We need both open standards and open source that support the standards. This is not an either/or situation and discussions that say one is more important than the other are silly, IMHO. Standards drive interoperability, open source drives ubiquity and lowers the threshold for those who want to get in the game. Therefore if standards are good and useful (and not all of them are), open source can help ensure that there will be widely available, good quality implementations. Adoption of the standards will then take place faster.

We cannot forget why SOAP/CORBA didn't have a great spread the first time it appeared, some coMpanie$ have made a broad campaign on unbelieving about such technologies. In a while they were fully used by free sw developers.

And what about the proprietary software hidden interfaces? Yeah, they keep secrets for they own. Look at the .net specification (which as published, is supposed to be public, so that anyone could port to other environments, etc), the try implementation "Mono" is still skidding because of lack of information, they cannot implement what is not said to be implemented.

So that's right that Open Standards help a lot (if really open), but Open Source may increase much more the progress of technology, since there are intellectual capital being carried with it, which may be rose by a community and companies interested on it.

And what about

The Linux Standards Base, or LSB, combines Open Source and Open Standards. In so doing the LSB has succeeded to:

1. Make application developer's lives easier
Application developers can spend more time building enhancements rather than verifying that their code works not only on the many Linux distributions but the different versions of a single distribution (i.e. Red Hat 7.3 vs. 8.0, etc.)

2. Make operating system vendors lives easier
By adhering to the standard, operating system vendors can reach a wider market, assuring that portability of applications is not an issue with their products.

3. Make end users lives easier
Customers are starting to demand LSB Certified solutions due to their promise of easier system administration, applications that run "out of the box" and lack of vendor lock-in.

4. Grow the Linux market
The LSB is setting the stage for the next wave of Linux market growth: commercial applications.

As of 2003 every major Linux distribution vendor in the world has voluntarily applied for and achieved LSB Certification. Additionally, the Free Standards Group, the organization responsible for the LSB, has been recognized by ISO and will be submitting the LSB for ISO transposition later this year.

The LSB is but one more example that the powerful combination of Open Source and Open Standards helps everyone.

All true, but OSS is one of the top implementors of open standards. Heck, Jakarta is the reference implementation for the Java servlet/JSP standards.

May standards were also created by these committees that haven't gone anywhere. PKI has myriad standards that are so complex and have interoperability issues that it's laughable.

Sure, HTML is great, but the OSI create SGML before it, and that failed big time because it was too complex.

TCP/IP is great, but the OSI came out with their own 7 layer communications stack and it never did take off. TCP/IP wasn't just a standard, but it was written into Unix and the code for it was publicly available.

LDAP has worked pretty well, though it's not truly the same as the proprietary vendors implementations, such as Microsoft's Active Directory. And again, the OSI created X.500 for this, and it went nowhere fast.

The XML standards are suffering and benefiting in a similar way. XML itself is quite easy and clean and is being used a lot. But the many other standards that surround it haven't been adopted anywhere near as quickly, and SOAP needs to be renamed because it's far from SIMPLE anymore.

And we don't need to mention that the standards bodies came out with ASN.1 for encoding and CORBA for interoperable object messaging, yet it was the early SOAP over XML, loosely based on the far simpler and more elegant XML RPC base.

OSS is great because it implements these standards in a way that people can actually check and verify. Most proprietary implementations of "open standards" often fail because they just can't help themselves making their products "better" than the standards by filling in the "missing pieces." What a joke.

Anyway, they play best together. What we're suffering with now is standards being created before they are developed and proven by working systems, and that's why many new XML standards getting more complex and less adoptable.