Yes I agree with you. The architecture needs more documentation, especially there are some really important API classes that have no JavaDoc at all. As for the maven-faces-plugin, I really dislike it. It does a nice job, but it is really hard to work with and modify (speaking from experience). There has been some discussion to move Trinidad to the new annotation builder plug-in, but I personally don't know the status of that and the opinion of the Trinidad contributors of changing the process.
Yeah, writing that WIKI gave me typer's cramp, it was the last straw that pushed me to switch from QWERTY to the Dvorak keyboard layout, so I agree that it should be more simple. FYI, I think that there are maven archetypes out there that make it easier, but I am not 100% certain there is an appropriate one. As for more components, it is hard to write something that has no ideas. No one has put any effort into suggesting, working on or submitting new components to the Trinidad sandbox. So basically, if there is no demand for specific components, why spend your own personal time writing them and trying to guess what people want? Skinning is a lot of work and I started one skin, but I lost my motivation, but I think there is some that are working on one (search the dev@ archives). One problem I find with working with Trinidad is still supporting IE6 which makes life extremely miserable. The JS layer needs some major refactoring (like no global functions and using a trinidad namespace for all JS code for example) as well that makes any JS involved component be more work than it should be. Oracle still puts a lot of effort in Trinidad support and it still uses Trinidad as a base for the rich client framework (if you want to see the public demo, you can surf here: http://jdevadf.oracle.com/adf-richclient-demo/faces/index.jspx) so that is the primary reason that Trinidad's core framework is constantly worked on, but since Oracle has their own renderers for these components, there isn't the same focus on the Trinidad renderers as there is on the framework. Like all open source projects, there has to be a good user and contributor community for it to truly prosper. Should people work on new sandbox components and volunteer and start submitting patches for the skin framework, progress would be seen. As for how to contribute, it is as simple as creating JIRA tickets at http://issues.apache.org and submitting patches. If a patch seems to be growing old, then it is typical to ping the users@ or dev@ mailing list requesting someone to look into it. It helps if there is adequate comments. For any API changes, it is always best to discuss the change on the dev@ mailing list. Before I became a committer I just helped out on the mailing lists and submitted patches, it really is not too difficult to get involved if the desire is there. -Andrew On Fri, Jun 5, 2009 at 7:25 AM, Luka Surija<luka.sur...@iytim.hr> wrote: > Andrew, > it's always the same reason. > Trinidad is too complicated do extend without strong knowledge: > 1. in it's architecture isn't well documented > 2. in maven-faces-plugin > > I'm familiar with your wiki page ( > http://wiki.apache.org/myfaces/Facelets_and_JSF_1.1_maven-faces-plugin_Getting_started > ) but, as you can see it takes too much effort to make a simple hello world > component and I think this is a main reason why there is almost no new > components in last 2 years. I'm using trinidad since m1-incubating version > and I'm tracking all the changes. In my opinion trinidad is potentialy > "the" best jsf framework currently available, but it lacks some features > to make it the best: > 1. More components > 2. easier component development > 3. new AJAX under-layer to track component changes (to allow push technology > one day). See ICEFaces as a example how to make a framework so popular with > this technology. But in the background it lacks so many thing and nobody > cares. I've tried ICEFaces, and for "hello world" application is ok, but for > anything more complex, trinidad is 100x better and more developer friendly + > has better browser compatibility. > 4. wow skin to make him more attractive. > > I'm sure that many trinidad users (developers) are willing to contribute to > the community starting from my self if they knew how to do it in some > easier way. > > Best regards, > > Luka Surija > > +385 1 61 99 140 > +385 98 434 061 > l...@iytim.hr > > I.Y. tim d.o.o. > Vrbik 3, HR-10000 Zagreb > www.iytim.hr > i...@iytim.hr > > > > Andrew Robinson wrote: >> >> I would not jump to just saying "that's totally wrong ..." when there >> is truth to the observation. I cannot speak for Tobago, but there are >> areas of Trinidad that have not significantly changed in years. This >> may give the impression that there is not much ongoing development. >> What you find with Trinidad is that the server side framework is very >> well supported, as well as the components classes, but the Renderers, >> skins and JavaScript of Trinidad are very much neglected. >> >> This is probably a result of many that extend Trinidad but do not >> necessarily contribute those extensions back. There is a Trinidad >> Sandbox, but unlike Tomahawk, there is no activity in it for the most >> part. I am not sure why this is and what we can do to motivate our >> users to provide new components and enhancements to existing >> components. >> >> So as a result, you will probably find that Trinidad is very solid, >> the server side keeps up to date with other libraries and with some >> new JSF technologies and there is a great community of support at the >> framework level. Just what is lacking is active support of the >> component offerings and the look and feel of Trinidad. >> >> As Apache relies on its users quite a bit, new patches, and new >> components are welcome, especially for the sandbox as it is a great >> testing ground for new ideas without having to perform all the >> architectural discussions up front. Then components can be brought >> into the core as they gain popularity and their architecture can be >> standardized if not already. >> >> -Andrew >> >> On Thu, Jun 4, 2009 at 9:00 AM, Matthias Wessendorf <mat...@apache.org> >> wrote: >> >>> >>> On Thu, Jun 4, 2009 at 2:21 AM, Christian Groove >>> <groo...@groovesytems.de> wrote: >>> >>>> >>>> Salut >>>> >>>>> >>>>> Hi All, >>>>> >>>>> We are developing a web application with myfaces. >>>>> >>>>> >>>> >>>> That's fine, JSF is the cool think. I worked with >>>> Tobago, that comes with a great layout manager >>>> and some cool widgets. >>>> >>>> It is not my intention to overwhelm that nice >>>> project but it seems to be dead. The development >>>> >>> >>> that's totally wrong ... >>> >>> >>>> >>>> of new widgets seems not to proceed so you may >>>> better look to other Taglibs like Richfaces, Icefaces. >>>> >>>> Groovy >>>> >>>>> >>>>> This application needs proper navigation and UI .I have observed >>>>> Trinidad >>>>> has good navigation practices and other utilities.Tobago has some good >>>>> layouts but not good navigation practices.both these technologies have >>>>> featurs like PPR. >>>>> >>>>> could some one plaese advice us which is the better technology and the >>>>> distinct features of these technologies. >>>>> >>>>> Thanks, >>>>> Srikanth >>>>> >>>>> >>>>> >>>>> >>>>> >>>> >>>> >>> >>> -- >>> Matthias Wessendorf >>> >>> blog: http://matthiaswessendorf.wordpress.com/ >>> sessions: http://www.slideshare.net/mwessendorf >>> twitter: http://twitter.com/mwessendorf >>> >>> >> >> >
