January 12, 2006

Formats Aplenty

As Ryan pointed out, the entire LabVIEW 8.0 Help is now posted on ni.com.

Because the LabVIEW Help .chm files are actually made up of individual HTML files, preparing those HTML files for the Web wasn't such a huge leap. Currently, the most challenging type of technical documentation to present on ni.com is printed or PDF content.

We make a PDF of every printed document that ships with an NI product and post the PDF at ni.com/manuals. But none of that content is searchable at ni.com. That's an awful lot of content to exclude from our site.

Google has a solution that works pretty well but there are also some big concerns with implementing something like it at ni.com. The biggest of which is that when you click an ni.com search result that happens to be a PDF, you then have to search for your terms again within the PDF. And some of these PDFs are quite large.

While one 300-page PDF might contain all your search terms, they don't necessarily appear on the same page or even in the same chapter. So exposing the PDFs through the ni.com search as is could lead you on a wild goose chase of clicking a link, waiting for a huge PDF to load, using the less-than-ideal PDF search feature to find each of your terms, realizing the document doesn't have what you need, and moving onto your next big PDF to repeat your search again. Not exactly the sort of user experience to instill trust in our technical documentation.

We've been investigating ways to convert the PDF content to HTML so we can display it in smaller individual documents, each of which would be indexed separately in the ni.com search. Even aside from the fact that we're starting from a not-so-flexible source format of Adobe Framemaker, this PDF content wasn't necessarily designed to be split up into smaller, disconnected chunks. Designing content for print/PDF is just fundamentally different than designing for online help.

The good thing is that all these problems are solvable, and we are pretty close to a good solution. A lot of folks point to the up-and-coming whippersnapper of the Format World as the solution to this problem—XML. But converting a Framemaker document to XML is no small feat, so it's hard to imagine a time in the near (or even distant) future when all NI printed/PDF documents will be sourced in the same format.

And who knows—by the time we've got most documents converted to XML, maybe a newer, better format will have emerged on the scene.