Open Source Commentary from Navica's CEO,
Bernard Golden
December 2007
In This Issue
-
Open Source in 2008: Predictions
Open Source in 2008: Predictions?
This year has seen enormous progress for open source. Large
end users have woken up and recognized that open source will
be an important part of their future: they now understand
that it's going to be an equal partner to proprietary software
within their infrastructure and, more tellingly, they's integrating
open source into their management processes.
On the commercial open source front, two open source companies
(XenSource and Zimbra) achieved gaudy liquidity events via
acquisition; the former by software vendor Citrix, the latter
by Internet pioneer Yahoo! All is not perfect on the commercial
open source front, though – more on that further into
the newsletter.
Proprietary vendors appear to be coming to an uneasy détente
with open source, recognizing that it is an unstoppable force.
Their main strategy appears to be grudging resistance, accepting
that customer demand makes open source inevitable but digging
their heels in all the way. By this strategy they seem to
be hoping that, while customers will move to open source,
their resistance will dissuade the open source choice in that
particular situation. This seems somewhat analogous to the
high profile legal strategy the record companies are pursuing
-- a holding action designed to stall the inevitable retreat.
However, we’re on the verge of a New Year and the approach
of a fresh start inevitably calls out for a set of predictions
about what 2008 will bring for open source.
Prediction #1: Open source continues marching into
IT organizations
I got an interesting comment back on last month’s newsletter,
in which I predicted a future of pools of proprietary software
surrounded by oceans of open source. One well-known open source
industry celebrity wrote to (mostly) compliment the newsletter,
but noted that he disagreed with this particular prediction.
I suppose he feels that most IT organizations will continue
to have mostly proprietary infrastructures with open source
sprinkled in for tactical uses.
Short-term he’s right. Long term he’s dead wrong.
While the sticking power of technology – particularly
software – should not be overlooked, all of us vastly
underestimate the long-term impact of free software. Without
beating the elasticity horse to death, it’s hard for
us – particularly hard for those of us who work in the
industry and are most knowledgeable about it – to understand
the revolution open source will bring to IT infrastructure.
As an aside, New Year also calls for resolution; one of mine
is to reduce my continuous trumpeting of elasticity –
not because it’s not true, but because I don’t
want to end up a one-note Johnny. Maybe I’ll adopt Alan
Greenspan’s notion of using a meaningless euphemism
for a term; in his case he proposed using the term banana
instead of recession; so if you see me referring to banana
in a newsletter in the future, substitute the word elasticity!
Furthermore, the deluge of open source will have a complementary
effect of driving the price of proprietary software toward
zero, which I applaud. I am not an open source zealot; I am
a software zealot, seeing it as a force reshaping the lives
we live.
As to why we don’t see more recognition of this impact,
it’s because most of the zeitgeist of the technology
industry is vendor-focused, and, as previously noted, the
proprietary vendors are clutching onto their existing marketplace
position for dear life – and part of their strategy
is to occlude and discredit the progress of open source.
Added to this the fact that the opinion makers – the
trade press and the analysts – are wedded to the current
winners both by sloth and by pay packet, and it’s no
surprise that little coverage is given to this profound trend.
Anyone who has ever attended a press briefing has seen the
dance: the vendor shows slides presenting their pitch, the
media dutifully takes notes that will be turned into stories
that reflect the vendor’s slant on things, all orchestrated
by a PR firm that urges the writers to give the best spin
on the story and favorably parrot the vendor’s message.
The analysts, of course, do a higher-brow version of the
same dance, being courted by vendors in elaborate “consultations,”
thoughtfully issuing market assessments and “magic quadrant”
listings for technology segments.
Inevitably, all of this activity is focused on today’s
winners and presents what John Kenneth Galbraith called “the
conventional wisdom.” No one is there to speak up for
the revolution, because the money all lies with the other
point of view. Nevertheless, open source is crashing into
IT infrastructures, despite what everyone says. As Winston
Churchill put it “A lie gets halfway around the world
before the truth has a chance to get its pants on.”
(Incidentally, one thing I won’t downplay over the next
year is my reliance on this lion of the English language;
Churchill has a quote appropriate to each occasion and every
situation).
The commercial open source companies don’t do themselves
any favors in this regard, since they seem to be reluctant
to discuss how much of their software is actually being used,
preferring to disclose customer numbers instead. In part,
this reflects a belief that presenting usage numbers would
be seen as a weakness, since there is far more usage than
paid installations. My view is completely different. While
generating paid use is a primary issue for commercial open
source companies, adoption is profoundly important and should
be trumpeted proudly. Software is a digital good, and open
source has the characteristics of an anti-tragedy of the commons:
the more people who use it, the greater the benefit, and the
more likely a company can prosper from a given open source
product.
Prediction #2: Not all Dogs Go to Heaven
I think I saw a prediction by Raven Zachary, well-known open
source analyst at 451 Group, that 2008 will see one or more
commercial open source companies go bust, and I think he may
be right (if he didn’t predict it, I’ll stand
out on a limb alone with the prediction).
Successfully running a commercial open source company is
difficult, and the mere fact that a company’s open source
product is widely adopted is not enough to guarantee commercial
success. The ability to entice users to pay is inextricably
intertwined with their perception of risk associated with
using the product, and if your product doesn’t qualify
as requiring risk-aversive behavior (i.e., payment to implement
risk avoidance measures like commercial support and indemnification),
your ability to generate revenues will be severely affected.
A number of commercial open source companies have been funded
on the basis that their products are or will be widely used.
This year will see that open source products and companies
are not Siamese twins, one unable to survive without the other,
but two different beings that may rely upon one another but
must be able to stand on their own.
It’s not all bad news for open source companies, however.
Many of these financial problems are associated with their
mode of capitalization – venture capital wants rocket
returns and the steep climb of revenues that supports outsize
returns. Just because an open source company is unable to
satisfy its investors doesn’t mean the company can’t
be successful, just that its success is inadequate for venture
capital expectations.
Any commercial open source company that goes bust will –
if it has any commercial traction at all – be restarted
by its founders the day after it closes its doors, much like
the rug merchants that are always “Going out of business
tonight” but somehow magically appear the next day with
the same rugs and staff but with a new name. The license for
open source products requires the source code to be available,
which means a failed company’s founders can pick right
back up where they were after a business failure. Just because
a company fails doesn’t mean the product has, and open
source means that successful products will generate some level
of commercial success, even if not a venture capital level
of success.
Prediction #3: Open Source Shifts Industry Value
and Power
I spoke a couple of weeks ago at the AFEI (www.afei.gov)
Open Source Conference (they actually called it an Open Standards
Conference, but everyone referred to it as an open source
conference and, if you look at their website, they even use
that term to refer to it).
One of the keynote speakers was Andre Boisvert, a longtime
tech exec and an early stage investor in open source companies
like Palamida and Pentaho, for which he serves as Chairman.
His presentation was very interesting, particularly as he
outlined his forecast for the IT industry: a massive shift
of power and revenue from software vendors to end users and
system integrators. With less revenue flowing to vendors,
the shift of power to end users is obvious. The prediction
regarding the future role of system integrators is less obvious,
however.
It makes sense that lower-cost software would free up more
money for services. It also makes sense that source code availability
would offer more opportunity for services used for customization.
From a politically incorrect perspective, it even makes sense
that services would be required to “civilize”
open source, since it often seems to be left in a semi-finished
state instead of the nicely polished form that we would prefer
and that proprietary software putatively delivers.
However, it must be said that the “name” system
integrators have not embraced open source to this point. It’s
no accident that the two who have – CapGemini and Unisys
– are the two weakest companies (and it’s no coincidence
that this week’s rumor is that Wipro is about to pounce
on CapGemini); the weakest players in a market always have
to be more innovative, since they’re locked out from
the biggest current opportunities, which are dominated by
the most successful competitors. This is somewhat analogous
to the just-about-to-go-out-of-business Studebaker releasing
the Avanti, a sports car far better than the overchromed barges
put out by the big three: they didn’t have to do anything
different (until the Japanese arrived, that is), but Studebaker
was desperate, so it was forced to turn to innovation.
The successful integrators have only paid lip service to
open source. They have a partner or two who is their “open
source global lead,” but in reality, he (or she) is
trotted out to conferences and customers who insist on talking
about open source, but, other than these ceremonial occasions,
is kept safely locked up in the back room.
The reasons the big integrators haven’t pursued open
source are primarily due to their commitment to their current
way of doing business:
- Big engagements on top of large software license sales
– if a customer buys $5 million of SAP, it probably
will swallow $10 million in integration fees. The big integrators
are afraid that the $10 million will shrink if the customer
has just bought $200K of Compiere. And they might be right.
- Cultural fit with the big software companies. Both integrators
and software companies do business the large way: direct
influence with the CEO and board, lavish dinners at trophy
restaurants to discuss “vision,” golf outings
at exotic resorts, and even far less savory offerings. By
contrast, open source companies offer a choice of tuna fish
or turkey at their Office Depot-purchased conference room
tables. If you were an integrator, which would you prefer?
- Channel conflict with current partners. If you’re
an integrator considering building an open source practice
focused on Pentaho when you’re doing hundreds of millions
per year with Oracle, aren’t you going to be more
concerned about Oracle than about potential customers for
Pentaho-based engagements? More germane, if you’re
an integrator doing significant Microsoft business and you
even begin to daydream about getting into open source, you’ll
soon have a call from Redmond advising you to get back to
reality.
Of course, this doesn’t mean there aren’t smaller
integrators out there who have embraced open source offerings.
Thus far, the integrator pool has tended to be mom-and-pop
open source enthusiasts as well as regional integrators with
specific technology focus (e.g., European business intelligence
integrators or Websphere-to-JBoss migration specialists).
I believe the regional integrators do hold promise for providing
open source services and they will be the beneficiaries of
the move to open source. In the future, when open source has
grown to the point where the big integrators can’t afford
to ignore it, despite the remonstrations of their proprietary
partners, these regional firms will be bolt-on acquisition
targets for the global integrators.
Boisvert’s overall theme is absolutely correct. Open
source is shaking up the power structure of the IT industry,
with most of the benefit redounding to users (after all, users
have much more power over service providers than software
vendors, so a shift of the industry revenue away from software
companies to integrators increases the influence of users).
Incidentally, Boisvert also described sales practices at
Pentaho. The company receives approximately 3500 sales leads
per months and turns them over to a team of about 5 telesales
reps. According to him, each rep spends no more than 30 seconds
on each lead; if it doesn’t look immediately promising,
it’s sent to the bit bucket. It struck me that this
sales model is destined for a head-on collision with standard
IT practices that are built upon the lengthy RFP, vendor-funded
proof of concept, and expensive sales rep personal visit mode
of procurement. Boisvert expects integrators to pick up the
pre-sales effort, but I remain unconvinced about this. Most
integrators focus on selling their services after the software
selection is complete; put another way, people generally focus
on selling their own stuff, not someone else’s.
Prediction #4: ODF Waxes OOXML
After the smoke-filled room debacle that was the initial
effort by Microsoft to ram its OOXML “standard”
through ISO, all eyes have shifted to the review meeting in
February where all the technical comments on OOXML will be
debated. Many people expect a successful effort by Microsoft
to pressure the review committee to accept its XML file format.
I think Microsoft is going to fall short at that meeting;
furthermore, even if their effort succeeds, it will be a Pyrrhic
victory. It is clear that a number of nations are moving to
ODF (Norway being the most recent to choose ODF as its document
storage format). These nations have a visceral dislike of
having their technology choices dictated to them and, for
sovereignty reasons alone, are selecting ODF. When a developed
nation like Norway standardizes on ODF, one has to wonder
if the United States will end up isolated on an island of
OOXML, much like it remains in the minority regarding metric
standards.
When such a large percentage of the world (which represents
the majority of the company’s growth potential) selects
a different standard, Microsoft will inevitably decide to
embrace ODF; after all, Office can support ODF and the company
can’t afford to lose that revenue stream in the forlorn
hope that everyone will join with it and standardize on OOXML.
Microsoft will embrace ODF for the simple reason that it can’t
afford not to.
Speaking of which:
Predicton #5: Microsoft Buddies Up to Open Source
I worked at the local telco (Pacific Bell) at the time of
the Bell System breakup. Early in the process, AT&T was
viewed as morphing from nurturing parent to detested competitor;
however, that tune changed later in the process, when PacBell
realized that an enormous amount of its revenue stream was
going to come from AT&T in the form of traffic termination
fees. AT&T rapidly became a valued business partner, because
PacBell needed the money.
We’ll see the same kind of process at work with Microsoft
in 2008. We’ve already had hints of it: Microsoft invited
the Firefox team to come and test with Vista, not because
it wanted people to choose Firefox over IE, but because a
significant part of the potential Vista market had chosen
Firefox, and if Microsoft didn’t ensure that Firefox
ran correctly on Vista, it was reducing its total potential
market. Microsoft couldn’t afford to ignore the potential
revenue base of Firefox users.
There have been other examples as well: Microsoft has focused
on ensuring SugarCRM and PHP run well on Microsoft Server.
Again, this is not because of love for those products, but
because customers who had already decided they were going
to use those products would have chosen a platform other than
Windows if Microsoft didn’t make sure Windows could
support them.
Heretofore, Microsoft has had the luxury of presenting a
pretty united front against open source: every group within
it has chanted the company line about open source not being
as safe/reliable/secure, etc.
That’s going to change. As open source represents more
of the IT user application base, Microsoft will have to embrace
it, even if it causes heartburn for the analogous Microsoft
applications. Microsoft won’t have a choice. This will
play out as tension between the different product groups,
executive jockeying, and attempts to sabotage open source-oriented
initiatives. It won’t be a smooth process, either. We’ll
hear contradictory statements from the company as individual
initiatives progress internally. But make no mistake about
it, Microsoft will exit 2008 much more open source-friendly,
because it can’t afford not to.
There you are. Five brave (if unproven) open source predictions
for 2008. Even if they turn out to be wrong, well, I'll chalk
it up to the world's shortcomings rather than a lack of prescience.
Navica News
Reaction to last month's newsletter
Last month’s newsletter discussing whether commercial
open source companies can scale generated quite a bit of attention.
Shaun Connolly of Red Hat blogged
about it, affirming his support of my perspective (which one
would expect, given that he works for a company pursuing a
support/services business model). Bill Snyder, erstwhile software
journalist at thestreet.com, discussed it in an InfoWorld
blog posting that was generally positive (I wanted to post
a link, but I can't seem to find the piece. I also received
a number of emails about the newsletter as well. The discussion
appears to be ongoing, but nothing has changed my mind: service
businesses can scale quite nicely, thank you, even if they
don't match the scale of the ideal proprietary software model.
You can hear me speak at these upcoming events:
I’ve been invited to speak at Sun about open source
(would have thought that was a bit like carrying coals to
Newcastle, but there you are). I had dinner with Simon Phipps,
open source main man at Sun, a couple of months ago and enjoyed
our conversation very much. Say what you will about Sun (and
I certainly have), of all the major technology vendors it
is the one least conflicted about boldly pursuing an open
source strategy. I don’t have the details about the
presentation date and time yet, but am looking forward to
it.
If you are interested in having me speak at your
organization:
Contact me directly via email.
Other publication news:
I’ve been invited to contribute an article on open
source governance to an upcoming DoD (Department of Defense)
publication on open technology. The AFEI conference I attended
demonstrated significant interest in using open source in
DoD projects, and this publication carries far more influence
than its relatively modest size might indicate. Being the
military, official approval counts for a lot, so the publication
will help legitimize DoD use of open source. I’m honored
to be invited to contribute.
New Book Availability
Finally, I’m happy to share the news that Virtualization
for Dummies is finally out and seems to be doing pretty well
– two copies were even sold on Amazon on Christmas Day,
which may say more about the book’s audience than the
book itself. A German edition will be released early next
year, as well as custom editions for Stratus, HP, and AMD/Sun
(a joint project).
|
