Twelve Tips about Starting a New Open-Source Project
2012-12-01
Here are some tips targeting developers who want to create an open-source project. They reflect my personal opinion and may not be all adapted to every situation.
Cyrille Rossant
Galry's Story, or the quest of multi-million plots
2012-11-30
About a month ago, I announced here the availability of a new experimental high performance visualization package in Python that I'm developing as part of my current research project. It has significantly evolved since then, but it is still experimental. Moreover, the interface is still not ready for a 0.1 release. I also need to do much more tests on various systems and graphics cards. In this post I'll talk about how the idea of writing a new visualization package came up in the first place. I'll also describe the new features that are coming to the library.
The Power of Shaders in Real-Time Graphics Programming
2012-11-05
I've been programming in OpenGL for a few months. Like a lot of programmers, I learnt the language by myself, thanks to various tutorials, books or e-books on the subject. One couldn't say there's a lack of resources on this 20-years old language since it's so widely used throughout the world. Yet, I was surprised to discover a few weeks ago that the vast majority of what I learnt has been obsolete for almost a decade. The reason is that too many textbooks and tutorials on the Internet about OpenGL refer to a deprecated way of programming and which relates to the fixed-function pipeline. The modern way of programming in OpenGL is to use the programmable pipeline through shaders. The free e-book by Jason McKesson is a very good resource for learning modern OpenGL programming using the programmable pipeline.
Introducing Galry, a high-performance interactive 2D visualization Python package
2012-10-24
I'm releasing today the code of a first experimental version of Galry, a high-performance interactive 2D visualization Python package that I'm creating as part of my current research project.
Profiling and optimizing Python code
2012-10-10
Premature optimization is the root of all evil.
There are two opposite directions a programmer can take when writing a piece of software: coming up with an elegant software design or with an heavily optimized code. A good design leads to better readability and maintenance, often at the expense of pure performance. Conversely, highly optimized code tends to be more difficult to read, and can lead to bugs that are hard to fix.