Posts tagged as: webservice

The rumors of Microsoft's death are greatly exaggerated

The Web 2.0 bubble has reached the predictable stage of assuming that Microsoft will inevitably fail to adapt to this latest tech craze. You can get the feel for it by reading the comments on Scoble's post about this subject. Where have we seen this before? Oh yes, in December 1982 when VisiOn was announced at Fall Comdex and the press (this was before blogs existed) fell all over itself predicting that Microsoft would lose control of the operating system market to the new wave of integrating environments and integrated products. We saw it again in 1995 when Netscape's browser was going to wipe out Microsoft, because Gates just didn't understand the Internet. Now ten years later we are again hearing that Microsoft is a dinosaur and can't possibly catch up with the latest web services wave.

Let's get serious for a minute. Microsoft has control over 90% of the desktops on the planet. I haven't seen any stats that web based apps are being used by even a few percentage points of the real users out there. As keeps happening, especially in Silicon Valley, the bleeding edge sees everyone they know going crazy over a new set of technologies, and they extrapolate that onto the general public.

History has shown repeatedly that first-movers do not always win, and often disappear. VisiCorp died within a few years of announcing VisiOn, and Microsoft won that round with two products that they hadn't even started work on in 1982: Windows and Office. Netscape failed under the weight of their own arrogance, and IE is now the dominant browser. It isn't just Microsoft who has beat the early market leaders. When Google first appeared, Yahoo was firmly entrenched as the dominant search engine.

I'm far from a Microsoft fan. I've made plenty of jokes about Gates being the Antichrist. I just don't see how the race for a set of technologies that may be exciting (I'm excited by the potential of Web 2.0 too), but haven't produced any products that real people (not bloggers) are using in any sizeable numbers is already over.

Microsoft keeps winning these races for two reasons:

• They keep plugging away at an application area until they do eventually get it right. They have the cash and the fortitude to keep retooling until the market starts adopting their solution.
• Their competitors ALWAYS f*** up. This is the part where I may believe in the supernatural aspects of Gates' success. I've seen it too many times, Fylstra, Kapor, Andreesen. There is something about becoming a billionaire, or at least a hundred-millionaire, that warps people's minds and ability to innnovate. Gates has avoided this, but look at Page and Brin with their new 767 toy. The drumbeat for Google as the new evil empire can be clearly heard.

Do I think Microsoft will inevitably win this race? Of course, not. Nothing is inevitable. But anyone who says the race is already over, and Microsoft can't turn the ship around fast enough is either a fool, has an axe to grind, has no idea of the history of this industry, or all of the above.

Time to start Ruby programming

I've been reading Ruby books for a week now and I'm getting antsy to do some coding. My way of learning a language is to pick a real target and then push my way through all the obstacles. That forces me to do the hard tasks, but keeps me from wasting time on features I'll never need. I know I've been saying that I want to use Ruby to run the CMS for this blog, but I want to hold off on that until I add enough features. For example, I still need to add calls to ping servers and tagging to this blog. It will be easier to just do that in FoxPro first, and not worry about the programming language at the same time. Besides, from everything I've seen, Ruby on Rails will be the best way to build a CMS with Ruby, so I'll wait until later to start that project. I'll pick an easier target for my first Ruby project.

I've had an idea for Amazon web service programming for a while that might make a good candidate. I'd like to build a "pick the best book on a subject" app. I'm not sure exactly what that means, but the general idea would be to enter a search term and have the app rank available books based on a set of criteria I develop. That will let me explore the Amazon database from the inside.

My first step will be to build a development environment. I could start by doing Ruby programming on my own computer, but that won't let you look over my shoulder easily. I could do the Ruby project on the server where this blog is hosted, but that brings up the security issue. Programming on the web in a new language always leaves open the possibility of security holes. I'd rather do my quick and dirty development in a separate location and then move them to a new server for production use. I also like the idea of a distributed environment for web development. It allows for better scaling and gives me more control later for rearranging the architecture. So my first task will be to build ruby.darwinianweb.com at a new host.

When I first started looking at Ruby I saw that there were some hosting sites that provided Ruby and Ruby on Rails installations at extremely low rates. They billed themselves as Ruby playgrounds, which is a great idea. I've been a programming language junkie for over 25 years, and the idea of playing with a new language is always a thrill. It looks like these hosts are aimed directly at that market.