Actually. I have no need to do so. OpenMalaysia blog did that for me. I am not even going to provide a link to the ZDNet article in order not to increase the popularity rating for Mr Jelliffe. However, I recognize that reader might be interested in the article, so what I will say is the article is available on the OpenMalaysia blog I referred to.
Only a few things that need serious/repeated debunking:
- Storing date as numbers in OOXML is not a problem and is not significant
Sorry mate, it’s significant. What is day 31003 in Gregorian Calendar? XML is suppose to be human readable as well. A lot of small developers depends on it.
There is a reason why ISO, when formulating its date format, decided to use human readable string rather than numbers.
- ODF itself refers to a lot of non ISO standard
First, some of the standards referred to by Mr Jelliffe is also referred to by OOXML.
Second, at least they stick to and use existing standards, rather than reinvent the standard, or worse, use one that was rejected by standard body.
Use of existing standards allows reuse of code and knowledge. Reinventing them, as in OOXML, requires more coding and relearning.
- OOXML is long because in the ECMA committee process, a lot of information had been requested by committee members and MS shows openness by adding them in.
And yet there is a LOT of necessary information NOT included in the OOXML, and the readers are requested by the standard to refers to legacy Office software.
This is washing dirty linen in public, not MS showing openness. If the committee manage to get MS to add a lot more meat to the standard as claimed and the standard is still “incomplete”, just imagine how awful the original MS submission is. Since the results is STILL so awful, the committee did an extremely bad job.
- (Partly as an argument to counter the accusation that OOXML is long) If you count references to standards used in ODF, ODF is at least as long as OOXML
I had already addressed this partly in the previous few points:
- MS uses some of the same standards as well and this will also expands OOXML size in parallel with ODF
- Reuse of existing standards help developers in terms of reduce the need to relearn and recode. At this point I will add that reusing existing standard means there is less likely the standard contained errors.
One other thing that you need to consider before deciding for yourself who is correct is “Should you redefine existing things, or should you reuse existing things”. Well, I can define a CTRAMBLER unit which is 1.5673 of whichever measurement unit you are using, be it centimeters, kilogram or Ampere. Its trivial to convert, but which would you prefers?
This reminds me of the “man-hour” measurement used in the service industry. If you send something for servicing, they are going to charge you by “man-hour”. Usually it means they say servicing your stuff will cost you N “man-hour”. The man-hour unit is different from company to company. The funny thing is, there is no resemblance to how long the company takes to service your stuff. They will quote you a fixed cost, which is a translation of the their “man-hour” unit, and that is all you pay. So, do you want to know about “man-hour”, or how much you have to pay? I am sure “man-hour” is useful to the company in more than one way, but should you care?