November 22, 2009 posted by Sarah Cockburn
In September this year, Guillaume Lasmayous spent 5 weeks in the US where he took the opportunity to meet with
some developers from the NetBSD project. On a Saturday afternoon Guillaume met Christos Zoulas to answer a few questions about NetBSD.
NetBSDfr: For those of our readers who don't know you,
could you tell us a little bit about yourself?
I was born and raised in Greece, and came to the United States to study
when I was 17. This is when I first touched a computer keyboard. Learned
Unix while an undergrad at Cornell on a VAX11/780 running 4.2BSD and
NetBSDfr: How long have you been using NetBSD? how did you discovered it?
What where you doing at that time?
I was working with CSRG re-writing the globbing code in csh so that
it would be free, fixing bugs in make, and writing libedit, when I
first saw the i386 code being committed by Bill Jolitz. I thought it
was really cool being able to run Unix on a small machine, and so
I was following both 386BSD and Linux. I did not like the code
quality in Linux back then and I was more familiar with Berkeley
Unix, so I had to choose one of the BSD's. Platform portability and
code quality is what made me choose NetBSD.
NetBSDfr: Which other UNIX or UNIX-like system did / do you use? In your
opinion, how do they compare to NetBSD?
I've used mostly everything. When I was at school there was a zoo
of computers, from VMS/Eunice, Domain/OS unix emulation, AUX, SVR2,
AIX/Locus, SunOS, HP/UX, 4.XBSD and the list goes on. Why do you
think tcsh is ported to so many platforms? I wanted to have a shell
that worked consistently on all the machines I needed to do work
on... So again, cross-platform portability. How do they compare?
None of them is perfect and all have their advantages and shortcomings.
There were some clear bottom feeders, but this was a long time ago...
NetBSDfr: When and how did you become a NetBSD committer?
I started writing the SVR4 emulation code and once I had it running,
Theo invited me to join the project as a developer.
NetBSDfr: How much time do you usually spend on the Project? How do you
manage to manage your role in the project and everyday life?
It varies a lot. I used to spend much more time in the past than
now. Now I fix the occasional bug, and port a driver/feature from
other BSD's here and there. Most of the work I do has to do with
dealing with NetBSD corporate paperwork and taxes. I would say
now I spend a few hours a week.
NetBSDfr: Can you explain to us what your role in core@ is?
The same as the other core members. We go over a list of items/projects
that need to get done, and try to nudge people to do them. This is specially
important during a release cycle. We also act as arbiters when there are
disputes between developers. For the most part we prefer that matters are
resolved in the mailing lists instead.
NetBSDfr: In 2006, you said in an interview
that NetBSD popularity was "stable or declining". After 5.0 release, has your opinion changed?
I think that 5.0 is a great release and I am planning to run it in my
production servers as soon as I get some new hardware. I don't know
about the popularity of NetBSD declining or not. I keep finding bits
and pieces of it in commercial products :-)
From your point of view, what are the
latest finest features of NetBSD?
All the new MP features, ZFS support, iSCSI initiator support (upcoming),
NetBSDfr: What were the last features you worked on?
The if_iwn.c driver. Importing openssl, openssh, cvs, file.
NetBSDfr: To you, what's still missing to the NetBSD Operating System?
A lot of things:
- NUMA support
- better net80211 support
- better internationalization support
- MP-safe networking stack
- A powerful interpreter in the base source
NetBSDfr: Where do you see NetBSD in the next 5 years?
I see it having many of the missing features implemented and ideally
as a reference implementation for Unix. I think that the clean header
files, the excellent documentation, and the well written implementations
of programs and subsystems will make more and more people wanting to
NetBSDfr: I know several developers are also working with other projects.
How are code imports from other projects discussed? Is this only based on individual
developer's decisions? Or is this a decision taken at core@ level?
For new packages there are discussions amongst developers. For existing
packages there is a list of maintainers and usually the maintainers of
a package decide if/when it is going to be upgraded.
NetBSDfr: We are starting to see the GSoC results on blog.NetBSD.org. How
are the GSoC subjects defined? Are they of the "targeted" projects by TNF?
Or some "nice to have"? Do all successful students get commit bit to the repo?
We have a list of projects in the Website, and people come up with their
own ideas. We rank them and then Google decides how many of those slots
we get. If they students get commit access is decided on an individual
basis. Usually they get commit access.
NetBSDfr: As a last word, what do you think makes NetBSD "special"?
I really like the people and the community around NetBSD. It is a group
of very friendly and smart people.
Many thanks to Guillaume Lasmayous from NetBSDfr for his hard work in preparing, conducting and translating this interview.
See the original post at www.NetBSDfr.org