Uncaught Exceptions

Today I isolated a particularly frustrating bug. During the event creation process, clicking on any form field was producing an “uncaught exception” in FireFox.

uncaught exception: Permission denied to get property HTMLDivElement.parentNode

I finally isolated the problem to be coming from Matt Kruse’s Calendar Popup but could not, for the life of me, figure out what was causing it or how to get rid of it. Luckily, I remembered that I had seen the same calendar on MyPunchbowl, so I ran over there and grabbed their copy of the javascript file. I’m not sure what they did to it, but it’s no longer throwing errors. Thanks, MyPunchbowl!

IE6 Compatibility: Back for Now

Between the Peekaboo Bug and some issues with prototype, we were having some fairly annoying issues with IE6. A few hours of tearing my hair out this afternoon seems to have produced a solution. For now.

 

Thanks to PositionEverything.net and Prototype’s latest stable release, things seem to be fine.

 

The only problem now is that some of the effects cause a “Stack Overflow Error at line 0″ in IE6. I think this is more a function of my “standalone” version of IE6, but I’ll check on some older computers at the office next week to be sure.

Still Cruising

Lots of work going on behind the scenes, but not much to speak of. The site design is coming together and we’re putting the finishing touches on the business plan. New features are still coming slowly, although we’re going to be pausing that soon when we start to deploy the sexy new design.

One thing worth mentioning, though, is that my problems with opera and Script.aculo.us/Prototype have been cleared up thanks to a commenter who suggested that I download the version of prototype that was being used on the Script.aculo.us site. That cleared the problems right up. Hopefully the forthcoming release of prototype doesn’t contain these same bugs.

Late Nights and Snags

Things are still on pace for tomorrow night’s LiveBlog, although last night was a late one. Every developer has come across a time when something stops working for what seems like no reason at all. Being mostly logical people, developers realize that it’s not possible for code to work and then stop working if nothing has changed… but finding what changed or, just as importantly, why that change matters, can be a stressful, frustrating experience.

Last night I was having a problem where a javascript function would not complete, but it would also not throw an error. It just seemed to die, leaving half of the function uncompleted. It ended up being a problem with a javascript for…in loop. The for…in loop iterates over every member of an object… every single one. This includes inherited members and (I think) methods. I’m not sure where, exactly, this caused the problem, but it looks like there are parts of the prototype/script.aculo.us library that, if they fail in a certain way, fail silently.

I also ended up having to re-tool one of the display algorithms because it choked when using the display effects on a large number of new items. Those things seem to be working now, and features are slowly being implemented in preparation for tomorrow’s demo.