BEA offers two new entry level editions of Weblogic

Discussions

News: BEA offers two new entry level editions of Weblogic

  1. BEA has announced two new entry level editions of WebLogic. Weblogic Express (announced in the past) will be targeted at JSP only applications and WebLogic Workgroup Edition gives the full features of Weblogic Server for 20 users or less. Both use the same codebase as the full version.

    The prices of these servers are at $694 per Weblogic Express server basic edition (web-based support, no clustering), $1145 for the premium edition (clustering + live support) and $4,000 per CPU for Workgroup.

    BEA's Press Release can be found at : http://www.bea.com/press/releases/2003/0204_wlx_entry.shtml

    Pricing information was pulled from this CNET article:
    http://news.com.com/2100-1001-983234.html?tag=fd_top.

    Threaded Messages (65)

  2. Tomcat forever :-)[ Go to top ]

    BEA WL has superior EJB/JMS/WS implementation, but JSP...

    Why not use Tomcat/Jetty for JSP only apps ?
  3. Tomcat forever :-)[ Go to top ]

    Because Tomcrap sucks too?
  4. Reasons please....[ Go to top ]

    Because Tomcrap sucks too?


    It's too easy to say Product X sucks. Doing so just makes the rest of us think you're incompetent and can't configure Tomcat properly.

    Why not offer up your reasons why you think Tomcat is not a suitable solution for simple JSP/Servlet based Web apps?

    ...
  5. Reasons please....[ Go to top ]

    It's too easy to say you're incompetent and can't configure Project X properly. Doing so just makes the rest of us think it sucks.
  6. Enough[ Go to top ]

    cmon guys. Don't start it again.
    I'm using Tomcat for production web applications. My customers are happy. There are hundred thausands of developers who build web applications on top of the tomcat.
    If you wanna pay money for servlet engine - that's your choice, not mine. But I'm not saying that your choice sucks. And I don't think mine sucks either.
  7. So much for that....[ Go to top ]

    So much for trying to make this a usefull and interesting discussion...

    Thanks Yury
  8. Tomcat forever :-)[ Go to top ]

    |
    |Why not use Tomcat/Jetty for JSP only apps ?
    |

    If you already use Weblogic, but want a lower cost alternative for such apps. Or because you are also using weblogic as the EJB container in a multi-physicahl-tier deployment.


    -Nick
  9. Tomcat forever :-)[ Go to top ]

    For non-mission critical, JSP only departmental applications,Tomcat forever makes good sense. more than enough!

    shirke
  10. Tomcat forever :-)[ Go to top ]

    I prefer J2EE web containers like Tomcat 4.1 or Caucho's Resin if I only need Servlets, JSPs, JNDI resources, and maybe JTA transactions - but no EJB, JMS, or JCA. They give me everything I need - in the case of Resin with good documentation, very good email support, and at a very reasonable price (500 USD per server, not per CPU).

    "Light" versions of J2EE enterprise containers tend to be as bloated and heavyweight as their full versions, just limited by their license file. Look at WebLogic or Sun ONE 7, >100 MB download compared to <10 MB for Tomcat or Resin! Installing or updating Tomcat or Resin is a breeze too, just download the file, unzip it, and off you go with the default config. Generally, I like to be able to use (and prepare) config files, and Tomcat and Resin are far simpler in this respect.

    So I'm not sure if both BEA WebLogic Express or Sun ONE 7 Platform Edition are appropriate for their supposed target market in terms of weight, especially when there are Tomcat and Resin around. I would only choose the standard editions of big iron servers for demanding enterprise projects, and stick to lightweight servers for web-centric projects.

    Ah yes, if I would have to choose a server with a big name because of my company's management, then I guess I would choose such entry editions, be it from BEA or Sun ONE, or Macromedia's JRun 4. But fortunately, that's not the case.

    Juergen
  11. Tomcat forever :-)[ Go to top ]

    Sophisticated XA support via JTA may be a special issue. Tomcat 4.1 needs Tyrex 1.0 as separate add-on, with special configuration involved. Resin only has a bare-bones transaction manager for distributed transactions.

    So if I had demanding requirements in this respect but no need for EJB, I would consider WebLogic Express and JBoss - and probably find the best combination of reliability and extensive documentation in WebLogic.

    Concerning such projects, I indeed welcome the availability of WebLogic Express. On the other hand, I guess that 99% of all the web-centric J2EE projects work with a single database...

    Juergen
  12. JBoss offers two new Versions[ Go to top ]

    FREE and FREE. And with each version you can get full support for J2EE 1.3 including clustering, support from "the source", and hooks into the container to add customized behvaior.

    I would hope that people would not pay $694 for Jetty or Tomcat, that seems foolish. IF you want a free version of clustered Tomcat or Jetty, download JBoss 3.x and rip out all the services except Tomcat or Jetty, and viola, you have Tomcat or Jetty with hot deploy.

    I am very interested to see who buys these limited versions. Where do you get $694 from? Are things that bad that you can not afford to give that away for free? I thought it was a forgone conclusion that the servlet / JSP layer has been commodity for some time.

    For all of you out there, before you spend $4,000 per CPU for a workgroup version of WebLogic, download JBoss, use it and realize you money can stay in your budget for something else. Spend money on knowledge not software. Invest in your people so they can write better code.
  13. Free and Free == Bad and Worse[ Go to top ]

    <ben>
     Invest in your people so they can write better code.
    </ben>

    Paying for WebLogic doesn't keep them from writing "better code", in many cases it *eliminates* the need to write code at all!

    compared to WLS, JBOSS leaves 10,000s of lines of code unwritten...
    (e.g. configuration wizards,
    scalable management architecture,
    application monitoring,
    self-healing services,
    multiple thread queues,
    flexible security model,
    clustered native JMS,
    3rd party product interoperability,
    enterprise-class CMP with cluster-wide caching,
    virtual hosting,
    XMLBeans support,
    native conversational webservices,
    seamless upgrade path to Portal or Integration or Liquid Data,
    etc. etc. etc.)

    Matt
  14. Free and Free == Bad and Worse[ Go to top ]

    Matt,

    Let's see here. Let's go one at a time.

    config wizards == xdoclet

    scalable management architecture == netboot + farm
    application monitoring == jmx
    self healing? bull, I would love to see it.
    Multiple Thread queues, WE GOT THAT
    flexible security model? We integrate with LDAP among other things
    clustered native JMS? Well, they got us there
    3rd party product interop? We are open source and based on JMX, so anyone who can add their software as an MBEAN can plug into JBoss. We have several software partners.
    enterprise-clas CMP? Sure we got that. Cluster-wide caching. yup, got that too in 3.2 plus, check out Caches on JBoss under development in 4.0
    virtual hosting? Jetty/Tomcat support that
    XMLBeans? Proprietary technology that could be interesting, yet to be determined if anyone cares about this.
    native converstaional webservicdes? Yup, jboss.net

    So, in general Matt, do your homework on your competition before you put your foot in your mouth again.

    I think you have a good product, just a shrinking market for people who are going to pay for commodity hardware. I guess you have never used JBoss. What do you do for BEA?
  15. JBoss and Weblogic Compared[ Go to top ]

    <ben>
    config wizards == xdoclet

    scalable management architecture == netboot + farm
    application monitoring == jmx
    self healing? bull, I would love to see it.
    Multiple Thread queues, WE GOT THAT
    flexible security model? We integrate with LDAP among other things
    clustered native JMS? Well, they got us there
    3rd party product interop? We are open source and based on JMX, so anyone who can add their software as an MBEAN can plug into JBoss. We have several software partners.
    enterprise-clas CMP? Sure we got that. Cluster-wide caching. yup, got that too in 3.2 plus, check out Caches on JBoss under development in 4.0
    virtual hosting? Jetty/Tomcat support that
    XMLBeans? Proprietary technology that could be interesting, yet to be determined if anyone cares about this.
    native converstaional webservicdes? Yup, jboss.net

    So, in general Matt, do your homework on your competition before you put your foot in your mouth again.
    </ben>

    As someone who has used both Weblogic and JBoss extensively I feel in some position to compare the two. When you pay your megabucks for Weblogic a large chunk of your money is going to technical authors - the Weblogic documentation is much, much more extensive and professional than the JBoss Group could ever provide on their own. Also the BEA staff are very quick and professional in their responses to questions on the newsgroups (all of which are archived by Google). When I had problems finding a solution was never that hard.

    JBoss, on paper, can do an awful lot that Weblogic can, but that's only on paper. You admit yourself that many of the features you are describing aren't in the mainstream version of JBoss. Also, getting JBoss to do the simple things (deploying basic beans) *is* simple, but doing the complex things seems the require you to understand a large portion of the JBoss internal workings. Encapsulation, anyone? What about an easy to use management console? No, JMX does not count - I never had to learn JMX to administer Weblogic.

    I don't want to be too critical of JBoss - I use it all the time and it's an incredible piece of software. But I'm afraid I'm fed up of hearing about all the great things JBoss will do and how it has an incredible architecture that goes beyond any of the competition. The bottom line is that most developers need to get things done and don't have time to learn about the internals of the app server or to care about what the marketing department is promising for future releases.

    What can we do today, easily? When you have a deadline that is what counts. If that is your measure then I think JBoss has a very long way to go before it beats Weblogic. Don't get me wrong, I'd like to see JBoss succeed, just don't expect my "love" to be blind.

    Ian.
  16. JBoss and Weblogic Compared[ Go to top ]

    Ian Fairman wrote:
    <Also, getting JBoss to do the simple things (deploying basic beans) *is* simple, but doing the complex things seems the require you to understand a large portion of the JBoss internal workings.>

    Hmmm...we are a software company using JBoss to deploy our application. It's not a simple app, it's kind of an ERP system with 400 ejb beans. 10 developers with only java and basic j2ee knowledge. I has done some digging in the JBoss sources and that helped a lot 2-3 years ago when we didn't know if the bug where in our code or in JBoss. Now JBoss is so mature and stable that I don't have any reason to creep under the hood (I do that but mostly because I love it... it isn't needed :). We have 400-500 JBoss's up-and-running WITH NO PROBLEMS AT ALL!

    And Ian also wrote:
    <the Weblogic documentation is much, much more extensive and professional than the JBoss Group could ever provide on their own>

    Yep, they certanly has a LOT of docs... problem is to find your solution :) Maybe if you have a really nice support contract with them you get fast solution - don't know :) But comparing of finding fast solution for free there is no doubt that you will get it faster from JBoss then from BEA. Last week I wasted a day or two on looking for docs telling me how to put up an evalutaion installation of Tuxedo and Jolt - What I had to do What I had to compile Why things didn't work... Shouldn't I be able to make a test by just running it out-of-the-box when downloading an evalutation???

    /Lennart Petersson
  17. JBoss and Weblogic Compared[ Go to top ]

    400 entity beans ? cmp ?
  18. JBoss and Weblogic Compared[ Go to top ]

    ok... counting them.... maybe 300 beans are more correct :)

    About 150 of them is entity beans and only CMP (still using ejb 1.1).

    /L
  19. JBoss and Weblogic Compared[ Go to top ]

    <lennart>
    Ian Fairman wrote:
    <Also, getting JBoss to do the simple things (deploying basic beans) *is* simple, but doing the complex things seems the require you to understand a large portion of the JBoss internal workings.>

    Hmmm...we are a software company using JBoss to deploy our application. It's not a simple app, it's kind of an ERP system with 400 ejb beans. 10 developers with only java and basic j2ee knowledge. I has done some digging in the JBoss sources and that helped a lot 2-3 years ago when we didn't know if the bug where in our code or in JBoss. Now JBoss is so mature and stable that I don't have any reason to creep under the hood (I do that but mostly because I love it... it isn't needed :). We have 400-500 JBoss's up-and-running WITH NO PROBLEMS AT ALL!
    </lennart>

    I've no doubt JBoss can scale to large applications. It just worries me that I have to understand so much of the JBoss internals to get it to work. I'm afraid the ability to look at the source code of someone elses large application is no appealling when you have a tight deadline. You can create a large EJB-based system without touching much of the middleware infrastructure, but sometimes you create small systems that require a much higher level of infrastructure support, i.e. authentication, security context propagation.

    Just as a matter of interest, what security features were you using? If you did was it easy to use? Personally I've found security setup a bit of a nightmare with JBoss (and something where the process seems much more straightforward in Weblogic).

    <lennart>
    And Ian also wrote:
    <the Weblogic documentation is much, much more extensive and professional than the JBoss Group could ever provide on their own>

    Yep, they certanly has a LOT of docs... problem is to find your solution :) Maybe if you have a really nice support contract with them you get fast solution - don't know :) But comparing of finding fast solution for free there is no doubt that you will get it faster from JBoss then from BEA. Last week I wasted a day or two on looking for docs telling me how to put up an evalutaion installation of Tuxedo and Jolt - What I had to do What I had to compile Why things didn't work... Shouldn't I be able to make a test by just running it out-of-the-box when downloading an evalutation???
    </lennart>

    We did have a support contract but as a lone developer of a subsystem in Weblogic using EJB/CMP 2.0 I never once put in a support call - I answered every problem with searches of the web, the Google archive or with the Weblogic docs.

    Maybe the real issue is one of team size. Ten developers can provide a lot of support for one another. I've never been on teams that size and we've never had more than a few developers working with JBoss, all wrestling with different features for different projects. With Weblogic the admin side is so much more straightforward - they don't assume you are a coder, whereas JBoss Group seem to.

    Again, to make this clear, I like JBoss, I just don't feel it is where it needs to be yet.

    Ian.
  20. JBoss and Weblogic Compared[ Go to top ]

    One of the main reason companies are afraid to adopt open source is support. I am sure companies with money and non critical applications will go in for weblogic just for the support and documentation.

    There is also a huge pool of talent to draw from for companies using weblogic.

    I am wondering at these prices weblogic will be able to cover the salary of the marketing guys. I think this is more to build brand loyalty and increasing customer base.

    Raghu
  21. JBoss and Weblogic Compared[ Go to top ]

    Raghu,

    Companies are not affraid to go open source AFAIK.
    http://www.sdtimes.com/news/063/story2.htm
    See market share of your vendor, vs top dog.

    Also, what other framework is more popular, lets use # of published books then Struts?
    http://jakarta.apache.org/struts/index.html
    How many books on your framework?

    I consulted on several convert comercial to open source projects.
    #1 reason is support is better on open source.
    Post a questoin on Struts/Tomcat and see the quality of response.

    Then enter an issue with a comerical vendor and wait a week. It's a test of wills.

    And last, major reason for open source is quality:
    http://www.opensource.org/advocacy/case_for_business.php

    For example lots of companie ban Windows for Internet becuase of security, and use Linux.
    http://www.netcraft.com/survey/

    It is ignorant to say that majority of people want to pay for inferior support and quality.

    Do you have ANY data to support that comrecial is:
    - better support
    - more market share
    - more reliable
    - costs less to operate in large enviroments

    Only newbie managers think that by paying more they get more. Sort of like somone wakling down the stree with Gucci shades, because they have better UV protection. They don't.
    But since you have paid a lot for a name brand, you fell good. You have less money.

    With open source, once can acctualy generate profit for their organization, since the cost of operations is low.

    .V
  22. EXPERT BACK OFFICE[ Go to top ]

    With Jboss, you get support from "the source" menaing the people who write the code. We do not tell you, sorry this is not a bug. Instead we say, let's have a look at your code and see if we can find this problem.

    The perception of support will be erased very shortly as their are some VERY large companies running highly mission critical systems on JBoss that are about to speak out in the press like business week, CIO magazine about the quality of the JBoss Group Expert Back Office.

    Companies like Worldcom, Corporate Express, GetThere (part of Sabre), and a lot of others have made the leap and gotten superior support.

    If you feel like to need to know more about JBoss, spend $3,000 on training instead of 4K on a single CPU license of a comparable product.
  23. JBoss and Weblogic Compared[ Go to top ]

    Well i consult for the US goverment. They will not install any software without support and License. Period.

    If you have been there done that, then it is easier to vouch for open source. But If I am a new manager, in charge of a critical project and my job depends on the failure or sucess of a project.
    - I will buy support from the vendor.
    - I will go with a proven product.
    - I will use products my team is familiar with.
    Not everyone has a very high tolerance to risk. Not every project has the luxary of learning from experimenting.

    Its not easy to get rid of microsoft. We get over 2 million hits a month on our website. 96% use IE. Thats a hard number to beat. Infact we build code to overcome known problems in IE. Even though IE is crapy its not going away any time soon.

    Vic if you look at JBoss they are trying to buld a credible support system. Even they recognize the value of support as a major selling point.

    raghu
  24. JBoss and Weblogic Compared[ Go to top ]

    Raghu,

    No one should use anything that is not supported! Agree.
    I am saying that open source has proven better support than comercial.
    Like I say post a question of a comercial vendor, then post a question here:
    http://www.mail-archive.com/struts-user%40jakarta.apache.org/

    See who gave you a better, faster more qualified answer. That is proven.

    What you I think are saying is that if you "buy/$pend" you will fell better with the less money, and worse support.
    You think you look cool in Gucci shades, and I think you look silly.
    (not personal, just arguing a point)


    .V
  25. nice[ Go to top ]

    vic,
    nice analogy, but i heard/read somewhere that "car analogy" is better.

    it says buying closed apps is like buying a ferrari(name any car you want here) with a bolted hood, and the company who sold it always argues that it comes with a support, but funny side is, if you call the support; most of the time if not 99.9%, you will hear "oh it's a known problem... it will be fixed in the next service pack, slated on so in so...!!!"

    and add to that, you can't even race the car with other cars and announce the result, because they will sue you hard. and worst it comes with a EULA, if you bother to read it. you don't actually own your car, they still own it. isn't that funny.
  26. JBoss and Weblogic Compared[ Go to top ]

    Cost of a Applicaton Software is very very very Minor. (especially for a mid-large scale projects)

    Just to try to save some bucks and use a Unsupported/Non-Benchmarked software (I am not pointing to any Freeware it can be commercial non-benchmarked software) and adding additional UNKNOWN overheads will propogate to the end until and including lot of other costs.

    So it is worth buying/licensing a Proven, reliabale, benchmarked software for creating a product.

    I am not against any Freeware products. Many have proven very reliable. But it all depends upon who uses it and how they support/maintain it and where it is being used and how it fits into the architecture.

    This may either require additional Support or a very skilled person to operate without any other guidance within the specified time.

    Not all big companies would like to go into such a mumbo-jumbo. For them this is minor charge.
  27. Free and Free == Bad and Worse[ Go to top ]

    Well, $694 is practically for free, at least for the majority of customers.

    In the choice between BEA and an open source project which has the following on their web site, and I quote:

    "That's quite alright (it is), if you want to play long run, pace yourself, do something relaxing, take a break, smoke a doobie, just look at a brighter day, heck!"
    (http://www.jboss.org/developers/join.jsp)

    Which would you choose if you were a manager responsible for picking the future software infrastructure of your company? I would pick JBoss, but then Im not making these kind of decisions, and I suspect that people who do might not...
  28. Free and Free == Bad and Worse[ Go to top ]

    Ben,
    Thanks for the kind remarks. However, your response conveniently missed the entire point of my post...

    To a reasonable person app monitoring != jmx and config wizards != xdoclet. Apache Axis (Jboss.net) requires MUCH more knowledge and coding than weblogic workshop and provides a fraction of the functionality
    (and http sessions != conversations).

    The cost/benefit of JBOSS goes away when you need to develop this additional functionality by hand and debug it, document it, maintain it, and train people on it. To claim otherwise is a diservice to the whole industry.

    BTW, Check out one of the "well-supported" jboss users trying to determine how to configure thread queues in jboss via the jboss.org forum... posted feb 3. 48 hours later -no response yet - and counting...

    http://www.jboss.org/forums/thread.jsp?forum=121&thread=27908&message=3763622&q=thread+queue#3763622

    regards,
    Matt
  29. For Matt Gunter[ Go to top ]

    Matt,

    What exactly is your role with BEA? The forums are FREE. If someone wants to get real support they can pay for it, for a lot less than a WL license.

    Once again, I have no problem with your product except for the fact that its' price is greater than ZERO.

    Something must be going on at BEA as we have talked to our fourth financial analyst of the week. They all wanted to find out more about JBoss and how it is taking customers away from BEA.

    Somehow Microsoft gets it and they are a lot larger than BEA. In a quote from their recent SEC statements,

    http://news.com.com/2100-1001-983373.html
    "To the extent the open-source model gains increasing market acceptance, sales of the company's products may decline, the company may have to reduce the prices it charges for its products, and revenues and operating margins may consequently decline," Microsoft said in a filing last week with the Securities and Exchange Commission.
  30. For Matt Gunter[ Go to top ]

    |
    |The forums are FREE
    |

    So are BEA's....

    And BEA's free ones are often monitored by some of their core developers.

    -Nick
  31. For Matt Gunter[ Go to top ]

    |

    >|The forums are FREE
    >|
    >
    >So are BEA's....
    >
    >And BEA's free ones are often monitored by some of their core >developers.

    So shall we all now go to the FREE BEA forums and start looking for questions that did not get answered in 48 hours and claim BEA users aren't well supported?

    Do you think we will not find such a case?

    BEA is going down. These people are getting desperate.
  32. For Thomas Mattson[ Go to top ]

    | So shall we all now go to the FREE BEA forums and start
    | looking for questions that did not get answered in 48 hours
    | and claim BEA users aren't well supported?

    No, thats not my point.
    The point is that their support is pretty good - and most of their newsgroups provide very good support. Many like yourself have painted a very rosy picture of OSS support and at the same time completely derided commercial software support. Its just a bit of a lopsided version of reality.


    |
    | BEA is going down. These people are getting desperate.
    |

    What is it with this attitude?

    If BEA/IBM dont bring their prices down - people say "they are too expensive".
    If BEA/IBM do bring their prices down and suddenly "they are getting desperate" "They are going down".

    Well, time will tell.

    The point that a lot of the anti-commercial software fanatics are making is that software is moving towards a service-oriented revenue model - where you pay for support rather than licenses.

    Given this model, what possible motivation is there for any vendor to make their product easy to use. If they make it really easy to use, no-one will buy their support - they wont need it. What sort of quality software will we be working with?

    Dont get me wrong, I am an Open Source software supporter - but I just dont get this "die bastard, die" attitude that some such as yourself have to vendors that produce good quality software and have the nerve to charge for it.

    -Nick
  33. For Matt Gunter[ Go to top ]

    |
    |price is greater than ZERO
    |

    (I wanted to post this seperately because I didnt want this confused with the post above - it has nothing to do with BEA or JBoss.)

    Ben,

    I really dont understand what problem people have with paying for quality software. A lot of the arguments I see ~for~ open source software start with this philosphy that to pay for software is stupid - and for vendors to ask for it is "stealing". I think it is kinda short-sighted and missing the point, in terms of what OSS projects should be concentrating on.

    For me, the primary question is the ~quality~ of the software. I have no problem at all with Open Source software - I use it all the time - I even contribute to some projects, but for me to choose it over commercial software, essentially it has to be becuause its a superior product for what I want to do, and it will save me time and trouble in the long run.

    The fixation on license cost, in particular, is one that gets my goat. In the scheme of the cost of a project, license costs are usually a very small percentage of the total cost of a project. Admittedly, there are some applications (esp CPU intensive apps) where the cost of per-cpu licensing is prohibitive. However, in many cases, its the development and production costs that are the REAL costs.

    More money, by far, gets burned by poor development than spent on server licenses. I would greatly appreciate it if more developers, project managers, and managers recognised this and focused on the big cost areas rather than the more "visible" cost areas. For me, this is the heart of the attitude problem: A purchase order requires a cheque and some kind of approval. To spend 6 weeks developing some crap software, of which 90% could have been downloaded for free, or paid for, seems to go largely unnoticed, and worse-still, its generally accepted.

    In some part, I see some of the anti-commercial software mantra as being "invest in people rather than licenses". While I agree with this on a certain level, there is a definite limitation to taking the "chinese approach" to problem solving (ie throwing more people at it). At the end of the day, if I can point a team of developers to a product, a website with some documentation, and they can get up and running with little fuss and no need for me to spend money on: a) product training or b) potentially expensive consultants, then I feel I am in front. If it also happens to be open source software, then great, but I dont make it my first priority.

    -Nick
  34. re:price is greater than ZERO[ Go to top ]

    Nick for president!
  35. For Matt Gunter[ Go to top ]

    | More money, by far, gets burned by poor development than
    | spent on server licenses.

    So rather than spending that money on a useless license that buys you jack, spend it on training your development team. You end up with less crappy development and in the long run saving a lot more money than what you spent on a piece of paper that restricts you from fully servicing your customers.

    | I see some of the anti-commercial software mantra as being
    | "invest in people rather than licenses". While I agree with
    | this on a certain level, there is a definite limitation to
    | taking the "chinese approach" to problem solving (ie
    | throwing more people at it).

    It is not about throwing more people at it, it is about making sure the ones you have know what the **** they're doing. Making sure they're writing quality software is the biggest cost saver. And I am more than happy to spend my license fees to train my developers.
  36. For Thomas Mattson[ Go to top ]

    | So rather than spending that money on a useless license
    | that buys you jack, spend it on training your development
    | team

    Apart from "useless license that buys you jack", I take your point. (Obviously, the license buys you a software product that saves you time - and hopefully results in a better quality solution.)

    Obviously having good people helps enormously to reduce cost. Whether you achieve this by training, or by hiring is another matter.

    I dont agree that spending money on training is always the answer. Reducing the amount you have to develop is usually a bigger and quicker win. Training can be quite expensive in fact (especially, when you consider the lost week of development). What is more, some of the skills you look for cant be acquired through training - it requires experience and it probably requires experience in a particular domain (and most training courses tend to be generic).

    I do take your point though, that money spent on improving your team is money well spent.

    -Nick
  37. you know[ Go to top ]

    I've said this many times, and I'll say it again: The JBoss folks would get a lot more converts if they'd just lose the patronizing attitude.

    I personally want nothing to do with JBoss because of the consistently arrogant and self-aggrandizing comments made by team members and representatives.

    Alternatively, BEA folks including Tyler, Cedric, and Joe tend to be very professional and helpful in all of their discourse.

    Technical competence won't win you sales, boys -- trust does.
    Take a cue from Linus.
  38. you know[ Go to top ]

    totally agreed.
    I mean I won't even download it because they may count it as a production release for their market share.
  39. Pramati ISV pack[ Go to top ]

    <vendor>
    You must check out the Pramati ISV-Pack.
    </vendor

    Common Ramesh, who in today's scenario wants to invest upfront in 50 and 100 license packs and keep inventory levels high or dead inventory.You know what is the cost of keeping floating inventories?Unless pramati helps reduce that by helping ISVs sell their products.I think that is a remote one..
  40. BEA,Small vendor[ Go to top ]

    <snip>In reality, BEA's technical qualifications are very below compared to a small-scale vendor such as "pramati". BEA's business vision is narrow-minded and there is not even a good candidate in BEA who knows what industry needed</snip>

    Mr/Ms. T Q : You totally forgot the otherside

    One can see,Some one who is small and say better technically qualified stinks or are hopeless in doing business. you can see that from the respective customer bases and developer forums.check out

    One needs to strike a balance!
  41. Pramati ISV pack[ Go to top ]

    <snip> Common Ramesh, who in today's scenario wants to invest upfront in 50 and 100 license packs and keep inventory levels high or dead inventory.
    </snip>
    You would be surprised! More so, even today there are companies going for site licenses. When someone needs bulk licenses, the value proposition from the volume packs is unbeatable. (Ofcourse, in todays market, noone would buy and stock them if they are not confident they can push them thru in a very short time).

    Cheers,
    Ramesh
  42. you know[ Go to top ]

    <snip>
    I've said this many times, and I'll say it again: The JBoss folks would get a lot more converts if they'd just lose the patronizing attitude.

    I personally want nothing to do with JBoss because of the consistently arrogant and self-aggrandizing comments made by team members and representatives.
    </snap>

    Exactly! Talking to them is like talking to asocials. You have to wear some protecting clothing to stay clean...

    But I'm sure they will leave to nowhere once Marc F has enough money in his pocket or sells his Ma' and Pa' company JBoss Group, respectively. We'll see.
  43. you know[ Go to top ]

    <stu>
    I've said this many times, and I'll say it again: The JBoss folks would get a lot more converts if they'd just lose the patronizing attitude.

    I personally want nothing to do with JBoss because of the consistently arrogant and self-aggrandizing comments made by team members and representatives.

    Alternatively, BEA folks including Tyler, Cedric, and Joe tend to be very professional and helpful in all of their discourse.

    Technical competence won't win you sales, boys -- trust does.
    Take a cue from Linus.
    </stu>

    Absolutely. I'd like to see the heat taken out of these discussions. It seems impossible now to say anything which is pro any other app server without been accused of some of the following:

    1) Being anti-JBoss
    2) Not being a "real" developer
    3) Being a stooge for the commercial vendors

    If we just share our experiences without insulting one another then we might just learn something.

    Ian.
  44. ian's comments...[ Go to top ]

    Ian - just a show of support for the points of view you are expressing in your posts... thanks for taking the time to articulate them so well, since they almost perfectly reflect the views of a certain other developer frequenting this board :).

    I just want to add - that while I hold both JBoss and Weblogic Server in high technical regard, Weblogic Platform is currently in a league of its own wrt the 3 pronged Server/Portal/Integration strategy. The 'leaked' news on Weblogic Platform 8.1 sounds dead on target with where BEA needs to be going - specifically an extended 'Workshop' that unifies portal and integration controls (and hopefully makes the Webflow diagram editor more respectable), finally JDK1.4 support and a greater focus on robustness and speed.

    In my current work situation we are like a poster child for their strategy... smaller, fast growing company, able to assemble only a small team of capable developers due to time and/or money constraints that has a hugely ambitious, longterm, standards based, service-oriented, workflow based, componentized and flexible IT infrastructure requirement. Not to mention the need to publish potentially dozens of portals from that infrastructure.

    Worries about vendor lock-in wrt Portal and business process engine can be dealt with later with the release of JSRs 159 (Java Process Components) and 94 (Portlets) - both of which BEA is a member of.

    To be sure - BEA has fallen down on several issues. Wrt the Portal and Integration servers... it's extremely annoying that they are not using BEA's own extensive rewrite of Server 7's security - but rather using the compatibility realms from Server 6. Read any Server 7 marketing brochure and you'll get an earfull about its new Security system - but BEA doesn't use it themselves??? Also JDK 1.4 support has taken an unacceptable amount of time - BEA has built a reputation for adopting new standards more quickly than their competitors, so what gives on this embarassing fumble? Parts of the EBCC application are useful, but extremely primitive and bugprone. Most amusing is the apparently O(N^N^N^N^N^N^N^N^N) search function in the webflow diagram editor. It's easy to get a Nullpointer exception in Weblogic Console... it needs to be more robust - especially wrt concurrency issues between the tree and the main web page view. Perhaps the EAR just clicked on in the java applet tree no longer exists? Portal "Developer's Guide" is the equivalent of a picture book for managers - very little useful technical information.

    What determines BEA's future? Can it convince enough other developers / IT departments on the merits of their Integration/Business process engine strategy? Do customers hold their Portal offering in high regards - for that matter does "management" even recognize the BEA name? Can it hold onto the grass-roots developer loyalty it created in the early J2EE days by not dropping the ball in the future wrt standards adoptance/speed/reliability? Do the Portlet and EJB component marketplaces mature enough to drive a general increase in demand for application servers? For me - showing management that our content management engine could also run as an EJB component alongside our own components helped bring them onboard on the whole app server concept. Bea must avoid the temptation to turn to the 'dark side' of relying on vendor lock in... there needs to be evidence that while Portal and Integration strategies are proprietary at the moment (for obvious reasons... there are no standards) going with BEA does not "lock you in" beyond any trivial level. In other words can BEA maintain an image of 'championing' the cause of the developer while avoiding that of an 'evil corporation' (something I consider them to have slid at during the last few years)? Can Bea establish a more attractive price point than their current one - and will that result in an increase in revenue? In my opinion they will eventually need to run on sales volume - and not be cherry picking the market to the degree they have been. Ironically its the "big companies" who BEA seems to be targetting that can support JBoss on their own, build their own workflow solutions, and probably even develop their own Portal frameworks.

    Bottom line?
    1) Difficult to imagine I would stand much chance of success without this product - or at least the measure of success would need to be considerably redefined.
    2) The product is not perfect and BEA can not afford to trip up in the future, because Larry Elison's LearJet is swooping in for the kill.
    3) Holding out hope that 8.1 will put to bed many of those previous 'trip ups'.

    cheers,
    Markus
  45. Very interesting comments[ Go to top ]

    Markus-

    I'm very interested to see your comments re: BEA Platform. I concur with most of your assessments and can identify with a number of the challenges that you indicate BEA must overcome.

    Technically, you are spot on about some of the technical alignment issues that exist with Platform 7.0. However, 8.1 is designed from the ground up to take these challenges head on and to provide a truly unified development, deployment and management platform.

    I can't say more now, but we'll have beta technology available very soon for the world to try. We will be having hands-on labs that cover most of this technology in depth at eWorld not to mention a couple dozen sessions covering different aspects of the platform (it's deep).

    But, just to give you a teaser: all development is done in Workshop including integration and portal, there is a clean security model that applies to all products and your concerns re: webflow development are tackled very nicely.

    Tyler Jewell
    Director, BEA
  46. Support[ Go to top ]

    <Ben>
    If someone wants to get real support they can pay for it, for a lot less than a WL license.
    </Ben>

    JBoss support costs $10,000 (source: http://www.jboss.org/services/consulting.jsp)

    I'll leave it up to the reader to decide whether it's "a lot less" than a WL license :-)

    --
    Cedric
  47. Support[ Go to top ]

    | JBoss support costs $10,000 (source:
    | http://www.jboss.org/services/consulting.jsp)
    |
    | I'll leave it up to the reader to decide whether it's "a lot
    | less" than a WL license :-)

    For any multi-CPU installation it sure is a hell of a lot less than the license costs of a BEA server.

    In fact, I can keep adding machines to my JBoss installation without having to fear of breaking any license agreements or the need to calculate whether I can afford to service my customers better with more hardware. I love that freedom. And so do my customers.
  48. Support[ Go to top ]

    Not having to go to budget meetings and having things approved: priceless.

    Beeing able to raise senior developers salaries, that's good too.

    .V
  49. JBOSS ,BEA[ Go to top ]

    Heard that JBoss ,Tomcats and JRuns already hold a dominant position as against a little known vendor in the India market ,with BEA's new offerings for low and medium market focus , will they exist in near future....begin count down?
  50. In reality, BEA's technical qualifications are very below compared to a small-scale vendor such as "pramati". BEA's business vision is narrow-minded and there is not even a good candidate in BEA who knows what industry needed.

       Examples to prove the above statement look into the technical details of award winning Portal framework from BEA. There was a lot of concern about how can we utilize the EBBC of WL portal in a corporate environment where multiple business specialists handling different sectors of a business portal which dealing with multi domain specific industry. When BEA was thinking about portal framework, they thought there would be only one business specialist who will update the portal look and feel and content according the business feedback from customers and users.
        What a fucking idea? How idiots there are to design a product with such a vision.
        
        Later I worked with a client who decided to use BEA portal in their company. My client have operation in 22 countries dealing with HR resource management, My client handling user driven information based on nationality, culture, language, regions etc. My client have special domain expert who know the business in each region they operate in this world. They wanted to bring the portal concept into their corporate world so that they can coordinate the business from multiple regions of the world from a single point of control.
        
        Management dream was to bring the corporate environment among business specialists and different departments into a common technical framework and make use of multiple business specialist’s work into a single portal application.

        Unfortunate to know that BEA selling Portal concept with a restriction that only one non-technical business specialist can update the portal application according the business feedback at a time. **** THEIR IDIOTIC IDEA AND WHO SUPPORT THEM. THEY DON’T KNOW HAVE ANY IDEA WHAT CUSTOMERS NEEDED IN CURRENT INDUSTRY. THEY HAVE TO GET EDUCATED IN BUSINESS IDEAS BEFORE TELLING OTHERS THAT THEY ARE THE LEADERS.
  51. For Ben Sabrin[ Go to top ]

    <ben>
    The [jboss] forums are FREE. If someone wants to get real support they can pay for it, for a lot less than a WL license.
    </ben>

    What great analytical insight! -just shortsighted...Typically, if that same someone factored in development, integration, downtime, and administrative costs they would happily choose the WL license.

    Ironically, they would then get real support ~AND~ A REAL PRODUCT (vs. a loss-leader for services)

    Matt
  52. For Ben Sabrin[ Go to top ]

    Matt -
    Curious as to what you mean by a loss-leader for services when refering to JBoss.

    Thanks
    Ray
  53. For Ben Sabrin[ Go to top ]

    Unfortunately, I'm going to have to disagree with you. We are moving from BEA to JBoss. Real support isn't really useful for the big items. Either a) it doesn't work or b) it will work in the next version. Niether of these answers is viable, especially if you don't have the source.

    JBoss is a real product. GE Supply just announced today that they migrated from BEA to JBoss. That's a big freakin' company.

    Steve
  54. bea in trouble[ Go to top ]

    Man, I guess JBoss is really killing you considering you have to drop your prices so much. Does this mean we will be seeing a revenue shortfall soon too at BEA?
  55. This is a welcome news !

    As a SME financial product vendor from Asia/India, we were really looking for these kind of Prices.We were using JBoss and JRUN since Pramati server is still v.expensive for what it had to offer.This is a good news from BEA.

    shirke
  56. It's About Choice[ Go to top ]

    I'm excited about the release of these versions of WebLogic. They give companies with a variety of budgets and IT departments a choice in how they leverage their infrastructure dollars.

    I don't necessarily think that the debate is JBoss vs. BEA. Every company has a risk and tolerance level for the products they choose to adopt. The reality is that there are pockets of companies and departments who have the flexibility to run their businesses on open source and some do not. These WebLogic price points (which include tiers of support) create options for companies to build a tighter relationship with BEA.

    Tyler Jewell
    Director, BEA
  57. It's About Choice[ Go to top ]

    Yep, these versions of WL seems excellent choice for those
    who dont want to use open source.

    good news, of course

    Maris
  58. Guys
       Comparing BEA product support with a big price companies have to pay for a servlet or ejb container is good idea for selling at the desk of some idiots managers in American cooperation. I would like you guys to read some of my comments about BEA's support on BEA products. After paying a huge amount for some garbage products + a good EJB container, what we got from New Jersey office is some misleading information which eventually lead us to the bugs in weblogic server and Portal applications.
       In my current project, 2.5 months ago, BEA send a senior WLS JAM specialist, who eventually made a support call to Houston office regarding a JAM installation problems. The lady on other side not even understands the requirements of setting up a startup class to enable the Gateway CRM communication.
       There is no reason or justice to add BEA support to justify the high price companies have to pay for BEA products.
       Wait and see, finally BEA will announce, WLS is free and no price any more.
       Understanding the less importance of middleware applications servers in market cuz of low economy, a large number of smart free vendors and tight competition from other major vendors, BEA decided to change color and names and bring the spirit of family or group of products under same name. Seems like nothing is clicking well and stock price is not going anywhere up. So this seems like a new game.
  59. Yes, welcome back Shaji[ Go to top ]

    So you had one bad experience with BEA support - and you used to work for them?
    Big deal.

    In contrast, we have had very good support on the whole.

    1) I have posted questions on their newsgroup, where most topics are monitored by their core developers for that subsystem (like JMS, EJB, CMP, SOAP), often received answers within a couple of hours. In some cases, there was already a patch - I raised a support call to get the patch and received the patch a couple of hours after that.

    2) I found a couple of bugs in the IDL generation. Again, I posted to the newsgroup late one night, had a patch jar in my inbox the next morning for me to try out.

    3) We had some serious concerns over C++ client integration via CORBA. One of their (founding) engineers read some of my posts and phoned me up to talk about it. They subsequantly gave us access to their pre-release CORBA C++ client.

    4) If we have had problems with software we were evaluating, we almost always ended up talking to one of their PS guys who actually used and knew the product.

    I am not saying that I have received this level of support ~every~ time - there have been a couple of times when the support was unsatisfactory - but, after a call to our account manager, it was sorted out.

    As for support cost, its important to remember that this support covers your version upgrades - its not just maintance. I know this is different to other licensed software we buy...

    In the end, the more pricing options IBM and BEA give us the better. It helps us with our case against the very cheap entry-level .net offerings.

    -Nick
  60. Yes, welcome back Shaji[ Go to top ]

    Seems like you are teaching business with people who worked in an out of this games.
    I am happy to know that your bugs got solved and you got the jar files next day in mail.

       I mentioned some minor cases to indicate that, it is not worth enough to justify the price of a product with its support.
       Bugs, bad coding standards, poor documentations etc are result of vendors who sell their ideas and products without proper vision.

       In 2001 first month, I was working in a project and had a question about commerce server, the case went to eastern region for support, that guy confused me for weeks finally I stopped following that case and found solutions myself.
       
       In 2002 end when I was working with another client, another questions unfortunately went to same guy, he tried to confuse me again and I redirected the question to Boulder office, I surprised to know that he is still working with your favorite vendor.
       There are number of times, I have to stop responding to their emails cuz the send me the URL to WLS, portal etc documentation in Internet.
       A high-level business question regarding the EBCC and multiple business analysts handling the business side for a corporate portal from a leading business firm (one of my previous client) was unanswered till I left the project.
       Once in Germany during a benchmark study for a famous banking firm, we were trying to contact Houston office for support, what we got, was not clear and satisfactory. Finally we could able to win the benchmark in favor of Weblogic application server.
       I prefer to give all the money to Google cuz more support and solutions are available in Google through its search features.
  61. Yes, welcome back Shaji[ Go to top ]

    | Seems like you are teaching business with people who worked
    | in an out of this games

    ¿Que?

    It seems that most of your problems have been with WLCommerce - which I cant comment on because I havent used it.

    We have recently looked at Portal 7, however, and it doesnt seem too bad (though, we are not interested in the WLCommerce aspect of it). I know it to be a whole lot more stable than 4.0 from people who are using it.

    In reality all vendors give shitty support at one time or other. I can tell you that a LOT is to do with how you deal with their support people. My advice is that the newsgroups are always the best first place to start (ie when you dont know what the problem is). I have worked in support, and its no fun dealing with people who cant explain what the problem is.

    On a different score (and I intend no offence) was the support contact in English? I suspect that English is not your first language - and I find what you write sometimes to be very difficilt to follow. I guess that wouldnt help if dealing with English support people.
    English, BTW, is a crappy language to learn - surpassed in complexity only by French IMO... ;-)

    -Nick

    >> your favorite vendor
    No. Not really. They just happen to be giving us what we want - if that stops, we look for a new vendor - thats the beauty of J2EE
  62. <snip>
    As a SME financial product vendor from Asia/India, we were really looking for these kind of Prices.We were using JBoss and JRUN since Pramati server is still v.expensive for what it had to offer.This is a good news from BEA.
    </snip>

    <vendor>
    You must check out the Pramati ISV-Pack.
    </vendor>

    Cheers,
    Ramesh
    - Pramati
  63. Hi PPL,

    IMHO SunONE Appserver 7 is a very good but underated app server and its free! The free edition is fully J2EE1.3 compliant. Granted this edition doesn't have clustering but for small shops with a load balancer (or none even) its a pretty good "buy" (Cuz its free!). Very Good docs, the management console is not bad (although further improvements can be made to the console), tight integration with SunONE 4 IDE (which is brilliant, abit sluggish sometimes but great all the same - but not free :-))

    Some orgs don't necessarily "endorse" the use of open source products so JBoss is not an option for them. IMHO I think the BEA announcement is a step in the right direction though. If your developers already have experience of BEA WL then a "cheap" WL is definately good news.

    But free is also good :-)

    Cheers

    Smythe
  64. Sounds Familar[ Go to top ]

    Sounds Familar
  65. 20 users ?[ Go to top ]

    Excuse my naive question but why would anyone want to buy WebLogic for a 20-user application ???

    In practice, 20 users will almost never justify the use of any EJB, so why WebLogic Server then ? For JMS ? JCA ? Web services ??? Isn't it rather a typical example of overengineering propaganda ?

    A simple servlet container should do just fine, ta very much.

                    Yann
  66. The jury is still out on BEAS[ Go to top ]

    Read the article here.