The idea of a worry-free computing experience is still obtainable, thanks to services that use Amazon S3 type methods of data storage. But the real value is in remote desktop support. I'm willing to bet this would make the idea of "software as service" suddenly quite viable for the non-enterprise and enterprise desktop Linux market.
Offering a "do it all" type of distribution isn't a bad idea. They main key is to simply provide better execution than past efforts while ensuring that the needs of many Windows users aren't being ignored. Considering the vast number of great applications available for the Linux desktop, I think a large number of existing Windows users could indeed make the switch with the proper support in place.
Linux brand peripherals
Printers, webcams and most other USB devices work great out of the box with today's modern Linux distributions. Sadly though, wireless chipsets remain hit and miss. This problem doesn't exist because of a lack of developer focus or a lack of driver support. No, instead this issue is ongoing because everyone buys from vendors that choose to support Windows users exclusively.
Imagine if Novell, Red Hat and Canonical got together and approached some big-name vendors to license a set of known-to-work chipsets in USB dongle format? The issue of compatible wireless devices would be solved overnight.
Remember that working chipsets do exist. It's just that companies use a revision number on top of a device type that then creates the problem. The same device type often comes with a new chipset. This makes the idea of hardware compatibility lists nearly impossible when it comes to wireless networking gear.
Will this alliance ever happen? No, not in a million years. It's much easier to have dated documentation point to poorly maintained compatibility lists with claims that NDISWrapper handles wireless compatibility just fine. Well, guess what? There are countless forum threads out there filled with individuals who might be inclined to disagree. Brand it, support it, and the problem can truly be solved in a duplicable manner.
Swallowing our pride and accepting reality
Many of you reading this are thinking "who cares?" because you have been able to overcome these issues just fine. And Im included in this list Ive found work-a-rounds and various ways of overcoming simple issues that should have been fixed years ago. However, despite the tremendous power the Linux desktop has to offer, there's always room to make things more fluid and logical.
I'm not talking about dumbing anything down, mind you. No, I simply want to see all of us decide that we either are going to start taking our platform seriously or opt to forgo the usual long-winded speech about how superior it is in comparison to the alternatives.
Why not provide a simple means of using already supported wireless chipsets? Why not decide that we're willing to reach out to people who want a Zonbu-like experience even if it means including less geeky users?
And lastly, let's stop buying computers only to then remove Windows and install Linux. This is the single biggest gripe I have. If you choose to ignore every other point, please at least hear me on this one. Build your own desktop PCs and buy your notebooks with Linux pre-installed.
If we're ready to stop talking and would like to put our money where are collective mouths are, perhaps more PC vendors would start taking us seriously in the future.
Remember, it's a numbers game. Even with the gains we've made over the years, too many of us are still buying products not supporting the Linux lifestyle. I happen to think it's pathetic and really needs to stop. Well, either stop or keep drinking the Linux Kool-aid about "community" and the "open source eco-system." It's put up or shut time, people.