Jef Raskin's Humane Interface

[eichin]

I finally finished Jef Raskin's book "The Humane Interface" and have promised enough people comments that I figured I'd just write a review - and livejournal turns out to be the most convenient place for me to put it up and get feedback, so here it is.

Jef Raskin's name has come up numerous times over the years, as a user interface design god -

Comments

[thomb]

Raskin's book is, in my opinion, simply brilliant. It transformed how I think about user interfaces, and except for insignificant nits, I have no argument with just about anything he says.

One issue with the LEAP approach that concerns me is that finding things (documents, interface elements) is just searching. What causes priority, or rather prevents "pollution" of the command namespace?I'm not sure that you find commands by searching; he suggest that those are all in visible buttons on the screen IIRC. The Cat had them as dedicated keys on the keyboard. But the general question is still on-target. The answer is hidden in the way the Cat worked. There is a special "document" character, which separates documents (just like LF separating lines, and FF separating pages, or whatever). So you search for "DOCUMENT T I T L E" if you want to find "title" at the beginning of a document. So you can't pollute it except by having multiple documents with the same title. In practice, this isn't so much a problem, I think: good isearch

[eichin]

I think I conflated the "applications were a mistake, verbs should work anywhere" aspects with the "text editor" aspects, as well as the "menu should just be chunks of text" idea. And yeah, I'd forgotten the "document marker" character, that simplifies things a bit - also, it occurs to me that in a system like this, "code words" would probably rapidly become useful. I suspect ordering the documents by some useful metric could also help.

As for evolving, yeah, I can see adding "about:python" and "about:books" strings in documents to make them more findable, techniques like that would probably evolve and first and then people would come up with Best Practices of some sort.

Keys are good

[andydent]

My first "aha" moment in UI design was back in the early 80's when I was supporting some Burroughs B20 aka NEC something aka TRW workstations which had a character-mode interface made highly usable by an office application oriented keyboard. There were navigation keys and command keys used by all the apps in a standard manner and the command-line interface was made easy because it was form based. You typed in a couple of letters and a command (or list of matching) was found with all valid params prompted and able to be scrolled through with the arrow keys in a simple "drawing characters" form. When you had filled in the params, you pressed the dedicated "go" key.

I think Raskin is mostly good value. Although a little sceptical about the zooming UI, there's been a lot of research (I think Shneiderman did some) by other parties and in some domains I think it works very well.