Avian’s Blog

Ever wonder what is the reason behind that big fat warning that you shouldn't use your halogen desk lamp without the protective glass? After yesterday evening I know why.

The 20 W 12 V halogen light bulb on my desk went out with a loud bang, disintegrating into tiny little pieces. They were mostly caught by the glass shield. Those that escaped at the edges were the size of fine glass dust and covered my desk, giving it a fancy shiny look.

This is how things looked when I removed the safety glass today:

There's a white residue on the safety glass that appears to come from the bulb pieces. I'm still wondering where that came from.

It's a little hard to see here, but the aluminum reflector is full of little dents. So counting in the apparent force of the explosion and the high temperature of the bulb, that glass shield most probably saved me a couple of burns on my hands.

The box of the new bulb I bought (made by Philips by the way) says it can be used in fittings without cover glass. Somehow I tend to disagree with that.

Update: here's a better photograph of the white residue on the safety glass.

A while ago I've seen a graph somewhere that showed ups and downs in one's life. It showed this gradual rise of happiness when graduation was getting near, when things are finally starting to fit together. The peak was followed by a steep and deep valley of depression with an annotation saying you realize you haven't figured things out at all. Maybe it's because working at Zemanta kept me busy immediately after leaving the University, but I have definitely been feeling this effect in the last months.

About the only thing that has been clear in my mind about what I want to do in my life is that I don't want to continue in the natural language processing and software development field that has been my regular job for the past two years.

I find it hard to stay optimistic about all the hype that surrounds the Web and the benefits it ought to bring to us all. I have a feeling that an immense amount of effort is being wasted on all kinds of web-related projects and products that are in the end only clever means to push advertising to people who don't need your service. I see little point in spending all those hours of work to try to keep up with ever-decreasing attention span of users and developing the next best way of telling their friends what kind of coffee they had this morning.

I've written before about my doubts about the way software business seem to run today. It's trying to build immensely complicated structures with processes and timelines which don't allow even simple programs to be written without bugs. The industry is moving towards the next best thing when the yesterday's project hasn't even been debugged properly.

At this point I should mention that my job at Zemanta was so far probably one of the best in this regard. There's the amount of actual research I've been doing and also the fact that the pace of development has been such that I've been able to keep the important parts of my software architecturally sound and relatively reliable. There is a quick-and-dirty hack here and there, but as a whole it's in a state I can be proud of as an engineer. But still there is this distinct feeling that it's not something I should be doing and a lack of feeling that I'm actually contributing something of long-term usefulness.

Then there's electronics, which I love as an engineering field. But I'm starting to realize that I've been looking at things from the wrong perspective. Maybe I haven't thought so when I graduated, but only now I see that I left the University excited about ways in the field that were popular 20 years ago. It's probably not the study program itself that's to blame - the knowledge it gave me is certainly not useless, it's only that I ended up with a like for something that doesn't appear to be in much need today in the outside world.

This kind of manual electronic design is largely limited to a hobby and an odd niche market. In fact electronics isn't that far from software today. Nobody is doing circuit design anymore. Digital circuits are build through complicated, abstracted interfaces like VHDL. There is a strong movement to software-defined devices. Those can be updated each time with the latest version of the circuit, which brings to the hardware world all the properties that capability has brought to the software world.

What isn't done by automatics is built on well tested, checked out basic circuit patterns that were thought out long ago. I've been told a couple of times that electronics is not a field you want to go in if you want to do research, to do something new. The real advances are made at the level of new devices. Physics of quantum dots, memristors, nanotubes and so on.

I tried to get excited by postgraduate study in the field, but right now I'm not even able to answer basic questions like why I want to do it, what I'm expecting from the study and what is the goal I'm trying achieve by it. I find it hard to limit myself to study of a single, narrowly defined field. I've always tried to keep a broad view of things, trying to understand to sufficient level all aspects of a project I'm working on. And it appears that most academic endeavors require much too narrow focus. At least for now, I haven't found an opportunity where I can be creative in this regard.

So, what does all that mean? Not that much it seems for the near future. I'll be spending less time working at Zemanta and more time keeping my eyes open and looking around for other opportunities. Maybe I'll even dig out something interesting.