Inversion of Control and New Architectures

Here is nice, compact article by Martin Fowler on the Inversion of Control pattern.  I’m revisiting the Spring Framework after stepping away from it for a while.

It’s always important when building a new products to learn from the “pains” of previous projects.  Reminds me of those poor suckers that implemented the full J2EE 1.0 stack back in the day.  Wait, I was one of them. *doh*

Anyway, what’s interesting is it seems that most new architectures are straying away from the Java stack.  According to the case studies at, most of the new architectures are LAMP, Ruby on Rails or Django based.  I don’t have practical experience on these platforms so I can’t tell you why but folks seem to tell me its a ton easier and quicker to develop.  My biggest fears are …

  • Long Term Manageability.   Well, this is a problem for any architecture but its seems that Java keeps things together cleaner.
  • Performance.  Java has a big head start in this area.

For a startup, these would seem like secondary concerns.  It’s probably a good thing when you start having these problems and by then you can throw tons of VC money at it.

Java verses LAMP verses Ruby on Rails Django

I’m still torn.

Dipity Timeline Widget

As I was poking around the Google YouTube API researching a side project, I stumbled upon this cool timeline widget by Dipity.  There seems to be an endless amount of these “widget companies” that are worth much more than the name of their space leads you to believe (Slide, RockYou, etc).

Dipity caught my attention only because they came up with a very cool use of the YouTube API.  However, they also came up with a very interactive widget that, if done properly, could be as compelling as Slide’s slide show widget.  Basically, they allow you to create a timeline of anything and share them with the world.  Here is a link to the Apple Timeline.  Even cooler, they allow various views on the same timeline and allow multiple people edit the timeline.

Their biggest problem is going to be performance.  I’m running Firefox 3 on a pretty beefy Ubuntu box and the rendering stinks.  They’ll need to figure this out.

Bottom line, a compelling startup can be as simple as a different way to visualize something.  It will be interesting to see if they are able to gain traction.

Apptio comes out from hiding …

A number of my old team members from the Mercury Diagnostics product line, which was eventually folded into HP Software went off and founded Apptio.  They describe themselves as:

Apptio is the leading provider of on demand IT Cost Transparency solutions. Apptio’s IT Cost Transparency solutions provide greater visibility into the cost, utilization and operations of IT products and services so that businesses can identify ways to reduce IT costs , make better IT decisions and benefit from more proactive IT demand management. Capabilities include template driven IT cost modeling, interactive reporting and analytics, and business facing  IT demand management.

I got a few IM’s today when their website went live with real content.  A very compelling idea if they can execute on it.  Based on the research I was doing at my previous product line, there is huge demand for this kind of software depending on the maturity of the IT organization. These guys are hunting for some big fish.

Now, take a look at their leadership team.  It’s stacked!  I predict these guys will get bought in a couple of years.

Twitter Technical Problems No Joke

I’m not a huge Twitter fan but have been playing around with their SMS/Twitter/Facebook integration which is pretty cool. I guess all of the blogs about how terrible their infrastructure is seem to be at least pointed in the right direction. They went down this afternoon and I believe are still down.

Then it got really serious.

Oh well, who needs good software? Bah!