+1 It will be selfish for someone to preventing T5 only because of the compatibility.
T5 is a renovation. When it out, everyone will be benefit. For the future of tapestry, T5 should going on. If you like tapestry, support it. 2006/7/29, Howard Lewis Ship <[EMAIL PROTECTED]>:
The current state of the T4 code is untenable w.r.t. to adding new features without breaking backwards compatibility. Each upgrade of Tapestry (2 to 3 to 4) has had major upgrade problems because of a number of factors, mostly the design of the APIs and the need to extend from base classes (as the base classes provide much of the component functionality). In the long view of Tapestry, the future users vastly outnumber the current users. Tapestry 5 is targetting that group of users. Existing applications coded to Tapestry 4 ... well, leave them on Tapestry 4! Many users prefer to stay on Tapestry 3. They don't want to face the upgrade from 3 to 4 if their application is working. The design of Tapestry 5 will facilitate easy upgrades from 5 on ... mainly because the API is minimal and flexible, consisting almost entirely of annotations, rather the classes to extend and interfaces to extend. Having an easy upgrade path from Tapestry 4 to 5 is a near insurmountable challenge. Is Tapestry 5 a new framework? Yes. Will it be an easy transition for developers (not code)? Yes. Tapestry 4 developers will see Tapestry 5 and understand it quickly and easily, and be happy about the new features. I see a lot of feedback from people I'm training in Tapestry 4. I see people stumbling because of inconsistencies of naming, because of the odd model (abstract classes), because of the need to understand too much of the internals of Tapestry. I do get upset when people act like I'm holding a gun to their heads. Now, in terms of the IoC container specifically ... I'm preaching about simplicity, simplicity, simplicity. HiveMind is simple, but still has a lot of mapping between Java and XML. As happy as I am with HiveMind and the HiveMind community, the core fact is that it is not as simple as I need it to be: the pure Java model of Tapestry 5 IoC is going to fit like a glove with the rest of Tapestry 5 in a way HiveMind (and certainly not Spring) could ever do. I've repeatedly spoken about, and blogged about, why Spring is simply not a good fit for Tapestry 4 or Tapestry 5. Spring exists to allow an application, a known entity, to be assembled. It simply doesn't inlcude the concepts necessary to configure a framework, allowing for the precise ovrrides and extensions that HiveMind and Tapestry 5 IoC do. I've seen some laughable postings about what it takes to even start to approximate HiveMind configurations in Spring ... huge amounts of ugly Java code and uglier XML. No thank you. My metaphor for all this is that I'm trying to raise the big pole that the circus tent will hang from. I can't wait to get the pole up so we can start having the circus! Under this metaphor, the circus will be the components and extensions that the community will put together. Finally, I've been at this for six years now. I enjoy working on the code, and I suffer through the things necessary to permit that ... Tapestry training, and marketting. Writing books. Creating demos. All that stuff. What I constantly hear is demands and suggestions about Tapestry and all I ask in return for all this time and effort is for people to give back a little something in kind. I don't ask for money (not for the software, not even as a donation), or time, or any other demand BUT that people HELP PUBLICIZE TAPESTRY. It is in the interest of the entire community that the word on Tapestry get out there. There are good books, better than average documentation, and healthy community but Tapestry just isn't getting the notice it deserves and a major reason for this is that people are failing to get the word out. How many of you have written a white paper on your Tapestry experience? How many have posted a supportive comment about Tapestry on a release posting at JavaLobby, TheServerSide or elsewhere? How many have tried to write an article on their Tapestry experiences for publication? Tapestry doesn't have a corporate sponsor like JSF or even Rails. Tapestry has its technical chops and its community. So if you are looking for reasons why Tapestry doesn't have the level of adoption you think it should have, don't look at the compatibility straw man. Look to yourself and what you have done to support Tapestry. On 7/28/06, Matt Welch <[EMAIL PROTECTED]> wrote: > Howard, if it really will be as difficult to move from Tap4 to Tap5 as you > suggest, and if this new code base is indeed mostly new, perhaps it might be > prudent to release what you are now calling Tapestry 5 as a new project > instead; one that is "inspired" by the Tapestry concepts and intentions. > Then Tapestry 4 could continued to be mainatained and if other contributers > we re so inclined, it could be upgraded to a more "migration friendly" > Tapestry 5. > > Personally, I don't have a problem with the migration difficulty as I dont' > have any Tapestry 3 or 4 projects that I would upgrade. I built what I've > built using the best options I had available at the time and while I > continue to maintain some of those appliactions, I dont' feel a pressing > need to upgrade them to the newest framework. I won't be upgrading them from > Spring 1.2 to Spring 2.0 either. What's the big deal? As far as I'm > concerned, there should be a migration path through all point releases, but > any easy migration is just gravy. > > I for one am thrilled to see that Tap5 is dropping some of the encumberances > of it's original implementaton. When I start a new project. I want it to be > using the best tools out available. Here's to Tap5 and all it's > incompatibilities! > > On 7/28/06, Howard Lewis Ship <[EMAIL PROTECTED]> wrote: > > > > Right now its impossible because there's nothing to convert to :-) > > > > It will be *VERY* difficult. This isn't a slap of new paint. Basic > > paradigms are shifting around in a major way. It would be comparable, > > or perhaps even larger than, converting between JSF and Tapestry 4. > > Possibly on the order of converting from Struts to Tapestry 4. > > > > On 7/27/06, Norbert Sándor <[EMAIL PROTECTED]> wrote: > > > I know that it's far away, but how easy/difficult will it be to convert > > > an application from 4 to 5? > > > > > > Regards, > > > Norbi > > > > > > --------------------------------------------------------------------- > > > To unsubscribe, e-mail: [EMAIL PROTECTED] > > > For additional commands, e-mail: [EMAIL PROTECTED] > > > > > > > > > > > > -- > > Howard M. Lewis Ship > > TWD Consulting, Inc. > > Independent J2EE / Open-Source Java Consultant > > Creator and PMC Chair, Apache Tapestry > > Creator, Apache HiveMind > > > > Professional Tapestry training, mentoring, support > > and project work. http://howardlewisship.com > > > > --------------------------------------------------------------------- > > To unsubscribe, e-mail: [EMAIL PROTECTED] > > For additional commands, e-mail: [EMAIL PROTECTED] > > > > > > -- Howard M. Lewis Ship TWD Consulting, Inc. Independent J2EE / Open-Source Java Consultant Creator and PMC Chair, Apache Tapestry Creator, Apache HiveMind Professional Tapestry training, mentoring, support and project work. http://howardlewisship.com --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
