Standards Panel at OSBC

Back to blogging. Too easy to get wrapped up in work and home and not put in the time for the blog.

On Feb. 15 I spent an hour on a panel moderated by Andy Updegrove (Gesmer Updegrove LLP), and joined by Tim Bray (Sun), Bob Sutor (IBM), and Stephen Walli (Optaros). The topic of the panel was the relationship between open source and open standards.

Andy did a good job of setting the tone of the discussion by pointing out the fact that there is little direct commonality between open source and open standards. The amount of hyperbole on the subject is frequently not helpful in that OSS is a method for creating software and standards a process for defining specifications. Those specifications may be built as open source (ignore licensing for a moment) or under any number of other development models.

Clearly, the overlap comes in when considering licensing. There are established licensing practices in the standards world whose goal was to create an environment of "reasonable" terms that were "non-discriminatory." In other words, participants and implementers would feel safe to engage and thus the standards would meet their ultimate goal of adoption. The question that comes to the forefront is how do OSS licenses interact with the terms associated with standards licensing? 

There is no easy answer to this question, and not being a lawyer I will leave it alone for the time being. Suffice it to say that there are concerns with patents, sub-licensing of modifications, conformance and a range of other items. Yet, it is clear that standards have been implemented in OSS projects so no one is overly clear on how compatible or not the licenses are.

Back to the panel session: each panelist gave their two or three minute spiel and each was along expected lines of thought given the companies that the panelists represent. In many ways, the points of view expressed follow the business models of the various companies. I will blog further on this in the future.

One point that seems to have been a real theme during the conversation was around the idea of multiple implementations. Stephen called out the details of the discussion in his blog of the panel. Everyone on the panel but me (hmmm...familiar territory at an OSS panel) emphasized the fact that there are multiple implementations of ODF and only O12 from MS for Open XML File Format. In fact, this was echoed in Gavin Clark at the Register piece today in the Register. From Stephen's blog,

"… in his explanation I thought I heard [Jason] say that Microsoft is defining the new Microsoft Office standard indeed to encourage multiple implementations, AND to encourage developers to innovate around the Microsoft Office document formats. In the first case, it would be bad business to encourage multiple implementations of a product that is responsible for 50% of the revenue stream. In the latter case, it's just another vendor specification to benefit the vendor regardless of the standards imprimatur"

The work at ECMA on the Open XML File Format will hopefully result in an open standard enabling multiple implementations of the file format. This means that non MS-Office applications will be able to generate documents to the standard format, thus making those documents fully compatible with office. Additionally, those non-MS Office applications will be able to consume the Open XML File Format docs from MS Office for use. That is very different than this standard enabling an independent version of the MS Office product. 

The working group at ECMA working on the Open XML File Format is made up of customers, competitors, and partners of Microsoft. The rules of ECMA enable any organization (including IBM or to participate in the process. Furthermore, the specification has already been altered based on suggestions from members of the working group. This suggests to me that there is strong interest in the future of this specification and that it is likely to be received well many customers and competitors alike. Thus, there are likely to be multiple implementations of the specification.

My best guess is that Microsoft will continue to be the only one building Microsoft Office.




Comments (4)

  1. Simon Phipps says:

    > The rules of ECMA enable any organization (including IBM or

    > to participate in the process.

    Actually, as Ecma has no "individual membership" category, and as is not currently an independent legal entity but rather an open affiliation of volunteers, can’t participate at Ecma as far as I am aware. But as the TC45 charter effectively forbids anyone changing the Office 12 format substantively during the process without Microsoft’s blessing, I have to agree with the position Bob Sutor of IBM asserted that participation seems largely pointless anyway apart perhaps for filter writers, especially on non-Windows platforms where some of the Office 12 format’s features will require reverse engineering the GDI.

  2. jasonmatusow says:

    Always a pleasure to hear from you Simon – and I stand corrected on the participation if they are not an incorporated non-profit entity at this point. If they are (or become one) then they could participate at no fee.

    I think you raise reasonable concerns about the O12 piece, and that is a topic that should probably get its own blog entry rather than a response to a comment.

  3. Sam Hiser says:

    <em>"This suggests to me that there is strong interest in the future of this specification and that it is likely to be received well many customers and competitors alike. Thus, there are likely to be multiple implementations of the specification."</em>

    Jason, this indicates wishful thinking…at best. If your colleagues are relying on the old "deployment power" of Microsoft’s new initiatives, they’ll need to check with the Zeitgeist. There are a few old market choke-points where you no longer hold the valve on the air supply:

    -file format

    -OEM / retail (the Google Desktop vig will bump you at Dell et al.)

    -API’s (Web applications are in)

    -Web pages (governments & banks have been scolded and are writing to Firefox [20% share in Europe).    

    You’re not dealing with a full deck if you & your colleagues continue to assert ECMA as ‘addressing customers’ needs’. It’s either cynical or delusional to believe that customers want to be tied to your stack(s). Office Live, maybe…

    I keep hearing glowing phrases about all the ISVs who are going to develop to the ECMA spec. It’s fantasy. Poppycock! There’s no basis for it.

  4. Wesley Parish says:

    Actually, with a hiss and a roar, a roll of trumpets and a blare of drums, and the terrified bleating of a totally confused sheep dog – not forgetting the confused yelping of an equally terrified sheep 😉 – I declare myself an ISV! – at least for the purposes of scratching together an unlikely assemblage and calling it "Office Miniatures", containing some of the oldest PD code and some of the latest unmaintained-but-GPLed application code there is.

    For what it’s worth, I’m going to be focusing on getting it compiling for a start; then I’ll be developing it to use ODF.  That way I can count on the ODF community – which I know is already in existence – to lick it into shape.  If and when Microsoft want to prove that OpenXML is for more than just Microsoft products, I’ll welcome their participation – on an equal basis with the rest of the team, which I hope will form up … but then, it’s not in my hands, unfortunately.

    How many people really want to write an Office Suite noted for small size rather than sheer bulk? 😉

Skip to main content