After all the mud slinging (disclosure: I did some of the throwing) in ODF vs OOXML in the period leading up to OOXML accepted as ISO Standard, things dies down as expected. I was hoping the argument will now shift to the technical merits of the two competing standards, but no. We have people still trying to make politics the front page.
First up, Rick Jelliffe. He attacked ODF openness by saying that people attending ODF meetings are 90% vendors. I do not think that is unusual. In fact, I am not surprised it is not higher. Standards are only useful for people who need it and joe public, who makes up the vast majority of users, could not be bothered. While slinging mud at ODF committee, I need to ask the question why Jelliffe had not attended the meeting himself? May be he is like me, i.e. cannot find someone to pay for the trip and the trip is too expensive to fork out from my own pocket.
An expose by Rick Jelliffe Mr Brown? I don’t really think so. ODF controlled by a “single big vendor”? Check out WG4 first. Every committee will benefit from being well-rounded, but virtually all committees are not. WG4 getting there? Me not that sure mate.
What I will agree with Mr Brown is that there should be scope for private conversation at Standard committee. Going all public like ODF means there are participants who want to discuss things in private simply bypass the committee by direct mailing. But I think OASIS knows this. It says everything has to be public to encourage transparency and encourage participants to open up.
As for WG5 (interoperability), just because you set it up does not mean ODF and OOXML camp will meet and participate. Quite frankly I do not see any need for such a committee as we all should just switch to ODF. OOXML is not open enough in terms of available implementations. We have implementations in Windows platform, but for others, the implementation is simply not good enough, i.e. read but no write. WG5 benefits OOXML by allows OOXML to “overcome” this problem by saying user simply convert to ODF as need be. This is to the detriment of ODF so why should ODF surrender the advantage of cross-platform implementation by participating in WG5?
From the ODF camp, I think this criticism of OOXML is a bit over the top. Mistakes happens. Moreover, every standard has defects so having problems with font specification is not something really unexpected.
CyberTech Rambler 
Please read the “over the top” criticism with attention, and also check the mentioned documents (some of them are already available on the Internet).
I didn’t wrote that post because I’m an ODF person, or a member of ODF TC, but as a Brazilian that spend more than a year working on OpenXML during its ISO evaluation. I was also at the BRM in Geneva representing my country and I saw things on that room that you cannot imagine (please search my blog and you’ll find a lot of good stuff there).
I agree that mistake happens, but do you really believe that a group of people could evaluate, debate and fix 1200+ technical issues on 5 days (during the BRM) ? (detail: those 1200+ issues was a summary of more than 3000+ presented, so we’re not talking about a single problem, but about thousands of problems…). The mistake here was the approval of an standard that wasn’t ready. Brazil, among several other things, official proposed that the OOXML should be presented at ISO as a normal project (and I would work to improve it on that basis, but no one at ISO listened us).
You may also check Alex Brown’s blog (http://www.adjb.net/post/SC-34-Meetings-Prague-Days-2-3-4.aspx), a guy that really supports OpenXML and had a major role on its approval at ISO (he was the BRM convenor), and he is telling that some decisions taken at the BRM will be simply ignored (and most countries changed their votes from NO to YES because somehow they believed that the BRM resolutions would fix the specification).
For me, this isn’t about ODF x OOXML, but about RESPECT (according to Microsoft, ODF clearly won: http://www.thestandard.com/news/2008/06/19/red-hat-summit-panel-who-won-ooxml-battle).
Best,
Jomar
Comment by Jomar Silva — April 11, 2009 @ 3:19 am |
Mistakes happens?
This looks pretty serious to me.
Quote from http://homembit.com/2009/04/for-the-skeptical-the-final-proof-the-openxml-wasnt-and-isnt-ready.html:
The document N1101/N1168 contains for example, several items in which they recognize that there are decisions made in the BRM (BRM resolutions) which were not incorporated into the final published text of the standard. In other words, even taking almost a year after the aproval of the standard to publish the text (yes, approved without reading), there wasn’t time/attention or anything else necessary to assure that the changes were published in the text (most of those changes, “conditioned” the approval).
End Quote.
In particular the last words above, ‘most of those changes, “conditioned” the approval.’
I don’t think this is usually covered by “Mistakes happen”.
Stephan
Comment by Stephan Wehner — April 12, 2009 @ 12:49 am |
@Jomar
“I agree that mistake happens, but do you really believe that a group of people could evaluate, debate and fix 1200+ technical issues on 5 days (during the BRM) ?”
No.
It also means the number of mistakes increases.
My definition of “mistakes” is very loose, i.e., anything agreed upon but does not make it into the standard. I will say everything that is agreed at the BRM should make it into the standard, unless one can demonstrate that it is wrong.
Incorporating 5000+ fixes into the standard document even at a “Standard” time frame is impossible, so acheiving through a fast track process will take a miracle. We know it, they know it, but they still push it through. That is wrong of them.
Alex Brown is definitely pro-OOXML, after the BRM his biasness certainly show through. However he is contributing to the next defnition of ODF by pointing out what he believe to be defects in the standard on the discussion list (office-comment_AT_lists.oasis-open.org)
“he is telling that some decisions taken at the BRM will be simply ignored (and most countries changed their votes from NO to YES because somehow they believed that the BRM resolutions would fix the specification).”
I am sure if (1)something is obviously a mistake and (2)the issue is at a stalemate (so someone have to break it) then decision at the BRM can and should be overturned.
I am sure a lot of countries voted with belief that the BRM resolutions would fix the specification (for the better) and they rightly expect the convener to carry it out. Now that they are finding that their resolutions are not carried out, they can always table it again if they feel strongly about it. Of course we know they won’t, coz the money from certain big vendor to attend meetings suddenly disappeared. 😉
Do you really think that when Microsoft says ODF clearly won they meant it?
Best regards.
Comment by ctrambler — April 12, 2009 @ 3:37 pm |
@Stephen
I will agree I am giving OOXML a lot of leeway when defining the term “Mistakes”.
The whole OOXML thing is a collection of mistakes:
(1)Mistakes in the time frame taken to approve it
(2)Mistakes by National Body to trust that BRM has influence
(3)Mistake by BRM for not attending to every concerns of national bodies
(4)Mistakes for not incorporating BRM resolutions in the published standard, even after one year. If you ask me the task (to evaluate and incorporate 1200+, I will say 5000+ changes) is impossible in a year.
Hmm…. so many mistakes, so is it really surprise that what I said fall into the “usual” mistakes category in the context of OOXML? 😉
Best regards.
Comment by ctrambler — April 12, 2009 @ 3:42 pm |
[…] their spots? Filed under: Uncategorized — ctrambler @ 4:34 pm After being prompt by Jomar Silva (someone I do respect and is surprised he took the time to comment on my blog) to re-read Alex […]
Pingback by Leopards finally revealing their spots? « CyberTech Rambler — April 12, 2009 @ 4:35 pm |
Hi. My comments were about the voting members present at ODF TC meetings, which are conducted by telephone (there is no ‘trip’), and how broader participation would address a problem of concentration of sectional interests. I think standards bodies need to reflect the market/bazaar of stakeholders: standards bodies with only suppliers (vendors) or only demanders (adopters/users) are unbalanced.
My blog articles identifying a dominance over the last few years by one group was not an “attack” or “slinging mud”, but definitely a call for more balance: more participation by non-vendors primarily, self-discipline by vendors secondarily. The only people who could complain about this would be people who think that domination by big business, cartels and sectional interests is a good thing for an important Open Standard. (Remember that ODF needs to meet a much higher bar than OOXML here, because OOXML is fundamentally intended to document a format for a pre-existing technology and feature set of recent proprietary systems. The more that standards like ODF get regulated and mandated, the more that their committee processes need scrutiny; indeed, it is clear that anti-trust regulators are interested in participation and processes for open standards.)
(Indeed, even though we disagree on many or even most issues, I think Jomar has been doing exactly the right thing by participating on the ODF TC: indeed, one of my proposals to the co-chair was to waive the membership fee for more people from the ODF Alliance or from, in President Lula’s phrase, non-“blue eyed” countries. The point of having broader participation is not to have more people who think like me, but to have more people who bring their own thoughts & priorities.)
There is no need to see my call for broader participation at the ODF TC as part of some ODF versus OOXML or even OASIS versus SC34 rivalry (the “next round” for goodness sake). All the standards groups need scrutiny, and they all need procedures to stop domination, and most of all, they all need broad participation. Join the group handling the standard of your interest or occupation. Join OASIS. Join W3C. Join ECMA. Participate at IETF. Join your national standards body and make comments on ISO standards. If you are concerned that committees are unbalanced, participate. (I participate in SC34 WG1 as best I can, and I also submit to the ODF comments lists sometimes.)
I agree with your final paragraph. Mistakes always happen. The BRM decision to adopt XML Schemas boolean types even for attributes that used “yes” or “no” etc was, in my view, a bad mistake. But there were delegations who did not consider than compatibility with the existing ECMA standard (i.e. what Office 7 implements, pretty much) was any kind of a goal: they wanted improvements that were not just clarifications or additions or profiles (subsets) but even incompatible replacements.
On Jomar’s comments: IS29500 now is being worked on as a normal ISO standard. It has a normal committee, WG4. Like any standard, any parts of it may be changed depending on the case being made in the committee and, ultimately, by the national votes. It would then have a new, dated version. So the inconsistent decisions of the BRM, such as the boolean case, are certainly up for grabs. However, SC34 is traditionally very reluctant to make changes to standards on parts of the technology that were consistent and widely supported: in fact, XML was born out of this reluctance (in that case, to revise SGML).
The OOXML defect report log from last month had about 165 defects logged, post BRM. And others are coming in: I submitted five or six yesterday, based on projects for a client’s (not MS) conversion project. It is still a mystery to me why people who do not need to use OOXML for document conversion have much of an opinion on IS29500: even IBM’s Bob Sutor could say “OOXML is about the past, and ODF is about the future.”
Ultimately the group that makes sure that the ISO standard reflects the information that the National Bodies wish to see in it are the National Bodies. They have to keep on participating, and not give responsibility to others. If the ITTF has gotten it wrong in implementing a BRM requirement, then this needs to be brought up at WG4 (as, indeed, it has been on a couple of issues.) The idea that the BRM is the last and only chance to fix problems was never correct. It may take years for IS29500 to have a really good debugged version, just as it is taking years for ODF to have a good, complete debugged version: that is just the nature of big standards. Freaking out about it too much is a waste of energy better spent improving the standards or the processes.
Comment by Rick Jelliffe — April 15, 2009 @ 12:11 pm |
Mr Jelliffe (you’re making today a busy day for me, hmm ?),
First, thank you for the comment about my participation at the ODF TC and I really think that more people would do the same.
I insist that OpenXML wasn’t ready to be approved as an IS, and I know you also know all the details behind OOXML development inside Microsoft (don’t you ?).
Please take a look at the recent SC34 meetings audience and you’ll have a more complete sense of what ‘unbalanced’ really means, and the sad about it is that the countries names are being used to ‘undercover’ real wishes (btw this is something that I really believe is shameful).
Unfortunately your dear friends from Microsoft are now trying to convince several governments that OpenXML, approved at ISO, is a 100% robust and solid specification, despite if you think that “…may take years for IS29500 to have a really good debugged version…”.
Honestly I only regret that the commercial needs of a company force the whole world to spend a lot of effort on an initiative like OpenXML, instead of unite the efforts developing something that is more useful, and I also think that the current financial crisis is showing the final results of redundancy.
About the BRM, the problem isn’t what the BRM was, but how the BRM was used.
First of all, read the JTC1 directives and you’ll see that the OpenXML BRM is completely unexplained (3000 technical issues on 5 days…). Before the BRM, OpenXML was REJECTED and after the BRM, your friends from Redmond had used the BRM results to convince people that everything was solved (please refresh your memory here: http://homembit.com/2008/03/finally-the-details-about-the-final-results-of-the-brm.html). Then OpenXML was “magically” approved.
After all this discussion about mistakes, I decided to prepare a document with all mistakes found on OpenXML process and decisions at JTC1, because I think that people need to know them all (and don’t worry, I won’t tell the details about people’s behavior and “tea breaks proposals” at the BRM… not yet). Most of those mistakes was presented to ISO TMB as the Brazil’s appeal, but without having “blue eyes” they simply ignored us (same to Venezuela, South Africa and India).
Comment by Jomar Silva — April 15, 2009 @ 5:53 pm |
@Jelliffe
I will go for a second round of ODF vs OOXML provided it is concentrated on technical issue, not politics.
In my opinion, neither OOXML nor ODF committee has the correct balance of vendors/users. ODF committee is too vendor heavy, and OOXML is too user heavy. In fact, OOXML minutes for the last meeting (http://www.itscj.ipsj.or.jp/sc34/open/1199.pdf) so user-heavy, i.e. only one office application vendor represented that one wonders why is it a ISO committee (or ECMA for that matter). It looks so much like a box-standard Microsoft feedback committee that it should be chaired by Microsoft, hosted by Microsoft and paid for by Microsoft.
What makes me sad about OOXML meeting is the main backers, i.e., Apple, Oracle and British Library did not even bothered to turn up. I hope they don’t make a habit of it. I am prepare to name-and-shame them if that is what it takes to get them off their chairs.
Between vendor-heavy or user-heavy, I choose vendor-heavy. It is after all, a office document format designed for office application. Linking with other systems is important, but it is not the ultimate aim.
I don’t think your view that “fundamentally intended to document a format for a pre-existing technology and feature set of recent proprietary systems” is shared by Microsoft. That view implies that the purpose of OOXML is to document anything Microsoft Office 2007 and below and nothing else. If so, no new features to support new functions of Microsoft Office 2009 and beyond should be accepted, as it is neither pre-existing or is a recent system, and, Alex Brown would not be blogging about a potential addition to the standard (http://www.adjb.net/post/SC-34-Meetings-Prague-Days-2-3-4.aspx) Certainly I believe like Apple. Oracle and British Library has that in mind, but not Microsoft.
Comment by ctrambler — April 15, 2009 @ 11:22 pm |
@Jomar
I share a lot of your views. I am sure he is aware of our view on the 3000+ technical issues and how the BRM should be conducted. I think you and I will just have to accept that Jelliffe has a different view.
I believe we should not let them forget about our view and point out every mistakes of theirs, and you are certainly one of the better contributor to the ODF side of the debate. Like Rob Weir and unlike me, you based all your points on facts that can be verified. I thank you and admire you for all your hardwork. Unfortunately there is the dreadded “BUT” here. I think it is time to reduce the word counts on the conduct of BRM and things before the BRM. Reducing the word counts will allow us to pay more attention to other issues, such as getting ODF approved and used by governments instead of OOXML. Note that I am not saying we should move on. I believe we should bring it up when it is warranted, such as in situation when we are attacked.
Comment by ctrambler — April 15, 2009 @ 11:23 pm |
@ctrambler “Between vendor-heavy or user-heavy, I choose vendor-heavy. It is after all, a office document format designed for office application. Linking with other systems is important, but it is not the ultimate aim.”
That statement bespeaks lack of familiarity with what an IT standard *IS.* But it is a lack of familiarity shared by all too many who work on IT standards. Standards are about uniformity, not variability.
An international standard must by law specify [i] all characteristics [ii] of an identifiable product or group of products [iii] only in mandatory “must” or “must not” terms. WTDS 135 EC – Asbestos, (World Trade Organization Appellate Body; 12 March 2001; HTML version), para. 66-70, http://www.wto.org/english/tratop_e/dispu_e/cases_e/ds135_e.htm
And IT standards in particular must “clearly and unambiguously specify all conformity requirements that are essential to achieve the interoperability.” ISO/IEC JTC 1 Directives, (5th Ed., v. 3.0, 5 April 2007) pg. 145, http://www.jtc1sc34.org/repository/0856rev.pdf
Absent such specifications, a standard is a standard in name only. A standard is intended to establish a market in standardized goods, creating economic efficiency and competition. This is perhaps most simply illustrated with weights and measures, where a pound of flour must weigh the same regardless which vendor sells the product. But we can also see it in the interoperability context, e.g., with standardized nuts, bolts, and wrenches.
Absent sufficient specificity to enable and require interoperability, ODF and OOXML create technical barriers to trade rather than promoting competition. And the Agreement on Technical Barriers to Trade unambiguously requires that national standardization bodies “shall ensure that technical regulations [includes international standards] are not prepared, adopted or applied with a view to or with the effect of creating unnecessary obstacles to international trade.” http://www.wto.org/english/docs_e/legal_e/17-tbt_e.htm#articleII
So while I agree that linking IT systems may not invariably be the ultimate goal, sufficient specificity in an IT standard to do so is in fact a threshold user and legal requirement. Otherwise, one has vendor lock-in and definition of the standard is controlled by the vendor with the largest market share, not the standard itself.
Neither ODF nor OOXML met than threshold for eligibility as international standards and still do not. In both cases, national standardization bodies voted to adopt the standards without paying heed to fundamental legal and user requirements.
Comment by Paul E. Merrell, J.D. (Marbux) — May 13, 2009 @ 3:31 am |