BPMN BPEL Roundtripping Revisited
+1. Good stuff, as usual, from Bruce Silver. The underlying rule of thumb here, I think, is something like “If you’re generating code, that code should never be touched, never be stored in SCM. It’s not source code. It’s a compile-time artifact, suitable only for further compile-time processing.” This rule is widely applicable, from generated MFC code right up to MDA / MDD approaches. If you find yourself unable to hold to this rule, then I think you ought to seriously step back and examine what you’re really trying to do, as I suspect that you may be misusing the tool of code generation.
There is a funny quote in Bruce’s post though — probably unintentionally so:
The way I look at it, BPEL is just like Java.
If there was ever a sentence designed to get me to run screaming into the arms of Ruby / Python / Scala / Whatever, then surely, that is it.
+1. Good stuff, as usual, from Bruce Silver. The underlying rule of thumb here, I think, is something like “If you’re generating code, that code should never be touched, never be stored in SCM. It’s not source code. It’s a compile-time artifact, suitable only for further compile-time processing.” This rule is widely applicable, from generated MFC code right up to MDA / MDD approaches. If you find yourself unable to hold to this rule, then I think you ought to seriously step back and examine what you’re really trying to do, as I suspect that you may be misusing the tool of code generation.
There is a funny quote in Bruce’s post though — probably unintentionally so:
The way I look at it, BPEL is just like Java.
If there was ever a sentence designed to get me to run screaming into the arms of Ruby / Python / Scala / Whatever, then surely, that is it.
You must log in to post a comment.