Wednesday I attended the W3C Technical Plenary Day festivities, which included a brief press conference with Tim Berners-Lee, interesting insights into the W3C's work in progress and future plans, and much more (you can view the agenda here). And it also gave me a chance to sit with Chris Lilley, a W3C employee whose responsibilities include Interaction Domain Leader, Co-Chair W3C SVG Working Group, W3C Graphics Activity Lead and Co-Chair, W3C Hypertext CG. What that combination of titles means is that he is the "go to" guy at W3C to learn what W3C's CDF standard is all about.
CDF is one of the very many useful projects that W3C has been laboring on, but not one that you would have been likely to have heard much about. Until recently, that is, when Gary Edwards, Sam Hiser and Marbux, the management (and perhaps sole remaining members) of the OpenDocument Foundation decided that CDF was the answer to all of the problems that ODF was designed to address. This announcement gave rise to a flurry of press attention that Sam Hiser has collected here. As others (such as Rob Weir) have already documented, these articles gave the Foundation's position far more attention than it deserved.
The most astonishing piece was written by ZDNet’s Mary Jo Foley. Early on in her article she stated that, "the ODF camp might unravel before Microsoft’s rival Office Open XML (OOXML) comes up for final international standardization vote early next year." All because Gary, Sam and Marbux have decided that ODF does not meet their needs. Astonishing indeed, given that there is no available evidence to support such a prediction.
So what is this CDF, and should it be considered to be an alternative to ODF? Here’s what Chris Lilley had to say, reconstructed from my notes (in other words, this is not a direct quote):
So we were in a meeting when these articles about the Foundation and CDF started to appear, and we were really puzzled. CDF isn’t anything like ODF at all – it’s an "interoperability agreement," mainly focused on two other specifications – XHTML and SVG. You’d need to use another W3C specification, called Web Interactive Compound Document (WICD, pronounced "wicked"), for exporting, and even then you could only view, and not edit the output.
The one thing I’d really want your readers to know is that CDF (even together with WICD) was not created to be, and isn’t suitable for use, as an office format.
Here are some other takeaways from my conversation with Chris:
- Although they would be welcome to become members, Neither Gary, Sam nor Marbux are members of W3C or the CDF working group
- The W3C has never been contacted by anyone from the Foundation about CDF. After the articles began appearing, the W3C sent an inquiry to the Foundation, and received only a general reply in response
- The CDF working group was not chartered to achieve conversion between formats
- Although he hasn’t spent a lot of time trying to unravel what Gary has written on the subject, he can’t make any sense out of why the Foundation thinks that CDF makes sense as a substitute for ODF
So how do we make any sense out of what the Foundation is saying?
Here’s a start: The Foundation has been very clear that it thinks that the OASIS technical committee has taken the wrong direction in its development approach with ODF. Disagreeing with an architectural approach is, of course, an opinion that any member of any TC is entitled to hold. Unfortunately, the Foundation wasn’t willing to take non-acceptance of its preferred approach lying down.
The simple fact is that the Foundation got out voted. No more, no less, no back story – end of story. Instead of accepting the fact that that’s how an open process can work, however, the Foundation decided to attack the process (It’s not open!), OASIS (It’s controlled by Sun!), and even ODF itself (It’s doomed to fail!). Meanwhile, those that had worked with them in the past fell away, and the only voices that are supporting them, the Foundation or their approach now are, predictably, the proponents of OOXML.
Here’s where things stand today:
- The Foundation’s former members are not supporting Gary, Sam and Marbux
- No one has spoken up to support their view of CDF as an alternative
- They are not working with the W3C at this time, and no one at W3C is working with CDF in the manner they are proposing
- CDF is not an appropriate substitute for ODF
- The CDF working group is not chartered to provide what Gary, Sam and Marbux want to try and accomplish, even assuming that what they want is technically possible
All of which takes us back to the question, What were Gary, Sam and Marbux thinking?
The answer, I think, can be found by applying the wise advice of a 14th century Franciscan friar named William of Ockham. The precept he offered is commonly referred to as "Occam’s Razor," and a common paraphrase goes like this:
All things being equal, the simplest explanation tends to be the right one
Here, the simplest explanation would appear to be simply that when the Foundation’s founders decided to turn out the lights, they decided to poke a sharp stick in the eye of those that had rejected their approach.
If that sounds like too harsh a judgment, we can fall back to the next most charitable one, which is that the founders are so convinced of their own insight that the rest of the world must be wrong – all of those community members in all of those countries around the world that rallied to the ODF cause – must be deluded and not capable of the same clear vision that the founders of the Foundation possess.
Whatever the explanation, the Foundation’s founders clearly do not understand what standards are all about. Standard setting is necessarily a consensus process – no one has to implement a standard, so everyone’s needs must be met to a sufficient degree for them to support the final product. Sometimes that means that the best technical approach is taken, and sometimes it’s the next best.
As with politics, it often takes a great deal of support from many constituencies to successfully launch a standard. In order to succeed against an entrenched product like Office, it takes an unusual amount of support from a very broad coalition of stakeholders. Absent that support, failure is assured.
Time will tell whether Gary, Sam and Marbux are right that ODF can’t succeed. Unfortunately, the way in which they chose to disengage has done nothing to advance their cause. But it has undermined, insignificantly, the good work of the thousands of people that continue to work, in a far more politically realistic fashion, to ensure the success of ODF – not to mention those that sent in donations to a foundation formed:
To support the community of volunteers in promoting, improving and providing user assistance for ODF and software designed to operate on data in this format.
What a shame.
For further blog entries on here , click
sign up for a free subscription to Standards Today today!
* * * * * * * * * * * * * * * * * *
Chris supplied the following information for those that want to learn more about what CDF is really about:
I mentioned four CDF specifications; here are their names and links
Compound Document by Reference Framework 1.0 (CDRF)
(this is the one that abstracts out generic material that could apply to any compound document)
WICD Core 1.0
(this is the main document, and allows interoperable combination of XHTML and SVG by reference, ie by linking)
WICD Mobile 1.0
(this adds a few mobile-specific items, such a s how to navigate around with a directional pad ona phone, but no mouse or other pointing device)
WICD Full 1.0
(this is a compatible superset of mobile, for example it uses CSS 2.1 rather than CSS Miobile Profile)
You also asked which organizations are involved in the CDF Working Group. Currently they are:
Helsinki University of Technology
Research in Motion (RIM)
There are some expressions of industry support at
Actually, I think it’s pretty easy to see why they think CDF is the answer.
Their requirement for ODF was for it to retain the "dark matter" – the proprietary RTF encodings known only to MS which make the round-trip of documents possible. It’s pretty simple: you write out a basic ODF document which reflects the main structure of the document, and all the bits of RTF you don’t know you stuff into proprietary tags and attributes on this data. The problem with that, obviously, is that any ODF-only app isn’t going to make sense of the proprietary stuff, and can’t use it: it’s not interop, it’s MS Office-only round-trip.
If you only output the basic bits of ODF, then actually HTML is very similar: the rough structure is the same. CDF is then your framework for referring to another XML format alongside the HTML which represents the RTF encodings that their software can’t make sense of. We’ve already seen their ACME software, which is simply RTF encoded as XML.
Again, not interop – it’s more round-trip. Office doesn’t care what format you load/save in, so long as you can convert it to RTF that it understands. That’s all the Foundation are doing.
The problem with that is that it’s not the job that ODF was really designed to do. If they’d said, "we’ve shifted our attention to working on CDF, and we’ll try and persuade folks at W3C to optimize in this direction," then who would care whether they were right or wrong on what people really want, and whether their approach would succeed or fail on the technical front? ODF and PDF, for example, have no quarrel with each other, and both serve a different purpose. And they could have turned the OpenDocument Foundation over to people with an interest in maintaining it – after all, even if you found a non-profit, you never own it.
The real problem I have is with their decision to attack ODF as they turned out the lights. That was immature, unnecessary, inaccurate…I could go on. I seriously doubt that Gary, Sam and Marbux have hit upon something that has a snowball’s chance in hell, and am very disappointed that they didn’t have the political savvy and the detachment of ego to stick with the community instead of launching a doomed and damaging fork. The only lasting effect of their efforts, in my opinion, will be to have injected more confusion and FUD-fodder into the marketplace.
"Again, not interop – it’s more round-trip. Office doesn’t care what format you load/save in, so long as you can convert it to RTF that it understands. That’s all the Foundation are doing."
In other words, this "Foundation" doesn’t comprehend the "Open" part of Open Document Format. Brilliant.
I do believe you mean DOC, not RTF, as RTF is fairly well documented (I’d link to a Google search instead, but I’ve been accused of being a spammer repeatedly) and of which the "scary confusing syntax" (which it has been accused of having) is actually similar to HTML+CSS.
And let’s not confuse RTF for things which Microsoft calls RTF at times, e.g. HTML e-mails, which aren’t RTF at all.
My impression is that:
This "foundation" (i.e: gary "the binary key guy" edwards, hiser ) was very enthusiastic about selling his Da Vinci plugin to Massachussets ( by the way a closed source and propietary project ).
When this project failed, they begun to "throw chairs" against ODF , to the format and its supporters
At this moment, they are making a big favour to MS, contributing to MS FUD machine.
Well, too much writing for me, this people are getting more words than they deserve.
Anyway, thank you Andy ,for this post.
Excuse me anony. Hate to interrupt your rant with the facts. By the time we arrived in Massachusetts, June 19th, 2006, CIO Louis Gutierrez did not have a budget. No secret there. If there was to be an ODF solution, it would have to be delivered at zero cost. We all knew that up front.
If you can handle the truth, here it is. ODF hung by a thread. A year long Massachusetts Pilot Study concerning the implementation of ODF had been completed, and the results were a total disaster. Every major ODF vendor participated in that Pilot Study and were unable to deliver.
There was no plan B, but Louis and his staff came up with this hail mary idea of issuing an RFi, "Request for Information concerning the feasibility of an ODF plug-in for MSOffice. A clone of the MS-OOXML Compatibility Pack plug-in if you will.
When we got there, as one of the few responses to the RFi, ODF truly hung by a thread. No money, no hope of success, a disasterous pilot, and one last shot – a plug-in of such high conversion fidelity that all of the ODF vendors believed it to be impossible to do.
But Louis had a plan. His plan was for us to contribute both our da Vinci plug-in and our InfoSet portable conversion engine as part of an open source ODF Community solution. He’s a very persuasive guy, and we agreed to what was then the unthinkable.
Although the design of the Massachusetts plug-in was complex, and expensive, Louis then persuaded IBM and Oracle to create a "benefactors" group that would sponsor this newly christened open source project. The targeted group included IBM, Oracle, Google, Intel, Novell, Sun and Nokia.
We agreed to contribute da Vinci and InfoSet under the GPL. IBM and Oracle agreed to organize the "benefactors" group. Louis Gutierrez persuaded California and the EU-IDABC to take permanent seats with Massachusetts on the new ODF Foundation Board of Directors, which was charged with managing and distributing the source code. Timothy Vaverchak was appointed project manager. The "benefactors" of course also held permanent BofD seats.
For our part, we delivered in early August the ODF iX version of da Vinci, proving our fidelity claims. We also wrote the plug-in road map of deliverables that became the outline and timeline of Louis’ mid August 2006 Report.
Key to the success of this project was OASIS approval of our ODF iX enhancements; OpenOffice implementation of support of the ODF iX enhancements; and, IBM-Oracle success with the "benefactors" group.
We failed in all three of these last requirements. OASIS never approved our iX enhancements, with April of 2007 marking the final declaration by the OASIS ODF TC that compatibility with existing MS documents and interoperability with existing MSOffice applications would not be part of ODF’s future.
A day after IBM and Oracle notified Louis Gutierrez that they were unable to pull together the "benefactors" group, Louis resigned. It’s been rumored that one member of the group decided not to participate, but would rather go it alone with their own plug-in.
So those are the facts anony. How it is we were going to commercialize the da Vinci plug-in in Massachusetts is a mystery to me. ODF hung by a thread, and many in the community were willing to do whatever it took to save it. You need to explain your accusation.
OBTW, where were you when this was going down? If there’s anything we learned in Massachusetts it’s that if we’re going to defeat Microsoft, it will be in the trenches, with real world solutions that are competitive alternatives to MS-OOXML. Blogging MS to death isn’t going to get the job done my friend. When the call goes out for real world solutions, as it did with the Massachusetts RFi, you’ve got to show up with more than your keyboard and blog.
We showed up. We put forward as an open source contribution something extremely unique. Something we and Massachusetts valued very much. Something an incredible amount of work went into. We put it all on the table because ODF hung by a thread. But where were you? What did you contribute? When the bell rang, and the fight was on with a ruthless monopolist, where were you?
So we didn’t get our iX enhancements through OASIS. So we’ve moved on to CDF. So what. The more salient point is that we are still in the battle with Microsoft. We’re going back into the trenches with CDF, where the battle of real world solutions will determine who wins and who loses. You’re not going to be there. No surprise. Very consistent as a matter of fact.
What we couldn’t do with ODF we think we can accomplish with CDf. Our objective is to neutralize and re purpose MSOffice to produce high fidelity CDF documents that are web ready for Exchange/SharePoint alternatives. What’s wrong with that? Does anyone have a better idea?
At 65% marketshare, the Exchange/SharePoint juggernaut will be difficult to stop. Microsoft has quietly launched a massive migration of MSOffice bound business processes to the Exchange/SharePoint developers hub. The connection between the desktop and server transition is such that both sides of the equation can read/write, with very high fidelity, MS-OOXML.
They’ve got a plan. Do you?
Today there is no alternative to MS-OOXML that can work within the context of the MSOffice business process migration to the E/S Hub. We failed with ODf because ODF was simply not designed for this purpose. So now we’re going to give it go with CDF.
Many people are sneering at this effort, including Microsoft. We think we can do it. But more to the point, at least we are willing to try. What’s your solution? Oh. You would rather sit back, do nothing, and let this juggernaut steamroll everything in sight without even attempting to provide an alternative. Saving your wrath for those who would try to provide an alternative. Any alternative. And the W3C’s CDf is every bit as legitimate an open standard as ODF. The only difference is that CDF was written for the web platform. ODF was not.
What i and my garage challenged partners care most about is NOT arguing that CDF is a better file format for OpenOffice than ODF. What we are totally concerned with is the migration of existing MSOffice business processes to MS-OOXML <> E/S Hub and the MS Stack. Once that happens, there will be no way to unlock those business processes. For the next fifteen years that segment of the market will belong entirely to Microsoft.
What we want to do is cut of that migration and re direct it towards CDF ready Hubs able to compete against Exchange/SharePoint. This is done through the neutralization of MSOffice and re purposing through our CDF conversion process.
If you have a better idea, than let’s hear it. Complaining about CDF without offering up a workable alternative is nonsense. This goes directly to my point. When the bell rings signalling that there is a battle with the great monopolist, you’ve got to show up with more than a keyboard and blog. This battle is going to be won in the trenches of real world solutions. Do you have any?
Didn’t think so,
Backend integration via document formats is all about (in the worse case) data transformation but more hopefully data binding and custom schemas.
You’re right that ODF 1.0 did not have custom schemas, but ODF 1.2 does if that’s what you’re talking about (some specifics would help everyone understand your arguments, unless you just want to classified as FUD).
Backend integration is of course important — what specifically are you talking about?
Mary Jo Foley has left us unsurprised as usual, with the knowledge that M$ spends millions on advertising with ZDnet publications.
I could have told you that; I post in the comments section, and the general trend there is that obvious Microsoft shills get to run around throwing personal insults at people who don’t agree with them, while anybody else gets modded to Hell and back.
Lets give her a break. She is not a technical person, more the journalist type. She is also not alone in this. We can argue that those journalists should had dig deeper before reporting.
She skeptical of Microsoft. I think the problem in this case is she just isn’t familiar with the issues involved.
Uh huh. So she doesn’t do the ten minutes of research it would have taken to learn than the OD Foundation does not represent the entire ODF universe? Feh. Total hackwork, and she should be ashamed.
havemore of an issue with the title than the content. You… do REALIZE that Associated Press, ABC, NBC, CBS, CNN and Reuters are all associated with members of the RIAA and MPAA, and that these media organizations have worked together with Microsoft to remove users control of their (the users) purchased content, software, or hardware? You are aware that the RIAA and MPAA are heavily aligned with Hillary Clinton and Liberal Leftist Democrats to REMOVE users freedoms? You are aware that all of the organizations have effectively formed a cabal for the past 20+ years insuring that there were no such things as unified or open standards?
Here’s a news flash. FoxNews is the ONLY major broadcast player that is remotely balanced. Now, I don’t care if you don’t like that, but if you want to state that Zdnet is imbalanced, you need to lump ZDnet into the "old media" who ARE heavily balanced in the favor of political and economic interests that have NO interest in Open Standards, Open Licenses, or User freedoms.
Um.. that’s kind of ridiculous. You do realize that the RIAA and MPAA is composed of media/music/entertainment companies and conglomerates, right? Thus, they would be aligned with all the major television networks. They tend to be right-wing, whereas the actors and preformers are more liberal. And Microsoft is quite conservative; see how they sleep together with the current administration. What’s your point?
Yes, as Stephen Colbert said, Fox News always presents two sides: the President’s side and the Vice-President’s side.
ODF is a done deal; ISO26300 is there ; and the implementations both no-charge IBM Lotus Symphony and for-a-fee IBM Lotus Notes are growing in quality and diversity all the time. Other vendors, too; Sun, Google, Novell, RedHat, everyone behind ‘KOffice’.
It’s only small-and-medium businesses who are ‘locked in’ to the One Microsoft Way.
Schools are locked in, too, but they pay so little to Microsoft as to make that market segment an unprofitable ‘business’, so it doesn’t count for anything in the quest for profit. You might view it as Microsoft suppressing competition by ‘dumping’.
Large businesses, governments, and globally-integrated corporations all have adequate resources to invest to dig themselves out of any Microsoft-related ‘hole’; certainly IBM will take their dollars and dig them out, if required. And for these market segments, it tends to be a question of ‘control of your own destiny’.
Can you be certain that your business documents, the text of agreements you commit to, your vital records, your public records, will be accessible to you and to those you choose to share them with for as long as you need them to ? When it could be ‘Decades’ ? ‘Centuries’ ? ‘Millennia’ ?
For that, it’s either ‘ISO26300’, or ‘pencil and paper’. Shifting standards don’t meet requirements.
First, thanks for coming by and offering your thoughts – dialogues are always best.
I’m not sure whether you segued from talking to "anony" to me part way through, but since you were talking about "a guy and a blog" I’m thinking maybe you were. If so, here is a response.
As I indicated in my piece, I personally have no problem with anyone pursuing any worthwhile technical strategy. My problem is with the tactics. You, Sam and Marbux have been rather inconsistent in your statements. Often, they are carefully crafted, and do no damage to the ODF campaign. Other times, though, one or the other of you goes into the attack mode, and talks about (for example) how ODF is doomed to fail. Whether or not that eventually proves to be the case, I think we can both agree that it is more likely to increase rather than decrease the likelihood of failure.
The other problem is that whether or not ODF is the best solution (I’ll leave that open for the sake of this comment), it is the one that continues to gather global support. The Foundation’s alternative, and past record, has been of diminishing support. That’s not meant to be a criticism, but only an observation that your tactics have been less successful than the ODF campaign, and – hey – no wonder, with all the resources the ODF campaign has. But the fact remains that as a result, the ODF campaign is way, way, way ahead, and therefore more likely to achieve success in the effort that we’re all committed to, despite differences in details.
If the Foundation is to succeed with its alternative, it has a long way to go. If it can be successful, I wish you every success. But until you get momentum and support behind you, please leave ODF alone.
As to what a guy and a blog can do, I think that there are quite a few people in all camps – government, major vendors, customers, community, who would agree that my blogging, news coverage, and getting the accurate facts to the media have had a positive impact in getting ODF more credibility and more wins.
So in answer to your question, yes, I do have a plan – and that’s to continue doing what I’m doing (which is also what I was doing for quite a while before Louis Gutierrez was even in office). So far, that plan seems to be having more success, so that sounds like the best decision to me. I do wish you the best of luck with CDF as well.
I also read Joe Wilcox, Rob Weir, Bob Sutor, Luis Mettler, and Groklaw.
I’m finding less and less tolerance on Groklaw toward diversity of opinion as many (if not most) of the long-time members are choosing to leave the site due to moderation policys that enforce PJ’s ‘party line’.
Groklaw will retain some residual value for me for a while yet as their ‘off topic’ posts provide good links to other open-source news activities, but I’ve given up on finding any pretense of consistency in PJ’s positions on current events.
On the other hand, I’ve found this blog to be very consistent WRT integrity in reporting and fact-checking. I like your writing style as it simply presents the facts as you know them without a lot of editorializing and ‘I told you so'(s).
The other sites I mentioned above tend to dilute their value with off-topic personal posts of little interest to me, but as a Linux Admin/Engineer/Hobbyist/Hobby-developer I am very interested in Microsoft’s attempts to derail open source and your articles and standards updates are quickly becoming more valuable to me for that purpose than Groklaw.
I read the post you’re responding to with interest. Especially the repeated "ODF is hanging by a thread". Perhaps 2005 was a bit early for a full switch to ODF, but I think not. I do think that the Mass attempt at ODF was both timely and ultimately doomed, but nevertheless a very critical and necessary step. MS and their proprietary lock-in products are deeply entrenched (much like a splinter). The way to work out a splinter is in several small steps. You cannot usually pull it out the first time, but must work it out little-by-little with the first several steps doomed to fail, but also necessary to set up subsequent steps that will ultimately succeed with the continued application of persistance.
I recently read through all the documentation I could find on the daVinci plugin and found (with apologies to G.E.) what I consider to be a critical and non-repairable flaw in their design. That flaw is simply this. They are depending for conversion on the Microsoft embedded conversion process that is neither documented nor public. This means that to convert to MSOOXML, using daVinci, the users must have already purchased MSOffice2007. This fosters (not impedes) takeup of MSOffice and furthermore (unlike the Sun plugin) is subject to being broken at any time Microsoft decides to ‘patch’ their office product. It does not help in translating older office formats in the absense of MSOffice 2007 to funcitionas part of their translation process.
I therefore, was surprised at my conclusion that the daVinci plugin – for all the good press and promotion it got was never a really viable product due to the dependencies on MS internal, undocumented API calls. Furthermore, the ODFix format extensions being proposed sounded like (since I never was able to successfully wade through all the content-free verbiage and actually find bullet points containing their proposed extensions) a series not of ‘proposed changes’, but ‘revised design’ to ODF. The upshot of these design changes appeard to be to essentially disregard the well-established guidelines and intent of XML as well as the existing structure of ODF and the add non-XML structures required to handle documents as streams instead of hierarchical trees.
When I read this on their site, I fully realized why OASIS chose to vote these changes down. When I read later that these were the SAME changes being asked for by Novell to ‘integrate’ their version of OOO with MSOffice, I got the distinct impression that the ODFF had (in secret) brokered a deal with MS to subvert any attempts on the part of customers to exit the MS document lockin trap. This impression was further strengthened by G.E.’s frequent assurances that daVinci would have 100% fidelity with MS conversions rather than the 85% that the OpenOffice project is claimed to have (also by G.E.). This statement not only contradicts the historical evidence that MS upgrades have historicaly been known for "messing up" conversion of older documents, but also for chaning displays based on which printer-driver is loaded. Open Office documents, on the other hand seem to be independent of printer settings when converting using their external conversion routines that do not depend on the good will of Microsoft.
It saddens me to see the ODFF’s recent attacks on ODF and the fact that they are giving up on ODF already. As I indicated above, the Mass attempt to escape the MS trap raised awareness of document retention problems as well as an awareness of the responsibilities of the state to continue to make documents available over time periods that exceed 1 Microsoft development cycle. It started many things into motion: it provoked the public interest of other states & the federal government in ODF, sparking an entire movement in Europe & 3rd-world countries in ODF that has *not* been forced underground by MS, it forced Microsoft’s hand to create the MSOOXML specification in which many weaknesses and inaccuracies have been found and published throughout the MSOffice suite and in which it has become obvious that MS has no real plans for sustainable document retention requirements. It sparked the entire sad story of the MS attack on the ISO. The Mass inquiries sparked the open-source world to recognize the needs of the disabled in a public way, resulting in a host of assistance technologies being integrated into the OS and the desktop API(s) instead of being left as 3rd-party addons that may or may not work with a given product and that are not natively supported by either the OS or the desktop API(s).
Granted, the use of ODF in Mass was defeated, but that was on political grounds – not technical ones. G.E. may not agree that there is a double standard in requiring ODF document applications to natively support the disabled, while giving MS a pass on support for disabled as long as taxpayers can pay through the nose to buy third-party products that do a second-rate job.
As to the much-vaunted ‘compatability with legacy documents", that is so much hog-wash. First, legacy retained documents in an ODF world only need a one-way transition – to ODF format – in exactly the same way that legacy documents would be converted (one-way) to MSOOXML. Second, It makes no sense to convert new documents to old, non-supported formats for the purpose of long-term-retention.
Third, current working documents can already be shared between OpenOffice and MSOffice (multiple versions) by simply saving in Work97/Excel97/Powerpoint(.ppt) formats in both OpenOffice and MSOffice2007. It’s not practically realistic to *assume* that an organization as large as the state of MASS will convert all their desktops to MSO2k7 overnight given their requirements for supporting their disabled employees and their requirements to exchange data with other agencies, contractors, and the general public – most of which will *NOT* have upgrade to MSO2k7.
Fourth, it is not the requirement of a document format to be ‘compatable’ with older documents as long as it can define and store sufficient elements that an APPLICATION can convert, store, retrieve, and render any legacy MS document without loss of fidelity. Again, I maintain that there is some leeway here in the ‘fidelity %’ that G.E is so fond of spouting since MS has such a long history of failing to properly convert documents between one version of Office and the next.
Bear in mind that there is a lot of FUD flying around as well:
"ODF does not support or document spreadsheet formulas"
— Given what we’ve seen of the MSOOXML spec, it does not matter that MS does or does not document their formulas. Their formulas are inaccurate and simply produce incorrect mathematical results. (2007 and 65536 bug, wrong results for ceiling functions, statistics, date functions due to the 1900 leap year implementation, etc). Therefore, no application based on MS Spreadsheets should be trusted. Documentation of their formulas notwithstanding. It’s much better for the IT users if they can have confidence in the results of a calculation without resorting to ‘stubby pencil’ validation of each new set of data plugged into a Microsoft Excel spreadsheet.
"ODF is a Sun Project and Microsoft was never invited"
— It may have started that way on day one, but there was nothing preventing MS from requesting participation or accepting invitations to joing at a later date. The simple fact is that MS chose not to participate in the OASIS process due to the fear that doing so may water down their later arguments for not complying with ODF – a standard that weakens their document lock-in hold on their customers.
"ODF & MSOOXML can be merged"
— As evidenced by the daVinci failure with ODF, the formats are too different in architecture and fundamental concepts to merge. While ODF defines a comprehensive document base format with the ability to add additional application-specific tags in an application-specific subtree, the MS specification defines a specific version of a document storage mechanism and is not able to expand beyond the single vendor’s implementation due to the lack of any mechanism for application or vendor-specific settings. Furthermore, the MS implementation uses several non-XML elements and techniques such as bitmasks and stream-oriented mechanisms that simplify the way that MS applications read/write the format, but that produce XML that violates both the design philosophies of XML and the ability of XML parsers to work with Microsoft’s version of XML.
To me, it is becoming more and more obvious that the only solution to the above problems is to:
A: convert all documents to ODF and immediately start using ODF only.
B: continue to work on harmonizing spreadsheet formula storage standards (OpenFormula).
C: continue to work on support for assistive technologies in office applications for all platforms and in all non-office aspects of Linux.
D: continue to hold Microsoft to the same technical standards that the open source community holds themselves. Publicize any and all instances of Microsoft attempting to improperly interfere with open source adoption as Microsoft surely would if the situation were reversed. (This is where blogs come in).
E: do not depend on Microsoft’s good will for the proper and continued operation of ANY open-source application, project or plugin. HIstory is littered with the carcasses of companies and the broken careers of people that did so.
F: do not accept any ‘gift’ from Microsoft. It will most likely contain unacceptable ‘strings’ attached. I know of no MS gift to date that has not had such strings.
That’s enough for one rant – thank you for listening. Andy has permission to use any of my comments for any purpose he sees fit.
Thanks for sharing all of these thoughts. There’s more here than I can respond to in detail, but I read it with interest.
I can vouch, from personal knowledge, for the fact that G.E.’s rendition of facts and events in Massachusetts is selective and not entirely accurate. For example, he skips in his rendition above from April to saying "and so Louis Gutierrez resigned" – when in fact it was October before that happened.
On Groklaw: I don’t read the comments there much, primarily because I simply don’t have time, and because the events that PJ and I cover only overlap in the standard area. That said, I think that PJ continues to play a huge role in keeping all kinds of people honest, and I’d hate to ever see her give up in that quest. She has a huge audience, is tireless in what she does, and is hugely savvy. Even if I don’t agree with every conclusion she reaches, I continue to have great respect for her dedication and integrity.
Regarding the idea of "merging" the formats (or as IBM call it "harmonizing") there is no technical reason why a future ODF can’t have OOXML extensions.
That "da-vinci" failed is not evidence that OOXML and ODF can’t be merged and I encourage anyone to look at the detail of the arguments to see that they’re false (hint: a mixed content model doesn’t affect merging in the slightest, it’s merely a different representation of the same content, and yet that’s the most popular argument)
I will admit to being very, very confused by all of this ODF vs ODFF posturing. I will try to put my current thoughts in short form, but it will be a muddled mess. I warned you!
From what I gather, the OpenDocument Foundation (ODFF) is attempting to create more of an interop format for working against a background MS server stack (Exchange/Sharepoint). They worry that MS is further cementing their business lock-in by moving more and more companies into dependency on not only the client -side software but also the MS business stack that has finally evolved into a serious competitive set. At that level, and in their view, the "atomic unit" is the whole document. The encoded content is not of immediate concern.
ODF is concerned with the actual document content, which ODFF is prepared to ignore. The "atomic unit" is the bits and parts in the document. They want to break the proprietary encodings that MS has that lock people into MSOffice. The stack is not of any immediate concern.
So, unless I misunderstand either camp, ODF is first attacking the client end of the stack, and ODFF is attacking the backbone server end of the stack. The former wants to break the MSOffice monopoly by allowing people to escape those proprietary encodings, and the latter wants to prevent the dependency on server software like Exchange and Sharepoint by allowing MS documents to travel to other destinations than MS "server" products.
Is this correct? I have yet to see anyone summarize the differences in any non-partisan way. The usual diatribe by people closer to the action is to go into the history of ODF or ODFF, talk about old slights and lost fights, and somehow try to pull at emotional heartstrings so as to gain mindshare. This is childish on both sides.
Furthermore, the world "orthogonal" comes to mind. I often see people too busy arguing their POV, and not listening to others, when there is no real argument to keep making. It’s apple-and-oranges. ODF vs ODFF seems like they are caught in this trap. Everyone wants to win an argument that has no possible win because the participants are not arguing about the same thing.
Why can’t the parties get along? I can see a cooperative that attacks the entire stack. Am I the only one?
Good questions and good message. I’ll leave it for others to rate how well you’ve summarized the technical side of things, but I can address the overall sense of what you’ve asked on the dynamics.
As I’ve said a number of times, including in my responses to comments in this thread, I have no problem at all with people finding value in CDF, if there’s value to be found there. If Gary and company are on to something, more power to them, and good luck.
What I am also saying, though, is that it’s unnecessary and harmful to trash ODF in order to promote CDF. At minimum you’re right in saying that there are two different strategies here and two different technical approaches.
What Gary, Sam and Marbux need to do first of all is to convince others that they’re on to something – if they can’t do that, tearing down ODF isn’t going to help them.