Re: Contribution

[Calum Benson <Calum Benson Sun COM>]

On Thu, 2006-08-31 at 13:51 +0400, Maxim Udushlivy wrote:

> >> Windows programmers manage to create successful applications without
> >> guidelines.
> >
> > Hardly.. the Windows guidelines are the thickest ones on my bookshelf:

> This is not a reply to my sentence ;) Even if guidelines exist that does 
> not mean they are widely used.

I don't know many "successful" Windows applications that don't broadly
follow the guidelines, though, apart from games.  (And Office, which is
a slightly special case, because it's where MS have always pushed their
latest UI developments first... but the successful ones all find their
way back into Windows and the guidelines eventually.)  Of course, your
definition of successful may be different from mine :)

> Arrived two years ago, so cannot compare... but may be it's just a 
> developers' professional growth, not tied to HIG's in any way?

I would say it's more the case that the HIG has been part of their
professional growth, to the extent that pretty much all the core GNOME
developers are now familiar with the parts relevant to them.  Prior to
the HIG, many of the developers may have been capable of writing
applications that were usable in isolation (although, frankly, some of
them weren't), but I would say the HIG has helped them to write
applications that are more usable as part of a consistent, coherent
desktop.

> >> So here is that principle that I think make Google successful: reduce
> >> number of UI controls and expand application functionality while
> >> preserving UI/functionality coherency. I think that consumer electronics
> >> inherently follow this principle (TV, video recorders, phones, etc.)
> >
> > Yet video recorders and phones have historically had some of the worst 
> > UIs imaginable... so there must be more to it than that.
> >
> You commented an illustration, not the principle itself ;) The principle 
> I expressed is in fact a modern GUI cornerstone! (oops...)

Well, I somewhat disagree :)  In general, it's just always not possible
to continue to reduce the number of controls while also expanding
functionality-- you often just end up with the typical nightmare
VCR/phone/remote control scenario where each control has multiple,
unmemorable context-dependent functions.

What you can certainly strive for is a simple UI that does a few things
well, progressively discloses more complex functionality if need be, and
interacts richly and predictably with the other (hopefully also simple)
UIs around it... much like the original Unix command line philosophy, in
fact.

The Google example you cite is a perfect example of that: their search
page user interface is so simple because it has precisely one function,
and the default behaviour and inherent complexity (as defined by the
search algorithm they use) is completely transparent-- and in most cases
irrelevant-- to the user.  If you head for the advanced search page, the
UI is a whole lot more complex, albeit still clean and well-designed,
but most users never need to see that.  That's pretty much the current
GNOME philosophy in a nutshell, too.

> My main thought about Gnome is very general: too much bureaucracy and 
> politics, not enough technology and real activity ;)

Well, I don't see it like that at all (although there are certainly
times when more cool stuff is happening than others, but that's just
natural), but maybe I'm just too used to working for bureacracy-laden
big companies :)  Which other large open source projects would you say
are doing things better?

Cheeri,
Calum.

-- 
CALUM BENSON, Usability Engineer       Sun Microsystems Ireland
mailto:calum benson sun com            Java Desktop System Group
http://ie.sun.com                      +353 1 819 9771

Any opinions are personal and not necessarily those of Sun Microsystems