Re: Roll-your-own vs. commercial software
I guess you don't buy Jerry Fiddler's argument about increased Time to Market pressure and increased complexity of the 32/64 bit environment as a reason for the Roll Your Own's to switch.
I buy it because I haven't made the investment to write my own kernal code. However, if I was in a similar position as many major developers of real-time systems, I'd use the internal code that already exists. Commercial software will continue to take market share from roll-your-own, but progress will remain slow due to the lack of overwhelming compelling reasons to switch. In the meantime, I think Wind River is on the right track trying to identify vertical markets where more value can be added, such as Tornado for Managed Switches. Nevertheless, entering new markets with more competition entails more risk and should be taken into account when making investment decisions.
You've met some poor engineers (or were one of them). I've never met a sw enginer that claimed he could write a better/more effiecient/more reliable/cheaper RTOS than Wind River, or pSOS or VRTX, for that matter.
Not many people will tell you to your face that he considers himself a genius, but actions speak louder than words, and 60% of the market is roll-your-own. I often ignore what people say and instead watch what they do when they're not trying impress or avoid offending me.
Any SW engineer who claims that starting from scratch he/she can write an RTOS that has more/better functionality and (this is the clincher) just as reliable and cost effective as Wind Rivers' (or even INTS etc.) is a fool and should be treated as such (or maybe a genius... and should probably be given as much berth as the fool in a time crunched sw project:)
It would be fair to say that 60% of engineers are fools or geniuses. The same could be said about most professions, since it's in the nature of the human beast to have a big ego. I would leave my ego at the door and choose a third party development environment to save time in a complex project (or anticipated additional future projects), but I'm obviously in the minority. I saw the same thing in the DOS world and was one of the few programmers who learned Windows in the late 80's, so I know first hand how difficult it is to move stubborn people.
If the speculation on this thread is accurate, Jerry Fiddler is like Bill Gates and Ron Ablemann is like John Sculley. Neither Fiddler nor Gates makes business decisions purely for money. Instead ego plays a big role and they want to see their creations exposed to the greatest audience. How else can you explain Microsoft paying $400m for WebTV? |
