>> I am working on a javascript version of the hypergraph >> similar to what you suggest. The key advantage of a javascript >> implementation is that I can embed it in the new Axiom Firefox >> browser front end. This will allow me to dynamically look up >> information during the graph traversal. > >One advantage to using Java is that the HyperGraph applet can be run >inside of FireFox, Internet Explorer, Safari, Opera, etc. without >modifications (why limit one's self to supporting just one popular >browser when you can support them all?). The applet can also be run >directly from the command line without the need for a browser at all.
One primary consideration is the number of languages we need to support. There is currently no Java in Axiom. Java is a heavy component to add to the build dependency. Other than that consideration, I have nothing against Java. My prior open source project (pinger) was all java code. Javascript is a different story. Given that we're using the browser as a replacement for the C code in graphics and hyperdoc, it seems reasonable to use javascript as a replacement language. I'm learning how to dynamically draw images using javascript and the canvas objects to replace the current graphics code. That code will get extended to handle things like hypergraphs or other kinds of graphs. And the hypergraph facility will eventually get reflected back into Spad code so we'll have new ways of drawing things like charts, histograms, hypergraphs, etc. I have hypergraphs "sort of" working (the transformations and drawing all work). I'm trying to make the nodes richer so we can get better information at each node. So Javascript seems well integrated while Java is "on the side". As to the question of supporting "one browser" rather than supporting them all, I'm taking the same approach as the rest of the Axiom system. That is, I'm not claiming that the code is specific to Firefox (or Linux) but I am claiming that I know the code will run on the supported platforms. I can make that claim because I build it and test it. As usual, I expect that Microsoft will be problematic. IE does not support the canvas object so there is very little chance of an IE front end. I understand that there is a plugin for external canvas support in IE but I have not tried it. However, Firefox on Windows will completely eliminate the whole graphics/hyperdoc porting issue as well as unify the two into a standard front end. So I suspect that Axiom on Windows will be much easier to port in the near future. The whole issue hinges on available time. Tim _______________________________________________ Axiom-developer mailing list [email protected] http://lists.nongnu.org/mailman/listinfo/axiom-developer
