Microsoft publishes 'incomplete' OOXML specs

Microsoft has been accused of publishing "incomplete" specifications for its Office file format binaries.

The Word, Excel and PowerPoint file format specifications, which were previously only available from Microsoft by request, were published on Friday, together with details of an open-source Office binary-to-Office Open XML (OOXML) translator project.

The binaries were published in response to concerns among national bodies voting whether or not to ratify OOXML as an International Organization for Standardization (ISO) standard, according to Microsoft Office program manager Brian Jones. The national bodies were concerned that third-party developers may have had difficulties converting Office binary formats to OOXML, referred to in the ISO ratification process as "DIS 29500".

"To make it even easier for third-party conversion of binary format to DIS 29500, Microsoft agreed to... make it even easier to get access to the binary formats documentation by posting it and making it available for a direct download on the Microsoft Web site no later than 15 February, 2008," Jones wrote in his blog on Friday.

However, the file format specifications were criticised as being incomplete by third-party OOXML developer Stephane Rodriguez.

In a comment posted on Jones's blog post on Saturday, Rodriguez said that Excel's internal format table, known as BIFF, had missing records and a "reserved" specification, while Office Drawing also had unspecified records.

"I first gave a cursory look at BIFF. 1) Missing records: examples are 0x00EF and 0x01BA, just off the top of my head. 2) No specification: example is the OBJ record for a Forms Combobox," Rodriguez wrote. "Then I gave a cursory look at the Office Drawing specs. And, again, just a cursory look at it showed unspecified records."

With the specs criticised as incomplete, Microsoft's Jones announced that the binaries were available under the company's Open Specification Promise (OSP), a more formal version of the "covenant not to sue".

"The binary formats have been under a covenant not to sue and Microsoft will also make them available under its OSP by the time they are posted," wrote Jones.

Microsoft's OSP is a form of licence agreement designed to give software developers peace of mind that Microsoft will not pursue them for patent infringement should they use the binaries, or the OOXML specification, to develop code.

However, the OSP has been criticised by intellectual-property law experts as legally inexact and untested. According to Ronald Yu, a US patent agent and academic, the covenant has never been tested in court, and it includes no mention of any court or tribunal in which to resolve a dispute. Patent coverage is also not explicitly defined, according to Yu, who stressed that the OSP is neither a release nor a contract, and, therefore, needs to be treated with some caution.

Microsoft responded to Yu's criticism by insisting that the FAQs on Microsoft's OSP web page clarify the OSP sufficiently. Steve Mutkoski, regional director of interoperability at Microsoft and one of the lawyers on the team that drafted the licensing terms, described the OSP as an "innovation" that enables a company like Microsoft to gain a quick and broad adoption of a technology without needing to negotiate 20- and 30-page agreements with every party that wants to use it.

Brett Winterford contributed to this article.

Advertisement

Talkback 8 comments

    "OOXML specs"?Dean -- 19/02/08

    This has nothing at all to do with OOXML. The specifications that were published recently were for the BINARY formats, not the OOXML formats. The OOXML formats were published a LONG TIME AGO.

    I think you need to seriously edit this article. The title in particular is very much misleading.

    binary specs part of ooxmlAnonymous -- 20/02/08 (in reply to #320095782)

    Dean-
    These specs are "required" by the OOXML spec becuase they are referenced material in OOXML and should have been include in it in the first place. the OOXML specs is not that old yet, hell there is not even a completed spec for it yet.

    It doesn't matterDean -- 20/02/08 (in reply to #320095861)

    Whether you believe the OOXML specs are complete or not is irrelevant. If you look at Brian Jones' blog, you'll clearly see that Stephane Rodriguez is talking about the BINARY specifications when he says they are "incomplete".

    This article is taking that statement and attaching it to the OOXML debate, presumably because people are more interested in OOXML than the binary formats.

    Besides, the binary specifications are not "referenced" by the OOXML spec. By that logic, you would expect Microsoft to also release the spec for the Word Perfect format -- it's "referenced" in the same way -- but Microsoft doesn't even control those specifications.

    OOXML require a mapping with the binary formatsAnonymous -- 20/02/08 (in reply to #320095866)

    Dean, you are the one that is misleading people.

    The truth is that serveral NB groups required that ECMA provide a mapping between their binary formats and OOXML so that OOXML can fulfill it mission charter about compability with the binary formats. ECMA responded by promising that Microsoft should make the binary formats available, and totally dodged the reasonable demand for Microsoft to provide the mapping they must have.

    Obviously ECMA:s response would be lacking even if the release of binary formats documentation covered everything since the mapping with the hard to use OOXML format is the hard step when to ensure compability with the legacy formats.

    The situation becomes even worse when you consider that the documentation is severly lacking and does not even explain how to parse the binary formats themselves.

    Let's hope that Microsoft and Ecma:s smokescreens will not trick NB representives to think that the problem with the legacy formats has solved, or even improved at all.

    You're missing the pointDean -- 21/02/08 (in reply to #320095915)

    You're totally missing the point here, Anonymous (if that is your real name).

    This article is talking about the BINARY formats as if they are the same thing as the OOXML formats. They're taking a quote by Stephane Rodriguez relating to the binary formats and applying it to OOXML.

    Whatever you think the problem with OOXML is, the particular quote used in this article has been taken out of context, and applied to a totally different technology. I don't see how you can possibly argue anything else.

    Are we reading the same article?Jeetje -- 21/02/08 (in reply to #320095934)

    Dean, from your seond paragraph I have the distinct feeling you missed the second paragraph of the article completely.

    To me this article seems to sum up the discussion at this point in time quite succinctly. With all due respect, what are you missing in it and what credentials do you bring to the table to tell a journalist from an established e-zine his way of writing articles is flawed?

    Credentials?Dean -- 22/02/08 (in reply to #320095985)

    Um, how about the ability to read and do a bit of research? That's all the credentials you need to realise the article is really misleading. Here, let me make it easier for you.

    The work "incomplete" in the article title implies that the word comes from something SOMEBODY ELSE said, right? But you can see the only quote where somebody uses the word "incomplete" comes from Stephane Rodriguez talking about the binary specifications.

    Also this:

    "With the specs criticised as incomplete, Microsoft's Jones announced that the binaries were available under the company's Open Specification Promise"

    Let's assume for a minute that "the binaries" refers to "the binary format specifications". This sentence implies that the paragraph above (quoting Stephane Rodriguez) happened *before* the binary format specifications were released. But that's obviously not true: Stephane was TALKING ABOUT the binary format specifications.

    Next, let's look at the title of the article, "Microsoft publishes 'incomplete' OOXML specs".

    OK, so the OOXML specs were published some time in 2006, if I remember correctly. So if people are critizing them as being incomplete, then in the least that would be *old news*. But the point is, nobody quoted in this article is talking about the OOXML specs.

    The closest thing that could possibly qualify is:

    "The national bodies were concerned that third-party developers may have had difficulties converting Office binary formats to OOXML"

    Which is certainly true, and a good reason to publish the binary format specifications. But that's not a quote from anybody.

    Don't get me wrong, I think it's about time Microsoft published the binary format specifications. And from what Stephane said it doesn't looks like it's all been published, which is a shame. But to take what he said about the binary format specifications, and try to make it look like he was talking about OOXML is bad journalism.

    To be honest, I don't think it was the result of anything malicious on the part of the author. I think he was just confused by what the binary format specifications were (as evidenced by the fact that he refers to them as "the binaries" at one point).

    (By the way, the blog post that the article is referring to is: http://blogs.msdn.com/brian_jones/archive/2008/02/15/binary-documentation-doc-xls-ppt-and-translator-project-site-are-now-live.aspx)

    More deep analysisAlex Ott -- 22/02/08

    I did more deep analysis of specifications, and found at least 8 not described records, that present in excel97 format
    I plan to write article with deep analysis of their specifications

Add your opinion


Latest Videos

Blogs

  • David Braue Will Rudd's bush backhaul bonanza deliver?
    Rural areas will be welcoming the government's decision to put its money where its politicising is, funnelling $250m into a regional fibre upgrade to six rural centres. Remedying over a decade of near-neglect at the hands of telecoms privatisation, the investment could be the firmest step yet for Labor's NBN dream — but with inevitable political questions and a looming election, Rudd and Conroy need to deliver, and quickly, to preserve the NBN's credibility.
  • Array Doing for AV what VoIP did for telephony
    Sydney-based start-up Audinate is making traditional analog cabling obsolete in favour of TCP/IP-based networking technology. And it's doing a pretty good job so far, with its technology used by World Youth Day and the Sydney Opera House.
  • Array WiMax in Australia: Part two
    WiMax could be the standard that drives the next phase of mobile broadband, it provides an opportunity for players wanting to establish a pure IP network to carry voice and data effectively — but is this what operators want?
  • More blogs »

Tags

Back to top

Featured