Given the sheer ubiquity of Microsoft Office documents, it may seem a bit quixotic to invent a competing set of document formats, and drag it through the standards bodies, all the way to ISO. The Open Document Format people have just accomplished that, and are now being hotly pursued by … Microsoft and its preferred Office Open XML specification.
If the creation of interoperability between similar but different programs is the sole purpose of a standard, office documents don’t look like much of a priority. In so far as there is any competition to Microsoft’s Office at all, the first requirement of such programs is to read from, and write to, Office’s file formats as if they are Office themselves. Most of these competitors have succeeded to such an extent that it has entrenched the formats even further. For example, if you’d want to use the same spreadsheet on a Palm and Google Spreadsheet, or send it to an OpenOffice using colleague as well as a complete stranger, Excel’s .xls is practically your only choice.
Yet the Open Document Format (ODF) has slowly wound its way from its origins in the OpenOffice file format through the OASIS specification body, and is now a full ISO standard; the first of its kind. But not necessarily the last: the confusingly named Office Open XML (OOXML) is already an Ecma specification, and the intention is to shift it to ISO too.
To understand why the ODF and OOXML standard steeple chase is happening at all, it is necessary to look beyond the basic interoperability scenario of lecturer A writing a handout that’s then downloaded and looked at by student B, and perhaps printed by lecturer C next year. That sort of user to user sharing was solved a long time ago, once Corel’s suite ceased to be a major factor. But there are some longer running issues with data exchange at enterprise level, and with document preservation.
Enterprise data exchange is perhaps the most pressing of the two. For an awful lot of information management purposes, it would be handy to have the same information in a predictable, structured format. Recent JISC project calls, for example, call for stores of course validation and course description data- preferably in XML. Such info needs to be written by practitioners, which usually means trying to extract that structured data from a pile of opaque, binary Word files. It’d be much easier to extract it from a set of predictable XML.
Provided you use a template or form, that’s precisely what the new XML office formats offer. Both ODF and OOXML try to separate information from presentation and metadata. Both store data as XML, and separate out other media such as images in separate directories, and stick the lot in a Zip compressed archive. Yes, that’s very similar indeed to IMS Content Packaging, so transforming either ODF and OOXML slides to that format for use in a VLE shouldn’t be that difficult either. It’d be much easier to automatically put both enterprise data and course content back into office documents as well.
The fact that the new formats are based on XML explains their suitability as a source for any number of data manipulation workflows, but it is the preservation angle that explains the standards bodies aspect. Even if access to current Office documents is not an issue for most users, that’s only because a number of companies are willing to do Microsoft’s bidding, and at least one set of open source developers have been prepared to spent countless tedious hours trying to reverse engineer the formats. One move by Microsoft and that whole expensive license or reverse engineer business could start again.
For most Governments, that’s not good enough. Access must be guaranteed over decades or more, to anyone who comes along without let or hindrance. It was this aspect that has driven most of the development of ODF in particular, and OOXML by extension. The Massachusetts state policy particularly, with its insistence on the storage of public information in non-proprietary formats, has led to Microsoft first giving much more complete description of its XML format, and later assurances that it wouldn’t assert discriminatory or royalty bearing patent licenses. The state is still going to use ODF, though; not OOXML.
On technical merit, you can see why that would be: OOXML is a pretty hideous concoction that looks like it closely encodes layers of Microsoft office legacy in the most obtuse XML possible. The spec is a 47 Mb whopper running to six thousand and thirty nine pages. On the upside, it’s single- or double-character terseness can make it more compact in certain cases, and Microsoft touts its support for legacy documents. That appears to be true mainly if the OOXML implementation is Microsoft Office, and if you believe legacy formats should be dealt with in a new format, rather than simply in a new converter.
ODF is much simpler, and far easier to read for anyone who has perused XHMTL or other common XML applications, and it is much more generalisable. It could be less efficient in some circumstances, though, because of its verbosity and because it allows mixed content (both characterdata and tags in a tag), and it doesn’t support some of the more esoteric programming extensions that, for example, Excel does.
All other things being equal, the simpler, comprehensible option should win every time. Alas for ODF, things aren’t equal, because OOXML is going to be the default file format in Microsoft Office 2007. That simple fact alone will probably ensure it succeeds. Whether it matters is probably going to depend on whether you need to work with the formats on a deeply technical level.
For most of us, it is probably good enough that the work we all create throughout our lives is that little bit more open and future proofed, and that little less tied to what one vendor chooses to sell us at the moment.