The following interview
with CSS inventor, Hakon Lie, took place on Wednesday, October
16th, 2002.
Eric Lerner: Do you have any favorite anecdotes
about the early days of the web?
Hakon Lie: I stumbled across it at an early
age, so to speak, in 1992. I worked in the cradle of the web.
I shared an office with Tim Berners-Lee at CERN, the laboratory
in Geneva. One interesting thing is that CERN was the birthplace
of the web, rather than the computer lab. It wasn't the computer
lab, or the places that were set up to do research on networks.
It was an obscure physics laboratory outside of Geneva.
Eric Lerner: Were you working with the particle
accelerator?
Hakon Lie: Exactly. We were looking for very
small things, and we had an accelerator beneath the ground.We
also had lots of computers to process all the data we got out
of the machines.
What struck me about CERN is the natural beauty of the surroundings.
It's incredible. You have Mont Blanc, the Alps, the vineyards,
and the pastures. It's really a beautiful place. But then you
get inside the CERN buildings. It's all of these very strict corridors
with doors lined up. Order prevails. I think that's a metaphor
for the web. This conflict between images and order that I often
see. HTML was conceived of as this very orderly language, but
people are using it to publish beautiful web designs and images.
That conflict plays out in the pages people publish today. But
that's not really an anecdote is it?
Eric Lerner: Actually, it seems to tie nicely
into some of the work you did with the CSS proposal.
Hakon Lie: That beauty got me thinking. We must
be able to combine beautiful designs with searchable text. Instead
of using images for example, to display textual messages, we designed
CSS so people could continue to send text across the web—but
to style it. So if you wanted 10pt Helvetica Red, you could specify
that in a language instead of making a picture of that string,
which is how people used to do that.
Eric Lerner: So, it sounds like you drew inspiration
from your natural surroundings. Were there other things that inspired
you?
Hakon Lie: Traditional typography, as in old
print-based typography-like typesetters, have been doing with
their lead based layouts for centuries. We tried to re-create
some of the beauty and feeling that goes in to a beautifully laid
out book. Aesthetically, I think the web is a nightmare, even
today. Perhaps it's slightly better than it used to be. You can
find sites that are aesthetically quite pleasing, but it's still
in the early days, and the aesthetics have room to improve.
Eric Lerner: Has CSS turned out to be what you
had hoped for when you were writing the original proposal?
Hakon Lie: It's probably not what I had hoped
for yet. The limitations are improving. It's now possible to design
a site entirely using CSS. If you've seen Wired
recently, they just switched to using CSS for layout a few days
ago. And the new Opera site, Opera.com
just did this yesterday. So, I think we're at a watershed with
regards to using CSS.
That's only a part of the solution. Lots of other things need
to happen as well. The people that set up the first websites were
not designers; they were scientists. Those who followed were hackers,
or the average homepage builders. They probably didn't have a
huge interest in aesthetics. They shouldn't have to. We should
be making the tools that allow them to create good looking pages.
Certainly, I think tools like Dreamweaver MX and Contribute are
very important in this regard. Together with CSS, I think we will
provide users with a solution. But aesthetically, I think the
web is a nightmare, even today.
Eric Lerner: If you look at HTML it almost seems
like, as a language, HTML doesn't have all the words you want
to use.
Hakon Lie: Exactly. This has to do with the
scientific heritage of HTML. In the corridors at CERN, the doors
are lined up. In HTML, you have the equivalent element with unordered
lists. You have your <UL>'s lined up, with items next to
them. The vocabulary declares rows of information, but not their
presentation. It wasn't meant for that.
At the same time as we started extending HTML, you got the <font>
tag and the <blink> tag—what I consider to be these
ugly add-ons. The table-based layouts people are using aren't
really intended to be used for layouts. It's certainly not the
right kind of tool. In my world, some of the motivation was to
create more of a beautiful web, but it was also to preserve the
structure of HTML. This allows people to continue to author structured
content that can be searched with Google or another search engine.
You shouldn't have to give up that structure to make it beautiful.
Eric Lerner: Are there certain areas you would
still like to see CSS tackle?
Hakon Lie: Along the way, we've had some requests
for changes—some of which we have been able to address. For
example, the hover effect allows you to change a link as the mouse
moves over it. This is something we didn't think about in the
beginning, but people liked it.
There are always more effects than people want. You can see some
of them in IE's implementation of CSS. They've added some things.
I understand the need for designers, who always want more effects.
But, I think it's very dangerous to let browser companies add
their own code for this. The style sheets you write will end up
non-interoperable between browsers and Operating Systems. CSS
was an attempt to standardize things. Not to give designers everything
they asked for, but to give them a subset that provided the rudimentary
tools, and was still implementable across browsers and platforms.
I was in great battles, first with Netscape and then with Microsoft,
for screwing up their implementations of CSS. That's one of the
reasons why, having worked for W3C for many years, I thought enough
is enough. We don't really need more sophistication, we need more
implementations of this. So, I joined Opera. We have, in a very
short time, been able to come up with a CSS implementation that
I believe is much better than both Netscape and Microsoft.
Eric Lerner: What web technologies do you see
as becoming most important over the next 5 years?
Hakon Lie: I think we'll see some development,
but all-in-all, I think the web is pretty much done. We've had
a decade of development now, and there are hundreds of millions
of browsers out there, which makes it tougher to innovate. In
the beginning, we had a few thousand people downloading nightly
builds of Mosaic. Back then, you could make changes pretty easily.
You could develop the protocols. It's much harder to do that now,
with all the infrastructure that's in place. CSS just barely made
it. The specification came out in 1996, and it's taken until now
that we're finally seeing sites switch to CSS based layout.
If you start now and develop something that depends upon browser
support, I think it will take even longer than 6 years. I think
many of the technologies that the W3C has been working on will
not make it into the web.
Eric Lerner: What about portable devices?
Hakon Lie: Many people have tried to get the
Internet on phones or other wireless devices. Here in Europe there
was a big push to do WAP. People were trying to convince the world
to use a new language, WML along with HTML to display content
on small screens. This failed, because content providers didn't
want to code several languages instead of one. Also, I guess they
were afraid that every new device would require it's new language.
They wanted to continue to use HTML.
Opera has taken on this challenge. How do we display HTML, as
it's being used on the web, on a very small screen? We think we
have the solution. We've found that we can reformat HTML pages
to display most of them on a small screen. We do some gentle massaging
of the HTML. We have to guess what is and isn't important. We
figure out which images are really needed, and which are there
for ornamental purposes. We download the images that are needed,
and ignore the ones that aren't needed. What you get on a cell
phone running Opera is much speedier pages than you would get
otherwise, but you still get access to the real web. I think people
are really becoming attached to the web. It has become a pillar
of information distribution, and you want to have access to your
daily newspaper, your dating service, and your bank—even
when you're not connected to the web. So, I think bringing information
to portable devices is very important, and Opera has made a significant
step in that direction.
Eric Lerner: Are there certain things that designers
can do to make Opera's job easier, as far as rendering pages on
small devices?
Hakon Lie: Yes. Designers should test their
pages. Opera 7 will have a hand held mode—a hand held simulator—so
that you can preview your pages to see how they would look on
a hand held device. Most pages should work without any changes.
Testing is the most important thing.
Eric Lerner: If you had to pick one historical
figure that you could have dinner with, who would it be?
Hakon Lie: Johann Sebastian Bach. I've been
listening to his works daily, and he's incredible. He's got a
very mathematical style of composing. How did he figure it out?
He didn't travel very much, but he somehow came up with this great
music.
Eric Lerner: Anything else you would like to
share with the world?
Hakon Lie: What started out as a volunteer project,
with open standards, where everyone could participate, is now
turning into...[what amounts to] a private mall. Since I plan
to spend the rest of my life on the web, I don't want to see that
happen. I think it's very important that people are conscious
when they use software. They should not just use what's on the
machine when they open the box. It's important that we maintain
the web as an open space where anybody can contribute. |
