Alex Brown came up with his 10 points on how to rejuvenate JTC1. [I prefer rejunevate over reform because to me reform means changing it to something else]. While I must admit that my knowledge of JTC1 working is restricted to my observation of OOXML passage through JTC1, Here are my take on it:
(1) Assert the worth of International Standard. With OOXML, JTC1 public image is tarnished, whatever the majority view of the participants actually think. JTC1 now have to either demonstrate its value, or parish. In the process, it might take ISO with it. No more “Standardization by Corporation”? Wish it was there before OOXML passage. This is one of the blatant case of “standardization by corporation”.
Solution to this is simple: Two independent implementations, i.e. the OASIS way. If it is too strong then lets put the proposed standard on probation, i.e., as long as someone commit to come out with a second independent implementation, then a probation period of say one year will be granted from the day of ractification. Until a second independent implementation comes out, the standard will be a provisional standard. It will be withdraw if no second independent implementation comes out within the probation period.
More importantly now does this means Alex Brown will campaign against XPS if and when it comes to JTC1 with only one major implementer?
(2)Recognise the distinctive requirements of ICT Standardisation. I do not agree that we should say that ICT standardization is unique and different from others. Every field can make such a claim. the question is whether ICT standardization meet the goals ISO is created for.
JTC1 as steward for ISO and IEC? No way. I prefer OASIS anytime over JTC1. Their procedures are stronger than the cozy relationship in JTC1.
(3) Redraft the JTC 1 Directives. Directives is OK. It is the implementation of the directives that needs to be tighten up. I disagree with Brown when he says current procedures “empower administrators over nations.” If you read the reason for rejecting the four appeals you can see the management of ISO and IEC squarely put the power in the hand of nations. How else can one interpret the sparkling of “Member nation had consider XXX and had chosen to approve it”?
If any, we need to strengthen the hand of administrators, staffing them with people familiar with the field and give them the right to exercise their professional judgment. Administrators and member nations should counter-balance each other.
(4) Move away from paper-based model. If I remember correctly, a lot of country requires elaborate procedure before some law is enacted. For example, it has to be written on a special type of paper by hand and carefully stored in a place where people can inspect it. Therefore, a total move away from all paper copy is a wrong move. We need at least one paper copy and several clones of it around the world. It is like the “Standard kilogram” stored in Paris and national copy of the said kilogram else where.
However, like my digital and non digital weighting machine, electronic version of standards has its place. Therefore I think JTC should explore learning money from wide distribution of digital copy of their standard. Take for example the PDF copy of C++ standard for less than USD25 when I bought it from ANSI.
Digital copy is no substitute for paper version. Hence it is important that the paper version is produced first as it is the GOLD standard.
(5) Widen International Participation. The current composition of JTC is a reflection the history of ICT evolution. As ICT becomes more important, you will see more nation take notice and becoming active in JTC.
Widening International Participation is always good. However, we do not want to see a repeat of the accusation that big corporation “buys” vote by agreeing to lavish money on smaller nations.
(6) Find a way for vendor-lead standards to mesh with JTC1 processes without compromising international control. I concur. I want to see a reform of the fast-track process. Currently, JTC1 cannot exercise any control over ECMA or OASIS submission. Everything they submit must be put to the member nation for voting. This is wrong. It is compromising international control. If they propose crap standards, JTC1 administrators must have the rights to reject it.
I do not think JTC1 should take over control of any standard the way it proposed to do with ODF. In fact, in my view, OOXML maintenance should be returned to ECMA and let ECMA maintain it. Both ODF and OOXML are developed elsewhere and it makes no sense for JTC1 to take over. For standard developed in JTC1 via standard ISO development process, then there is a case for JTC1 to be in control.
JTC1 should be the oversight committee overseeing both maintenance regime and provide facilities to reconcile/arbitrate differences between the different standards under its auspice.
(7) Periodically change the nation having the Secretariat and Chair appointments. I sense some national rivalry (and jealousy) here. Are we going to suggest ISO rotate its secretariat around the world? No. Look at the European Parliament. Having facilities at different sites make things more difficult.
Some ability to change secretariat and chair appointment is needed, however. I would not recommend a two terms only chair and secretariat as imposed on US Presidents. I will require the chair and the secretariat to be up for election at fixed term.
And if the majority wants the USA to remain chair and the secretariat so be it.
(8) Balance transparency and confidentiality. Transparency first, confidentiality last. If one want to earn respect ISO Standard, then one have to demonstrate transparency. Confidentiality should only be granted as the very last resort.
(9) Clarify intellectual property policies. This is extremely important. I hope JTC1 use its weight and expertise to craft an IP policies that will be held out as the GOLD standard for everyone else to follow.
(10) Encourage best practices at national level. JTC1 can certainly nudge member nation to raise their game. However, it should start by setting an example. Currently, it is a BAD example.