In a recent blog entry, Microsoft fanboy Rick Jeliffe tries to compare SQL to OOXML by quoting ISO SQL editor Jim Melton, who said:
Or perhaps most people were somewhat intimidated by the prospect of (thoroughly) reviewing a 6,000 page document. To put this in perspective for those who know SQL’s size and complexity, the sum of all nine parts of SQL is about 3950 pages. A ballot on SQL frequently receives several thousand comments, and we’ve been balloting versions of SQL for 20 years!
Rick Jeliffe then says:
Under that kind of criteria that our Big Blue friend is proposing, the ISO SQL standard which is one of the most widely implemented and important and mission-critical of all ISO IT standards would not be of high enough quality to make the grade!
A little later, Jim Melton comes back with this less than happy response:
Whoa, there, Rick. If you're going to quote me, then I want to be sure that the context is available to your readers.
One relatively important fact that didn't show up in the words you quoted is that the standardizers of SQL weren't so arrogant that we thought we could rush (and it's hard to deny that the phrase "Fast Track" implies hurry) 6,000 pages in one go, without it having not been visible to the vast, vast majority of the world until it started its FINAL ballot.
So, it took the SQL world some 20 years to write 4000 pages of standard, to root out the serious bugs (and thousands of smaller, mostly editorial, bugs at the same time), and to reach a genuine consensus on the content. I don't know Rob Weir, but I think it's misleading and unfair to extend what he said by concluding that SQL is not of sufficient quality to "make the grade". I do not believe that anybody thinks that there are still 200 serious errors remaining in SQL's 4000 pages, must less a thousand or two. Why? Simply because we have taken the time…years of it…to carefully root them out and fix them, giving the world at large plenty of time to review our bug-fixing efforts.
What I see DIS29500 doing is exactly the opposite. You've written 6000 pages of specification largely in secret (and, I understand, recently added over 1500 more pages) and given the world five months to read, absorb, understand, review, critique, and establish informed positions on it. Worse, whether it happened because of unreasonable methods, pure random chance, or genuine and unexpected interest, the fact that the size of the JTC 1 Subcommittee that was to vote on the document suddenly exploded gives the appearance that somebody was trying too hard to stack the deck…almost as though it wasn't really desired to have too much real review. Please note, I don't know any facts at all about the membership changes in SC 34, except that it happened. I'm not accusing anybody of anything, merely stating what people have inferred from those facts.
In my not-so-limited experience, if a 5-month (or 9-month) review of a 6000 page spec revealed "1027 unique issues", then a truly open process in which the document went through the normal WD, CD, DIS, FDIS process would almost certainly reveal upwards of 5000 unique issues. Also in my experience, fixing even 1000 non-trivial bugs is a very daunting process that takes many months, perhaps two or three years (given a reasonably high frequency of meetings — say, three 3-week meetings per year).
In short, I want to emphasize that I think a Fast-Track process for any standard of this magnitude is a monumental mistake and a serious perversion of the entire concept. I was wary of that process when it was introduced, but saw that (initially, at least) it was being used for moving well-established, very widely implemented specifications into the ISO world for maintenance and possible additional development. Speaking solely for myself (and I EXPLICITLY disclaim any intent to imply an Oracle viewpoint, a USA viewpoint, or a North American continent viewpoint!), I find the whole thing appalling.
Jim
P.S., Please note also that I have taken no position at all on the merits of standardizing the technology in the spec, nor even the merits of the technology itself. I am ambivalent about whether the world community would be better served by one standard in this space or two or more (I know that the world is a Better Place for having one standard for relational database management, and a Better Place for having more than one standard for programming languages). I object solely to the process by which this has taken place.
P.P.S., Again, without making accusations about anything, are you aware that the international standards community generally views ECMA now as a wholly-owned Microsoft subsidiary? I offer no opinion about the validity of that view, but almost everybody to whom I have talked about ECMA dismisses it as little more than Microsoft's bought-and-paid-for channel for submitting documents for pretend standardization. That sounds a bit harsh, but that's what I hear.
P.P.P.S., One last thought and I promise I'll close: You refer in your text to "IS29500". That's rather premature, isn't it? At the time of my writing this response, the standard has not been ratified. So it's still DIS 29500 at the moment.
