Why Windows?

Dave Cohen said, via Twitter,
People who complain about Apple products running poorly on Windows crack me up. It's Like saying "This Rolex looks bad on my ankle."

It got me thinking about why Apple and Linux are not doing as well as Windows. There's the typical answers, ranging from boring to trite:

• Windows is the market leader and has done a good job of maintaining their established dominance
• Windows uses its dominance in a way similar to monopolies; it uses dominance in one market to maintain or establish dominance in another
• Windows has better support for office applications
• Windows is cheaper
• etc.

None of these hold water for me.

• Market domination is important, but the numbers haven't budged in decades, indicating it's more than simple momentum.
• The number of markets in which Microsoft dominates is shrinking or losing importance. Example; dominance in graphics processing (DirectX) doesn't really matter much, since OpenGL is good enough for most games.
• Mac and Linux both support great office applications.
• Linux is cheaper.

Despite all this, Microsoft is still by far the most dominant player with market share staying more or less steady. So, let's look closer at Apple. I'll dissect Linux later.

For me, it comes down to a misunderstanding on what is a true qualitative comparison and what is a threshold metric. "Qualitative" means more is better. "Threshold" means that the platform merely has to meet a certain level (the "threshold"). Anything more than meeting the threshold doesn't matter.

In the 2008 primary elections, I read an NYTimes article about Obama's foreign policy experience. The author claimed that Obama will never prove to the country that he's better at foreign policy than Hillary Clinton. But, all he had to do was prove he was good enough, not better, then he can win on the truely qualitative metrics. I believe a lot of Apple's problem is that they're competing on quality when all the market needs is a minimum threshold.

Apple is easier to use
Yes, but this is a threshold issue. If people find Windows easy enough, easier doesn't really matter. It's Windows which has to be so bad that people look for something new.

Apple breaks less
Fewer viruses, fewer blue screens, etc. Again, this is a threshold issue. As long as people are happy to reboot, run a virus scanner, whatever, people won't feel compelled to leave

Apple has better integration
iPhone, iPod, iTunes, etc. Apple has some great web products. But, most run on Windows and none are really compelling enough for people to buy a new computer.

Apple has better user interface design
People love to point out how much friendlier Mac OSX is than Windows. No doubt it's true. But, Windows is good enough for most users. Again, it's a threshold issue and not a qualitative one.


So, how do you beat Microsoft? 1) Attack threshold issues where they fail and 2) Attach truly qualitative issues. Unfortunately, MS has done an excellent job of "good enough" and both approaches will yield limited results. Microsoft has a history of meeting these issues slowly.

So, you can see, most of the people who have attacked MS successfully have done so by attacking new features/requirements before MS had the chance to establish dominance. Like, Google's search, SalesForce's SaaS, Amazon's cloud deployments, etc.

Where is Reliable Response attacking Microsoft? Mass mobile marketing. MS doesn't have a strong mobile play, and where they do it's focused on the Windows Mobile OS. RR plays in the mobile integration for feature and plain phones...those without a smartphone OS. By the time MS wakes up to this as a viable market, we hope to be entrenched.

Unless I advertise my plans on blogger.com :)

Federated Payment Services

This is something that's probably already been thought of. I think we're probably already moving in this direction...

One of the reasons the Apple App Store is so successful is the ease of which you can make micro-payments. Once you set up your billing info once, you can buy apps quickly. If the app's only a buck or two, it's easy to buy the ones you like. The seller can sell the app to a couple millions users and make some real money.

Other phones, like the Google Phone, and sites like Scrib are mimicking this model with their own micro-payment sites. They're all good ideas, and I think the micro-payment concept can be extended to a lot of products.

But, having many different micro-payment sites means having to trust each site with your credit card, and to take the time to fill in the information. If it's just for a magazine or a spur-of-the-moment purchase, I probably won't bother. Without there being a critical mass of products I'd like to purchase on any particular site, there's no point to signing up.

That's why I think it's important to federate the micro-payment services. Let me choose any particular payment service; paypal, google, authorize.net, amazon, visa, mastercard, whoever. When I go to Scribd/Google's Market/Apple's App Store, have it log me in automatically. Let PayPal/Google/Amex/Whoever take it's 2.whatever percent, Scribd take 10% and the rest go to the author. Have me sign up once and never bother me again.

Then, I'd be happy to blow $5/day downloading my newpaper, KenKen, weekly gossip rag and every Larry Niven book they'll sell me.

Case in point. I love the New York Times, but I won't sign up for their service. It's $50/year, which is too much for me to pay for a newspaper. And, I don't want to have to sign up with another e-commerce server. However, I'd happily pay 50c every weekday to download it to my Google Phone, if it was easy and secure. 260 weekdays == $140, $123 bucks of which goes to NYTimes.

Effect of Gas Prices on Emergency Notification

Gas prices are affecting everything. From the obvious, like the decrease in sales of SUVs and the increased charges for deliveries, to the not-so-obvious, like the increase of the cost of milk and the resurgence of the locally-grown-produce trend. One of the not-so-obvious affects is on emergency notification, business continuity and disaster recovery.

Gas prices affect where people live and how they work. People bought large houses in the exurbs during the 1990s because gas was cheap and it was easy to get to work. People got used to commuting 15 miles in a large SUV. Since many of these cars get 10 miles to the gallon, that meant people were paying for 3 gallons of gas a day. At $2.50/gallon, that's $7.50, or roughly the cost of taking the BART from Oakland to San Francisco. Or the cost of commuting from New Jersey to New York along the Erie Lackawanna train system. Or maybe double the cost of taking any large city's metro system from the 'burbs.

Even with a small difference in price, people will often opt for using their car. It's just a lot easier than riding the bus or light-rail. I take the bus sometimes. I have an incredibly easy route, just one bus to take and it stops within a block on both sides. Still, if it's raining or snowing or really cold, I'll drive because it's better than waiting for the bus in the rain. So, at $7.50/day for driving and maybe $3.50/day for metro service, it's easy to justify the extra $4.

Fast forward to 2008, where gas is $4/gallon, and now you're looking at $12/day. At that point, it makes sense to look for either alternative transportation or increasing the amount of time you work from home. Either way, you're spending more of your day outside the office.

The purpose of emergency notification is to contact the responsible parties during an emergency, wherever they are, whatever time it is. Any notification that relies on your employees being in the office is bound to fail. This is only exacerbated by the new trends towards mass transportation and increased telecommuting. The ability to respond to those emergencies, even if it's just to escalate the alert, from whatever mobile device you are currently using, is critical.

My product, Reliable Response Notification, addresses this problem. By supporting many device types, and multiple devices per person, your monitoring, ticketing and security systems are able to reach your employees on the bus, in their car, at home, at the local coffee shop, wherever. Store-and-forward technologies, like email and SMS, allow the employees to answer at his or her convenience. They don't have to miss the bus to take the call, they can get on the bus and read the message when it suits them. Escalation allows the system to adapt to times when an employee is simply unavailable, whether that's in a tunnel on the BART or because they're on a bike and can't hear the ring.

My Ideal Hosting Environment

Before I even start, let me say that I haven't found the perfect hosting company, but eApps comes pretty darn close. They hit all the big points, and only miss a few smaller "nice to haves". This is why I promote them to all my consulting clients and use them exclusively for hosting Reliable Response Notification.

#1 - Reliability
This is more than the server not crashing. I've been on hosting companies that switch around the file system structure without telling me. Since RRN uses a few absolute paths in the config file, moving these paths have negative affects. I've also had hosting companies simply switch off Tomcat, or exceed the allotted maintenance windows, or change IP address. I need to know that my hosting company won't break my app on me.

eApps does require you go through some hoops to get everything working reliably. You need to set Apache directives through the control panel...just editing files is a no-no. I got caught by this. But, they patiently explained how and why it works, so I'm on board.

Also, I don't mind being on a shared server, but there needs to be some protection against the other customers. I've been a bad shared customer at times, too. I had a nasty bug where Yahoo IM access would occasionally drive CPU utilization to 100%. Some hosting companies happily allow you to continue driving CPU up. Others will stop your app and boot you off the server. A good hosting company simply reduces the CPU cycles you have available.

#2 Support -
I *HATE* when someone tells me "I rebooted it and it works now". I need to know why it failed, so I can avoid that in the future. Customer support isn't about answering emails. It's about populating those emails with useful information. Phone support is nice, too. Online chat doesn't work well for me.

#3 Features -
Virtuozzo/OpenVZ is a wonderful thing. I suppose Xen and VMWare are, too. Most of my apps don't require an expensive hosted server. OpenVZ gives me the ability to purchase a small, cheap server that acts like a dedicated OS. Plus, it can dynamically scale, restrict resource usage, and migrate between physical servers.

Linux has a lot of good software for it. Some of that should be available to the hosted server. MySQL, PostgreSQL, Tomcat, Apache, PHP, curl/wget, Nagios, and some web stats package. If you don't support those, then at least allow us to use the standard package installers, yum, apt or whatever, to install them. Failing that, then, we should be able to install from source.

#4 Scalability
With most hosting plans, you can move up and down plans fairly easily, ie scaling vertically. eApps takes it to an extreme. You can move up and down, most of the time without stopping your application. 0 downtime. As well, they support clustering, using excellent hardware load-balancing. This allows you to scale horizontally, by adding new servers. Combine that with their super-cheap, super-flexible OpenVZ servers, and you have one of the most affordable, scalable solutions on the planet.

All in all, I've been very happy with eApps. They have some competition from Amazon, which I imagine will get tougher soon. They also have competition from the classic providers like Verio, GoDaddy and Network Solutions. None of these companies provide the individualized support, attention to your particular app, and general excellence that eApps does.

Privacy in a Twittery World

Reliable Response is in an office building across from the Capital of Colorado. We get a lot of political groups in here. The most excellent Colorado Statesman is downstairs. The Colorado Right to Life group is next door. My favorite is Jim from the Colorado DLC. He and I talk occasionally. He lets me rant, and I help him with technical issues. Coming soon...webcam-based interviews with politicos around the state. Maybe he'll let me rant on camera.

Jim was asking me what I thought about the phone companies eavesdropping on us. Obviously, no one wants their private conversations to be listened to by anyone, especially some low-level government or telecom flunky. As a former low-level flunky, I can attest that we flunkies *love* listening in on internet traffic.

The fact is, to quote Scott McNealy, "Get over it. You have no privacy." I couldn't agree more. The government can't give it back to you. We can sue whomever we want, but the RBN will always be willing to sell your information to anyone who's willing to throw down a couple bucks for it. Try asking Putin to help...let me know how that turns out.

Anyone who knows me knows that I have worked in security, including a long-ish stint for RSA Security. I believe in encryption. I love it! Mostly because it puts the control over your privacy back in your hands.

To answer questions about telecommunications security, I will always say that counting on the government and the telecommunications companies to keep out of your business is naive at best. There's only one person you should trust. You. Make sure you have the best encryption you can use.

On the other hand, the government can, and should, make it easy for you to do that. To this end, the IETF has released RFC 3711, the Secure Real Time Protocol. This is an extension on top of the most popular VoIP protocol to provide real security. People should start using it. We should demand it's support for all VoIP phones, including Skype!

Reliable Response Notification doesn't use any of these mechanisms. It's a problem we're looking to address. We're publishing some pretty private information. Stuff like IT outages, purchase requests, and internal marketing communications. It keeps me up at night to think that someone might at AOL might be looking at these IMs streaming past. The problem is that the communication methods people use simply don't support these encryption standards. Even support for PGP, an encryption standard that was old 10 years ago, has seen so little uptake as to be considered effectively dead. This is a problem.

The government can help. Richard Clark's cybersecurity efforts are a start. The government should lead, promote and market security solutions. But, it should never make them mandatory. When there's critical mass, people will use them. I hope one day, people will tell me that they won't purchase Notification without built-in security.

WorkFlow and Notification

I've been working with an OEM partner, Advantage Integrated Solutions. They have been using Reliable Response Notification for an interesting purpose. Notification has been taken out of its typical role as an emergency notification tool, and they've been using it integrated with a workflow system. This has been very interesting to me.

It brings notification out of the IT horizontal market. IT is a notoriously difficult horizontal to sell new products into. They've been burned by poor quality, fast changing requirements, large bills, and difficulty maintaining the software they install. I can promise that my software is easy to use, bug free, etc. I can even show them, with demos, trials, references, etc. But, nothing will remove a deeply felt belief.

It introduces an SAAS model. I've come to love software as a service. It's cheaper for the customer, while providing Reliable Response with a reoccurring revenue stream. Support is 100x easier, and the product appears more reliable because Reliable Response can handle problems before they show up with the customers (for instance, hardware or network issues). I get better usage statistics so I can further improve the product. Best of all, it opens up a market to a whole new breed of customer, the small company that doesn't have a lot of money. Instead of a one-time $10k purchase, you can do a $40/month and buy only the features you need.

It's a new market. I have competitors in the IT space. At least one is a good competitor. And, another competitor has the vast market share. Competitors are good because they let you know the market is proven. Even better are customers. Pretty much every vertical out there requires this sort of product, but the IT horizontal is the only market that's being addressed. Same product, more customers, fewer competitors. Even though I get less money per sale, the sales are easier.

It's easy to get technology partners. When Notification is an IT product, I can partner with monitoring, help desk and security packages. Most of these are closed systems with few easy integration points, and a difficult partnership program. Once I get outside of IT, I can partner with a much broader range of customer. Pretty much every SAAS vendor has the hooks I need. Heck, I don't even need the buy-in from the SAAS vendor for the most part (although it's nice to have).

So, why hasn't Notification taken off in this space? Well, I only got the religion a few weeks ago. Mostly, though, I am looking for partners to bring into the company who know particular verticals that I can attack. Anyone who's interested, email david@reliableresponse.net

Reliable Bunnies!!!

David Hauser of GotVMail sent me a gift. The Nabaztag WiFi Bunny. I posted a video of my bunny working with RRN and Nagios at: