[Arcana] Being clear about this.

Karl Fogel kfogel at red-bean.com
Sun Jul 29 21:35:04 CDT 2012

Jim Blandy <jimb at red-bean.com> writes:
>Now, this I can agree with *wholeheartedly*.
>I think the problem really started in earnest when the iPhone appeared
>and showed everyone else up. I imagine other phone makers crying, "Our
>phone has all the same features [if you're a checklist-driven idiot],
>so why don't people like it as much? And our phone doesn't drop calls!
>[Nokia, for example, was much better at actually making calls.] Apple's
>so great at hype! [True, and necessary, but not sufficient.]" But it
>forced people to recognize that they really had a blind spot that
>consumers really cared about. And it wasn't just phone manufacturers
>who began to wake up, but people developing web sites and writing
>desktop software as well.
>The expertise supply that has appeared in response to that demand is
>generally underwhelming, in my experience. 

[Dammit, you're saying this so much better than I said it.  Stop that!]

>I think a good designer has to have all three of these:
>- a developed visual sense, like an experienced graphic artist (and
>not all graphic artists seem oriented towards pleasing people)
>- a developed sense of semantic models and schemas, like a software
>architect (and you know many programmers aren't good architects;
>perhaps most aren't)
>- a developed sense of, I guess, humans' cognitive abilities, like a
>cognitive scientist(???). This is what Jef Raskin went off to study
>when he realized he didn't know how to answer the questions that kept
>coming up.
>But mostly, the designers I meet are people who've got the first, see
>that we hackers suck in this area (absolutely correctly), but who don't
>even have the wherewithall to *recognize* their limitations in the
>latter two.
>We do have people at Mozilla who've got all three. But we also have
>the more ordinary type.
>And I agree that the expertise which is out there is usually directed
>at designing low-investment interfaces. It's very consumer-oriented.
>And Emacs isn't, and that's good, nay, *precious*.
>But I also see a lot of clever solutions in the consumer stuff that I
>miss in Emacs. We have no app store; the Emacs wiki is almost like
>that, but the install experience is horrible. Our display model is
>pathetic. And there are places where Emacs itself doesn't do
>incremental-search things that it should: why isn't ibuffer (I think
>that's what it's called?) the default? Firefox's Awesomebar does it,
>as does Sublime's buffer-switching stuff.

Oh, gosh, so true.  Emacs doesn't learn from the non-Emacs world nearly
as fast as it could :-(.  There was a brief period of time when browsers
(having finally gotten completion at all) actually had IMHO better
completion behavior than Emacs did, until Emacs finally changed its
completion to not be purely prefix-based.

I cry when I think of what Emacs would be today were it not run in the
most dysfunctional manner a project has ever been run in, with the
possible exception of the first half of the Carter administration.

>An aside: there's extraordinary resistance to even the most minor
>changes. One time we reordered two items on a context menu (it made
>sense; it was more consistent with other software) and our support
>organization (our feedback organ) was flooded with complaints for
>weeks. Partly that's just an effect of having 200 million users, but
>still, you can see that it takes guts and faith on the part of
>management to follow along.

Yeah -- it makes me kind of respect the Gnome 3 team too, even though
(so far) I really dislike what they did with Gnome.  They knew there
were going to be complaints, and they forged ahead anyway because they
thought they had a winning interface for the long term.  (Jim, if I
understand correctly, you rather like it, so who knows, I could just be
the house interface curmudgeon).


More information about the Arcana mailing list