[nycbug-talk] BSD in the enterprise....

[G.Rosamond]

On May 29, 2004, at 2:03 AM, Nigel Clarke wrote:

> What will it take to have *BSD in use in corporate environments? Linux
> was not successful until companies like Solaris and IBM started to
> endorse it. When I say successful, I'm speaking of making it into
> corporate environments.
>
> One of the other advantages that Linux offers is that it is economical.
> That and stable vendor support make it an easy decision for management.
>
> How does BSD compare? How will BSD make it into corporate environments?
>

First things first.

I think this thought about the corporate environment is incorrect in 
some ways.

The presence of BSD in the hosting and ISPs is enormous.  That may be 
the most challenging 'corporate environment' on a technical level.  I 
also think that the BSD presence in the world outside the US east coast 
is much more significant and recognized.

That seems like the 64k question, but I think the base reason for the 
'tardiness' is the ATT court case, which dragged the BSDs down from 
getting a quick start in the years before the dot com period.

What does Linux have?

First I don't think that the popularity of Linux is at the expense of 
the BSD's in general.  When companies migrate to free, open source 
software, it's in the favor of the BSD community.   Sure, some migrated 
*from* BSD to a Linux distro when BSDI, BSD/OS when south, but the 
anecdotes you hear seem to show there's some fluctuation between the 
BSDs and Linux.  I think that people have accepted FOSS in the 
corporate environments helps us all.  Just as when SCO attacked Linux, 
it was also an attack on the BSD community, even before we got a 
mention.

Linux has a real political appeal due to the GPL for a layer of people 
also, and that means an increase in advocacy.

That point is ever so clear on Slashdot, where lots of barely technical 
users troll the BSD stories as a hobby.  It seems to set a tone that 
few people openly take seriously, but many seem implicitly impacted by.

As everyone knows, the BSD community has never taken advocacy seriously 
as a whole.  This may or may not change, but the role of NYCBUG is 
certainly important in this equation.  What city is more important than 
NYC?  Maybe Rome and Athens in the classical period, London in the age 
of the British Empire, but that's the past.

The vendor question is also important, as you raised it.  Because of 
BSD licensing, there's no need for a vendor to advertise the code being 
at the core of their closed source and or embedded system.  Snap 
Appliances, for instance, broadcasts that one of its product lines runs 
on Linux.  But try to find on their www site that their other product 
line uses BSD.  I tried it once, but it was fruitless.

What are the strengths of BSD?  A long history.  Innovation in whole 
numbers of area, from TCP/IP to DNS and firewalling.  A level of 
obsessive seriousness among its developers.  The incorporation into OS 
X.  It's recognition as a serious project, something everyone notices.  
Tell someone you hack Linux, and the average tech thinks, cool.  Tell 
them you hack BSD, and you get immediate respect.

In comparison to Linux, the advantages are clear to me.  Stability, 
security and performance before bleeding edge, which certainly matters 
in a corporate environment.  The licensing is better for businesses 
(although I think it's better for developers even more).  Finally, when 
you use BSD, you use a total system, and probably some apps that are 
GPL.  When you use a Linux, you're using Linus' kernel, some 
distribution's userland and system, third party packages, and someone 
or another's drivers.  All along the way, you're banking on things to 
work like a nice mixed salad.  You just have a greater chance of 
someone throwing in a radish too many, and you'll have a hard time 
finding out who it was.

This became an issue with the SCO case, or whatever you call that 
stock-pumping fiasco.  Is there code that SCO claims to own in Linux?  
If so, where, and who put it there?  These are questions that are 
difficult to answer, since there's no real accountability in code 
contributions in the Linux model of development.

Open corporate backing is good, without a doubt.  Particularly if it's 
a big name (and I don't mean Wind River).  Apple isn't going to run BSD 
commercials like IBM has for Linux, but that's fine.

But corporate support is about technical support and development, as 
you stated.  *That* to me (and I know many others), is a major handicap 
for the growth of the BSDs.  As I've mentioned before, we are 
attempting to setup some basic support through BSDMall.  It will start 
out as M-F, 9-5 email and phone support, and hopefully develop from 
there.  BSDMall has the name recognition and credibility, and we think 
it can work.  And it will greatly enhance the prospects of the BSDs in 
the corporate environment if things develop.

I don't believe that the BSDs are some "red-haired step child" (no 
insult meant to all the red-haired step children on the list).  Things 
are moving along quite well on a development level, and I know that 
there's been a serious upswing in interest in the BSDs over the past 
year.  Ask Chris Coleman on the state of BSDMall comparing today to a 
year ago.

Woah, it's late. . .

g