Once More unto the Breach

"IT Departments Running on the Edge of Collapse"

There's a (short) interesting interview with Sun's James Gosling on C|Net today.  On the second page, discussing his desire to develop tools to help cut software development costs, he depressingly observes:

If you come up with a good software development tool, that makes life easier for the developers and they can get their job done quicker, then the first thing the manager says is "Oh you've got free time on your hands. Do this extra thing." ....

So pretty much every IT department will always be running right on the edge of collapse because if you ever get beyond the edge of collapse, you collapse, things fail, things fall apart. If you ever get on the other side, things get a little bit easy and people say "Oh we can do more."

In some sense I've resigned myself. In the land of tool builders like me, it's not about cutting IT costs as much as it is inevitably about enabling IT departments to do more.

He should be celebrating.  It's all about enablement.  Two ratios we seem destined to live since we started software programming are:

• The average number of lines of code written by an average developer per day
• The number of defects per thousand lines of code in a system

All our advances in languages and tools are about developing more (better) software by writing less code.   From higher level (compiled) languages, to modules, libraries, objects (and distributed objects), scripting languages, web services, visualization, configuration management, and automated build and delivery, it's all about enablement.  Every time we double the speed or transistor density of a chip, or the bandwidth on the Net, we come up with new and interesting applications to devour it.  So too with human capacity.

This is actually why free and open source software is such a fundamental software engineering friendly practice, and why sharing software (eventually through copyright) predates the FSF et al.  Good engineers develop systems to optimize for the scarcest resource.  The engineering efficiency and expediency of sharing known good building blocks and tools is obvious.  It's about preserving the value in the software, not destroying it.

Dr. Gosling is hopefully wildly pleased with his contributions! 

Open Standards, IPR and Innovation Conference, Beijing 2006

The China National Institute of Standardization and Sun Microsystems co-hosted a one day conference on "Open Standards, IPR and Innovation" in Beijing.  I was an invited speaker.  The event's goal was to bring together a set of speakers that would provide different perspectives and practices on how standards and intellectual property rules play together inside and out of China, such that China can determine what will best work for them. 

There is obviously a reasonable amount of tension between a handful of countries and China when it comes to claims of piracy, and as China's economy continues to grow, that tension has broken out in the standards arena as well with competing standards for such things as RFID technology and WiFi complicating the discussion.

The great thing about the conference was that one felt the Chinese speakers and attendees were definitely attending to learn, but it was NOT to learn how to be more "western", but rather to learn so as not to make the same mistakes.  The growing sophistication of business and business practices is palpable in Beijing in general, and definitely present at the conference.

One thing that was also surprisingly apparent was the lack of patience with the multinationals and their pricing, Microsoft in particular.  This was not an undercurrent in the presentations.  It was right out there in the open. 

I remember getting into the debate a few years ago while still at Microsoft.  Some managers took the position that countries like India would obviously want to mimic the  "American" success of companies like Microsoft.  The debate continues along the lines that India would obviously NOT want to use open source software because it's IP hostile, and they really need to build lots of proprietary software businesses because it's a demonstrated way to "make money" and grow the economy.  Of course they would want to do so on a Microsoft software platform.

First, open source software isn't IP hostile.  It depends upon strong IP law.  Second, countries such as India are service-based today, and that means EVERYONE gets to take home some money rather than focusing the wealth in the hands of a few.  (The last thing India wants is a return to the days of the Rajahs with a handful at the top.)  Lastly, (and this is the kicker), when you scale out the amount of economic growth that's taking place in countries such as India and China, the last thing they want to do is ship massive amounts of hard currency OUT of their economy to the U.S. for the privilege of developing software to improve business.  (And Microsoft is a veritable Hoover in its structuring of its subsidiaries.  No amount of in-country R&D re-investment can make up for the amount of money at risk here.)  Open source rules in a case like this.

Leap forward to China.  Same story.  Same debate.  Same mistaken thinking from the Western multinationals that somehow China wants to be like them or worse yet, ship all that money out of the economy (government perspective) or the market (corporate perspective).

I understand a naive Western marketing person's desire to believe that China is their next growth market but they fail to understand the economics involved.  This is why Linux rules here -- the cost of goods sold is significant when you add up the royalty payment times the population of customers.  I have held a Motorola Ming phone.  Developed in China for the Chinese market.  It is the sexiest little piece of hardware on the planet, and incredibly functional.  It's Linux-based. 

Why would Motorola want to send that much money in royalties to Symbian or Microsoft?  Yes, yes, I can hear the "we're so valuable" statements from those marketing managers already.   Frankly, I don't really want to read Word documents on a two inch by three inch screen.  The Ming is a work of art.  But I digress.

Several interesting items came out at the conference:

• China has its own document format standard called UOF.  It is somewhat consistent with ODF.  There is to be a convergence.  I learned at dinner one night that Red Flag has already built UOF support into Red Office, so hopefully the support will rapidly ripple back out through the OpenOffice.org community and the rest of the ODF supporting products will soon support UOF as well.  (Andy Updegrove also attended the conference.  He has made the connections and will hopefully have a proper article on UOF soon on his blog.)
• China is very respectful of IP.  They value patents.  They seem to be uninterested in supporting software patents.  They do however understand how patents might make a mess of the standards world, and are very interested in work like the new ex-ante based IPR policy from VITA.

This was the first event in which I've participated where there was simultaneous translation going on.  My slides needed to be handed in a week in advance such that they could be translated as well.  Wireless handsets were available so you could listen to the translation.  While one screen had the slide in English, the other had the Chinese slide.  Attendees and speakers simply worked in their native language.  It was fascinating to see it all just work.   

I've posted all my photos on flickr from the conference and the conference dinners.  I still need to get some names tagged, but it's a start.  All in all a great conference.  I learned a lot, and it was a privilege to participate.

Post to del.icio.us
digg it

www.flickr.com

This is a Flickr badge showing photos in a set called Open Standards, IPR, Innovation Conference. Make your own badge here.

Sun and Open Source Java versus the Microsoft .NET Efforts

I attended the small event Sun hosted this evening to talk about the next steps towards an open source Java world.  On hand for the discussion were Sun Software EVP Rich Green, VP of Developer Programs and Products Laurie Tolman, and VP of Mobile and Embedded Systems Alan Brenner.  Dan Farber covers the details well here.

There were generally supportive statements around open sourcing javac and the hotspot VM by the end of year, as well as the JavaME (Mobile Edition).  This is all good.  The web site to follow for the work is http://community.java.net/jdk/opensource/

The interesting thing to watch here is how Sun works this opportunity as it's sort of backwards.  Standards exist to enable multiple implementations.  This is certainly how the JCP has functioned, and the Java certification program protects the brand.  Open source software describes a licensing model that ensures a collaborative community can be developed around a software project, and Sun has a long history of participation in such communities, including its own, so should be able to develop a new community here. 

Historically, standards often arise out of collaborative technology communities, and the shared technology snaps to the standard once created, enabling all the community participants to rapidly develop products.  Here, the development community is coming somewhat out of order, behind the specification, but lets set that aside for a moment.   

So what will open source Java mean?  First remember that this effort has been driven not by Sun, but demanded by the community around Java.  Whether the demands are valid or not, or indeed politically motivated or not, there are still good things to be had from the process. 

Sun has driven the Java standardization process through the JCP for some time and has a strong collaborative community and process, supported by a strong certification and branding program.  (The certification will always protect the brand and expectations of integrity, and Sun needs to finally stop the hand wringing over open source projects "drifting" from the specification.  It doesn't happen in the real world, and it won't happen with Java.) 

Delivering Java technologies as open source still makes sense, however, even if the standard has led the implementation so to speak.

As a primary reference implementation, it will provide the following benefits to the entire Java community, Sun included:

• It will harden the primary implementation for Sun's and the community's benefit.  Allowing others to tinker and explore will find new and interesting problems.  They can be addressed.
• It will enable new innovation.  Many claim Java's day is done.  Allowing new implementors to explore the primary production base will invariably lead to new ideas and innovation on the platform beyond JCP participants, and to everyone's benefit.
• As new code enters the source base, it will likely come in at a very high level of quality.  Even if Sun developers act as the primary committers for the foreseeable future, a high level of inspection will be brought into play on code coming in from the outside.  For new work delivered from the inside, the inspection will likely be equally vigorous from the community.

That is not to say there won't be challenges.  As with any large code base that exists in a commercial product, all will need to be inspected carefully from a number of angles.  Sun won't want any obviously embarrassing code released, but as well, they need to ensure all code licensed in from outside can be released, and manage that process. 

Sun already supports a strong development community around OpenSolaris, and hopefully that experience can be leveraged by the “OpenJava” team.  (Certainly the questions I asked of Sun people tonight indicated the Java folks were absolutely sharing notes with the OpenSolaris community.) 

Likewise, Sun already supports a strong collaborative community in the Java Community Process, so they have a great channel to begin their open source efforts when they figure out how they intend to publish what sources. 

Perhaps it would be better to contrast this against other similar efforts.  Microsoft developed C# and the Common Language Runtime (CLR)  of .NET as a product first.  The specification for the C# language, and the Common Language Infrastructure (the CLR plus the base class libraries) was then taken through ECMA International, an industry consortium, to develop a standard, and then through the international standards process at ISO. 

During the same period as the initial standardization, a team at Microsoft created a shared source project around the standard, aka Rotor, providing what was essentially a reference implementation of the standard by cutting down the product code base to the standard.  The project, however, was not open source.  It’s license restricts commercial distribution.  It was essentially developed to drive academic research. 

Neither are changes accepted back from the community around this "Shared Source" project.  The code base is still linked to the product code, and Microsoft has concerns about intellectual property taint.  It’s unfortunate.  One of the first contributions from the outside world, within a day of the initial release of a million lines of code of implementation and test harness, was a fragment of code that improved the just-in-time compiler performance by 10%.  It of course couldn’t be accepted.  Nor the first robust bug fix the was contributed the next day.  The community quickly got the message and stopped contributing. 

A second implementation sprang up after the ECMA standard completed.  The mono project was started by Ximian as a complete ground-up implementation of the ECMA C#/CLI standard, and distributed as open source software under the GPL.  Ximian has been since acquired by Novell. 

Microsoft continues to position mono as “an interesting science experiment” in its market commentary, despite its growing success, and to maintain an arm’s length ambiguity about the state and status of non-essential patents that may be infringed by mono. 

Instead of encouraging multiple implementations of a standard they instigated, they discourage them.  Instead of embracing open source collaborative development, innovation and contribution, they keep the community hobbled.

So Sun has a huge opportunity to “do it right” with Java. They began the release of Java EE 5 with the GlassFish project, and continue the work in the context of a culture shift that has delivered OpenSolaris.  Now time will tell if they can harness all their collective experience in open source software, standards, and the JCP to bring about a complete open source Java world.

Post to del.icio.us

Sun's Changing of the Guard and Open Source Software

I've been interviewed several times in the past couple of days for my opinions on Sun's changing of the guard from McNealy to Schwartz (CNet article), and what that means for open source.

First, Jonathan Schwartz has been running point on executive announcements from Sun on open source participation and initiatives for some time now, so I wouldn't expect his support for such initiatives to change as he takes the helm.  Indeed, Sun has a long history of involvement with collaborative development since long before we called it open source software.  SunOS was a BSD derivative.  They've participated in Apache, the Mozilla project, and made substantial contributions to the Gnome desktop with respect to accessibility. This was all before the creation of their own projects and the release of OpenSolaris. 

Second, Linux has certainly caused Sun no end of grief over the past 5-6 years as customers moved from Big Iron UNIX systems as represented by large multiprocessor SPARC systems running Solaris to scalable server farms of less expensive Intel-based systems running Linux.  Customers wanted to support adding compute power by the slice at marginal extra cost.  Sun dabbled with Solaris on Intel and that failed because no customer wanted it.  Solaris was synonymous with Big Iron UNIX.  Customers that wanted Intel server based "UNIX" weren't going to call Sun, and Sun's field probably wasn't getting compensated properly on small margin sales of Intel Solaris.  This was as brilliant as DEC trying to sell PCs (on several separate occasions). 

Third, while Linux was the competition, Sun didn't think open source software was the competition.  Creating the OpenSolaris community earlier, however, wouldn't have helped. The entire company was still tooled up around its high margin big iron business.  The creation of the OpenSolaris community is the right answer as the company moves forward with the new hardware lines, designed specifically for web-centric compute loads.  The OpenSolaris community has been highly successful by all accounts (March newsletter, Redmonk analysis).  It doesn't compete with the Linux community, and was never intended to be that sort of play.  It is a direct conversation, however, between Sun and its customers and partners as they develop the next interesting architecture play.  (The Network is the Computer from the customer's perspective — and Sun will have the next computer to support specific sorts of Network workloads.)

Sun has the opportunity to reinvent itself around the new hardware line.  They have made such changes in the past: they started as the desktop UNIX workstation company.  As the Internet boomed, they became the big iron UNIX server company.  McNealy was around through that transition and is still certainly engaged with the company's daily operations as head of Sun Federal.  That's not to say there aren't big challenges ahead and some of them will no doubt be painful. 

But really, it's all good news.

Post to del.icio.us
digg it