In this post, cloud computing is defined as the use of web applications such as Google Docs.

I’ve written a post about why cloud computing is dangerous.

I’ve written another post about why cloud computing is dangerous.

And now I’ve written EVEN another post about why cloud computing is dangerous, and that post is the one you are reading right now.

But regardless of my opinions on cloud computing, I felt that completely abandoning cloud computing or creating one’s own miniature cloud would be too difficult for nontechnical users. Perhaps the nontechnical users should avoid trusting random startups with data, but keep using the services of larger providers, like Google. Such larger providers couldn’t possibly be a poor choice to keep one’s data, right?

Well, apparently I’ve been proven wrong.

A security flaw has been found in Google Docs. Don’t get the idea that this is a small flaw; a minuscule crack in a great wall. Sticking with the wall metaphor, the security issue with Google Docs is a massive, gaping chasm where a wall should be. Sure, Google has well-read engineers that can quickly patch up security holes, but apparently the same excellent engineers could not have patched up the security issue in the first place¹.

This raises an important question: If Google’s world-class engineers cannot prevent a user’s data from being stolen, who can?

Well, actually, anybody can prevent a user’s data from being stolen. Not putting the data online (even in a private online place) results makes it far more difficult to steal the data.

Granted, data often has to be put online for collaboration purposes, but there are still more secure ways to hold important data online. For a few USD a month, one can obtain a shared web hosting plan to install wiki software on². The shared host should also provide support and features to help one secure his or her data³. Keep in mind that with shared hosting, one’s data is still on far-away servers, but one gets partial control of the system.

Now, I would like to finalize in saying that there is nothing wrong in using Google Docs or even a random startup’s cloud computing services; neither must be completely abandoned.If you are fine with anybody in the world reading what you put on far-away servers beyond your (at least partial) control, and are prepared for the unlikely situation that you lose your data, there is no harm in cloud computing.

¹ I would like to point out that the engineers are Google are far better programmers than me, so this should be considered as an insult or a claim of superiority. I deeply respect the Google engineers, but do not think that far-off servers are a good place to store one’s data.

² It is possible that there is a security flaw in the wiki software that one may put on a shared hosting server, but choosing open-source wiki software and using good security practices should hopefully mitigate this.

³ Granted, with a shared hosting plan, it is possible for the system administrator to look at your files, but the risk of having your data stolen is still lower with a reputable hosting company than a random startup.

So, a few days ago, I got a spankin’ new T-Mobile G1. This has resulted in a lot of hands-on experience with the user interface of Google’s Android platform. I can’t help but compare the Android OS to Openmoko Linux. After all, both use Linux kernels.

No offense to the Openmoko developers, but I think that Openmoko pales when compared to Android. Below are reviews of two Openmoko user interfaces.

OpenMoko Train Wreck from Dave Fayram on Vimeo.

More OpenMoko Train Wrecking (Now with Qt!) from Dave Fayram on Vimeo.

The main problem I see with Openmoko is that it has attempted to borrow too much from desktop Linux. For example, Openmoko uses the X11 protocol, which is far too bloated for a mobile device. Because Openmoko borrows so much from desktop Linux, some of the problems of desktop Linux have been inherited by Openmoko, and the less-powerful hardware of mobile devices makes the problems worse.

In contrast, the Android platform uses more of their own code, and I think the results are visibly better.

At the same time, there is talk of Android on netbooks, because Android beautifully does many of the things that you would expect from a typical netbook.

Combine all of this, and you get an interesting question.

What if, Android, having shed the problems of an Openmoko-style phone operating system, overtakes traditional flavors of Linux in netbook marketshare?

It could happen.

If it does happen, could the same Android-style wheel-reinventing bring Linux more marketshare in the area of more powerful laptops and desktop computers? After all, didn’t Apple do the same thing by creating Mac OS X from FreeBSD?

Perhaps desktop Linux needs to see the same Android-style revolution. Android itself will probably not expand beyond the netbook.  It remove the need for “creative” methods to bring a system to visual consistency.

I believe open source software can easily be superior than a costly, proprietary equivalent, but perhaps the fact that many communities are involved in working with desktop Linux has been harming it’s ascent to the common user’s desktop.