Home > Projects, RTOS, SKC++ > Simplifying RTOS

Simplifying RTOS

Wednesday, September 9, 2009 Leave a comment Go to comments

Most of today’s RTOS offerings are too complicated. Here are some reasons why this is so:

  • Insufficient thought , before designing, as to what is really essential and what is not, leading to redundancy and confusion.
  • The perceived need to offer everything already offered by competitors, then to add more features in an attempt to steal a march on them.
  • The fact that some products, as in all walks of life, are just badly designed. Unfortunately, the power of heavy marketing to foist mediocre products onto customers has been proven over and over again.
  • The fact that some people, including software designers, have a natural tendency to complicate rather than simplify. This, I’ve noticed, happens particularly with younger engineers who are anxious to shoehorn everything they know into a project rather just what’s needed. I probably did some of that, too, but it’s so long ago that I can’t remember!

Well, I’m bucking the trend by writing what I believe will be a simpler and better pre-emptive kernel (I prefer not to overstate what I’m doing by calling it an RTOS). Furthermore, I’m going to blog about the project, from now, as it progresses. The possibility of public scrutiny and feedback will, I’m sure, help me to do the job properly and, above all, will drive me to keep doing it until it is finished.

I wrote a pre-emptive kernel in the early 90’s. It worked very well in my original client’s application and I remain proud of the achievement. However, that kernel was constrained by having to be “feature-compatible” with the kernel it replaced (VRTX), which is not a good model for what I have decided to create now. Furthermore, it was implemented in C, as are most kernels in use today, but the trend is increasingly to use C++ in embedded work. In contrast to the popular technique of wrapping a C API in a set of C++ classes, this, the first of my planned projects, is being implemented directly in C++. This gives a performance edge over the wrapper approach but, equally importantly, it allows greater freedom to design, from scratch, using an OO approach.

But enough rambling. The project is called SKC++ (Simple Kernel in C++) and further postings about it will follow apace.

Advertisements
Categories: Projects, RTOS, SKC++ Tags:
  1. No comments yet.
  1. No trackbacks yet.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: