Once More unto the Breach

On "open", standards, and open source software

I pull a small news feed from ConsortiumInfo.org (in the left hand column).  On a monthly basis they also produce a bulletin that covers some convergence of standards and other relevant topics.  The March issue is all about "open" as it relates to standards and open source.   Andy Updegrove provides a short collection of balanced articles and opinions as always.  What provoked this issue was the protest letter from the open source community leadership on the new OASIS patent policy that I've also written about here.

The protest letter still bothers me.  To be clear I am not refering to money with Corwin's Razor, so much as the desire of a group of people outside a community placing economic requirements on the community without actually contributing.  Imagine for a moment any of the following scenarios:  a collection of enterprises (think something like POSC, the Petroleum Open Systems Consortium) or comp. sci. academics (from CMU SEI) or lawyers (the U.S. National Society for Public Interest Law maybe) starting to go from OSS project community to OSS project community delivering an ultimatum that the customer wants to acquire only "professional" open source (for some group centric definition of professional) that must be licensed under the CPL (because it's clear about patent licensing rights and viral like the GPL) and the project must use subversion for all it's configuration management, and have module level documentation in Z notation to demonstrate its "correctness" in a "defensible" manner, for equally group centric definitions of correct and defensible.  I can imagine the response to such an ultimatum from the leadership of any of the communities. 

Updegrove's articles ask us to take the necessary steps to understand one another's points of view then start the dialog to at least learn from one another, and hopefully find the compromise positions (and rightly assumes there will be more than one).  This is a community oriented idea if ever there was one. 

Perspectives on Microsoft's Shared Source Initiative

I wrote this O'Reilly Network essay on perspectives on Microsoft's Shared Source world a couple of weeks ago based on a discussion with Nat Torkington.  Since then Jason Matusow (Microsoft, Director of Shared Source) has blogged his view on his Shared Source world, and Stephen O'Grady (Red Monk) has blogged a different perspective. 

Doomed to repeat it ....

I just found the ZDNet article "EDS: Linux is insecure, unscalable" via slashdot.  It's funny.  But I wasn't sure how to title this blog entry:

• "Four Year Old Rhetoric"
• "Those that don't pay attention to history are doomed to repeat it"
• "It didn't stop us choosing UNIX when *I* worked at EDS"

The thing that is stunning is the level of tired vapid vendor rhetoric here.  Are we surprised at the opinions of an alliance that contains Sun (really Solaris is better than any Linux), Microsoft (Windows Über Alles), and Oracle (Pay no attention to the MySQL behind the curtain)  — probably not. 

The interesting bit for me comes from personal history.  I was at EDS in the late eighties.  Our core team knew more about VAX/VMS than many places.  (We weren't learning much at DECUS any more.)  We worshipped at the Wall of Orange.  And we were very tired of the DEC sales team trying to push us up the hardware life cycle to support their market capitalization, and the licensing arguments when we tried to use cheaper memory, or disks-other-than-Digital-disks. 

Almost any vendor's "UNIX" in 1990 was less secure, less scalable, less robust than ANY other mini-computer of it's time, and especially VMS.  But every vendor had one.  And so as much as we thought VMS was an incredibly engineering friendly product that we knew inside out, and the obvious deficiencies in the UNIX systems of the day, we started writing new applications on UNIX systems, and migrating other applications over, using the [then] new POSIX.1 and ISO/ANSI C standards.  We didn't actually care about Digital's market cap.  They unfortunately didn't care about our systems problems.  They made the mistake of forgetting that as customers, it was our money, and that they weren't "special", even if they had superior technology to sell. 

A lot of people don't realize that the POSIX Wars weren't about standardizing UNIX.  It was standardizing the mini-computer.  Customers like choice.  And as Christensen points out, when a market begins to be over-served by the dominant vendor, the call for standards happens to encourage implementation choice. 

Update:  Always good to see confusion reign supreme in a large organization.  Here's EDS stating how great Linux is as well, both on its own site and in the news from Australia. 

An Interesting Juxtaposition of Patents and Microsoft

Two interesting reports appeared in the past few days.  First there's a good article on Microsoft encouraging patent reform.  As a company spending outrageous amounts of money on lawyers filing patents, it makes sense to reform the system.  The top companies applying for patents in our industry tend to top the 2000-3000 applications mark.  At an estimate of $12,000-$15,000 per patent to the lawyers for writing clever applications, that means these companies are each spending on the order of US$40M/year to add to the patent portfolio.  Sort of takes your breath away. 

Then they get into the other side of the coin so to speak with their intellectual asset strategy, and how they use intellectual property to best advantage.  As reported in the article, Microsoft is in a fairly constant state of patent litigation, much of it inbound.  Deep pockets makes one a good target. 

Among the patent changes Microsoft proposed directly quoted from the article:

• Halt the diversion of fees earmarked for the U.S. Patent and Trademark Office (PTO) to other government uses.
• Allow third parties to submit "prior art" information to patent examiners during the patent process itself, rather than only after a patent has been issued.
• Allow third parties to challenges patents "administratively," rather than just through litigation in order to help weed out questionable patents, as currently is permitted in Europe.
• Create a special court that would consolidate and hear all patent cases at the federal district level in order to improve consistency and predictability of patent litigation.
• Reform the standard used to measure "willful infringement," via which claimants are currently allowed to collect treble damages.
• Increase "harmonization" and collaboration across international boundaries.
• Move to a "first-to-file" system, rather a "first-to-invent one," thus following the patent procedures to which most other countries adhere.

While the list of proposed changes is pretty good, they tend to focus on getting good patents out of the system.  It's unclear that they address the problems around the length of time to get a patent out of the system, which means they will still be caught so to speak shipping products not knowing whether they're infringing or not.  So they may still be getting sued as much.  While such litigation may close more quickly under a reformed system, there is still that pesky need for lawyers.

The other report of note in the space is Stephen O'Grady's blog discussion of Microsoft and patents around the Mono project.  He points out that it is conventional wisdom that is getting in Microsoft's way with respect to things they could do with patents to support mono. This would encourage C# to be seen as a cross-platform environment as the language continues to be hammered by the likes of java and cross-platform open source environments.  Indeed, why can't Microsoft make the same sort of strategic intellectual property play that IBM has recently.  There are even more creative intellectual asset plays Microsoft could make in this space.  Unfortunately, they are bound by the very practices that made them successful.  And this might be why they see patent reform as more better patents (not necessarilly) faster as the way forward, rather than getting truly creative in their intellectual asset strategy. 

 

Does not Compete

Everyone has opinions about the next wave of open source software investment.  Obviously so do I, as I came to Optaros.  Unfortunately, everyone sees the words "services" and "open source" and assumes there is but one sort of service.  It would be like confusing a medical doctor providing a medical service, and a lawyer who specializes in medical law providing a different sort of service.  The Mercury News Silicon Beat reported "Optaros takes on SpikeSource, other open source start-ups".  This isn't so. 

SpikeSource is delivering a LAMP stack with particular support and maintenance offerings.  SourceLabs is delivering a tested and certified stack, focusing on the rigor of the integration testing.  These companies can be compared to RedHat and SuSE in terms of business model, product offering differentiation, and differentiated positioning.  (When you look at SuSE the positioning is all about ease of management as opposed to RedHat positioning around security and being current.)

OpenLogic is delivering a tools architecture for open source development, and it's services  are around supporting and maintaining its products, and training, customization, and consulting on the development space using their products.  Gluecode is delivering a similar open source offering differentiated towards the java developer.

Optaros is a traditional consulting and systems integration company delivering applications level solutions using open source regardless of underlying platform or toolset.  We don't compete with SpikeSource.  Or SourceLabs, Gluecode, OpenLogic, or anyone else delivering open source products.  Indeed, if they all succeed wildly, it will make our jobs easier satisfying our customers. 

Updated 03-14-2005, 13:41 PM:  The Mercury News Silicon Beat has updated their article here.  Thank you. 

r0ml Joins Optaros

I am very very happy.   r0ml joined Optaros today.  This is why I love startups.  Paradoxically, you have the opportunity to reach more people than you can in a large corporation.  ("Reach" is a myth in large corporations.)  And you get to work with your friends.