DragonFly BSD
DragonFly users List (threaded) for 2004-10
[Date Prev][Date Next]  [Thread Prev][Thread Next]  [Date Index][Thread Index]

Re: Our fearless leader on video!


From: "Martin P. Hellwig" <mhellwig@xxxxxxxxx>
Date: Tue, 26 Oct 2004 14:31:59 +0200

Jeremy Messenger wrote:
<cut>> It will be nice if someone can quick write down a summary of it. Before
you ask me why, because I am deaf. ;-)

Cheers,
Mezz


Cheers,
Emiel



Sure, here is the exact thing as what I could hear, maybe somebody would be so nice to fill in the gaps and don't knows so that it can be extracted into an summary.


(0:00)
BSDmall.com & - Daemon News
Present - An Interview with
Matt Dillon - Creator of DragonFly BSD
(0:17) OK, I should talk about DragonFly and so
(0:20) I guess a little more background is required before I talk about DragonFly ehm
(0:25) FreeBSD is not the first operating System I worked on, I actually worked quite a bit on linux
(0:30) eh before FreeBSD and even eh during
(0:35) my Linux days I also wrote a number of eh embedded operating systems
(0:40) for work I was doing up <...can't understand part 0:43-0:43 ...> systems , so we had
(0:45) linux as a base station and later a Next machine as a base station and finally now
(0:50) eh a pared(?) of FreeBSD machine as a base station but the units on the field world,
(0:55) this embedded operating system ,ofcourse, the interface with linux or
(1:00) Next or FreeBSD we had to use open standard protocols so to talk to dp (?)
(1:05) ehm so the-the operating system, eh the
(1:10) c-concept of working on operating system is something that is really been embedded in my life
(1:15) eh for a very long time and FreeBSD 4 ehm
(1:20) my work on FreeBSD 4 was a wonderfull experience, despite having-to to, you know,
(1:25) to pull (laughs) to get things in, push-n-push-n-push-n-push
(1:30) ehm but with FreeBSD 5 they kind went of
(1:35) in a direction that eh didn't-wasn't really compatible with my own roots
(1:40) and own eh desire in terms of future direction of FreeBSD
(1:45) in the embedded world, the operating systems I wrote where very heavily threaded
(1:50) they where very heavily event orianted and messaged and I knew that
(1:55) that type of operating system could be written very efficiently to operate very efficiently I mean
(2:00) these embedded systems where running 10 Mhz cc (?) at thousands and there where doing alot of stuff
(2:05) eh so thats really, you know that-that-that
(2:10) division ehm his-what-ir-what caused me start
(2:15) contemplating during DragonFly and I didn't just announced it I actually started working on
(2:20) the light weight kernel thread scheduler almost immediately in January
(2:25) last year and I didn't announced eh DragonFly
(2:30) until June eh until June last year and I spent all that time really
(2:35) building the LWKT-stuff eh from a fort-audex(?)
(2:40) base and making sure that it did what I wanted and that it worked 'n that the project could be really announced
(2:45) and go for ehm and in fact thats what happend I
(2:50) threaded the interrupts I got L-ub-u-double-WKT working and announced DragonFly
(2:55) and got eh enough interest that I felt that it could be forwarderd as a project
(3:00) and-eh from then on it-it just been great and 'n pretty much eeh overloaded (laughs)
(3:05) ehm it-its I can still be programming I'm not that you know
(3:10) I'm not 100% manager now like eh like Linus Torvalds's aught to be
(3:15) ehm and its-beh-it's been a great ride and its continuing to be a great ride
(3:20) and we got lots of very interesting things planned for DragonFly you know-our our ultimate goal
(3:25) as asside but to get there eh we have to go a long eh parallel track
(3:30) with ef-eh what FreeBSD 5 is doing for quit a while ehm
(3:35) but-we-s but we will start to diverge I mean t-the basic things like getting rid of the big GIANT lock
(3:40) and eh our approach to that is very different from FreeBSD 5
(3:45) eh but ones we get passed that eh it gets far more interesting
(3:50) it gets into virtualised machines, virtualised resources ehm cache coherency
(3:55) across clusters and so it will be interresting <fade out> (4:00)


Special Thanks to Matt Dillon
George Rosamond and
bsdmall.com

--
mph



[Date Prev][Date Next]  [Thread Prev][Thread Next]  [Date Index][Thread Index]