Google and the Open Wireless Networks Goal

This morning, as I drove 10 miles out of my way to avoid an accident that wasn't reported on traffic radio (which is a gripe for another day), I tuned into NPR for Morning Addition. I love public radio, because the news is probably the least biased that I have ever heard (yet still very biased). It's refreshing to hear about events that most major news groups would see beneath them, such as the existence of Rose Shows.

This particular morning they had This article on Google's efforts to urge the FCC to open wireless networks to more competition outside of the "Big Four". The reason behind it is sound: In the wake of a possible failure of the Net Neutrality situation, Google wants to secure the future of wireless networks by making their applications, search engines, etc. available for everyone who want to use it. It's a boost for business in a growing market, and provides good PR for Google.

Another really good reason I heard Google mention on Friday was the need for real competition within the Wireless Network community. Right now, you have the Big Four: Sprint, Verizon, T-Mobile, and AT&T. There are other, smaller networks, but they are marginalized by these big competitors. A lot of it has to do with existing network presence, limited radio bandwidth availability because of funding, and so on. Sound familiar? Ma Bell had the same setup, and continues to have the same setup with service monopolies in areas where they own the infrastructure.

So what's the answer? provide a public infrastructure. Google is trying to convince the FCC of this through their auctions. It would force the winning bidders to provide access to these air waves to allow other networks to utilize them. Hence better network support, better customer service, better consumer experiences.

Sounds very similar to the arguments that initiated the UTOPIA project in various Utah cities. And the same arguments are being made against Google's initiative: Bidding will not go as high, and the FCC will take a hit on how much money they will make, ultimately hurting the tax payer in the short run. So Google offered the answer: Guarantee a bid that will be attractive to the FCC and make the potential loss go away. The same attempt was done by various businessmen in Salt Lake to convince the Salt Lake Council to go with UTOPIA.

I wish Google success in this endeavor. Perhaps they can create a change in the wireless community that will make me want to have a cell phone. Perhaps they can convince an increasingly consumer-hostile business to release their strangle-hold on the market to allow for real competition to emerge, and by doing so provide an air of innovation that will make wireless a network worth innovating in.

Mac OS X Server 10.5: A Nice Setup

June is coming up rather quickly, and that means that the World Wide Developers Conference for Apple is coming as well. Now, while I don't usually get excited about trade shows, this one is different: Mac OS X 10.5 is said to be announced. There are a lot of reasons to get excited about 10.5, but I want to focus on the Server. That, in my opinion is more important than the client upgrades.

But what's so great about 10.5? 10.4 already has a great Directory service, integrated with Kerberos, and has a lot of functionality that can be easily implemented within the service itself. It also supports SMB, AFP, NFS, and FTP natively, allowing for easy file storage both within the network and outside of the DMZ. So what makes the new implementation so exciting for me?

Well, it's best described by John C. Welch from InformationWeek. In his article Inside Apple's Leopard Server OS, he gives an outline based on Apple's own information about their new implementation of server. Here is how it boils down for me:

iCal Server
One of the best products that Apple could have implemented was an iCal server. Not only does it allow for resource and calendar sharing, but it provides the information as part of an open standard set by CalConnect, which pushes for open standards within a calendaring server. It also allows for server clustering out of the box, so there is no additional fees to set up a Calendar server. Add that to no per client fees, and easy setup, and you have one powerful tool in your hand. Once we get the latest upgrade to 10.5, we already plan on setting up a test server with the iCal system running.

Wiki Server
I haven't jumped on the Wiki bandwagon as of yet as far as collaboration, since I don't need to collaborate much in my current position. I did work with a type of Wiki at eBay for an internal knowledge base and document server, but it wasn't anything fancy. But, because the server can be integrated with iCal Server, it can provide details about requirements on a specific date. This would be invaluable for Technology classes, because you can link an existing class requirement entry in the Wiki with the class date. Now the support staff know exactly what needs to be supplied for the class based on a single entry. I could see a lot of educational institutions utilizing this method for their support staff.

iChat Server 2
This new implementation of iChat allows for a hookup to other Jabber services (like Google Talk) without actually signing into Google Talk. It also is rumored to have Kerberos support built in, so single sign-on is more of a reality with a chat service. Why is this significant? Because setting up a single sign on environment is really a pain if everything doesn't integrate with it. So, because chat can be very important, Kerberos integration without special back-end coding makes the internal iChat server that much more attractive.

Mail
The Mail server in OS X 10.4 is fairly standard, but 10.5 gives you clustering at no extra charge. While I don't think I'll set up another mail server anytime soon (at least until the Utopia project comes to my area), it's something that I may consider in the near future.

Open Directory 4
Ever since I have started working with Mac servers, I have been very impressed with Apple's implementation of a Directory system. Open Directory integrates well with Active Directory, eDirectory, or any other LDAP implementation. It also allows for a lot of control over access.

But the big news that I like is RADIUS integration with Open Directory. So now, for the price of an Xserve, you could easily set up your own dial-up network connection, integrate services through Kerberos, and provide an excellent connection setup. This also would go for Wireless or Wired networks, which would require some sort of authentication for access. Having worked in that field not too long ago, I think this is perhaps one of the most powerful tools that Apple could have provided.

So what does this all mean? It means that Apple has made it easier for me to implement a well built network in my own home without high-cost software. But why Apple? All these tools are available through Open Source technologies on Linux, right? Well, as much as I like Linux (and I really do, don't get me wrong) and it's enterprise level tools, I'm too lazy to set it all up. Even if it were all through RPM or DEB installs, it would still take a lot of set up time, which I really don't want to invest. Using the Apple tools, I can set up the minimum requirements for these powerful tools in a short amount of time, and then they will just work.

But if I wanted to do a real enterprise setup, I probably wouldn't use the Apple servers as my primary system. Not that it isn't robust enough, but I could do it cheaper with Linux and Blade servers, and higher the support staff to manage the system. Apple allows me to manage the system myself, which is ideal for a small business anyway. One support person, or even a support consultant, can set up your system that looks and acts very professional, without having to have them on staff full time.