

4. what do you consider the major technical and social challenges associated with meta-design?

4.1. technical

The major technical challenges would be in finding a set of features to map to the meta-design, which the users will be able to manipulate so easily while keeping the learning cost low. If the learning cost cannot be kept very low, it is then important to have a good set of documentation, so other co-designers can come along after you and follow your trail of work easily.

4.2. social

Depending on how low the cost of learning is, there will be different levels of learning involved. When dealing with something like an Open Source project, there are also the social issues which exist from having a leader.

5. describe your own experience with meta-design

5.1. with ideas associated with meta-design

I have contributed to a few Open Source projects, and currently work on the MANTIS Operating System, an Open Source project writing an Operating System for sensor networking devices here at CU. I have worked with Open Source projects as a designer, developer, user, and bug reporter. It is from this experience I am basing my concerns about the learning side of things.

5.2. with systems built on meta-design?

With the Linux kernel, the GNU utilities, GNOME, KDE, freedesktop.org, openoffice.org, and more there is an impressive amount of quality Open Source software, enough to run a completely Open Source computer system, which I do. I have pretty much exclusively used Open Source software for several years.