Monoculture is bad. I love to see so many different version control systems pop up lately. While Subversion is an excellent and popular choice that will get most tasks done, darcs, git, Mercurial and all work to provide new approaches — primarily when it comes to distributed versioning rather than a fixed client-server architecture. While they all steer in a similar direction, they’re different in the details. You might agree with one design decision here and another there. And perhaps in the long run, the VCS market at large will evolve into one where DVCS comes standard and is implemented well.

Until then, we have a problem. Because we haven’t quite settled yet on the “right” implementation — and, in particular, the one right interface, aforementioned ones such as git all go for their own little ways, and end up being incompatible.

Via David: “the new portability nightmare”.

echo 1 > foo
$vcs add foo
echo 2 > foo
$vcs commit

What was committed, “1″ or “2″? Depends on which $vcs you use.

That says it all.

I must admit I’ve only followed the OASIS OpenDocument vs. Office OpenXML debate from a distance, but the little I did hear left me with the hard-to-believe impression that OOXML proponents largely either work for Microsoft, are paid/bribed by Microsoft or are just plain clueless, and that a general aura of evil hovers over the spec.

Put another way, it’s been very polarizing with little middle ground. There are certainly valid criticisms (particularly jarring are the bizarre, hardly-documented legacy compatibility attribute such as lineWrapLikeWord6) worth discussing, but what I’ve generally found to be missing is a voice of reason.

A voice that actually has a bit of an insight on the process. A voice that can say this:

What Was Good · The people. With a very few exceptions, everyone really tried hard to work together and make the document better. Everyone freely acknowledged that the job was way too big, but there we were for a week to take a run at it anyhow. I include the nations’ representatives and the ISO people and the Microsoft people when I say this; they were, by and large, a pleasure to work with.

This idea that Microsoft is a giant Mordor-like work camp of mindless, soulless ones is far too common, and I’m thankful for different perspectives, like the above one by Tim Bray. And yet, he points out, most of the commentary on the recent ballot resolution meeting is once more “spin”-filled.

It’s not hard to imagine that Microsoft tries as hard as they can for Office’s dominance to prevail. It will, however, be hard to have constructive, productive, useful results when you spend much of your time bickering. There’s a Microsoft-funded BSD-licensed ODF add-in project, there’s ODF and OOXML support in any OS X 10.5 NSAttributeString, and there’s plenty of other examples of how developers can just get to the point, work together, and accomplish something where the user wins.

I see no chance in OOXML going away, and I don’t want OpenDocument gone either. As ideal as a world with one document format to rule them all would be, it strikes me as unrealistic. Instead, let’s get pragmatic, and work to improve both, as well as interaction between the two.

It would be an interoperability win (which is what we really want, right?), and a lock-in loss.