[TAG] Linusability

[Ben Okopnik]

On Thu, Jul 08, 2010 at 09:03:17AM +0530, Dr. Parthasarathy S wrote:
> Linux usability -- an introspection (Code name ::  linusability)
> 
> In spite of all claims and evidence regarding the superiority of
> Linux, one aspect of Linux remains to be its major weakness --
> usability. This may be the reason for its slow acceptance by the
> not-so-geeky user community. I am launhing a serious study into
> various aspects of the usability aspects of Linux, so as to list
> out the problems and hopefully help some people offer solutions.

I suppose it depends on how you define "usability". If you're talking
about simple ease of application use for the average user, then you may
be operating from an outdated model of the situation.
 
As you probably know quite well, one of the major parts of any study is
designing the experiment environment - part of which is neutralizing any
observer bias. In this case, the strongest bias that I see is your
belief that Linux has usability as a "major weakness". I might have
agreed with you, say, 10 years ago; you could still make somewhat of a
case for it even 6 years ago. Today? I'd say that Windows has a very
serious usability problem (I just mentioned the screaming frustration of
trying to deal with a simple task under it the other day.) Linux has, in
my estimation, surpassed Windows in usability - and continues to improve
constantly.

On the other hand, if you define "usability" as the total package - that
is, application usage, administration, program installation, etc. - then
Linux may well come up a little short by comparison for the moment. It
won't always stay that way, but for now, Linux definitely has some
places that could stand a lot of improvement. 

On the positive side, I've done a number of Linux installations (Ubuntu,
mostly) over the past several years. In most cases, these people had no
understanding of computers - they were just simple users - and simply
wanted to get on with their tasks: writing letters, sending email,
browsing the Web, using Skype. Once I set up their systems, walked them
through using the applications, and showed them how to install new
software, they were on their own... and guess what? None of them had any
usability problems.  One woman called me up because she was having
trouble using Skype...  turned out the problem was the microphone gain
on her daughter's computer on the other end of the connection.

This is not to say that Linux is problem-free - but then, we don't ask
that of any other OS, so that's not a rational metric. Here are the
problems that I currently see:

1) Windows and MacOS have _very_ slick UIs: once you learn how to use
one app, there's a large hunk of interface dynamics that carries over to
every other app. This is, unfortunately, not true for Linux: given the
"bazaar" method of app creation in the FOSS community, just about every
program outside the standard KDE/Gnome set has its creator's idea of
"the perfect interface" - whatever new and different thing that may
happen to be. Worse yet, the FOSS competition model doesn't work, or
works only peripherally, for refining that part of the system - so a
Linux user has to learn and remember each app's UI. This is rather
inefficient, as well as being a pain in the ass. :) This is a big one,
and I don't know that much improvement is possible.

2) The Linux methodology for program installation is a wonderful thing
(especially when compared with the "infectious diseases ward"
environment of Windows software installation); however, it *is*
different. As such, it needs to be made more explicit and more obvious
to the new Linux user. Fortunately, this wouldn't take much: in my
opinion, an "Install New Software" icon on every freshly-installed
desktop would do a great job of pinging a user's interest and making it
obvious. Also, working to make that installation interface as intuitive
as possible - it's pretty good now, but could be improved - would be a
good idea.

3) System administration. This is a rather technical issue under Linux,
and is usually blown off with "oh, just go to the Ubuntu forums -
they'll explain everything!" THE AVERAGE USER IS NOT GOING TO DO THAT.
They'd rather go back to their comforting - albeit infected - Windows
world (spyware and viruses don't *usually* wipe out a computer these
days, and anti-virus software mitigates much of the pain anyway. Beware
the slippery road of "graceful degradation"...) I applaud Ubuntu's
effort in creating their "System->Preferences" and
"System->Administration" menus - but it's not enough. That UI - here's
that issue again! - needs a lot of improvement, and a lot more polish.
Perhaps giving it a look and feel that's similar to the Windows Control
Panel (which, in turn, looks similar to their file manager interface...
that's pretty good UI design, there) would be helpful.

4) As a competitive issue, there aren't enough Linux programs yet. Yeah,
sure, $BLAH thousands of programs... not nearly enough. Windows has
_millions_ of programs that have been written for it, filling nearly
every niche, including lots of free apps. Many of them are *really*
good. With Linux, once you get away from the mainstream, you're mostly
out of luck. E.g., as a sailor, if I want a good charting and navigation
program, I have... um... well, there's 7Seas ENC (proprietary app, very
rough UI, badly thought-out) and OpenCPN (http://www.opencpn.org/).
There used to be 'xmap' - proprietary and longer supported. That's about
it. If you compare them to one of the many, many available Windows apps
- "The Cap'n", "SeaClear", "Fugawi", the excellent "MaxSea", or the
highly-professional "Tsunami"... well, there's really no comparison.
Niche business apps, like - say - restaurant POS software, medical
practice or childcare center management? Not happening under Linux
(although some people are working at it, and might have a usable product
in the next few years.) Not to knock the efforts of the people who are
writing the Linux apps - I'm really, truly grateful for their efforts -
but there's just a much larger software universe for Windows, and Linux
hasn't caught up yet.  This one will take some years - and, in my
opinion, this is where the real battleground lies.

The average user knows nothing (and cares less) about licenses,
philosophies, and larger ideas/issues behind Linux; most of them have
"free" (illegal/stolen) copies of Windows, so cost isn't a real issue
for them. What they do care about is getting what they want, with
minimum effort and minimum time investment.  In most cases, for most
users, Linux supplies that at least as well as - and sometimes better
than - other OSes. In some situations, it's not quite there yet.


-- 
* Ben Okopnik * Editor-in-Chief, Linux Gazette * http://LinuxGazette.NET *