the Design Experience Weblog Archive

Ted Shelton comments on Lawrence Lessig's talk. He argues that Lessig is wrong about copyright holders having to divulge the method of creation, that is, disclose source code for software after the copyright limit expires. Shelton says that Coca-Cola has never been required to release the "source" or recipe of the popular beverage. This is wrong because copyright is only granted for published works. The recipe itself is the product, not the beverage. If the recipe were published, they could not stop you from producing the drink. Software is instructions for a computer just as a recipe is instructions for a chef.

Also I am still annoyed at everyone who says that there are viable software products over ten years old. Sure, the original Microsoft Word may have been released over ten years ago, but every time a new version of the source code is made, there is a new copyright. In fact you could copyright every code check in seperately if you really wanted to. So yes the copyright on Word 1.0 might expire after ten years under a revised copyright rule, but so what? Who would that really hurt? Microsoft would be ten years ahead by then. The purpose of copyright is to establish a rich public domain of knowledge. All I know is I would never have learned to program except for reading source code. It wasn't the only thing I have done, but without that little extra, I could never have made the progress I have in the last yaer.

I started this weblog so I could practice writing. And now that I am so involved in open source programming projects, I have something to write about. That is documentation. OpenACS needs help in many areas of its documentation. Mainly because the source we took it from was never quite documented. Now that the code is in decent shape a documentation effort if very important. So, very soon I will be posting snippets of docs in progress that I am working on. First up will be an expanded guide to using ad_form.

Besides the documentation, I will be writing some design ideas for packages I am working on, mainly ETP and Survey, although you never know what will come up.

Here I am, back from being way to busy to even remember that I have a personal web site. Since the middle of July I have taken on two rather time consuming, interesting and altogether educational projects. The first was rebuilding a survey package for Sloan/MIT to work with their new dotLRN system. The second project was the migration of openacs.org from OpenACS 3 to OpenACS 4. This is still ongoing, so if you follow that link, don't expect much. For now the new stuff is at dev.openacs.org (expect that link to break when we finally switch over.)

Both projects were distinguished by the great people I got to work with. I can't say enough about the people at Musea Technologies, Sloan, the OpenACS hackers who worked on the new site, and the entire OpenACS community that supported us during the migration.

The OpenACS project is interesting because of the simple, organic way it developed. I was chatting with Roberto Mello on the #openacs irc channel and we decided to start working on migrating the site. Roberto set up a CVS repository and AOLserver and we were ready to go. From there we took some sample migration scripts and the original design work Musea did for the redesign of the site and started working. Along the way we got volunteers to help out with the design and migration. I ended up "organizing" these volunteers. I never expected to become a project manager of a major web site migration, but that is what happens when you decide something needs to get done in a volunteer organization.Unfortunately, after the excitement of a new design wound down, the "fun" stuff was all taken. I got really busy finishing up the survey package and the new site is still almost ready to go. Hopefully we will be able to lauch the new site very soon.

Working on the survey package I learned even more about building OpenACS 4 packages and also learned a great deal about the new dotLRN application. dotLRN includes a new portal system that will eventually make it back into OpenACS. I also learned that most of the really cool and powerful tools in the OpenACS toolkit are hidden or unknown. One thing we need to do is get better documentation of the APIs that are available to a developer. All the knowledge of these features is currently locked up in the brains of the few developers who have worked extensively with the toolkit.

I also learned about a new form building tool in OpenACS called ad_form. I promised to write a document explaining how I used it in the survey package. Ad_form saved me many hours of work developing the new forms. It is very easy to create templated forms that interact with the database. I'll write more soon.