Is moving from a proprietary OS to Linux the same thing as DSO? I'd call it close, but no cigar.
As a news story this week from sister site Linux Pipeline explains, PalmSource is migrating its Palm OS--used in PDAs and smartphones--to a Linux kernel. Though the first devices running on Palm Linux are probably two years away, the move is important for embedded-software developers mulling the costs and benefits of DSO.
The motivating force seems to be trouble with the Palm OS middleware. As reporter David Beers writes, because PalmSource in early 2004 started using a new middleware framework running on a proprietary kernel that had never seen silicon outside of a lab, either PalmSource or its licensees had to create completely new drivers for every smartphone hardware element. This became what a PalmSource executive calls a "critical time-to-market issue." PalmSource's hope is that Linux will fix this problem by streamlining product development.
While this is not a DSO implementation, it does bear certain resemblances. For one, there's a new reliance on a standard platform (in this case, Linux) rather than a home-grown solution. For another, the shift appears to be enterprisewide; PalmSource has not said anything about keeping selected models on a non-Linux OS. For yet another, PalmSource's expected benefits include lower costs and faster time-to-market. About the only thing that's missing, I'd argue, is the adoption of a formal DSO framework. No cigar, maybe, but pretty darned close.
What do you think? Send your thoughts and opinions to email@example.com.