Posts Tagged ‘programming’

Dropping a Stone and Watching the Ripples

My post yesterday on the Flash v. HTML 5 “debate” managed to get this blog a lot more eyeballs than it’s had in years – probably since 2003 if I’m honest. And yeah, that was kinda fun – I haven’t really stirred up a hornets nest like that in a long time. What I really appriciated was that some people actually wanted to TALK about the issue and just yell over each other. I actually learned a lot from the discussion – both in terms of better understanding other’s views and in finding out about some pretty damn cool projects.

On the views side – I heard from a Flash “hater” that was very, very clear in why he disliked Flash down to a use-by-use level. He’s on linux, and while that is a distant minority platform, his experience with Flash has been appalling and I totally understand his frustration. When Automata builds Flash-based projects we always work to provide the best solution we can for users who can’t or don’t want to run Flash, but this is far from a universal technique. To see or do many things online you just HAVE to have Flash – many times when it, quite frankly, makes no damn sense (I’m looking at you Chipotle, and no, you don’t get a link). I think we, as Flash developers, need to advocate better for the PROPER use of Flash – since quite frankly, every bad use of Flash makes all of us, and our favorite platform look bad in the eyes of a growing number of users.

On the project site, one of the most interesting projects that was brought to my attention was RaphaĆ«l – a JavaScript library that wraps up the various vector drawing/animation functionalities present in modern browsers (VML in IE and SVG in everything else) into a single library that works just about everywhere. It looks quite easy to use and it is quite suitable for production work (as evidenced by this site for the the new Nissan Leaf).

Opening a Can of Bugs

I spent a good amount of time tonight working on getting a new release of HYPE ready. I quickly smashed through the bugs reported up on GitHub and even managed to track down a nasty series of three bugs that were all hiding behind one another (I fixed bug A, which exposed bug B, which, when fixed, exposed bug C!).

The clencher though was when I discovered that a null object error that would only crop up when I would save a file out from my test SWF – and even then, it wasn’t consistent. I was able to fix the bug, but I haven’t been able to nail down exactly what was happening. As best I can tell, opening a FileReference dialog interupts the normal process of the garbage collector and it appears that the bug was causing my code to be dependant on the rapid cycling of the garbage collector to hide it’s mess and prevent it from throwing runtime errors.

And after all of this I just discovered a particular edge case that breaks the RLE compressor I wrote for my TGA (Targa) image exporter. Ugh. Gonna have to deal with that one in the morning.