My iPad ship notification came early this morning, and I along with about a quarter million other overgrown little boys and girls started anticipating the early (late?) Christmas Day coming this Saturday.

I'm excited, but not because I have a portable TV to look forward to. I'm also not excited for the price inflation that app vendors (who have barely scraped by at $0.99 and $1.99) are going to try to foist upon us with the new form factor. In fact, I'm not excited to buy any apps at all.

What I'm most excited about is seeing how well an A4 optimized Mobile Safari can behave as a full featured application host— for the only type of apps that will actually matter long-term on the iPad outside of the mail client: fully native cloud apps.

Unlike the iPhone, we don't have to worry about the poor DOM support for things like the camera and the accelerometer, or the coarse and kludgey location support that the browser provides. Those things make a ton of sense for a device which is always with you and frequently needs to be used from hostile environments (at weird angles when you are not sure where you are). In the iPad's case what will matter is the ability to get near native speed and stability from the browser for the following things:

1. Multitouch. Mobile Safari for the iPhone has supported multitouch for a long time, albeit a bit too slowly especially when compared to the speed of the native app gestures. I sincerely hope that the A4 chip and the larger screen area (thus larger digitizer) closes this gap.

2. CSS3 transforms. Productivity style work doesn't need to be cinematic, but there is a certain amount of eye candy that can make a big difference. Dynamic highlights, accordion effects, etc.

3. Clientside storage. This is good enough on the iPhone. All they need to do is not break it (and maybe move the default upper limit on SQLite above 5MB)

4. Offline support. App-cache works on the iPhone, but it is a little squirrely sometimes. In order for the iPad to replace a laptop it needs to be rock solid on planes, trains, and automobiles (in tunnels) which means this needs to work more reliably.

5. Websockets for true bidirectional communication with a server for all of those emerging realtime apps. Ok, this one is a total pipedream and the truth is long polling works pretty well on the iPhone, but one can hope.

There are rumors floating around the Internet that Apple is looking to break all of this stuff in Mobile Safari (in order to encourage developers through their tollgate), but I doubt it. More likely the worst case scenario is that like any big company, Apple is a multi-headed beast— albeit one that is often tamed by Jobs— and as such, there will be a pocket of cluefulls that will do the right open web thing and let Safari make the iPad the first truly novel cloud access device.

We can hope.

In the meanwhile, if you're going to hold an iPad hack party in the coming days, consider targeting the partial HTML5 implementation on its Mobile Safari instead of falling victim to the sirensong of Interface Builder and its jazzy drag-and-drop goodness. You'll be doing yourself a huge favor when the other large OEMs start pouring tens of millions of cheaper Linux and Windows-based tablets into the market over the coming years.

And Merry Christmas.