Wafer Comparison of Open Source Web Frameworks

Discussions

News: Wafer Comparison of Open Source Web Frameworks

  1. Wafer Comparison of Open Source Web Frameworks (36 messages)

    There are so many web frameworks out there, it is hard to know what they all do. The Wafer research project tries to keep them in track with a feature matrix, and by having example applications running, noting the differences. By seeing one application, and how the framework fits in will be very interesting.

    Wafer Framework Home

    Framework Feature Matrix

    Threaded Messages (36)

  2. I found this meaningless. Also, at least for some of the frameworks, it's also incorrect. E.g. contrary to the matrix, Cocoon has i18n, EJB, Web services support etc.

    You won't compare product by simply matching their feature lists. Why would you do it with OSS?
  3. Frameworks![ Go to top ]

    Recently I've downloaded most of sourceforge.net open source java/j2ee frameworks ...waste of time .They all repeat what has been said in Struts Framework. Sun's WAF is original but it not practical!
    Faisal
  4. Frameworks![ Go to top ]

    They all repeat what has been said in Struts Framework.


    yes, I feel the same for more than 10 years. Actually, since I've learnt about Turing's Computer. Everything's a repetition of TC. So, boring :P
  5. Frameworks![ Go to top ]

    Saying "All useless, all just a rehash of Struts" is like saying "All other cars are just a rip off of the Model T". Each framework has it's strengths. Personally, I think Struts is to messy, it tries to cater to too many people, and when it takes 5 different files to make one page, it takes too much.

    At this point, you would expect me to say X is better than Struts. I won't, because each framework has it's place (even Struts :-)
  6. Frameworks![ Go to top ]

    Struts is based on JSP. JSP sucks, so Struts sucks too. :P
  7. Frameworks![ Go to top ]

    <quote>
    Struts is based on JSP. JSP sucks, so Struts sucks too. :P
    </quote>

    In one statement you have managed to demonstrate your complete ignorance of Struts. Struts is based on the Model 2 architecture. It provides support for doing the view using JSP, but is not tied to it. A lot of people have successfully using other technologies like Velocity and XSLT.

    IMHO a great strength of Struts is that you can selectively use what ever pieces of the framework you like. It does not force you to have to have to use them all.

    And like other posts, I do not think that Struts is the be all and end all of web frameworks. There are many other interesting frameworks out there. The advantage of Struts is the big user community, and hence available tools, training and books.
  8. Frameworks![ Go to top ]

    Actually, Model 2 sucks too.
  9. Frameworks![ Go to top ]

    First, you say Struts sucks because it is based on JSP. When it was pointed out that you were clearly wrong, you say Model 2 sucks.

    Why does it suck? What is a better approach?

    Ryan
  10. Frameworks![ Go to top ]

    And I am not even using Struts! RELEGION!
  11. Shut up Faisal[ Go to top ]

    Someone please mark Faisal for deletion.
  12. Frameworks![ Go to top ]

    Struts is based on JSP.

    > JSP sucks, so Struts sucks too. :P

    In fact, I use Struts with Velocity and I'm quite happy.

    Struts isn't based on JSP.
    You didn't know that, so you suck. :)

    Knurpsfloh
  13. Frameworks![ Go to top ]

    me bad. Struts sucks by itself without any relation to the fact that JSP sucks :)
  14. Frameworks![ Go to top ]

    Perhaps you can provide some more qualitative analysis to back up your assertion that most of the frameworks on SourceForge are a "waste of time". I believe you are way off base and have not really analyized the different frameworks because there are significant differences in how they approach web application development.

    It is posts like this which prompted me to create the Wafer project in the first place. Granted the project is far from complete but it is better than "All frameworks suck except the one I use" analytics.
  15. The product feature matrix is only one part of the Wafer project. In addition to the feature matrix there is an example application which will hopefully be implemented using each of the different frameworks. Also planned are more detailed analysis of each of the frameworks with their strengths and weaknesses. Naturally any contributions to the Wafer project are welcome. If you feel that the matrix is incorrect then please send an email to the wafer development list at wafer-developer at lists dot sourceforge dot net explaining the features that are supported yet missing from the matrix and how they are supported in each framework.

    Sincerely,
    Anthony Eden
  16. This is a good idea but hard to do in reality.

    Quite often the concepts being compared are not directly comparable.

    Also, a flexible framework might have less features but be more useful/friendly/productive due to the architecture of how its features work together.

    Perhaps, the site could focus on examples of doing the same thing on each of the frameworks under the feature categories that currently exist in the matrix? This would require quite abit of time and understanding of each but might provide enough information to allow comparison.

    Dan
  17. I agree, this is a good idea, but unfortunately it is not terribly meaningful. Just because a certain framework has a particular feature does not mean that it is powerful or easy to use.

    Take for example the "Designer/Developer separation" category. One person may consider JSP Model 2 to provide such a separation, but another may argue that this is not a real separation at all, and other approaches (that exist in several of those frameworks) are much better for this purpose. It seems to me that using "checks" inevitably leads to oversimplification to the point of making the information useless.

    The second part of the table (EJB integration, etc) seems to be even less useful. Most of the features there are typically provided by outside libraries. Take for example "O/R" -- there are so many "O/R" frameworks with their own advantages and disadvantages, that this choice should really be made by the developer, rather than by app. Forcing you to use a particular one is a negative in my book rather than a positive. It is more important how well a particular framework can be integrated with other O/R libs. Most of the other sections in this category seem to suffer from the same problem.

    Perhaps sending a "questionaire" to the framework developers that asks them besides everything else to provide example code and collating this would be far more useful -- developers can then make a relatively informed decisions as to whether what these frameworks really mean and whether they would be useful or not.

    Oh, and why isn't there a "Components" section in the top part?

    -mb
  18. I don't recall ever claiming on the Wafer Project site that having a feature makes a framework powerful or easy to use. The feature matrix is only designed as a quick way to see the frameworks side-by-side with an indication of what kinds of things you will have to either a.) implement on your own or b.) utilize another tool to handle.

    It is true that these checks are a simplification, and anyone who decides on a product based on a feature matrix such as this is not making an informed decision. However, given the large number of frameworks this is at least a starting point.

    As for the second section, the point of this section is to provide information about whether or not the framework provides some integration with those technologies beyond what is provided in the Java APIs themselves. Thus, for O/R it would indicate that the framework provides some sort of a simplified integration with one or more O/R tools. This can be a big factor in time to market if you have to integrate tools into the framework yourself.

    Sending a questionaire is a very good idea. Perhaps you would like to join the mailing list and contribute a questionaire?

    Finally, "Components" was going to be a category but it was hard for me to determine what that means (at least more so than with the other features). What does component support mean? Does it mean that you can have

    Sincerely,
    Anthony Eden
  19. Whatever![ Go to top ]

    Give me names ...if Stuts & jsp sucks what's left in J2ee that does'nt suck. If u think so name me a component in .Net or whatever programming lang that has the power JSP's Custom tag .


    Faisal
  20. Professionalism ...[ Go to top ]

    It's funny how much more eloquent and grammatical the optimistic people are from the jerk-reaction naysayers.
  21. Whatever![ Go to top ]

    Well, if JSP and Struts are the only things "left in J2EE", then "I'm speachless"

    Servlets, JMS, MDB, Stateless Session EJB are not bad at all.
  22. Whatever![ Go to top ]

    Faisal,

      You sounded so narrow minded and cynical. This is NOT a RELIGION. From your inputs, you obviously have very little knowledge of struts. In case you are wondering where you can find this resource, http://jakarta.apache.org/struts/.
  23. You are correct, comparing these frameworks is extremely difficult due many factors including the various approaches of the frameworks to web application development, developer mentality (what works for one developer or team may not work for another) and other factors.

    There is already work being done on implementing an example web application with each of the frameworks. It is up to the framework community to contribute implementations though as I do not have anywhere near enough time to implement the example in each of the different frameworks.

    Sincerely,
    Anthony Eden
  24. I too agree that it will be very difficult for a comparison like this to yield much. I've found that to really compare different products, it's best if you have used one of them fairly extensively. That will help you identify all the different things the products are trying to achieve. With that knowledge, it's often possible to read through the documentation of a similar product and immediately identify the key pros/cons that the product has.

    Rob
  25. How did this story even get on TSS? I have been monitoring the Wafer Project for about 6 months now, just to see if there was any activity because I was interested in what they would find. There has been no activity. The project, from the surface, is dead. Also, it seems there is no scientific analysis of ANY sort to explain the matrix.
  26. Try http://www.waferproject.org/features/definitions.html for a brief explaination of what a check mark means. There will be more detailed explainations of how each framework implements the specific feature in the future.

    As for lack of activity, perhaps you should contribute instead of complaining. Wafer will only provide a good analysis of competing frameworks through contributions from the community.
  27. JStateMachine promising. Of course it reminds me WL Portal Webflow concept. But portal webflow is too heavyweight.
  28. already interesting ... but ...[ Go to top ]

    It is already interesting...
    ... but :
    The very important information that is missing here is the learning curve.

    In these times, most projects involves multiple technologies. Even subsequent projects in the same organisation use different technologies. On top of that software engineers have a tendancy to switch jobs.
    All this implies that a software developer has to learn new technologies at a continious rate.
    How fast can a developer be brought up-to-speed with the technology or framework is therefor one of the most import aspects for me.

    I realize that this kind of information is hard (or even impossible) to obtain. Compared to information on the learning curve, a list of features is 'very easy' to collect. Probably due to this, it is a pitty to see that some of the frameworks do not look at all at the trade-off between features and complexity.

    In my opinion the art of designing frameworks is to design frameworks in a layered approach. The first layer contains all essential elements : the beginning user is only confronted with a minimal complexity to learn the framework. The second layer adds some more advanced features and therefor some additional complexity... That way it is possible to build a feature-rich framework that is easy to learn.

    hope you found my thoughts interesting enough to read.
  29. already interesting ... but ...[ Go to top ]

    Hi

       I do agree with you,they have missed out the learning curve.Struts is a very good framework.But it takes at least 1 week of learning curve for most of the developers.

    Regards

    Tanmoy
  30. already interesting ... but ...[ Go to top ]

    Hi

       I do agree with you,they have missed out to mention about the learning curve.Struts is a very good framework . But it takes at least 1 week of learning curve for most of the developers.And its become more steep when you need to handle a complex UI.

    Regards

    Tanmoy
  31. already interesting ... but ...[ Go to top ]

    I agree with you that learning curve is a huge factor in how a framework stacks up against its competition. Unfortunately the time it takes for you to learn a framework will probably be different than the time it takes for someone else to learn the same framework. Additionally, some frameworks may be easier for you but harder for someone else.

    One of the ways that this information can be provided is through a more in-depth analysis of each framework's strengths and weaknesses.

    Sincerely,
    Anthony Eden
  32. I have been thinking of a comparison idea like these for many open source and java solutions. The power of Java and open source has amplified solution invention and we need ways to assess these solutions and make the best of them - combine the best and throw away the rest while allowing for choice.

    I applaud Wafer's move and pls, note that you may not get it right the first time but in the long run it can be achieved. all we need to do is join in making the clarification of technologies easier so the dream may come true.

    'Bioye
  33. I have been thinking of a comparison idea like these for many open source and java solutions. The power of Java and open source has amplified solution invention and we need ways to assess these solutions and make the best of them - combine the best and throw away the rest while allowing for choice.

    I applaud Wafer's move and pls, note that you may not get it right the first time but in the long run it can be achieved. all we need to do is join in making the clarification of technologies easier so the dream may come true.

    'Bioye
  34. I have been thinking of a comparison idea like these for many open source and java solutions. The power of Java and open source has amplified solution invention and we need ways to assess these solutions and make the best of them - combine the best and throw away the rest while allowing for choice.

    I applaud Wafer's move and pls, note that you may not get it right the first time but in the long run it can be achieved. all we need to do is join in making the clarification of technologies easier so the dream may come true.

    'Bioye
  35. I think that its a good idea, but needs to get deeper. A good start.
  36. Whatever![ Go to top ]

    Mr OO
    How the hell do u know how much I know about Struts.I did not say that Struts is the most perfect framework nor did I dimiss the rest nd if u check my replies -I DID NOT SAY STRUTS IS THE BEST .As far as I am concerned there is no perfect & off the shelf framework for everybody.Had i been norrow minded I would't have downloaded most - if not all the frameworks and checked them !
    Faisal
  37. Whatever![ Go to top ]

    nd if u don't mind could u tell how did u find about my ignorance of Struts Framework
    Faisal