IT Investment Research

AP Reports Open-Source-Based Software Failed to Meet Obama Team Needs

As I blogged about on October 26 on itbusinessedge, the Associated Press (AP) is providing the only independent primary-source for a report dated October 24 that whitehouse.gov has changed its underlying content management software (CMS).  This is a category of software that includes EMC Documentum, IBM Filenet, Oracle (ORCL) Stellent, OpenText (OTEX), dozens of open-source-licensed products and projects, and software sold as services associated with Microsoft (MSFT) Sharepoint. For most IT staffers, the whole subject is kind of yawner—IT shops change underlying software every day.

No one would probably even write or blog about the change except for the mangled set of facts put out on the AP wire, and the difficulty trying to clear them up. Neither General Dynamics (GD), the large federal contractor that did the work according to the supplier of the new CMS software, or the supplier itself can or will say what the old CMS was or what other new CMS options were looked at by GD or the White House or even answer simple questions about how or when the contract to do the work was awarded and how much the new change cost. 

• Acquia, which is funded by North Bridge Partners, Sigma and Partners and O’Reilly Media’s venture arm, is the developer of the new CMS. But it deflects questions to its PR agency which says, we’re not “at liberty to say” in response to the above questions.
• “I will forward your question to our customer (presumably meaning the White House) as they have requested,” says the General Dynamics’ spokesperson, the prime contractor for the White House Communications Agency but better known as the tanks and submarine manufacturer that did over $12 billion in business with the U.S. government in its Fiscal 2009 just ended.  (General Dynamics got into the information technology business by acquiring Anteon in late 2005.)

GD won't even say if it has ever done a press release about the subject. Acquia won't even say how many developers it has working on the product.

So falling back on the AP report, here’s what we know--or more appropriaely don't know--with some educated-guess commentary:

• “White House opens Web site programming to public.” -- Presumably that headline, which appears hundreds of time on the web in publications that picked up the AP story, refers to the fact that the new CMS is an open source licensed project supported by Acquia, in the way that Fedora is an open source project supported by Red Hat (RHAT). It is unlikely that the White House is inviting random programmers to change its code base.
• “Security is fundamentally built into the development process because the community is made up of people from all across the world, and they look at the source code from the very start of the process until it's deployed and after." -- We don't know from the AP story what the security issue was at the Bush White House that General Dynamics had to fix.
• “It will be a much faster way to change the programming behind the Web site. When the model was owned solely by the government, federal contractors would have to work through the reams of code to troubleshoot it or upgrade it. Now, it can be done in the matter of days and free to taxpayers.” -- Presumably the revenue loss to General Dynamics will be minor as federal contractors will be replaced by thousands of volunteers.
• “It didn't let the tech-savvy Obama team build the new online platform it wanted. For instance, 60,000 watched Obama speech to a joint session of Congress on health care. One-third of those stayed online to talk with administration officials about the speech. But there are limits; the programming used to power that was built for Facebook, the popular social networking Web site.” -- As with other sections of the AP article, the antecedents of the many “its” and “thats” are hard to follow but apparently the open-source-based Facebook was not up to the challenge that the Obama team wanted or needed to tackle. (However I don’t believe the AP meant to say that Facebook was the outdated CMS technology used by the Bush administration, since the AP refers to that technology as the 'proprietary Bush technology.')
• “It comes more cheaply than computer coding designed for a single client, such as the Executive Office of the President. It gives programmers around the world a chance to offer upgrades, additions or tweaks to existing programs that the White House could - or could not - include in daily updates.” -- I am almost certain that the Obama administration does not want programmers sending upgrades or tweaks to the White House—send them to Acquia.

Other independent web sites, in particular techpresident, the Register and CNet, have taken some good educated guesses on the subject as well.

-- Dennis Byron

Open Standards Bigots Still Putting Out the "Big Lie" about Massachusetts and ODF

About once a year, some blogoblatherer peddles the “big lie” that earlier this decade all State of Massachusetts government information-technology (IT) groups were major supporters of Sun's (JAVA) Open Document Format. So once a year, being Massachusetts born and bred, I feel the need to knock down the Massachusetts/Open Document Format (ODF) 'big lie" by providing links to the relevant public records.

This year the offending blogger is Paula Rooney of ZDnet (see her paragraph about "brave Massachusetts technology execs") but if not her, it would be someone else before long. It is unfortunate that she did not check her own web site before writing about the "brave… execs." In addition to the relevant official state documentation on the subject in the links below, there was some pretty good contemporaneous reporting of the back-room dealings between Sun and a few state IT employees right on ZDnet. One of the more telling sentences about what appears to me to be a Sun-ODF effort to manipulate Massachusetts was this information from David Berlind of ZDnet in October 2005:

"Although it isn't clear what Massachusetts' test for openness was back in the early 2005 timeframe, it is clear that "the test" was officially revised... by the time (Massachusetts' Enterprise Technical Reference Model) ETRM was ratified on Sept. 23..." 

Why (and how) the test was revised was answered in 2006 but the open standards blogoblatherers conveniently leave out that fact. The information about a flawed process developed by Berlind was confirmed in a Massachusetts legislative report, which said that that particular version of the ETRM (3.5) was put together by a "kitchen cabinet," which not only failed to give outsiders such as Microsoft (MSFT) a reasonable chance to comment on or react to its plans but also froze out all other relevant Massachusetts IT groups, including the ones involved with aiding the disabled and those in charge of public records.

The legislative report lays out how this small group—actually one department in one part of one branch of the Massachusetts government—took numerous underhanded (even by very generous Massachusetts state government standards) to try to wire the procurement process for Sun. The “brave… execs” issued--possibly illegally according to the legislators--ETRM 3.5 saying that Massachusetts government would only use documents produced in the following four "standard" formats: HTML, Sun ODF, Adobe (ADBE) PDF and TXT. Note that there were four “standard formats” listed, not just ODF. "Standard" is in quotes because PDF was no more a "standard" at the time than Microsoft’s Office formats but clearly the "brave... execs" had to make their "open" version of the ETRM look somewhat open.

The effort was part of a multiple-year process that began in 2002 with a study by a state-sponsored group called the IT Commission, apparently orchestrated by the same "brave... execs." The IT Commission was “facilitated” by IBM (IBM) consulting employees under contract and included a Sun employee but no Microsoft employee. IBM uses ODF in one of its less popular Lotus products or at least did so at the time. Microsoft was a major supplier of technology to Massachusetts at the time (and still is I believe) and should have been represented on the commission or all vendors should have been barred.

The legislative committee concluded that Rooney's "brave... execs" were:

“… not aware of the cost of the ETRM, the impact it could have on the state’s public records, limitations on IT accessibility for persons with disabilities, that the agency excluded key governmental and advocacy groups, and that the proposal was issued in violation of state law.”

The above is the summary.  The details run on for 30-40 pages. The finding of disregard by ITD and the IBM-Sun group for the disabled bordered on immoral, a much more serious charge than illegal here in Massachusetts.
 
Today as in 2005 the issue of document format standards is, as it should be, a miniscule part of Massachusetts’s ETRM (now in its fifth version). New IT management in state government gave Microsoft a chance to make the list and gave Sun a chance to make ODF handicapped accessible. Massachusetts considers HTML, ODF, OOXML, PDF, RTF and Text to be acceptable document formats. The issue of accessibility for the disabled is now a key part of the IT procurement process.

Common sense triumphed over the attempted Sun-ODF coup in Massachusetts earlier this decade. Unfortunately other Massachusetts politicians went on to try to extort money out of Cognos.

-- Dennis Byron

MySQL to Oracle? Marten Says “Kylla!” Monty Says “Ei!”

In dueling blogoshots released in October 2009, the two Finns most associated with the MySQL database software company, which they sold to Sun (JAVA) in 2008 for a bazillion markaas, are urging the European Union in opposite directions when it comes to Oracle’s (ORCL) takeover of MySQL (by virtue of Oracle’s pending acquisition of Sun). 

Marten Mickos, the CEO of MySQL from 2001 to 2008 and now an entrepreneur in residence in Silicon Valley at big MySQL cash-out winner Benchmark Partners, says

"Oracle has as many compelling business reasons to continue the ramp-up of the MySQL business as Sun Microsystems and MySQL previously did, or even more. Even if Oracle, for whatever reason, would have malicious or ignorant intent regarding MySQL (not that I think so), the positive and massive influence MySQL has on the DBMS market cannot be controlled by a single entity--not even by the owner of the MySQL assets. The users of MySQL exert a more powerful influence in the market than the owner does." (Source: CNet)

Ulf Michael Widenius (called Monty), co-creator of the MySQL project, founder of the company of the same name in 2001, and now founder of the MySQL fork project called Maria, says (on his blog post quoting himself from a press release that was not released):

“… Oracle should resolve antitrust concerns over its US$7.4 billion acquisition of Sun by committing to sell MySQL to a suitable third party… MySQL needs a different home than Oracle, a home where there will be no conflicts of interest concerning how, or if, MySQL should be developed further."

Mickos wins this Arctic Circle standoff on the numbers from an information technology (IT) investment perspective.  Oracle can’t sell MySQL because it’s sitting on Sun’s books at a half a bazillion markaas and is probably only worth about as much as a markaa is today given the trivial number of paying MySQL customers estimated by Mickos in his blog post.  

Monty's argument against Oracle makes no sense in that he has already forked the project. And anyone else can as well because of the license terms and conditions he used to initially release it.

Of course Mickos' trivial numbers kind of fly in the face of Mickos’ claim to the EU Anti-competition Czar Neelie Kroes that MySQL has a “dramatic effect” on the database market. Unless the dramatic effect is that non-paying customers are not buying database software, something that disagrees with what I see in the marketplace. But Neelie wants out of this job so badly she already gave Microsoft (MSFT) everything it wanted in a recent decision.  She probably wants out of there before Ellison starts tearing into her. He won’t meet her for a quiet breakfast of Belgian waffles some morning the way Ballmer did.

-- Dennis Byron 

Ballmer Does Boston

People that live in first-tier towns like NYC, London, San Francisco and Chicago always say they cannot believe how provincial Boston is. As a native, the charge always bothers me but Boston keeps proving what a one-horse parade it runs--even in terms of the enterprise-software world--when it gets all a twitter (literally and figuratively) over the fact that Microsoft’s (MSFT) Steve Ballmer met some people in Boston (actually across the river in Cambridge I think) for breakfast on Friday morning October 16. 

The home-town Boston Globe is all over the “story,” except that there does not seem to be a story.  They just had a business breakfast like the last 5000 business breakfast you’ve attended.  According to the Globe, the “guest list” included an eclectic bunch for sure:

• Two old enterprise-software dinosaurs, now VCs, Dan Bricklin from the 70s (Visicalc) and John Landry from the 80s (Lotus)
• A much younger dinosaur, Bill Warner, who founded Avid Technology (AVID) in the 1990s and now seems to mentor enterprise software folks
• A bunch of younger VCs (Microsoft is always on the prowl; it picked up its Chief Technology Officer Ray Ozzie in Massachusetts, not that Ballmer needed a VC breakfast to find Ozzie)
• A college president, Gloria Larson (not a college whose name you’d recognize unless you follow Division II basketball where Bentley is nationally ranked perennially; Larson has the college presidency because of politics, probably the same reason she was at the breakfast)
• A couple of leading local Internet entrepreneurs (and the head of the Mass Technology Leadership Council that designated them “leading local entrepreneurs”)
• A twit named Laura Fitton; read it for yourself; I can’t make this stuff up (actually Steve Mann is advising her so she might have something going under the twill)

The Globe is hard at work trying to find out what was discussed other than “Do you believe it snowed in Boston on October 16th?” 

Looking at the guest list I see two themes: mentoring and old and new software start-ups.  Take a guess at which subject Ballmer was probably interested in.

-- Dennis Byron

What's Rocket Software Going to Do with $90 Million

The October 13 squib by Kara Reeder on IT Business Edge about Rocket Software sent me on scavenger hunt the morning of October 14. The Boston Globe article to which Kara referred and subsequent links and Googles implied there was a mystery with questions such as

• “How large an enterprise-software company is Rocket?
• “Where did the new investment come from?”
• “Who are the 24 companies it has acquired?”
• “What was the patent lawsuit with CA settled in February 2009 all about?”

Questions I was asking myself after following the threads from Kara’s piece included

• “Why does a Dutch company have to file with the U.S. Securities and Exchange Commission (SEC)?”
• “How did that acquisition of Seagull turn out, any way?”
• “IBM UniVerse… the UniVerse name rings a bell but I don’t associate it with IBM?”

There’s a lot of speculation about the timing of the announcement that Rocket acquired IBM UniVerse on October 1 and the SEC filing on October 9 but I don’t think it’s likely that Rocket Software, doing $50 million or so in software sales a year, paid over $90 million for IBM’s UniVerse business. Maybe Rocket sought financing so as to pay off CA—and have a little left over to afford a tony (or is it spelled toney?) Massachusetts address like Newton. But why bother if it has five years to pay back CA.

So here (I apologize for the open-barn-door nature of it) is some belated information technology (IT ) investment advice that might still be actionable. Work your way down the Software 500 list looking for slow-growing publicly held infrastructure enterprise-software suppliers, likely with an IBM mainframe connection. 

Rocket Software, a company that has apparently been quietly funding acquisitions from its profits for a decade, now has some real money in its pockets.

-- Dennis Byron

(P.S. The answers to all my questions follow if you care:

(The answer to the SEC-related question was easy, just a matter of bad memory on my part.  It wasn’t Rocket Software that was Dutch. It was good old Seagull, the mainframe screen scraper guys acquired by Rocket in 2007. To be less flip, Seagull described itself at the time of the acquisition by Rocket as specializing

“... in technology that transforms "legacy" applications into SOA-compliant Web services, helping enterprises achieve exponentially faster IT support for business change, governance and compliance.” 

(Seagull continues to market its software separate from Rocket. Rocket is based in Newton, MA and was formed in 1990, making it probably the “youngest” IT/enterprise-software company with a “Route 128” address. Route 128, which circles Boston, was of course the Highway 101 of its day.

(The SEC filing is a Rule 505 Form D event. Rule 505 allows some companies offering their securities to have those securities exempted from the registration requirements of the U.S. federal securities laws. As the SEC site explains:

(“Form D is a brief notice that includes the names and addresses of the company’s owners and stock promoters, but contains little other information about the company.”

(That’s for sure when it comes to the Rocket filing.  I remember chasing down Rocket when they acquired Seagull but hadn’t paid much attention before or since.  They apparently liked their privacy and that was OK with me since no one would be investing in them. Well that was only half true.  You were certainly investing in the many companies Rocket acquired for dimes on the dollar and with little outwardly publicized funding—until this recent SEC filing—except for a line of credit from Wells Fargo.

(Catching up with Rocket now, it looks like its acquisition bandwagon has been at the head of a little known parade for a decade. Rocket is to infrastructure enterprise software what Infor is to applications enterprise software (that is, what the old Computer Associates used to be to all enterprise software). Acquired companies and brands in addition to Seagull (Bluezone/Legasuite) include Areotext (from Lockheed), ARKIVIO, CorVu, Gentia Assets (from Open Ratings) some IMS utilties from Telcordia, Mainstar, NetCure, Peritus, Servergraph, SmartDB, StorageNetworks STORos Storage Manager, SystemsSoft, and TSCI.  Rocket tried to scoop up NetManage, a Seagull competitor, in 2008 but NetManage ended up going to Micro Focus.

(As for the implied mystery, the privacy fetish is apparently still a Rocket Software company characteristic. How large is Rocket Software? I found that although Seagull and CorVu were listed on the Software 500 before they were acquired, they don’t seem to be separately listed in the latest John Desmond bible.  And Rocket is not there either that I can see. My guess: maybe this smorgasbord is doing $50 million a year in software sales.

(As for the law suit with CA, Inc.--not to be confused with the old Computer Associates, two years ago CA said that certain Rocket developers misused intellectual property from one of the CA-related entities and improperly incorporated software code from it into certain software products being developed at Rocket as far back as 1999 and 2000. Rocket said at the time that it believed that “the claims are meritless,” but in Febnruary of this year Rocket agreed to license technology from CA, including source code authored several years ago and related trade secrets, for $50 million to be paid in increments through 2014--see the latest CA, Inc. 10-K.

(As for my final question, over the years, Rocket had purchased the Visionary brand from IBM back in 2000 and has many other mainframe-oriented products as noted above. That, I guess, is what led up to the October 2009 acquisition of IBM UniVerse.

(My memory cells clicked in since I began writing this post 15 paragraphs above. My thought association exercise went like this:

• UniVerse…
• Jim Capeless…
• Paris in the early 1970s...
• Dick Pick

(Some of the story belongs on my travel blog, not here. Most of it belongs in my unwritten “People’s History of Information Technology,” where people like Pick and Capeless get treated with the same depth as Gates and Jobs. If I ever write it, it will be a history where the winners (Highway 101 again) don’t get to write the history books.

(But the nut of it is that UniVerse is a remnant of IBM's acquisition of the database part of Informix in 2001, part of which had previously been called Ardent, which was formed by a merger of Unidata and VMark in 1998. VMark, where Capeless was President in the 1980s and 1990s, after working for Honeywell Bull in the 1960s and 1970s, had written its own version of the legendary Dick Pick’s widely licensed data management concepts back in the 1970s.

(Making this even more confusing, what was left of Informix was re-named Ascential, which was also acquired by IBM in 2005.)

What’s with these SaaS Enterprise Software Fanboys?

Saugatuck Research is out on October 8, 2009 with one of its periodic hallelujah pieces about a so-called class of enterprise software suppliers it calls Software as a Service (SaaS) software providers. (This link requires you to sign in but I believe there is no charge to see the summary.) To be fair, Saugatuck is not the only firm putting out this poorly structured view of the enterprise software industry. Look at Saugatuck's and similar SaaS-slanted research very skeptically from an information technology (IT) investment and IT market analysis point of view.

According to Saugatuck, the SaaS class is posited as different than “On-Premise software providers,” which it periodically also calls "ISVs." The SaaS class is regularly described as a new phenomenon. In one Saugatuck report, this SaaS class is even said to be “…Starting to Feel the Effects of Tough Economy” (see Saugatuck report RA-602, published on 28 May 2009). But the October 2009 announcement says,

“By YE2010, our SaaS basket will be growing at a 20 percent year-over-year rate (with the slope suggesting a potential path to achieve 30 percent or higher segment performance in 2011).”

There is no better way to understand the Saugatuck research’s flaws than to look at its so-called “baskets” of different classes of enterprise software suppliers

On-Premise ISV Basket: 
• Autodesk (ADSK)
• BMC (BMC)
• Compuware (CPWR)
• Epicor (EPIC)
• Lawson (LWSN)
• Oracle (ORCL)
• SAP (SAP)

SaaS Market Basket:
• Blackboard (BBBB)
• Concur (CNQR)
• DemandTec (DMAN)
• Kenexa (KNXA)
• NetSuite (N)
• RightNow (RNOW)
• Salesforce.com (CRM)
• SuccessFactors (SFSF)
• Taleo (TLEO)
• Ultimate Software (ULTI)

It is the most misleading bit of cherry picking and revisionism I’ve seen among this sort of SaaS fanboy research. It is absurd to include BMC and Compuware in a listing that is otherwise made up of enterprise applications suppliers.  Or if there were any logic in including BMC and Compuware, why not include Red Hat (RHAT) and VMware (VMW) also?  Maybe Saugatuck left them off because it would skew the results of Saugatuck’s share-price comparisons in a direction that invalidates Saugatuck theories?

It is equally absurd for Saugatuck to leave off Microsoft (MSFT) and IBM from a list that is supposed to accurately portray the so-called “on-premise software community.” Saugutuck says that’s because the two market leaders do not segregate “their reported revenues between new license sales, maintenance and services revenues” but of course neither do most of the so-called SaaS providers.  So you cannot compare apples to apples. 

And on which list should I put Google (GOOG)?

Then you have the research's logic problem that all of the so-called on premise software community also offers its software functionality as a service--that is, SaaS--and has done so for years.  In fact, although I don’t buy Oracle's actual claimed dollar amount, I agree directionally with Oracle’s claim that it is already the number-two provider of software delivered via SaaS delivery mechanisms.  Of course that is what SaaS is: an enterprise-software delivery mechanism, not a market. That misunderstanding is the biggest flaw in Saugatuck’s and similar SaaS research perspectives.

Conversely, as explained in my IT Business Edge research, almost all of the so-called SaaS basket entrants also offer or used to offer “on-premise software.” That’s another way to say they are all or mostly also 10 or more years old so any share-price time-series analysis similar to that included in the Saugatuck research should span a few decades, not a few quarters.

There are all kinds of finer points to watch for but bottom line: don’t use this stuff to decide on how you invest in the enterprise software market.

-- Dennis Byron

Judge Microsoft-EU Deal Investment Implications by Its Opponents

In reading about the almost Microsoft (MSFT) agreement with the European Union (EU) announced October 7, I feel Microsoft investors have finally got a good deal from the EU regulators. The mounting opposition to the deal is my proofpoint. Just as the Opera (on Oslo stock exchange) company executives that started the anti-Microsoft (MSFT) browser complaint in EU now feel chained to an EU Competition Commission (EUCC) boat anchor being thrown overboard by Neelie Kroes abandoning ship, Free Software Foundation Europe (FSFE) executives are also worried by what looks like a done-deal agreement between Kroes and Steve Ballmer.

The oddly named European Committee for Interoperable Systems (ECIS) is supposedly against it but has not posted a formal statement. [It's oddly named in that the leading members of the "European Committee" appear to be the North-American-domiciled corporations Adobe (ADBE), Corel, IBM, Oracle (ORCL), Real Networks, Red Hat (RHAT) and Sun (JAVA).]

The FSFE and ECIS, rightly so, seems more concerned about the interoperability part of the agreement than the trivial "browser ballot" bulloney. On September 28, Karsten Gerloff—FSFE President—wrote in a blog:

“We are concerned that the Commission may end up reversing years of successful antitrust work if Neelie Kroes settles for far too little in order to close a deal, any deal. That would mean that Europeans remain stuck with the present Microsoft monopoly in most areas of the desktop. Even worse, that monopoly would have the Commission’s official seal of approval, effectively ruining many years of outstanding work by Ms Kroes and her team.”

“Of the desktop” is the key phrase in Mr. Gerloff’s complaint. Microsoft no longer cares about the desktop.  It’s just a cash cow it will milk as long as it can. Technologically and from a business model perspective, Microsoft has moved on, while the EUCC, ECIS and FSFE continue to look backward.

At least the FSFE--looking backward--has a hint that something’s about to hit it in the back of the head. On October 6 (apparently after the story of the impending Microsoft-EUCC deal began to spread), Gerloff wrote Kroes in an open letter saying:

“Microsoft must be required to provide interoperability information either royalty-free or in return for a one-time payment. Running royalties are incompatible with Free Software. The PFIF agreement, which resulted from the Samba case, provides a tested and working instance of such an agreement.
“Microsoft must provide a legally binding assurance that it will not assert those of its patents which relate to the interoperability information against Free Software. The lack of such assurance would let the company use Fear, Uncertainty and Doubt (FUD) to discourage competitors from making use of the interoperability information, leaving the remedy ineffective.”

On October 8, the FSFE posted an even more strident complaint to the EUCC.  With all guns blazing, it says

"There is no clear commitment from Microsoft to adhere to web standards in the future, nor to end the company's habit of adding proprietary extensions to standards."

Bingo, the FSFE is on to something even though it does not appear to know what it is. Microsoft is totally willing to let the European Free Softies re-invent all the desktop spreadsheet and word processors they want, and let them waste even more cycles making their “free” spreadsheet and word processors work with deskbound Word and Excel if they want.

While the FSFE and the EUCC is worrying about 20th century technologies, Microsoft is moving on to augmented reality that makes browsers prehistoric, clouds that make permissive licenses and desktop applications irrelevant, and new (or repurposed Entertainment and Device Division) hybrid consumer/enterprise products that bundle software down in the silicon, where the FSFE cannot find it to re-invent it.

While the FSFE and EUCC is worrying about standards no one in the industry cares about (including all the Microsoft competitors that pay lip service to the ECIS), Microsoft is not only not promising to adhere to any standards, Microsoft is making all these standards irrelevant.  Steve Ballmer explained why and how this works at his July 2009 Microsoft Financial Analyst presentation (see second bullet).

-- Dennis Byron

Be careful what you wish for, Opera!

The failed Norwegian browser provider, Opera, says it is not sure about the Microsoft proposal for a European-Union (EU)-mandated “browser ballot” to be used with Windows 7 as an alleged means to lessen Microsoft (MSFT) Internet Explorer (IE) market power. According to one of the UK information-technology rags (I now can't find the link) Håkon Wium Lie, Chief Technology Officer at Opera, said:

"Opera Software supports the concept of a ballot screen to give users easy access to better browsers. The important question is how this ballot screen is implemented. We are still studying the announcement from the European Commission and will have further comments at a later stage."

Opera (on Oslo Stock Exchange) of course is the sore browser-market loser that started this whole waste of time and investor value.

I can see why he is withholding comment. The example ballot available on the Microsoft site implies that Opera makes it into an all-important top 5 listing of browsers by “share” (term in Microsoft document) or by popularity (term in EU terminology). But neither “share” nor popularity is defined in the accompanying legal blather that I can see. Nor is the time period being measured. Opera’s “share” at given moments in time by various counts hovers around 1%, a third the size of the fourth place browser (which is of course years behind IE. But Opera could actually be surpassed by Netscape or Mosaic for all I know if some more meaningful cumulative criteria were used. Cumulative usage would seem to be the best metric. For example if I started with Netscape years ago, I still want to use Netscape.

It would be poetic justice if after this collosal waste of effort the product didn't even make it on to the ballot?

-- Dennis Byron

Enterprise Software Lock-in: Red Hat to Use Bungee Cord, not Handcuffs

At its financial analysts briefing on October 6, Red Hat (RHAT) reconfirmed that it still plans to

• gain half of server operating software and middleware market share during the 2010-2019 decade
• stay focused on the lower layers of the stack, as opposed to developing or acquiring applications and database products

CEO James Whitehurst explained the uberthinking and then to explain how Red Hat would reach those goals, Product and Technology VP Paul Cormier took attendees through a 30-year history of and 10-year look-ahead into how information technology (IT) systems/software suppliers try to lock-in IT users.  For example, he explained:

• the total systems-supplier lock in of the 1980s, broken by Intel (INTL);
• the Microsoft (MSFT)/Intel Windows lock-in of the 90s, broken by Red Hat (according to Red Hat of course);
• the IBM/Oracle (ORCL) SOA lock-in of the current decade, broken by JBoss.

I would contend that an SOA lock-in never really occurred (or has not yet occurred and may never occur).  But I like the symmetry. Coming next, according to Cormier is the emerging virtualization lock in led by VMware (VMW). See What Virtualization Means for Enterprise Software Tomorrow.

Cormier made an interesting observation after explaining how Red Hat Enterprise Linux (RHEL) saved the world from Wintel and JBoss saved the world from IBM WebSphere and Oracle/BEA WebLogic.  His question: why were users so willing to turn around via virtualization and go back to the total single-supplier stack lock-in of the 80s?

• His answer: It’s easier for independent software vendors (ISVs) to deal with an integrated stack.  That has some merit but shows his Digital Equipment Corp. heritage.
• My answer: Everyone, including IT management and staff as well as ISVs, want an integrated stack.  Except they don’t think of it that way; that is, they don’t think of it technologically.  What they want is ease of use and one throat to choke.

Of course, despite all its good-marketing hoopla about breaking up lock-ins, Red Hat wants to tie up IT users also.  In fact it will have to meet those 50% market share goals. It looks like Red Hat will offer a bungee cord rather than handcuffs but it’s still lock-in. 

And that’s OK, because as I say, that’s what the market wants. It's human nature not technology.  I bought a Chevy recently.  I did not want to buy an engine from Taizhou Jili and a transmission from Dyna and a power train from Getrag and try to put them together myself using my old beat-up Chevy as the body (and it wasn't even a clunker).

Whether you like his answer or mine, the symetrical timeline is already marching on. Next Cormier sees other suppliers, especially Microsoft (who he notes he also actively partners with) trying to lock in users in the cloud.  The technology to accomplish this is found in messaging and grid computer science and Red Hat has been actively working at that layer for two or more years already.

(As a sidelight, Red Hat didn’t take on Microsoft the way Microsoft took on its competitors--including Linux suppliers--at its Financial Analyst day in July 2009. As I noted in the Red Hat middleware post linked to above, I miss Marc Fleury.  Has Red Hat decided to leave it to its lawyers to play hardball?)

-- Dennis Byron

Red Hat Still Wants to Own Half its Markets

Red Hat (RHAT) executives briefed financial analysts October 6.  The key question I brought to the session is

“Where does Red Hat stand against its previously announced plans to gain half of market share in server operating software (see Feb 14, 2008 entry) and middleware during the 2010-2019 decade?”

Red Hat CEO James Whitehurst addressed the goals out of the chute by explaining differences between revenue and sites when it comes to operating software, and the difference between “paid share of application server workloads” vs. the many unpaid JBoss instances in the market.

Red Hat estimates JBoss already has 30% share of application server workloads vs. a like amount for Oracle (ORCL)/BEA WebLogic and IBM (IBM) WebSphere. Whether or not this is somewhere between too optimistic and outright delusional, the problem is that most of the JBoss instances result from product-only downloads licensed via open source terms and conditions (Ts&Cs) with no associated subscription maintenance. Which means no accompanying Red Hat revenue. Red Hat recently trifurcated the JBoss/Red-Hat middleware lineup so as to potentially increase the paid instances.

I was surprised to hear--and assuming that I heard correctly--that Red Hat also believes there is as much unpaid Red Hat Enterprise Linux (RHEL) in the market as paid RHEL. That’s not the same as unpaid JBoss, because—while users can get JBoss under open source Ts&Cs—users should deal with the open source Fedora project, not RHEL, if they do not want Linux subscription maintenance. Red Hat is actively trying to “recover” revenue where RHEL is present at no fee. Whitehurst said Red Hat recaptured $2-$3 million of such RHEL potential in its last fiscal year.

Whitehurst provided this data as part explaining three key Red Hat board-level initiatives in meeting its goals:

• build subscription value

• increase routes to market, and

• encourage mainstream adoption

There are two or more key Red Hat people associated with each of these initiatives, each of these with specific subtasks (e.g., using both the indirect and direct channels to reach market, playing nice with Microsoft to encourage mainstream adoption, etc.).  It sounds like there is a very good program for slapping tasks on people's foreheads and taking names when objectives are not met.

Whitehurst also noted that in addition to market and shareholder goals, Red Hat--from a brand point of view--wants to be known as the“’trusted’ open source leader,” trusted being the key word according to Whitehurst.  I feel that users don't care what Ts&Cs Red Hat's products are distributed with and that users don't go into the market specifically looking for open source software. Users care that these products are good (and Red Hat's are). This branding approach is especially an issue if the goal is half the operating software and middleware markets.

I was good to hear Red Hat re-confirm that it is going to stay focused on operating software and middleware (as expanded with virtualization and cloud-computing-related technologies). Whitehurst says he is not looking for apps or database acquisitions. 

-- Dennis Byron