Archive for September, 2009

Photographing the Earth from above

September 25, 2009

I was impressed by this:

Two MIT students have managed to take pictures or the Earth curvature by launching a photo camera + mobile phone attached to a balloon. I was really impressed. If I was a kid, this would be very motivating stuff to get me into science. I would probably try to replicate it too…

Being a software engineer, I’m thinking … what if you get a “smartphone” and you write some software for it … what software would you write? Well, let’s see:

1. Software to control the camera integrated in the smartphone, so you don’t need a camera and a phone
2. Software to connect via gprs and submit live images (or hey, a live streaming) … question is, would you have any signal at 28km above the earth?
3. Software to monitor the GPS trajectory and altitude, so that you can plot it on a map and you can draw all sorts of graphs
4.  A IM chat client to tell your friends you’ve done it 🙂



September 22, 2009

I’ve added recently another language to my repertoire: MQL. It’s the native language in the MetaTrader, which is a FX trading application. I’ve started FX trading only recently but I’m really attracted by the algo trading and trading in general.
MQL is quite basic, it doesn’t support complex data types, I could really use with a dictionary or an object of some sort.

What can you do with it? You can write an Expert Advisor, this is a algo trading application.
It has a 3 methods: init, deinit and start. The first 2 act as constructor/destructor, whereas the second is called on every tick, which you can process and then decide whether to open a trade, close a trade or do nothing.


September 7, 2009

Does outsourcing works? It’s a question I had to revisit recently because my company is thinking about outsourcing. The main reason is cost and I think that’s the wrong reason to outsource for. And because you start outsourcing for the wrong reasons, it ultimately ends up in a failure.

Most companies look at it as a purely tactical move, although they call it strategic. The attraction is cost. However, because the move is tactical, the integration is also at a very tactical level. Usually the team assigns unattractive tasks like defect fixing or other uninteresting work to the offshore teams. Which ultimately leads to other problems like lack of knowledge, specialization and ultimately productivity (which is labelled as failure, and rightly so). Also, because the work is so un-attractive, it only appeals to mediocre employees, the best ones leave for greener pastures. In the end, it becomes a problem of loyalty.

An important problem is communication and the fact that there’s a bigger disconnect betwen the onshore and offshore (or nearshore) teams. Specs have to be more detailed and knowledge needs to be transferred more formally (which in IT doesn’t work that well – a lot of knowledge is tacit and not necessarily easy to capture in documents and diagrams).
This in the end results in more tech lead and PM time, which is quite expensive. Tech leadership and PM become full time jobs dedicated to the offshore team. If you add the cost of lower productivity to do with poorer communication and disconnection from the business/front office and you add also the travel costs, I’m not sure it’s such a good deal anymore.

Nevertheless, waves and waves of management professionals and consultants will fall for it, because on paper it represents such a good deal. You can boast about cutting cost which is music to the ear for the upper management, especially in recessions. And then they can cash in their bonuses and claim success in the short term, whereas the company (and shareholders ultimately) are left to pick up the pieces.

I’m not against outsourcing though, I’m just against it when it’s done for the wrong reasons. I also think that doing it for the right reasons would also mean doing less of it overall. Just because it’s not a universal panacea.