Last year I wrote about some aspirations for our web applications, and as I look back over the year, I'm forced to concede that they largely didn't come true. In particular, we haven't used or integrated any third party applications into our web services; if anything we've done the opposite because we decided that the one web service that we currently out-source – Search – should be brought in-house, and spent a chunk of the summer working to do just that.
Why is this? In part, it's just a demand question; integration with third-party services just wasn't something which our users wanted as much as they wanted various other improvements to our tools and services. It's also because after I wrote last year's notes, we decided to do some big pieces of work which largely ate up the year – new single sign-on, SiteBuilder 2, new look-and-feel for Forums – and those rather superseded my original thoughts.
So what about this year? Well, the first and most obvious point should be that making predictions or resolutions is a foolish and ineffective thing to do, but that doesn't worry me, so here to be checked back on in twelve months' time are my thoughts for this year:-
- Get SiteBuilder closer to Blogs in terms of the ease and frequency with which new features can be added to it. Even though we didn't work for many months of the year on Blogs, we got some good new stuff into it quite quickly, but by contrast the rate of new stuff in SiteBuilder has slowed and slowed since we first released it. In part this is a strategic not a technical issue; we can try things out in blogs to see how well they work, and if they don't work well or they slow things down we can fix or change or withdraw the new features. SiteBuilder is much more institutionally important, so we can't experiment with it quite as freely. But even so, I believe we should be able to make it easier to deploy new releases and easier to add new features than it currently is (indeed this work is already underway, so this is a relatively low-risk prediction ;-).
- Work some Ajax-y stuff into the user interfaces for our applications. We've dabbled with this a little bit already, with bits of Flash and some nice LiveSearch features, but I intend for this year to be the one we start doing more to let people edit and manipulate their content easily with clicking, dragging, and instant, visible changes on the page.
- Keep looking for ways to minimise the burden of application/service support on developers. The more we can build our applications so that people who specialise in support (such as our helpdesk people) can fix things for people without developers needing to be involved, the better. If that means taking a bit longer to write a web UI to do some tasks rather than requiring a developer to go and fiddle with a database, then maybe that's a worthwhile upfront trade-off.
Back in twelve months.