Ubuntu’s Overlay Scrollbars

by on May 28, 2011

It’s not often you get a radical change in the WIMP model, but the mighty Christian Giordano has tried just that with the introduction of “overlay scrollbars” in Ubuntu 11.04.

Unfortunately, I think this is what might be called a “misfire”. The main problem is that in hiding the thumb of the scroll bar by default, you are immediately up against Fitts’s Law because the reduced size of the target will slow its acquisition. That’s an HCI fail – and one that will ensure you’re never going to work for Bruce Tognazzi.

The other problem I have with it, from a methodological point of view, is that Giordano is taking his cue for the design from current, mostly mobile, touch interfaces. These sometimes exhibit similar scroll bars in order to reduce clutter on the screen. Clutter is of course a good problem to solve for in the highly constricted world of mobile and tablet UI. But desktop interfaces are a completely different kettle of fish. For starters, the vast majority of people running Ubuntu will be geeks with high resolution screens with oodles of real-estate available. Indeed, even if they’re not geeks, it’s hard to find anyone with a screen of less than 19″ at 1280×1024 these days. So that’s a UCD fail in not considering your users. The aforementioned Fitts’s Law issue is also aggravated by large screens with high resolutions because of the large distances between pointers and targets.

So it’s a nice try. But no cigar. I’m turning them off, and so should Ubuntu, I’m afraid.

Comments

Hi Jonathan, surely this is a brave change (see Fitt’s Law visual target issue), but you are very wrong on judging the method.

In facts:

1) Ubuntu is a universal OS which could run on MANY devices: http://www.omgubuntu.co.uk/2011/05/install-ubuntu-on-the-htc-desire-hd/
Some companies are already selling tablets based on Ubuntu.

2) Unity, the new UI for Ubuntu, comes originally from the netbook edition. Netbooks, which have very small screens, are a good platform for Ubuntu because the recent versions of Windows run just too slow there.

3) Ubuntu is trying hard (surely not an easy task), to jump the chasm. So our main target is definitely not geeks (which could also easily tweak or disable them).

Best, chr

How long have you used it for before deciding you don’t like it? I’ve not used it, so I can’t really comment, but it seemed like a brave design which, if successful, removes quite a lot of chrome (always desirable, no matter how many pixels you have available.)

What I found clever was how it dealt with the usual problems of only-visible-when-moused-over solution to remove clutter – well described by Cennydd here http://www.cennydd.co.uk/2010/end-hover-abuse-now/

By retaining a small scrollbar, a visible target remains. The brain just has to re-train itself to aim for the predictable appearance of the actual, larger target.

Anyway, curious to see how natural selection treats this particular little WIMP evolution.

I used it for about a month. It’s also worth noting that it only turns up on some apps (gedit, terminal and a few other). Firefox, OO and most major apps use normal scroll bars. So you have to re-train your mind in some cases but not others.

BTW in a mobile context the Fitts’s criticism wouldn’t apply since all scroll bars would presumably be hard up against the edge, thereby having infinite hit area.

So don’t get me wrong – I’m not saying I dislike the idea per se. I just think it’s a misfire on the desktop. The problem of extra pixels there isn’t a big enough problem to mess around with the utility brought by traditional scroll bars.

I also have to say that imposing one platform’s limitations on another is a somewhat odd way of doing design. What next? Banning indicated states on the desktop because they’re not possible on touch screens? Oh, wait:

http://trentwalton.com/2010/07/05/non-hover/

Sigh.

Leave a Reply

Your email address will not be published.

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>