Pros and Cons to Open Source Business Models

I just read this NY Times article on Open Source as a Model for Business is Elusive which claims that successful open source companies have more societal and strategic value than financial and cites the effect that MySql is having on the EU's evaluation of Oracle's acquisition of Sun as an example. They worry that if Oracle acquires Sun, then they will kill MySql which is a low cost alternative to Oracle's database product.

The article lists the usual suspects in their case both for and against open source; Red Hat, XenSource, and Mozilla being the three top reference models.

They also show how open source companies are supported by large technology vendors with deep pockets in order to wage proxy wars with their competitors. Google with Mozilla and IBM with Linux against Microsoft are two examples cited. This is actually Oracle's defense against the EU's concerns in that Oracle claims that they will continue to support MySql in order to eat into Microsoft's Sql Server market.

It seems that the exit strategy for open source companies these days is in acquisition over going public. Witness the recent purchase of SpringSource by VMware in order to gain more control up the API virtualization stack as the corporate world turns towards cloud computing. The acquisition of XenSource by Citrix is another example.

Giga OM countered to this article with an article of their own with the claim that Open Source Business Models Aren't Dead End Streets. They cite Android, Acquia, and Cloudera as example companies.

The Browser Wars Circa 2009

If you have any understanding of computer technology and you haven't been in a coma for the past twelve years, then you already know that there has been a very significant trend in software application development from windows based applications to web applications.

The drivers for this trend aren't very hard to comprehend. A traditional windows application incurs a lot more development costs in terms of installation and testing on the various different types of client computers (i.e. PCs) than the same app delivered as HTML over the web. While there have been impressive advancements in reducing windows application testing and deployment over the years, there is still a higher TCO for windows apps than for web apps.

Not that web apps will completely take over windows apps. Some areas, such as graphics manipulation, VoIP, and video capture, will most probably always be in the province of windows apps. At a minimum, you will always need a web browser running on the client machine as a windows app in order to get access to the web apps. Without the web browser, the web apps are useless.

This is nothing new to the major technology vendors. As competitors over gaining IT market share, they have known this for quite some time. Own the web browser and you own the web. That is why Microsoft aggressively went after Netscape back in the mid 90s. Netscape was the corporation that formed around the original inventors of the web browser. This competition between Netscape and Microsoft eventually led to Netscape being acquired by AOL in 1998. Round one of the browser wars goes to Microsoft.

But the founders of Netscape were not willing to give up so easily. Even as the company was being sold, they created a non-profit foundation devoted to the proposition that innovation on the Internet would thrive only if there was available a web browser that was not so directly controlled by any single vendor. This Mozilla Foundation eventually spun off a for profit subsidiary in order to gain the revenue needed to continue to provide a quality web browser.

This "phoenix from the ashes" strategy worked well. While continued development of Microsoft's web browser languished, the Mozilla browser (called Firefox) continued to enhance and innovate on the web browsing experience. Mozilla was able to do this because they used the open source model to keep their development costs low. Recently, there has been much speculation about the mass migration of web browsing from Microsoft's Internet Explorer web browser to Firefox. Round two of the browser wars goes to Mozilla.

What about the other players in this war? Well, Apple has always had a place on the battlefield with their Safari browser. They don't have much in the way of market share, however. There's a few other minor players but their low market share numbers make it such that they are really not worth mentioning here. What is newsworthy is when Google announced their entry into this war with their web browser named Chrome. For one thing, a lot of the revenue for Mozilla comes from Google. The concern is that revenue stream will dry up now that Google and Mozilla are direct competitors.

Google Chrome currently doesn't have a lot of market share yet so why the concern about Chrome? Google is a big company with deep pockets. This coup has been tried before by another big company with deep pockets, Microsoft. Google is ratcheting up their marketing machine over Chrome.

There are also some noticeable difference between what Google has done with Chrome and what Microsoft did with Internet Explorer. The biggest difference is that Chrome is based on open source.

Recently, the NY Times published a story on the latest turn of events in the web browser wars. The war is very lukewarm now. Not a hot war at all. Google will continue to fund Mozilla, at least until 2011. Google's funding accounts for over three fourths of Mozilla's revenues. Mozilla recently moved their physical office away from the main Google campus.

Why do you care? If you are a software vendor or IT shop that makes and publishes web applications, then you want to make sure your applications run smoothly in the most popular web browsers. If your web applications suddenly stop working, then you have a serious problem.

That is why industry watchers keep up with the web browser wars. They don't want to be caught by surprise by the threat of a web browser upgrade or patch that was purposely designed to destroy the competition.

Oracle to Acquire Sun

Sun Microsystems has been on the ropes since the first decade of this millennium when some of the largest Internet properties (e.g. Google, Yahoo, Ebay, Amazon) switched from Sun's Solaris to GNU Linux. That, and the two most recent stock market crashes, have really beaten this server hardware vendor and inventor of the Java programming language and platform down from a share price of $233 to $3.

A few weeks ago, it looked like IBM would be the one to purchase Sun. Now, it looks like database vendor and services company Oracle is stepping up to make the purchase.

This could be of particular concern to the large number of web companies who use MySql as their database. MySql is owned by Sun and will fall under Oracle's control once the acquisition is complete. Will Oracle poison or sunset MySql in order to make their own database product more compelling? Oracle has tried to acquire MySql before and has acquired one of MySql's back-end storage systems so you can't tell me that they aren't thinking about it.

That's the big news. Oracle makes a few other tools that compete with Sun tools too. Will Sun's NetBeans take a back seat to Oracle's JDeveloper?

May Update: Sun shareholders sue to block Oracle Acquisition.

June Update: Oracle fails to get regulatory clearance by the deadline.

September Update: DOJ gives Oracle approval to proceed with acquisition. Only barrier now is EU Commission approval.

The Second (Life) Coming

Some recent announcements show movement on the part of Linden Labs to open up their protocols for Second Life such that others can host servers to participate in a federated, 3D version of the world wide web.

This is not Linden Labs first journey into open source when they released an open source version of their HUD. What's different this time is that they are taking steps toward releasing server code and protocols such that others can host virtual worlds on their networks which you could teleport to using the Second Life HUD. It still has a long way to go but it is a step in that direction.

Linden Labs isn't the only one using open source as a way to leverage themselves into a larger market where they have to share. This larger trend of moving from content provider to content platform provider means that Linden Labs has to give up a share of the revenue pie in order to be a founding participant in the vision of a World Wide Web where web pages are replaced with 3D virtual spaces and surfers who happen to be visiting that place at the same time get to text message each other.

Considering the impact of Web 2.0 and social networking on the web, I believe this to be a very natural and obvious evolution of the web. There are plenty of competitors who would love to beat Linden Labs to this punch so perhaps that is their "bottom line" motivation to move in this direction.

Facebook Joining the Open Source Bandwagon

It is expected that predominant social networking web site Facebook will soon announce the re-release of its platform as open source. Competitive pressure is credited as the reason behind this move as competing web sites are already positioning to get behind another open source platform for exchanging social information. This is driven by a larger trend in which content providers leverage Metcalfe's law by transforming themselves into content platform providers.

Microsoft Boosts Support for Rival Formats in Office

In a recent news article, Microsoft has announced plans to provide native support for the Open Office file formats in its own office productivity suite.

Open Office is an open source office productivity suite whose development is mostly under the guidance of Sun Microsystems.

Microsoft has made a lot of money off of their own office productivity suite. Over the years, this profit has been used to underwrite Microsoft's attempts to acquire increased share in the markets that Sun traditionally occupies.

I find Open Office to be an excellent office productivity suite. Because of this, I have neither personally used nor purchased a copy of Microsoft Office since 1995. I have used later versions of MS-Office while working for various corporations where that productivity suite is part of the standard, corporate desktop image.

Although MS-Office still enjoys predominance in the marketplace, the list of companies that have switched to Open Office continues to grow. I believe that more and more companies will re-examine their position on MS-Office and its premium pricing as the market of office productivity suites becomes more and more of a commodity.

The Return of the Semantic Web

About a month ago, I blogged on how the semantic web was having a hard time getting any traction. I mentioned then that the Reuters News Agency was starting to endorse it. Now it appears that Yahoo is also starting to embrace the semantic web through the announcement that they will begin to parse certain microformats with their search engine. Perhaps it is premature to count the semantic web as DOA. I hope so.

Some of you may have noticed that, technically speaking, microformats are not really a part of the semantic web specifications. To me, the semantic web is all about marking up content in a machine understandable way such that the semantic context of searchable content is also specified and searchable.

As the project leader for an open source content publishing system, this is great news for me. I would love to add RDF Site Summary and microformat support to improve the SEO features of this project. There is no reason to add these features unless the popular search engines start using them. I wish that Google would make a definitive announcement of what microformats and/or RSS modules that their search engine would support. Right now, it appears that they are supporting hCard, hAtom, XFN and FOAF.

What is your opinion on this? If you are a content publisher, then what SEO enhancement features and formats would you like to see in your software? The most viable contenders seem to me at this time to be hCard, hCalendar, FOAF, and DC.

The Top 20 Proprietary Tools That Drive You Poor and Their Open Source Alternatives

Occasionally, I write here to advocate the use of OSS. I ran across a blog entry recently that recommends open source alternatives to popular proprietary tools. I made some open source recommendations for developers but never spelled out the proprietary tools that these were intended to replace. I thought that I would do that here.

OSS Alternatives To Proprietary Tools
Proprietary	Open Source
VSS, SOS	Subversion
IIS	Apache
Weblogic	Tomcat
VS.NET	Eclipse
SQL Server	MySql, PostGreSql
Sharepoint	Plone
Essbase	Mondrian
Internet Explorer	Firefox
Microsoft Outlook	Thunderbird
Maya	Blender
Cakewalk	Lillypond
Goldwave	Audacity
Half-Life	Nexuiz
Vista	Ubuntu
Photoshop	GIMP
Illustrator	Inkscape
Microsoft Office	Open Office
Microsoft Visio	Dia
Magic Draw	Umlet
ERWin	ddt

Sun Acquires MySQL

In an earlier blog about Open Source Software, I had posited that MySql was perhaps a more relevant open source relational database than PostGreSql by sheer virtue of the fact that Oracle keeps trying to kill it by acquiring the vendors who provide MySql's transactionally aware back ends.

A recent announcement on the MySql site names SUN Microsystems as successfully acquiring the MySql AB corporation. This is a surprise because they usually promote competing open source relational database vendor PostGreSql. Here's hoping that SUN isn't buying MySql in order to kill it so that PostGreSql can grow into MySql's market.

February 2009 Update: MySql founder Michael Widenius has resigned from Sun.

Son of Open Source

It has been a month since I wrote on my favorite open source projects. Some of you have been very kind to point out some cool projects that I missed. I feel like now would be a good time of introduce these worthy projects.

In the original article, I talked about Java and some very nice libraries that extend its functionality and power. One project that I missed was google collections which extends the Java Collection Framework. Though still in alpha, this project is one to watch for.

I talked about the popularity of using IoC/DI which is a lightweight way of building applications from components without introducing of lot of static dependencies thus allowing complex systems to be resilient to change over time. I listed Spring as a technology that empowers this lightweight approach. Another is Google Guice that also allows you to easily build complex object hierarchies from a configuration. The difference between Guice and Spring is the format of the configuration file. For Spring, that is a single external XML file. For Guice, it takes the form of special meta-data comments embedded within the code itself which could be spread across potentially thousands of different files. There are pros and cons to both so which approach appeals to you is a matter of personal choice.

I talked about JUnit which is a unit testing framework. One advantage of IoC/DI is that you can easily plug in a data access tier for the real application that uses a relational database and a mock object tier that return canned responses for the purposes of unit testing. EasyMock is a way to dynamically generate these mock objects.

I can't believe that I completely forgot to mention workflow in the original article (it was in the mind map). This term has been overused so allow me to say that there are two types of workflow, internal workflow and external workflow. External workflow is compelling because it allows for a standardized way to automate the interaction and integration of different systems to provide a complete solution. Internal workflow is an easy way of making your application more flexible and configurable by allowing you to bolt on a workflow engine that empowers easier customization of business processes. There are two parts to workflow, the engine and the editor. The editor is used by workflow architects (if internal) to customize business processes. It is also used by integration specialists (if external) to specify web services that are aggregates of other web services, keeping satellite systems in sync. For internal workflow, the engine is embedded within the application. For external workflow, the engine is a web service that is considered separate from, but dependent on, the other systems. There are a few really good proprietary workflow engines and editors out there but my favorite OSS workflow is YAWL.

I talked about web browser scripting and listed my favorite libraries to help accelerate development. I mentioned how hard java script is to debug but completely forgot to mention my favorite java script debuggers. For Firefox, I use the Venkman debugger. Though not OSS, I should mention that VS.NET has an adequate debugger for java script on Internet Explorer.

The Rise of Open Source Software

The Open Source Software movement has been gaining so much momentum within the Information Technology community recently that it is almost impossible to ignore. Technology players like IBM and Google have contributed to open source. Even Microsoft, whose CEO displays open hostility towards OSS, have contributed. Now media companies such as Yahoo and the New York Times have contributed to OSS.

CTOs and other key decision players in any company's IT department no longer consider OSS as an "if" but rather as a "when." It's not so much a decision as to whether or not to embrace OSS but rather to decide what their OSS strategy is and the timeline for executing on that strategy. It is also very likely that OSS is already being used within their organization.

Why is it that OSS is so compelling? It has a large community supporting it and the price/performance ratio can't be beat. I have blogged on this subject in more detail elsewhere.