What to do now

by Rick Jelliffe

Now that ODF and OOXML are both set to be on the ISO/IEC books, it is useful to consider what the next productive steps are.

For genuine ODF Supporters who are concerned that ODF has languished a little out of the limelight during 2007, there are a lot of useful things to be done. You don't even need to join the OASIS groups or your local National Body or SC34 to begin.

I suggest here are some things that will help the ODF effort coming into ODF 1.2.

  • Lobby the component standards groups, notably the W3C, to have official RELAX NG schemas available. Without schemas, there is no validation, and without validation there is no conformance testing, and without conformance testing there is no interoperability. (Or, at least, it becomes significantly more difficult in each case.) I believe SMIL is an example of this. If possible, actually have the schema ready and waiting, to make it easy: you will feel more of an achievement to have part of the standard that you can say "I contributed that"!

  • In a similar vein, lobby the component standards groups to harmonize their standards with ODF. SVG is the one in particular that seems needed. It would be great if not only would W3C SVG group add the few missing attributes and so on, but perhaps also make a profile of SVG to match ODF better (this is not a concrete suggestion, just something whose usefullness could be checked up by someone wanting to get involved.

  • Speaking of SVG, some open source XSLT transforms for going from ODF's "SVG" to standard SVG would be good.

  • Join in the KOffice and Open Office efforts, especially in areas that effect you or for which you have expertise. Maths is a good area, for example.

  • Check through the IS29500 spec that are of interest, when it comes out, and figure out whether they are things that are decorations (which can be handled merely by foreign elements in ODF) with the current ODF behaviour an adequate fallback, or features that are currently unsupported in ODF, that will need attention. Share your results with the SC34 committee and with the OASIS and ECMA committees.

  • Patrick Durusau has made a request that he thinks the area of checking how well some of the detailed descriptions of formula functions in IS29500 accords with the reality of Office as currently implemented, would be really helpful. This would help both IS29500 get improved and provide better information for IS26300.

  • Join in a conformance testing group: make up test documents. Ideally a test library will have some tests that test one thing per document, which makes a very large number of documents, and others that test cascaded errors. So I wonder if algorithmically generating test documents from schemas is viable.

  • Get you National Body to submit more Defect Reports, so that SC34 does not lose impetus. Remembering that when something becomes a standard, maintenance becomes a community job not "their" job.

Of course, if you were not interested in being constructive but in trying to frustrate yourself there are other things you could do. You could for example, mount a court action asking for something that you know to be impossible (e.g. withdraw a vote on a ballot that has been closed), with reason that you know won't stand up (e.g. that a committee of long-term experts changes it vote after being satisfied that there have been enough changes to proceed with a standard), with odd legal ground (if the voluntary standards group is not subject to administrative law, not being under the government), and where you know that your standards body's final vote is a credible one (because it was shared by more than an absolute majority of other National Bodies around the world.) Why would someone do that, my readers might be asking themselves? Embarrassment? Sour grapes? Vindictiveness? Marketing?

I certainly hope that national standards bodies will stand by their committee members and provide financial support during court cases, for time and expenses the private individuals will be dragged away from their work. This kind of intimidation, to use courts and the threat of legal action to force a result after you have lost the technical argument, should be seen for what it is.

Now please, I am not saying that I have confidence in every NBs votes. While I believe that every NB acted intra vires and therefore legal overturns are futile, I was not pleased with the Norwegian national vote (for just the same reason as I was not pleased that several NBs voted for ODF bypassing their technical committees too) and the Brazillan vote (after an IBM representative blogged that he had convinced them that if they had *any* outstanding technical problems they should vote no: if he is true, the NB secretariat should have picked this up in committee and told its members that perfection is not a requirement for a standard.) But, I don't see them as acting outside their powers.

And, most importantly, it is a different class of problem to have a standard accepted than to have it blocked.

National and international standards bodies are highly aware that their activities and importance is tolerated and encouraged only because they create markets. The minute a national or international standards effort becomes a servant of some clique or cartel, to the exclusion of others, it loses its fundamental justification. (I say "effort" because a body may have thousands of efforts on the boil at any time.) For standards bodies, exclusive behaviour is a mortal sin; in comparison, too much inclusiveness (i.e. by having multiple standards where in a perfect world we could imagine having only one) is only a mild (and bearable) fault. (And, indeed, in most cases I consider support of plurality, to allow the market to choose, a positive virtue.)


2008-05-02 10:42:41
The web adopts enablers and abjures disablers. The right to exist is in the utility, not the signature.