Keynote: 5 a Day, and Session: Failing to Succeed
- I know that Robert Lefkowitz has a following, but I had never seem him talk before. He is a dynamic speaker, fun to listen to, and obviously knows his stuff. The keynote was an extended metaphor about how open source is like tomatoes. I cannot possibly give it justice in a blog post, so I won’t. The session on failure was also very interesting (and packed); it is important to note that he’s not giving the trite “failure is ok” mantra that I sometimes hear from some speakers. It was a much more nuanced “failure is ok, as long as you identify and acknowledge the failure quickly, and learn from it”. I’ll try to blog more about this later.
- Session: Python in Mozilla
- Mark Hammond gave a straightforward summary of how Python could be used in Mozilla. There wasn’t really anything I didn’t know: basically the deployment problem is mostly unsolved, but there is at least a plan to have a single, separate Python install for the Mozilla runtime. We’ll have to figure out how to make this work in practice.
- Lunch: Mark Hammond and other Python developers
- I didn’t realize the extent to which, among some, python is a religion, not just a programming language. I must say I’m not impressed. Mark had a much more pragmatic approach than others at the table. And the animosity towards JavaScript was astounding, considering the extent to which, when coded properly, Python and JavaScript have common design patterns and object orientation.
- Session: Building Internet Applications with Mozilla XULRunner
- My own talk went well; there was a good and varied crowd, and they asked good questions.
- BOF about the Mozilla platform
- We had a small group of people interested in Mozilla, conversation wandered from Calendar and email apps to the use of the JavaScript language.
This entry was posted on 28 Jul, 2006 at 3:05 pm and is tagged with , oscon06.
Trackback.
July 29th, 2006 at 9:25 pm
Ben, can you explain the state of Python on Mozilla a bit more?
I had always thought that the goal was to be able to script xul in python and bundle pyXPCOM for Firefox 3.0 and xul-runner.
Is that not the case?
Thanks
-Jed
July 31st, 2006 at 3:27 am
into an ambitious proj…wud luv to get ur views on it…
gimbalz
A web application platform, presentation and integration system which will give new dimension to WebApps correlation, usage and viewing…
A browser killer application…
Web browser UI is not meant for accessing WebApps…
Web browser doesn’t provide synchronization among WebApps…
Gimbalz provides the solution to the above problems and much more…
Gimbalz enables WebApps to dock…synchronization among WebApps thru API…central repository storage option…cutomization…
A small summary of the plan…firefox platform can be reused with essential changes as the model to build a new WebApps friendly UI and docking station…develop an effective WebApp API for better synchronization, control and flexibility…
In a world where the future we all know would be WebApps…we need a platform for WebApps like we had with Linux to tackle proprietary dominance…
Developers plz dont think its not fair to develop for a new platform…docking XUL Apps can be achieved with minimal effort…some of the existing firefox extension can also be reused with little effort…the syncronization among WebApps only will take extra effort for the WebApp developer…but synchronization among WebApps is something which will work wonders…
A seizable community of firefox developers,extension and theme writers do exist…if we can all work together and build Gimbalz…I beleive this would be the beginning of the next war(WebApps)…after the much hyped OS and the browser one…
if u have any queries…plz do get in touch with me thru get2navin@yahoo.co.in or navinsylvester@gmail.com…
Lets build the next Linux…
July 31st, 2006 at 11:46 am
Jed, the Mozilla 1.9 codebase has changed so that we can support multiple scripting languages in chrome pages. But we are not going to be shipping an entire Python runtime, for sure. We may end up shipping some glue code that downloads a Python runtime if necessary, or this glue may live in an optional extension.
Navin, your proposal is short on detail and long on rhetoric. How is your Gimbalz platform different from webrunner (or is it merely an extension of the webrunner concept)?