Recently, we here at delphic sage having been starting down the road to adopting jQuery as our client side framework of choice with regards AJAX and DHTML deliciousness. Part of this process included identifying which .NET controls we needed to replace with jQuery plugins. One of these that came up was the ASP.NET validators, which although highly effective and easy to use, offer a rather dated UI experience. I did a bunch of searching around and discovered these wonderful jQuery validators provided by bassisstance.de (Jörn Zaefferer has written many great jQuery plugins, check them out!).  These validators are just what we were looking for, but it didn't take long to discover that (A) they are not very .NET friendy because they rely on elements having a "name" attribute, and (B) they are not nearly as easy to implement as the asp.net validation controls from a developer perspective.

OK, So Let's Eat Our Cake, too

The jQuery validators look and behave greatly, but the asp.net validators are much easier to use. My mission here is to extend the asp.net validator controls to highlight and focus on invalid form fields in way that is comparable to what Jörn has done with his jQuery validators. The key advantage that I will try to attain is that I want the validators to be completely transparent to our developers when implementing in forms. They should continue to implement ASP.NET validation the same way they always have, and I should be able to upgrade the validators accross various legacy projects with very little effort. Because of this requirement, using inheritance to override the control behavior was out of the question since it would require developers to use new control tags and references for the new controls. Now that the goals have been defined, it's time to get hacky with some javascript.

In the last few years, major developments have happened in the application world that have enabled me to completely abandon many of the win32 client programs that I have depended on for years. I have been reading about this phenomenon for a long time now, but we seem to be at a significant tipping point. At first, I never really bought into the fact that that web applications could even compete with traditional win32 and mac desktop applications.

The Astronauts

Sometime around 2001, I was working on my first AJAX enabled web application and saw first hand the limitations of trying to do anything overly interactive within a web browser (heck, we were dealing with netscape 4.7 and IE 5 at the time). In fact, it wasn't even yet called AJAX at that time, I think we were calling XML Sockets or something to that effect. Anyhow, around this time many of the big thinkers (or architecture astronauts, as Joel calls them) out there on the internets were beginning to tout the end of desktop computing as we know it. They talked about how in the future, we would store all of our files and applications online, and that we could access them from anywhere. I think "Passport" was actually supposed to be a significant step toward this utopian computing ideal. At the time, I thought it sounded like a great direction, but I wasn't about to hold my breath due my experience in developing rich web applications and first hand knowledge of how terrible and inconsistent the world of web browsers was (and still is, although improving).

Fast Forward a Whole Bunch of Years

Here I am sitting at my desk, minding my own business, when across the room Todd gets very excited. He grabs his stapler from his desk, as well as a random stack of paper. After futzing with it for a moment, I hear the stapler do its thing, followed by a gasp of happiness.

T: Did you know a stapler can go the other way, so you can take the papers apart without ripping them?
J: Yes.
T: Did you know it was called pinning?
J: Yes.
T: How did I not know this?!

As it turns out, 3 out of 9 people surveyed did not know anything about the pinning action of a good, old Swingline.

We're learning new things here everyday, people.