Revisiting Dogfood – Companies not using their own products

Discussions

News: Revisiting Dogfood – Companies not using their own products

  1. A long time ago, I blogged about Spring not eating its own dogfood, using Drupal for their website instead of a Java product, much less a Spring component somewhere in the mix. They're not the only high-profile Java site using non-Java – and it's telling that people didn't care. Spring had a lot of valid points in response to the original post. Mostly, they said that their focus was on improving Spring, not on creating a website with Spring, and using Drupal allowed them to have a flexible website with little work – and they felt that the irony of a Java site using PHP was not a fair comment on Java. I obviously felt otherwise. That said, even Dion Almaer, editor of TSS at the time, commented that the work involved might not be worth it:
    Do you think that TheServerSide.NET should be running on a .NET codebase in parallel with the Java one (which was the case in the past)? At some point you need to just get something done, and having to bend over backwards JUST to get something running on your technology doesn't make sense.
    Dion's dead right, and dead wrong at the same time, in my opinion. The key here is that image may or may not matter. To me, it does; a site serving as an evangelist for a given technology should itself use that technology in as many situations as can be made feasible. That means that my perfect TSS would use Java... for serving web requests, for managing the forums, for classifying data, for transforming input, for sending emails, for managing security, for using bookmarkable urls... for everything that is within scope, where the return on invested effort is reasonable. Even if the return on investment isn't 1:1 or better (meaning that it takes more effort than it should for a given result), I feel it's important to put forth the extra effort – and golly, folks, document it. If JavaMail is too much of a pain to use (which it probably is, for what most people want to use it for), then find or create an alternative (like, say, commons-email, which provides an excellent abstraction for most purposes) and document it, so the whole community improves and learns. However, I started off by saying that image may or may not matter. In a lot of cases, honestly, it doesn't. Consider: I complained about Spring not eating its own dog food, which was met with resistance; ever since then, I don't know of others sharing that opinion, or even of people actually being affected in any measurable quantity. (People saw it, shrugged, and moved on, a completely fair response.) I'd like to point out another site with a lot of visibility: Javalobby. You may notice that the look and feel of Javalobby (rebranded under the DZone network) has changed quite a bit; one thing you may not realise offhand is that DZone's CMS is itself running on Drupal. By saying it's about developers and not just Java – a thought TSS itself is echoing - the rebranding helps... but for long-time members like myself, it's a little sad. However, the key here is that until Rick mentioned the new look and feel in a newsletter, I didn't even notice. I read DZone stuff every now and then, of course - and I even post stuff to DZone occasionally - but I didn't realise that the underlying CMS platform was Drupal even after they'd switched over. (They do, however, use Ruby and Java for elements of their site – a commendable approach, and they're working even more on integration as time goes by.) The purist and evangelist in me rebelled at the thought. However, that impulse still ignores Dzone's primary mission – which is to communicate with its audience and provide a forum for its users. If Drupal enables that, then they're pragmatic enough about it to recognize that the effort in using a Java platform might not be something they care about. Based on their user statistics, I don't think that their readers – you and I, in a lot of cases – care either. So what does this mean? In one sense, it's disturbing: quoting a reader, it's a flaw with "the long tail of zealotry, everybody pimps their own dog food niche but they don't evangelize the bigger picture." (It should be noted that in the case of the Pet Store, this is a Good Thing. Come on, Sun, get rid of that dog.) With everyone focused only on their niche and not the bigger picture, the Java ecosystem ends up being fragmented out of the gate. At the same time, even after more than ten years, maybe the ecosystem still isn't strong enough to stand on its own. Why? It should be noted: at a TSS redesign meeting recently (yes, we do have them, and yes, TSS is scheduled for reimplementation and modernization at some point), it was suggested that TSS itself move to Drupal – a three week effort, ignoring data conversion, according to the project team. (Personally, guys, and no offense: riiiiiiiight. If you can do it, go for it.) What would TSS readers' opinion be of this change? (Do you think you'd notice?)

    Threaded Messages (76)

  2. Dogfood or steak?[ Go to top ]

    I think the best way to evangelize technology you support or are responsible for is to use it yourself. This really boils down to a moral dilema: Sure, the easy path is easier, but is it the right path? A slippery slope indeed. It's easy to serve up a plate of dressed up dogfood to someone else and tell them it's steak. It's quite a bit harder to sit down and enjoy the same meal with them.
  3. Re: Dogfood or steak?[ Go to top ]

    I think the best way to evangelize technology you support or are responsible for is to use it yourself. This really boils down to a moral dilema: Sure, the easy path is easier, but is it the right path? A slippery slope indeed.

    It's easy to serve up a plate of dressed up dogfood to someone else and tell them it's steak. It's quite a bit harder to sit down and enjoy the same meal with them.
    That's silly. Creating a framework and writing a java portal are two different items. By your logic, the engineers who design Ford 2 seater cab trucks, should only drive trucks. They shouldn't drive cars or ride motorcycles. I mean, if they don't drive trucks, who should? Too bad if they have a family of 4. Use the truck...you designed it...its easy to *say* that the truck is good, harder to prove it, right?
  4. Re: Dogfood or steak?[ Go to top ]

    I think the best way to evangelize technology you support or are responsible for is to use it yourself. This really boils down to a moral dilema: Sure, the easy path is easier, but is it the right path? A slippery slope indeed.

    It's easy to serve up a plate of dressed up dogfood to someone else and tell them it's steak. It's quite a bit harder to sit down and enjoy the same meal with them.


    That's silly. Creating a framework and writing a java portal are two different items. By your logic, the engineers who design Ford 2 seater cab trucks, should only drive trucks. They shouldn't drive cars or ride motorcycles. I mean, if they don't drive trucks, who should?

    Too bad if they have a family of 4. Use the truck...you designed it...its easy to *say* that the truck is good, harder to prove it, right?
    I think the closer analogy would be whether a guy who designs trucks for Ford drives a Ford vehicle... or a Toyota.
  5. Re: Dogfood or steak?[ Go to top ]

    I think the best way to evangelize technology you support or are responsible for is to use it yourself. This really boils down to a moral dilema: Sure, the easy path is easier, but is it the right path? A slippery slope indeed.

    It's easy to serve up a plate of dressed up dogfood to someone else and tell them it's steak. It's quite a bit harder to sit down and enjoy the same meal with them.


    That's silly. Creating a framework and writing a java portal are two different items. By your logic, the engineers who design Ford 2 seater cab trucks, should only drive trucks. They shouldn't drive cars or ride motorcycles. I mean, if they don't drive trucks, who should?

    Too bad if they have a family of 4. Use the truck...you designed it...its easy to *say* that the truck is good, harder to prove it, right?
    I think the closer analogy would be whether a guy who designs trucks for Ford drives a Ford vehicle... or a Toyota.
    I don't agree. Here's why - Spring didn't create a website that uses Guice. Spring is not equivalent to drupal but it is equivalent(more or less) to Guice. At least that's how I see it. The Ford truck guy may design a truck, but a truck isn't always the best vehicle for the job. He has to also drive a car. It should be a *Ford* car. Spring is like, what was it, 3M? "We don't make the webapp, we make the webapp better."
  6. Re: Dogfood or steak?[ Go to top ]

    It should be a *Ford* car.
    I feel sorry for whoever that guy is.
  7. Re: Dogfood or steak?[ Go to top ]

    It should be a *Ford* car.
    I feel sorry for whoever that guy is.
    Well, it'd be just desserts, that's for sure. I have a feeling the Ford guys *do*, in fact, drive Hondas or Toyotas.
  8. Re: Dogfood or steak?[ Go to top ]

    Joe -
    It should be a *Ford* car.
    I feel sorry for whoever that guy is.
    Well, it'd be just desserts, that's for sure. I have a feeling the Ford guys *do*, in fact, drive Hondas or Toyotas.
    Go to a Ford factory and check. Better yet, drive there in your Honda or Toyota, park in the parking lot, and then go count all the Hondas and Toyotas in the parking lot. By the time you get back to yours, they'll have "taken care" of it for you. (FWIW - It's a really really really bad idea to park a foreign car in the parking lot of a Ford factory.) Peace, Cameron Purdy Oracle Coherence: Data Grid for Java and .NET
  9. Re: Dogfood or steak?[ Go to top ]

    Well, it'd be just desserts, that's for sure. I have a feeling the Ford guys *do*, in fact, drive Hondas or Toyotas.


    Go to a Ford factory and check. Better yet, drive there in your Honda or Toyota, park in the parking lot, and then go count all the Hondas and Toyotas in the parking lot. By the time you get back to yours, they'll have "taken care" of it for you.

    (FWIW - It's a really really really bad idea to park a foreign car in the parking lot of a Ford factory.)
    Well, having been a Ford owner in the distant past, I'd rather they "took care" of my foreign car by making a product that makes me willing to buy American again, instead of by defacing where I found more reliability and value. Sure, I understand that they'd resent my parking in their lot with a competitor's vehicle, but hey... if my Granada hadn't kinda sucked, I'd probably still be driving a Ford, wouldn't I? Besides, Cameron, what are you doing on this thread? There are other sites, you know... ;)
  10. Re: Dogfood or steak?[ Go to top ]

    .. having been a Ford owner in the distant past, I'd rather they "took care" of my foreign car by making a product that makes me willing to buy American again, instead of by defacing where I found more reliability and value.
    Your suggestion isn't debatable; building a better product is always a better strategy. (Since my wife's Honda was made in the states and her old Ford came from Mexico, I don't know what it means exactly to "buy American again", but I think I get your drift ;-)
    Besides, Cameron, what are you doing on this thread? There are other sites, you know... ;)
    I'm just not feeling the love, Joe .. Peace, Cameron Purdy Oracle Coherence: Data Grid for Java and .NET
  11. Re: Dogfood or steak?[ Go to top ]

    It should be a *Ford* car.
    I feel sorry for whoever that guy is.
    Well, it'd be just desserts, that's for sure. I have a feeling the Ford guys *do*, in fact, drive Hondas or Toyotas.
    They don't get to park close to the plant though.
  12. Re: Dogfood or steak? Steak please.[ Go to top ]

    I never got this analogy of "eating your own dog food" anyway. Dog food companies should serve dog food for lunch?
  13. Re: Dogfood or steak?[ Go to top ]

    I don't agree. Here's why - Spring didn't create a website that uses Guice. Spring is not equivalent to drupal but it is equivalent(more or less) to Guice. At least that's how I see it.

    The Ford truck guy may design a truck, but a truck isn't always the best vehicle for the job. He has to also drive a car. It should be a *Ford* car.

    Spring is like, what was it, 3M? "We don't make the webapp, we make the webapp better."
    There seems to be an inordinate amount of categorizing here. Spring is a DI framework? Java is for building enterprise apps only? Are you all errant time travellers visiting from the distant past? The point is that Java has grown to the point where it can support more than just enterprise development. Spring and its various projects have grown beyond a simple DI container. This is old news. Putting aside the fact that there are applicable projects under the Spring umbrella nowadays that could have been applied to the problem of developing a community site, additionally there are plenty of frameworks, applications and tools written and supported by the Java community in general. I think the point is that it's appropriate to use technology drawn from the community you're involved with when building a site to address the needs of that community. Drupal is not the only contender out there for CMS/Portal.
  14. Re: Dogfood or steak?[ Go to top ]

    I don't agree. Here's why - Spring didn't create a website that uses Guice. Spring is not equivalent to drupal but it is equivalent(more or less) to Guice. At least that's how I see it.

    The Ford truck guy may design a truck, but a truck isn't always the best vehicle for the job. He has to also drive a car. It should be a *Ford* car.

    Spring is like, what was it, 3M? "We don't make the webapp, we make the webapp better."


    There seems to be an inordinate amount of categorizing here. Spring is a DI framework? Java is for building enterprise apps only? Are you all errant time travellers visiting from the distant past?

    The point is that Java has grown to the point where it can support more than just enterprise development. Spring and its various projects have grown beyond a simple DI container.

    This is old news.

    Putting aside the fact that there are applicable projects under the Spring umbrella nowadays that could have been applied to the problem of developing a community site, additionally there are plenty of frameworks, applications and tools written and supported by the Java community in general. I think the point is that it's appropriate to use technology drawn from the community you're involved with when building a site to address the needs of that community. Drupal is not the only contender out there for CMS/Portal.
    I'm not saying that Spring or Java can't be used to create a CMS. I just don't think the Spring guys need to write one *just* to host their site. It simply isn't necessary. We write Java apps. Are you using HotJava for a browser? Well, Java must suck for writing software.
  15. Come on you can't compare software to trucks and cars! And you shouldn't, on Graham’s Hierarchy of Disagreement http://blog.createdebate.com/2008/04/07/how-to-write-strong-arguments/ I put you at Contradiction. If you promote Java software, yes you should use Java to create your own site, because a site is software, and you can supposedely use Java to create sites Unless of course you think using Java to create a site is analogous to driving a truck in a formula one race! Java is general purpose it's not domain specific, trucks are domain specific, this is why your analogy doesn't work. And by making that previous statement I think I have risen to counterargument or even refutation on Graham’s Hierarchy of Disagreement
  16. Come on you can't compare software to trucks and cars!

    And you shouldn't, on Graham’s Hierarchy of Disagreement
    http://blog.createdebate.com/2008/04/07/how-to-write-strong-arguments/

    I put you at Contradiction.

    If you promote Java software, yes you should use Java to create your own site, because a site is software, and you can supposedely use Java to create sites

    Unless of course you think using Java to create a site is analogous to driving a truck in a formula one race!

    Java is general purpose it's not domain specific, trucks are domain specific, this is why your analogy doesn't work.

    And by making that previous statement I think I have risen to counterargument or even refutation on Graham’s Hierarchy of Disagreement
    You absolutely can compare java to a truck. Both are tools. Both are vehicles. Both were created by humans. Both consume resources. Both can be designed and promoted by the creators and STILL not be used for each and every task that confronts the creator. If you don't think Java is domain specific, try using it to cook or brush your teeth. Java's only possible use is to write software. Unless the website you displayed found a different use.
  17. Redhat, Tomcat or JBoss?[ Go to top ]

    The same thing occurred with Redhat and JBoss. Once the good ole' boys are gone, the name of the game is to sell JBoss the brand. When Redhat interviewed me for an architect position, the interview went south and I asked what position is this for? The answer was, "well, I'm not sure." That was comforting. The position was through a recruiter and Redhat was looking for contract developers to convert Tomcat apps into one's that utilize JBoss. It was a question of image. Developers out there be loyal to your technologies. If Spring is good for you, then convince the management to use it. Most of the managers are more concerned with bottom line than what jar dependency you need.
  18. Re: Dogfood or steak?[ Go to top ]

    The "SpringSource Enterprise Bundle Repository" at http://www.springsource.com/repository/app/ is Java based.
  19. I think it depends on the context. *If* your technology is touted as being good for public-facing web developement and you aren't using it yourself, then thats an issue. In the case of Spring, I think its sweet spot is medium to heavyweight complexity enterprise business apps, so its not unreasonable that they arent using it. But at the same time, if they ever explicitlty or implicitly market Spring as a "one size fits all" type application framework and then claim it would be too much work to use that technology to create their own site, that would be very telling. Has Spring made the claim either implicitly or explicitly that it is a one size fits all framework? Their current banner on their website reads "We deliver the world's leading Enterprise Java Platform" so at the current moment there seems to be truth in advertising.
  20. In the case of Spring, I think its sweet spot is medium to heavyweight complexity enterprise business apps
    I disagree. I think it's useful for all types of webapps. Spring is cool because you can use as little or as much of it as you want. But one feature I'll always want are its transactional unit tests. I'll NEVER do that by hand again. So, as soon as you want to write 1 test for a DAO, Spring is worth considering.
  21. Humor[ Go to top ]

    Here are the adds I found at the bottom of the discussion: Is this the kind of dog-food Spring should consume? Dog Food Comparison Tool Take the True Blue Test. See How Your Brand Compares. Learn More. xyz.com Dog Foods To Avoid Find Out Which Dog Foods May Potentially Kill Your Best Friend xyz.com Free Pet Food Samples Life's Abundance Dog & Cat Food Order Free Samples Today xyz.com Pure All-Natural Dog Food Free range meat & organic veggies. No grains, fillers or additives. xyz.com/Organic Premium Dog Foods Healthy, Holistic Nutrition Compare your dog food ingredients xyz.com
  22. In the case of Spring, I think its sweet spot is medium to heavyweight complexity enterprise business apps
    I disagree. I think it's useful for all types of webapps. Spring is cool because you can use as little or as much of it as you want. But one feature I'll always want are its transactional unit tests. I'll NEVER do that by hand again. So, as soon as you want to write 1 test for a DAO, Spring is worth considering.
    I have to respectfully disagree that Spring is hammer that easily fits all nails. I think alot of people who only do enterprise developement or sophisticated transactional websites have a hard time seeing how lighterweight technologies are applicable to less complex tasks. But really, it comes down to a matter of opinion, if Spring is useful to you for small projects you should use it. For me, if we are talking about a web app and Spring we are talking about Spring MVC. In my opinion this technology is overkill for a basic app that mostly just pushes data in and out of a database. Another thing to consider that while Spring is extremely elegant, it isnt "simple". Ive taken programmers with just basic Java exerience and had to get them up to speed on Spring, and it would make me realize that from a practical perspective Spring isnt the easiest thing to learn depending on the context. One has to udnerstand alot of concepts to make a meaningful contribution to a project. Don't get me wrong, I think Spring is a great tool and was clearly needed to within the enterprise Java community. If only because it made unit testing and AOP a reality compared to what was possible with J2EE. That said, if someone came to me and wanted a web app prototyped in a few days and a finished product not long after that, Spring MVC wouldnt be my first intuitive choice.
  23. In the case of Spring, I think its sweet spot is medium to heavyweight complexity enterprise business apps
    I disagree. I think it's useful for all types of webapps. Spring is cool because you can use as little or as much of it as you want. But one feature I'll always want are its transactional unit tests. I'll NEVER do that by hand again. So, as soon as you want to write 1 test for a DAO, Spring is worth considering.
    Agreed, I used Spring for some very simple apps. For one, which took 8 days, 6 of that was spent detangling atrocious SQL in the app. Yet the simple app had sophisticated caching, security, and transaction support thanks to Spring. It scaled down nicely.
  24. Interesting issues and questions in your post, Joe. I can only speak for the DZone/Javalobby part of it, but in some ways we profoundly f**ked the whole thing up and are still working to identify the recipe. A lot of people think it's so simple, and they make blithe comments about how you just , but in reality these transitions are much harder than most of the commenters realize. Whoever estimates three weeks for a Drupal transition is hitting the crack pipe too hard and may need an intervention. Anyway, all of us today work in a field where multiple technologies intersect and interoperate. There's no great purpose served by blind adherence to any particular technology as a matter of principal or allegiance. Certainly it may be the case that some would avoid specific technologies on principal. Unless one's worldview is exceptionally narrow, however, there are still likely to be multiple technologies available to solve any given problem. IMO, it pays to evaluate the options, even if you end up selecting the technology where you feel strongest. Good luck with wherever this goes. Rick
  25. Spring is in the business of offering services around a world-class framework. They are *NOT* in the business of creating little website portals. Drupal is in the business of offering a world-class CMS/Portal. Do you really expect Spring to invest its own internal resources on creating a portal from scratch? Its not a trivial task to design and code. Furthermore, looking around the CMS market, you will find few better than Drupal (and maybe some other PHP CMS) that offer such a deep feature-set , reliability, and community support/extensions. As a business, you learn to use the best tool for the job and focus on your core competency. In contrast, at TSS you *ARE* in the business of offering a website portal for Java developers. I would expect that you do have some talent in-house that can execute on this, using Java. Roy Russo http://www.loopfuse.com (OMG! Our site is written in PHP, but our app is Java!) ;-)
  26. In contrast, at TSS you *ARE* in the business of offering a website portal for Java developers. I would expect that you do have some talent in-house that can execute on this, using Java.
    Actually, no - our mission is providing an enterprise development community. Our expertise in-house is on implementing media-oriented content management systems, so technically, our in-house talent is not in this area.
  27. Actually, no - our mission is providing an enterprise development community. Our expertise in-house is on implementing media-oriented content management systems, so technically, our in-house talent is not in this area.
    Ok so "Your Enterprise Java Community" tells me you run a portal with Java "stuff" on it. Fittingly, your expertise in-house should be geared to managing a PORTAL with JAVA STUFF on it. I think we're saying the same thing, you're just choosing to use fancy-editor-speak. ;-) Roy Russo http://www.loopfuse.com
  28. Actually, no - our mission is providing an enterprise development community. Our expertise in-house is on implementing media-oriented content management systems, so technically, our in-house talent is not in this area.
    Ok so "Your Enterprise Java Community" tells me you run a portal with Java "stuff" on it.

    Fittingly, your expertise in-house should be geared to managing a PORTAL with JAVA STUFF on it. I think we're saying the same thing, you're just choosing to use fancy-editor-speak. ;-)
    Nonsense. The "java stuff" there is irrelevant to our in-house staff- it's just (hopefully) relevant to the readers.
  29. Nonsense. The "java stuff" there is irrelevant to our in-house staff- it's just (hopefully) relevant to the readers.
    Joe, an "Enterprise Java Community" portal, running on PHP/Perl/RoR would look mighty stupid. I think we all agree on that, as the perception by your readers alone should dictate that you strive to remove technology considerations that are not Java. Its all about street cred. ;-) To Dion's original point, Yes your .NET site SHOULD run on .NET. Its a freaking .NET community! To JavaLobby, Yes a site called JAVAlobby running on PHP is boneheaded. Just like MSDN runs on? kthnx. Regards, Captain Obvious
  30. Nonsense. The "java stuff" there is irrelevant to our in-house staff- it's just (hopefully) relevant to the readers.


    Joe, an "Enterprise Java Community" portal, running on PHP/Perl/RoR would look mighty stupid. I think we all agree on that, as the perception by your readers alone should dictate that you strive to remove technology considerations that are not Java. Its all about street cred. ;-)

    To Dion's original point, Yes your .NET site SHOULD run on .NET. Its a freaking .NET community!

    To JavaLobby, Yes a site called JAVAlobby running on PHP is boneheaded.

    Just like MSDN runs on? kthnx.

    Regards,
    Captain Obvious
    MSDN is backed my MS and they are, for practical purposes, of limitless resources. Should a Javalobby run on EJBs simply because that's what they report on? Now *that's* boneheaded!! Who cares what the portal site runs on? Who cares if the portal site uses DI? I'd rather them spend time(in the case of Spring) making Spring better and have sample apps using AOP than spending time trying to figure out how to shoehorn AOP into something of which it isn't suited.
  31. MSDN is backed my MS and they are, for practical purposes, of limitless resources. Should a Javalobby run on EJBs simply because that's what they report on? Now *that's* boneheaded!!

    Who cares what the portal site runs on? Who cares if the portal site uses DI? I'd rather them spend time(in the case of Spring) making Spring better and have sample apps using AOP than spending time trying to figure out how to shoehorn AOP into something of which it isn't suited.
    I'm lost here - JavaLobby and EJBs? Where does an EJB feature here?
  32. MSDN is backed my MS and they are, for practical purposes, of limitless resources. Should a Javalobby run on EJBs simply because that's what they report on? Now *that's* boneheaded!!

    Who cares what the portal site runs on? Who cares if the portal site uses DI? I'd rather them spend time(in the case of Spring) making Spring better and have sample apps using AOP than spending time trying to figure out how to shoehorn AOP into something of which it isn't suited.
    I'm lost here - JavaLobby and EJBs? Where does an EJB feature here?
    At some point Javalobby listed articles about EJBs. Should they then be forced to run the Javalobby site on technology featuring EJBs to be considered a legitimate Java site? I don't think they do. By the same token, I don't think the Spring guys need to write a CMS tool to legitimize their technology. I think the Google guys can use Thunderbird instead of Gmail and still have Gmail be considered good technology.
  33. By the same token, I don't think the Spring guys need to write a CMS tool to legitimize their technology. I think the Google guys can use Thunderbird instead of Gmail and still have Gmail be considered good technology.
    The difference here is that Spring's core competency is NOT in building portal/community sites. JavaLobby and TSS DO offer portal/community sites for JAVA stuff.
  34. The difference here is that Spring's core competency is NOT in building portal/community sites.

    JavaLobby and TSS DO offer portal/community sites for JAVA stuff.
    JavaLobby and TSS don't offer portal/community site SOFTWARE for Java stuff, they offer INFORMATION. I don't see that it matters what they publish on. If there isn't a free Java publishing solution which matches a free non-Java solution for ease-of-setup and features, why shouldn't they be able to use a non-Java solution to publish? It's not that relevant to the market they're addressing which is enterprise. Typically this market is engaged in in-house systems development and rarely in choosing which free publishing tool to use.
  35. At some point Javalobby listed articles about EJBs. Should they then be forced to run the Javalobby site on technology featuring EJBs to be considered a legitimate Java site? I don't think they do.

    By the same token, I don't think the Spring guys need to write a CMS tool to legitimize their technology. I think the Google guys can use Thunderbird instead of Gmail and still have Gmail be considered good technology.
    I don't think the "articles about EJBs" implies that they (or we) should use EJBs. For one thing, I'm actually kinda neutral here - both sides have merit. For another, I think there's an issue with the overarching technology not being used, not whether EVERY POSSIBLE PERMUTATION is applied - although I also think there'd be value in a shootout that applied as much technology as it could, where appropriate.
  36. Nonsense. The "java stuff" there is irrelevant to our in-house staff- it's just (hopefully) relevant to the readers.


    Joe, an "Enterprise Java Community" portal, running on PHP/Perl/RoR would look mighty stupid. I think we all agree on that, as the perception by your readers alone should dictate that you strive to remove technology considerations that are not Java. Its all about street cred. ;-)

    To Dion's original point, Yes your .NET site SHOULD run on .NET. Its a freaking .NET community!

    To JavaLobby, Yes a site called JAVAlobby running on PHP is boneheaded.

    Just like MSDN runs on? kthnx.

    Regards,
    Captain Obvious
    So you think the guys who run the site on Assembler should have to write it using Assembler ? What about Java Swing sites ? Should they even have web pages or should the build their sites as downloadable Swing apps ?
  37. So you think the guys who run the site on Assembler should have to write it using Assembler ?

    What about Java Swing sites ? Should they even have web pages or should the build their sites as downloadable Swing apps ?
    Yes, and people who build Ford cars, should write their websites using car parts.
  38. Yes, and people who build Ford cars, should write their websites using car parts.
    Haha... ma-an, you've made my day. Thanks, Nikita Ivanov. GridGain - Grid Computing Made Simple
  39. is there not a Java based CMS/Portal they could use? Or are we saying Java isn't up the task of building a CMS/Portal system (funny in a past job I was going to use Spring to do that myself. I'm glad I never did since I guess it's not good enough)
  40. It never ceases to amaze me how many people can't grasp the concept that PHP or Java or Spring or whatever is not the solution to every problem and there's absolutely nothing wrong with that. This is a development community and we should be smart enough to understand that the tools/technologies used to build an enterprise banking application are not necessarily the tools to build a content-only web portal.
  41. A long, long time ago...[ Go to top ]

    ...there was a fine book about web development, that distinguished between "sites that really are programs" and "sites that really are pages". I find the idea quite bizarre that, because I develop a tool to do dependency injection and that supports a bit of "web flow" - sites that are programs - I would be expected to build a web site - that actually is a site -with these very tools, instead using, say, Plain HTML, which is probably better suited to what I am trying to achieve.
  42. I think overall this is very valid point. In most cases (where it is even remotely applicable) companies should use their own products. I’m not sure I agree w/Spring example here… Web framework is largely irrelevant right now in their line of services and has been for quite some time now. As an example, we in GridGain ship one in a kind grid-enabling of JUnits… and we use it to run our own JUnit suits on daily basis (on for every commit, actually). We found a whole slew of instabilities early on in development and it allowed us to release very stable version of that code. More over, it shrunk the JUnit build from 1 hour to ~10 minutes and it gave us a good visibility in how this integration is going to be used by the end user – the developer. Best, Nikita Ivanov. GridGain – Grid Computing Made Simple
  43. Crack me up...[ Go to top ]

    Here's the news. The folks at TSS, DZone, etc. are in the PUBLISHING business. They publish and distribute content. Their goal is to make that process as efficient, pain free, and profitable as possible. No where is there any requirement that they use their publishing mechanisms as a platform to gain expertise in the field that they are publishing in. Why should TSS, DZone, or anyone else limit their tool choices for a task this unrelated to the field? Who cares? The Spring guys get their expertise from developing their software, and through using that software in the field via consulting and training. That's where their real world experience and knowledge comes from. Why should they waste their time writing a CMS if an off the shelf system works just peachy for them? Are you going to condemn them for using MS Office instead of "hand crafting" some presentation software for their training sessions? Are you MAD?? You don't look to Spring for their CMS and website building skills. You look to them for their expertise in the enterprise space. The value that TSS and DZone etc. is in their editorial policies, their network of authors, and the community that flutters around the lights they lit. As long as the site comes up promptly when I hit the bookmark, the details are unimportant to me. On the Internet, no one has to know you're using PHP. Would they be able to contribute perhaps a bit more to the community if they were written in Java? Maybe, for a little bit. It would be as much commentary on their back end design skills as much as on the implementation language as anything, and it would be a short term gain. Once the site is stable, the story stops being interesting save as a "look they've been running java for XXX bleems of time!". Man, slow news cycle or something.
  44. If you want to make a web based app you'd have to be forced to use Java. No intellegent person or collection of people would use Java EE for any part of a web application unless they wanted to waste time and money. It is not surprising that these morons who continue inventing the endless collection of vaporware that Java EE has degenerated into, would never dream of using it themselves. What is also not surprising is that they are so arrogant as to make excuses, and believe them, for why they use PHP or ASP, or anything but Java EE. No really big players use Java EE either. The question is, when will this abortion that Java EE has become end?
  45. If you want to make a web based app you'd have to be forced to use Java. No intellegent person or collection of people would use Java EE for any part of a web application unless they wanted to waste time and money. [..] No really big players use Java EE either.
    What do you define as a "big player"? Just to name a few well-known companies that I've visited with in the past year that use Java EE technologies extensively for their web sites: * Amazon * eBay * Google * Yahoo! * AOL * Sony * AT&T * BT * Salesforce.com * Hotwire * Orbitz * Expedia Almost every major bank and many "etailers" use Java for their web sites. It doesn't mean that everything has to be Java -- look at Google and Yahoo! as examples of companies that successfully use many different languages, frameworks and tools. Anyhow, your claims are as factually baseless as your religious debate is nonsensical. Peace, Cameron Purdy Oracle Coherence: Data Grid for Java and .NET
  46. Not only the below mentioned companies but almost every lage enterprise uses Java in one way or the other .OSS or commercial is immaterial . We tried memcached but later on shifter to coherence . I don't care even if its written in assembly language . The debate is not worth the time . Look at alfresco completly written in Spring framework and is a great CMS tool but does that mean people stop using drupal ? . These products have their space . its like saying oracle as an organization can't use my sql in any of their developments .
  47. They use Java EE sparingly..[ Go to top ]

    Cameron, It is my understanding that none of the companies you mentioned use Java for more than a tiny percentage of their web traffic. Probably to hedge their bets. And it is not religion, it is just that I noticed a long time ago that Java EE is nearly impossible to get anything done in.. And the end product may cost millions to develop, but it sure is slow.. It just seems to be an exercise in complicating something simple.. I'm also upset that I have to go around coding in Javascript, PHP, Perl, and Python and take a big paycut because I can no longer stomach the insanity of a 1000 line stack trace in JBoss and 3 days to change a date on a form.. That's probably why everyone uses something else when its their time and money.. Steve
  48. Re: They use Java EE sparingly..[ Go to top ]

    Cameron,

    It is my understanding that none of the companies you mentioned use Java for more than a tiny percentage of their web traffic. Probably to hedge their bets. And it is not religion, it is just that I noticed a long time ago that Java EE is nearly impossible to get anything done in.. And the end product may cost millions to develop, but it sure is slow.. It just seems to be an exercise in complicating something simple.. I'm also upset that I have to go around coding in Javascript, PHP, Perl, and Python and take a big paycut because I can no longer stomach the insanity of a 1000 line stack trace in JBoss and 3 days to change a date on a form.. That's probably why everyone uses something else when its their time and money..

    Steve
    I've worked on large high volume websites in the past. I can say from my experience, being productive and producing good code using J2EE is a matter of skill. In my case, it was JSTL, servlet container, a few custom tags and custom caching solution. The website handles millions of page views each day without any problems. We had features like multi-lingual support, and cobrands. Seriously, if it takes 3 days to change a date on a form, something is wrong. Don't blame J2EE for user error. My bias 2 cents. peter
  49. Cameron,

    It is my understanding that none of the companies you mentioned use Java for more than a tiny percentage of their web traffic.
    Each company is different. I can't disclose any internal information about customers' systems, but eBay (for example) has done many public presentations explaining how their overall infrastructure is set up, and how their applications work. It is largely Java EE for the web content and for many back end systems, but that doesn't mean that they do stupid stuff (i.e. they don't build eBay like Sun built PetStore). Same goes for Amazon. It used to be a C++ web app ("obidos" or something like that), but they moved it to Java. For Google, the last time I looked, over 98% of their revenue comes from a Java application (AdWords). The list goes on ..
    .. it is just that I noticed a long time ago that Java EE is nearly impossible to get anything done in.. And the end product may cost millions to develop, but it sure is slow.. It just seems to be an exercise in complicating something simple..
    Java EE, used well, isn't complicated. It's when people use crap from Java EE when it's not called for (e.g. the old "EJBs for everything" approach) that makes things complicated. Remember, if you're using Servlets or JSPs, you're building on Java EE. You don't need to be using J2CA or EJBs .. Peace, Cameron Purdy Oracle Coherence: Data Grid for Java and .NET
  50. laughable....[ Go to top ]

    Cameron,

    It is my understanding that none of the companies you mentioned use Java for more than a tiny percentage of their web traffic. Probably to hedge their bets. And it is not religion, it is just that I noticed a long time ago that Java EE is nearly impossible to get anything done in.. And the end product may cost millions to develop, but it sure is slow.. It just seems to be an exercise in complicating something simple.. I'm also upset that I have to go around coding in Javascript, PHP, Perl, and Python and take a big paycut because I can no longer stomach the insanity of a 1000 line stack trace in JBoss and 3 days to change a date on a form.. That's probably why everyone uses something else when its their time and money..

    Steve
    It can take a long time to get relatively simple things done in straight J2EE if you don't understand when/how to use it, don't have a good IDE, and or don't utilize frameworks built on top of J2EE to make web site development easier. I wonder how much you really understand about enterprise concerns. If you did you'd find that all these "simple" things tend to add up to quite a bit of complexitiy. Even choosing "simpler" frameworks or tools will not get you away from the complexity that real business demands can drive. You seem to be stuck with a relatively narrow point of view on what people are doing with computers. There is more to IT than web sites. Trying doing real backend enterprise level integration with PHP or something else "cool" that you can do simple web sites in. Let me know how poor your tool of choice is when you realize that no tool is the best for every job. Travis Retzlaff
  51. Re: laughable....[ Go to top ]

    It can take a long time to get relatively simple things done in straight J2EE if you don't understand when/how to use it, don't have a good IDE, and or don't utilize frameworks built on top of J2EE to make web site development easier.
    Not to speak for Cameron but I think you guys may be talking past each other here to some degree. You mention "J2EE" and presumably mean the EJB/Servlet technologies of that version of the spec. If you take that narrow definition then yes you are absolutely right, those are incomplete and often inapropriate tools depending on the task. If you expand the conversation to JEE 5, then it becomes less black and white. The move towards annotations and POJOs has gone a long way towards simplifying the technology stack. If you expand the scope past the official specs and include widespread technologies like Sping, Hibernate and Seam then I think your statement is less true. There still are spaces where the Java ecosystem doesnt fully deliver a solution for, but the situation is alot different than was just a few years ago.
  52. Re: laughable....[ Go to top ]

    It can take a long time to get relatively simple things done in straight J2EE if you don't understand when/how to use it, don't have a good IDE, and or don't utilize frameworks built on top of J2EE to make web site development easier.


    Not to speak for Cameron but I think you guys may be talking past each other here to some degree. You mention "J2EE" and presumably mean the EJB/Servlet technologies of that version of the spec. If you take that narrow definition then yes you are absolutely right, those are incomplete and often inapropriate tools depending on the task. If you expand the conversation to JEE 5, then it becomes less black and white. The move towards annotations and POJOs has gone a long way towards simplifying the technology stack. If you expand the scope past the official specs and include widespread technologies like Sping, Hibernate and Seam then I think your statement is less true.

    There still are spaces where the Java ecosystem doesnt fully deliver a solution for, but the situation is alot different than was just a few years ago.
    Very true. I was referring to the narrower definition of the official specs, and you are correct enhancements in JEE 5 do simplify the stack. Thanks for your polite response to my perhaps somewhat off topic vitrolic post :)
  53. "pure" j2ee[ Go to top ]

    I was referring to the narrower definition of the official specs, and you are correct enhancements in JEE 5 do simplify the stack.
    There was a trend, early on, to be "pure" J2EE, which meant "the only things I can write are those types of components spec'd by the J2EE spec." The result was disaster. For example, EJBs weren't really _that_ bad for encapsulating some transactional stuff (i.e. they were only 5x harder than they should have been), but if they are the only tool you are allowed to use, then almost everything will seem hugely difficult and very painful. I think somewhere along the line, we just took the whole "spec" thing a bit too seriously with J2EE. Peace, Cameron Purdy Oracle Coherence: Data Grid for Java and .NET
  54. Steve, JEE (in various component incarnations) is responsible for billions of dollars every year and is utilized by almost all major online businesses on the planet (Cameron mentioned Top 10 – there are hundreds and thousands more). JEE requires certain skill and had its own crop of problems. Nonetheless, this is a de-facto standard platform for enterprise development (including Web-related systems) and with .NET they both cover >95% of all enterprise software development in the last 5 years in my estimation. Saying that JEE is irrelevant is like pissing against the wind… Best, Nikita Ivanov. GridGain – Grid Computing Made Simple
  55. 95% in Java EE and .NET[ Go to top ]

    Nikita, I do also think that Java EE is the way to go for enterprise applications, but I think that Java EE and .NET are far away from 95%. If you talk to banks and insurances you would see that COBOL and PL/1 still cover a lot of enterprise application development on mainframe applications (zSeries). And C and C++ is established in the embedded market like in the automotive or aerospace area. I think that Java EE is THE integration platform for enterprise systems. If you would ask me for a coverage rate, I would say that 20-30% for Java in the enterprise software development area would be a fair number... Mirko Mirko
  56. DesiredNewFeatures.java[ Go to top ]

    If you want to make a web based app you'd have to be forced to use Java. No intellegent person or collection of people would use Java EE for any part of a web application unless they wanted to waste time and money. It is not surprising that these morons who continue inventing the endless collection of vaporware that Java EE has degenerated into, would never dream of using it themselves. What is also not surprising is that they are so arrogant as to make excuses, and believe them, for why they use PHP or ASP, or anything but Java EE. No really big players use Java EE either. The question is, when will this abortion that Java EE has become end?
    +1. I still think J2EE can still be bettered- that is if Sun and co. come off the dream of 'Enterprise Java','Java in the Enterprise' when the lift off has not yet taken place. They need to listen to developers, preempt I should say.This where MS shines.The name of the game of 'PRODUCTIVITY'.You prove nothing by using complex stuff to build more complex stuff. I think we need a laid-back, relaxed kind of Java.The current J2SE and J2EE specs are simply over-engineered and way too complex.The complex machinery involved in getting a simple one page functional site attests to that fact. I think they should bake RAD right into the JVM for JEE to compete in today's market.Java the language needs to be reworked from the ground up-closures is a must,currying, a reworked generics implementation,.Net style delegates or anything that can act as a function pointer,and most importantly functional programming features are what Java needs to transition to the next level.It must be neat and easy to use( and smell good too).Borrow if they have to and give developers a bang language and platform for building next-gen apps. Till then...
  57. Elephant in the room?[ Go to top ]

    I'm going to commit the ultimate OSS sin and mention money. Imagine you own a busines, and you have to cough up for payroll each month. If one of your teams tells you that something non-core (i.e. your web site, not your prodct) can be done in one month using technology A, then are you really going to spend your own money, doubling the cost, just because you're feeling a little bit religious about technology B? I'd give Magnolia, or some other JCR based system Java system a shot - because everyone benefits if the Java ecosystem extends it's borders, but I wouldn't suffer for the cause. M.
  58. Re: Elephant in the room?[ Go to top ]

    I'd give Magnolia, or some other JCR based system Java system a shot - because everyone benefits if the Java ecosystem extends it's borders, but I wouldn't suffer for the cause
    I totally second that. Also I would say that companis like magnolia, alfresco, nuxeo and dotCMS should compete to setup spring's website for free - just to prove their point. Guys, convince the spring team (is it Colin Sampaleanu that takes care of the site?) and I will be convinced, too. Regards, Raffaele
  59. As we have read here recently: "Java is losing the battle for the modern web" ...
  60. Java VS PHP[ Go to top ]

    We do lot of dev in JAVA and some work on PHP , our site runs on typo3 , wordpress , mantis (bug tracker) .i think PHP folks are more practical like linux guys where they want a system to work and be usable and Java folks are freebsd guys more therotical but useless for vast majority of users , now installing wordpress is damn easy , takes less memory (even Sun Jonathan ' blog uses wordpress) with more that 2 thousand plugins !!!! not like so called EJB or java monster which is JSR x, y, complaint with with 20 plugins which require 1 gb of dedicated ram etc , with 50X more expensive hosting than php I guess java folks should concentrate on building piratical application more than those stupid x,yz JSR Maybe building Rich application with Dwr , GWT , Echo3 with light weight container like jetty and great apache commons, even new frameworks GridGain,Grails are great . I do hope java something equivalent of PHP pear repository or extensible java Lets Keep the Bullshit out Build Stuff which we can actually use !!
  61. Re: Java VS PHP[ Go to top ]

    check out Coldtags suite for example: http://www.servletsuite.com/jsp.htm
  62. If JavaMail is too much of a pain to use (which it probably is, for what most people want to use it for), then find or create an alternative (like, say, commons-email, which provides an excellent abstraction for most purposes) and document it, so the whole community improves and learns.
    If Oracle, as a content backend, is too much of a pain to use then would you find or create an alternative and document it, so the whole community improves and learns ? Reinventing the wheel ? I think that the image could matter for the key issues, but even in this case I have still doubts. Would you expect that all the database interface of Hibernate web site being implemented with Hibernate itself ? All ? Forum, content management, mailing list, everything ? Guido
  63. I don't think anyone is saying code your own CMS or Portal in Java.. But if they are a company shilling Java products would it hurt them to use a Java based CMS/Portal?? Especially Spring.. Maybe one of the many CMS systems that use Spring under the covers might make sense.. I understand that the Spring folks think EJB is no good but apparently they feel the same way about Spring based CMS systems.. Remember perception is way more important than fact.. (and that's a fact! :-) )
  64. I don't think anyone is saying code your own CMS or Portal in Java.. But if they are a company shilling Java products would it hurt them to use a Java based CMS/Portal?? Especially Spring.. Maybe one of the many CMS systems that use Spring under the covers might make sense.. I understand that the Spring folks think EJB is no good but apparently they feel the same way about Spring based CMS systems.. Remember perception is way more important than fact.. (and that's a fact! :-) )
    And Spring has sure been damaged by not running its site on a Spring based tool. Perception is way more important than fact *to you*. Others, I guess, prefer to make their choices based on fact.
  65. Others, I guess, prefer to make their choices based on fact.
    I'm not sure what color the sky is in your perfect world but it's sure not blue.. The vast majority of people basing things on facts leaves far too many unexplained opinions in the world. Most people think they know the facts when they make an argument but most of them really don't. For instance: We know the Spring site is not Java/Spring based.. But we don't know why. Some of us believe we know why (you) others are just guessing (me). Either way it's our perception of the facts that is driving our opinion of the matter. Honestly I never really noticed what they (Spring) used for the web site. I guess I know now but I still don't really care.
  66. Others, I guess, prefer to make their choices based on fact.


    I'm not sure what color the sky is in your perfect world but it's sure not blue.. The vast majority of people basing things on facts leaves far too many unexplained opinions in the world. Most people think they know the facts when they make an argument but most of them really don't. For instance: We know the Spring site is not Java/Spring based.. But we don't know why. Some of us believe we know why (you) others are just guessing (me). Either way it's our perception of the facts that is driving our opinion of the matter.

    Honestly I never really noticed what they (Spring) used for the web site. I guess I know now but I still don't really care.
    I just looked outside. The sky is blue. The issue is that most of us, I suspect, who use Spring don't care what their site runs on. I've gotten years of use out of Spring despite the fact that I never gave anything considering to whether or not their site uses Spring, Java, or anything else. The question, to me, is simple: Does it solve my problem?
  67. The issue is that most of us, I suspect, who use Spring don't care what their site runs on.
    I agree.. As I said, I really don't care what they use. But people on this thread were answering the question with "why should they build a CMS/Portal in Spring if they are framework people and not CMS people." I agree they shouldn't. But they didn't pick a CMS based on Spring. So if you are reading this thread it might make you wonder: Is Drupal better then a Spring based CMS like Alfresco? If so why? Is it the underlying stack? or did the Drupal guys just kick butt and make a better CMS despite the fact the Alfresco guys had a superior stack. Basically I was trying to push us back to a better discussion than the "they're not CMS experts" stuff that was flying around. Because if the answer is that Drupal is better because they had a better stack to build upon we might want to discuss that and figure out how to fix it.. Personally and this is just my opinion, but I see CMS/ECM and Portals to be the type of enterprise apps that Spring/Java should do well with..
  68. I think that Spring's dogfood is an underlying enabling technology, not a community portal. If they had a community portal product in their offerings and didn't use it, I would worry. But they don't. Being worried about what community tool the Spring team uses is like being worried about Sun employees using any program that isn't build in Java. It's unrealistic and completely pointless.
  69. I think that Spring's dogfood is an underlying enabling technology, not a community portal. If they had a community portal product in their offerings and didn't use it, I would worry. But they don't. Being worried about what community tool the Spring team uses is like being worried about Sun employees using any program that isn't build in Java. It's unrealistic and completely pointless.
    ... although it's worth noting a few things here: Sun also has C compilers and UNIX environments in their product lineup, so them not using Java hardly implies that they're not using their own products. Even there, it's worth pointing out how often Sun does use Java for system administration. Of course, we've left some of the original surface material far, far behind. I don't think Spring doesn't use Java - I know they do. My concern - yes, concern, not complaint - is that it's not public-facing, and from a technology-evangelism standpoint, that's not such a positive thing - but at the same time, as the multiple viewpoints in this thread illustrate, it's not such a negative thing either.
  70. Rather unfair to dog food[ Go to top ]

    I think using dog food as an anology to compare Spring is unfair to dog food. After all, dogs have to eat, otherwise they'd die. No one needs to eat Spring to survive and depending on the person, Spring might taste like rotten eggs. Dog food is atleast nutritious and gives dogs energy to run. Spring is like sprinkling magic dust on a project and calling it "light". Really it's just a freakin tool. Who cares if Spring source actually uses it to build their own website. Spring it out there to hawk their wares and convince the world it's sliced bread. Take it or leave it. Having used it for the last 14 months on a large complex project, I'm not impressed. It's no better or worse than anything else out there. If DI floats your boat, any DI tool will do, you don't need Spring. peter
  71. No mention of Seamframework.org?[ Go to top ]

    It's a very intentional attempt to eat one's own dogfood!
  72. There is no reason for Spring or JavaLobby to use only a Java based CMS, when Drupal does the job much better. Spring community should spend time on their framework instead of building their CMS website on Spring. I am actually glad, spring is not forcing down their throat to build a cms site on java. That is time wasted in my mind. Their core is building the framework, as long as they do good job in that, I dont care if they run their website in Java or .NET or PHP or RoR or Flash... GO DRUPAL.
  73. While springframework.org does not use Java, Spring's commercial company SpringSource (springsource.com/springframework.com) uses Liferay, a JSR-168 compliant Java portal. For details see: http://www.liferay.com/web/guest/about_us/news/springsource According to this Press Release, "Spring Framework is a key component of Liferay Portal." BTW, I am not affiliated with either SpringSource or Liferay.
  74. While springframework.org does not use Java, Spring's commercial company SpringSource (springsource.com/springframework.com) uses Liferay, a JSR-168 compliant Java portal. For details see:
    http://www.liferay.com/web/guest/about_us/news/springsource

    According to this Press Release, "Spring Framework is a key component of Liferay Portal."

    BTW, I am not affiliated with either SpringSource or Liferay.
    Very interesting. Looks like they took this to heart.
  75. Inaddition to spring framework Liferay Portal use several other opensource Java technologies from Apache.org and others. Download liferay, unzip it and do a search for *.jar to see the list of the Java technologies it uses. Liferay has everything to get started including CMS, Forum, Wiki, Chat, Calendar and over 60 useful portlets available.
  76. I think dogfooding is very important. As a Java developer, I intentionally went to Java technologies to drive my blog. Check out Pebble at SourceForge. It works great, and it's free.
  77. Fabric3[ Go to top ]

    Fabric3 one of the OSS SCA implementations is built enirely on top of the SCA programming model. Fabric 3 Source Repository Thanks Meeraj