[Cialug] Xen Source Acquired?

[Josh More]

Nathan,

Note: this comes across somewhat more pro-Novell than I intended. 
While I like Novell, the point of this is to discuss the "companies in
open source" issue.  Due to my company's relationship with Novell, I
have the greatest familiarity with them, so I used them in a lot of
examples.  Please feel free to discuss other companies as well.


I agree that Novell's past marketing hasn't been great, in that it's
been hard to see what they offer.  Citrix is similar.  However, I'd
caution you to against using historical marketing to reflect today's
businesses.  Novell has been marketing their Linux under their SUSE
brand, and it seems to me to be *very* well known in the market place. 
They've also just started (as of this week) pushing on their new
ZenWorks software that does Enterprise-level desktop/server management. 
To my eyes, Novell's current marketing is very straight-forward and easy
to understand.  (Their website, however, is not.  Guess we can't have it
all.)

Citrix isn't there yet.  Much as Novell took a few years to re-work
their branding, I think that Citrix needs to do the same.

To address your other comment about "strange niches" and "solve
problems you didn't realize were a problem", I see your point.  However,
I'd like to point out that many of us don't see these as problems
because both Novell and Citrix focus on problems that arise due to poor
scalability.

If you don't manage an environment with over 1000 users, then it's
likely that you won't see the true costs of password resets (Novell's
identity play), the costs of maintaining and updating patches and
polices (Novell's ZenWorks play), or the problems of remote access and
telecommuting (Citrix's remote desktop play).  Sure, we all see these
are problems, but in the small business world, they're minor annoyances.
 In larger organizations, these same problems can completely paralyze an
organization.  You often wind up with networks that have grown
organically, and the folks that built them are the only ones that can
support them.  When a company has their most senior people walking
around thousands of PCs with a CDRom, there are big problems.

Moving on...

I don't know much about Citrix's offerings right now.  I'm just at the
beginning of exploring their lines.  Thus far, I know a bit about their
Application Firewall Appliance (interesting idea, but I prefer AppArmor
on SUSE), and something about their acceleration technologies (cool, but
only useful if you have megabucks to throw around).  All I know about
their desktop streaming is that it competes with VMWare's desktop
streaming.  I don't understand either of those technologies very well. 
Perhaps someone can talk about that at next month's
virtualization-themed meeting (hint hint).

I do understand you desire to have open source projects stay "scrappy".
 However, there's another issue here.  If a project stays too small, the
developers will often tire of it and move one.  Planner is an excellent
example of this.  It could have been an MS Project killer, but
development is moving at a snail's pace (if you glued the snail to a
turtle doing the moonwalk).  Contrast this with a project like Compiz. 
Compiz was floundering as a project until Novell took it in house.  The
resulting controversy was painful, but we never would have gotten Beryl
or Compiz Fusion if a company hadn't expressed interest.

There are some projects out there that do just fine as community-only
(OpenSSH, for example) and others that do well as hybrids (OpenOffice). 
Other's are killed when a company backs away from them (Hula).  Yes,
some are killed when companies take them over, but some thrive.

I think that point is to look at the entity that is ultimately
responsible for the success of a project.  Based on history, I'd say
that IBM and Novell are great at being leaders of open source projects. 
In some ways, Novell leads a bit too much, because when they back out
projects sometimes die.  IBM is better at that.  Then there are the
companies that buy/acquire open source software and close the source
(Tenable), which often has a poor outcome for all involved.

The big questions that stand out in my mind are:

1) Will Citrix be a good steward of the Xen project?
2) If Citrix is not a good steward, will Novell and Red Hat play nice
and lead open source Xen?
3) Is VMWare in violation of the GPL?

I think that the only way to know is to give the companies time, watch
them like hawks (we're the hawks in this analogy (even though hawks
don't flock like open source folks to (sorry for the bad analogy (and
the nested parens)))), and scream if we see them doing something wrong.




 

-Josh More, RHCE, CISSP, NCLP, GIAC 
 morej at alliancetechnologies.net 
 515-245-7701



>>> "Nathan C. Smith" <nathan.smith at ipmvs.com> 08/15/07 3:49 PM >>> 

> 
> First of all, it's inaccurate to say that Citrix hasn't done
anything
> innovative in years.  It would be more accurate to say that Citrix
has
> been amazingly bad at talking about their recent innovations.
>
Do they share a marketing with Novell?

I will say one issue with their marketing is that their products tend
to fit
strange niches and often solve problems you didn't realize were a
problem
until you use and understand the solution- which may involve a
paradigm
shift.
 
> Among one of Citrix's innovations is having a completely separate
> desktop offering that is based on, but not directly tied to Windows.

> They offer this desktop over the wire to thin clients, Linux
machines,
> Macs, and other Windows workstations.  This puts them in direct
> competition to VMWare's streaming desktop offering.  The problem was
> that VMWare offered more than just desktops.

Did Citrix develop this or aquire it?  It sounds a little like the
"desktop
anywhere service" they aquired with GoToPC and now offer.
> 
> What will hit us is that Citrix needs Xen to support their latest
and
> greatest appliances, which could mean better cutting-edge hardware
> support in Xen.  Also, since XenSource, Novell, Red Hat (and 
> others) all
> contribute to the Xen open source project, it could mean that 
> XenSource
> now has more resources to put into it.  That could be good as well. 
> $500mil can go a long way to improving a project.
> 

I may be alone in this opinion, or lack some data points, but it seems
like
money and industry support does a good job of wrecking many good open
source
projects.  I prefer them when they are still scrappy.  The latest
developments with MySQL come immediately to mind.  OF course Digium
and
their hold on Asterisk via hardware.  I'm probably lacking some data
points
or a wider point of view so feel free to enlighten me.

Some companies (IBM?) seem to do a better job of nurturing and
cultivating
rather than acquiring and dominating the technology.

Feel free to educate me, my opinions are still forming.

-Nate
_______________________________________________
Cialug mailing list
Cialug at cialug.org
http://cialug.org/mailman/listinfo/cialug