Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 20 Apr 1998 19:54:12 -0400 (EDT)
From:      "Adrian T. Filipi-Martin" <atf3r@cs.virginia.edu>
To:        Irving Popovetsky <irvingp@puck.nether.net>
Cc:        freebsd-advocacy@FreeBSD.ORG
Subject:   Re: Screen Shot
Message-ID:  <Pine.SOL.3.96.980420194439.20973A-100000@mamba.cs.Virginia.EDU>
In-Reply-To: <Pine.LNX.3.96.980418153007.5116B-100000@puck.nether.net>

next in thread | previous in thread | raw e-mail | index | archive | help
On Sat, 18 Apr 1998, Irving Popovetsky wrote:

> Some of the most impressive screenshots that I have personally seen have
> been on the webpages of various Window Managers:
> 
> www.enlightenment.org  (enlightenment wm)
> www.afterstep.org  (afterstep)
> www.kde.org   (the K Desktop environment)
> 
> in their screenshots gallieries.  While most of these screenshots were
> taken on Linux boxen, its still just XFree86 that they are running, and I
> have been able to achieve similar on my FreeBSD box.
> 
> But some of those look real pretty, and make just as great of an argument
> for us as they do for the linux people, maybe even a better one :)

	Yep, they look great.  I just about wet my pants when I installed
enlightenment for the first time.

	The problem as I see it is that screen shos still do not show of
what makes FreeBSD a superior alternative in many cases.  As you
mentioned, most of the window manager screen shots were taken under linux.
Would they look different on a FreeBSD box.  I would hope not!

	So what can we show _in addition to_ the above?  I don't know how
to pictorally represent a Linux or NT box gobbling up ever CPU cycle in
sight while a FreeBSD box calmly services its clients.

	I guess I wih we had some "reputable" benchmarks on various
measures of system throughput.  i.e. compelling graphs.  Anyone read the
papers from Usenix two years ago in which FreeBSD, Linux and Solaris were
compared.  I thought it was a decent start.  There were deinite flaws in
the testing methodology, e.g. the NFS server was always a linux box, but
that could have been remedied by providing the bench markers with enough
hardware to do the tests properly.  I think they were doing a best effort,
if not a perfect one.

	Are there any people interested in publishing refereed benchmark
papers?  I find most people are compelled to look closer when you show a
graph where the competetors have context switch time grow exponentially
as a function of number of processes and FreeBSD has one that grows nearly
linearly.  This is a valuable difference worth paying money for on a busy
system.


	Adrian
--
adrian@virginia.edu        ---->>>>| If I were stranded on a desert island, and
System Administrator         --->>>| I could only have one OS for my computer,
Neurosurgical Visualization Lab ->>| it would be FreeBSD.  Think about it.....
http://www.nvl.virginia.edu/     ->|      http://www.freebsd.org/


To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-advocacy" in the body of the message



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.SOL.3.96.980420194439.20973A-100000>