LWN.net Logo

Erosion of trust

Erosion of trust

Posted Nov 25, 2007 21:20 UTC (Sun) by ncm (subscriber, #165)
Parent article: The GNOME Foundation on OOXML

Is erosion of trust in the standards process a bad thing when the standards process itself is
being corrupted?  We've already seen the whole ISO debacle from start to finish, and the
effective destruction of that committee by Microsoft manipulation.  With all due respect to
Jeff's coding skills, we've seen no evidence that he has the strategic awareness needed to
protect Free Software from Microsoft machinations.

Microsoft doesn't need education.  They understand perfectly what they want to obtain, and
they have people plenty smart enough to figure out how to get it.  We already know, whenever
it seems like they are behaving as if enlightened, that it is a trick, because we already know
their goal is (as it must be) the destruction of Free Software.


(Log in to post comments)

Erosion of trust

Posted Nov 25, 2007 22:43 UTC (Sun) by jdub (subscriber, #27) [Link]

Please don't respect my coding skills. :-)

The erosion of trust in the standards process is definitely a bad thing, and Microsoft should
be held accountable for the horrifying impact they're having. If I had more of a bent for
conspiracy theories, I'd suggest they were abusing the process not just to get OOXML through,
but to make the process laughable too.

Unfortunately, the community are helping them abuse the process by making OOXML a political
issue -- which actually helps Microsoft, because they can say all disagreement with OOXML is
"political" in nature. Instead, we should be aiming to defeat OOXML purely on technical
grounds (and repudiate Microsoft's actions in the process), in line with the standards
process.

Erosion of trust

Posted Nov 26, 2007 1:55 UTC (Mon) by gdt (subscriber, #6284) [Link]

Unfortunately, the community are helping them abuse the process by making OOXML a political issue -- which actually helps Microsoft, because they can say all disagreement with OOXML is "political" in nature.

The problem here is that the major objection to OOXML is political -- that the world is best served by one standard document format, developed cooperatively between vendors, users and experts in a collegiate atmosphere, with technologies and intellectual property rights which allow care-free implementation by all.

Erosion of trust

Posted Nov 26, 2007 2:04 UTC (Mon) by jdub (subscriber, #27) [Link]

<blockquote>The problem here is that the major objection to OOXML is political -- that the
world is best served by one standard document format, developed cooperatively between vendors,
users and experts in a collegiate atmosphere, with technologies and intellectual property
rights which allow care-free implementation by all.</blockquote>

I totally understand that, and you'll find that almost all GNOME contributors are politically
against OOXML -- but that has no bearing on the ISO decision. In fact, the more "political" we
are, the more Microsoft can say that we're politicising the process... and that has done quite
a bit of damage already. We need to aim for the absolute rout of OOXML on technical grounds,
without all the shrill voices and politicisation. It doesn't make us look good, and it is not
at all convincing as part of the process.

I just hope that people think about and try to understand why we're doing this without making
emotional, dismissive judgements beforehand. Happy to agree to disagree, but highly offended
by the demonisation and insinuation from some quarters. (Not saying you're doing that, just
mentioning it.)

Erosion of trust

Posted Nov 26, 2007 19:16 UTC (Mon) by stevenj (guest, #421) [Link]

I totally understand that, and you'll find that almost all GNOME contributors are politically against OOXML -- but that has no bearing on the ISO decision. In fact, the more "political" we are, the more Microsoft can say that we're politicising the process... and that has done quite a bit of damage already. We need to aim for the absolute rout of OOXML on technical grounds, without all the shrill voices and politicisation. It doesn't make us look good, and it is not at all convincing as part of the process.
Your underlying assumption seems to be that "politics" is bad, and only "technical" discussions are good or appropriate. But politics (from the Greek πoλιτικα, "civic affairs") is simply the process of diverse interests trying to decide on common policy for the common good. The whole question of choosing an international standard is intrinsically a "political" process. It is nonsensical to deride those you disagree with as "political" and "politicizing the process".

With OOXML, as with any standards, the fundamental question is: does annointing this as an international standard serve the public interest? Will making it a "standard" improve interoperability and hence spur competition and communications?

This is not solely a question of whether the OOXML specification, considered in a vacuum, is an adequate technical description of a particular format. The context is also important: the fact that ISO has already recently standardized a document format for essentially the same task means that adding a second format, rather than attempting to fix any deficiencies in the first, will further splinter the market. It will force governments, companies, and vendors to support both formats and will reward Microsoft for ignoring what the rest of the world already agreed upon—an approach that is certain to have repercussions on future standards efforts (e.g. PDF vs. Silverlight).

Erosion of trust

Posted Nov 26, 2007 19:23 UTC (Mon) by stevenj (guest, #421) [Link]

(Whoops, I should have said "PDF vs. XPS." Silverlight is MS's competitor to Flash.)

Erosion of trust

Posted Nov 26, 2007 19:31 UTC (Mon) by jdub (subscriber, #27) [Link]

Your underlying assumption seems to be that "politics" is bad, and only "technical" discussions are good or appropriate.

Not on the whole. The political issues at stake are incredibly important -- I absolutely understand that (being a contributor to the FLOSS community for many years now, it is part of who I am). Microsoft are doing this to king-hit their competitors, most importantly to us, FLOSS.

But: Few of those issues are relevant to the ISO process. That is what I am referring to when saying that the politics are not productive. Microsoft gets to cry "zealot" when we attack OOXML on political grounds, we should not let them manipulate the process this way, on top of all the other ways they've done so already.

I am of a similar opinion to you. I don't want to see OOXML as an ISO standard. What I have been talking about in this thread is the method by which we fight it. I do understand the context of the battlefield.

Artificially limiting the debate in Microsoft's favor

Posted Nov 27, 2007 18:56 UTC (Tue) by stevenj (guest, #421) [Link]

But: Few of those issues are relevant to the ISO process.

To the extent that this is true, we should be vigorously pressing ISO (whose motto is "One standard, one test" after all) to acknowledge that they are relevant.

It seems to me that you've fallen hook, line, and sinker for Microsoft's main tactic to limit the debate: they want to exclude all consideration of whether ISO should have two incompatible standards for office-type documents (which is really the most serious objection to OOXML), and instead focus on extremely narrow technical questions. In much the same way, they reduced the "contradictions" phase of the process to the meaningless criterion that two standards only "contradict" one another if they cannot physically coexist on the same computer.

This is very similar, by the way, to how Microsoft tries to limit the debate about using free/open-source software in government procurement policies. In their "software choice" campaign, they insist that the procurement criteria be artificially limited to price and technical fitness for a short-term task—hoping you'll ignore the fact that vendor-independence, transparency, etcetera are also merits that can and should be taken into account.

Artificially limiting the debate in Microsoft's favor

Posted Nov 27, 2007 19:10 UTC (Tue) by jdub (subscriber, #27) [Link]

To the extent that this is true, we should be vigorously pressing ISO (whose motto is "One standard, one test" after all) to acknowledge that they are relevant.

That motto was used in 2002 and is not relevant to ISO's processes. There are plenty of ISO standards that cover the same ground. The crucial point is that they don't conflict with each other, and generally build upon standards that have gone before.

So let's not argue OOXML on these grounds. They're just irrelevant. The one area this might work for is us that OOXML reimplements absolutely everything and doesn't build on existing standards (ODF uses things like SVG, etc), but attack it under the terms defined by ISO.

I'm not falling for anything -- I'm certainly not falling for some of the uninformed but popular opinions about these things in the FLOSS community.

Erosion of trust

Posted Nov 26, 2007 4:31 UTC (Mon) by jwb (guest, #15467) [Link]

I'd say that a "care-free implementation" is exactly what we already get :)

Political vs. Technical Arguments

Posted Nov 26, 2007 10:04 UTC (Mon) by Felix.Braun (subscriber, #3032) [Link]

The problem here is that the major objection to OOXML is political -- that the world is best served by one standard document format

Sorry to be nitpicking, but to me it seems that the pre-existance of another standard capable of representing essentially the same information is an eminently technical argument. As far as I understand, an introduction of a new standard only makes sense from a purely technical standpoint, if no competing standards exist. Therefore, I took it that Microsoft was argueing that ODF was a technically inferior standard which could not be fixed to include all required information.

Political vs. Technical Arguments

Posted Nov 26, 2007 10:23 UTC (Mon) by jdub (subscriber, #27) [Link]

Sorry to be nitpicking, but to me it seems that the pre-existance of another standard capable of representing essentially the same information is an eminently technical argument. As far as I understand, an introduction of a new standard only makes sense from a purely technical standpoint, if no competing standards exist.

Possibly, but it has no bearing under the ISO rules. It's just not relevant. If OOXML contradicted ODF, then sure.

It's dangerous for us to take the "one standard" argument, because if we wish to introduce a fully Free, unencumbered technology as an ISO standard after setting the precedent (not that it can be set, and not that it's relevant) that there should only be one standard for a particular purpose, we'd be screwed.

Consider, for example, a fully Free, unencumbered video codec (such as the upcoming replacement for Theora) vs. MPEG4. An ISO standard already exists, and it's encumbered up the wazoo. Let's not make long term sacrifices for short term wins.

Therefore, I took it that Microsoft was argueing that ODF was a technically inferior standard which could not be fixed to include all required information.
Microsoft has defined OOXML as a standard for Microsoft Office document archival and interoperability, which cleverly allows them to suggest it is somehow different to ODF.

We need to nuke OOXML in ISO purely on its technical merits -- and there's definitely a lot of room for that!

Political vs. Technical Arguments

Posted Nov 26, 2007 10:51 UTC (Mon) by grouch (guest, #27289) [Link]

We need to nuke OOXML in ISO purely on its technical merits -- and there's definitely a lot of room for that!

Been there; done that. The irreconcilable technical deficiencies of MSOOXML have been thoroughly documented. The GNOME Foundation should cease apologizing for Microsoft's scandalous corruption of the ISO process, cease transferring blame for that corruption to the victims, and begin researching the historical record. I suggest Groklaw's ODF/MSOOXML page as a starting point. It provides a chronology of related news beginning in January 2005 as well as "Resources", "Objections" and "Blogs".

Political vs. Technical Arguments

Posted Nov 26, 2007 10:57 UTC (Mon) by jdub (subscriber, #27) [Link]

Been there; done that. The irreconcilable technical deficiencies of MSOOXML have been thoroughly documented.
I know, and we're facing them down around the world. I'm involved in that process here in Australia.
The GNOME Foundation should cease apologizing for Microsoft's scandalous corruption of the ISO process, cease transferring blame for that corruption to the victims, and begin researching the historical record.
Excuse me? Read the statement! We are nowhere near apologising for Microsoft's scandalous corruption of the ISO process. We raised this issue explicitly in our statement.

Political vs. Technical Arguments

Posted Nov 26, 2007 19:06 UTC (Mon) by grouch (guest, #27289) [Link]

Excuse me? Read the statement! We are nowhere near apologising for Microsoft's scandalous corruption of the ISO process. We raised this issue explicitly in our statement.

You're excused. I did read the statement, including the bizarre bit about "FLOSS implementations of OOXML". (What makes this bizarre is the fact that the specification for MSOOXML is so non-specific; see, for example, WordProcessing ML -- unspecified application behavior).

The 5th point in the "Position" section of the GNOME Foundation statement chastises "community" for the erosion of trust in the standards process and thereby transfers blame from the victimizer, Microsoft, to the victims. You are apologizing for Microsoft's corruption of the process and blaming those who exposed that corruption for the resultant erosion of trust. The process did not anticipate the calculated, deliberate attempts to subvert it which Microsoft used. To borrow the characterization from the GNOME statement, this most certainly is a "black and white" issue, or, more correctly, a good versus evil issue, in that a single vendor has and is attempting to subvert what is supposed to be a global consensus-building process into a means "to control and restrain" any who seek to extricate themselves or their data from utter dependency upon that single vendor.

By seeking the impossible goal of "FLOSS implementations of [MS]OOXML" at a time when there is only one partial implementation of that format, very few documents existing with even that partial implementation, and no established dependency upon that partial implementation yet, GNOME is assisting in creating a dependency on that as-yet deviation from standards and thereby undermining global efforts to produce a document format by true consensus.

Political vs. Technical Arguments

Posted Nov 26, 2007 19:13 UTC (Mon) by jdub (subscriber, #27) [Link]

You are apologizing for Microsoft's corruption of the process and blaming those who exposed that corruption for the resultant erosion of trust.

No, we pretty firmly blamed Microsoft, and also raised the danger of the community's adoption of similar behaviour, because this will fall straight into Microsoft's hands.

The community does have a role to play in uniting our efforts and the demonisation of GNOME has not been productive in that regard. (Let alone the non-practitioners who don't understand how the politicisation of the process -- by either side -- is damaging to us.)

I'm okay to agree to disagree here. There are clearly differing views on the topic.

Erosion of trust

Posted Nov 26, 2007 10:53 UTC (Mon) by ncm (subscriber, #165) [Link]

Unfortunately, the community are helping them abuse the process by making OOXML a political issue -- which actually helps Microsoft, because they can say all disagreement with OOXML is "political" in nature. Instead, we should be aiming to defeat OOXML purely on technical grounds (and repudiate Microsoft's actions in the process), in line with the standards process.

There's nothing political about rejecting a standard proposal that is unimplementable except by reference to code that is a trade secret. It's that way for an obvious reason: to help enforce a monopoly. Is helping to enforce a monopoly "political"? Is pointing that out "political"?

You're being played. Get advice from sound strategic thinkers before it is too late. Follow it even if that means admitting you did wrong. Follow it especially if that means admitting you did wrong.

Erosion of trust

Posted Nov 26, 2007 10:59 UTC (Mon) by jdub (subscriber, #27) [Link]

<blockquote>There's nothing political about rejecting a standard proposal that is
unimplementable except by reference to code that is a trade secret.</blockquote>

That's absolutely right, and this is one of the very effective *technical* criticism we're
using around the world to defeat OOXML. That is not a political criticism under the terms of
ISO, so it's not relevant to what I'm talking about.

<blockquote>You're being played. Get advice from sound strategic thinkers before it is too
late. Follow it even if that means admitting you did wrong. Follow it especially if that means
admitting you did wrong.</blockquote>

I think you misunderstand the situation.

Erosion of trust

Posted Nov 26, 2007 11:06 UTC (Mon) by tzafrir (subscriber, #11501) [Link]

Could you give some specific examples of how it helps for you to be the representetive of the
GNOME foundation "on the inside"?

I think all the posts in this discussion have been over-generelizing ;-)

Erosion of trust

Posted Nov 26, 2007 11:18 UTC (Mon) by jdub (subscriber, #27) [Link]

Well, an easy start is that Jody has pursued our goal: He's worked hard to make sure Microsoft
document as much of the standard as he could get them to during the ECMA process. This will
help FLOSS products implement OOXML, so users can move to FLOSS without sacrificing their
documents or cutting off compatibility with their friends and colleagues. It has also helped
improve the implementation of the binary formats to a certain extent, owing to their
similarity with the XML dump of the format (OOXML). ;-) This is like being given the
opportunity to grill Microsoft about the CIFS protocol.

Secondly, Jody's work has increased the size of the OOXML specification, which has helped
cement a useful technical argument against OOXML as an ISO standard -- one that is
particularly relevant in Australia (where I live), because we're a small country, and 80% of
our companies are small businesses for whom understanding or implementing the OOXML spec is
hideously out of reach.

So there are some good outcomes that are obviously based on our objective for being there, but
there are also some amusing benefits that we didn't expect. ;-)

Erosion of trust

Posted Nov 26, 2007 11:53 UTC (Mon) by tzafrir (subscriber, #11501) [Link]

How much is Microsoft actually commited to follow the changes in the OOXML standard? IIRC
there has been actually some statements of Microsoft seniors of the reverse.

If that is the case, interoperability with ISO-OOXML does not imply interoperability with
MS-Office 2007 (or later).

Erosion of trust

Posted Nov 26, 2007 12:16 UTC (Mon) by jdub (subscriber, #27) [Link]

How much is Microsoft actually commited to follow the changes in the OOXML standard? IIRC there has been actually some statements of Microsoft seniors of the reverse.

Yeah, they've already revealed some weaselly intentions on that front. I'm not too worried about that, we've fought a much tougher interoperability battle with them over the binary formats, but this time we'll be better prepared.

It will also be damaging for them to not be standards compliant (with their own standard!) from a government perspective, so we as a community can hold their feet to the fire there as well.

Note that I'm not suggesting that our participation in TC45-M solves everything, it's just another front in a much larger battle... of which the ultimate goal is Software Freedom!

Copyright © 2012, Eklektix, Inc.
Comments and public postings are copyrighted by their creators.
Linux is a registered trademark of Linus Torvalds