ECMA-376 Implementation Notes for Office 2007 SP2

Implementer Notes Just Make Good Sense

Stephen Peront — Fri, 16 Jan 2009 18:47:39 GMT

I am pretty excited about our release of the ECMA-376 Implementer Notes . These notes provide a wealth

Microsoft Releases ECMA-376 Implementation Notes

Eric White's Blog — Fri, 16 Jan 2009 22:54:22 GMT

Today, Microsoft released ECMA-376 implementation notes for Office 2007 SP2. These notes are an invaluable

Great News for Open XML Developers: ECMA-376 Implementation Notes for Office 2007 SP2 are here!

Erika Ehrli — Fri, 16 Jan 2009 23:32:00 GMT

We've published today a new set of Open XML implementation notes. The ECMA-376 Implementation Notes helps

ECMA-376 Implementation Notes for Office 2007 SP2

chris.strevel [blog] — Sat, 17 Jan 2009 03:33:13 GMT

El día de hoy muy temprano me desperté y leí sobre este post que acababa de escribir mi buen amigo Doug

[Open XML] Le détail de l’implémentation Open XML d’Office 2007 SP2 en ligne

Julien Chable — Mon, 19 Jan 2009 17:41:16 GMT

La nouvelle n’est pas vraiment nouvelle puisqu’elle avait été anoncée jeudi dernier (ayant été en déplacement

re: ECMA-376 Implementation Notes for Office 2007 SP2

hAl — Mon, 19 Jan 2009 18:40:09 GMT

How can I distinguish incoming Office Open XML files between 1st edition and second edition/ISO versions ?

re: ECMA-376 Implementation Notes for Office 2007 SP2

rebound — Tue, 20 Jan 2009 17:17:35 GMT

Ambiguity and errors in the specification are essentially poison for implementors, since the Microsoft Open Specification Promise only covers a third-party implementation "... to the extent that it conforms to a Covered Specification ...". Conformance becomes a risky and dangerous proposition for third parties, where protection against copyright infringement and/or patent claims vanishes if their interpretation of the specification is wrong.

I believe that Microsoft needs to explicitly extend its promise to clarify this case, as otherwise its promise sounds nice but is of very limited value to third parties.

--rebound

re: ECMA-376 Implementation Notes for Office 2007 SP2

dmahugh — Wed, 21 Jan 2009 01:13:29 GMT

hAl, I understand that Switzerland has submitted an IS29500 defect report regarding that issue, and it will be a topic of discussion at the WG4 face-to-face meeting next week in Okinawa.

Rebound, this is an interesting question that I assume applies to many standards implementations. I’ll ask around and try to understand it more. Do you happen to know how other tech companies like IBM and Sun address these sorts of issues?

re: ECMA-376 Implementation Notes for Office 2007 SP2

hAl — Wed, 21 Jan 2009 15:27:10 GMT

@dmaghugh

If not explicitly found in the spec then can you tell us how Office 14 will distinguish the versions as office 14 is already found in alfa version and should already be using both versions I asume.

Document Interoperability Initiative, document-format implementation notes, and more…

Interoperability @ Microsoft — Thu, 22 Jan 2009 01:31:57 GMT

Hi, my name is Jas Sandhu and I am an evangelist on the Microsoft Interoperability Strategy Team. I manage

re: ECMA-376 Implementation Notes for Office 2007 SP2

Jesper Lund Stocholm — Thu, 22 Jan 2009 11:16:31 GMT

Doug, hAl,

I agree with you both that this is quite essential and that we should take care of this as soon as possible.

However - as I understand the ISO-process, we (SC34 Working Group 4) will need to create a suggestion for a COR and have it approved by SC34. If this is correct there is sadly a time-span of maybe 60 or 90 days before it can be approved.

Maybe it will can be approved by the SC34-plenary in Pragh in end of March?

rebound:

The /practical/ application of an OSP, ISP or CNS is to ensure that the legal foundation of licensing of whatnot technology has been formally written down. With this implementers will start to dig through the covered specification and interpret it any way they see fit to be able to implement it. If there are ambiguities they will choose whatever "makes them feel good" and then get on with their work.

These problems exist in all standards and most implementations of ODF, PDF or OOXML will to some extent be "semi-nonconformant" since interpretation-decisions will have to be made during implementation.

The above is not to say that OSPs, ISPs or CNS's are not important and should not be required. I am just, to the best of my knowledge, not aware of a single implementer ever being sued for implementing e.g. ODF in a non-conformant way (or implementing DOC, for that matters)

re: ECMA-376 Implementation Notes for Office 2007 SP2

André — Fri, 23 Jan 2009 01:39:22 GMT

For an commercial implementer license uncertainty is a massive problem. For someone who implements software for a company obviously not, as his company takes the risk.

Microsoft did not show any good will to address concerns raised against their OSP. Contrary to what has been claimed Randz licenses are not offered as they were "not necessary". What matters is what the courts say. You cannot build an investment decision on shaky legal foundations. It is a simple application of Murphy's law to assume that potential legal risks are real.

re: ECMA-376 Implementation Notes for Office 2007 SP2

rebound — Sat, 24 Jan 2009 17:21:40 GMT

Thanks for the responses to my query.

I'm taking the liberty of imagining myself as a manager in charge of a team that's implementing ISO29500, looking for the risks attached to that undertaking, and seeing if the available information and legal framework are sufficient to let me manage the risks.

@Doug: Sorry, I haven't researched the similarities or differences between Microsoft's position and other companies' positions regarding legal protections relating to standards conformance.

@Jesper: If Microsoft and the implementor have differences over the interpretation of a portion of the specification, Microsoft's significant legal resources, large and growing IP portfolio, and the specification ambiguity may generate a significant risk for the implementor.

(In general): While thinking about the responses to this issue, I've found it interesting to note that the GPL is a "bottom-up" approach to distributing documents with a disclaimer-based system for managing risks that could be incurred by the act of distribution, whereas the OSP is a "top-down" approach, not proscribing how the software should be written, but supplying risk mitigation through an evaluation of how the resulting object functions.

The functional approach has some interesting twists: Supposing a change by a supplier (e.g. an API change in an OS) results in a third-party's implementation becoming nonconformant: Where does any incurred liability lie?

Finally, another angle: If Microsoft's implementation doesn't conform to the specification, but varies in a way that is covered by a Microsoft-held patent, how can a software manager deal with the risks of this situation? (This is a variant of the "bug-for-bug" compatibility issue that can sometimes occur when a dominant supplier's product becomes a standard.)

Thanks again for your consideration of these comments.

--rebound

re: ECMA-376 Implementation Notes for Office 2007 SP2

dmahugh — Sun, 25 Jan 2009 22:12:07 GMT

Andre and rebound:

I'm not a lawyer so I don't think I can add much to these sorts of theoretical discussions of IP concepts. I will say that we believe our approach provides appropriate protections for real-world scenarios and is commensurate with what other vendors are doing in this area. I'm curious, have you guys expressed similar concerns about the approach other vendors are taking?

re: ECMA-376 Implementation Notes for Office 2007 SP2

rebound — Mon, 26 Jan 2009 23:41:17 GMT

@Doug,

Thanks for the reply. Underlying my position is basically "software consumer fatigue" -- I'm sick and tired of writing software to reinvent the wheel ever since I started programming in the late 1970s. I want to be able to think mostly about my client's needs, rather than wrestle with changing interfaces to underlying services. An example of this is that I've settled on using OpenGL as a portable 2-D graphics interface, as it gives me the flexibility to change hardware platforms and/or operating systems without much hassle.

On the subject of the OSP and ISO29500 conformance versus errors/ambiguity: Rather than a vague notion of "IP concepts", I'm looking specifically at patents, and in particular wanting to clarify what the OSP means by the phrase "to the extent that it conforms".

@To all: Do you have any suggestions for an alternative place (another blog, perhaps?) where these discussions might be able to move forward?

--rebound

Standards-Based Interoperability

Doug Mahugh — Fri, 05 Jun 2009 17:02:29 GMT

There has been quite a bit of discussion lately in the blogosphere about various approaches to document