Royalty Free RTOSs mean “Money Up Front”

In the embedded systems market Royalty Free RTOSs are the rage today. I recently had a chance to evaluate several RTOSs and this is what I found.

A Royalty Free RTOS does not necessarily mean cheap. Generally what it means is that for a large up front fee you can use the RTOS on a "Royalty Free" basis afterwards. This pricing model may work well for some high volume products like cell phones and PDAs, but for low volume products it can be much more expensive.

Also, royalties are usually preferrable to high entry costs for products that have a real chance of failure. The traditional "Shared Success" model has definite appeal for these endeavors. If a project fails and few units are ever sold, then the costs for the RTOS and associated development tools are also low.

Obviously the RTOS companies need to get paid somehow whether it is through royalties or otherwise. However, I think that their general depiction of "Royalty Free" as being a low cost model is quite misleading in many cases. In fact I think it is time for some of these "Royalty Free" companies to start offering some plain old royalty based pricing for the many low volume products out there!

Also, it would be great if some tool vendors began to offer "Royalty Only" RTOSs with no up front costs. I believe that this would encourage more companies to undertake risky new product developments.

Java - Great Technology but Overhyped and Oversold

Most non-trivial, Java based desktop applications that I have used are frustratingly slow. Now I know that the Java evangelists will state that Java applications can approach or even exceed the performance of similar C++ applications. However, I have seen few examples of this in the real world. So I have to ask a very blunt question: Is Java inherently slow or are Java developers just incapable of producing a complex desktop application with even reasonable performance?

I like Java. It’s a relatively clean and simple language and is also quite powerful and portable. But don't try to sell me about its performance characteristics because I have seen little evidence of it in the real world.

The Java crowd should start selling Java for applications where it is best suited such as Web applets; rapid prototyping; and light weight, portable user interfaces. Selling Java as the solution to all development projects simply confuses evaluators and frustrates adopters.

PDAs - Too Small and Too Big

I know that I am in the minority here but am I the only one who thinks that the current "mass market" PDAs are still not the right size. I find them too big to be really portable but too small to be really useful.

If I want an organizer that can keep all of my contacts and schedule information and is really portable then I would opt for something like the Xircom Rex. The Xircom Rex is about the size of a credit card and can be easily synched with a PC which is basically required for any non-trivial data entry.

If I want an organizer that is powerful and flexible then I would opt for something like the NEC Mobile Pro 780/790/900 or the Psion Series 7. These devices provide a nearly full size keyboard and a large display that allow you to efficiently take notes and/or view documents.

It appears that the majority of users disagree with me and both the REX and HandHeld PCs are fading away. However, there is still hope as I have recently been hearing a buzz about the Alphasmart Dana which provides a full sized keyboard with a wide display for note taking. Also, many PDA vendors are now selling full sized keyboards for their PDAs.