Automation is helpful, and I’d highlight the point you made in passing:

Its often very easy to lose sight of the end users and this is a big problem

That’s about not designers, nor developers, but people who have to work with this stuff day after day, under circumstances most others can’t guess at, and to deadlines that are inconveniently near.

… I’m with Martin on this one. I also don’t believe that complexity is an inverse function of usability (i.e. the more complex, the less usable). To a large extent, many CMS tasks can be automated on the first draft and refined by an reporter / editor subsequently. Tasks like automating the tagging and properties of articles could be a great help for example.

As a developer, I realise how easy it is to create something that ‘you know how it works’. Its often very easy to lose sight of the end users and this is a big problem. Usually, designers only get involved at the front end of the site and never pay a moments notice to what goes on under the covers. I’ve found that getting the designers involved at the level of interface designer to the CMS is extremely important. Just because it is an internal tool doesn’t mean that it doesn’t warrant a similar level of design as the front end.

So what’s my point? I think that the combination of looking at where tasks can be intelligently automated by the system itself and getting the designers involved at an early stage can dramatically reduce the complexity of systems. Granted, it will increase the amount of budget for the development, but the benefits in the reduction of time and stress on the end users should justify this … hopefully.

I picked up on your points about usability because you were (and are) advocating the excellence achieved in personal CMSs to be carried over to enterprise CMS, and I wanted to say that it’s difficult to do because the task in hand is more complex, and they are different beasts. Of course, that doesn’t mean championing usability isn’t worth doing, and I agree that it’s a cause worth championing.

As an aside, I have been preparing these articles simultaneously for this blog and for Guardian’s own “Inside…” blog using the R2 CMS, and there were clearly more clicks needed to get the work done on the R2 CMS. (I didn’t have the luxury of a production staffer to help me). It would have been easy to blame the tools, and possibly the usability of the tools. Of course, I knew I was working for outlets with very different standards, and the result on the Guardian site is a much more sophisticated offering. Still, in the heat of the moment it’s easy to lose sight of that bigger picture.

But I don’t think the frustrations journalists (including me) often express about the CMSs they use have much to do with the debate about “lightweight vs. heavyweight” software as you’ve described it in the past. I think the issue is primarily one of poor backend usability and design.

The weakest point of editorial systems usually has nothing to do with the core software and everything to do with basic user interface design: Crucial options that are hidden away in counter-intuitive places, unnecessary navigation elements that require extra clicks for heavily-used or required tasks, menus that that don’t anticipate what you are most likely to achieve, pages that take ions to load, and so on. When you’re trying to break a story online while the next print deadline looms, every extra click and cumbersome procedure is a serious obstacle.