Lars Nooden wrote: > Mathias Bauer wrote: > >> I can even imagine a much simpler approach: just keep the old UI as an >> option and use usage tracking to find out how many users switch to it. > > Studies for the courts found in 1996 that even back then 60% kept the > default settings. By now it's close to 100%. I've tried paying people > to customize Xfce and KDE UIs. It's just not in popular culture anymore > do to more than whine about defaults.
Well, we are not talking about whether people use settings, we are talking about *the* setting. If a default setting creates so much pain as people think, users will want to know how to change it and they will find out how to change it. Of course it's necessary to take the facts mentioned by you into account (and it's backed up by results from the Microsoft research that IIRC less than 10% of their user base use customization at all). So if e.g. 20-30% of the user base (don't nail me down on exact numbers now) would change the default of something it would be a clear sign that something is wrong (IMHO). Or viewed from the other side: that means that saying "70% of our users use the default, so it must be right" is self-deception as long as you don't have a clear sign that those 70% have chosen the default by intent. > 1) Rather than find out how many change it, find those who *do* change > it and find out how *they* change it and then study why. Yes, but if you have to focus your work it's first necessary to get some numbers about how many people do that at all. > 5) Have a taser charged and ready for anyone proposing to copy the > failure known as "the ribbon". Ditto for vague, nebulous, or diplomatic > speech that could be interpreted as such. Sorry, but that is nonsense. "Could be interpreted as such" is an excuse to suppress every contribution that you don't like by alleging it falls into that category. Or explained in a Pythonesque way: If you want to throw stones on someone, allege he has said "Jehova"! You have modernized that concept by using a taser instead of throwing stones, but in fact it's the same. > 6) Interview experienced trainers. I can propose at least two. > > 7) Bring in real HCI specialists to do a) time-motion studies, including > eye-tracking. Use as subjects educated people *who've never used a > computer before* to gather data. Growing markets have lots. Then bring > in novices and power users. Compare and analyse. > > etc. The interesting part is that Microsoft did exactly all that (except no. 5 of course ;-)). And the result is not what you like and it's also not what I like. So there are some more things to consider. I will talk about that in another mail. >>> I'm very open to improving the UI, but only in the context of greater >>> efficiency. Change for the sake of change is not good either. >> >> Agreed. I hope you don't want to say that Renaissance is about the >> latter. > > I don't want to but, of necessity, have to. Work out a scientific > approach, and the assessment can change. In the mean time, #5 above is > probably the most valuable in the near term. Proof by allegation. >> IMHO it is pretty clear that our current UI needs a change, > > Of course. However, survival of the fittest random changes doesn't > really work in a population of one program. Yes, but I can't see where this concept is applied. Collecting feedback for an applied change isn't "making random changes and see which survives". It's just the realization that absolute truth can be achieved rarely (if at all) and that it's never a bad idea to plan for the chance that you have made a mistake. Regards, Mathias -- Mathias Bauer (mba) - Project Lead OpenOffice.org Writer OpenOffice.org Engineering at Sun: http://blogs.sun.com/GullFOSS Please don't reply to "[email protected]". I use it for the OOo lists and only rarely read other mails sent to it. --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
