In a series of maneuvers between companies and standards bodies that highlights a growing friction around industry guidelines, Microsoft has dropped out of the W3C choreography working group focused on establishing rules for how businesses will communicate via Web services. Recently, IBM and Microsoft have been at odds with other companies around standards submissions.

The working group is focused on choreography or orchestration, which refers to the use of XML-based Web services to build software for sharing data and processes in complicated business scenarios that may span several companies and computing systems, such as financial transactions.

If the two market giants spearhead a choreography group that runs parallel to the W3C working group, there is concern that rival standards submissions may hinder the progress of software interoperability through Web services standards.

Critics contend that Microsoft and IBM, which also is not participating in the WS-Choreography Working Group, are causing confusion with attempts to exert their influence over an increasingly contentious process to define agreed-upon methods for exchanging information using Web services standards.

Sun issued a statement condemning the move, saying that "IBM and (Microsoft) have now moved away from a leadership position in Web services standards and become a disruptive force in the industry."


Read Microsoft breaks with standards effort from ZDNet

I believe standards are nothing but a monopoly from big vendors and once again they proved it. What is standard? Any thing that already been developed in their (IBM and MS)labs:-) Just think how can you beet/compete them if you always have to follow them. I think it is good if we don't have any standard on webservices choreography so we can live with our standards not on the big giant corporate.

However typical of them, anarchy is not the solution. Without standards, we wouldn't even be here in this forum, posting our views and thoughts.

Henrique, I completely agree.

The whole crux of the issue with Web Services is interoperability. Truly effective interoperability on the Web Services scale absolutely depends upon standardization. Otherwise, the promise of Web Services as the building blocks of the web for distributed computing gets mired down in little proprietary islands of one-off standards, dialects and "flavors" for Orchestration, Security and the like.

Very frustrating; no matter who you like (IBM, Sun, Oracle, Microsoft), all the big vendors have the same problem advancing the industry on this front. Very funny that the success of the very technology they've put so much stock in carrying them into the next generation of computing (i.e., Web Services) requires them to work and play nicely if it's to realize its fullest potential.

Sun cannot expect IBM and Microsoft following them. Given a chance IBM will take over Java too. beware SUN :)

Any thoughts on what this may do to the BPEL4WS initiative?

Standards are certainly the most important aspect of the future technologies especially WS. At the same time, there is no harm in having a few competing standards especially in the upper layer standards. By upper layer standards, I mean a standard that is based on certain other core standards. And choreography certainly falls under that category.
Sometimes all the people agree on something (so only one standard) and sometimes people don’t agree on one thing (multiple standards). By standard, I mean both types of standards that are specified under some consortium body or a defacto standard. At the same time, the IT ecosystem will not allow huge number of competing standards so I don’t agree that there can be anarchy in the standards world.

No wonder why every one ends up hating Microsoft!

But on the other hand, I am sick of all those WS-* acronyms and standards. It is time for the vendors, large and small, to stop writing specs and start delivering implementation.

BPEL4WS is a good example of that...It is supposedly the best thing since slice bread and will be the equivalent of SQL for workflow but there is not a single vendors beside IBM alphaworks BPEL4J that is able to back up marketing claim/hype! This is sad.

What made J2EE successful was that there were specs (far from perfect), a lot of implementations and a large developer communauty that could organically improve the solution.

There is a lot Web Services can learn from that organic, implementation based, evolution process.

Valery

Valery,

I hear your frustration. As a vendor, we sometimes feel overwelmed with the number of specs too. But I think that in the long run things will end up converging rapidely into a new stack will reduce the pain of having multiple applications talk together (at least if Microsoft and IBM end up remaining on the same side of the fence).

If you are still looking at BPEL4WS engine, we/Collaxa have a J2EE based product that is available for download[1] and can be installed on top of JBoss, WebLogic, WebSphere or SunONE.

We very much agree with your view that developer community is a key driver for innovation and usability so be vocal!

For those who have never heard of Web Service orchestration or BPEL4WS, here is a mini-tutorial[1]. Here again, if you have any questions, please submit them to the web service forum or email me directly edwink ATNOSPAM collaxa.com

Edwin

[1] Download BPEL Orchestration Server
http://www.collaxa.com/developer.download.html

[2] BPEL4WS Tutorial
http://www.collaxa.com/pdf/collaxa-bpel101-a.pdf

Gentlemen,

  Pardon me but I must vent, since we are on the topic of treachery.

http://www.newyorker.com/fact/content/?030331fa_fact1

The terms workflow and business process management (BPM) have been overloaded with very different meanings.
On top of that, a BPM systems can be seen from different points of view : inter-business communication, process modelling, enterprise application integration (EAI).
There's a lot of combinations in there and in my opinion, that explains the big differences between the tools and hence in the standards, pushed by their respective companies.
One thing that fascinates me is how the big players keep on trying to reverse engineer their tools to create standards for it.

It seems to me that most of these tools lack a theoretical foundation. Some very interesting work in that area is being done by professor Wil van der Aalst. He distilled a catalog of workflow patterns analogue with the design patterns for object oriented programming. See http://tmitwww.tm.tue.nl/research/patterns/

For the industry, the lack of standardisation is more dramatic then most people imagine. It prohibits a lot of companies to make serious investments in the area.

Has anybody got an idea on how to get out of this standardisation mess ?