- Words to live by
- Taking Excel to the Grid?
- Just what constitutes an open source company? (via RightNow Technologies)
- OSDL Desktop Linux Client Survey Report Coverage
- OSDL Desktop Linux Survey Report
- Learning from the entertainment industry
- If the software market is saturated...
- How many software jackpots can there be?
- The Windows Idiot Tax
- The politics of control
November 29, 2005 | Comments: (0)
Business 2.0 has an interesting collection of wisdom from some of the business community's biggest names in this month's issue. My favorites? Penn Jillette and Eliot Spitzer (though this man is so full of himself that it's hard to take him seriously...until he files charges against you :-).
This was one of the few things I learned at Stanford Law School: you never want to have to enforce a contract, because the clear interpretation of a contract becomes much less clear the minute both sides stop agreeing on that interpretation.Make Deals With People, Not Paper
PENN JILLETTE, magician, author, and producerThis was the hardest thing to learn when I was 19. When we first started doing Penn & Teller shows, I thought that if you had a contract, it was enforced. I thought there were the contract police -- so you'd sign a contract that says you're going to give me a million dollars, and if you don't have a million dollars, someone will step in and give me my million anyway. Right.
That's one of the hardest lessons for a guy like me who has no interest in business but now runs a multimillion-dollar enterprise. A contract is not much of a legal document. It's just an agreement that two people who trust each other have made. You can't enter into a contract with anyone that you wouldn't make a handshake deal with, because everything comes down to a handshake deal.
The more experience I got in showbiz, the less I read the contracts. Now I don't bother. If I can't make the deal in a phone call, and have them understand it, then it's not a worthwhile deal. You're making a deal with the people, not with the contract. That's a mistake that people make a lot: "We've got it in writing now." The contract is clarification, but it's not enforcement.
Which brings us to Mr. Spitzer's counsel:
Never write when you can talk. Never talk when you can nod. And never put anything in an e-mail.I need help with this last one. I write waaaaaaaaaayyyy too much.
Posted by Matt Asay on November 29, 2005 09:46 PM
November 29, 2005 | Comments: (0)
I have become obsessed with grid-ifying massive Excel models that we use for financial analysis. This is a very real business problem in financial, retail and anywhere you are doing serious data analysis. We have models that take hours to run thanks to Excel's single-threaded design and we at the point of bringing down servers. Consider this a plea for help-and don't give me the "you should cluster the servers" response. It's not that easy, I need some kind of conversion mechanism.
I am ready to start an open source company to tackle the need to go from Excel to something else in a grid-in-a-box type of situation. The goal would be to make it all hidden from the end-user who just wants to increase processing horsepower. I figure you make it open source and you'll get people to use it and add extensions to other apps. I think it could be built on Globus.
Let's start a company. I know lots of VCs. Email me if you are interested.
Comments and ideas are welcome.
Here are a few things that I have come across so far and why none of them really work for me:
Platform Computing and Data Synapse: Too expensive and require mucho pro services to get up and running-but they do solve the problem, essentially acting as a plug-in to Excel that distributes the workload.
Excel Grid: A .Net plugin that processes file onto a Grid with Gridbus middleware. A few problems: runs on Windows; development seems to have stalled; many other grid components needed.
Turbo Excel: Haven't tried it yet, but could be helpful. Turns excel into C++ so that computers can process faster. Seems like the best idea-now to do it on a Grid.
Clustering/Beowulf etc: Unfortunately requires developers (defeats my hidden from end-user goal) to redo applications.
Azul Systems: Love the idea of a compute appliance, but don't need Java processing.
Posted by Dave Rosenberg on November 29, 2005 08:04 PM
November 29, 2005 | Comments: (0)
Just what constitutes an open source company? (via RightNow Technologies)
I received an unexpected, but welcome, surprise in my email in-box today. Greg Gianforte, CEO of RightNow Technologies, emailed me to suggest a broader definition of an open source company. I've blogged this before, but not with the thoughtfulness that Greg's email evinces. (Tim O'Reilly has written on this before, as well.)
Is RightNow any less an open source company than Red Hat, which layers a "proprietary" certification/service offering on Linux? Or SugarCRM, which layers proprietary bits on open source? Or Novell? Or IBM? Or Alfresco? Etc. It seems to me that the only appreciable difference may be in how these companies choose to market themselves, and how they choose to provide incentives to prospects to become customers.
Anyway, to Greg's email. Thanks for letting me post it, Greg. RightNow is clearly in good hands.
Matt, Imagine a car company that got its steel for free. So, instead of spending its money on costly raw materials, this company could invest in high-value differentiators such as better vehicle design, build-to-order manufacturing, and superb customer service. Now imagine that this company also offered customers completely care-free car ownership. Its customers would never have to worry about gas or oil changes or insurance, because that would all be taken care of for them. All they'd ever have to do would be to get in and drive. Obviously, that would be a great deal for the customer. Plus, with its cost-of-materials reduced to zero, the company could operate quite profitably while offering its extremely compelling value proposition. This, in a nutshell, describes the business model of an on demand software vendor using open source technology. By eliminating the cost of databases, operating systems, and other infrastructure components, open source technology allows an on demand vendor to invest more in development, hosting, and services. And by providing software as a service, the on demand model frees customers from the valueless, budget-sucking burdens of IT ownership. In other words, on demand software delivery is an extraordinarily effective way to monetize open source.
To understand how the "water" of open source is transformed into the "wine" of business value, let's first review the case for on demand applications. Customers are embracing on demand applications because they prefer to spend their money on application functionality (which has lots of value to the business) rather than the ownership of IT infrastructure (which has none). On demand vendors enable customers to achieve this objective by hosting and managing the supporting infrastructure for the application-delivering functionality where and when it's needed via the Web.
The customer doesn't care which operating system or database the on demand vendor is using in the hosting facility, as long as the application is scalable, reliable and secure. So on demand vendors are free to leverage open source solutions such as Linux, MySQL, and Apache to keep their infrastructure costs low-and thereby invest more in important value-adds such as R&D, multi-version hosting architecture, and implementation support.
RightNow offers a prime example of how this formula works. We've pioneered a wide range of CRM innovations, especially in the way our software automatically learns about customers from their behaviors. We've built a uniquely sophisticated hosting environment that has supported over 1 billion customer interactions on behalf of our clients in the past few years at 99.98% reliability. Plus, our enterprise-class hosting capabilities let customers upgrade when they want to-unlike other on demand vendors that force their customers to upgrade simultaneously. We allocate significant resources to ensuring the success of our customers through a highly differentiated engagement model, closely tracking the effectiveness of their implementations and pro-actively pinpointing opportunities for improving their ROI. We've also grown revenue for 31 consecutive quarters and have been profitable since we went public.
All of this has been made possible because we decided to use open source-rather than become just another distribution channel for monopolistically priced technologies from Microsoft and Oracle. As operating systems, databases, Web servers, and other infrastructure components become increasingly commoditized, there is simply no good reason to pay through the nose for proprietary solutions that offer no discernible functional advantages over their open source counterparts.
So, while RightNow is not an open source vendor per se, it's hard to name a company that's more effectively taking advantage of the economics of open source.
Every business should strive to maximize the value it delivers to the customer and minimize the cost of doing so. The combination of on demand and open source allows us to do exactly that.
greg
---------------------------------------------------
Greg Gianforte, CEO and Founder
RightNow Technologies
Posted by Matt Asay on November 29, 2005 07:15 PM
November 29, 2005 | Comments: (0)
OSDL Desktop Linux Client Survey Report Coverage
One thing that the survey report has shown is that people are far more interested in deploying (not just talking about) Linux on the desktop than I had originally thought.
Users, peers lead firms to consider desktop Linux
Companies have been slow to embrace Linux as a desktop operating system, in sharp contrast to the rapid adoption of the open-source software as a server operating system. But that situation is set to change, with market analyst IDC forecasting that 17 million PCs worldwide will be running desktop Linux by 2008. At present, Linux has a 2.5 percent share of the desktop operating system market based on shipments, according to IDC, with that figure set to reach 9 percent by 2008.
Trying for the Top: Desktop Linux Aims for the Big Time
OSDL has reached out to other open-source organizations to work in a small group of less than a hundred developers on how to dramatically improve the Linux desktop.
Fans aim to get Linux on desktops
OSDL has two goals for the gathering: identify areas that need improvement; and boost collaboration among the various groups that make versions of desktop Linux.
What Users Want from Desktop Linux
Download the full report.
Posted by Dave Rosenberg on November 29, 2005 11:05 AM
November 28, 2005 | Comments: (0)
OSDL Desktop Linux Survey Report
The OSDL Desktop Linux Client Survey is live for you to download. Since I wrote it, I offer a snippet here for your enjoyment. More insight tomorrow.
Top reasons for deploying Linux on the desktop:
1. Employees requesting Linux (user demand)
2. My competitors have successfully deployed Linux
3. TCO (Total Cost of Ownership)
Top inhibitors of Linux desktop adoption:
1. Application support
2. Peripheral support
3. End User Training
I did a few interviews on the report already. Check out eWeek's write-up.
Posted by Dave Rosenberg on November 28, 2005 07:55 PM
November 24, 2005 | Comments: (0)
Learning from the entertainment industry
When I've spoken at Open Source Business Conference, LinuxWorld, OSCON, or other open source conferences, I like to use the entertainment industry as a perfect example of what is happening to the software industry, or should. With every new technology - from the player piano to radio to cassette tapes to CDs to p2p file-swapping - the music industry has fought technological progress (and consumer convenience). The movie industry has been the same - from The Jazz Singer to betamax to DVDs to simultaneous day-and-date release - Jack Valenti and Crew have fought technology progress (and consumer convenience).
Fortunately for both parties, they've lost. Today, the movie industry makes a large majority of its money from the rentals aftermarket (60%+). The music industry? The shift to the dreaded CD, as just one example, vastly inflated the size of the market. It's unclear, given these industries' tremendous errors of judgment over the past few decades, that every one of them isn't fired. (Perhaps starting with the Sony executive who came up with the brilliant "rootkit" idea. They win the Bozo Award for 2005.)
In software, as you can see from this slide I use, it's no different. However much Microsoft et al. may want to control how/when/why software is used, the world is moving past them. If they're lucky, they'll get slapped around, get wise, and then get on board. If they're unlucky, they'll get mowed over.
Back to the entertainment industry (this time, the movie industry, in particular). As December's Fast Company reports,
The curtain is rising on Hollywood's "third act," and the drama is most stunning when you briefly recall the plot points of the first two. In the Old Hollywood of the '20s through the '40s, the seven studios...exerted awesome control ovre talent as well as movie theaters. Every week, about two-thirds of all Americas went out to the movies to see...well, whatever was playing at the neighborhood movie house. Because the moguls tightly controlled the costs of making films, and since they hardly had to spend on advertising, nearly every release made money....In the software context, this is roughly analogous to the shift away from large up-front license fees to an emphasis on maintenance and support revenues.The Old Hollywood system fell apart in the '50s when the studios were forced to divest the movie houses and let the stars become free agents. The studios' fall was accelerated by the rapid and inexorable rise of television, which decimated the movie-going audience. But eventually a new Hollywood emerged, and it was as brilliant and astonishing an invention as the original. [The industry consolidated into massive conglomerates] that combined them with TV and cable networks, cable and satellite connections, and video-rental stores. Now...the oligarchs have a 96% market share of the box-office take. They collect 98% of the national prime-time television advertising dollars. And they have 80% of the subscribers to pay-TV services....
This system has been imploding for two reasons. First, it depends on mobilizing a mass audience through mass marketing, and that's a daunting task these days [given how fragmented the viewing audience has become, with so many competing options]....Second [and hugely important for open source, nearly every movie loses money on its theatrical run these days....The theatrical release has become a loss leader to promote the studios' real money-maker, the DVD, which in turn often serves as a loss leader for Wal-Mart to draw in shoppers.
In other words, the software industry is shifting onto open source's home territory, and the smaller, more nimble competitor with a ready-made overhead to live off this kind of revenue stream will win. Ties goes to the disruptor.
Posted by Matt Asay on November 24, 2005 12:03 PM
November 23, 2005 | Comments: (0)
If the software market is saturated...
...then open source may well be succeeding because it allows for "small bites" of IT. Subscription-based business models fuel this. So does free and open access to the software in advance of a buying decision. Ease of access and ease of payment makes for ease of adoption. In a saturated market where enterprises have already bought more IT than they know what to do with (NIck Carr has been interesting and controversial on this), rip-and-replace really isn't an option. But incremental additions through open source? Yep.
Posted by Matt Asay on November 23, 2005 11:15 PM
November 23, 2005 | Comments: (0)
How many software jackpots can there be?
It's late and the pumpkin pie is still baking (egg nog, by the way, is the secret to a killer pumpkin pie), so I'm reading Fast Company opine on Mark Cuban's foray into the movie industry (December 2005 - not yet online). Cuban's partner, Todd Wagner, says something that struck me as similar to what's happening in the software industry:
The duo [Cuban and Wagner] admit there was no initial plan to build a vertical media machine. The acquisitions [of movie theatres, production companies, etc.] were designed to tip the movie industry's daunting odds in their favor. "If you just invest in [producing] movies, that's like going to Vegas," Wagner says. "You might get lucky and win at blackjack. But if you go to Vegas often enough, the casino's going to beat you." With 2929 Entertainment, he says, "we wanted to start making ourselves look more like the house. We could make a movie, put it on our screens, play it on our TV stations."After an interesting lunch conversation with Charles Forelle of the Wall Street Journal today, I'm wondering if this doesn't describe the software world (including its nascent open source component) perfectly.
Larry Ellison certainly seems to think so:
"We are no different than any other industry," he said. "You have good years ... and suddenly demand falls off, and there are more companies, there are more sellers than there are buyers for technology."More vendors than buyers. Supply exceeding demand. Thus leading to heavy consolidation such that only a few vendors with mega-operations can more efficiently serve customers.
I'm not a big fan of big, but there's something to this. Name the top 10 largest software firms. Now name the next 10. The first time I did this I was shocked at how quickly the numbers fall. Microsoft, IBM, Oracle, SAP, and then you plummet to those few scraping $1 billion in revenues, and fall off the $1 billion cliff to find a maze of much smaller companies. For all the talk about amazing margins and big money in software, there really aren't that many companies striking it rich in software.
Including open source. This isn't a critique of open source, but rather a defense. I've wondered before about whether we over-hype open source by making believe that there will be thousands of winners. It seems that it's not a question of over-hyping open source, but rather of over-hyping software companies, generally. And not because there's something wrong with software (or open source), but rather because the world doesn't need 100 $1 billion software companies. There's not enough demand.
All that said, I'm witnessing firsthand surging demand for open source in the enterprise. It has been surprising in its scope, scale, and depth. Perhaps this means my employer, Alfresco, is one of the chosen few. Perhaps. But I think it's actually an indication that the buying community is tired of being treated as a captive consumer. "I'm Vendor X and I own this market - take my product or leave it."
Again, this is something I've witnessed firsthand at Alfresco: large, Fortune 500 companies approach us (and, I presume, our open source peers) as much for the open style of doing business that open source engenders as for our code. They want to take smaller bites of technology, at their own pace, rather than sucking down multi-million dollar feasts.
So, I still think there's opportunity in open source for consolidation. I don't view consolidation as a bad thing. As with Mark Cuban in his rage against the Hollywood machine, sometimes you have to play by the industry's rules - slightly tweaked in your favor - to win. The important thing is that open source companies not lose touch with the communities that keep them honest and make them different - consolidation of technology to the detriment of community is a losing strategy.
Posted by Matt Asay on November 23, 2005 11:07 PM
November 22, 2005 | Comments: (0)
For those who still believe that running Windows instead of Linux is cheaper or more cost effective let me give you a real world scenario I discovered today.
We have an application that runs on Windows and needs a new staging environment. I set out to order two simple HP servers. Once again I was dumbfounded by the additional costs associated with simply trying to run a server. So, after chastising one of my team over the fact that we use anything from Microsoft I did a simple cost analysis and came up with hard evidence of the Windows Idiot Tax, the extra costs imposed by using Microsoft products on a server.
| Windows | Linux |
| Server hardware 2 x $1250 = $2500 Windows Licenses 2 x $700 =$1400 User Cals 3 x $30 = $90 SQL Server 2000 Standard 1-cpu $1850 Windows TOTAL= $5840 |
Server hardware 2 x $1250 = $2500 RedHat download = free MySQL = free (for clarity, this is non commercial use) Linux TOTAL=$2500 |
If you are wondering about the 3 user CALS, you have to buy packs of 5. So I have some spares if you need them.
Plain and simple, Microsoft is punishing it's customers for choosing their products. The Idiot Tax applies to many more Windows scenarios than just the server software. Think about the extra costs for virus protection, recovery from IE exploits, and heaven forbid you dare to cluster SQL server. I'm all for capitalism, but this is abuse.
Here's another interesting thing I came across recently: a number of people kept telling me that doing offshore development with .Net was cheaper than doing Java. And they assumed that the TCO would be less as well since that is what the offshore companies were pushing. TCO aside, I finally realized that what they meant was that the offshore companies charge less for .Net than they do for Java. I don't want to start a conspiracy theory, but you think that Microsoft might be seeding developers with Visual Studio? You think maybe Microsoft has convinced the offshore groups to push .Net? I know of a number of large financial institutions who are offshoring and seemingly overnight decided to switch to .Net for development coincidence? I think not.
Every minute that businesses rely on Windows a penguin sheds a tear.
Posted by Dave Rosenberg on November 22, 2005 10:14 PM
November 22, 2005 | Comments: (0)
I'm still plowing through David Daniell's biography of William Tyndale, translator of the Bible into English (from Greek/Latin/Hebrew). As I read tonight, I came across this quotation from the Bishop of London, Cuthbert Tunstall, commenting on Tyndale's (and Luther's) translation:
Many children of iniquity, maintainers of Luther's sect, blinded through extreme wickedness, wandering from the way of truth and the Catholic faith, have craftily translated the New Testament into our English tongue, intermeddling therewith many heretical articles [though he was never able to name one, except when taken out of context] and erroneous opinions, seducing the common people; attempting by their wicked and perverse interpretations to profane the majesty of Scripture, which hitherto had remained undefiled, and craftily to abuse the most holy word of God, and the true sense of the same. (190)For such alleged misdeeds, the Protestant believers were "scourged with whips, to make [them] revoke [their] opinions" (185) or, in the case of Tyndale, were burned alive.
Why? Once you parse through the various writings and reported speeches of the high-ranking clerics and government officers (like Sir Thomas More, John Stokesley, etc.), the answer is plain: because translation of the Bible democratized access to God or, at least, to His word (as held by the Christians). It took the clerics somewhat out of the picture. And so they killed to keep control over the minds of the people. (Stay with me here. This has open source implications.)
Except once the genie was out of the bottle, they couldn't put it back in. As the priests persecuted the translators (and readers thereof), it sowed the seeds of its own undermining:
There is detectable in London a reaction of puzzlement. Politically, the Church, led by Wolsey, was in fact making a great mistake, sowing the dragons' teeth which would before long yield destruction, and not just through Henry's [VIII] divorce and the break with Rome. Luther had torn down thousand-year old walls hiding the Bible. The bishops in England were building them higher. London men and women were not going to be soothed with the sops of Nicholas Love, the best the Church could manage, with the addition of the Lord's prayer in English. Luther was giving the German people the Epistle to the Romans, raw.... (194)All of which makes me think of open source. There is a real struggle today over this same issue: how much control to leave to the common people (you, me, customers, partners), and how much to retain in the hands of vendors. It's not a matter of life and death, but it is a highly important political (and business) debate. Now that IT buyers have had a taste of the "open source Bible," will they be content to go back to "proprietary Latin" and "obfuscated Greek"?
On the one hand, you have the Luddites (who, as Dave Dargo reminds us, were not so much opposed to technology qua technology as they were determined to keep their jobs/livelihood intact). These are the companies that cling to their proprietary business models, hoarding their intellectual property and declaring that the commoners (customers) couldn't possibly benefit from it, anyway.
On the other hand, you have the "left wing" of the open source world who believe that everything should be free and open. Full, unfettered access to the "open source Bible," as it were.
The most interesting group, however, stands in the middle. These are the people and companies who are inclined toward open source, but cling to measures of control. There's a big part of me that wants to tell these people, "Get over it. Just abandon efforts to control the code, let go, and see what happens." But this is a scary thought, and it's not always practicable. And maybe it's not the right thing to do, anyway....
For example, in MySQL's case, it could surely go to a 100% "pure" GPL play. No dual license. The problem, however, is that this would immediately decimate its popularity (read: use), the company's profits, and we'd be back to Oracle, IBM, and Microsoft land. So, MySQL exercises some control over its code by retaining the right to dual-license, and give companies that want to embed the MySQL database a way to do so without open sourcing their code through GPL "infection."
I think the benefits of having MySQL widely used far outweigh any negatives that come from it being controlled by MySQL, and the code having the potential of being closed. Without this mechanism of control, MySQL would have a much harder time making money, which would limit the amount of development it could do, innovation it could foster, etc.
There are a range of open source companies today that give away most of their code, but keep some proprietary. In no case is this done because of a lack of open source commitment. In every case it is to help remind customers to pay, given that the products are good enough that they require little or no support after the first year. Thus, customers could take the code without giving anything (code or cash) back.
As we move up the stack, we're going to run into this "quandary" more and more. Applications are different from databases (often embedded) and a decentrally-developed operating system like Linux (which needs to be synthesized together, certified, and packaged). I'm not convinced (though I want to be) that the Red Hat model necessarily works higher up the stack.
Still, there's more Luther than Wolsey in me, and I'm trying to figure out if, in fact, the answer really is, "Just let go." What do you think?
Here's a thought, btw, which I just saw on the Hobu blog, and quite like:
Remember that the community around your open source project provides most of the value. The software is what brought people in, but the people make the project. Don't screw it up by thinking you can control them or by thinking that what you say is the final word. You get what you give.
Posted by Matt Asay on November 22, 2005 06:57 PM
November 22, 2005 | Comments: (0)
On Pentaho and buying community
This press release "Pentaho and Mondrian: Open Source Business Intelligence Consolidation Begins" passed by me a few weeks back and only just recently caught my eye.
Today, Pentaho Corp., the world's leading provider of open source Business Intelligence (BI) software, announced that the Mondrian On-Line Analytical Processing (OLAP) project is joining forces with Pentaho to become a permanent feature of the Pentaho BI project, which provides comprehensive reporting, analysis, dashboards, workflow, and data mining capabilities.
Pentaho is not without challenges. There is not a visible vibrant community (though I am sure someone will tell me otherwise.) They are building a BI on application on top of a set of open source components that they don't control (except Mondrian.) And ironically Pentaho uses JasperReports as their reporting engine.
It's also unknown if they acquired the Mondrian copyrights when Julian Hyde joined the company. While I think it's great that there is so much interest in open source BI, I fail to see how this move benefits the existing user base. It appears as if only Pentaho gets anything out of this.
I struggled with this same issue back when JasperSoft changed models into an open source company. But at least that was clearer in terms of the benefits for all sides. This one doesn't make much sense to me and in fact may be a detriment to those using Mondrian.
Posted by Dave Rosenberg on November 22, 2005 01:40 PM
November 21, 2005 | Comments: (0)
Microsoft-where open standards aren't actually open
Microsoft said it will offer its Word, Excel and PowerPoint document formats as open standards. But, can an open standard be owned by one corporation?
[Asay comments: I think they already do. Standard by consensus or standard by fiat - I'm not sure it matters how it comes about. (Well, of course it does, but reality is reality....) Jason Matusow's comments are interesting and important to this.]
On one hand I applaud the move as a strategic monopolistic tactic to protect market share, but on the other hand find it typical of Microsoft to pay lip service to open-ness while not giving up any control. Maybe they should have called it "shared standards"--oh, wait there is no sharing involved, hence my previous sarcasm.
I suppose it's better than nothing...?
[Asay comments: The real kicker in all this is that we've been promised XML before by Microsoft. And got it. Do you use it? Of course not. True, this time XML is supposed to be the default format for Office 12, but I wonder how many people will actually use it, given that most corporations will lag years behind and stick with Office 97, 2000, or XP. In short, does it matter?]
Am I the only one who didn't miss hearing from Microsoft's Alan Yates?
Microsoft misses the mark---again
Microsoft Leadership going down the tubes
Posted by Dave Rosenberg on November 21, 2005 03:28 PM
November 19, 2005 | Comments: (0)
It has nothing to do with open source (except the tenuous "community" linkage), but Pandora is one of the coolest "apps" I've used in some time. Put in the name of a song or a band that you like, and it starts feeding you songs/bands that have a similar sound. You listen to them in your browser (sounds great), and are one click away from buying them if you like them. (It lets you listen to the entire song.)
Consider it a highly tailored radio station.
It's very cool. Try it out. (I'm now listening to Death Cab for Cutie because of its help. Need to find a good cross between Smashing Pumpkins and The Smiths now....)
Posted by Matt Asay on November 19, 2005 02:35 PM
November 18, 2005 | Comments: (0)
EnterpriseDB CEO on Sun's support of PostgreSQL
EnterpriseDB Andy Astor knows my obsession with open source databases and anticipated my questions about how this might affect his company. Andy provides some insight on Sun's support of PostgreSQL below.
As you will remember, EnterpriseDB is an enterprise-class database that is built on PostgreSQL. We have enhanced PostgreSQL to create a database that is compatible with many Oracle applications and also better performing, and we price our product to reflect its open source foundation.A natural question to ask is whether EnterpriseDB now competes with Sun, at least to provide databases on the Solaris platform. The answer is "No", for the same reason that EnterpriseDB never competed with PostgreSQL itself. EnterpriseDB is substantially different from native PostgreSQL. Customers choose EnterpriseDB because those differences, including compatibility with Oracle applications and better performance, are important to them. In that way, EnterpriseDB's business model is different from the many other commercial open source companies that simply provide support for open source software.
I usually begin to describe EnterpriseDB's advantages by listing PostgreSQL's many merits, including its maturity and reliability and the huge open source community of talented developers that creates and maintains PostgreSQL. Sun's selection of PostgreSQL will likely mean that many more of our potential customers and others interested in EnterpriseDB will be familiar with PostgreSQL, so we'll be able to focus our communications efforts on the added value we provide on top of native PostgreSQL.Sun's selection of PostgreSQL to complement its high-end, enterprise-class operating system, rather than MySQL or another open source database, also highlights PostgreSQL's superiority over other offerings in terms of reliability, scalability and suitability for enterprise use. Again, because our product is built on PostgreSQL, we're pleased with that result.
That's my $.02 on the latest in a series of announcements, including recent news from Oracle, Microsoft, MySQL and CA/Ingres, that together illustrate the transformative effect of open source on the enterprise database market.
Sun's support of PostgreSQL is a good move, though a bit late to market. Nonetheless it shows that Sun is taking the right steps to be an open source company. I wonder if they should have left Solaris closed and opened all the sundry software.
Posted by Dave Rosenberg on November 18, 2005 08:44 PM
November 18, 2005 | Comments: (0)
Infectious licensing: helpful or hurtful?
Despite the title, Greg Vetter's paper on open source licenses [PDF] ("'Infectious' Open Source Software: Spreading Incentives or Promoting Resistance") has less to do with a rigorous analysis of the pros and cons of infectious licenses, and much more to do with an explication of what, exactly, the GPL/LGPL mean. Vetter does a great job on this and should be a required read if you're hazy on open source licensing.
Vetter does, however, make a claim that interoperability should trump infection in open source licensing:
Some estimate that infectious terms promote open source software growth by supporting community development norms and preventing proprietary poaching of the software, or converting proprietary software to open source. Even if some of these viably benefit open source, my analysis is skeptical that infectious terms, on balance, have beneficial effects. They adversely increase legal uncertainty for open source licenses and produce incentives detrimental to interoperability, compatibility, and coexistence between open and closed source software. As a result, open source licensing should precisely define infectious terms in order to support open source development without countervailing effects and misaligned incentives. (58-59. Footnote references omitted.)I agree. Once you move out of operating system land, it becomes increasingly important to clearly delineate how different software can interoperate with yours.
Indeed, it is the operating system (or rather, the Linux kernel) that most clearly exemplifies this. Linus Torvalds (in The Linux Edge, page 108) identifies his decision on linking (allowing programs to make normal system calls into Linux without being GPL-infected) as one of the most important non-technical decisions to contribute to Linux's success. (See Vetter 114, as well as 158-159.)
While we live in a heterogeneous software world (and I assume we will for some time, and probably forever), it's important that one's license reflects this heterogeneity. This is why, for example, Alfresco opted for the MPL (Mozilla Public License) rather than the GPL. We wanted a more clearly written license that "infects" individual files, and not the "whole" of a program (as the GPL does). As Alfresco is both an application and a platform, this was critical for us - we wanted to maximize third-party integration into Alfresco.
I believe there are excellent reasons to use the GPL (as I've argued), but easy embeddability into other applications is not one of them. Consider MySQL. A large percentage of its revenue derives from those who want to embed the MySQL database into their applications...and so pay MySQL (under its dual-license scheme) to get a non-GPL license to the code.
Does this mean MySQL (and Alfresco, and Sugar, and...) is not open source? Of course not. No less so than Linux. What it means is that such companies recognize, like Vetter, that the road to open source is not through force/compulsion but rather through interoperability.
Posted by Matt Asay on November 18, 2005 04:14 PM
November 18, 2005 | Comments: (0)
Job posting: Alfresco looking for Valley-based senior sales engineer
I'm looking for someone to join Alfresco's Silicon Valley-based team. Senior sales engineer willing/able to wear multiple hats while we're in startup mode. (i.e., Good with customers/partners (Demos, presentations, answer technical questions, provide partner training/certification), also probably provide some second-line support.) It would be ideal for the person to have ECM experience, but I'm more concerned with general technical aptitude and good attitude. It would also be great if they have some (even nascent) open source experience.
Most of all, I want this person to care as much about the company as I do. This is someone that will work closely with me on large customer and critical partner opportunities. I want them to be aggressive but not obnoxious. A hard worker but not self-absorbed. Etc.
In my experience, a good sales engineer can have a huge positive impact on sales. If you know someone with whom I should talk, please email me. I'm looking to hire this person before the end of the year.
Posted by Matt Asay on November 18, 2005 12:52 PM
November 18, 2005 | Comments: (0)
A matter of harm, not a matter of taste (Nick Carr on porn)
So, I'm a silly prude. I don't find porn ennobling (as unthinking feminists apparently do), titillating, or otherwise worthwhile. I think it is harmful, and not just to children. It has no redeeming value. It has no justification.
All of which makes me glad to see someone like Nick Carr attacking tech's cosy familiarity with porn. You've heard the drivel before: porn drives innovation in technology. Even O'Reilly's eTech conference shamefully had an entire session devoted to the issue. What losers we must be. Turning to Carr:
A week or so ago, the U.S. Senate held some hearings on pornography, including the internet's vast and various store of the stuff. Orrin Hatch, the Utah Republican, called porn a "problem of harm, not an issue of taste." Nobody, though, paid much attention to the proceedings. Popular blogger Jeff Jarvis, in a post titled A Nation of Hairy Palms, dismissed it all as "silly crap" from "conservative prudes."Well, I am a prude, and with good reason. And I agree with Nick. I'm not ashamed to say that, to the extent that technology is driven by a vulgar debasement of women (and men), I'm not sure I want to be on board. We can do better. Surely we can do better.Jarvis's reaction is typical of the blogosphere's, and, for that matter, the whole country's, laissez faire attitude toward online porn: Yeah, there's a whole lot of it out there, but it's basically harmless, even kind of amusing. Anyone who has the temerity to criticize it, or even call attention to it, is just a prude or a loser who deserves to be ridiculed and ignored.
Another common view of digital porn is that it's useful - as a case study for internet businesses. Paul Kedrosky, in a recent post, rehearses this theme: "I think that a valuable startup exercise would be to do a wholesale survey of all emerging technology in the promotion, selling, and distribution of online porn ..."
But maybe the most common reaction of all is simply denial. When Icann recently proposed setting up an online red-light district, under the .xxx domain, many politicians around the world, led by President Bush, attacked the idea, and Icann shelved the plan. Establishing a porn domain would have acknowledged the fact that the web is crammed with naughty pictures and videos. Without .xxx, we can pretend it doesn't exist - or at least distance ourselves from it.
I don't think I'm a prude (and I like to pretend I'm not a loser), but I'd like to suggest that internet pornography is bad. Very bad, in fact.
Posted by Matt Asay on November 18, 2005 12:30 PM
November 18, 2005 | Comments: (0)
Open Invention Network: Shamefully underfunded
LinuxGram is reporting that the Open Invention Network (OIN) - the patent-buying co-op founded by IBM, Novell, Red Hat, Sony, and Philips - only has $40 million to spend (despite initial intentions to credit it with $280 million). This is a shockingly low number, especially when one considers that Red Hat put in $20 million of it. That's a big chunk of change for Red Hat - it's putting its money where its (business) model is.
But what about the others? IBM, which had $1 billion to put into funding Linux, can't find a few tens of millions of dollars to protect that investment? (Of course, IBM no doubt continues to buy and build Linux/open source-relevant patents on its own, but apparently these are not to be shared?)
As for Novell, my alma mater, it's unclear what cash, if any, it put in, but it did contribute its $15.6 million CommerceOne patents. In-kind donations are always nice but, as I've learned with OSBC, cash is king.
Still, if it were me granting patents that would protect open source, I'd be putting those into the common pool that relate to an office suite that Novell once owned. Unless I'm much mistaken, Novell still owns highly relevant patents that would serve (and, indeed, serve today) to keep Microsoft at bay, because they go to the heart of Microsoft's Office cash cow.
Novell, however, isn't alone in owning patents that would be relevant to this open source and the OIN. IBM has them. Sony has them. Philips has them. (Red Hat probably doesn't, so it's exempt from this criticism.) Nokia and Yahoo!, who were also supposed to join OIN but backed out, also have relevant patents.
Let's call a spade a spade. $40 million is a symbolic gesture. It is not even a finger in the dike (if you remember the Dutch fable). These companies can do better. If they're serious about OIN and what it represents, fund it. Fund it with cash and with their highly relevant patents. If they're not serious, they should stop wasting our time and bandwidth on press releases.
Posted by Matt Asay on November 18, 2005 11:54 AM
November 17, 2005 | Comments: (0)
Intellisync gets snapped up - Funambol next?
News today is that Nokia is buying Intellisync to power its email service. What does this have to do with open source? Funambol.
Funambol, with 20,000 downloads/month of its Sync4j project, offers Intellisync-like sync functionality for any SyncML-enabled phone (which includes most phones on the market). Funambol is open source. Funambol is run by a Juventus (Patrick Vieira-stealing) fan, but that is its only deficiency right now. (And there's always hope that he'll come around to Arsenal. :-)
Now, I'm an open source bigot, so my perspective is suspect. But still I have to ask, Why would anyone want to build on Intellisync when they could build on a community? Why wouldn't you want to build on an open source, and open standard sync/email technology? It's not as if it's open nature would challenge your ability to charge for it. How many suit-inhabiting Wall Streeters do you think are going to download the code and set up their own RIM-like email service?
None.
Funambol is one of the most interesting open source companies today. If you're an Outlook user, you should head to its site to get set up on its sync service. Free, works great, and open.
What's not to love?
Posted by Matt Asay on November 17, 2005 05:13 PM
November 17, 2005 | Comments: (0)
Tyndale's Bible and Open Source
I'm in the middle of David Daniell's biography of William Tyndale, translator of the King James Version of the Bible. (Tyndale's work represents ~90% of the New Testament and the first half of the Old Testament.) 
Fascinating stuff, whether you're religious or not. As Daniell notes, the English language today is rife with Tyndale's phrasings ("the powers that be," "the spirit is willing," and "scapegoat," among others) and English cultural life (and, subsequently, the world's) was heavily influenced by the Bible (and Tyndale's translation of it.
It's hard to talk about Tyndale without bringing up words like "Christianity," "Bible," and "faith." I'm not going to try, and hope you'll excuse me. The commonalities between Tyndale's Bible and software's open source are fascinating, and worth a perusal here. (Besides, the disastrous effects on me aside proving the contrary, a little religion won't hurt you. :-)
In reading the book, I've been struck by a question: How does one person, against all tradition (and ultimately at the expense of his life), determine to Think Different (to borrow Apple's tagline)?
Tyndale thought very differently for his time. Consider:
Tyndale wrote in English. Most writing of his time was in Latin. The two most famous books in Europe, Sir Thomas More's fantasy known as Utopia and Erasmus's The Praise of Folly, were in Latin. To be a scholar and not to write in Latin was odd to the point of standing condemned....All university work and most printed books were in Latin; in 1605, of six thousand volumes in the Oxford University Library, only sixty were in English....Erasmus, so influential on his three visits to England, sharing the life of warm households, debating, staying in Magdalen [at Cambridge University], inspiring and being inspired, teaching Greek as Lady Margaret Reader at Cambridge, listening, commenting, writing letters, maintaining his friendships, spoke not a word of English. (45-46)In other words, there was an intelligentsia built up around Latin (and Greek) - inaccessible to all but the elect (many of them priests) - that circulated Big Thoughts without sharing any with the little people.
Small wonder, then, that economic and intellectual riches were held by very few - the social "anorexics" that starved themselves out of sheer snobbery and ignorance. Sir Thomas More, a bit of a snobbish putz, refused to have Utopia translated into English because "it might fall into the hands of the simple and unlearned who might misunderstand and take harm from it" (69). How very thoughtful of him.
This, of course, led to religious practices that had little to no foundation in the Bible, but which served the (Catholic) Church's needs. More from Daniell:
...[Tyndale's English Bible] could be read and understood, without censorship by the Church or mediation through the Church, as it was written to be read as a coherent, cross-referring whole. Such reading produced a totally different view of everyday Christianity: the weekly, daily, even hourly ceremonies so lovingly catalogued by some Catholic revisionists are not there; Purgatory is not there; there is no aural confession and penance. Two supports of the Church's wealth and power collapsed. Instead, there was simply individual faith...found in Scripture. (58)By making "public" the Bible, Tyndale enabled lay readers to discover Christianity for themselves, unmitigated (for better or for worse) by the Church. (Open source, as I'll explain below, accomplishes much the same thing, and with eerily similar results for the "Proprietary Church."
Back to Tyndale....Why did he choose to democratize access to the Bible?
...[T]he impetus for Tyndale to write in English when he began to do so had several origins: the expressive everyday phrases of the Vale of Berkeley [i.e., local language was rich - he didn't need Latin to express himself], and the Lollard pastoral concern that the Word of God should be in the vernacular; with...the example of Luther. (46)A message for the people in the language of the people. Sounds obvious, but it wasn't (and Tyndale was put to death for his heretical act of democracy).
And it isn't obvious today in software development, either. (Btw, if it feels like I've taken a major step down here in theme, it's because I have. However much we may want to pretend that software freedom is as important as intellectual and religious freedom, it isn't. Not even close). We still have our High Priests of C++/Java/etc. that want to restrict access to code. This is expressed in proprietary software companies holding back code (the equivalent of requiring it all be in Latin), insisting that "the Plebians" don't want it, wouldn't know what to do with it, etc. While this is almost certainly true on one level and at a given period of time (now), I'm not convinced that it's valid in the long term.
When Tyndale's Bible came out, most of England couldn't read it (including the clergy!!! See page 78). Today, hundreds of millions have read his translation, and billions have been influenced by it.
Perhaps increased access to source code will be "translated" into easier and easier ways of interacting and manipulating it, such that a time will come when even "code Plebians" like myself will be able to hack. (I believe this is already starting - I can't code, but I've learned how to use HTML tags for this blog, and use Google and the "View Source" feature in the browser to look at other websites to help me find pointers on how to do different tags all the time.) HTML, open standard. Websites, open source.
Ultimately, in its own sphere, perhaps open source will have an analogous effect to Tyndale's expectations for his translation of the Bible. A flowering of democratized knowledge or, in this case, software. As Tyndale told a learned man before leaving England to commence his translation:
"If God spare my life, ere many years I will cause a boy that driveth the plough shall know more of the Scripture than thou dost." (1)Let's hope so.
Posted by Matt Asay on November 17, 2005 05:06 PM
November 16, 2005 | Comments: (0)
Under the Radar Event: Companies I liked
Funambol-was already a fan, still a tough market, but with 20k downloads a month they will crack it sooner or later.
Soonr-supercool remote access to desktop apps and files via cell phone. One of the coolest things I have seen in a long time.
Akimbi-crazy new capture and restore virtualization thing
Coverity-Great product, boring preso.
I thought a few of the companies really sucked-or at least their presentation skills sucked which turned me against them. Lots of talk about open source, but really Funambol was the only open source company. A direct quote I overheard in the lunch area: "I downloaded the podcast and was able to realize the long tail that we're seeing with the growth of our business." Seems he forgot to mention Web 2.0 and AJAX, or maybe I missed that part.
If you want gory details or are afraid you were one of the suckers feel free to email me: openresource at infoworld.com
Posted by Dave Rosenberg on November 16, 2005 07:08 PM
November 16, 2005 | Comments: (0)
Hanging with Scoble at Microsoft Campus
I spent most of the day down at the Under the Radar event, hosted at the MS campus down in Mountain View. I feared holy water and cries of "Infidel" but my open source leanings were hidden underneath my Butterfly suit.
Before I actually headed into the company demos I was fortunate to spend an hour with Robert Scoble, who is single-handedly taking on the MS establishment and trying to shake the place up. If you don't already read the Scobelizer blog, you should. Not only is he a good writer (and typer ~100 wpm) he may actually have the most relevant observations of anyone in the tech industry. We talked a lot about Google, Web 2.0 companies and the fact that MS is so big they have had a hard time seeing the forest through the trees, while Google continues to build the forest around them.
Scoble contends-and I largely agree, that any startup that draws an audience that would look at advertising along side content could simply be feeders for Google, Yahoo or MS who are battling it out. I need someone to give me a stupid idea to build and sell to those three. Some kind of mash-up/map/podcast/ROI tool perhaps.
My biggest realization from our conversation is that I tend to think of MS as one giant company when really it's like 100 small companies all trying to be relevant-hence the market confusion and multiple personalities. That also speaks to the seemingly random announcements and focus, one minute it's Windows Live, then it's supercomputing-Scoble pointed out that there are so many projects going on that its difficult to manage the communications, which I agree with, but still they could do a better job with.
We also hung out with Rajesh Setty who just published a book called Beyond Code all about how developers need to become more socialized in order to move beyond just being part of a team. Then Stan Murawski from the MS Mainframe Group told me that he is trying to figure out a communitized tool set for users migrating from mainframes to Windows server system-he even laughed when I called B.S. on his attempts to give me the Microsoft party line. It was really interesting talking with those guys. I feel like it's gotten harder to have those conversations mostly because the conferences I tend to go to are all open source oriented.
Posted by Dave Rosenberg on November 16, 2005 06:47 PM
November 16, 2005 | Comments: (0)
I stumbled across an interesting post by Mike Bergman this morning. He starts by quoting an exchange at a Union Square session between Tim O'Reilly and Martin Nisenholtz:
NISENHOLTZ: I sort of call it the lottery syndrome. There was a Powerball lottery yesterday. Tons of people entered it. We know that someone won in Oregon...we also know that the chances of winning were one in 164 million....I guess what I'm struggling with is how we measure the number of peer production efforts that get started versus Wikipedia, which has become the poster child, the lottery, the one in 164 million actually works. Now it may not be one in 164 million. It may be one in 10. It may be one in 50, but I think that groups of people like [prominent Web thinkers] tend to create the lottery winner and hold the lottery winner up as the norm.Bergman then does just that, diving into the Sourceforge data to explore this "lottery effect."O'REILLY: Look at Source Forge, there's something like 104,000 projects on Source Forge. You can actually do a long tail distribution and figure out how many of them...but...I would guess that one in like...154 million are probably out of those 100,000 projects, there are probably, you know, at least 5,000 who have made significant reputation gains as a result of their work. Maybe more. But, again, somebody should go out and measure that.
Great data, with an interesting (implied) question: Do we sometimes overstate the success of open source (and Web 2.0, for that matter) by glomming onto the few winners, pretending that they are a widespread phenomenon? I think open source is a major trend, not a fad, and not relegated to a few chance successes. But I also feel that its future will be helped as we enable commercial open source to flourish.
Posted by Matt Asay on November 16, 2005 02:19 AM
November 15, 2005 | Comments: (0)
Microsoft rejects IBM strategy and open source 'dorks'
Via ZDnet--If this is not a joke, then we should all just accept that Microsoft is not going to change. The upside is with management like this maybe they really will go out of business.
Even as Microsoft continues with the most aggressive product rollout in its history with the launch of the Live series of hosted applications, it is showing no inclination to consider alternatives to its long-standing packaged software approach.IBM's on-demand model is "crazy" and Open Source is "really a developer phenomenon" that does not stand comparison with "customer experience of Windows Live", said Charles Fitzgerald, Microsoft's general manager of platform strategy, last week.
Open source is really a developer phenomenon that speaks to infrastructure. With Windows Live and Office Live we are talking about customer experiences, whether it is a personal set of services for Windows Live or things that are aimed at helping people grow and manage their businesses. I doubt if you talk to users of Office Live that they have any interest in dorking around with source code. This is about customer experience rather than developer experience so it's largely irrelevant.
I really don't even know what to say. The absurdity confounds me. I can't believe the Microsoft PR people allow this guy to talk.
Posted by Dave Rosenberg on November 15, 2005 02:36 PM
November 15, 2005 | Comments: (0)
OSDL launches Patent Commons Website
OSDL launched the Patent Commons Project website today. The Commons compiles more than 500 patent pledges housed in a central location where they're searchable by anyone. CA, Ericsson, IBM, Intel, Novell, Red Hat, and Sun are all supporting the effort.
The Patent Commons Project launch comes on the heels of the Open Invention Network announcement. It's good to see that a variety of steps are being taken to address the flawed patent system. OSDL is focusing on the patents that have been pledged publicly and that benefit all open source software. It then catalogs these pledges made by companies who want to retain ownership of their patents. OIN is buying patents that apply specifically to Linux and is cross licensing them with industry players.
At the end of the day, these efforts combined, along with others, begin to reduce potential threats to open source innovation. More info at http://patentcommons.org
Posted by Dave Rosenberg on November 15, 2005 12:20 PM
November 15, 2005 | Comments: (0)
My two weeks with Alfresco has opened my eyes somewhat to the perils of starting an open source "commercial community." The development of a community de novo for a commercial entity entails all sorts of pitfalls that an organic community simply doesn't need to be bothered with. Or, rather, they do, but the stakes (reputation instead of financial) aren't the same.
For example, when a commercial open source's project is in its nascence, it is fragile. Subject to death by boredom (no one cares about the project), fork, and misappropriation. If you're Red Hat with millions of downloads and a growing business, this latter concern is, well, not very concerning. But if you're a new player like SugarCRM, it is.
In fact, SugarCRM dealt with this very problem late last year when vTiger apparently misappropriated its code, causing SugarCRM to add an attribution clause to its license. Foul! Some in the community cried, and maybe so. But consider what SugarCRM was trying to do: protect its ability to develop a community. Not one devoid of parasites (these will always exist), but one devoid of parasites who do actual harm to the project SugarCRM set out to build (and, hence, to the code/community vTiger was trying to leverage).
Beyond community-dilution issues, there is the issue of monetization of one's project. It's easy to point to Red Hat as proof that open source business naturally thrives, but I think Red Hat is a poor example. The company is clearly kicking tail. But it's not clear that open source companies will share this fate simply by being open source. Red Hat was able to grow its brand almost completely unobstructed (Who at the enterprise level cared about Linux then? And what other Linux company had the kind of market valuation and consequent cash in the bank to spend years figuring out a business model and branding itself?), whereas so much attention is paid to open source today that few companies will be able to grow their brand unnoticed in a largely non-competitive environment.
This is not intended (in any way) as a slight on Red Hat. Rather, it's meant to give pause to those who think that open source is logically a good idea for this or that product niche that is still dominated by closed-source companies. When you have millions of downloads (JBoss, MySQL, Linux), it's OK to only monetize a small percentage of those. But if you're in an application space, you (almost definitionally) have a smaller potential audience. This means that you must do things to increase the conversion rate.
So, you see companies like SugarCRM, Alfresco, JasperSoft, and virtually every new open source company offering both open and closed-source software (with the closed-source bits meant as a way to entice enterprises to pay). This is one of the biggest issues for JBoss and other well-architected software - paying for support is fine and good for the first year or two, but after that, once your users are highly familiar with the system and able to support themselves? (Red Hat found a clever way around this by effectively locking customers into a support contract, inseparable from the bits. I think this was a stroke of genius.)
In sum, it's not easy to launch an open source company. It's not easy to get people to pay for value. It's not easy to get would-be competitors to play fairly. And so, open source companies today are experimenting with nuanced approaches to open source which reveal not a lack of desire to be good open source citizens, but rather a desire to promote open source without being mowed over by some aspects of open source licensing that may not be naturally conducive to commercial success.
We could say, "Tant pis!" but the better answer is to look for ways to help. The more companies, the more (and more diverse) the open source code out there. Surely this is a good thing.
Posted by Matt Asay on November 15, 2005 09:24 AM
November 15, 2005 | Comments: (0)
So, Host Marriott is buying $4 billion worth of Starwood Hotels properties. Note, however, that it was Host Marriott, and not Marriott Hotels, that bought the hotels. In other words, Host Marriott (which owns Marriott's physical properties) will now own Sheratons, etc. As described in a great Financial Times [Subscription !$!#!$ required], hotel companies increasingly see themselves as marketing/branding companies, and not property owners.
Who doesn't, these days?
Seriously, Google piggybacks on content owners. Red Hat piggybacks on Linux kernel hackers [though it is now actively leading that development). ESPN (Mobile) piggybacks on Sprint's cellular network. I could go on.
Doesn't anyone actually build things anymore? What will happen if we all turn into marketing machines? Will this leave us in a world devoid of substance or, in the case of open source, of software?
The false premise of open source is that there is a massive global community happily hacking, sharing code with all who want it. The reality is that there are actually communities of very small communities, creating somewhat narrow products that fit these somewhat narrow communities. This is as it should be. If no one is paying me to do the work, why should I write code that you want, rather than what I want?
This implies, therefore, that even open source companies need to build. Need to invent. Need to create.
Those VCs looking to fund existing communities should be equally anxiously engaged in trying to build out these communities through acquisition of development talent. Organic communities can (or, rather, will) only go so far. And so it's important to invest in people/projects/companies who know how to develop communities. Frankly, I know a fair number of such companies that aren't open source at all (like Microsoft).
There will only be so many Red Hats in the world. (Last I counted, there was only one.) But there can be many MySQLs, SugarCRMs, Alfrescos, and JBosses. Do not be deceived. There is no free lunch, or free community. Value still derives from tangible innovation. Not marketing.
Posted by Matt Asay on November 15, 2005 02:06 AM
November 14, 2005 | Comments: (0)
MySQL 5.0 hits 1 million downloads-Interview with Zack Urlocker
MySQL's latest release hit the 1 million download mark this week. A quick Q&A with MySQL's Zack Urlocker follows.
Why is interest in MySQL 5.0 so high?
Without question, people are tired of paying boatloads of their hard-earned cash to the big database vendors. But it's not just about low-cost. The LAMP stack gives you a lot of flexibility and scalability and reliability. It just works. Its convenient, and its fun.
The difference is that the power is no longer in the hands of the vendor--it has been transferred seemingly overnight to the consumer. It's a mature market and most people don't want or need the latest bell-and-whistle features.
I think that many Silicon Valley software companies have fallen into the same trap -- they are stuck asking themselves, "How can we make the most money?" instead of, "What do users want?". Now that users have more choice -- all vendors need to be on their toes. That produces better products.
Is MySQL ready for the enterprise?
Of course. We are being used by enterprises ranging from two employees to 20,000. Embedded in small devices, significant departmental apps and powering the Web's largest sites.
The cool thing that we see a lot of the hottest start-ups going with the open source LAMP stack right out the gate. Sure they save money doing so, but they also are planning ahead of time for huge growth and they know they can always scale out with MySQL down the line with commodity hardware without missing a beat. They want to be tomorrow's Google.
With MySQL 5.0, we have added a lot of old-school database features like stored procedures, triggers and views -- but we're zealously focused on the features that the proprietary vendors aren't: high-performance, reliability, affordable scalability and overall ease-of-use. We're more than happy that other people want to battle over who has the most complicated database. We strive to simplify that complexity for our users.
How does this new release impact the MySQL ecosystem and community?
We are really dedicating ourselves to the community these days. Not that we ever took them for granted, but we realize that this is really a key differentiator for us. We have hired three new community relations people in the past three months -- including a VP -- and are currently looking for a North American open source person [http://www.mysql.com/company/jobs/communityrelations.html if anyone's interested!]
We're getting a lot of fantastic contributions and exchange of ideas on our forums (forums.mysql.com) and blogs (www.planetmysql.org/). We're also working hard to evangelize MySQL 5.0 with database tools vendors like Quest and Embarcadero, enterprise software packages like Agresso and SSA Global (formerly Baan), as well as a new generation of open source projects/companies such as Zimbra, JasperSoft, Alfresco and SugarCRM.
Open source has now achieved a level of adoption that now many companies can build their businesses around it. We see this as the ultimate proofpoint that open source is ready for very dramatic growth. It's hard to say which startup companies will be most successful, but it is clear that all of them will be based on open source.
Posted by Dave Rosenberg on November 14, 2005 06:30 PM
November 14, 2005 | Comments: (0)
SAP's Agassi responds: I love open source
Shai Agassi responded on his blog and in the media to my post (and others) last week regarding his comments about open source. I took his advice and went back and listened to the ZD podcast to make sure I wasn't overreacting. I was able to find all of the quotes I used from the VNUnet article and they weren't out of context enough to alter their meaning, contrary to what Agassi stated.
In the interview Agassi says that SAP is actually a generation beyond open source-one where those who have hacked the code are now going back and trying to fix the issues they created. I am willing to accept that statement in the case of behemoth software apps like SAP and to an extent with Linux, but would argue that the benefits available by being able to modify the code are still greater than the risk of having to go back to a "standard" app or distribution. I talk to many financial institutions, I've never been told by anyone that they regret making changes to Linux or other open source software.
Meanwhile, SAP runs on Linux, uses MySQL, and invested in Zend, and yet fought for the EU patent directive. Is SAP a study in contrasts, or opportunistic? Hard to say. I think the point Agassi was trying to make is that software-or any IP based commodity needs to have a leader (a commercial entity) guiding and supporting it. That notion has been proven both right, by SAP, and wrong, by Apache.
"Open Source" is a conceptual umbrella underneath which many products, projects and companies sit. I am not a zealot. I am too cynical, jaded and educated to completely believe in anything-software or otherwise. I am a pragmatist about the benefits of open source for the greater good and take offense when business leaders use their bully pulpit to make comments that reflect a lack of understanding or marginalize the global movement to fit into their worldview.
I should have chosen another word than fool for Agassi. He's a really smart guy and someone future business leaders should learn from. I just think his perspective is skewed and one-sided. Oh well, I am over it.
Note to SAP-feel free to call me or email me. I am always happy to tell people why they are wrong ;>
Links:
Dan Farber: SAP's Agassi explains his open source stance
Dana Blankenhorn: Rhetoric vs. Reality at SAP
Posted by Dave Rosenberg on November 14, 2005 09:57 AM
November 11, 2005 | Comments: (0)
Modern Management Expert Peter Drucker dies at age 95
Odds are you have spoken any number of Peter Drucker's ideas and theories without even knowing it. Drucker is arguably the most important marketing mind ever, revered as the father of modern management for his numerous books and articles stressing innovation, entrepreneurship and strategies for dealing with a changing world.
WSJ is running a fantastic article about Drucker written on his 90th birthday.
Mr. Drucker is famous for a series of questions: What is our business?Who is the customer? What does the customer value? The answers to those questions, asked by generations of managers around the globe, became known as "the theory of the business." The most distinctive hallmark of the managerial mindset is that it operates from that theory. Major decisions and initiatives all become tests of the theory. Profits are important in part because they tell you whether your theory is working. If you fail to achieve the results you expected, you re-examine your model. It is the managerial equivalent of the scientific method, starting with hypotheses which are then tested in action, and revised when necessary.Mr. Drucker's insights about "knowledge work," a phrase he coined decades ago, grow out of his broad understanding of work as a human activity and management as a liberal art. He has always understood that people are deeply--and rightly--resistant to being "managed." This point is especially relevant for knowledge workers, who know more about their jobs than their bosses do. For them, supervision is a special kind of hell. This is why good managers help people manage themselves by focusing consistently on performance and results and by teaching them, often by example, to think about what they are good at, how they learn, what they value. Such self-knowledge is essential to performance.
A few Drucker ideas that relate to technology and open source: (via Wikipedia)
The need for community. Early in his career, Drucker predicted the "end of economic man" and advocated the creation of a "plant community" where individuals' social needs could be met. He later admitted that the plant community never materialized, and by the 1980s, suggested that volunteering in the non-profit sector might be the key to community.
A desire to make everything as simple as possible. Drucker has suggested time and again that corporations tend to produce too many products, to hire employees they don't need (the better solution is contracting out), and to expand into economic sectors that they should stay out of.
The need for "planned abandonment." Corporations as well as governments have a natural human tendency to cling to "yesterday's successes" rather than seeing when they are no longer useful.
(more at Wikipedia)
Posted by Dave Rosenberg on November 11, 2005 04:50 PM
November 11, 2005 | Comments: (0)
Gartner: Hold off on Vista until 2008
Via ZDnet: Companies shouldn't rush to upgrade to Microsoft Windows Vista, according to analysts at Gartner, who believe most firms could safely hold back until 2008.
The majority of improvements in Vista will be security-related and most of this functionality "is available via third-party products today", Gartner claimed in a research note published on Friday.
While Vista will feature "offer incremental, evolutionary improvements" over its predecessors, XP users should "pursue a strategy of managed diversity" only bringing in Vista on new machines and not until 2008, the analysts recommend.
Posted by Dave Rosenberg on November 11, 2005 03:46 PM
November 11, 2005 | Comments: (0)
Where are they now? (Executive moves in the open source business community)
A call this morning clued me into something that I found a bit shocking.
Moshe Bar is leaving XenSource. I say "shocked" because I had seen Moshe as XenSource. That may not be fair, given that XenSource has other great people (including the Cambridge UK-based development team originally responsible for Xen), but Moshe was the public face. I really like Moshe - never had a bad experience with him - and wish him well. He's apparently moving down to Texas and will be working more on OpenMosix, his open source Linux cluster project. (If you're down in Texas and want a great founding member of your open source team, here's your guy. If you can get him.)
Also, the indescribable r0ml (who left Optaros a few weeks ago) has landed, and I think it's a great landing place for him. Root Markets. Don't ask me what they do. R0ml explained it all to me, but all I understood was "root" and "markets." What little I did understand is that it's a perfect vehicle for r0ml to use his finance and engineering backgrounds, and bring an open source flavor to it all (though the company is not an open source software company).
Lastly, a question. Does anyone know where John Young (formerly VP of Marketing at Red Hat) has gone?
Posted by Matt Asay on November 11, 2005 12:49 PM
November 11, 2005 | Comments: (0)
Web 2.0: Free love or pricey lawsuits?
Just when it seemed everything about Web 2.0 was cheap and easy (dirt cheap infrastructure (software and hardware), build on others' networks/bandwidth, etc), along comes a wake-up call. Nothing's free. Including, as reported today, user reviews, thanks to Amazon.
Haven't heard? I'll let Susan at InternetNews tell you:
The online retailer of books and just about everything else was awarded three new patents, covering its purchase circles, search and consumer reviews. While Amazon.com's patent police would go after Web publishers, not consumers, the review patent could put the kibosh on the social networking components of many search services.Uh-oh. Suddenly, it looks like someone just might own a fair amount of that "architecture of participation." (As for negative feedback on companies, Dave's got that one locked up, too. So don't even think about posting a negative reply to this! :-)Amazon.com's purchase circles can be based on everything from a hobby to an employer, or they can merely peep to see what's the hot book among Oracle employees. The patent covers methods of forming circles and marketing to them, for example, by showing one person looking at a book detail page and who else in the circle has bought that book.
The second patent covers a method of discovering and delivering as search results related products from multiple categories, such as books written by Steve Martin, as well as DVDs of movies in which he appeared.
The third patent is the real kicker. It covers methods for encouraging consumers to write reviews of items they've purchased by determining the optimal times to send them e-mails or reminders.
Posted by Matt Asay on November 11, 2005 11:26 AM
November 10, 2005 | Comments: (0)
Forrester on where open source is booming
Michael Goulde is Forrester's go-to guy on open source. I like his reports - definitely worth buying.
In a report he did in June - Open Source Usage Is Up, But Concerns Linger - Michael surveys a range of large enterprises (95+), and comes up with some interesting data (below). I've included some of it here, but highly recommend that you buy and read his full report to get the full story. (Btw, you can find some of Forrester's data from last year's report here in this eWeek article.
First, Goulde's research clearly shows that open source is rising.
Of course, "use" doesn't necessarily translate into "paid use." This is made somewhat clear when Goulde follows up this question with a drill-down on what open source products/projects enterprises are using:
For those who try out open source, it appears that there's no going back. Goulde's survey reveals a high level of satisfaction with those who have tried out open source. For those who aren't satisfied, I'm guessing that many (i.e., those who still think open source is about cost) simply bought on a somewhat false premise. Open source is not a price tag. It's much more than that.
If you correlate this with the "What are they buying" data above, then it appears that Linux (and, really, the LAMP stack) is paving the way for great open source adoption up the software stack. Linux and Apache paved the way for MySQL (and the Ps - Perl, Python, PHP), which paved the way for JBoss/Tomcat, which are paving the way for...open source applications. As the enterprise gets progressively comfortable with each new stage of open source encroachment, I think we'll see much wider use of open source software.
Now correlate this growing use of open source to the growing areas of enterprise software spending (from a different Forrester report):
Clearly, there is a massive open source software market waiting to happen. While I think there have been some weak investments in open source, I definitely don't think it's weak to invest in open source right now. Open source has a lot of room to grow.
Posted by Matt Asay on November 10, 2005 07:54 PM
November 10, 2005 | Comments: (0)
Does the world need another CRM company?
In Dave's last harangue (no, the one that came after spanking Shai @ SAP), he asks:
I think SugarCRM is proving that replacing proprietary apps is really not that big of a deal, but does the world need 5 more open source CRM companies?My answer? Absolutely.
SugarCRM is helping to level-set expectations of what is really needed in a CRM system, and what it should cost [Note: I'm on SugarCRM's advisory board.]. Same with Alfresco in Enterprise Content Management [Note: I work for Alfresco], JBoss in application servers, MySQL in databases, etc. Some might call it commodification, or lowering the bar. I think it's the exact opposite. It's raising the bar, returning the software business to one focused on service/customer support, and not on 1s and 0s.
Having reset expectations, open source companies (and their associated communities, but mostly the core project leads) are able