The GPS is my favourite gadget on the G1, smooth integration with Google maps is great. I’d love an Android netbook, esp. if it had a built in phone!

Android is all about users customizing their devices. And for those users that are not capable or willing to do this there will be pre-fabricated compositions or templates to make this task painless. In addition to Android components ported to TVs and Vehicles I also expect to see the the Java Applet reincarnated. Users will not have to learn to jump through new hoops for every device they own.

Also, as a developer I don’t care if my Java code runs on a JVM or on Dalvik. And don’t count out anything. Skype announced they are coming out with an Android client.

as for your other questions:

“why does Ubuntu feel the need for MID and UMPC specific projects?”

probably because the UI needs improvement anyway, and this an area particularly well suited to it. i’d still rather issue tweaked flavors of linux than try to recreate its application library.

“Why are Mozilla doing Fennec?”

i don’t see this as relevant, given that netbooks are – in all likelihood – going to be running regular FF rather than Fennec.

true. but the only organization i’ve seen that does a really good job with UX is Apple; Android hasn’t really distinguished itself in that regard, from the reviews i’ve read. is it easier to use than say Netbook Remix? sure. but it’s generally been implemented on handsets, so the design scope is a wee bit narrower.

“It isn’t nearly as appealing a market for developers of new applications (and that’s leaving aside the superior tooling and modern APIs/language of Android.)”

if we’re talking about the handset market, i’m with you. but i’m not sure reinventing the wheel with regard to desktop applications is all that attractive.

“I’d also note that iPhone has the same basic problem (substitute ObjC for Java) but it already has more interesting end-user oriented applications than (say) gtk or QT do.”

i view the iPhone as fairly unique here, and i think the development kit and, perhaps more convincingly, the app store, have been successful primarily because of the device. i’m not sure that the iPhone’s 10K strong application base is going to be duplicated any time soon by another product, because the device itself simply occupies a space of its own in the market.

If netbook on the other hand implies some combination of: ARM processor, good battery life, small yet high-DPI touchscreen, instant-on/sleep, small and/or soft-keyboards, highly mobile and pocketable device, location aware via GPS, permament internet connection via 3G and a range of price points then I see plenty of exciting opportunities.

Obviously this moves into iPod Touch and Nokia N800 territory but on the other hand the people who led the netbook charge seem to be reatreating back into boring Windows laptop-land. Possibly because kinda-cheap, kinda-underpowered Windows laptops turned out to be a good yet unexpected market for them.

Optimistically though I see several signs of others entering this new market where rule number one is no windows and the opportunities that arise once that leap has been made.

Looking at it from the other direction: why does Ubuntu feel the need for MID and UMPC specific projects? Why are Mozilla doing Fennec? At what point does the effort of rewriting things for the unique UI, hardware features and use cases of these devices meet the cost of starting from the other end and building on what’s provided by Android or Maemo?

Handheld platforms are different. Too different to hope desktop apps can just be ported over.

Not sure about netbooks tho (don’t have one). Seems like I want to run desktop software on those puppies, not handheld.

OTOH, I suspect there will be some interesting Android apps that I’d like to run as first-class applications on my desktop, just like I run other little gee-gaw apps on it today. Which you can’t do. Running in the emulator doesn’t count.