Frank at Google Earth Blog is already doing an excellent job looking at and explaining the new features of Google Earth, and so is Google Lat-Long blog, so no need for duplicate posting here. What you’ll find instead is some more esoteric observations, ranging from the philosophical to the mundane.

The philosophical:

It occurs to me that the individual improvements that comprise this latest revamp fall into two categories — let’s call them brains and beauty. I get a lot more excited about the improvements to the brains of Google Earth, though I realize that it’s beauty wich turns heads, and which people “want” (or “cling” to ;-).

How to define those two categories? Brainy improvements improve the quality of the information that Google Earth delivers, or improve the efficiency of access to existing information — they improve the function of Google Earth as an atlas. Beauty improvements, on the other hand, improve the function of Google Earth as a plausible mirror world, an ever-more accurate simulacrum of Earth.

Atlas or mirror world, what’s the difference? Their functions are in fact poles apart. Atlases filter out as much as possible that which is not information or which obscures information. Mirror worlds, on the other hand, aim for the accents and details that provide a sheen of reality — precisely that which atlases strip off in the pursuit of clarity. Atlases try to augment reality by pushing high-information content to the fore. Mirror worlds do not.

These two functions can conflict if they exist in one application, and indeed I feel Google Earth is acquiring something of a split personality with this release. I’m not really complaining; you can turn off many of the improvements that are propelling Google Earth to its mirror-world destiny, but it is still the case that resources are being expended by the team on making Google Earth pretty, sometimes even at the expense of clarity.

I’ll explain, but I’ll start by categorizing the most recent release’s improvements into brains and beauty:

Brains:

Revamped navigation controls

easily viewable acquisition dates

Street View in Google Earth

12 new languages

Flash support for Mac

Beauty:

More and better buildings

New sunlight control

New atmosphere

I think most of the brainy improvements are uncontroversially brainy, but why do I regard the new buildings and the new sunlight controls as cosmetic? Let’s take each in turn:

More and better buildings: It’s true that the rendering of 3D buildings is now much more efficient, and that there are a lot more of them, but why doesn’t this constitute a huge informational boost? Because the satellite imagery already tells us there are buildings in those places; there is precious little else added by a 3D representation without metadata such as: Are the buildings residential, office, factory? Who owns them, when were they built, how high, who built them, how much was paid for them, what businesses are inside? Do they have websites, do they deliver?

A lot of this no doubt will arrive down the line, but until then, the buildings are pretty rather than informationally dense.

New sunlight control: When NASA World Wind got a feature just like this last year, I remember biting my tongue lest my critique be seen as partisan, but I did not really regard that functionality to be something that increased the informational quotient of World Wind much; I feel the same criticism now holds for the implementation in Google Earth.

The sunlight control is useful in one specific way — to see where on Earth it is night and day at a specific hour. But when it comes to lighting landscapes and buildings at sunrise or sunset, I don’t see the value besides eye-candy appeal.

First of all, the sunlight control is a bit of a misnomer. We’re not actually seeing the effect of sunlight: Buildings and mountain peaks do not generate actual shadows on neighbors when the sun hangs low on the horizon. Instead, we get shading as an inverse function of how much a surface faces a light source we’ll call the Sun.

Shouldn’t I be happy that Google Earth isn’t going all out for realistic shadows but instead is giving us a shading tool, which we can use to tease out terrain features? I would be if the light source were movable to ad hoc locations other than where the sun can be — for example, if I could make it revolve around the sky at 10 degrees above the horizon at locations other than the North Pole currently, I’d be happy, indeed. (Do go to the North Pole and try it, though. I really works, there.)

What I’m trying to say is that the current implementation falls between two chairs. It’s not quite accurate in terms of mirror worldliness, but is still limited in scope by its mirror-world aspirations.

New atmosphere: One place where I know I disagree with Google Earth Blog’s Frank Taylor is that I like to be able to see the little rectangular strips of high resolution imagery across the face of Google Earth when zoomed out, whereas Frank would prefer the look to be seamless and realistic, and prefers the zoomed out Google Earth to look just like the real thing.

For me, however, those rectangular strips represent information. They promise a lot more information if I zoom in on them, and also hint that there’s something there worth taking a high-resolution image of. I feel that this kind of information should trump a realistic view of Earth. And while the new atmospheric look certainly helps with the realism, it filters out too much of the kind of detail that I feel makes Google Earth useful when zoomed out. Yes, you can turn it off, but it’s turned on by default. I feel version 4.2’s atmosphere had a much better information/realism balance.

Priorities:

So what would I have prioritized instead of buildings, sunlight and a thicker atmosphere? Well: A projection that is friendly towards the poles; 3D bathymetry; better tour creation support; GeoRSS support; better search filters. These functions would all tilt Google Earth back towards the atlas end of the spectrum. Of course, that’s not necessarily where the money is. Google Earth is ad-supported, and nobody searches for pizza at the poles or in the Mariana Trench. And when Microsoft Virtual Earth does things like add lots of new buildings and everyone applauds, it’s hard not to want to compete. But that’s okay; in the end I much prefer a free mirror world with atlas functions than nothing at all.

That’s all the philosophy I can muster tonight. Now for the mundane stuff.

The mundane stuff:

Bear in mind that this latest version is still beta, so some stuff doesn’t work well. On the Mac, I’ve found the keyboard controls to sometimes lock up, even on a new install on a new account. If you’re using GE for a must-succeed presentation, best to stick to what you know works.

Another reason to ease into the new version is that the keyboard controls have changed. [On the Mac,] Command + up-arrow or down-arrow previously let you zoom in and out, but have now been mapped to the new look-around control. Zooming is now linked to the function + arrows combo. It takes some getting used to, unless of course you use the on-screen controls or a SpaceNavigator.

There’s one new preference item: You can turn off the new auto tilting feature that you get by default as you zoom in by right-clicking and drag the mouse up/down or using the on-screen controls.

Another subtle change: You can no longer filter the default layers by Core/All/Active layers only. Considering that it was a little-used feature that could generate confusion, this simplification of the UI is welcome.

One thing I think is a bit of a loss: No longer do we get a precise percent figure in the status bar showing how far along the download of current view’s imagery is. Now it’s a growing circular arc that slowly grows into a completed circle. Frustratingly, the arc sometimes gets smaller!

Turn on layers or placemarks in Google Earth, switch to Google Sky and you still get to see that content, floating among the stars. Surely this bug can’t be that hard to fix?

Small gripes, really. To be honest, I’m having way too much fun watching and waiting for the next move in the Iran-Google war of wills. It sure helps that Google has little or no business interests Iran, and that it still has a lovely concession up its sleeve: Offering to make a localized Farsi version that omits the Arabian Gulf reference, in accordance with their “primary, common local” doctrine. (Just don’t call Farsi Farsi:-)