OMGIF! An Animated GIF Encoder in JavaScript
Dean McNamee wrote an animated GIF encoder in JavaScript, working anywhere JS runs including the browser and Node.js. You can add multiple frames and encode your GIFs right in the browser. Woo!
MEGA’s JavaScript Encryption
You may or may not have heard that Kim Dotcom’s new file hosting site MEGA (née Megaupload) has launched over the weekend. They are claiming that it uses a lot of new HTML 5 features (and suggest you use Chrome with the site), and also support widely touted encryption support. I had a look through the JavaScript source code to see what was up and it turns out they really are using a lot of cutting edge stuff.
First, I noticed that all of the encryption is being done client side, in JavaScript before uploading and after downloading the files from the service over XHR. There has been some discussion about how secure the service really is, but it is interesting to see cutting edge JavaScript used on such a high profile site. The encryption algorithms are all implemented using JavaScript, and they attempt to capture entropy using your keyboard and mouse movements while you’re on the site. It doesn’t use the Polycrypt library that I posted about a few days ago, but once browsers implement the WebCrypto API, I’m sure MEGA will be one of the first major sites to adopt it.
As for other HTML 5 features, they’re using the Filesystem API (only available in Chrome ATM) for local file caching during uploads and downloads, as well as the drag and drop API, the FileReader and writer APIs, and more. Check out the source code and the site for yourself over at mega.co.nz.
LaTex In JavaScript
As I mentioned on Twitter over the weekend, the popular LaTeX typesetting program has been compiled to JavaScript with Emscripten. Texlive.js takes a LaTeX file as input, compiles it, and then generates a PDF document. Because it uses really large data URLs, it actually crashed Safari, so check it out in Chrome which seemed to have no problems for me. It uses WebWorkers to do the compilation and generation in the background.
When you hit the compile button, it actually downloads all of the fonts and other resources that it needs from the server so it might take a little bit of time depending on your connection. It actually downloads the LaTeX source code for each of the packages being used (the code that actually generates the PDF, reads in the font files, etc.) and compiles those packages to JavaScript on the fly, before actually running them to generate the output document. This means that you could use any LaTeX program on the fly, without first compiling it with Emscripten: only the LaTeX compiler/interpreter itself needs to be compiled with Emscripten. I’m a little fuzzy on the details, so if I got anything wrong or missed anything major please let me know in the comments!
More
Ariya Hidayat, who wrote the impressive Esprima JavaScript parser in JavaScript, has blogged about a JavaScript Code Complexity Visualization which generates really pretty graphs showing the complexity of your JavaScript code. It uses JSComplexity (which uses the Esprima parser) to do the heavy lifting and Plato to do the visualization with Raphael. You can see an example output report for jQuery here and more linked from the Github page.
Marijn Haverbeke has blogged about an error resilient JavaScript parser/static analysis tool that he’s working on to help enhance a JavaScript code editor with better errors, autocomplete, etc. It’s an interesting read, so check it out!