On 14/01/07, Paul Downey <[EMAIL PROTECTED]> wrote: > > > > > > > > On 13 Jan 2007, at 22:28, Steve Jones wrote: > > > > As Anne points out, it's a standard in name only. > > > > Given that BPEL 1.1 wasn't really a standard and that BPEL 2.0 (the > > first actual standard) hasn't made it to be a standard yet I think > > this might be writing it off a little bit too soon. > > Er, I think you just made my point more strongly. > > Watching the TC is a little like watching a divorced > couple who continue to share a nice house. They > hate each other, live separate lives and can only > agree on the nice address. > > > Standards have a value, standards create markets > > "innovate, implement, standardise" is the most successful model
Is it? 802.11x is a good example of where standardisation is the bit that actually created the market. The first two phases didn't achieve that. CDMA v GSM is another example, CDMA might have more "innovation" but the market belonged to GSM. > > > and while BPEL may, > > or may not, be the execution standard that helps to create a more > > active business process market there doesn't appear to be much else to > > compete with it. > > I actually prefer pi-calculus such as W3C CDL, or SSDL, but any > one of a number of proprietary solutions is going to be a much better > bet > - they've not been badly designed by committee. Proprietary solutions do not create markets, unless they are monopolies. If we truly want to get to interchange of process and the ability to execute my processes within other people's domains then we need standards. Standards create markets, proprietary solutions create silos with minimal value. > > > Even a crap standard is better than nothing, > > Standards are agreements, and if they have commitment they have > power often above the value of what is being agreed upon. > But it's the agreement we're questioning. My advice, FWIW, is > avoid it. So your argument is that BPEL 2.0 doesn't have the agreement of the group. I'm certainly concerned about MS as Windows Workflow doesn't use BPEL in anyway shape or form, who else are you worried about. > > > OLE was a great example of such a thing. > > OLE was a standard? Maybe you'd like to update this page: > http://en.wikipedia.org/wiki/ActiveX It was a de-facto standard created my a monopoly and it was dreadful, it was terrible but it was still better than hacking it yourself and getting barfed at when the next e-fix arrived. Standards don't have to come from standards bodies, they just have to be agree upon by the market. > > -- > http://blog.whatfettle.com > > -- > http://blog.whatfettle.com > >
