We all come up with good ideas. Open source developers are often driven by their creativity to provide us with the software we know and love. But behind the success of every open source effort, there is motivation. In the early stages of a project, the developer team, eventually a team of one, has close to no support or exposition. So it all comes down to stubbornness and belief, which have to last enough for a first interesting version to emerge.

Now what's "interesting"? In this context, it means interesting enough for users to adopt, and outside developers to get their eyes in the source code. Users get you feedback, developers get you technical expertise, and these are sure to skyrocket your motivation. But before they pick up, the software should be at least a bit usable and its source code understandable: even if it's not shiny or polished, it has to show its potential.

I've got through a bunch of audiobooks since my last post. Deception Point was, well, a small deception as it seems to be a draft for Digital Fortress (or is it the other way around?). It almost felt like reading a book for the second time. As usual, the characters are not as detailed as I would have liked them to be, but the writing and narration were simple, and very effective. As with all of Brown's books, Deception Point is a captivating thriller that lacks the depth that differentiates books from movies.
A friend of mine recommended Mort by the witty Terry Pratchett. A short yet very enjoyable, book which introduced me to the author's popular Discworld series. Full of humor and great quotes, this first step through Terry Pratchett's work will definitely make me come for more.

I've also been playing a bit of Polarium, a great tetris-ish puzzle game that got some of my friends and colleagues hooked. I'm going to try and develop a Qt-based clone of it, if I can just get rid of my nascent addiction.

I just read some really interesting essays by Jack W. Reeves about software design. They point out several important aspects of software design, including the source code. The author wrote that code (coders?) help design the software as a whole and its importance should not be underestimated as it is very often the case.

Funny quote I heard yesterday: "We use a secure file transfer protocol at work. People call it secure because when a file is transferred, we are certain that the file written on the other side is the same as the one that was sent". Now that's security :)

I just finished listening to A Short Story of Nearly Everything, by the popular author Bill Bryson, and narrated by Richard Matthews.
A great work of vulgarization (in the good sense!) about the history of science, with a special emphasis on physics and geology. Many interesting facts, stories, and anecdotes about both who and what made science what it is today. The light and humorous tone of the writing (and narration) made the book a real treat to listen to, turning what could have been a boring exposé into a captivating story.

I'm now starting Deception Point by the now massively popular Dan Brown. After listening through his three other books, I feel this author is largely overrated, with poor characterizations but definitely enthralling plots. We'll see how things turn out in this one.

There's been quite some stir lately, regarding multi-core dies in the next generation of mainstream processors from the likes of Intel, AMD, or IBM. As contenders seem to shift from the gigahertz hype, upcoming personal computers will likely feature powerful SMP capabilities with a relatively slow evolution of frequencies.

The shift towards multithreaded, concurrent applications is a turn most developers will have to take as this great article by Herb Sutter convincingly exposes. Sutter writes, among many other things, that "Concurrency is the next major revolution in how we write software". Performance improvements will not be as easy as just waiting for the next gigahertz monster: there will be some major efforts to be made by the developers in order to make full use of the upcoming hardware technologies.

As a side note; people already had trouble choosing their processors: which brand? What are those numbers beside the processor's name? Explaining multicore to your mom won't be easy, but who needs to understand when you've got questionable marketing?