Linked by Thom Holwerda on Thu 6th Aug 2009 22:04 UTC
Thread beginning with comment 377620
To view parent comment, click here.
To read all comments associated with this story, please click here.
To view parent comment, click here.
To read all comments associated with this story, please click here.
RE[3]: This ballot isn't needed.
by hyper on Fri 7th Aug 2009 06:19
in reply to "RE[2]: This ballot isn't needed."
If one was cynical, it would seem that Microsoft sought out as many parts of the ODF specification that could possibly be interpreted in different ways as it could, and then implemented ODF 1.0 in an incompatible and non-interoperable way under that different interpretation.
Then the problem is in both ODF specification and its Microsoft implementation. What kind of "specification" can be interpreted in an "incompatible way"? The answer IMO is: a very bad one. Oh well...
RE[4]: This ballot isn't needed.
by JohnFlux on Fri 7th Aug 2009 07:03
in reply to "RE[3]: This ballot isn't needed."
Then the problem is in both ODF specification and its Microsoft implementation. What kind of "specification" can be interpreted in an "incompatible way"?
All of them. If you take any non-trivial specification and start "language-lawyering" it (as Linus like to say) you can make an implementation that isn't compatible.
To give a random example, Microsoft's .docx format didn't specify in the specification whether the sin(x) function, in Excel, should take radians or degrees.
Any computer scientist would know that they mean radians, and it can always be checked against Excel, but someone wanting an incompatible implementation could just take it to mean degrees.
Even in the POSIX world where specifications were made as specific as possible there are still some ambiguities - just see the kernel mailing list.
RE[4]: This ballot isn't needed.
by Vanders on Fri 7th Aug 2009 10:12
in reply to "RE[3]: This ballot isn't needed."
If Microsoft (who apparently employ some of the smartest developers in the world, so we are told) had problems interpreting parts of the ODF specification, they only had to ask the Oasis group for guidance.
Everyone else managed to implement ODF themselves without the problems Microsoft apparently had.
RE[4]: This ballot isn't needed.
by segedunum on Fri 7th Aug 2009 10:29
in reply to "RE[3]: This ballot isn't needed."
What kind of "specification" can be interpreted in an "incompatible way"? The answer IMO is: a very bad one. Oh well...
That statement is so stupid and devoid of logic that it isn't even funny. Anybody with two brain cells to rub together knows the weak link there is the interpreter.
RE[4]: This ballot isn't needed.
by lemur2 on Fri 7th Aug 2009 11:44
in reply to "RE[3]: This ballot isn't needed."
"
If one was cynical, it would seem that Microsoft sought out as many parts of the ODF specification that could possibly be interpreted in different ways as it could, and then implemented ODF 1.0 in an incompatible and non-interoperable way under that different interpretation.
If one was cynical, it would seem that Microsoft sought out as many parts of the ODF specification that could possibly be interpreted in different ways as it could, and then implemented ODF 1.0 in an incompatible and non-interoperable way under that different interpretation.
Then the problem is in both ODF specification and its Microsoft implementation. What kind of "specification" can be interpreted in an "incompatible way"? The answer IMO is: a very bad one. Oh well... "
Three problems with your reasoning:
(1) There are a number of independent implementations of ODF in various Office suites other than Microsoft's implementation, and all of them managed to be interoperable, and
(2) Microsoft itself has offered a specification to the ISO, which has been accepted as ISO 29500, and there is not one single implementation of that "standard" anywhere on the planet (let alone a number that can interoperate), and
(3) There are files that one can load and save in ODF formats that allow one to TEST compliance. Microsoft's implementation of ODF fails these test cases miserably.
QED.
The defence rests.
Edited 2009-08-07 12:03 UTC





Member since:
2007-02-17
It doesn't work.
Of all the existing implementations of ODF, Microsoft's is by far the most non-compliant and the only one that doesn't interoperate.
There are at least two plugins from other companies that work far better.
If one was cynical, it would seem that Microsoft sought out as many parts of the ODF specification that could possibly be interpreted in different ways as it could, and then implemented ODF 1.0 in an incompatible and non-interoperable way under that different interpretation.
Such a pity that Microsoft now has to undo all that sabotage effort they have put in if they want to be compliant with ODF 1.2.