GWT Shared Entry Point with iPhone UIWebView - iphone

Question

I have an app which works fine at the moment and uses geolocation. I have a method which calls com.google.gwt.geolocation.client.Geolocation.
I extended the existing entry point and overriden this method to use the native Geolocation library so I don't have any permission popups.
How can I avoid having two modules which takes twice as long to compile?

Answer 1

In GWT normally you have an entry-point per application, or perl .html page, but you can have other scenarios: you could have a module file (.gwt.xml) with multiple entry-points, or a page loading multiple modules (.cache.js), or even you could load the same module (.cache.js) in different pages.
So, in your case, you could maintain one entrypoint and one module file and include the same compiled module in both pages. In this case you have to write some code in your entry point to know what to do in each page:
public void onModuleLoad() {
if (Window.Location.getPath().matches(".*page_1.html.*")) {
// do something
} else {
// do another thing
}
}
Think that in this case you will have all the compiled stuff in both pages, but you could take advantage of gwt code-splitting and make each page load just the stuff it needs:
public void onModuleLoad() {
if (Window.Location.getPath().matches(".*page_1.html.*")) {
GWT.runAsync(new RunAsyncCallback() {
public void onSuccess() {
// do something
}
public void onFailure(Throwable reason) {
}
});
} else {
GWT.runAsync(new RunAsyncCallback() {
public void onSuccess() {
// do another thing
}
public void onFailure(Throwable reason) {
}
});
}
}
The advantages of this approach is that you compile just once for all your site, and you share the same code in all your pages taking advantage of the cache. There are disadvantages as well like your final code is higher, dev-mode could be slower, etc.