Document Formats Discussions

It is worth recapping some interesting conversations related to a blog post I made last week about the OpenDocument Foundation support of CDF (Compound Document Format). The points I was making were picked up by a few reporters - Mary Jo Foley at ZDNet and Martin LaMonica at CNET for example. There was also a brief article by Lucy Sherriff (my blog was not mentioned there) at The Register about the question of suspended development for the da Vinci project from the OpenDocument Foundation. Moreover, Gary Edwards has put up a big blog posting (long read) explaining his position on these issues. You might want to read the Rob Weir posting on the OpenDocument Foundation as well. Also, you should really look at Tim Anderson's comments about Adobe to round out the discussion a little.

Ok, where does this leave us? First, the real issue is being to surface in the threads above. ODF originally was a representation of the features in OpenOffice. Open XML originally was a representation of the features in Microsoft Office. Adobe's Mars project is likely (I have not done any digging on this one) a representation of their applications. This means, all of these formats have different strengths and weaknesses (performance, file size, custom schema, other capabilities...). It means that developers making decisions about what format to take advantage of has a rich set of choices open to them today. The fact that XML underpinnings are being married to standardization of the high-level formats is a good thing. That standardization should not become a mechanism for slowing/stopping innovation in those applications. That does not serve customers well.  

To highlight my points - two quotes from those links above:

Gary Edwards points out about ODF and Sun:

Sun has already made it clear at the OASIS ODF TC that they are not going to compromise (or degrade) the new and innovative features and implementation model of OpenOffice just to be compatible with the existing 550 million MSOffice desktops.

Tim Anderson points out about Adobe:

It [Adobe] wants the world to accept its runtimes and formats as standards, while preserving its commercial advantage in controlling them.

My take-away on this is that commercial competition, not in document formats, but in the applications space - is the real juice behind the document format discussion. I'm fine with that, but it seems unsavory to many and takes away from the purity message of the ODF advocates. But, IBMs investments in ODF are not about making Sun wealthy, nor are they about their customers. Their investment on ODF is meant to create market opportunity for their products and services...period. Again, I'm jiggy with that - but let's talk directly about these issues rather than around them.

Next, I have seen a few comments about the Open XML Ballort Resolution Meeting (BRM) recently that I wanted to address today as well. None of the blogosphere postings about ODF/CDF/Open XML will directly affect the BRM process. The Ballot Resolution Meeting will move ahead based on the quality of the work done by the Project Editor and Ecma TC45. I did read one incorrect statement from someone - so let's make this clear. There were more than 3000 comments with the votes (yes, no, and abstain). The work being done now is to respond to all comments in a professional, and thoughtful manner. The National Bodies will then need to consider the dispositions presented in order to determine how it affects their vote. If they feel the dispositions are satisfactory - and had previously voted no, they may change to yes (or of course choose to leave their vote as no). Generally speaking this is the accepted practice. The rules do not exclude the possibility of a yes moving to no, but that is not common practice. As for the process at the BRM itself - they will clearly need to think about how to categorize the issues and how to efficiently work through them. My understanding is that approximately 50% of the comments are duplicates, and a significant portion are editorial in nature. The real work, and discussion, will focus around a much smaller number of points. I am sure that will be more than enough for some very serious, high quality discussion at the meeting.

Comments

• Anonymous
  October 30, 2007
  PingBack from http://msdnrss.thecoderblogs.com/2007/10/30/document-formats-discussions/

• Anonymous
  October 30, 2007
  Just to be clear, Jason, the assertion Gary makes in that quote is ridiculous. As you know Sun and Microsoft work (at a suitably professional distance!) on interoperability all the time, so a general stance opposing it is unthinkable. The reason the three-man-Foundation's ideas were not adopted at the ODF WG at OASIS were because they were either not proposed, not thought through or not appropriate - I checked with multiple parties. I have no idea why the trifecta are so poisonous about Sun, but I suggest it's not a valuable data point.

• Anonymous
  October 30, 2007
  Jason, Over the past week or so, I've been coming to the conclusion that the whole file format discussion is slightly misfocussed.  It's a long argument, so I'll just post the foundation of my house of cards for now. In my opinion, free market economics should be seen as a two-stage process: first, producers decide what's technically possible, and create a set of alternatives; second, consumers decide which alternative is most valuable to them, and vote with their wallets.  The key here is that producers shouldn't be making value judgments, and consumers shouldn't be making technical decisions.  When producers decide amongst themselves what the consumer wants, you get a situation like in video storage - after 30 years, they still haven't realised that consumers want convenience, not picture quality.  When consumers are asked to make technical decisions, you get a situation like Unix in the 80's - it creates uncertainty and arbitrary differences that drive value through the floor. In the file format discussion, all the alternatives we're discussing violate these rules.  If producers were to agree on a single format, they'd have to decide which features were valuable to consumers.  But if they just present the current alternatives to the market, consumers won't understand the technicalities well enough to make a good value judgment.

  • Andrew

• Anonymous
  October 31, 2007
  This is straight shooting from you, Jason. Where we differ: we consider that a Universal Document Format is quite different from your conception in that conformance of all applications is mandatory and yet the format is conceived to hold and preserve all elements. Proprietary, invisible or otherwise rogue elements created and fostered by any single or subset of applications would be excluded by the design of a rich and multifarious market-place of FULLY-CONFORMANT applications, all first-class citizens. It pre-supposes the possibility of a wide choice of applications, diversity of applications and a rich diversity of features & functionality around a single format. This diversity excludes the likelihood of a single dominant application emerging. The more diverse the <em>application</em> field, the more secure the ecosystem. In documents, we haven't seen an ecosystem like this or a format like this yet; but we are -- perhaps Quixotically, you might argue -- working on it. The Web of course is the Jungian archetype for this model around (X)HTML. Not coincidentally, Office 2.0 provides an early snapshot of this kind of rainforest, and we welcome Office Live and all the interesting services you will provide through that platform with equivalent access alongside so many other promising solutions that Microsoft doesn't control. In view of this egalitarian outlook, Microsoft's strategy of cementing the shift of control of customers' data and business processes to around OOXML and Sharepoint | Exchange makes me belly-laugh.

• Anonymous
  October 31, 2007
  The comment has been removed

• Anonymous
  November 01, 2007
  Sam, I wish you good luck with your utopian ideas.  They fall into the category of "if you build it, they will come", i.e., if you define a fixed Universal Document Format, then applications will magically appear to support it.  (And, you undoubtedly hope, displace Microsoft Office or render it just one of many such applications.) You should consider that:

• People don't buy document formats, they buy office applications, i.e., they will look first at an office applications capabilities first, and its document format second.
• There are tremendous costs to switching office applications that have nothing to do with document formats.   Even if all office applications used your UDF, that is probably not incentive enough to make switching any easier than it is now.
• People's needs vary greatly, so it is extremely difficult to come up with just one format.  How would a fixed UDF address this, without the use of a defined extension capability (e.g., like custom schemas in OOXML)?
• The needs of individual users and large organizations are very different.  The former will want to just run the office suite and don't care about the format; the latter want to run their current applications (i.e., MS Office) but want the capability to easily interchange data with great fidelity (i.e, they want the capabilities that are offered currently only by OOXML).  How does a new UDF address these facts?
• People's needs change over time.  How does a fixed UDF set by a committee address this? Regards, Ian

• Anonymous
  November 01, 2007
  The comment has been removed

• Anonymous
  November 03, 2007
  Ian, Jason - these are truly ignorant things to write - "... the user (writ large) is not interested in the format, they are interested in the capabilities of the apps" "- People don't buy document formats, they buy office applications, i.e., they will look first at an office applications capabilities first, and its document format second." Nearly EVERYONE that buys an application checks to see if it will be compatible with the formats they deal with. What other reason would Microsoft have had for supporting Lotus 123 to the point of including a defective calendar? Because they were being nice to Lotus? Hardly - because Lotus users would not change applications until compatibility was ensured. All the early home adopters of pc compatibles (oh yes, that format thing again) 'borrowed' the applications purchased for the office ot use at home for the reason that the format for executing the applications was compatible with their home machines and so were the  formats used by those applications so they could bring work home or use the office printers to print the work they did at home. Just for Ian - "- There are tremendous costs to switching office applications that have nothing to do with document formats. " That is certainly true, so why do would you support Microsoft in making such huge changes in format and interface? Is it to increase training costs? Is it to ensure that more .docx documents are made, forcing non-Office 2007 users into buying an application with features they don't use? The following quote shows a below ignorant  grasp - "- People's needs vary greatly, so it is extremely difficult to come up with just one format.  How would a fixed UDF address this, without the use of a defined extension capability (e.g., like custom schemas in OOXML)?" The defined extension capability in MSO-XML is realizeable entirely by zipping the XML content along with whatever document it might relate to. This can be added to -any- application. No-brainer. Zip your custom XML to the file(s) you want and voila - all except for embedding that content into MS apps  - because MS didn't want to enable that capability. As to the variability of user's needs, Microsoft has provided the -one- format (well, several, but generally backwards compatible) for office products for at least a decade. Perhaps you haven't noticed. Somehow, though, governments and individuals have tired of that single-sourcing and the license raids and the difficulties caused when Microsoft changes MS formats - e.g. the scientific publishers who went to the trouble to create a converter that the new MS formats break, as well as the learning curve involved with the interface changes.

• Anonymous
  November 04, 2007
  There has been quite a bit of Open XML discussion in the blogosphere recently, and I've not posted on

• Anonymous
  November 04, 2007
  There has been quite a bit of Open XML discussion in the blogosphere recently, and I&#39;ve not posted

• Anonymous
  November 05, 2007
  Ian - gosh! "People don't buy document formats, they buy office applications, i.e., they will look first at an office applications capabilities first, and its document format second." Actually Ian, people look at formats first and application capability second. Few people want an application that is an island, incapable of sharing with others. Microsoft supports my position. They made Word compatible with Word Perfect, at least enough to satisfy Word Perfect users so they could safely move to Word and not lose all their work. Microsoft even carefully copied the Lotus 1-2-3 bug for dates, just to encourage 1-2-3 users to move to Excel. And what, do you suppose, is the reason for the HDVD vs Blu-Ray selection difficulty? No player capabilities are considered by the average shopper as much as they consider whether media using that -FORMAT- are likely to be available for a sufficient time. "People's needs vary greatly, so it is extremely difficult to come up with just one format." Why does Microsoft push .docx with XML as the default for Office 2007? Are MS users an entirely homgeneous bunch who need only the capabilities that MS can supply? When some users try to choose another format in the goverment sector, the MS formats become the -only- solution to -all- the problems a government might have, so say MS men-in-black. You say many users, many formats. MS acts to say, all users, our format only, except for customer appropriation purposes.

• Anonymous
  November 06, 2007
  Deux membres du comité technique OpenDocument à l’OASIS, Gary Edwards (fondateur et président de la fondation