Cart
[Home] [By Thread] [By Date] [Recent Entries]
> It is still a good idea. Process orchestration is > overkill for most business processes that have to be > flexible at the level of the desktop. There likely is > a 20% solution that hits the right combination of > width of process and depth of task. Right, in our product we found very little use for orchestration outside of the "default". What we did find is that people are very interested in inputs and outputs. By creating additional input sources and output formats (like plugging rule results into an SVG) we definitely approached something more akin to 80/20. The user gets the functionality of orchestration that she is looking for without the hassle of having to define it. > On the other hand, those rules are typically in code, > augmented by human reasoning, are not always prudent > to expose, and might present a contractual risk. Enterprise > engineering is a very difficult task if it is coupled > to organizational control instead of simple interfaces > among external partners. Right, that is why you work with the "business experts". For us we developed a pretty front end, and added an option to export to XML on a server and have a standard webservice/website that allows you to push data through your "hosted" rules. > The problem of standardization coupled to publication > is forgetting that business is competitive and openness > provides opportunities to competitors. Sagacity rules. Right. Jeff Rafter
|
