"And in this corner ..." - can you hear Michael Buffer introducing ODF vs. OOXML in the boxing ring?

May 6 2009

Last week, Microsoft released 2007 Microsoft Office System Service Pack 2. And boy are the sharks swimming in the water.

Before I go on, I want to state where I stand - I really do not care about what games anyone wants to play when it comes to interoperability. Period. I have always been the guy who played in both sandboxes - Lotus and Microsoft. It started out as being the guy who figured out how to make Ami Pro drive Excel via DDE. Then it was Word Pro and Access. Office and Notes. Visual Studio.NET and Domino Web Services. SharePoint and Lotus Symphony. VSTO and XPages (coming to a user conference this August). I love being someone who isn't pigeon holed into one technology or camp. So as application automation morphs into document generation in the next couple of years - I want to work with both OOXML and ODF. PSC is already working with OOXML and ODF with our customers. In the end, I do not care which 'wins'

Since the release of SP2, both IBM and Adobe have come out and been very critical of the release. Adobe is taking aim at the PDF tools that are now included 'in the box' ... which you could download from the Microsoft web site for the past two years. Adobe's complete Acrobat toolset has add-in's for creating PDF within Office - which is a very good tool. Gray Knowlton covers a bunch of the details. Read for yourself and decide what you think.

With ODF, Rob Weir of IBM took Microsoft to task with issues that have come up with formulas in spreadsheets. Doug Mahugh answers back. After being poked by Nathan with his tests, I did some of my own and have found that both scenarios, the one that Rob and the one that Doug lay out, are repeatable.

I am not going to get into the middle of either discussion. I just want to point out something Doug has done - Microsoft

posted their guiding principles and their implementer notes. Microsoft invited folks to Redmond to discuss both ODF and OOXML - I know this personally because I was there. IBM decided not to show up even when invited. I do not remember if Adobe was in the room.

So yes, I think Microsoft is creating some issues with interoperability when it comes to a document standard. I also think that OpenOffice.org has bugs in it's ODF implementation - and everyone has decided to copy the bugs to make things work. Should Microsoft have done that - perpetuate known bugs just so files open? I know people who are on both sides of that discussion. I am not sure which side I am on - but I do know who is the loser.

There is a quote that I am reminded of here (not sure who said it) "It doesn't matter who wins the war, it is the average person that always loses" - and right now, the only thing for certain is that the average user is losing. Microsoft nor IBM is not winning here. Neither is ODF vs. OOXML. This is about a whole slew of folks who are not communicating. I wish both IBM and Microsoft would get in a room, ignore the freaking standards board bull shit and solve the problem. Yes, at the end of the day, it needs to come down to standards so someone has a blueprint to implement to. But right now, the lack of communication and professionalism is only going to hurt the end user. To everyone involved - please stop playing 'the game' and solve the problem.

It will be interesting to see what happens with ODF 1.2 becomes a ratified and Office 14 / 2010, OpenOffice.org 3.X, and Lotus Symphony 2.0 fully support it. When Open Formula is a real standard people can support. When the user can stop worrying about the format the file is in - all of the tools open the document no matter the format. Yes, that is my head in the clouds dreaming of a document format utopia, but hey, I can have my dreams :-) Until then, I will support both formats, both companies, three pieces of software, and build a document generation business that supports them all.

3 Responses to “"And in this corner ..." - can you hear Michael Buffer introducing ODF vs. OOXML in the boxing ring?”

  1. 1) Rob Weir says:

    1) Sure, Microsoft posted their interoperability principles. But did they follow them in SP2? Is stripping formulas really what you would call "Preserving Users Intent"? I've been an engineer long enough to know that the best way to figure out a user's intent is to ask them. Silently losing data is not a reasonable solution. How is this "Preserving Editability" when you can not edit the spreadsheet without having it return incorrect answers because your formulas are now missing? How does this meet the "Adhere to the ODF 1.1 Standard" when the output from Excel 2007 SP2 does not follow the required cell addressing formats? Sure, having principles is great. But following them is even greater.

    2) You mention that OO has bugs and that that Microsoft should not need to copy them. This is true, but totally irrelevant. Excel 2007 SP2's inability to read ODF spreadsheet formulas from other vendors has absolutely nothing to do with bugs in other applications. The spreadsheets that Excel 2007 SP2 cannot read conform to the ODF standard. The documents that Excel 2007 SP2 writes out do not conform to the ODF standard. The bug is on Excel's side.

    3) You say that there are a "whole slew of folks who are not communicating". But the tests, which evidently you have replicated, show that all of the latest ODF spreadsheet implementations seem to work pretty well with each other, except for Microsoft. So who is not communicating? Everyone else? Or Microsoft? This is not hard to figure out, especially when the Microsft/Clever Age Add-in for Office, which Microsoft funded and provided architectural support for, got this right.

    4) As for OpenFormula, it doesn't cut it to say support will come after it is a standard. That isn't the way it works in OASIS or the W3C or in many other standards bodies. We're required to have implementations before we can even put ODF 1.2 and OpenFormula up for vote as an OASIS Standard. Those are the rules and it is a good thing, IMHO, to have demonstrated implementations before a standard is approved. So the argument that Microsoft cannot implement it because the standard is not approved yet is hooey. Not only can they do this, as an member of the ODF OASIS TC they are especially encouraged to so.

  2. 2) Doug Mahugh says:

    Rob, I think John has some great advice here, and we should both do what we can to follow it. In that spirit, a few thoughts on your comments ...

    As I’ve said many times on my blog and in person, our #1 guiding principle for our ODF implementation was conformance to the ODF 1.1 standard. I think it’s reasonable to assume that you would be very disappointed if we didn’t make that a top priority, right? Preserving user intent was also a goal, of course, and in the case of something as complex as formulas, that’s nearly impossible to do without complete documentation of the syntax and semantics of those formulas.

    You mention the concept of ODF TC members being encouraged to implement standards before they are approved. Is that an OASIS policy, or your personal opinion? If there has been communication on that concept, I missed it.

    Regarding your comment about the ODF 1.1 standard's requirements for cell addressing formats, I’ll look into that. (My day job involves many things other than blogging, and I only have a minute at the moment.) Are you talking about the markup that we write in our own namespace for formulas? Or something else? I'm glad to provide a specific answer to a specific question on this matter.

    As to your point about the need for standards to be preceded by implementations, that’s a very surprising perspective coming from you. During the DIS29500 process, you and I often debated this concept, and if I understood you correctly you were then saying that the existence of an implementation made ECMA-376 not a true open standard. Do you see that differently now?

    We'd love to work with IBM on interop issues, Rob. We've invited you to come see what we're working on and give us your feedback. You're welcome at any and all DII events, as everyone is. Meanwhile, from our end, we're active in maintenance of both of the XML-based document format standards we support, and we're eager to participate any time there's a public or private discussion of interop issues.

    Yes, I didn’t make it to the Beijing event, but that was before SP2 was out, so we weren’t even shipping an ODF implementation at the time. I felt that it wouldn’t be appropriate to attend and talk about our implementation if the attendees couldn’t actually see it for themselves. That has changed, and I look forward to participating in many such events going forward. Please keep me on the invitation list for events you're coordinating, and I'll do the same with you for events that we're coordinating.

  3. 3) Arioch says:

    Hmm... interoperability between Lotus and Office... Someone is working on it they say...

    Ok, let take simple office configuration:

    Win XP 32-bit (russian-localised, where it counts) (same in Vista and Win7 though)

    legacy Lotus Notes - version like 6.5 or 7.02

    eclipse Lotus Notes - 8.0

    MS Excel 2003 with latest updates

    OpenOffice.org calc 3.1.1

    The very simple thing now: i open some table in Calc or Excel. I want to mail crop of it to my co-worker, so i open the table (in OOo or Excel) select needed cells and press Ctrl+Ins, then i go to Lotus, select New Memo and press Shift+Ins (Paste).

    And what we have now ? we have a picture, inserted into the mail! when my co-worker would receive the mail - he would never be able to edit this crop or to copy somethign from it - it is just picture !

    Ok, really, i can go Edit|Paste Special and select RTF or HTML format, so hat Lotus would not trash the table data... And it is Windows basics in a way (though only few users knows it). Yet for all heavens, what do Lotus thinks about, when it pastes DATA-less picture instead of DATA-full HTML or RTF ?

    And such a images, instead of data from tables, that i receive regularly from co-workers is not a great thing about interoperability.

    One more example, though i am not sure you'd be able to reproduce it - this times you need russian-language in Windows/Excel. Open Excel, make some table with russian words in it, save as XLS. Open that XLS in Lotus 7 - all russian words are thrashed to be unreadable completely - Lotus just looses information about locale and fails to guess it from Windows settings. That is a known issue in OOo - but since i talk of Lotus 7 we cannot blame OOo/Symphony core there.

    Above exampels are simplistic ones, i can hardly believe no one tried them, if they really tested for interoperability!

Leave a Reply