Project FlaXe update: RIP Project FlaXe

haxe logoBack in November, just after Adobe announced they were donating Flex to Apache, I posted a suggestion for a Flex salvage project. This project would involve porting the Flex AS3 code to haXe and using the latter’s ability to target JavaScript to turn Flex into a framework that could run on HTML5.

The project was called “Project FlaXe”, for haXe’s creator (Nicolas Cannasse) pronounces haXe as “hex”, thus FlaXe would be pronounced “Flex”. Sadly it was a joke that almost no one got. Far more people worried that FlaXe implied it would be flakey. This taught me an important lesson: don’t suggest a joke name for something, as many people will worry more about the name than the viability of the project!

I spent some time during late November and through into December experimenting with haXe and existing AS3 to haXe converters as the obvious start point to the project. Here I learned another important lesson: rudely barging in on an existing language community, criticising many of its features and demanding changes to make it better for a large project like Flex, does not win one friends! I won’t go into details, but there are aspects of haXe that cause big problems for a Flex port. The upshot of this was that by Christmas, I’d lost all enthusiasm for using haXe as a replacement to AS3.

During December, Apache formally accepted Adobe’s offer to donate Flex to them and the Adobe Flex podling was born. The email list for this project has been incredibly busy, which suggests there are many in the Flex community keen to keep Flex alive. Something that I didn’t expect, but was very pleased to see, was that many on the list support the idea of porting Flex to JavaScript. The route we will be investigating will be modifying the new Falcon Flex compiler to output JavaScript, rather than flash player bytecode. A very handy blog for reading more details on how this might be achieved is Bernd Paradies’ “All my circuits.”

There are still potential problems with targeting Flex at JavaScript directly, not least being the disapointing snail’s pace at which Adobe are fulfilling their promise over what will be donated to Apache. In the last two months, we have received a semi-working copy of the Flex 4 SDK only. There is a real risk that Adobe will backtrack on handing over the automation classes in Flex. There is even a slight risk that they’ll not hand over the Flex name. The tools, test frameworks, Flex 3 etc are still nowhere to be seen and could be month’s away from delivery. At best, we might receive Falcon sometime in the summer. These delays risk people losing interest in Flex the meantime. However, these worries and delays aside, I think the Apache Flex initiative remains the best hope for the future of Flex.

So to that end, I intend putting my “Future of Flex” efforts into Apache Flex and so Project FlaXe has shuffled off this mortal coil…

One thought on “Project FlaXe update: RIP Project FlaXe

  1. This would be a great news if Flex SDK is ported to JavaScript with the support of the team under Flex Apache. I would definitely considering using it. Thanks for those infos David.

Comments are closed.