wifi shield woes - should i just do xbee instead?

I’ve got several project ideas that involve building appliances or ‘art objects’ that will be in my home and need an internet connection. I’ve fiddled with the ASyncLabs WiShield (now defunct so not a practical option anymore) and now the WiFly which if you read some posts has become a very long nightmare for many smart people. I fired one up last night and so far it’s not working properly. And this is just the plumbing for my project, not the core of the project, so I’m spending all this time messing with the peculiarities of these chips, software that is only in Alpha, WPA-PSK, screwing with my router, etc. I feel like I’m in the stone age and I haven’t even reached the core business logic and hardware of my project yet!

So my question is: can I expect a relatively easier path by ditching wifi altogether and instead going with an approach that uses 1) on the ‘base’ side an arduino with an Ethernet shield and a XBee, and 2) on the wireless end an arduino + XBee? Yes, it’s probably more expensive but at least I can get a project finished using proven stuff, right? And the XBee will perhaps beat out wifi for range and robustness, right? I can hardly get a decent wifi signal in my basement. Thanks for any comments on this!

Do you really need an internet connection, or is wireless connectivity the issue? If the latter, you should check out the Pololu wixel (if intermodule range of up to 70 feet is sufficient). They are cheap, reliable and the available apps seem to work well. In addition, the folks at Pololu are extremely responsive to questions posted on their forum. See http://www.pololu.com/catalog/category/110

I really need Internet, I am retrieving things like weather condition data.