darcusblog

Ed has two recent posts that ought to get one thinking of the possibilities of libraries—and in particulaar big data providers like OCLC and the Library of Congress—getting on board the semantic web train. The first is a more high-level goal of the open data movement, complete with nice diagram. The second is a much more grounded example of the kind of practical things that can make it happen that he and I put together. Allow me to illustrate from my command-line:

$  xsltproc \
http://inkdroid.org/data/identity-foaf.xsl \
http://orlabs.oclc.org/Identities/key/lccn-no99-10609 | xmllint --format -

<rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns:srw="http://www.loc.gov/zing/srw/" xmlns:foaf="http://xmlns.com/foaf/0.1/">

<rdf:Description rdf:about="http://orlabs.oclc.org/Identities/key/lccn-no99-10609"> <rdf:type rdf:resource="http://xmlns.com/foaf/0.1/Person"/> <foaf:name>Berners-Lee, Tim</foaf:name> <foaf:made rdf:resource="http://worldcat.org/oclc/041238513"/> <foaf:made rdf:resource="http://worldcat.org/oclc/048753874"/> <foaf:made rdf:resource="http://worldcat.org/oclc/040278766"/> <foaf:made rdf:resource="http://worldcat.org/oclc/044933478"/> <foaf:made rdf:resource="http://worldcat.org/oclc/045065386"/> <foaf:made rdf:resource="http://worldcat.org/oclc/044281610"/> <foaf:made rdf:resource="http://worldcat.org/oclc/075964549"/> <foaf:made rdf:resource="http://worldcat.org/oclc/044721973"/> <foaf:made rdf:resource="http://worldcat.org/oclc/036040597"/> <foaf:made rdf:resource="http://worldcat.org/oclc/040938943"/> <foaf:made rdf:resource="http://worldcat.org/oclc/051662536"/> <foaf:made rdf:resource="http://worldcat.org/oclc/122918124"/> <foaf:made rdf:resource="http://worldcat.org/oclc/034829358"/> </rdf:Description>

</rdf:RDF>

It took all of about 30 minutes to do this. Now imagine if each of those target URIs also served up (either directly, or via GRDDL) RDF descriptions of those resources …

A few days ago I commented positively on Apple’s Pages styles UI. This is without having actually used it. Having just tried the latest version of the application, I’m rather appalled at how limited the styles support really is.

1. they effectively deprecated the style-based UI in the new version, switching to the more familiar (and broken) direct formatting approach
2. one can only edit styles by directly editing text, and then telling Pages to redefine the style based on those changes
3. no support for style hierarchy it seems (!)
4. whoever designed the default templates and styles did a horrendously poor job of it. The default font almost everywhere is Helvetica (!). This is despite the fact that OS X ships with a really excellent text font (Hoefler Text).

Sigh … so much for good examples. The result would be a nightmare for users really, forcing them either to rely completely on presentational formatting if they didn’t want the defaults, or having to modify every single style (since there’s no inheritance). For the limitations I’ve noted in OOo, it’s way ahead of Apple on this count.

Eric Lai has an article about tension around technical and governance issues within OpenOffice.org. He asked for my thoughts on the matter, and included most of what I said in the article.

I’m struck by how defensive OOo supporters are of any criticism. I don’t really think they fully appreciate the immensity of the task of not just providing a reasonable alternative to Microsoft’s desktop monopoly, but a superior one that will actually draw users to it. In terms of market share, OOo is utterly failing. Just to give my own experience, I have yet to meet a single colleague in my field that uses anything but Word to author their documents. I would be surprised if any of them have even heard of OpenOffice.

Clearly this failure is not the fault of OpenOffice.org alone; indeed, it’s as much as anything a consequence of monopoly. The question raised in various places about the technical and governance infrastructure of OpenOffice.org, then, is not about pointing fingers and assigning blame. It is about recognizing the enormity of the challenge and figuring out how to address it. Our model should be something like Mozilla, which has managed to steadily chip away at both the market share and the mind share of Internet Explorer by providing a superior product, with demonstrable innovations, backed up by a vibrant user community, organized in an effective open community. So the question many of us are raising is: what changes can we make to OpenOffice.org to see similar results?

IBM has released a free office suite based on OpenOffice called Symphony.

It seems they’ve done a nice job overall. The UI looks nice and clean and the website includes a list of easily accessible templates.

One problem with the UI, however, is its emphasis on presentational formatting. IBM is hardly unique here (Google Docs, for example, has no support for user-defined styles at all), but I’d like to think UI designers can do better.

Consider this screenshot:

Styles are present in the right-hand panel, but they are grouped within the local presentational styling section called “Font.” An average user will, not surprisingly, tend to fall back on the presentational attributes to get the formatting they want.

We know that semantic document authoring has all kinds of benefits; from easier document reuse and repurposing, to enhancing accessibility, and so forth. What if instead the panel had a top-level and more prominent “style” heading:

Style
-----
paragraph: ________
character: ________

Provide a wide-variety of excellent templates with a full gamut of possible (semantic) styles, make them available on the internet with previews and browseable from within the application, and users can instantly see benefits from this.

Obviously one needs to make it easy for users to quickly modify formatting, but must this require a “font” panel? Is there really not a better way?

Here’s a nice example from Apple’s Pages:

Styles are front and center in the UI, users can instantly see what they will look like when chosen, and each of them are about the meaning (heading, caption, etc.) of the content, rather than what it looks like (big, bold, etc.). I’d like to see something like this from the OpenOffice universe.

Great to see IBM finally step up and join the OpenOffice project. Of course, there are still some issues to attend to around organization and community to ensure the long-term health and viability of the project.

In related news, we have someone over on zdnet asking:

IBM to give OpenOffice the Outlook e-mail killer it needs?

Oh god, please no. This is exactly the kind of “follow Microsoft over the cliff” thinking that kills the possibility for real alternatives. My university has just implemented what I consider a disasterously poor decision to move to Exchange. I have yet to hear a single person express happiness about this move. The kind of integration that its promoters promised, of course, really only works if everyone uses a Microsoft stack: Windows, Office, and Outlook. I and many others simply won’t do that, and it turns out that the mail server works rather inconsistently with IMAP clients such as Apple Mail. The promised wonder system is, in my experience, a decided step backwards, and it is so precisely because of its monolithic integrated design.

No, the solution is open, decoupled systems; disparate services and applications tied together though open standards (think efforts like CalDav). OpenOffice does not need yet more integrated functionality; it needs to be stripped down. OOo will not have much of a future unless its developers look forward and ask how personal computing and productivity might look different in the internet age.