Share via


“I am rubber, you are glue…”

Rob Weir is apparently offended that someone would suggest that ODF is controlled by a single vendor, going so far as to accuse Burton Group of some influence over the Open XML ISO voting process. He says:

"Waiting until after February, after the DIS 29500 process concludes, to make corrections is unacceptable. Since your stated purpose in making this report public was to "advance the debate" in the current OOXML ISO process, withholding factual corrections until after that process concludes would imply that you and the Burton Group see no problems with knowingly persisting in influencing an ISO ballot with false information published under the Burton Group name. I don't believe that is the image that the Burton Group would want to project. So I urge that a correction is in order now."

https://www.robweir.com/blog/2008/02/punct-contrapunct.html

This is feedback to three posts from the Burton Group that are responses to the ODF Alliance.

https://ccsblog.burtongroup.com/collaboration_and_content/2008/02/burton-groups-r.html
https://ccsblog.burtongroup.com/collaboration_and_content/2008/02/burton-groups-1.html
https://ccsblog.burtongroup.com/collaboration_and_content/2008/02/burton-groups-2.html

But as we see with Rob over and over again, he seems to be on both sides of the argument. Here's Rob throwing those same sticks and stones at ECMA TC-45.

"This should sound familiar. OOXML is nothing more than the preferences of Microsoft Office. Whenever Word changes, OOXML will change. And if you are a user or competitor of Word, you will be the last one to hear about these changes. ISO does not own OOXML. Ecma does not own OOXML. OOXML, in practice, is controlled and determined solely by the Office product teams at Microsoft. No one else matters." https://www.robweir.com/blog/2007/12/those-who-forget-santayana.html

"So what Ecma is offering SC34 is nothing close to what was promised. Ecma is really seeking to transfer to SC34 the responsibility of spending the next 3 years fixing errors in OOXML 1.0, while future versions of OOXML ("technical revisions") are controlled by Microsoft, in Ecma, in a process without transparency, and as should now be obvious to all, without sufficient quality controls."
https://www.robweir.com/blog/2007/12/bait-and-switch.html

"First, Microsoft has managed to get JTC1 to clamp down on information. What was a transparent process is now mired in multiple levels of security leading to delay, denial of information to some NB participants and total opaqueness to the public."
https://www.robweir.com/blog/2007/12/662-resolutions-but-only-if-you-can.html

These three (along with a lot of other) statements are of course inaccurate. So if we're recommending retracting things for the sake of accuracy, it might be appropriate to start here: https://www.robweir.com/.