I'm going with Taha's approach but I'm having some difficulties and questions. In Taha's article he referred to his contributed page this way " debitCardModule/instructionblocks" where "debitCardModule" is the page's module name in camel case (first letter is lowered) and then the page name, all in lower case. I tried contributing the page name as a string as the article states and using it's .class type but both didn't work, it's always returns this:
Render queue error in BeginRender[Index:layout.delegate]: Failure reading parameter 'to' of component Index:layout.delegate: Unable to resolve 'adminModule/adminmodulemenus' to a known page name. I have some questions: 1. The "ComponentSrouce" service, should It be injected by the plugin module to be able to resolve pages contributed by the same module (I can't imagine how can I inject a service in a static method to contribute my configuration - page name and block id) 2. How does the main app know where to find my modules package to locate the pages and components packages ? Like in the web.xml, the app's package is stated but how is it done with a contributing module ? 3. How can a contributed block, have links to pages in the main app ? On Sun, Aug 26, 2012 at 11:55 PM, Muhammad Gelbana <m.gelb...@gmail.com>wrote: > The blog post is fantastic but It lacks the point that Taha tackled which > is contributing blocks to the main application. My goal here isn't to > contribute blocks (which seems to be the best option so far because it > solves the problem of having links displayed as sub links to another top > link. Igor tried to solve this using annotations which is really nice. I > still have to compare between using annotations and blocks to categories my > page links. Although for some reason, I feel contributing blocks will be > easier, may be I'm wrong. > > May be I can work around contributing blocks by having components act as > the contributed tml part and annotate the component with a reasonable > annotation to indicate that it's template has to be displayed ? Then Igor's > loop would have a "delegate" component instead of the hyperlink one ? > > > On Sun, Aug 26, 2012 at 9:21 PM, Richard Frovarp <rfrov...@apache.org>wrote: > >> On 08/26/2012 10:23 AM, Muhammad Gelbana wrote: >> >>> After reading Taha's >>> article<http://tawus.**wordpress.com/2011/04/20/** >>> tapestry-magic-3-plugin-**blocks/<http://tawus.wordpress.com/2011/04/20/tapestry-magic-3-plugin-blocks/> >>> >, >>> >>> I'm now trying to dissect my application to be able to distribute it to >>> different clients. So it will be distributed as one slim web application >>> with really nothing to offer. And to use one feature, a tapestry module >>> (.jar) file will just be dropped in the classpath and it should be >>> auto-loaded (using tapestry). >>> >>> I find Taha's article highly innovative but I'm trying to find out if >>> anyone tackled the problem with a different approach that could be >>> "simpler" or capable of handling what Taha's solution probably didn't. >>> >>> >> This might prove useful: >> http://blog.tapestry5.de/**index.php/2010/01/19/tapestry-** >> ioc-modularization-of-web-**applications-without-osgi/<http://blog.tapestry5.de/index.php/2010/01/19/tapestry-ioc-modularization-of-web-applications-without-osgi/> >> >> ------------------------------**------------------------------**--------- >> To unsubscribe, e-mail: >> users-unsubscribe@tapestry.**apache.org<users-unsubscr...@tapestry.apache.org> >> For additional commands, e-mail: users-h...@tapestry.apache.org >> >> > > > -- > *Regards,* > *Muhammad Gelbana > Java Developer* > > -- *Regards,* *Muhammad Gelbana Java Developer*
