Tue, 08 Mar 2005

Prescript, added later: I want to emphasize that this isn’t about dumping on the Abi guys, Marc, Alan, Dom, Martin, etc.- IMNSHO, they make the best word processor on my platform right now, and they appear to be doing great work to extend it and add features that are pretty cool and darn useful (like the grammar checking mentioned below, and the very useful image wrapping.) I’m just lobbying for them to take a step back and think big picture before taking their next big steps, much as Alan is doing, just with a totally different focus ;)

Marc:Grammar checking isn’t feature creep ;) Really, the focus is not whether or not Abi is getting bloated- it is about how you take the features you have (or may add, like the cool image wrapping stuff, or potential features to make it a better simple text editor) and present them to the user. Even the stock toolbar (I’ve updated the link) has a lot of stuff that is useless most of the time, and doesn’t reflect thinking about how users actually use the features that they request. This is what is great about Pages, and about Apple in general- they write applications that reflect what people actually do, not what they say they want to do when writing feature checklists. That’s why we had grip for ages (‘we’ beat apple to the basic idea of a ripping/DB tool) but they got iTunes- same core functionalities, but one is considered to be the standard against which everything else is judged and the other gives you an entire tab for what options to pass to the encoder.

As another example, users say they want buttons for bold, and italic, and what have you, but what they really want, 99.9% of the time, is formatting for different text types:

Compare that one dropdown to six entries in my stock abi and ooo toolbars. [OOo at least has such a dropdown, but as is typical, it is suboptimal, not actually showing you what you’re selecting.] Which actually gets at what the user needs to do to produce a decent looking document? Do they want to remember what font and size they used for every other header? Or do they just want to select ‘header’, change the font/size once, and move on with actually writing headers? Adding such a menu to Abi and nuking most of the text formatting options from the default toolbar wouldn’t necessarily reduce the feature count (all those things could still be there), but it might (along with other similar changes) radically simplify how people use Abi, and make them happier users.

I don’t think OOo can turn into a tool that actually answers user needs like Pages appears to- even adding single, stable, uncontroversial features is a killer (as all of you involved in the wordperfect stuff have noted) so I can’t imagine a radical change in philosophy being at all realistic (though I’d love to be proven wrong.) Abi is a small, active community, with a smaller, nimbler codebase- so it can think about this kind of radical rethink of options and features.

As an aside, this is deeply related to what irritates me about the thread currently raging on desktop-devel (and, as usual, killing usefulness for actual developers): the assumption that polls can somehow tell you want users want. If we’d polled users around gnome 2.0, we’d still be chasing KDE down the ‘more options! more options!’ route, instead of actually focusing on things that are easy to use and functional. Now, actual market and user research done the right way can be useful- I’m intrigued to see that Alan is doing something along those lines, such a thing has been proposed on the marketing list, and Novell certainly has mulled serious user testing (as opposed to polling.) I’d love to discuss that. But instead we’re just encouraging more people to unsubscribe from d-d-l. Blah.

Postcript:Alan: I don’t think that, in the end, ‘Desktop Publishing’ is all that much different from what most people really need in a word processor- they have a handful of formats they typically use, and they want to crank out content in those formats quickly, repeatably, and easily, focusing on the content mostly, and having the machine take care of formatting as much as possible. That’s not really all that different from ‘Desktop Publishing’, is it?