The big news for Drupal this week is the relaunch of whitehouse.gov on their software. Somewhat more quietly, though, recovery.gov abandoned it. So, congratulations are in order -- but maybe also condolences?

I'm certainly happy to see U.S. Federal Government take open source software seriously (and commercial support from Aquia will certainly have helped.) However, seen from this side of the pond, that isn't particularly spectacular (we're used to a more active push for open source than that.) What's puzzling, though, is how recovery.gov was first launched on Drupal ten months ago; then relaunched on SharePoint several weeks ago. How do you go from nothing, to open source, to Microsoft; in one year?

I can appreciate the fact that with a change of government, there's a need to act fast. And there isn't really time to gain knowledge over several lengthy project cycles (make mistakes, learn, try again). But considering the amount of money involved, I wouldn't say doing several cycles very quickly is really an alternative. (The new recovery.org cost some $9 million to implement to date, certainly not as much as the $46m I've seen on Twitter, but still, a lot of money.)

This is not to say that going with Drupal is a mistake. Slate has pointed out some of the issues with the system, but looking at the other products we cover I think I could come up with equally caustic comments on each of them. And recovery.gov's choice of SharePoint would be an easy target, but we usually try to be a bit more nuanced.

The real issue, to me, is how you select software when you don't have time for mistakes. For a distant outsider to the U.S. Government like myself, the current process looks a lot like "we know this, let's pick that; oh no, it doesn't work, quick, let's use that, we already own it." Recovery.org may have swapped Drupal for SharePoint in a remarkable few months' time, but ambitions for the near future are to add "social networking capabilities." Is SharePoint 2007 really the obvious choice for starting in WCM, then building out to social? (I particularly liked OhMyGov's quote: "The government wants to go into the social networking world, but they don't necessarily know what that means.")

The more sensible shortcut would have been to read up on other people's experiences. And of course, as you'd expect, I'd recommend spending some days with our Web CMS Report and Social Software & Collaboration Report, rather than accelerated trial-and-error. Then again, can you blame me? We did try to warn them pretty early on...