April 1, 2005

One-on-one with Miguel de Icaza

Author: Joe Barr

One of the
highlights of my trip to Salt lake City last month for BrainShare
was the opportunity to interview Miguel de Icaza, a
mercurial star of the free software movement who has been responsible
for hugely successful projects and also founded his own company
with Nat Friedman that was later acquired by Novell. Like everyone
else at BrainShare, Miguel was on a tight schedule, but he took time
out from preparing a Friday morning keynote to do this interview.

one-on-one schedule with the press was slotted for talks about Mono,
but I asked for something different.

Can we talk about a couple of things other than Mono?

Yeah, sure. Go ahead.

You're one of the brightest stars in the free software world, so I
want to do a story about you. Besides, you've already been
Slashdotted once this week talking about
Mono. How about a brief outline of your history with free software,
and your career to date?

A friend of mine, we had a couple of Unix machines, and a friend of
mine recommended that I check the GNU stuff out because they had
Emacs and a free compiler that we could use on our school UNIX and
SPARC machines. This was in 1990, I believe. 1990, yeah.

That's when I
read about the GPL, also things like the Manifesto, so you start to
look at things differently. You begin with that, and then the
programs are really neat, because they give you the source code. But
then there are a lot of other benefits, and the thing passes by you.

How old were you then?

I guess i was 18, about to turn 18. I was 17. I was a student at the
National University of Mexico.

Weren't you a System Admin there?

Yes, but that was later. I was a Sys Admin for the Nuclear Science
Research Institute.

OK, that's how you got started. What was your first involvement with
an open source project?

I cannot remember. I think it was that I was contributing the
routines that ran in route, without any port of Wine, I think. Then I
worked on Midnight Commander.

That was your first big splash.

I think I was working on Wine, turning Wine into a library, back in
the day. No, it wasn't a big splash back in the day. It was just a
little tool that I wrote, that was it. It was a tiny tool.

Midnight Commander?

Yeah, but that was in 1992.

And it's still in Linux distributions today.

Yes, yes, I mean, yeah. I no longer maintain it, someone else does.
It's just that it's grown, and it has a community
of its own, and everything. But when I released it it was a tiny
little program.

When did you get started with Gnumeric? Was
that your next project?

Oh, no, no. Then I was working on Linux, on the Linux kernel. I
ported Linux to SPARC, with David Miller. Linux and the SGI, we wrote
a bunch of the device drivers for that. Linux RAID, so you could hot
plug a RAID card, worked on that with Ingo
Molnar and Gadi Oxman.

Click to enlarge
Miguel in meditative state; photo by Joe Barr

Who started GNOME?

I did, with Federico Mena. Federico was
already contributing to the GIMP, and I was busy with the Linux on
the SGI, and I was trying to get Federico to do it, and Federico
wouldn't do it. So then I said, I'll stop all the stuff on the SGI,
let's do this thing together. So we launched GNOME in August, 1997.
And it was the summer after, that I did Gnumeric.

OK, so GNOME has turned into a very big deal.


And then you and Nat started your own company?

Yes. In 1999.

Susan and I ran across you and Nat in the Computer Museum..

That's right, in San Jose. That's correct.

I think you were plotting the company then.

I don't think we had launched the company when we saw you there.
Because although the company was incorporated in October, we had been
talking about doing a company since April, so it was from April to
October, and we probably went to a ton of shows during that summer,
trying to raise funds, or hire people, or stuff like that. It was
during that period, so we probably met before the company was

OK, so your history is pretty well known from that time forward, but
I wanted to get your earlier history. So then you founded Helix, and
that became Ximian.


And then you adopted the Evolution project?

No, well, yes. Evolution was initially just a library to do email. It was written by Bertrand Guiheneuf,
he's now at Apple. Bertrand wrote that library, and then I sent an
email to the GNOME committee saying that we really needed a GNOME
mail program.

outlined what I wanted to have on that. It was initially called GNOME
Mail, and then Bertrand remained with Evolution. He wanted to call it
"E dash volution," or something
like that. I forget what happened. So he actually renamed it
Evolution. Bertrand was an employee of Helix for about the first
year. But he wanted to go back to France, so he left the company. I
think he finished his PhD, or something like that.

And Evolution, of course, has come so far as to -- if Andreas, my
peer from Austria (Ed note: not Germany, as was originally written) was right this morning in observing the demo in
the keynote-- has come so far as to have been copied in the new
GroupWise features.

Yeah, there is some cross pollination of ideas, but I don't think at
this point you can claim, really, that anybody can claim... They
probably came from Outlook, they came from some weird mail interface.
I mean, all of these things are incremental feature updates.

And now, Ximian has been purchased by Novell.
Given your philosophy, and your obvious care and concern about free
software, are you happy with your current situation with Novell?

Yes. I am still in charge. I still believe it is hard to be a pure
open source company, if you want to do software development. If you
want to do consulting, that's fine, but the kind of companies that
just produce software, as pure open source, is incredibly hard.

Ximian we tried initially, and we failed, so we did half open source,
half proprietary. And Novell is very much along those lines. So, I am
thinking here a little bit conservatively, and Novell still has big
lines of proprietary code, and we do have also a large investment in
open source.

the Ximian side of the house still continues to do a lot of software
development, so that works out. All the Linux group, KDE, the
Evolution group, continue to do open source. They open sourced the
Connector, that was proprietary in the Ximian days, Mono is open
source. all of the stuff done on the desktop is open source. In that
particular play, they are trying the services thing, so if you get
NLD, if you pay a service to get update for enterprise customers,
right? Or you can get the non-enterprise edition, the SUSE 9.x
releases. The enterprise ones are SLES 9 and NLD.

The KDE/GNOME rivalry. The competition has sometimes gotten pretty
nasty, with different people making comments on Slashdot,
that sort of thing. The community can get pretty rowdy. How do you
see that today when you -- representing GNOME and Ximian -- and SUSE
are all part of one big, happy, family now? Will there continue to be
both environments?

You have to ship both environments, because people might have already
applications built for one or the other. You have to ship both or you
will be at a disadvantage, particularly in markets where one gets
there before the other. So we have to ship both.

software development-wise, at least the Mono group and the Cambridge
group continue to develop exclusively using GNOME libraries, if only
because there is a long licensing discussion and it all came down to
using the same tools that our customers use without having to pay
extra money, so we continue to do that.

also does development but they mostly do development around the
kernel, the libraries, X server, but not much in the way of new
applications. The SUSE side of the house is pretty much a packaging

I heard yesterday, in an interview with Debra Anderson, the Novell
CIO -- we were discussing the Novell migration -- that people were
telling her that they couldn't migrate to Linux because they've got
all this dot net stuff. She said she tells them, "Well, Mono is
being ported to NetWare, so you can."
So Mono is serving as an exit strategy for Windows users to cross
over to Linux. And still, people have cursed you for making a "pro
Microsoft" tool like Mono.

Click to enlarge
'Bell Curve' sketch by Miguel de Icaza

(Miguel picks up my notebook and begins sketching in it.)

In every population, you have this behavior, where it doesn't matter
what you're plotting, but every time you deal with a population you
have this curve, right, the standard distribution.

The bell curve.

Right. So you always have two extremes, and it doesn't matter what
you ask them. It may be, for example, the height of a population. Or
it may be the preference towards salt or sugar. It doesn't matter,
you always have this tiny percentage who respond at each extreme. So
if you have a perfect curve, this would be 11 percent, this is 78
percent. So there are people who are very vocal in terms of not
liking Mono, because it comes from Microsoft, and anything from
Microsoft, they hate. I am not wasting my time with that 11 percent.

I just find it ironic that now Mono is going to be used as an exit

That was always the goal. It was always the goal to use it as a way
of migrating from Windows to Linux. The reason we wanted it more, was
because it was a better development platform for our own stuff. Like,
for example, NetWare 7.1 just announced, on Monday, it includes Mono.
A photo management application, built with Mono. Our Beagle desktop
search, built with Mono.


Beagle is built with Mono, yeah. Really nice. The new Internet
sharing, is built with Mono. Eugenia Loli
has about 45 applications running with Mono. The music player,
research into usability of music playing, built with Mono.

I overheard part of this story yesterday, but I want to hear more
about your vacation.

Beirut? I went there because I read this book from Robert Fisk, a
journalist, because I liked his articles. The kind of articles that
they didn't want to read in the U.S. after September 11th. So, I read
a bunch of his articles, and then one day, I noticed he actually had
a book, so I purchased the book. It's called "Pity The Nation,"
and it was fascinating.

a book that tells about the Lebanese civil war, from 1975 to roughly
1990. His friend was kidnapped, an American called Terry Anderson, I
think. So he was kidnapped, and this is the story of how he tried to
track him down, how he was kidnapped. He was always in the same city,
but he never knew where he was.

one of the stories, but as I was telling my friends, it's like the
1,001 Nights, that book. A really big book, has all these stories for
15 years of what happened there. It is very much like 1984. It has
all these alliances being cooked, and dismantled, and remade with a
different opponent. So at the beginning, the Syrians are the allies
of the Christians, and they come in to save the Christians. But then
the ultra-right wing don't like them, so they bring in the Israelis.
And the Israelis attack these guys, and the Syrians ally with the
Palestinians and the local Muslims. It's everyone against everyone.
So, anyway, it was a fascinating book.

was going to a conference in Istanbul, a Novell conference in
Istanbul. Since I was there I figured I would spend a couple of hours
in Beirut. So I went to Beirut, my wife came with me.

How long were you there?

For a week.

What I overheard was that you were there for both the pro- and the
anti-Syrian protests.

Yes, the first day I woke up there, I go to the lobby and everybody
is watching TV, and I go, "What is that?" When I was coming
to Beirut, I was reading in the newspaper that Hezbollah
had come close for a big rally, and so I didn't know how big it was
going to be.

could do something touristy, or I could go directly to the action. so
I went there. As it turns out, it was an eye opener because it was
not what you expect. What gets reported in the U.S. about the Syrian,
about the pro-Syrian rally, has not been true to reality.

wasn't... They show these pictures of all these outraged people,
right, all sorts of pro- Syrians. And then here are all these
beautiful women, and there were people yelling and screaming in both.
It was obvious they had their own agenda.

anyway, I also sent a letter to Robert Fisk, the author of the book,
and i said i was coming to Beirut, and he gave me a call when I was
in Istanbul. He said when I arrived in Beirut we could have dinner or
something. So we met him, we had lunch with him. He wined us up, then
sent us to the sobering refugee camps. They were showing me this
horrible place, this massacre place, and I'm trying to not be just
completely drunk, and ask some rational questions.

really really nice guy. My wife said, "Honey, seems like he
knows us all of our lives, he treats us so nicely."

stands, prepares to leave)

Miguel, I want to thank you very much for taking the time to speak
with us today. I know you have to get back to preparing for your

Click Here!