IsComputerOn

Mark Hellegers Talks About Themis.

The Themis project (additional link with screenshots and icons here) has always fascinated me somehow. The concept is good, and with some work it can bring forth a fantastic BeOS-native browser that goes way beyond the capabilities of NetPositive, and is open source as well. I already knew Mark Hellegers, one of the browser's developers, from earlier BeGeistert events as he is one of the few core Dutch visitors. I think most of you would like to know a bit more about Themis and its goals. That's why I sat down with Mark for a short while, and under read more you'll see what came out.

If what you are about to read draws your interest and you are considering to join the project, drop a line at z3r0_one at users dot sourceforge dot net. IsComputerOn: So let's start with the beginning. How did the whole Themis project take off?

Mark Hellegers: Raymond [Rodgers] started the project. He wanted a more modern browser than was available at that time, and wanted a different approach than the standard, a completely modular approach with even separate tags being modules. But that last thing we dropped because that wasn't going to work with modern standards. I myself joined in after about a year, which was by the end of 2001. Now we have four active developers.

ICO: What's Themis eventually supposed to become?

Mark: A small fast BeOS-native browser.

ICO: Like NetPositive?

Mark: Yes.

ICO: And what will then be the real difference between NetPositive and Themis?

Mark: Themis will be more modular, so people can add their own parts to it, and Themis will support most modern standards, like HTML 4.0 and CSS.

ICO: So it will support plugins, will people also be able to throw in plugins other than parsers?

Mark: Yeah, you could think of Flash, Java, JavaScript, whatever one feels like adding.

ICO: Do you have a certain kind of release date, like do you want to get it out as a more or less complete version within a reasonable amount of time?

Mark: I think we can have something rendering on screen in at most six months, and to have something usable in a year. But this all depends on how much time we all have.

ICO: What are the delaying factors to the progress of the Themis browser?

Mark: Implementing the full specification of a protocol, for instance HTTP and FTP, or that of markup languages, like HTML and CSS.

ICO: So what you're saying is, the Themis browser will also support FTP?

Mark: We eventually do intend to, but we first want to implement all components so that HTML can be rendered. After that we will add other things, including FTP.

ICO: Have some things gone wrong already, or did you run into some kind of problem or difficulcy so far?

Mark: Oh yeah, plenty. SSL is tough to support on NetServer, getting the HTTP protocol to work on most web servers is also difficult, and parsing sloppy HTML files is a challenge.

ICO: I've just seen Themis parsing an HTML page, and it does take some time to do that. Do you think you can speed this up easily?

Mark: Well, you've seen it parsing on a really old laptop with unsupported graphics, which slows it down quite a bit. But it does still need to get faster. There are still some optimizations to do that will cut down on the parsing time, but exactly how much of a difference that will make, I don't know.

ICO: Final question. Are you letting yourself be inspired a great deal by other browsers, or do you really try to come up with ideas and solutions all by yourself?

Mark: Part of the fun of working on Themis is learning how a browser works, and I personally try to find a flexible solution to the problem I'm working on. How the other team members work, you'd have to ask them.