Home Messages Index
[Date Prev][Date Next][Thread Prev][Thread Next]
Author IndexDate IndexThread Index

Re: Microsoft Memo to Partners in Sweden Surfaces: Vote Yes for OOXML

On Aug 29, 9:39 am, "ness...@xxxxxxxxxxxxxxxxxxx"
<ness...@xxxxxxxxxxxxxxxxxxx> wrote:
> Quotes are from groklaw article, below are some links.
>
> <Quote>
>  IDG in Sweden is reporting the contents of a leaked Microsoft memo
> sent to Microsoft partners there, telling them to join the Swedish
> Institute of Standards and vote yes on OOXML.

So at this point, Microsoft has even admitted to bribery.

> IDG hence calculates that it cost Microsoft a little less than $30,000
> to get a Yes vote in Sweden. Is this kind of pressure to tip a vote
> allowed under the ISO rules, by the way?\

Unfortunately, yes, this is not so uncommon.  Vendors often recruit
clients and customers to vote in favor of their initiatives.  It's one
of the reason that the ISO OSI specification failed and so many
(billions) of computers use TCP/IP instead of OSI.  For
telecommunications, the OSI and it's lobby effort cost them their
vote, and IETF which only requires that a specification be
sufficiently complete in specificatiotn that an undergraduate can
implement the specification within a few weeks - became the dominant
standards body.

OOXML is a "pig in a poke" and an ISO endorsement won't change that.
On the other hand, ODF has been implemented in Open Source (Open
Office), which not only assures that the standard is implementable,
but also that it is easy to validate proprietary implementations of
the standard against the OSS implementation.

Conversely, opening Word documents is a crap shoot.  If you open the
document with the version used to create it, and you haven't enabled
the security features such as disabling macroviruses, then you might
get the expected result.

> If so, maybe it's time to
> tweak? Shouldn't voters at least have to understand the specification
> they are voting on?

They should, but it is not required.  When OSI votes came up, IBM, HP,
DEC, and Novell were all lobbying to get their proprietary protocols
included as part of the specification, either as core or supersets.

If you do NOT have ISO certification, it's much harder to get the
technology approved, but even then, the ISO Subsets and supersets can
allow a vendor to implement a subset that only nominally applies.
Remember POSIX?  Posix was approved at levels 1, 2, and 3.
Practically everybody was level one compliant - even Windows NT, VMS,
and MVS.  On the other hand, Level 2 and 3 pretty much mandated a UNIX-
like kernel, and assured true compatibility.  Most Unix vendors
offered full posix compatibility, including Linux.

When customers specified that a platform be POSIX compliant, they
meant full complaince, for government contracts and such.  This would
get lost when the decision to choose NT 4.0, VMS, or MVS was being
made.

In many cases, where full posix compliance really was needed, the
failure to check for level 2 and level 3 compliance often cost
$billions in rework over the subsequent years.

One of the big "features" of OOXML is the ability to include
"blobs" (binary large objects), within the XML message itself.  You
can wrap an entire word document in an OOXML wrapper, and be
"compliant".  You can also wrap WMA, PPT, XLS, and VSD objects within
OOXML, and still be compliant.

Of course, this pretty much nullifies all of the reasons for using XML
in the first place.

> The public is going to be seriously impacted by
> the results of this vote. Shouldn't we expect voting to be based on
> having a clue, not just based on a list of talking points provided by
> a vendor insisting its standard be approved?
> </Quote>
>
> http://www.idg.se/2.1085/1.118337http://www.groklaw.net/



[Date Prev][Date Next][Thread Prev][Thread Next]
Author IndexDate IndexThread Index