4 ms·

That blog post is from a year ago. In the intervening time, Asana's actually moved away from developing Lunascript. [1] Instead, they're just coding in Javascri

by aston 16y ago

That blog post is from a year ago. In the intervening time, Asana's actually moved away from developing Lunascript. [1] Instead, they're just coding in Javascript. There's a lesson in there somewhere [2] [1] http://www.quora.com/Why-is-Asana-developing-their-own-programming-language-(Lunascript)/answer/Jack-Stahl http://www.quora.com/Why-is-Asana-developing-their-own-progr... [2] http://www.joelonsoftware.com/articles/fog0000000018.html http://www.joelonsoftware.com/articles/fog0000000018.html

jackstahl 16y ago
:P It's important to note that we abandoned (or "put on hold") Lunascript, but not Luna. The vision for Luna always revolved around making it easy to write bug-free, real-time, responsive, collaborative, interactive webapps. The syntax of Lunascript over directly using the framework in Javascript was always a nice-to-have. I'm not denying your lesson, of course. To this day we still have some Lunascript in the codebase, and it does create some developer friction that costs real time. In retrospect, I'm sure it seems obvious that we should skipped the language and just built the framework. However, it's important to understand that the language itself was built mostly by Justin very early in the company's history, and basically no work was poured into it after we hired employees and built out the product. One framing of Lunascript is two guys taking a few months to try a few big ideas out before building a company they want to do the right way. We always knew we would have to compromise on ideals once we moved forward, but we wanted to see what stuck. I think a lot of the reason we're under scrutiny here is because people look at the time since D+J left Facebook and are like "what did you guys do with all that time?" The answer is the company didn't get started in earnest until almost a year later when we hired Greg and Malcolm.
alttab 16y ago
Polish and strap this baby to Node.js and we got something interesting! This functional model is something I've been striving towards in Rails, but over the years I feel the "rails way" more emcompasses development idioms instead of application idioms.

by aston 16y ago

jackstahl 16y ago

:P It's important to note that we abandoned (or "put on hold") Lunascript, but not Luna. The vision for Luna always revolved around making it easy to write bug-free, real-time, responsive, collaborative, interactive webapps. The syntax of Lunascript over directly using the framework in Javascript was always a nice-to-have. I'm not denying your lesson, of course. To this day we still have some Lunascript in the codebase, and it does create some developer friction that costs real time. In retrospect, I'm sure it seems obvious that we should skipped the language and just built the framework. However, it's important to understand that the language itself was built mostly by Justin very early in the company's history, and basically no work was poured into it after we hired employees and built out the product. One framing of Lunascript is two guys taking a few months to try a few big ideas out before building a company they want to do the right way. We always knew we would have to compromise on ideals once we moved forward, but we wanted to see what stuck. I think a lot of the reason we're under scrutiny here is because people look at the time since D+J left Facebook and are like "what did you guys do with all that time?" The answer is the company didn't get started in earnest until almost a year later when we hired Greg and Malcolm.

alttab 16y ago

Polish and strap this baby to Node.js and we got something interesting! This functional model is something I've been striving towards in Rails, but over the years I feel the "rails way" more emcompasses development idioms instead of application idioms.