IBM Purchases GlueCode Software

Discussions

News: IBM Purchases GlueCode Software

  1. IBM Purchases GlueCode Software (72 messages)

    IBM has announced its purchase of GlueCode Software, one of the new firms supporting open source integration, today. The purchase price is estimated at being under $100 million.

    GlueCode has a product, called JOE, that packages Geronimo with a number of other open source products to create an enterprise portal, and employs a number of Geronimo committers. IBM's product line for Java centers around their Websphere application server and products tied to it, including WSAD, Rational, and WebSphere MQ, targeted towards large enterprises.

    According to the press release, IBM will become an active contributor to Geronimo, and will contribute software to Eclipse to allow its use for developing, debugging, and deploying Geronimo-based application software.

    One industry insider commented that GlueCode's acquisition reflects pressure on the larger players to build on an open-source foundation, and focus on adding value to that, at least for the lower-end markets.

    What do you think? The open source support space is very active right now, with at least three companies (including GlueCode) being very aggressive in marketing integrated open source solutions. With IBM's purchase, does this change the nature of open source at all?

    Further, what do you think Geronimo's future will be? Will WebSphere gain characteristics of Geronimo, or vice versa? After all, most Geronimo and Derby contributors are now employed by IBM. Does IBM have too much power?

    Additional Resources:

    Threaded Messages (72)

  2. wow....or as cheech said "I can see clearly now"
  3. All,

    I'm a bit confused at to the value-add that GlueCode provided. I recall that they used to sell an enhancement to JetSpeed, however when I look at their most recent web site, I don't see any mention of the Portal server.

    Did the company transition from being a Portal server provider to one that productized Apache Geronimo and Pluto? If so, there this is really impressive marketing. I won't dig into this more, just draw your own conclusions.

    Regards,

    Carlos
  4. What did GlueCode sell?![ Go to top ]

    All,I'm a bit confused at to the value-add that GlueCode provided. [..] Did the company transition from being a Portal server provider to one that productized Apache Geronimo and Pluto? If so, there this is really impressive marketing.

    Gluecode had a good chunk of Apache-related engineers on staff. Among other things, they would be buying a good chunk of the Geronimo team, for example.

    Peace,

    Cameron Purdy
    Tangosol, Inc.
    Coherence: Cluster your POJOs!
  5. $100 million ?[ Go to top ]

    Is there a link that confirms the amount? the reuters article says financial terms were not disclosed, so where did 100mill come from ?
  6. What did GlueCode sell?![ Go to top ]

    Cameron,

    In other words, GlueCode in essence was more like a recruitment agent for IBM? Interesting business model.

    Regards,

    Carlos
  7. What does Gluecode Sell?[ Go to top ]

    Gluecode provided two things. First, they provided a "commercial" packaging of the Open Source projects and took on the responsibility to support it. Second, they also added a set of Administrative tools based on Portlets to provide additional GUI confiugration / administration functions.
  8. What did GlueCode sell?![ Go to top ]

    All,I'm a bit confused at to the value-add that GlueCode provided. I recall that they used to sell an enhancement to JetSpeed, however when I look at their most recent web site, I don't see any mention of the Portal server.Did the company transition from being a Portal server provider to one that productized Apache Geronimo and Pluto? If so, there this is really impressive marketing. I won't dig into this more, just draw your own conclusions. Regards,Carlos

    Hi There,

      If you look at the Harmony Announcement(http://blogs.cocoondev.org/dims/archives/003095.html)
    there are many gluecode employees. From BEA's jRockit commitment to Harmony and IBM's jikes and the newly borrowed Geronimo experience, IBM(and BEA) can influence J2EE in a much better way. IBM quickly moved to influence Harmony significantly.

       The ability to control(influence) J2EE on Linux(which is going to be the major deployment platform(or already is) ) is what IBM/BEA are after. Currently, Java cannot be distributed as part of a Linux distro. I believe that in 1/2 years Harmony(with IBM/BEA's contribution + other open source) should have a decent JVM that is comparable with Sun'. Once that happens, then distributing whole applications specific distros should be easier.

      Also, Java doesn't have a decent open source portal. If these(geronimo) guys have a portal too, that would fill out IBM's entry level offerings.

      Again, as BEA's executive mentioned yesterday, in a week from now we would be more knowledgeable as to the significance of Harmony in today's purchase.

    BR,
    ~A
  9. What did GlueCode sell?![ Go to top ]

    If you look at the Harmony Announcement(http://blogs.cocoondev.org/dims/archives/003095.html)there are many gluecode employees.

    There only was one Gluecode employee - me - and Apache Harmony has nothing to do with Gluecode. Gluecode focuses on Java application servers. It has nothing to do with J2SE implementations, and has no interest in J2SE implementations.

    -geir
  10. I think IBM is positioning itself as they did with derby recently.

    derby/geronimo - standards compliant & free for developers to get something up on - then move to WS/DB2 for production where the value adds make sense - with a nice easy migration path between the OS/Free version and the corporately respected/supported but not so free version.

    david
    javabarista.com.au
  11. I think IBM is positioning itself as they did with derby recently.derby/geronimo - standards compliant & free for developers to get something up on - then move to WS/DB2 for production where the value adds make sense - with a nice easy migration path between the OS/Free version and the corporately respected/supported but not so free version.davidjavabarista.com.au

    Things may go the other way: many people will find that they do not need Websphere/Db2. Geronimo/Derby is good enough and move down. More people will move down than move up.

    Ibm's move is more strategic one. It wants to have stronger influence on j2ee. This move will start a war. Jboss will react. Bea will react. Now, bea's development license is free. They may give a slim down version for production for free. That may attract more people that Geronimo. If you own a pet store and want to set a web site, do you care the source is open or not? If it is free and good, it is good.

    Sun is facing tough challenge. Sun needs some real influence, not just a spec leader.

    Why Geronimo has got so much attention? Because it is Apache. If Geronimo stays in Apache, any one can provide a j2ee server based on Geronimo by provide nice user interface on the front and optimize it at the back. The technical support is more on how to use it, not on how a jndi name finds it object.

    Wei Jiang
    Perfecting J2EE!
  12. IBM software online[ Go to top ]

    How IBM did the purchase of GlueCode Software? What companies can you recommend me for E-Commerce? I've read a guide about how to sell software online without any long range experience. the above website from Plimus is sort of outsourced spot for selling. Any idea about Plimus anyone?
  13. I believe GlueCode product is more zipping J2EE opensource projects. I would say vale added packaging business. They added sevice, management and performance tuning matrix.

    Only difference between JBoss and JOE is JBoss has their own core open source products.

    IMO, IBM is buying GlueCode Software as a prototype business model which is based on opensource projects. Thats the way, their strategy to gain the market in lower-ends and consulting business.
  14. Mmmh seems like a bad news for Jboss Inc. as IBM plans to directly compete with JBoss in the future to provide a free low-end J2EE server:
    http://yahoo.reuters.com/financeQuoteCompanyNewsArticle.jhtml? duid=mtfh65519_2005-05-10_12-32-13_n10729375_newsml
  15. Less than 100 millions? For what?

    Bad for JBoss, yes. That's the LGPL, boys. Otherwise you were certainly acquired long time ago.

    -- Andreas
  16. That or IBM just validated JBoss' business model ...
  17. or invalidated
  18. A more compelling analysis:

    http://www.forbes.com/markets/2005/05/11/0511automarketscan09.html
  19. This reminds me of about 6 years ago when IBM said "heh, why are we wasting time and money building our own web server when this Apache web server works and is superior?" Then they integrated Apache into their offerings and dumped their own web server.

    I can see IBM at some point saying "heh, why are we wasting time and money building our own J2EE application server when this Apache appserver works and is [in the future] superior?" Then they might make Geronimo their default server and build the rest of their tooling on top of it.

    When IBM puts it's weight behind a particular open source implementation, that could be a signal to the major enterprises to start looking at Geronimo over other alternatives.

    Floyd
  20. I don't expect we will completely replace WebSphere with Geronimo. WebSphere has been around a lot longer than Geronimo and has added a lot in the way of QoS and Scale. I don't know how long it will take for Geronimo to grow that same set of capabilities. It's important to understand that Geronimo is it's own independent community and set it's own directions. But if it continues in the path its on then I expect we will start consuming parts of Geronimo (just like I already consume elements of other Apache projects) in WAS -- alot of that depends on the engineering effort and other factors, but I am looking at it. For now, I view Geronimo as complimenting the WebSphere edition set -- providing for occasions that require basic J2EE function at a small footprint. Geronimo will continue to grow and evolve and we'll have to work out how best to balance between its small footprint characteristics and the desire to increase synergies between it and the things we've done in WebSphere. In the mean time, we will focus primarily on bridging Geronimo and WebSphere to make it easier for customers to move back and forth for the occasions that motivate the use of one or the other. We're going to do our best to become good members of the Geronimo community (building on the contributions that Gluecode have already been making) and contribute where we can. I think it makes sense to help this community be successful. We will all benefit from that.
  21. For those of you who don't know Rob, Rob is the chief architect of WebSphere at IBM.

    Billy
  22. Rob is the chief architect of WebSphere at IBM

    ok thanks,

    .V
  23. And it'll work on something more up to date than J2SE 1.3 too!

    -John-
  24. Ooouuccchh,
    John, those cheap shots hurt :) Give me a phone call btw, wondering what you're up to lately.

    Billy
  25. IBM's J2EE[ Go to top ]

    it seems that IBM has been scuessfull selling Apache HTTP.

    Now they also want to repleace WS (it has a bad rap) w/ code that alegedly derives from jBoss.

    Bye WS, good ridence.

    .V
  26. don't think that is true[ Go to top ]

    it seems that IBM has been scuessfull selling Apache HTTP.Now they also want to repleace WS (it has a bad rap) w/ code that alegedly derives from jBoss.Bye WS, good ridence..V

    I don't believe that is true. I've used older version of Websphere back in 99. I could be wrong, but I thought the code started from the original RI codebase.

    peter
  27. IBM's J2EE[ Go to top ]

    it seems that IBM has been scuessfull selling Apache HTTP.Now they also want to repleace WS (it has a bad rap) w/ code that alegedly derives from jBoss.Bye WS, good ridence..V

    Yep, in fact WAS has such a bad rap that it now has more than double the marketshare of the next-highest vendor.

    Vic, I understand you may have had a bad experience with WAS 3.5, but we're up to v6.0 now.

    Randy
    I work for IBM but don't speak for them.
  28. IBM's J2EE[ Go to top ]

    it seems that IBM has been scuessfull selling Apache HTTP.Now they also want to repleace WS (it has a bad rap) w/ code that alegedly derives from jBoss.Bye WS, good ridence..V
    Yep, in fact WAS has such a bad rap that it now has more than double the marketshare of the next-highest vendor.Vic, I understand you may have had a bad experience with WAS 3.5, but we're up to v6.0 now.RandyI work for IBM but don't speak for them.

    Nobody ever got fired for buying IBM, but I'm sure plenty
    of people have moved jobs rather than work with it.
  29. We all know IBM's manipulations with marketshare numbers...

    Try products of the next-highest vendor and enjoy the difference. Or trust to the survey of the theServerSide.com.

    Really - glue is definitelly what IBM needs in WebSphere...
  30. IBM's J2EE[ Go to top ]

    WS has such a bad rap that it now has more than double the marketshare of the next-highest vendor

    I think Tomcat, Resin and 1 and 2.

    I guess Oracle/Orion and jBoss are 3/4.

    And BEA/IBM at bottom of production market share. (Maybe tops in money spent?)

    (no I will not give out the same link again and again, but if you have one, great, i'll learn)
    Try to deploy a War to Resin or Orion. Then repeat w/ WS.

    Anyway, Geronimo will help. WS gave you a reputation, and this is a good way to fix.

    (don't worry about me, I would not use EJB 3.
    http://chris-richardson.blog-city.com/read/1212333.htm - it still has no support for lists. I use iBatis and hessian to SwingX. Maybe EJB 4 can return a list or rows )

    I wonder which IBM J2EE version will be deprecated in years to come, hmm.

    .V
  31. IBM's J2EE[ Go to top ]

    Yep, in fact WAS has such a bad rap that it now has more than double the marketshare of the next-highest vendor.Vic, I understand you may have had a bad experience with WAS 3.5, but we're up to v6.0 now.RandyI work for IBM but don't speak for them.

    If you don't speak for IBM then don't come on here with the marketing crap.

    Plenty of us have used WAS 4-5 and ran away screaming. Maybe 6 is better, but there were so many problems that I doubt you could have addressed half of them between 5 and 6.

    The marketing numbers are for $$$... And who knows how IBM breaks out WAS sales. They don't say. Probably some percentage of every hardware sale ends up on the sheet for WAS.
  32. WAS 5[ Go to top ]

    Well... I've used WAS5, and went into its use and administration with a fairly jaundiced eye. While it's not my favourite application server, I don't think it's incapable, and it's not unpleasant, when you consider the J2EE roles and deployment model.

    In fact, from that standpoint, outside of the, uh, startling speed of the admin console, it's pretty good. Those who "ran away screaming" perhaps had their own ideas of what "good" meant, and perhaps WAS5 didn't meet those ideas... and they overreacted.
  33. WAS 5[ Go to top ]

    Well... I've used WAS5, and went into its use and administration with a fairly jaundiced eye. While it's not my favourite application server, I don't think it's incapable, and it's not unpleasant, when you consider the J2EE roles and deployment model.In fact, from that standpoint, outside of the, uh, startling speed of the admin console, it's pretty good. Those who "ran away screaming" perhaps had their own ideas of what "good" meant, and perhaps WAS5 didn't meet those ideas... and they overreacted.

    Did you try automating it? Did you try versioning the WAS repository? Did you try using it without using the (crap) IBM tools?

    We had IBM consultants in for 2 MONTHS helping us to automate the build/deploy system from ant. Don't tell me about IBM's tools, I don't want to hear it. My automated build should NOT have to use a headless version of WSAD to deploy.

    I roughly estimated (and had agreement from the team) that developer productivity went down roughly 50% when we had to develop against WebSphere instead of WebLogic. Ant builds took over twice as long. Starting the server took over twice as long.

    In production it's only about as painful as WebLogic, but in development? If you want to have automated non-vendor-tool builds? I hope you have a book to read.
  34. Yes, we tried automating builds, no problem at all. I think it took one of our people just one or two days to do it, without using the headless WSAD jars. Why would you hire IBM consultants to do this basic stuff ?

    I have no idea where your productivity loss came from. In our current project, starting up the WAS test environment may take a minute, building the whole system from ant takes 2 minutes now I think: how would these kinds of numbers significantly influence programmer productivity ?!?

    Even better: often we don't have to restart the server at all during development, because the (not crap at all imho) IBM tools allow us to change the code from the debugger and simply continue. The same for the web front end: we use Tapestry which (in development mode) allows us to make changes to webcomponents without having to restart anything. You change something and have a new result within a second.

    Cheers, Luc.
  35. I'm sorry, but I don't believe you. Maybe MAYBE in WAS 6 they made things easier, but I don't think the turnaround was long enough for them to make this much improvement.

    I'm not the only one who's had these experiences... I got an email from someone after my last set of comments asking how to get this working.

    If your application can come up in 1 minutes and building your whole app with Ant takes 2 minutes then I don't think we're talking about the same type of apps... Are you sure you need WAS for that small a project? For us, building with Ant to auto-deploy to WAS took > 15 mins and just bringing up the app in WAS took 3-4 minutes.

    You don't work for IBM Global Services, do you?
    Yes, we tried automating builds, no problem at all. I think it took one of our people just one or two days to do it, without using the headless WSAD jars. Why would you hire IBM consultants to do this basic stuff ?I have no idea where your productivity loss came from. In our current project, starting up the WAS test environment may take a minute, building the whole system from ant takes 2 minutes now I think: how would these kinds of numbers significantly influence programmer productivity ?!?Even better: often we don't have to restart the server at all during development, because the (not crap at all imho) IBM tools allow us to change the code from the debugger and simply continue. The same for the web front end: we use Tapestry which (in development mode) allows us to make changes to webcomponents without having to restart anything. You change something and have a new result within a second.Cheers, Luc.
  36. I'm sorry, but I don't believe you.

    Well, I cannot do much about that. It is fully up to you to filter out unwanted input.

    I was just posting this as a datapoint that does suggest there may be other explanations than 'product A sucks' for the fact that some projects have problems with product A.

    The project we are working on is not small, and I don't work for IBM (never did either).

    Cheers, Luc.
  37. IBM's J2EE[ Go to top ]

    If you don't speak for IBM then don't come on here with the marketing crap. Plenty of us have used WAS 4-5 and ran away screaming. Maybe 6 is better, but there were so many problems that I doubt you could have addressed half of them between 5 and 6. The marketing numbers are for $$$... And who knows how IBM breaks out WAS sales. They don't say. Probably some percentage of every hardware sale ends up on the sheet for WAS.

    People on this site (always the same people) go on and on about how WAS is terrible, people run screaming, unsubstantited claims about how IBM must be cooking the numbers, etc etc. When someone asserts that today's announcement is because "WebSphere has a bad rap", that clearly deserves a response. The marketshare numbers and trends from Gartner and IDC argue against that assertion. Those firms have an interest (future reputation and credibility) in making those numbers as accurate and relevant as possible.

    Randy
  38. I have read the latest Magic Quadrant. It was an eye opener for me that Gartner does not verify vendors' statements at all. Gartner asks a question (e.g. if the server has feature X) and the vendor answers. That's more or less all ! NO VERIFICATION ! Observing how arrogant and noisy Larry Elison from Oracle is and having real life experience with IBM & Oracle consultants (they always answer 'yes, the product has the feature and it is the best'), it was no surprise for me that Oracle app server was so high.

    So, I simply do not respect Gartner's Magic Quadrants anymore. I believe much more in TSS survey about app servers and my own "I will play with it for 1-hour" test.
  39. IBM's J2EE[ Go to top ]

    People on this site (always the same people) go on and on about how WAS is terrible, people run screaming, unsubstantited claims about how IBM must be cooking the numbers, etc etc.

    I wouldn't say that WAS 5 and especially 6 are that bad. They are, well, OK.
    The only problem with all and every IBM product I've ever encountered is that IBM just doesn't know what the word "usability" means.
    One could look at their website to figure that.
    Or use that AS/400 half-assed DB2 database that says 'file is in incorrect state for this operation', which, translated into human language, means actually 'you trying to execute transaction for a table without transaction journal turned on'.
    And finally that nasty WAS web console that always brings you back to the beginning, as if it's very hard to remember last page you were on.
    No, I'm not saying IBM is doing bad software. It's not that bad. It's just not for humans, but rather for cyborgs. One could even see a Global Services $$$ conspiracy here - the less usable our products are, the more $$$ customers have to pay to our consultants.
  40. Usability[ Go to top ]

    How can you complain about usability? I'm shocked. On every page that is blank IBM prints "Page intentionally left blank." ")
  41. Usability is one story. The other is they make their products overly complex. Look at Portal or Workplace. To install them you must read at least 300 pages of install guide and usually around 350 pages of Release Notes (all the latest 'surprise-suprise behaviours' !). Or - go to 5-day installation course. Forget about actually using functionality of the products.

    Finally, WebSphere is a bunch of 300-400 products under one name. But these hundreds of pieces are rather poorly integrated (most of them are not at all). So, where is the benefit of buying all the pieces from a single vendor ? And do not forget IBM will add to the picture even more pieces - some of them not from WebSphere brand. So, more headaches for YOU.

    IMHO, IBM's software is like a glass vase, in which it is YOU who must try to glue all the pieces together. You rather won't be able to, so - after a few failures - you come to IGS anyway. Yep - it means more $$$ to IBM.

    I like Eclipse. My theory is Eclipse is successful because it is not IBM's project any more. So, good foundation (idea), which got finished and useful.
  42. Java Fan -
    It is interesting that you only started posting this week. If you feel so strongly, how come you post anonymously? If you have a vested interest you ought to reveal it.
    I work for IBM (but I not speak for IBM either)

    Tom
  43. Tom[ Go to top ]

    I respect you, but does your comment add any merit ? I do not favor any other product here. I am just pointing out what pisses me off in WebSphere. And btw not only me...
  44. Java Fan -It is interesting that you only started posting this week. If you feel so strongly, how come you post anonymously? If you have a vested interest you ought to reveal it.I work for IBM (but I not speak for IBM either)Tom

    Well, I've been posting here a long time, and I'm using my name. The WebSphere team had 2 people from the team I was on down to talk about usability and the problems we had.

    Some companies "get it" that a real development shop is going to want to automate builds and not have to have everything done through the vendor tool. They "get" that versioning the server configuration so that developers can just sync from version control and be ready to go creates big productivity wins. They "get" that taking an ENTIRE DAY to install the app server on your machine is unacceptable. IBM is not one of those companies.
  45. FUD[ Go to top ]

    They have listened. 6 supports rapid deployment, no need for WSAD/RAD. Just plonk your annotated code in a directory and off it goes.

    Why would it take an entire day to install? The installer takes about 15 mins.

    Why does all rational comment go out the window whenever websphere is mentioned? Yes, I know some people are still in therapy after bad 3.5 experiences - it has vastly improved since then.
  46. RE: FUD[ Go to top ]

    Have you tried installing WebSphere Portal or Workplace ? Did you managed to have it up and running in less than 1 day ?

    Regarding WAS: the installer might take 15 minutes. But then, you have to install all the patches. And some of them are on the install CD, most of them are on IBM's support (long, long list) and for some, hmm, you should ask your good friend at Big Blue to grab for you. Usability ?
  47. FUD[ Go to top ]

    They have listened. 6 supports rapid deployment, no need for WSAD/RAD. Just plonk your annotated code in a directory and off it goes.Why would it take an entire day to install? The installer takes about 15 mins.Why does all rational comment go out the window whenever websphere is mentioned? Yes, I know some people are still in therapy after bad 3.5 experiences - it has vastly improved since then.

    Don't label it as FUD because your experiences are different. In my experience, WAS 5 took up most of a day to install. There was the horrible installation routine plus the script we had to follow to get our repository all set up to be the same as what everyone else was using.

    Maybe 6 is better, I haven't had to use it, but I've heard really good things about WebLogic 9, so I doubt they've caught up.

    Personally, I prefer running Resin from within IDEA... takes about 1 second to start.
  48. [snip] Some companies "get it" that a real development shop is going to want to automate builds and not have to have everything done through the vendor tool.
    IBM gets that, they provide a custom ant task to build your projects. Or if that's too easy, you write your own. Maybe a page or two of Ant script. Really not sure what the difficulty here is.
    They "get" that versioning the server configuration so that developers can just sync from version control and be ready to go creates big productivity wins.

    We versioned our server config to CVS, new developer starts, they sync up with CVS, and they are off and running. Again, not sure what the difficulty here is. If you're suggesting that somehow versioning should be built in to WebSphere, I would disagree.
    They "get" that taking an ENTIRE DAY to install the app server on your machine is unacceptable. IBM is not one of those companies.

    Sounds like whatever method you are using needs improvement, and probably you would have trouble irregardless to the Web server. Have you bought the server, but are skimping on IDE's etc? Maybe spending a little money there, would save you much more money in your configuration issues.

    Granted, WebSphere is complex. It's hard to build a product that will run eBay, and also easily install on your desktop so you can run Hello World.

    Haven't RTA, but I'm guessing this is part of the reasoning behind the purchase.
  49. They "get" that versioning the server configuration so that developers can just sync from version control and be ready to go creates big productivity wins.
    We versioned our server config to CVS, new developer starts, they sync up with CVS, and they are off and running. Again, not sure what the difficulty here is. If you're suggesting that somehow versioning should be built in to WebSphere, I would disagree.

    The WebSphere repository is a huge and complicated directory structure. Making the smallest change can break the whole thing unexpectedly. By comparison, the WebLogic configuration is ONE xml file, which is pretty straightforward to understand and modify. The repository pretty much HAS to be edited via the slow and confusing admin panel.
    They "get" that taking an ENTIRE DAY to install the app server on your machine is unacceptable. IBM is not one of those companies.
    Sounds like whatever method you are using needs improvement, and probably you would have trouble irregardless to the Web server. Have you bought the server, but are skimping on IDE's etc? Maybe spending a little money there, would save you much more money in your configuration issues.Granted, WebSphere is complex. It's hard to build a product that will run eBay, and also easily install on your desktop so you can run Hello World.Haven't RTA, but I'm guessing this is part of the reasoning behind the purchase.

    So you're saying you shouldn't use WebSphere without WSAD? We used IDEA and ant, so having to switch to WSAD for WebSphere would have been another productivity hit on top of the problems with using WebSphere in general.

    For some reason you assume I'm comparing this to something trivial, like Tomcat. I'm comparing it to WebLogic, which, IMO, is a MORE feature rich and dependable enterprise deployment environment, so don't tell me it HAS to be this complicated, because I've seen the competition.

    The reasoning behind the purchase, as far as I could figure out, was that the IBM salesrep took the decision maker at our customer out to play golf, or some other equally technically-driven reason.
  50. Java Fan -It is interesting that you only started posting this week. If you feel so strongly, how come you post anonymously? If you have a vested interest you ought to reveal it.I work for IBM (but I not speak for IBM either)Tom

    He does sound a lot like the anonymous accounts JBoss use(d?) to make themselves seem more popular than they were...

    Mark's whining about how the purchase is a direct attacked aimed at destroying open source (a.k.a. his product...) is along the same lines.
  51. Please ![ Go to top ]

    Please do not stick any badges to me. I do not see JBoss (and other OSS app servers) mature enough yet (for my needs at least, read 'enterprise applications'). I do not like M.F.'s way of communication - I think he is in the same group of noisy, arrogant, 'black PR' people as Larry O.

    I am not saying "I do not like OSS at all", either ! I respect all the communities work. I just think not every OSS product is good yet (neither every proprietary product is). I love Linux, Apache Web Server, Eclipse. Those are mature enough for my needs. I respect what Spring brings to reduce complexity of some of J2EE's aspects (but I would prefer to stay within 'standard' J2EE, just 'made simpler').

    I believe as time flows, the communities' efforts will push some of the project high enough to fulfil for my needs (like with mySQL - the new version with stored procedures, triggers, etc. might finally grow enough to break into middle-level enterprise apps).

    You see, I have not mentioned JBoss - I do not like the JBoss Inc. company (I stress 'company', not the community). They are for profit company (which is good, but this should not be hidden under 'love, peace, open source' PR mantra they do). Do they bring any real innovation to the J2EE community ? All the time they have this "me-too" approach - look at their shift from app server to platform. Aren't they following someone ? JBoss Portal - hey - how many full time people are developing the product - 2 ? 3 ? Give me a break ! I need enterprise class solutions which are innovative and do not break ! Not toys consisting of glued pieces.

    I want to stress again my approach - read the TSS surveys, believe in the evaluations you (not others - read Garter) do. That's all.
  52. IBM's J2EE[ Go to top ]

    People on this site (always the same people) go on and on about how WAS is terrible, people run screaming, unsubstantited claims about how IBM must be cooking the numbers, etc etc. When someone asserts that today's announcement is because "WebSphere has a bad rap", that clearly deserves a response. The marketshare numbers and trends from Gartner and IDC argue against that assertion. Those firms have an interest (future reputation and credibility) in making those numbers as accurate and relevant as possible.Randy

    Who do you think they have credibility with? With the people on this site who know the technology, or with the CTO's and CEO's who like bright and shiny whitepapers?
  53. WS 5 ain't much better[ Go to top ]

    We're using WebSphere 5.1 and it isn't much better. It's got some nice features, like the fact that the messaging provider is built in and works for XA transactions. But overall the development tools are so flakey that they end up creating as many problems as they solve. I positively become nauseuous just talking about or thinking about WSAD. The server itself, though, doesn't work with any garbage collection strategy except stop-the-world mark and sweep, which is very limiting. We've had IBM "experts" out here and we have an entire team that has tested WS 5.1 with every single GC strategy and they all eventually melt down the entire cluster except for the most primitive mark and sweep. There are a lot of other issues, too.

    I would be very surprised if all of these issues are addressed in WS 6, but hey, anything's possible, especially with the big bucks at IBM. For us, a major bank in the US, we're trying to get off of WebSphere ASAP and be totally container independent for the forseeable future.
  54. IBM's J2EE[ Go to top ]

    Vic,

    You really need to keep your opinions upto date on WAS. XD and 6 are superb.
  55. $100 million ?[ Go to top ]

    $100 million sounds like an obscene amount of money for just a packaging of a bunch of open source projects. IBM could have given *me* the money and I would have created a ZIP for them with the whole stuff.

    As for replacing WebSphere with Geronimo, the very idea sounds laughable if you compare the feature sets or maturity of Geronimo with WAS 6.0 or especially WebSphere XD. Puzzling...
  56. $100 million ?[ Go to top ]

    Don't fall for this old trick.... "They" could have equally said "analysts suggest the deal was less than $1 Billion", and been technically correct. Right? I bet they could have said "less than $10 million" and still been technically correct...

     - Don
  57. Old trick[ Go to top ]

    Would they say "less than $100 million" if they actually bought the company for a fiver and some change ? Technically, it wouldn't be a lie.

    As for the product positioning, the graph in the press release makes it clear: it's for losers not even willing to pay $1000 or some more for WebSphere Express! :-)
  58. That one caught me off guard, but it makes sense. I've been saying for a while that as open source solutions squeeze the J2EE app server piece for the major vendors, they need to move up the value chain.

    On another note, this could add more validation to Spring, because they've done more and more integration with Geronimo.

    It will be interesting to see how this plays out...
  59. Very interesting, this should stir things up a little. :-)

    -John-
  60. I am glad to see IBM recognizing there is a market even below their "Express" line of products, which supposedly should be cheap enough for clients with fewer than 1000 users.

    But if then keep moving to OSS strategy (and they will, because there are good competitors like JBoss, Jonas, etc), their main revenue will change from selling software products to selling _services_, ie, a complete solution whatever commercial or OSS products used.

    And to deliver projects, well, as a multi bilion software company with CMMx certification, does will IBM compete well with others companies using agile much less bureaucratic processes ? I don't think so. So I guess they will reach a situation where they will realize they need start building agile teams instead of spending 40% or 50% of project efforts just to make sure team is following every CMM process, which can make sense for big customers but not for OSS customers, definitively.

    And for people here comparing WAS to Geronimo, this is totally incorrect. Sure, Geronimo is much better if you compare it against WAS 3.5 but this is not the case anymore since WAS 4.x. The current versions are too far superior both in reliability and features terms.

    Just my personal thoughts.
  61. But if then keep moving to OSS strategy (and they will, because there are good competitors like JBoss, Jonas, etc), their main revenue will change from selling software products to selling _services_, ie, a complete solution whatever commercial or OSS products used.

    Well, in fact they do it already and a lot of it.
    IBM (Global Services?) even hires freelancers and resells them for EUR 120-150 per hour to their big customers (==banks, insurance companies, etc.) and also runs its own huge offshoring facility in India.
    Considering all that, I would say IT services should definitely account for decent share of IBM income. It looks especially lucrative, as most of customers have strong intentions to outsource all and every IT activity.
    And what's important - obviously those big customers wouldn't even consider running something on open-source containers, but the container blessed by IBM can be even open-source - it doesn't matter, as long as IBM stands behind.
  62. Makes Sense.[ Go to top ]

    All philosophical considerations of open source aside, this makes perfect sense:

    (Apache Brand) x (Global Services Channel) = $$
  63. Interesting development[ Go to top ]

    This is very interesting development. It is good for j2ee.

    The interesting thing is that Geronimo is not as "pure-non-commercial" as apache web server. Many key players of it have directly link to commercial activities. Now, Sun gives Geronimo a free j2ee license and certification (will). Will that change?

    Wei Jiang
    Perfecting J2EE!
  64. Interesting development[ Go to top ]

    Now, Sun gives Geronimo a free j2ee license and certification (will). Will that change?

    We have to wait and see what excuse sun comes up with to revoke the license. If sun revokes the license, i am guessing IBM won't mind spending few more dollars to get Geronimo certified.
  65. Interesting development[ Go to top ]

    This is very interesting development. It is good for j2ee. The interesting thing is that Geronimo is not as "pure-non-commercial" as apache web server. Many key players of it have directly link to commercial activities. Now, Sun gives Geronimo a free j2ee license and certification (will). Will that change?Wei JiangPerfecting J2EE!

    No, it won't change. Apache Geronimo is an open community of the Apache Software Foundation, governed by Apache rules, and anyone is welcome to participate. In fact, everyone is invited to participate.

    This is how it always has been, and will be.

    -geir
  66. you can read Fleury's pathetic response in his blog
  67. This is awesome. Congrats. Let's just hope that Geronimo will speed up a little bit through this, will ya ?! :-D . Just joking. I'm just happy you guys have earned some serious backup. The model has proven to work pretty darn well (look at hibernate/jboss, spring/interface21 etc.) I just simply love some pieces of software from you guys (read ActiveMQ for example).
  68. This is awesome. Congrats. Let's just hope that Geronimo will speed up a little bit through this, will ya ?! :-D . Just joking. I'm just happy you guys have earned some serious backup. The model has proven to work pretty darn well (look at hibernate/jboss, spring/interface21 etc.) I just simply love some pieces of software from you guys (read ActiveMQ for example).

    Note that IBM haven't bought the main company providing services on ActiveMQ - yet :)

    James
  69. Bickel Blog Blown Back[ Go to top ]

    The jBoss groups original response to the IBM/Gluecode news was posted by Bob Bickel on the jBoss blog. Bob openly ranted about IBM's move to purchase GlueCode and accused IBM of viciously defending their market. The blog posting must not have been the spin Fleury wanted, as it was removed from the site. In its place is a softer spin from Fleury which (properly) relates IBM's move as a favorable one in that it brings some credibility to all open source software, especially in the corporate markets where IBM rules.

    The retracted "permalink":


    In its place is Marc Fleury's softened spin:

    [Fleury isn't taking comments on his post, unlike nearly all other posts in his blog.]
  70. Bickel Blog Blown Back[ Go to top ]

    The jBoss groups original response to the IBM/Gluecode news was posted by Bob Bickel on the jBoss blog. Bob openly ranted about IBM's move to purchase GlueCode and accused IBM of viciously defending their market. The blog posting must not have been the spin Fleury wanted, as it was removed from the site. In its place is a softer spin from Fleury which (properly) relates IBM's move as a favorable one in that it brings some credibility to all open source software, especially in the corporate markets where IBM rules.

    Ah, love that Google cache: http://www.google.com/search?q=cache:wjVeceFJg7wJ:jboss.org/jbossBlog/blog/bbickel
  71. Cached version
  72. Bickel Blog Blown Back[ Go to top ]

    from google cache:

    ------- start quote -------
    Blog/bbickel/

    Big Blue Gets Religion?

    IBM announced it has acquired Gluecode (http://home.businesswire.com/portal/site/google/index.jsp?ndmViewId=news_view&newsId=20050510005542&newsLang=en). In so doing, IBM is acknowledging the overwhelming momentum behind open source in general and the success of JBoss in particular.

    What we at JBoss see every day is that CIOs want to create strategies around open source. They increasingly trust it to run their core business applications and the technology is available today to support mission-critical deployments. IBM’s embrace of open source will be a shot in the arm for JBoss and the rest of the open source movement.

    At the same time, IBM’s acquisition is an overtly defensive move (http://www.nytimes.com/2005/05/10/technology/10blue.html). JBoss is being adopted so rapidly and in such volume (BZ Research reports increases in market usage for JBoss from 13% to 26% to 34% over the past 3 years in terms of production usage) that it is becoming, in effect, the next Linux at the middleware layer. This naturally makes WebSphere (IBM was at 33% in the same study by BZ Research) the next Solaris.

    IBM has in many ways welcomed Linux, but is hostile toward open source products that may affect the cash cow that is the WebSphere business. This intent of the acquisition, we believe, is to slow down JBoss and to try, with Geronimo, to capture the high-volume J2EE space.

    But IBM is betting on unproven technology. Today, JBoss is the high-volume J2EE leader. The JBoss ecosystem is active and enormous, and as a result we’ve got great technology. We intend to keep our open and large community and ecosystem happy and thriving.

    IBM customers should be wary of a potential bait and switch ­of IBM suggesting Geronimo for low-end deployments, then pushing WebSphere to do the heavy lifting. This is also creating issues for customers and partners who may feel pressured to support yet another application server. We question how robust Geronimo can be given that its backing has primarily come from Gluecode, a company that’s been known for some time to be in financial trouble and lacking in community support (Less than 100 posts in 7 months of existence compared to 3,000 per month on the JBoss forums- http://www.gluecode.com/forums/index.jspa). Geronimo is obviously not J2EE certified. It is obviously low-end. It obviously is not tracking the EJB3 spec. Gluecode has written far less code (for example, in Q4 of last year Geronimo had 800 total new commits, while JBoss had over 7,000). Geronimo also does not offer customers the full vision of the JBoss Enterprise Middleware System.

    JBoss customers and partners have a single, unified, always free license to our software. JBoss users will not be presented with what we anticipate will be a forced trade up from Geronimo to WebSphere. JBoss software is capable of running high-end applications; Geronimo is not. And JBoss does not force customers to purchase expensive long-term contracts. Frankly, we question whether IBM's dual interests will keep Geronimo from reaching the full capabilities of open source that JBoss has reached.

    It will be interesting to hear IBM articulate a strategy on how they merge WebSphere and Geronimo and handle Gluecode’s dual-licensing strategy. IBM will also need to explain its strategy for the Gluecode Portal and IBM’s products in this area.

    JBoss will continue to work with IBM Global Services and the IBM Hardware teams because of strong customer demand. We will also continue to work with IBM on various standards.

    Overall, we see this as a move that should clarify things for customers and partners: open source is the here and now. Companies are implementing open source middleware­ - not just Linux­ today. The JBoss ecosystem (contributors, partners, customers, users) is central to that movement and will continue to be the technology and market leader for open source middleware.

    Posted on Tue, 10 May 2005 12:46 by admin
    ------- end quote -------
  73. Blog[ Go to top ]

    Bobby Woolf's blog discusses the Gluecode acquisition:

    http://www-128.ibm.com/developerworks/blogs/dw_blog_comments.jspa?blog=392&entry=81035