IT Investment Research

The EU's "Profound Misunderstanding:" MySQL Dominates Database Market

According to a Bloomberg story released November 21 concering the European Union (EU) statement of objections against the Sun (JAVA)/Oracle (ORCL) merger, it's hard to see why the homeboy EU open source project MariaDB, a fork of MySQL, needs the EU's statist protection. And it's clear why Oracle says the EU Competition Commission has a

"profound misundertanding of both the database marketing and open source dynamics."

The open source thing is a side show designed to get the open source blogblatherers up in arms but it is clear, if the Bloomberg story is accurate, that the EU really misunderstands enterprise software market dynamics. Apparently—according to Bloomberg’s reading of what it says is a 155-page document—the Eurocrats have decided that MySQL is the fourth leading database in the world. Also according to Bloomberg, the EU finding is based on the number of daily downloads of MySQL.

So now we know how the EU analyzes competition in the enterprise software market.  It’s known by software market analysts as the Compuserve or AOL method.  That is, it proposes that one size a software market not by the number of productive instances of a piece of software or by the revenue generated by that software but by the number of annoying free floppy diskettes sent out to consumers by snail mail every month. That is the way AOL and Compuserve executed try-it-you’ll-like-it marketing before Internet downloads became the norm. Billions of the diskettes of course immediately filled up landfills pre recycling days. Most never made it into a floppy drive never mind doing anything useful.

By the EU Competition Commission’s logic, MySQL must be the leading database in the world, not number four.  At the reported 60,000 downloads a day, there must be more than 20,000,000 copies in productive use just from downloads in 2009.  That’s something like five for every database administrator in the world.  Figure that MySQL has been around for 10 years or so and that means 200 million instances are running. Figure that no reasonable program development person would download a copy every time he or she began a new application development project (that is, they would use the copy they previously downloaded) and you have to assume, using EU logic, that MySQL is in use in billions of productive situations.

And yet according to IDC, according to the Bloomberg article reporting on the leaked EU statement of objections document, Sun only realized $40 million in worldwide software revenue last year from MySQL. Typically IDC software market statistics do not include professional services revenues—such as for training and implementation—but just license and subscription maintenance revenue. I do not know the methodology or source that IDC used in the statistic quoted by Bloomberg. Another widely quoted source in last few weeks said Sun netted $300 million from the first full year of marketing MySQL. MySQL executives were widely quoted as saying the company did between $50 and $100 million in revenue in the last full year before it was acquired by Sun.

So why does MariaDB need the protection from the EU that its public policy expert is lobbying for?  All it would seem Maria needs to do is peel off 10% of the vast MySQL installed base and it would have a multi-hundred-million dollar software and support business. Or maybe what is really happening is that the Maria guys know--because they were the MySQL guys until recently--that the project actually has very little business viability.  And they have to stop the train for a decade or so until Maria might theoretically catch up to Oracle database functionality and popularity.

That's all I can figure from this EU-bureacrat-full-employment extravaganza. 

Oracle, do the right thing by shareholders (yours and Sun's). Get out of the EU.

-- Dennis Byron

Are EU Objections to Oracle/Sun Really About Maria?

In a November 16 blogpost, despite the cursory view of some analysts who say withdrawing from the European Union (EU) enterprise-software market is not feasible for Oracle (ORCL), I wrote that Oracle's withdrawal was not only feasible but something Oracle’s board is really obligated to consider. Microsoft (MSFT) in fact mentions the option specifically in its SEC filings in discussing its always contentious dealings with the Eurocrats. The catalyst for the blog posts is the EU Competition Commission’s statement of objections against the Oracle acquisition of Sun (JAVA), objections reportedly based on Sun ownership of the MySQL relational database product. (I say ‘reportedly’ because the EU, with all its fondness for openness, does not openly release its statements of objections.)

It’s interesting the way the debate spins out from that point.  Over on Information Week, Bob Evans has already written Larry Ellison’s withdrawal letter to its EU customers, calling European information technology (IT) professionals to the barricades against Neelie the Sixteenth (or Neelie Antoinette, whichever you prefer). At InfoWorld, Bill Snyder talks to a lot of open source gurus and finds that no one except the MariaDB MySQL fork guys stands to gain.

When Bill called me about my November 16 post, he asked,

“But why would Oracle want to give up a four-to-six-billion-dollar annual revenue stream for the trivial revenue stream MySQL generates?”

It’s a good question but my answer to Bill is

“It doesn’t matter to me why Oracle wants MySQL” 

The issue is that the EU shouldn’t be able to tilt the playing field in favor of the EU homeboys trying to start the company that Snyder references. The EU has similarly tried to tilt the browser market playing field in favor of some Norwegian whiners with their free-market-loser product called Opera.

This distortion of market dynamics by politicians is bad for EU IT consumers and it is bad for enterprise software suppliers in general. And its probably even bad for the EU. Oracle’s opening position at the meeting on November 25, if it hasn’t told the EU this already, should be: 

"Approve the Sun deal or the new Snoracle does not do business in the EU." 

Close down that factory in Scotland and the warehouse in the Netherlands. Send your citizens over the border into the Ukraine to get new software or to the UAE for training. (I assume they are already getting their software maintenance calls answered from India.) Take out maybe 20,000 EU jobs. At such a high percentage of the total workforce as compared to percentage of the revenue stream the EU will contribute to Snoracle, it would be well worth it even with onerous severance payments. 

And Oracle shareholders will also save the multi-billion-dollar payments that the EU weaseled out of Steve Ballmer.

And then as Bob Evans points out, listen to the EU citizens scream at its Competition Commission.

-- Dennis Byron

Pulling Out of Europe Not Only Feasible for Oracle; Maybe It's a Board Imperative!

Over on Internet News, Andy Patrizio mentioned my November 13 post about Oracle (ORCL) dumping the European Union (EU) rather than MySQL and then tried to come up with some reasons to shoot down the idea.  I have no problem with that. The idea is fair game for argument, rebuttal or even disdain.

But when Patrizio says “It just isn’t feasible,” that’s not the case. It might not be good idea, it might be hard to execute, it might cause problems for customers (but see below for how easily they are handled), it might be more trouble to manage than it is worth (such that the proposed savings do not materialize), etc. But it is definitely feasible. 

In fact, in some cases it is an imperative. Public companies must either implicitly or explicitly take the following position quoted on page 14 of the 2009 Microsoft (MSFT) 10-K:

“We may have to choose between withdrawing products from certain geographies and…”

Boards of directors are obliged by fiduciary responsibility to constantly ask themselves whether they should do it. Is Oracle's board currently so obligated?

As for feasibility, don’t fall into Patrizio’s trap of equating Europe with the European Union. Not all European countries belong to the EU. Perhaps the EU might tell its citizens that they can’t go to (or call) Switzerland or Croatia (or India or Belarus) to get some software fixed.  But I would guess that would be as hard a task as the state of Massachusetts has trying to keep its residents from buying goods over the border in sales-tax-free New Hampshire

As an aside, I did not mean to imply that “Europe's biggest nations (England, France and Germany)” share of Oracle’s revenue is “shrinking because Europe's economy is in worse shape than America.”  I don’t think anything I said even hinted at that. The economies of those countries are shrinking because the countries’ populations are shrinking or growing very slowly. It is a long-term issue and not simply an outgrowth of the Financial Panic of 2008.  

-- Dennis Byron

The Numbers Explain Why Oracle Should Abandon EU

I said in my post on November 12 Larry Ellison should ask himself whether it is even worthwhile for the new Oracle (ORCL)/Sun (JAVA) combination to do business in the information technology (IT) backwater that is the European Union (EU). Why go through the hassle of dealing with EU regulators when it appears the return on investment is so low? My suggestion is to just consummate the merger without EU approval and don’t do business in the EU.

A bunch of Euroblatherers on seekingalpha.com commented that my idea was anti-Europe bashing. One even criticized the quality of my photograph, which I love because it is a snapshot taken by friend in front of an old restaurant on Helsinki harbor. I love to go to places like that in Europe and look at the old cathedrals and decrepit palaces.  It’s just that the over-regulated EU is an increasingly bad place to sell enterprise software and IT.  Look at the numbers.
 
According to its 10K, Oracle received less than 15% of its revenue from the EU big three—the UK, France and Germany—during fiscal 2009.  And that’s down from 16% in fiscal 2008. Given those countries’ contracting economies that number is bound to keep dropping. Because the revenue is based on GAAP accounting, it would take some work to figure the real compare (e.g., would need to factor in the percent of business BEA and other acquisitions did in the EU) but maybe—best case and counting Slovenia and Wexford and all those other great bastions of industry—Oracle gets 20% or 25% of its revenues from the IT consumers regulated by the non-tech-savvy bureaucrats in Brussels.

According to its 10-K, Sun (JAVA) on the other hand gets almost as much revenue from Europe (that's not the same as EU but that’s the best number we have) as it does from the U.S. And a third of Sun’s U.S. revenue comes from Avnet. In fact, while Sun revenue in the U.S. dropped 25% in the last few years, it only dropped 15% in Europe. The trendline of Sun’s EU numbers vs. the U.S. numbers should forever be considered a good leading indicator of failure in the IT market for all companies. Clearly when everyone in the U.S. is bailing out of a company’s IT products and services as with Sun, but the IT consumers of the EU haven’t got the word, sell your shares. 

As with the Oracle numbers, there might be some falling/rising/falling dollar issue in those statistics. But I bet it wouldn’t take a lot of analysis to figure out that that revenue—whatever the apples-apples revenue numbers are—is no where near as profitable as doing business in real markets where politicians don’t try to regulate the shape of fruit. For example it looks like half the employment of the combined company would be European even though only 25% of the revenue comes from there.  If Snoracle gets out of the EU, that salary expense drop to the bottom line.

-- Dennis Byron

Memo to Oracle: With Sun Comes Widgets

With Oracle (ORCL) going to the mattresses with the European Union (EU) over its proposed Sun (JAVA) acquisition, many analysts are suggesting Oracle walk away from Sun.  I am suggesting Oracle walk away from the EU. Larry Ellison should ask himself whether it is even worthwhile for Snoracle to do business in such a backwater.

The EU is the place that 50 years after Plan Calcul is still trying to launch a nativist information technology (IT) company to compete with IBM. Larry, ask Neelie about Tulip.  Remember Cii and that crazy Siemens, ICL, Nixdorf, Bull, Olivetti thing in the mid 70s. No prolonged effort to get approval of the Sun acquisition is going to change that basic dynamic. No three-year effort to get to the Court of First Instance of Last Resort in Strasbourg is going to be worth spending all that time walking around in Stalinist architecture during days of little sunshine (the EU does not work in the summer and it rains all the time there in the spring).

So why go through the hassle of dealing with EU regulators when it appears the return on investment is so low? That’s especially true if Oracle is serious about keeping and growing Sun’s systems/storage business.  Software—the market that Oracle knows well—is portable. But message to Redwood Shores:

"With Sun, comes widgets."

So Larry this should be the opening position at the meeting on November 25, if you haven’t told the EU already: 

"Approve the Sun deal or the new Snoracle does not do business in the EU."

(Update on my post of November 10 about not being able to analyze the EU’s statement of objections, I received the following from the EU’s crack PR folks:  “We regret to inform you that Statements of Objections are not documents of public domain. They are sent to the parties concerned in the transaction.”  So I’ll have to wait until later this week when SAP leaks it to the Wall St. Journal.)

-- Dennis Byron

European Union Database Market Takes on Oracle: They Call the Wind 'Maria'

This is the battle I have been waiting for (at least until November 15 when New England takes on Indy).

I must admit part of my lust is just for the sport of it… to see Ellison do what Ballmer should have done. But it’s deeper than that. The statement of objections reportedly filed by the EU Competition Commissioner against the Oracle (ORCL)/Sun (JAVA) merger pits the statism of the 50-year European Union (EU) attempt to nationalize industries and markets, including a score of failed attempts to spawn an EU information technology (IT) industry and various EU IT markets, against a future world order totally tilted to the western part of North America and the rest of the Pacific Rim in all things economic and political, not just IT.

I’d really like to dig deeper into it but I cannot find the actual statement of objections being raised by the EU.  This makes it hard to comment but the paragraphs below summarize what Oracle says in rebuttal.  I am clearly not going to be disappointed because Oracle’s bombast is as usual seriously misleading and yet cuts to the quick. For example:

"Oracle's acquisition of Sun is essential for competition in the high end server market, for revitalizing Sparc and Solaris… "

Well that’s clearly not true because Sun has been a nonentity in the server market for some time so there is no essential competition.  I guess you could say it might be important to “revitalizing Sparc and Solaris……and for strengthening the Java development platform” but the IT market has already said it doesn’t care.

“The transaction does not threaten to reduce competition in the slightest, including in the database market."

This is apparently the heart of the matter and is true because no merger or acquisition reduces competition; bad products decrease competition. As usual, a bunch of European guys with an apparently bad database management product called Maria (forked out of MySQL but see paragraphy below)--unable to compete on a level playing field, the worldwide IT market--are trying to tilt the EU database software playing field in their favor.

“It is well understood by those knowledgeable about open source software that because MySQL is open source, it cannot be controlled by anyone. That is the whole point of open source."

This is a side issue, or at least it should be, but the statement is true.  Apparently the EU marches on because of the EU losers who were not able to take over the world with their low-functionality EU database software.  It’s the same meme as the loser Norwegians that screwed up Windows 7 because no one used their Opera browser.

“Oracle and MySQL are very different database products.

Well that’s not true.  Why would these EU losers who are marketing Maria start this battle if it was a different product than Oracle's database software.  The whiners know that it’s just that Maria is 30 years behind Oracle’s relational database research and development.  In theory, Maria could catch up if the EU tilts the playing field enough.

“The U.S. Department of Justice carefully reviewed the proposed acquisition during the normal Hart-Scott-Rodino review and considered it again when the European Commission initiated a second phase review.

Well Larry, that's not a winning argument.  Don't you know that nothing makes the EU feel more comfortable than telling them the American government thinks it’s OK.  Look how well the EU is coming through in Afghanistan. I hope Ellison has already made the threat Ballmer should have: 

"Approve this or Oracle is out of Europe.  Who needs your dying statist economy anyways?"

-- Dennis Byron

Did Romney Plan to Ride the Microsoft Anti-Trust Lawsuit to the White House?

In preparing my recent annual blog post correcting the record concerning Massachusetts' so-called adoption of so-called open standards by so-called "brave Massachusetts technology execs," I came across some academic research on the subject.  It is written by three Universitiy of Illinois (UI) professors and is in some unclear manner funded by the National Science Foundation (NSF).  The UI professors' research depends heavily on the open standards blogosphere propaganda lie that Microsoft (MSFT) killed the open standards effort here in Massachusetts through aggresive lobbying in 2006. 

That's all Sun (JAVA)/IBM Kool Aid of course but the academic paper got me thinking: mabye this really had more to do with the Massachusetts lawsuit against Microsoft over Internet Explorer and the fact that Mitt Romeny was running in the 2008 U.S. presidential election from the minute he became Massachusetts' governor in January 2003. In reading about the role the UI academic research ascribes to former Mitt Romney aide, Eric Kriss, I decided to re-analyze my opinion. I believe that the effort to foist "open standards" on Massachusetts government was simply the misguided effort of a few low-level information technology (IT) bureaucrats in one department of one part of one branch of Massachusetts' three-branch, multiple-part, multi-departmental state government, aided and abetted by relatively low-level IBM and Sun guys. 

But maybe I should put my Massachusetts politician hat on.The Kriss connection spurred me to read some of my own contemporaenous analysis of the Massachusetts IT Commision, which met throughout 2002-2003, predating all the publicity about open standards by two years. This was also during the period when Massachusetts was the lone holdout or one of the few holdouts against settling the Microsoft anti-trust action begun years earlier by the U.S. Department of Justice (DoJ).  Perhaps Romney's aides thought they could run that issue--or the good-government (known as goo-goo here in Mass.) idea of openness--into a presidential campaign issue.

The "open standards" crusade had actually begun with a legitimate open source idea to preserve Massachusetts' government intellectual property (IP).  It had nothing to little to do with public records or open document formats at the beginning in 2002. But upon taking over the governorship as the IT Commission was wrapping up its work in 2003, the Romney administration twisted the obviously good idea of preserving Massachusetts' own IP into a potential mandate to acquire only open source licensed IP.  A Senate committe quickly nixed that idea for a lot of reasons (most important of which is that the Romney administration could not demonstrate a return on investment) and the open source idea slowly morphed into an open standards idea related to public records and announced in 2005. 

The problem, as explained elsewhere on this blog, is that the Governor has nothing to do with public records in Massachusetts. The rest is history as they say, albeit in this case one of the few historic records written by the losers at IBM and Sun. And of course the Romney campaign failed--if that was its intention--to light a fire with this goo-goo idea or any other and lost almost every presidential primary he entered in 2008.

By the way, I can't tell if the UI academic research is just unintentionally error-ridden or intentionally misleading.  It admittedly depends on what it calls "anecdotal" evidence when--as explained in my previous blog post--all kinds of clear, crisp Massachusetts public records, including all the material related to the 2002-2003 IT Commission noted above, are available to explain what really happened.  One of the professors told me he read the IT Commission materials but I see no mention of it in the academic paper.

I have written the professors to see if I can get some of these issues cleared up.  How I get them retracted is more of a problem. Having this bull forever out there on the web sites of already discredited media outlets is one thing.  Having it preserved forever in some NSF-sponsored official looking document is another.

-- Dennis Byron

IBM, Oracle, Microsoft Not Worried about a Drupal White House

If you follow the growth of content management software (CMS) in the enterprise software market for a living and also follow United States (US) politics as a civic responsibility, your worlds intersected during the week of October 26.  The leading related headlines in the blogosphere in terms of most frequent appearance went something like this:

• "White House Switches to Drupal, Making Obama Our First Open-Source President.."
• "WhiteHouse.gov goes Drupal: can open source lead to open government?"
• "White House opens Web site programming to public."

So if you invest in IBM Filenet, Oracle (ORCL) Stellent, Microsoft (MSFT)  Sharepoint or dozens of other CMS products marketed by publicly traded companies [e.g., Autonomy (LSE:AU), EMC (EMC) Documentum, OpenText (OTEX), and more], should you be following these developments closely? 

The answer is yes, but don’t depend on any of the stories or blog posts that appeared on the blogosphere during the past week for your background information.

You should pay attention because CMS is a hot enterprise software market area that should buoy many of these public companies’ share prices over the next decade as the need to manage unstructured data (like rich web site content, email databases, etc.) becomes more prevalent than managing structured data (like numbers, names and addresses in an ERP system, supply chain logistics, etc.).  Just as the ERP market matured and consolidated during the 2000-2009 decade, CMS will mature and consolidate between now and 2019.  So there will be winners and losers just as with ERP, a market that could not support the dozens of separate suppliers that existed 10 years ago and that has now effectively been winnowed down to a few choices (Intuit, Microsoft, Oracle, Sage and SAP).

As for the whitehouse.gov stories that have permeated the blogosphere this week, I have not seen so much drivel and deception since 2005 when IBM and a few low-level information technology (IT) bureaucrats in Massachusetts tried to convince you that Massachusetts was some kind of leading-edge proponent of Sun (JAVA) Open Document Format. The Drupal postings whose headlines are noted above are full of the same kind of incredible misinformation as the Massachusetts ODF articles of 2005:

• As explained here on capitalgig.com, Obama was NOT the first open source president when it came to CMS at least (Clinton probably was but no one from that era has weighed in)
• Open source—at least commercial open source marketing that involves red, white and blue for-profit companies such as Acquia, the sponsors of Drupal, along with General Dynamics (GD), the developers and implementers of the current whitehouse.gov—has nothing to do with open government and GD’s and Acquia’s market dynamics are exactly the same as IBM’s, Oracle’s, etc.
• The White House did not open its website’s code to the public; that would be just plain dumb

As for all of the blogoblather about Drupal there were a set of anti-Drupal headlines during the week as well.  For example

• "Why Drupal in the White House is a Bad Idea"
• "Obama's biggest blunder - Drupal?"

Less you think these posts come from George W. Bush supporters mad that he wasn’t given the credit he deserves (as if he even knows what the discussion involves) for putting up an open-source-Perl-based CMS in 2001-2002 or from IBM, Oracle, etc. CMS proponents worried about Acquia and its 300 support customers, neither is the case.  All the anti-Drupal vitriol comes from other open source CMS products.  Go figure.

And only one blog post that I saw in the past 10 days pointed out that Drupal got kicked off the recovery.gov web site at the White House the same week that it got put on the whitehouse.gov web site.  Not much investment potential in a marketing machine that loses one for every one that it wins.