From Volunteer Open Source to Professional Open Source

Discussions

News: From Volunteer Open Source to Professional Open Source

  1. Marc Fleury has written a blog entry summarizing his business model called "From Volunteer Open Source to Professional Open Source", in his own words (as opposed to paraphrases in BusinessWeek and on forums such as Slashdot). In it, he explains a shift in terminology (from "amateur" to "volunteer" open source) and his commitment to open source business models.
    One parting word of advice to developers who do aspire to earn a living from their FOSS contributions, along the Professional Open Source model. Devote serious thought to your license because it can seriously impact whether you will be able to earn a living writing and supporting your software. If this is your ambition, I seriously recommend GPL-style licensing (JBoss projects are LGPL) because this licensing family encourages a positive feedback loop and discourages forking. While competing software vendors, for obvious reasons, love BSD-style licensing and will encourage you to adopt these more liberal licenses, the impact for the FOSS developer is that IF you are writing BSD-licensed software AND YOU ARE SUCCESSFUL, your only option may be to go to work for your proprietary competitor who is only to happy to incorporate your work in their proprietary product. Beware of the man offering you candy.
    What do you think?

    Threaded Messages (70)

  2. I believe that the open source part of the industry is up for a bloodbath, and pretty soon, sure, some companies will survive, but the wast majority will go belly-up akin to 2001 dotcoms.

    Why?
    Heres why:

    Compelling customer benefits:
    Very few, sure you get the source, but that is a very small part, especially if you are selling to customers that don't consider software development a core part of their activities.

    Customer power:
    Massive, they can do whatever the OSS companies offer on their own, go to their direct competitor etc. The customers basically have the companies balls in a nutcracker.

    Threat of substitutes:
    Are you kidding? Your product IS a substitute to you. As is any moron who can understand it and knows how to register a company.

    Competitions:
    Fierce, the space is quickly becoming overestablished with developers wanting to make a quick buck on open source, and their mothers.

    Barriers of entry:
    None, anyone smart enough will be able to duplicate most of what you do in no time.

    Sustainable advantages:
    None, no patents, no proprietary code no nothing.

    Economics:
    You are competing against "we´ll do it ourselves" or "let the lowest bidder competent enough do it". The software is commodity. Wake-up call: so are your services. You are competing 90% on price. And on another note: how do you scale a business that is so dependent on man-power? Where will you find people that are smart enough? And how do you justify (and motivate) really smart and not exactly cheap resources to do mostly support work?


    I am not contesting that a few companies will survive, but at least 95% of them will be extinct this time in two years time. Most of the companies don't even HAVE a business model to begin with! Unless you count loosing money as a business model. I recall some company selling services (and add-ons?) for Maven, why? In the words of Duece Bigalow: "No, I won't pay, YOU pay me $5!".


    And there is no questioning that Open source has changed the face of the software industry. It is not just open source that will find it hard to make money, traditional software vendors will find it hard too.
  3. Well said, Wille. I'll give it another 2 years and the hype is over. I'm neat to know what the next hype will be. May be selling first class closed source software the old fashioned way with license revenue, stable and healthy company behind without VC money?

    I don't get it what the value is of open source. Sure, I use open source s/w but I don't look into it nor do I fix bugs. If it doesn't work, may be I drop a line into a mailing list. However, most of the time I dropped it and look for alternatives or I write it myself.
  4. I don't get it what the value is of open source.

    There is loads of value in open source for companies such as proprietary software vendors, consultancies and other businesses heavily relying on internal development.
    No license fees, managing and being in control over your own upgrade cycle, not being locked into one vendors services etc are all good value and benefits. But those are benefits of open source software, _not_ benefits provided by open source businesses.

    And furthermore if you look at the industry and market forces I mentioned in my previous post, open source is as lousy an industry you can be in. It's probably way worse than the airline industry (and the airline industry as a whole has never made a profit throughout it's history).

    Looking at it from an investors viewpoint, the smart money should go into industries that have forgiving characteristics (like pharmaceuticals 20 years ago).
  5. There is loads of value in open source for companies such as proprietary software vendors, consultancies and other businesses heavily relying on internal development.No license fees, managing and being in control over your own upgrade cycle, not being locked into one vendors services etc are all good value and benefits.

    These "benefits" some would see as "dangers". No one to be held accountable for miserable performance, bugs in the software. Security weaknesses of the software immediately out in the open, require much quicker response in patching applications. Need to manage third patry source code that does not fit your corporate standards.

    Actually, a lot of users of open source are just that. They load the binary (or the source, compile it once) and never upgrade. After all "never touch a running system" has been good advice for generations of developers when concerned with their own code only. The need to keep the "expertise alive" in your cooperation can be quite tricky. Imagine you installed something qmail, patched it up to support imap, certificates and so on and two years after continuus zero downtime operation you hit a security advisory that requires immediate action. If the guys who set it up are still around you are lucky - if not, bad luck.
    It's probably way worse than the airline industry (and the airline industry as a whole has never made a profit throughout it's history).

    Well, PanAm did well before the market was opened up :-).
  6. Crosslink[ Go to top ]

    Looks like the same debate as on this thread:
    http://www.theserverside.com/news/thread.tss?thread_id=35031
  7. Customer power:Massive, they can do whatever the OSS companies offer on their own, go to their direct competitor etc. The customers basically have the companies balls in a nutcracker.

    This is a sadly typical proprietary software company mindset. Empowering your customers and giving them choices is considered to be a huge negative. You are in a zero-sum game with your customers, and the only way you can win is to lock them into your product without any options.

    Look at professional services in other industries. They have figured out that if they develop relationships with their clients based on trust and confidence, their clients don't jump ship to realize moderate price savings. Those companies call it "delivering world-class customer service." We call it "letting them put our nads in a juicer and pressing the puree button."

    Here's a fun exercise:

    1. Start a new company

    2. Advertise that you are offering the same services that Interface21, at HALF THE PRICE!

    3. Report back in three months to let us know if you have put Interface21 out of business yet.

    4. If you haven't, consider why that might be. After all, Rod simply allows customers to put his Johnson in a weed-whacker.
  8. Here's a fun exercise:1. Start a new company2. Advertise that you are offering the same services that Interface21, at HALF THE PRICE!3. Report back in three months to let us know if you have put Interface21 out of business yet.4. If you haven't, consider why that might be. After all, Rod simply allows customers to put his Johnson in a weed-whacker.

    You forgot step 2.5:

    2.5. Call yourselves the real core developers at Interface21.

    Man, this post gives me feelings of deja vu ;-)

    Bill
  9. Empowering your customers and giving them choices is considered to be a huge negative.
    It's not huge negative, however offering no compelling customer benefit and no reason to stick with you is a huge negative.
    Look at professional services in other industries. They have figured out that if they develop relationships with their clients based on trust and confidence, their clients don't jump ship to realize moderate price savings. Those companies call it "delivering world-class customer service."
    Open source can probably be a good and lucrative "consulting gig" for the top guys like Rod Johnson. But the question remains, if you have the ambition to grow into a large company, how do you scale it? How do you work around the industry dynamics I listed? Sure you can say "you only hire the best", but I think 10 other consultancies are already saying that, an I am betting most of them are lying their teeth out.

    As long as the business is relatively small, clients will buy into the superstar name, but once you start growing, there is no avoiding the industry dynamics and forces.

    And believe you me, all those VC's funding open source start-ups, their not in it to help some nice fellas build a 5 man consultancy where a bunch of nerds can hack away on their pet projects.
    They are in it to build large companies and exit profitably. And they will pressure the company to go there,and if they don't see it going there, they will raid the place for all it's worth and even sell the stapler, leaving the founders with nothing.
  10. "They are in it to build large companies and exit profitably. And they will pressure the company to go there,and if they don't see it going there, they will raid the place for all it's worth and even sell the stapler, leaving the founders with nothing."

    How is this different than any other company started with the help of VC funding? Do you think if I got this dream of owning my own steak house that the VC companies care about my dreams? No. They care about profit. If they can bend you over, they will. Apparently you think the geeks working on these projects are so retardedly stupid that they aren't aware of this.

    From what I've read so far, none of your arguments hold any water at all.
  11. Apparently you think the geeks working on these projects are so retardedly stupid that they aren't aware of this.
    No, I do not think they are retarded.
    From what I've read so far, none of your arguments hold any water at all.
    Here's a challenge for you: let's make a list of all Open Source focused companies in the Java/J2EE space today, and let's revisit that in 2 years time. I'll eat the pants I am wearing right now if not at least 90% are out of business or have been swallowed by other businesses for peanuts.
    Wishful thinking will not make the market forces and economics go away, no matter how hard you wish.

    Note: I am _not_ saying 100%, there will be a few survivors, some that will probably become quite large and successful with viable business models. Even some dotcoms survived the dotcom shake-out (eBay, Yahoo! and Amazon come to mind).

    But that does not take away the fact that there will be a brutal shake-out and consolidation sooner or later, if you don't believe that, well, we'll just have to see who's right in two years time.
  12. Here's a challenge for you: let's make a list of all Open Source focused companies in the Java/J2EE space today, and let's revisit that in 2 years time. I'll eat the pants I am wearing right now if not at least 90% are out of business or have been swallowed by other businesses for peanuts.
    Quite likely you're right. Still, that's no different from any emerging (but important) area of the economy, now or in the past. Take auto manufacturing or railways. The only real difference may be the pace of change being more rapid these days, but that's not confined to just this area...
  13. Here's a challenge for you: let's make a list of all Open Source focused companies in the Java/J2EE space today, and let's revisit that in 2 years time. I'll eat the pants I am wearing right now if not at least 90% are out of business or have been swallowed by other businesses for peanuts.
    Quite likely you're right. Still, that's no different from any emerging (but important) area of the economy, now or in the past. Take auto manufacturing or railways. The only real difference may be the pace of change being more rapid these days, but that's not confined to just this area...

    No, I am not exactly being Nostradamus in my predictions, am I? :)
    But then again revisiting the market forces of the industry, right now is a lousy time to enter the market for "generic open source company inc.".

    Unless you have a superstar name to sell (as in Rod Johnson), or have already been around for a good while and thus have gained some market momentum (as in JBoss I would guess, without knowing their financial situation), the odds are pretty much stacked against you (and there really are no guarantees with the former either).

    This is _especially_ true if quick, unprofitable growth into a large corporation funded by VC money is on the agenda.
  14. I wonder how much money are these business actually making RIGHT NOW. I'd personally NEVER buy consulting from JBoss, but I'm not managing any process-critical system. Would I if I was? Would I even deploy on JBoss if that was the case? I don't know, since there are many factors to analyze, all of particular to each business scenario and some of them absolutely subjective.
    Just saying "No, because there's no one we can blame when the whole thing crumbles" is a silly argument. On the other hand, saying "Yes, because all those fellas at JBoss are quick-ass bedroom programmers who code on their free time, so they MUST be good" is equally dumb. A more professional opinion would be "We analyzed a number of products (open source of not) and JBoss is/isn't appropiate for our scenario + budget + capabilities."

    Cheers and happy coding,
    Martin
  15. Well, I think that Open Source companies have a failure rate that are about in line with typical startup failure rates, as mentioned above.

    One of the things that people keep implying is that OSS is somehow trying to fundamentally change economics. It's just not. All it's doing is shifting economic value to a different part of the economic loop. Pure and simple. There are four primary ways to create economic value with Open Source, and people are making money off of them:

    1) Licensing and Operating Cost (the one that gets the most press)
    2) Complementary Product impact. Oracle, IBM, and plenty of startups are making money here.
    3) Collaborative Return on Investment. Projects like Kuali and Sakai are nailing this space, and companies like RSmart are doing well to make money off it.
    4) Services associated with Open Source (IBM was prescient here).

    Of course, for every one of these there are more failures. But that's just normal business: Poor execution, leadership, operations, technology, marketing whatever.

    My group, the Open Source Lab, has taken advantage of most of these for our clients in various ways. And we are getting paid to develop products, not just services around them. We just completed our report for the last 18 months. Our primary investor's (Oregon State University and the Oregon University System) return was over 400% for that time period. Now we're not a private sector company, we're a unit of OSU. However, we're not doing anything that unique. It simply works. And private companies are making money from the 4 areas just fine.

    However, like anything, it has to work appropriately. There are plenty of times when I tell people that they should just go buy something proprietary. It'll save them money.

    It's not rocket science, and it's not even difficult. People forget that open source is both a technological innovation as well as a means for economic and technological delivery and collaboration. It's NOT trying to re-write any economic rules. The problem comes when too many people "square-peg-round-hole" open source into places where those economics aren't served.

    I did leave out "control of destiny." Control of destiny is an issue that we see that sometimes outweighs the economic issues. We do see people sometimes moving to Open Source when they feel they'll have better IT control, even when the economic/financial value is not in their favor. That, in itself, is a marketable opportunity for some companies, but it's slightly out of scope here.

    Jason McKerr
    The Open Source Lab
    "Open Minds. Open Doors. Open Source."
  16. Well, I think that Open Source companies have a failure rate that are about in line with typical startup failure rates, as mentioned above.

    I think it will go up well above the average in the next two years to then fall right back down.

    Why?
    The reasons I mentioned earlier, the space is getting crowded and over-hyped, everyone and their mother are starting open source businesses.
    That and the other factors collaborating are making right now a lousy time to enter the industry. 2 years ago or 2 years from now will probably be a better time to enter..
  17. Like I said, I'm in no decision-making position... so it's no surprise mi appreciation is wrong.
    It seems you know what you are talking about. Do you have some paper or article with precise data about this? This whole thread is about speculation, and that's cool, but it would be nice to have some real information.

    Cheers and happy coding,
    Martin
  18. Unless you have a superstar name to sell (as in Rod Johnson), or have already been around for a good while and thus have gained some market momentum (as in JBoss I would guess, without knowing their financial situation), the odds are pretty much stacked against you.

    I don't think building a company based on a superstar is a good approach. Rod Johnson is not a good example anyways. Even if Rod Johnson left Interface21, they would still have a viable business model because Spring is a successful project, there's a bunch of good people over there besides Rod, and they own the Spring website (trademark too?). Really, the same strengths JBoss has had since day one.

    So, there's a few important things when building an open source business:

    1) Trademark
    2) Control over the OSS project
    3) Business website and project website are one. This goes with branding. This is why it will be so hard for companies to build any kind of business off of Eclipse or Apache based projects. The website is a key to market your services. I can't stress this enough
    4) License. Bring on the flames, but a FSF license has many business advantages that BSD does not. a) It protects you from your closed-source competition b) It protects you from organizations like Apache or Eclipse from forking parts or all of your project (although this may change) c) It allows you to make more money by dual licensing. For instance, MySql ships GPL, but dual licenses if you pay for it.
    5) Believe it or not owning the IP is not as important as Cameron thinks. If you are not doing the dual license strategy, owning all the IP is much less important than to actually controlling the project. JBoss Inc. *still* does not require contributors to sign over copyright as we feel it is important for contributors to retain rights over their code. Makes everybody feel they own a piece of the pie and it ensures that there can never be a license change.

    Really, the most important things in OSS business is 1-3. 4-5 are much less important, but they do help solidify your business.

    Bill
  19. I don't think building a company based on a superstar is a good approach.

    Aah, but that depends on _what_ type of business you are building! :)
    If you are building a small consultancy where clients can depend on being serviced at least in part by a "superstar" it might work successfully.
    If you want to build a business that is going to get large and scale, it is less of a good idea.
  20. I don't think building a company based on a superstar is a good approach. Rod Johnson is not a good example anyways. Even if Rod Johnson left Interface21, they would still have a viable business model because Spring is a successful project, there's a bunch of good people over there besides Rod, and they own the Spring website (trademark too?).
    But Spring is not a successful project because of "good people", and Rod is not a superstar "just because". Any successful project, of any kind, needs a good leader with a vision of where to go, and Rod is one of those. Simply having "good people" is not enough. You can get all the talent you want, but without direction not much is going to happen.

    Also, in terms of recruitment having a "superstar" makes things easier. If there are two OpenSource consulting companies, and one of them has the lead developer of what you want to work with, which one are you going to choose?
    So, there's a few important things when building an open source business:1) Trademark2) Control over the OSS project3) Business website and project website are one. This goes with branding. This is why it will be so hard for companies to build any kind of business off of Eclipse or Apache based projects. The website is a key to market your services. I can't stress this enough4) License. Bring on the flames, but a FSF license has many business advantages that BSD does not. a) It protects you from your closed-source competition b) It protects you from organizations like Apache or Eclipse from forking parts or all of your project (although this may change) c) It allows you to make more money by dual licensing. For instance, MySql ships GPL, but dual licenses if you pay for it.5) Believe it or not owning the IP is not as important as Cameron thinks. If you are not doing the dual license strategy, owning all the IP is much less important than to actually controlling the project. JBoss Inc. *still* does not require contributors to sign over copyright as we feel it is important for contributors to retain rights over their code. Makes everybody feel they own a piece of the pie and it ensures that there can never be a license change.Really, the most important things in OSS business is 1-3. 4-5 are much less important, but they do help solidify your business.Bill
    'tis interesting to note that 1-5 are all self-serving aspects of a business model. Ever heard of "customers"?

    I would say that without condition 0) thou art lost: you need to make sure that you are serving a customer need, and in a way that those customers want the thing you are selling. Do that, and your business will have a much greater chance to "solidify".
  21. 90% of ALL businesses fail within the first 3 years so your proposal is spurious.
  22. 90% of ALL businesses fail within the first 3 years so your proposal is spurious.

    FWIW, this is urban legend. Stats from "Startup Business Risk Index" (Brandow Company Inc) show the 4 year survival rate on most businesses at 55%, and in software it's about 45%. This compares with other sources too. Reputable stats also disagree that most companies that accept VC money go bust soon after. It may have been true for the years 1999 and 2000, for obvious reasons, but otherwise it's not true at all. In fact, the stats I've seen for local VC funds are that about 85% have a "successful exit", which I suppose could mean almost anything, but I think generally means the VC fund at least got their investment back.

    Some people like to lump a lot of the success/failure with "luck". I don't agree with the "luck" statement. Fact is that people have a very difficult time distinguishing between a "good idea" (of which there are LOTS) and a "good opportunity" (of which there are few...) A good idea is rarely a good opportunity, and many people explain the difference as "luck". Are certain OSS models a "good idea"? Personally, I think yes. Is it a "good opportunity"? Personally, I'm not as convinced on most models, although some like eclipse seem like great opportunities to me.

    - Don
  23. In fact, the stats I've seen for local VC funds are that about 85% have a "successful exit", which I suppose could mean almost anything, but I think generally means the VC fund at least got their investment back.

    Hmmm, the local VC's I've talked to (specializing in software) put that figure at around 15-25%.

    I think your neighborhood has much better entrepeneurs than my neighborhood. :)
  24. Statistics...the great satan. I'm willing to admit that the 90% failure rate is exaggerated. However, my point still stands and that is most businesses simply don't last. Take your stats out to 6 years and you will see the failure rates increase (at least according to the government).
  25. Not everything is money...[ Go to top ]

    Open Source Software is IMHO a good sign for collaborative people sharing knowledge. I don't think open source is here to challenge propiertary vendors. It's something diferent. I use open source software because a I disagree with intelectual property. Knowledge should be free.

    If you want to startup an opensource company for benefiting you only I think you are making a mistake.

    Information can not be owned. It is not a posession. The software industry should be based on services.

    Excuse my english. I am from Argentina.
    Charly
  26. Take your stats out to 6 years and you will see the failure rates increase (at least according to the government).

    That's clearly true... they certainly couldn't go down farther out, anyway.
  27. I'll eat the pants I am wearing right now if not at least 90% are out of business or have been swallowed by other businesses for peanuts.

    Now there's some real incentive to see all these open source companies succeed ;-)

    Peace,

    Cameron Purdy
    Tangosol Coherence: Clustered Shared Memory for Java
  28. And believe you me, all those VC's funding open source start-ups, their not in it to help some nice fellas build a 5 man consultancy where a bunch of nerds can hack away on their pet projects.They are in it to build large companies and exit profitably. And they will pressure the company to go there,and if they don't see it going there, they will raid the place for all it's worth and even sell the stapler, leaving the founders with nothing.

    I agree. I think these companies can be successful, but I think very few will deliver VC type growth and returns. Too much VC money out there.
  29. And believe you me, all those VC's funding open source start-ups, their not in it to help some nice fellas build a 5 man consultancy where a bunch of nerds can hack away on their pet projects.They are in it to build large companies and exit profitably. And they will pressure the company to go there,and if they don't see it going there, they will raid the place for all it's worth and even sell the stapler, leaving the founders with nothing.
    +1 ...well said.

    Nikita.
    GridGain Systems.
  30. And believe you me, all those VC's funding open source start-ups, their not in it to help some nice fellas build a 5 man consultancy where a bunch of nerds can hack away on their pet projects.They are in it to build large companies and exit profitably. And they will pressure the company to go there,and if they don't see it going there, they will raid the place for all it's worth and even sell the stapler, leaving the founders with nothing.

    And how exactly is this different from a closed source start-up?
  31. And how exactly is this different from a closed source start-up?

    I would suggest that it is not about whether one publishes one's source code or not, but rather whether one owns (retains the IP rights to) one's own source code.

    For example, how can an investor be relatively certain that Merck and Phizer will still be in business in ten years time? It's because when they build something, people pay them for it.

    Quick pop quiz: Name ten popular drugs. Now tell me how many of those were discovered / created / designed in India, a country that does not protect the designs of medications as intellectual property.

    Nonetheless, I think it's pretty obvious that companies can make money providing just service and support for open source software.

    Peace,

    Cameron Purdy
    Tangosol Coherence: Clustered Shared Memory for Java
  32. And how exactly is this different from a closed source start-up?
    I would suggest that it is not about whether one publishes one's source code or not, but rather whether one owns (retains the IP rights to) one's own source code.For example, how can an investor be relatively certain that Merck and Phizer will still be in business in ten years time? It's because when they build something, people pay them for it.Quick pop quiz: Name ten popular drugs. Now tell me how many of those were discovered / created / designed in India, a country that does not protect the designs of medications as intellectual property.Nonetheless, I think it's pretty obvious that companies can make money providing just service and support for open source software.Peace,Cameron PurdyTangosol Coherence: Clustered Shared Memory for Java

    Nice to see Cameron having other areas of expertise - the Indophile, the patent expert, the IP expert, the Copyright expert ... :)
  33. IMHO, there is very little money to be made by just supporting open source software. If you are new to JBoss for example, you may need a couple of weeks of training and a 6 months support to get up to speed and running. Clients who are smart enough to choose open source software are smart enough to know when the "support" is not needed. The real money probably is in developing systems for clients using open source software. So, a start-up needs to have experise in a broad array of technologies (say, Spring + Hibernate + Tomcat + MySQL, for example) AND domain knowledge (finance, health care, etc.) to build something or help clients build. This is probably how startups could survive in Billion dollar service industry. Most people quote how RedHat and JBoss make money by supporting Linux and JBoss respectively, but that cannot be universally applied to all software. If you just support Hibernate for example, you wouldn't survive for more than a few months because it forms just a small piece in the overall solution.

    Open source software is a maze for most companies because you don't know which component is most suitable for your need and how it works with rest of the system. So, startups focusing on building frameworks using a myriad of smaller frameworks that solve the problems in a perticular domain may end up being successful.

    My 2 cents.

    Peace,
    Raghavendra.
  34. Clients who are smart enough to choose open source software are smart enough to know when the "support" is not needed.


    +1 well said

    Successful OOS does not mean successful open source business, but successful career life (get hired with high payment or being an individual consultant). Open source business model depends on the type of open source software itself. Working with useful but complex OOS may need training and professional service support, but not the case for simple OOS. For example, Struts is the most popular MVC OOS and widely used in web-based applications, but how many people here need Struts training and support? JBoss and Interface21 may survive because they offer useful and complex software similar to RedHat, and the existing, popular open source software business model is good for them.
  35. Keeping Perspective[ Go to top ]

    I think we are crossing over the peak of the hype curve and entering the trough of disillusionment on open source. The theory that Open Source is going to overrun commercial vendors has been completely overblown. They both have a time and a place.

    You have to give credit to the JBoss guys to get the Wall Street Journal, Business Week and others to print their story. It is classic David vs. Goliath but won’t have the same ending. Neither side will completely win or loose. Open Source development will find its place among other technology products and processes.

    There is no doubt that Linux, MySQL, Apache, JBoss, Struts, Spring and other projects will carve out ideal use cases, but to assert that they are all things to all projects is a wild exaggeration. In 1996 people were saying that Linux was going to run over the UNIX platforms. There was a frenzied scramble across the industry. Here we are ten years later and while Linux has steadily made inroads, it is still a relatively small % of the server OS market. It is growing faster than other segments, so it is important, but it took ten years to get this far and will take another five to ten years of growth to cement leadership in the OS space. That’s a long time, especially if you factor in the backing from IBM and others.

    The net is that people need to put these things in perspective. Open Source is new and interesting but is not the nuclear event that some make it out to be.

    Eric
  36. How many times do we have to go through this? We're not even completely over the hangover from the last hype-driven bloodbath and we're arguing again over why we are right and everyone else is wrong.

    This is the same crap everyone was shoveling four years ago and you are still wrong. For a bunch of smart people we really act like we haven't learned a damned thing.
  37. Great analysis! I hope Fleury's talk becomes a wake-up call.

    I don't care if these OSS start-ups will succee or fail.

    What we should worry about is that the abundance of open source software will graduately erode the value of all software and we, the practitioners in this industry, will all suffer!

    Just take a look at Borland and BEA. Today, Borland is heading to the exit; tomorrow maybe BEA.

    Yesterday, nobody trusted Linux as a viable OS for business; today many Fortune 500 run on Linux. The same will happen graudately to all other types of "niche" software. There is no need for Borland JBuilder now. There will be no need to BEA WebLogic. Who knows, maybe one day SAP will have a creditable OSS equivalent as well.

    The "niche" will be kept pushing to the corner.

    For programmers, while you can have accesses to a great number of free tools, your programmiing expertise becomes commoditized. Any outsourcing providers in India can just use, copy, learn from many of the OSS product out there and no need to hire a talent programmer or software designer at all.
  38. What we should worry about is that the abundance of open source software will graduately erode the value of all software and we, the practitioners in this industry, will all suffer! Just take a look at Borland and BEA. Today, Borland is heading to the exit; tomorrow maybe BEA.Yesterday, nobody trusted Linux as a viable OS for business; today many Fortune 500 run on Linux. The same will happen graudately to all other types of "niche" software. There is no need for Borland JBuilder now. There will be no need to BEA WebLogic. Who knows, maybe one day SAP will have a creditable OSS equivalent as well. The "niche" will be kept pushing to the corner.For programmers, while you can have accesses to a great number of free tools, your programmiing expertise becomes commoditized. Any outsourcing providers in India can just use, copy, learn from many of the OSS product out there and no need to hire a talent programmer or software designer at all.
    I have not gone through the original writing of the Bill which has lead to this discussions , but looking at the community future view and my analysis is that the Services based on the Open sources will generate the revenue for the OSS companies .
    I believe the Spring does generate the revenue based on the services and same is true for Jboss or any other popular opensources .
    Any outsourcing providers in India can just use, copy, learn from many of the OSS product out there and no need to hire a talent programmer or software designer at all.
    I believe to understand the internals/design of the OSS you need to have talent and the proper approach .
    Regards
    Vicky
  39. I would disagree[ Go to top ]

    Great analysis! I hope Fleury's talk becomes a wake-up call.
    I don't care if these OSS start-ups will succee or fail.What we should worry about is that the abundance of open source software will graduately erode the value of all software and we, the practitioners in this industry, will all suffer! Just take a look at Borland and BEA. Today, Borland is heading to the exit; tomorrow maybe BEA.

    Yesterday, nobody trusted Linux as a viable OS for business; today many Fortune 500 run on Linux. The same will happen graudately to all other types of "niche" software. There is no need for Borland JBuilder now. There will be no need to BEA WebLogic. Who knows, maybe one day SAP will have a creditable OSS equivalent as well. The "niche" will be kept pushing to the corner.

    For programmers, while you can have accesses to a great number of free tools, your programmiing expertise becomes commoditized. Any outsourcing providers in India can just use, copy, learn from many of the OSS product out there and no need to hire a talent programmer or software designer at all.

    I'm bias here, but I disagree that OSS might make programming expertise a commodity. I would argue that at any given time, there's a "relatively constant" percent of programmers that have enough skill to read a given piece of OSS code, understand it and become an expert. Most programmers do not have those skills. As an OSS project grows and matures, the learning curve becomes more steep. What OSS gives you is the option of diving into a piece of code and patching it yourself.

    With closed source, that is limited. If I find a bug in a commercial product, they are not going to let me see the code and debug it myself. With OSS I have that option, but I have to have the skills and desire. Without the desire, trying to fix a simple or complex bug in any piece of software is going to be tough. Knowing where to look isn't always obvious, so expertise will still be the same.

    As OSS matures, the need for entry level positions is going to shift to off-shore, but I don't do entry level stuff. It does affect people and OSS does have it's draw backs.

    peter
  40. OSS and jobs[ Go to top ]

    I'm bias here, but I disagree that OSS might make programming expertise a commodity. I would argue that at any given time, there's a "relatively constant" percent of programmers that have enough skill to read a given piece of OSS code, understand it and become an expert. Most programmers do not have those skills. As an OSS project grows and matures, the learning curve becomes more steep.

    From my experiences, as an OSS project grows and matures, the more free help and documentation there are and the less it is needed for diving deep into the source codes. If that OSS project is not matured, then you should not use it at all and look for another mature OSS projects.

    In a company I worked for, we ran Tomcat fine for years without ever looking into its source codes. I looked into Hibernate source code once only because I didn't want to wait until the next day to get answer from the forum.

    It used to be that they maintained their own O/R engine but it would take quite a dedicated talent programmer to maintain it, so they switched to Hibernate and "free" the talent programmer to high-level task such as UI.

    Yes, you still need talent programmers. But if it used to be that you need at least 4 or 5 gurus and 20 average programmers in a paticular shop, now you need just 1 or 2 gurus, 5 average ones and hired 30 new grads in India to do small bug fixes.

    Yet, as the software value has been eroded, fewer investments would be poured into software projcts, especially productized software, so the demand would not grow much.
    What OSS gives you is the option of diving into a piece of code and patching it yourself.With closed source, that is limited.

    This is indeed a positive thing about OSS. I have never disputed the advantages of open-source software.

    The problem is that: there is no such thing as a free lunch and what you give away may be your job.

    The final question is: are you really a guru or an average one?
  41. OSS and jobs[ Go to top ]

    I have no doubt that the limitless ressource that FOSS represents benefit overall growth. The fact that FOSS may drive a specific company or country out of the market is just something we may have to accept.
  42. OSS and jobs[ Go to top ]

    From my experiences, as an OSS project grows and matures, the more free help and documentation there are and the less it is needed for diving deep into the source codes. If that OSS project is not matured, then you should not use it at all and look for another mature OSS projects.In a company I worked for, we ran Tomcat fine for years without ever looking into its source codes. I looked into Hibernate source code once only because I didn't want to wait until the next day to get answer from the forum.It used to be that they maintained their own O/R engine but it would take quite a dedicated talent programmer to maintain it, so they switched to Hibernate and "free" the talent programmer to high-level task such as UI.

    Yes, you still need talent programmers. But if it used to be that you need at least 4 or 5 gurus and 20 average programmers in a paticular shop, now you need just 1 or 2 gurus, 5 average ones and hired 30 new grads in India to do small bug fixes.Yet, as the software value has been eroded, fewer investments would be poured into software projcts, especially productized software, so the demand would not grow much.
    What OSS gives you is the option of diving into a piece of code and patching it yourself.With closed source, that is limited.

    This is indeed a positive thing about OSS. I have never disputed the advantages of open-source software. The problem is that: there is no such thing as a free lunch and what you give away may be your job.

    The final question is: are you really a guru or an average one?

    Good question. I am far from being a guru. As much as I would like to have that much expertise, I don't. In some areas I have a fair amount of depth. In other areas, my experience is weak. Absolutely there's no free lunch. My definition of guru is insanely high and reserved for the top 5%.

    I'm comfortable enough with OSS code that I have no problems diving in and fixing a bug when needed. I think the reality is the IT world is changing much faster than people can cope. That in itself does not validate or invalidate OSS. Everything is impermanent, so sooner or later the need for lots of entry level and mid level programmers is going to diminish. Sure it's scary, but I like change and prefer to shake things up regularly.

    Take the domain of rule engines. Even though RETE algorithm has been in the public since the late 70's, there are few open source RETE rule engines. Aside from CLIPS, there are no high performance RETE rule engines that are mature and have been road tested for 2 decades. There are several high quality commercial RETE rule engines, but their lead will eventually diminish. Gaining deep expertise takes a lot of time, but contributing to OSS doesn't have to necessarily mean giving away everything. often contributing to OSS leads to collaboration and creates a better product.

    there have been plenty of times when my ideas for a jmeter plugin were limited or stupid. After some discussions with other people, the idea is improved and results in a better plugin than if I developed in a bubble.

    peter
  43. Comments on BEA???[ Go to top ]

    The comments about BEA are off base. Open source, as great as it is in many areas, does not have products that compete with the core of BEA's business. The heavy lifting we do with WebLogic and Tuxedo is our bread and butter and in those use cases we have a sustainable lead.

    To reiterate our Q1 results: highest revenues ever in a Q1, highest services revenues ever in a Q1, highest operating profit ever in a Q1... Total revenues are up, cash in the bank is up ($1.6B), cash flow is consistently strong...and so on. BEA has weathered the industry storm well over the last few years.

    Also, check the IDC market share data. BEA is #1 in the largest segment (UNIX deployments) and is #1 is the fastest growing segments (Linux deployments).

    The FUD that the Astroturfers and our competitors spread should be filtered accordingly. The financial and market share data speak for themselves.

    Eric
    BEA
  44. ... Also, check the IDC market share data. BEA is #1 in the largest segment (UNIX deployments) and is #1 is the fastest growing segments (Linux deployments)....EricBEA

    I thought JBoss was NUMERO UNO! Last I heard, every fortune cookie comes with Jboss installed on/in it. Some market penetration, won’t you agree?

    buffet overeater
  45. Comments on BEA???[ Go to top ]

    Also, check the IDC market share data. BEA is #1 in the largest segment (UNIX deployments) and is #1 is the fastest growing segments (Linux deployments). The FUD that the Astroturfers and our competitors spread should be filtered accordingly. The financial and market share data speak for themselves.EricBEA

    Though, to be fair IDC numbers are based purely on vendor reported software license revenue and therefore disadvantage lower priced or free products (unless they recently changed to include support and services revenue).

    I obviously mention this because Sun has been shipping a *FREE* J2EE Application Server for years which the IDC reports do not recognize.

    Rich Sharples
    Sun Microsystems Inc.
    http://blogs.sun.com/roller/
  46. Comments on BEA???[ Go to top ]

    Also, check the IDC market share data. BEA is #1 in the largest segment (UNIX deployments) and is #1 is the fastest growing segments (Linux deployments). The FUD that the Astroturfers and our competitors spread should be filtered accordingly. The financial and market share data speak for themselves.EricBEA
    Though, to be fair IDC numbers are based purely on vendor reported software license revenue and therefore disadvantage lower priced or free products (unless they recently changed to include support and services revenue).I obviously mention this because Sun has been shipping a *FREE* J2EE Application Server for years which the IDC reports do not recognize. Rich SharplesSun Microsystems Inc.http://blogs.sun.com/roller/

    I do find it funny that people think JBoss makes up our market share numbers. Various market share surveys have us between #1 and #4.

    * BZResearch has us at #1.
    * Gartner magic quadrant
    * O'Reilly's OnJava.com did a Reader Survey We come out on top there too.

    Bill
  47. Comments on BEA???[ Go to top ]

    I do find it funny that people think JBoss makes up our market share numbers. Various market share surveys have us between #1 and #4.* BZResearch has us at #1.* Gartner magic quadrant* O'Reilly's OnJava.com did a Reader Survey We come out on top there too.Bill
    I do find it funny that you mention the O'Reilly survey, since their readership isn't exactly representative of all users of application servers. One might say that readers of O'Reilly are, so to speak, slightly slanted in the general direction of OpenSource solutions. Regardless, a survey with 660 replies isn't exactly statistically verified. The rule of thumb in statistics is that anything less than 1000 entries is pretty much pointless.

    So while you might not be "making up numbers", per se, you are using numbers that doesn't represent what you say they are representing. Not sure which is worse.
  48. Comments on BEA???[ Go to top ]

    The Gartner Magic Quadrant is not a measure of market share.

    Eric
    BEA
  49. Comments on BEA???[ Go to top ]

    I do find it funny that people think JBoss makes up our market share numbers. Various market share surveys have us between #1 and #4.

    * BZResearch has us at #1.
    * Gartner magic quadrant
    * O'Reilly's OnJava.com did a Reader Survey

    Bill, Jboss is widely used .. but pointing to opt-in suveys and web polls on open source sites is hardly a way to prove it ;-)

    It's like trusting a poll on slashdot that shows that Linux has 90% of the desktop OS market share, or a poll on MSDN that shows that Windows has 95% server OS market share.

    You're not going to get a fair study, because Jboss will lose all the revenue comparisons (maybe having 1% the revenue of BEA and 0.01% the revenue of IBM).

    Websphere will never get a fair shake if you compare download numbers, because the Internet infrastructure itself can't handle more than 3 concurrent Websphere downloads.

    BEA and Sun will never get a fair shake because they're the reporters' and analysts' favorite whipping boys for the last couple of years. I swear if Sun cured cancer and solved world hunger, somebody would be kicking the 5h1t out of them in a front-page article the next day for something McNeally said.

    In the meantime, we all know that Microsoft is independently paying some independent third party to independently create an independent study to independently show that .NET is independently 28x as fast as J2EE .. which is so timely, what with J2EE being out of fad anyway.

    The more things change .. ;-)

    Peace,

    Cameron Purdy
    Tangosol Coherence: Clustered Shared Memory for Java
  50. Comments on BEA???[ Go to top ]

    I don't think we worry that BEA would go under this year or the next. (In particular, I have never bought a single share of BEA.)

    We should worry about the trend!

    Solaris used to be dominant serer OS. Today, it may still be faster than Linux, but it is now forced into a OSS model.
  51. I believe that the open source part of the industry is up for a bloodbath, and pretty soon, sure, some companies will survive, but the wast majority will go belly-up akin to 2001 dotcoms.Why?Heres why:Compelling customer benefits:Very few, sure you get the source, but that is a very small part, especially if you are selling to customers that don't consider software development a core part of their activities.Customer power:Massive, they can do whatever the OSS companies offer on their own, go to their direct competitor etc. The customers basically have the companies balls in a nutcracker.Threat of substitutes:Are you kidding? Your product IS a substitute to you. As is any moron who can understand it and knows how to register a company.Competitions:Fierce, the space is quickly becoming overestablished with developers wanting to make a quick buck on open source, and their mothers.Barriers of entry:None, anyone smart enough will be able to duplicate most of what you do in no time.Sustainable advantages:None, no patents, no proprietary code no nothing.Economics:You are competing against "we´ll do it ourselves" or "let the lowest bidder competent enough do it". The software is commodity. Wake-up call: so are your services. You are competing 90% on price. And on another note: how do you scale a business that is so dependent on man-power? Where will you find people that are smart enough? And how do you justify (and motivate) really smart and not exactly cheap resources to do mostly support work?I am not contesting that a few companies will survive, but at least 95% of them will be extinct this time in two years time. Most of the companies don't even HAVE a business model to begin with! Unless you count loosing money as a business model. I recall some company selling services (and add-ons?) for Maven, why? In the words of Duece Bigalow: "No, I won't pay, YOU pay me $5!".And there is no questioning that Open source has changed the face of the software industry. It is not just open source that will find it hard to make money, traditional software vendors will find it hard too.

    Hmmm. A hint: "The middleware market therefore appears doomed to fail even more drastically than other software markets. The success of free/libre/open-source software, originally created for ethical reasons, may be explained by interpreting it as a new paradigm that provides effective answers to the structural flaws of the market. Now that mainstream industry players realize that this approach makes economical sense, we propose a new rationale for middleware development. A business-neutral meta-organization federating vendors, customers and governmental agencies shall target the sustainable development of a business ecosystem where stakeholders could develop beneficial middleware strategies in line with their business and societal requirements."
  52. Then again, if you use GPL/LGPL a lot of potential users will walk right past you nowadays. So your actual odds for being successful may be worse. Sad but true.
  53. This is a rather honest blog entry, I think.
    It's perfectly OK IMO to make monkey out of your open source product. I'll have to take Marc's word when he says that's how JBoss got started, though...
    People make money out of propietary software all the time, under the form of training, support, magazine articles, plain reselling, and many others.
    In spite of that, I believe you take an inherent risk when you license your software as open source: you make easier for others to make money out of your effort. If you feel comfortable with that, then go ahead... and THAT is altruistic.

    Cheers and happy coding,
    Martin
  54. Your only option may be...[ Go to top ]

    IF you are writing BSD-licensed software AND YOU ARE SUCCESSFUL, your only option may be to go to work for your proprietary competitor who is only to happy to incorporate your work in their proprietary product. Beware of the man offering you candy.

    Following this cheery train of thought---

    If you are successful (regardless of the license you choose), then someone with deep pockets will probably accuse you of infringing on their patents and sue you.

    Your only option may be to go to work for somebody who has the where-with-all counter-sue (or move to Europe).

    My, isn't that a pleasant thought (moving to Europe) ;-)

    --John
  55. how many commenting contribute?[ Go to top ]

    As a regular contributor to open source projects, how many commenting have been participating for several years?

    I have my own reasons for contributing to open source and I sure has heck don't consider myself a hobbiest. I do it because often I need tools to do my job. Contributing to JMeter makes my life easier when I need to load/stress test an application. Contributing to drools is technically challenging and allows me to implement novel algorithms and techniques, which commercial rule engines have not considered.

    I for one just ignore the noise and keep on contributing. Open source is many things to many people. my bias 2 cents.

    peter
  56. how many commenting contribute?[ Go to top ]

    I have my own reasons for contributing to open source and I sure has heck don't consider myself a hobbiest. I do it because often I need tools to do my job.
    This is certainly an important motivation, and it's great that open source tends to empower users in such a way. People's experiences feeding back into projects is one reason that open source products often do a good job of meeting real-world needs. Nevertheless, people with an itch to scratch don't always commit to maintenance of their code once their itch has been scratched and they've moved on... Typically some corporate entity will need to emerge if that's to happen in the long term.
  57. I have my own reasons for contributing to open source and I sure has heck don't consider myself a hobbiest. I do it because often I need tools to do my job.

    This is certainly an important motivation, and it's great that open source tends to empower users in such a way. People's experiences feeding back into projects is one reason that open source products often do a good job of meeting real-world needs. Nevertheless, people with an itch to scratch don't always commit to maintenance of their code once their itch has been scratched and they've moved on... Typically some corporate entity will need to emerge if that's to happen in the long term.

    There in lies the crux of all software in general. Even business phase out or throw out old code. No piece of software will last forever and will have a death at some point.

    I would argue that major projects get a regular infusion of new developers and old developers move on to new things. The same is true of the commercial world. Developers join a company, work for a while and then move to a different job. Often a new developer takes over that component, and rewrites it, so the code changes or dies off all together.

    I define professional developer as someone who develops and supports the software they write. software that's just opened and isn't maintain is abandonware to me.

    peter
  58. This thread seems to have diverted to the viability of the Professional Open Souce business model. I just wrote up a blog myself (http://jboss.org/jbossBlog/blog/bbickel/2005/07/11/Will_Professional_Open_Source_become_dominant_in_Middleware.txt) myself describing the viability of Professional Open Souce business model. It will not work for every open source case, but it can work in some. The returns so far for JBoss have been positive.

    Bob Bickel
    JBoss
  59. I'd like to comment about the difference between Open Source Project being successful and Open Source Business being successful.
    The first is about having a stablished community around the open source project, active commiters and good steering by few good leaders. Of course, having a successful business on top of that is helpful, but not mandatory for the success of the open source project itself, as most have lived without a business until now. I've seen many arcticles questioning the success of a open source project based on the success of the business created on top of it, take JBoss for example: in case JBoss Inc. gets out of business, nothing will impede the open source project from going on happily without it, so that questioning is invalid.
    The other part is the success of the Open Source Business, which obviously depends heavily on the success of the open source project. And if the owners of OS project are the owners or the "active" employers of the OS business, the greater are the chances of success.

    Regards,
    Henrique Steckelberg
  60. +1 to this, but we should wait some time in order to get actual facts to back it up... ;)

    Cheers and happy conding,
    Martin
  61. False Dichotomy[ Go to top ]

    Marc is essentially proposing a false dichotomy between the BSD and GPL camps. I.E., either choose BSD, and run the risk of people taking the code private, or belly up to GPL and scare people with viral licensing.

    There's a third choice, which is growing in popularity, and that's the Mozilla style licenses. Copyleft without the viral. Of course, that's the one I'd choose to back, since we just released Marc's newest competition under a Mozilla-style CDDL license. (I'm speaking about GlassFish of course).

    As for BSD style licenses, Marc should remember that most OSS people are still doing it for love, not money. I think he keeps forgetting that.
  62. False Dichotomy[ Go to top ]

    Marc is essentially proposing a false dichotomy between the BSD and GPL camps. I.E., either choose BSD, and run the risk of people taking the code private, or belly up to GPL and scare people with viral licensing.There's a third choice, which is growing in popularity, and that's the Mozilla style licenses. Copyleft without the viral. Of course, that's the one I'd choose to back, since we just released Marc's newest competition under a Mozilla-style CDDL license. (I'm speaking about GlassFish of course).As for BSD style licenses, Marc should remember that most OSS people are still doing it for love, not money. I think he keeps forgetting that.

    My USD0.02...

    Projects at the Apache Software Foundation as well as others that use the BSD style license have proved that there's no real reason to legally compel contribution to the projects via some kind of virality in the license. People will do so for what I like to describe as "market forces". People who build on collaborative open source projects (like what we do at the ASF) realize that in general, they benefit too when they help make what they are building on be stronger or better. When you don't contribute back, you basically wind up with your own private branch. If that's what you want, great! Enjoy the freedom (and the maintenance overhead)! But I think you then are losing out on one of the major benefits of collaborative open source, sharing the work among many.

    Yes, the BSD-style license gives you the freedom to choose what you do with your intellectual property, and by and large, people tend to choose to give back. (For those that don't... enjoy the freedom!) This is one reason why Apache webserver, Apache Tomcat, Apache Axis, Apache Maven, etc, etc etc continue to grow and be so successful.

    That said, I do think that the Mozilla-style license is a reasonable thing too, as from my POV, a licensee still has the right to choose the fate of their original work, which I think is an important freedom. Yes, they have limits on changes they make to the existing work, but that's not entirely unreasonable, and they have the option to replace parts (in entirety) with replacements under a license of their choice.

    Open source projects provably thrive under all of these licenses - I suppose the choice just comes down to personal tastes and goals of the licensor.

    geir
  63. Marc Fleury has written a blog ...

    REMEMBER WE LOVE YOU

    Well, I don't want you to love me. And again. I don't want it :-))
  64. Why FOSS Contributors dislike commercial open source.
    Marc Fluery seems a bit puzzled why some open source developers sees commercial open source as betrail. I'd like to offer an explaination, but instead of addressing that question directly. I will start by discussing the potential revolution that free open source could become.

    Unlike most ressources free open source software is a rare kind of limitless ressource, that doesn't give each person that shares it a smaller share, no matter how many people share it. Ecconomic theory stipulates that society can maximize benefit of such ressources simply by ensuring unhindered access to it.

    Some free open source software contributors may see their work as a small revolt against capitalism, whereby they deny capitalists right to make a profit of the fruits of their labour. Free open source software can be seen as an alternative to commercialization where everything, even core services of the public sector are outsourced to private enterprizes for profit. Free open source could represent a turn in the tide whereby the people take back a sizeable chunk of the world ecconomy from commercialization. If you think about it the symbolic value of Linux stealing market shares from Microsoft can not be overestimated. It can be seen as an attack at the heart of capitalism, which would explain why neo-conservatives feel so strongly about open source and refer to open source as neo-communism.

    Contribute according to ability and receive according to needs. Free open source atleast has the first part allready and if the free open source ecosystem continues to grow, so it can provide competive solutions for most generic problems, the second part could be satisfied in the future.

    Commercial open source software does not give unhindered access(1) and as such can not fullfill the full potential that free open source software might. Commercial open source software erodes the open source software potential and it steals volounteers from projects that might be more deserving.

    1) Since support is a potential revenue stream for commercial open source. Commercial open source has a confict of interest in providing unhindered access to documentation for example. I'll ignore the fact that free open source documentation is often lacking as well.
  65. Since support is a potential revenue stream for commercial open source. Commercial open source has a confict of interest in providing unhindered access to documentation for example. I'll ignore the fact that free open source documentation is often lacking as well.
    I don't believe this. Certainly our attitude always has been with Spring that we provide the best documentation we can for free. We have put a lot of time into this, and continue to do so. We are also committed to providing free forum support. Currently 2 of the top 3, and 6 of the top 10, posters on the Spring forums are from Interface21. The result: people feel good about the software and find it easier to use. More people use the software, and speak positively about it to others, so the ecosystem grows. I don't think there's any conflict here. The wise course is to grow a bigger pie, not ensure that any time anyone takes a bite they need to pay something. So there's no conflict between growing a sustainable business and doing the right thing (helping users).

    I think respect for the community is essential, regardless of whether people are paying clients or not.

    Rgds
    Rod
  66. Since support is a potential revenue stream for commercial open source. Commercial open source has a confict of interest in providing unhindered access to documentation for example. I'll ignore the fact that free open source documentation is often lacking as well.
    I don't believe this. Certainly our attitude always has been with Spring that we provide the best documentation we can for free.

    I'll put another spin on your comments. Open Source projects may have quality documentation, but it is *very* difficult to maintain up-to-date documentation.

    JBoss really couldn't get any volunteers initially to maintain documentation until we forced users to pay for it. Once the documentation became such a tiny percentage of our revenue, we could make it free and also hire full time professional documentors like Norm Richards.

    So, in summary, at least for JBoss, quite the opposite happened. We didn't get quality, up-to-date, professional documentation *until* we commercialized.

    Selling documentation is a great way to bootstrap a business though. It enabled us to fund or subsidize the salaries of a few developers.
    We are also committed to providing free forum support.

    Providing free quality forum support gives potential customers a taste of your full, commercial support offerings.

    Bill
  67. What's value of service?[ Go to top ]

    If the product is openly available for free, how much can you really charge for services and documentation? Everybody with some talent can provide same or better services and documentation. Not believing? Just check out the JBoss books in your local bookstore.

    I would never spend more than $100 to fix my old TV, because a brand new TV costs only $200; but I would spend a lot more money to fix my car.

    SAP consultants used to charge $200/hour (not sure what's rate now) because not too many people have accesses to the SAP products or even documentations.
  68. What's value of service?[ Go to top ]

    If the product is openly available for free, how much can you really charge for services and documentation?

    As much as it is worth, I would assume. I still think there are various skills that will always pay. SAP consultants (well good ones eventually) can charge rather large amounts of money because what they know is perceived expensive in terms of the time it takes them to learn the technology and in terms of its impact on the business.

    Much the same with Oracle consultants. Consider the vast amount of documentation available for oracle. They are expensive because what they do is both critical and, well, hard. Oh, and don't forget, even if there are hundred of libraries out there, there is also a significant amount of time to learn them and learn them well, and that needs to be reimbursed.

    No mate, the more open source frameworks, the more libraries, the more unskilled workforce, the more money is there in fixing things. I really don't worry too much......
  69. I'll put another spin on your comments. Open Source projects may have quality documentation, but it is *very* difficult to maintain up-to-date documentation.
    From what I have seen, the same is true of closed source projects. They have trouble with quality too.

    I have had a lot trouble finding good/extensive documentation from a certain large software developer, either from books, manuals, their online website(s) and google. On the other hand, I usually can google up a solution on OSS projects or get it from books (usually multiple books if it is a good project). This holds true for some closed source projects like those associated with Java.
  70. The best open source projects - Apache, Mozilla, Open Office aim to create free and open software for basic components.
     Then commercial vendors can pickup these components and use then in commercial products that have license fee, support etc
    which users want.
     Also users can install via a free download and get suport by searching google.
     I don't believe we will ever see a Microsoft, Oracle or SAP equivalent company in Open Source but I am sure as software matures certain open source products will become the standard as Apache has for web servers.
     Professional Open Source will only be IBM, Sun et al contributing free development resources to Apache etc so they can get cost-effective components for their own software architectures instead of buying licenses from Microsoft.
  71. Hmmm. Did not know that such announcement was "newsworthy" according to TSS standards. Just my 2 cents here: http://os3g.blogspot.com/2005/08/economic-perspective-on-open-source.html (and another perspective on "professional open source".