I'm constantly learning: programming languages, data science, machine learning, music. You see that sailing pic? I have no idea what I'm doing there. But, I haven't drowned yet, or wrecked the boat. So it goes with learning...
My beloved macbook air and pro. I have a number of workstations, both real and cloud based, but my macbooks generally run the show.
You can see my entire collection of software, tools, knives, and lint in my dotfiles repo.
In terms of editors, I'm a somewhat recent vim convert, and use it for Haxe. I've caught up with it pretty quickly, and have even written a nice Haxe plugin for it called "vaxe". I like the ideas and apis behind git and bitcoin as well. Great opportunities to do some hacking.
Most of programming is "gradient descent". You can iteratively improve things, but eventually you run out of cheap/intuitive gains. To move beyond that, I always look to nature and cognition: fractals, graph theory, information theroy, and dimensional scaling. Most of my work involves dealing with data, so I tend to draw more from the "data science" side of things.
There's a few folks that I follow ahead of topical interests. That means, if they do something, I'll check it out even if I'm not interested in the subject. To that end, Mike Bostock is a pretty amazing combination of coding data science, and aesthetics. His d3 library is great.
Tim Pope has some wonderful vim tools that I didn't even realize that I needed. Hadley Wickham, likewise, has some wonderful data manipulation tools that have made certain aspects of my work much easier.
I've done a lot of things I'm proud of with Haxe, but I'm not always able to share it as I wished. I've contributed to some of Franco's work with thx, and contributed a lot of plugins for the editors I've used over the years, including textmate and vim.
Not really, but I haven't tried very hard with my Haxe work.
I wanted to do some front end work that called for flash. But, I quickly learned that I hated flash. Haxe was an easy switch.
The language itself is simple if you have any experience with statically typed langauges. I put together a 15 minute guide here that links on to some other useful resources.
I ended up learning a lot from that mailing list. There's a ton of bright, accessible people there doing things I never thought was possible. Haxe ties together a lot of programming techniques from different languages and platforms. You may have to "unlearn" some things from other languages, but once you do a unified picture of web development becomes much clearer.
I would also recommend thx and polygonal ds for folks wanting to see how to write "idiomatic Haxe" to solve familiar data structure and algorithm problems. I'd also recommend tink for folks wanting to move beyond that.
Haxe has the capability to do some pretty amazing data/orm configuration thanks to its powerful macro functionality. I'm working on this a bit with postgrehx, but don't have anything killer yet. The Mongo driver from Matt Tuttle looks promising too.
I also think we need a solid cross-platform UI framework. It would also be great if the library provided the ability to utilize native UI where it was possible. Most mobile phone apps these days are just variations on the different UI sensibilities that exist in native Android/iPhone/etc. Most of the current crop of UI frameworks try to recreate the same UI on every platform. That's understandable, but not what most users on those platforms expect.
Some of the games have been amazing. Papers, Please was a great use of Haxe. Perhaps Haxe just lets the creativity of the author shine through.
A cross platform IDE suite for Haxe is about 75% there. We're missing good ways of refactoring, and some other common IDE utilities that rely on deep compiler integration. After that I'll be completely happy.
I also think we'll see some great test/deployment functionality. Andy Li has already done Haxe a great service by setting up a testing profile for Haxe on travis-ci.
Once we have all that, I think we will see Haxe developer working together more easily in business contexts, and a whole different side of Haxe will emerge.
I'll be talking about promhx, a promise and functional reactive programming library for Haxe. The library provides a way of dealing with asynchronous operations with values, rather than callbacks.
There's a couple of solid benefits to this approach. The results of an asynchronous operation are separated from the operations that rely on it. This lets developers avoid binding all of the logic for a conventional asynchronous operation inside a single callback. The second benefit is an error handling mechanism that allows for flexible error management, a common pain point in asynchronous programming. Finally, a full suite of utilities and methods enable developers to reason about asynchronous workflows without getting bogged down in syntax and callback indentation.
I get to meet all these European Haxe folks that I've known only through e-mail handles.