They’re not announcing the results officially until tomorrow, but it appears that enough countries have changed their votes since September 2007 for OOXML to become an ISO standard. Some of the key ’switchers’ responsible include the UK, Ireland, Denmark, Norway, Japan (all from No to Yes) and France (from No to Abstain).
We need document standards to preserve business data over long periods and thus as a core principle it’s a good thing to have an ISO standard used by Microsoft Office, the dominant business office suite - and for good reason, it’s been a great product over many years. Personally I find the features added to it since around 1997/2000 to be entirely optional in all practical real-world use, so I jumped off the upgrade treadmill some time ago and am quite content with Open Office now for both personal and business use, but that doesn’t detract from the fact that Office has always been good. Milked until it’s drier than the Goby desert maybe, but still good 
The problem that most (sane & rational) detractors have with OOXML is not that it’s from Microsoft (although that justifiably promotes extra scrutiny from the off, given their track record) - it’s that, as a standard, it’s borderline unimplementable. The whale-crushing 6,000 page spec has proved somewhat inpenetrable to most people outside Microsoft, even with the changes since September. Ambiguity, references to existing Office behaviour, patent concerns (remember, Microsoft’s open specification promise excludes all commercial implementations) all make OOXML something of an encumbered standard at present, and encumbered standards are potentially worse than no standard at all - because they can give an impression of openness without actually delivering on the benefits that it’s supposed to bring; i.e. a lack of vendor lock-in for your critical business data. There is exactly one implementation of OOXML right now - Office - and without a clear, unambiguous, no-nonsense spec there won’t be any others that you can trust either. Which the cynical among us would say is exactly the way Microsoft wants it.
I don’t have anything against OOXML in principle if it’s truly open and genuinely implementable by someone outside Microsoft. Sure the world doesn’t need 2 document standards, but if MS insists on having their own (what’s new, I think ‘Not Invented Here’ must be in their training manual), let them if that standard is truly open. A 6,000 page spec with 1,100 comments still largely unaddressed at the last reading doesn’t seem to fit the bill to me - maybe if more time had been spent hammering those things out I might have felt like it’s respecting the spirit of openness, but this just feels like a wave-through by vested interests to me. Let’s see what happens over the next year or so, but I’m not confident about OOXML doing anything but sowing confusion and doubt in the document interchange arena, all the while being spun by the PR machine as an interchangeable standard. Business as usual then.
This is a good write up of the process so far and why it’s been seriously lacking.
April 1st, 2008 at 5:50 pm
It wasn’t until reading this that I realised just how technically horrible OOXML is. I’ve lost all faith in my country after voting yes for this monstrosity.
April 1st, 2008 at 8:03 pm
Great example. They can’t even be internally consistent, no wonder it takes 6000 pages - not even counting the ambiguities and vague references to ‘behave like Excel’, that’s a nightmare
April 1st, 2008 at 8:30 pm
lol @ “not invented here”… I’m sure it might have something to do with incompatible / unclear licencing of some standards / and source code. Can you imagine how many people would love to uphold software patents against them? Or would love to sue MS? I mean, it’s almost fashionable already, isn’t it?
You could also argue that the MS C++ compiler, and the upcoming IE8 are *very* standards compliant. IE8 has the problem right now that if it implements standards very, very strictly that it will mean many websites would no longer work.
April 1st, 2008 at 9:45 pm
You can’t be sued for using source under the proper licensing, nor for implementing a common standard from scratch. You’ll get sued for patents whether you reinvent your own version or not - it’s the idea that counts there. So I’m not seeing that as a particularly strong argument for constant reinvention (of specifications and standards) - except that MS likes owning everything they work with. It’s a control thing, and I’m sure a lot of it is down to nervousness, but a lot of it is also down to culture. You think they’re any different from IBM at the end of the day when it comes to legal exposure? And yet IBM supports common standards considerably more often rather than going off in its own direction. Culture.
The MS C++ compiler is much more standards compliant yes (although it does lag gcc still), and it sounds like IE8 might be the best standards version yet - these have taken some years to evolve to that stage.
I don’t buy the ‘we’ll break websites’ excuse at all though - any website worth its salt already works with more standards compliant browsers, 90% of the hacks that exist are there precisely because IE has made such a pigs-ear of it in previous iterations. Sure you might break some sites where the writer is clueless enough only to ever target IE, but they’ve had that coming for years, maybe they’ll finally wake up to their responsibilities and also come to realise *why* standards compliance is important. Opera is my best test of sites so far, if IE8 does it better and exposes valid holes in my CSS, then I should fix them (and perhaps learn something in the process).
April 2nd, 2008 at 2:12 am
It is April 01, no?
April 2nd, 2008 at 9:44 am
@futnuh: I truly wish that was it.
April 2nd, 2008 at 4:27 pm
“I don’t buy the ‘we’ll break websites’ excuse at all though - any website worth its salt already works with more standards compliant browsers, “
Ain’t gonna happen.
There is a reason why a binary app compiled around 1997 still has a very good chance working on the latest Windows,, while every RH or Fedora release requires a new set of RPMs because shit breaks all over.
It has something to do with “The Customer Is Always Right” thingy …
April 3rd, 2008 at 9:40 am
Hey Steve, re IE8… Joel has written an interesting piece on this:
http://www.joelonsoftware.com/items/2008/03/17.html
April 3rd, 2008 at 10:04 am
I get the pragmatism argument, but I also think the ‘damned if you do, damned if you don’t’ argument is actually in favour of change. If you’re going to be damned, you might as well be damned for doing it right rather than wrongly. There’s a grey area here of course, but I think Vista is a perfect example of what happens when you let compromise rule your thinking. Vista started out with some good ideas of what to do with the next Windows desktop, and over the years most of those things got whittled away, perhaps many of them because of backwards compatibility concerns, or just because the effort couldn’t be expended on them because of effort going elsewhere. The result is a deeply underwhelming copycat OS that most people feel breaks enough to be annoying, but doesn’t add enogh new & exciting to be worth the hassle. Getting stuck in that no-mans land in the middle is the worst possible scenario - you need to decide direction you want to go. In my mind, server OS’s should lean toward stability & compatibility, desktop OS’s can lean more toward innovation because they’re less critical and people like innovation there. Perhaps part of the issue is MS trying to make one general OS strategy address both things.
Both windows an IE have suffered from their own successes, but also because of poor engineering decisions were made during their development - speed or convenience over good security or standard compliance usually. Those decisions accumulate and make things harder in the long term, particularly if you want to keep all that backwards compatibility, and at some point someone has to say ‘enough - we have to do it properly this time’. If you don’t, you’ll never be able to move beyond history because there’s an inertia there which will hold you back - that’s certainly one of the factors in Vista’s poor reception I think.
I lost all trust in IE many years ago and am very unlikely to go back to it, but I will be glad if in future I don’t have to butcher my CSS so badly to make it work properly on the browser most people use.
April 3rd, 2008 at 10:15 am
That said, I am in favour of the ‘I’m tested with IE8′ tag. This is nothing wrong with standards, it’s because IE has been so messed up in the past people have had to code sites specifically for it, and those hacks will take a while to weed out. My guess is that ‘upgrading’ sites to IE8 will mean removing a ton of crappy CSS and Javascript hacks we never should have had to write in the first place. What I am saying is that it shouldn’t for a second be used as an excuse for IE8 to dodge implementing standards correctly. If it has a fallback mode AS WELL, preferably controlled by the site rather than the user (and probably defaulting to ‘on’ if the site doesn’t specify, at least in the medium term), that’s fine. I just want to be able to remove my IE hacks long-term, because they suck. I’ll have to keep them for years anyway because people will take ages to upgrade, but at least if the standards-compliant implementation is there, some day I may be able to get rid of them entirely. That can’t happen if backwards-compatibility is used as an excuse to dodge the issue and the standards-compliant implementation isn’t in there at all (albeit in ’stealth’ mode for old sites).
April 3rd, 2008 at 2:51 pm
It was a while ago I read Joel’s article on IE8, but if I recall correctly I was unimpressed, mainly due to an argument to the effect that there is “no such thing” as a Web Standard.
Joel argues that Mozilla, Opera, Webkit, etc. are all, like IE, a “standard unto themselves” and that the Standards themselves are riddled with ambiguities. As someone - and I am so totally not alone - who has written a fair number of web sites and applications based on the W3C specs, this is total BS. Non-IE browsers are in most cases about 99% identical in how they interpret the Standards and they cover any remotely significant part of them with commendable consistency and intuitiveness. Sure, things aren’t perfect, but the inevitable quirks are utterly, insignificantly trivial compared to the hell one experiences when trying to work with IE. Relative to how divergent IE is, Mozilla, Opera and friends all work pretty much exactly the same.
Some of the misgivings about Web Standards I think are due to old attitudes from the latter part of the Netscape era, and while vestiges of that horrific period still haunt us, for the public Web the realities of this ancient time can be safely forgotten.
More generally, Microsoft has billions of dollars riding on continued vendor lock-in horrors. I’m not saying they’re some greedy cartoon villain and their competitors are perfect angels, but it’s simply not in MS’s interest that people are able to take their own data, that is in a Microsoft-entangled format, and use it on a different platform. They will go - and have gone - to great lenths to intentionally (and sometimes artificially) maintain as much incompatibility as possible.
April 3rd, 2008 at 4:25 pm
I agree. I had to make perhaps 1 or 2 changes to my sites to cope with the fact that Opera is a bit more strict than Firefox, the vast majority of the crap is always to handle IE. Sure there’s no 100% standard, but at least most other people get close. If IE8 has a mode which is as close to the standard as FF2, I’d be happy that we’re on the right footing for the future.