I have a new blog so this blog is now closed down! 

Wait a sec and you should redirected automatically, if not click below 

http://steves.businessblog.com/

Subscribe here

http://steves.businessblog.com/blog/index.xml

If you want to know why I switched have a look here

http://steves.businessblog.com/blog/_archives/2004/8/25/129522.html

Jonathan Schwartz writes another great article about what's important about the word Open in an IT context, he does this by comparing and constracting Open Source with Open Standards.  he goes further by showing the great work Sun has done to create reference implementations of their J2EE standard, and provide tools to verify compliance.  He provides a few real world illustrations of how the difference affects real business decisions.

Definately worth a read.

http://blogs.sun.com/roller/page/jonathan/20040808#rewriting_history_and_vocabulary

I was talking with some senior guys from Red Hat last week about their potential move beyond platforms towards solutions.  We were actually discussing collaboration solutions.  There view at the time was that their focus was to take what was available in the Open Source community and productionise it.  Its interesting therefore to see them release an application server.  When you look at the potential though to address the collaboration market Red Hat would do well to consider packaging a solution for email, IM, document management etc.  At the moment they ship the bits, but the bits don't make a solution.  If you look at a previous post about Microsoft and their, "integrated innovation", marketting there is probably as much scope if not more to do the same thing in the Open Source world.  Start thinking Solution guys, you seem to have Platforms and Component packaging fairly well sown up.

eWeek reports that Wyse, (a long term user of Windows embedded technologies), is now moving into Linux in a bigway.

"Linux has really grown, and has become 20 percent of the worldwide thin-client marketplace,"

What actually suprised me was that the market share was so low, given that Linux seems to be a perfect fit for the embedded market, but clearly it takes time for things to catch on.

This article describes some of the characteristics of XML that make it so powerful  The bottom line is that its generic nature means that generic solutions can be created tat can be applied to all manner of different problems.  The example quoted is a tool that compares two XML files, produces an XML file showing the differences.  The article then goes on to explain how this generic tool might be applied to different types of problem.  Well worth a read if you are trying to get your mind around some of the things that will be possible in the future.  Reminds me a bit of Unix pipes!

Jonathan Schwartz, provides a very interesting perspective on the pickle that IBM have got themselves into over Linux.  Whats even more interesting is that he does not discuss Sun's simillar predicament!  Maybe thats because his series of blogs on this and related topics is building up the background for Sun's position which is likely to see a return to the promotion of Solaris as a viable alternative to Redhat which is gradually being positioned by Jonathan as a kid of "proprietary solution", certainly one that locks you into RedHat.

Update: I am not the only one who is puzzling over what Jonathan is up to with his blog.  Look at this eWeek article that pulls lots of opinion together, it appears Jonathan is playing a pretty risky game.

Although not strictly contradictory, it makes for a nice title.  This article is about one of Microsoft’s reactions to Open Source and one way in which it is delivering on its “integrated innovation”, marketing strategy.  

 

The basic concept is that Microsoft takes a collection of their products, and applies them to the solution of a particular business need.   They publish for free standard architectures, processes, templates etc.  You can populate these architectures with some products of your own choice.  In a way whilst this is not Open Source it’s a sort of Open Solution. 

 

The concept is quite interesting to me because one of the challenges with Open Source software, due in the main to the way it is created, is how to build a coherent solution from the many different components, without some over-arching architectural vision.  Where does this vision get created in the current Open Source development model?  It happens within IBM, Red-hat and Novel etc and it probably happens in a proprietary way.  Even if all of the source for the components in the architecture are Open, the architecture itself is likely to evolve in directions specific to the motivations of its creator and be effectively proprietary.

 

So I am left thinking should the emphasis shift from Open Source to Open Standards and Standard Architectures.  Maybe this in the long run is more important.  If the software that implements the standard happens to be Open Source then that’s all well and good, but at the end of the day possibly of only transient importance.   In their own way, (Microsoft always do things their own way), Microsoft is giving us an example of Standard Architectures, implemented increasingly with Open Standards.  Not quite what I had in mind, but it’s what got me thinking in this direction.

Johnathan Schwartz writes a nice article about Sun's dilema, now resolved, about how to compete against Linux.  Linux is not a product, its a social movement that Sun applauds, so how can they compete?  He goes on to explain that in reality Linux is delieverd as many incompatible distributions, and that its not viable to test against them all.  Out of this confusion, in the server space especially, Redhat have appeared as the clear leader.  Now says Johnathan, Sun has someone to compete with, and their relationship can be a normal business competition without the complexity of the social movement.

He talks a little about Open Source Solaris, gives a download link for Solaris and I provide a link to a Whitepaper that provides more details on Solaris.

In this article, webservicespipeline.com discusses the rate of adoption of .NET compared to J2EE.  Its conclusions are quite suprising.  It seems that the rate of .NET adoption continues to grow at quite a rate, and puts usage on a par or slightly greater than J2EE.  It puts .NET success mainly down to increaded developer productivity and ease of deployment and management. 

This is signifiacnt for three main reasons:

1. In the hard nosed business of IT software development, even with all of Microsoft's woes, when it comes down to making business decisions, many IT companies still seem to make decisions based on rational criteria, and long term strategy and architectural elegance or portability don't win out in many cases.
2. There is likely to be a lot of new software developed for the Windows platform
3. Mono is going to be a pretty important Open Source project

This may seem like a simple question to answer, i.e. is written, just like any other software!  It also might seem a strange sort of question to ask, but you will hopefully get my point if you read on!

 

NOTE: No thorough analysis supports the observations I report here.

 

It seems to me that the vast majority of the important Open Source Software comes to be through the following mechanisms:

 

1. Cloning or reproducing in some way an existing design specification or similar.  Examples of this route being Mono(.NET), Linux(Unix) and Wine(Win32).  This technique is usually to force a product or interface into the open, by creating an alternative.
2. Donating, i.e. some third party gifts pre-existing Open Source to the community, examples of this being OpenOffice, Zope and Niku.  This route is often taken by closed source product companies with an old product that is not generating much revenue.  The closed source community uses this old product line to, improve their image, generate services revenue, stimulate demand for optional closed source products, kill off a competitor etc.  In some cases the original developer continues to have some involvement in the development, in other cases not.
3. Sponsorship, i.e. some third party, usually a commercial company, for reasons of their own, sometimes benevolent sometimes not, sponsors the community to develop a product or improve it in some way.  Perl and Python are examples of this.  Most often those sponsored are development leads.  This is often taken by companies wishing to safeguard investments that depend on the continued evolution of the Open Source product, or to stimulate demand for a relate product like training or books.
4. Academia, i.e. some academic project deliverable evolves or develops some Open Source software.
5. Community demand, the final and by far the rarest mechanism is that the community sees a need, appoints a leader and builds a community to solve the need.  The best examples of this are extensions to the four categories above, for example a driver for Linux or a filter for OpenOffice.  I believe the most important example however is Apache.
6. Individual demand, i.e. a person pursues an interest, or a personal need, programming languages are a good example of this for example boo and pylon.  The passionate individual becomes the leader by default and may gather around then a small band of co-developers who share the interest.

I hesitate to put them in order of importance, because of the importance of the exceptions, however if we set aside Linux, then Donating seems to be the most important way for "market shaping" software to become Open Source.  The order of importance is then probably:

 

1. Donating
2. Cloning
3. Sponsorship, (because it funds the community leaders)
4. Academia
5. Community demand
6. Individual demand

Of course these mechanisms are not the sole domain of Open Software, Free Software, (i.e. free to use), has also been developed extensively and uses all of the same mechanisms to come into being.  An additional very important mechanism for Free Software being bundling, i.e. its free with this book, magazine, operating system, training course, paid for product etc.

 

Future posts will reveal why this analysis is important, in my view at least.