Framework needs a mobileready event.

[jblas]

Historically folks have been waiting for the first pageshow or pagechange event to figure out when they can run their code to initialize their application. This is fine for some folks, but some developers need to know is when the framework is actually done initializing itself. Since set up of the first page, resulting in the first pageshow and pagechange, is just part of the framwork initialization, it would be nice to know definitively when it is done initializing itself.

What we really need is a mobileinitcomplete or mobileready event to tell folks the framework is initialized and ready to go.

[scottjehl]

this is before the first pagecreate, or..?

[jblas]

This is after all is said and done ... the point being that as PART of mobile initializing, it creates the pages, as well as other init types of things ... so folks want to know when it is all done.

[jblas]

I took a look at what it would take to add this, so I'm just jotting down some notes here of what I noticed when looking at jquery.mobile.init.js.

• We still fire off a DOM Ready handler even when $.mobile.autoInitializePage is false.
• In the DOM Ready handler does 3 things:
  • Scrolls the document by 1 pixel and attempts to initialize $.mobile.defaultHomeScroll based on the results.
  • Calls $.mobile.initialzePage() conditionally based on the value of $.mobile.autoInitializePage.
  • Sets up an onload handler that executes a silentScroll() to hide the browser's nav bar.

There are a couple of problems I'm seeing that we'll need handle to accurately tell when we're all done initializing:

• Even after initializePage() is done executing, the first page may not be ready, especially in the case where the document URL contains a hash, since it fires off a hashchange event to trigger what ever page it refers to, to load, which is an async operation.
• We should execute the silentScroll() after the initial page is shown, or onload, whichever comes last. The reason we need to do it this way is because, the size of the page may not be accurate before the onload handler fires, because any inline images may not have loaded yet.

[toddparker]

@scottjehl - Mind weighing in here?

[jaspermdegroot]

In what way is this different from the pageinit event?

[pillex]

i need to scroll the page when first loaded. sometimes works, sometimes jQM scrolls page back to top. probably a race condition. i guess a mobileready event would help to solve this without checking for race conditions.

@uGoMobi - probably after pageinit the framework is working heavily. after mobileready it's doing nothing

[arschmitz]

@pillex is there a reason you cant use pageshow for this. your example seems like a situation where this would be what your looking for.

[pillex]

i'm using lab.js which loads .js async. don't have the problem currently. but back then i tried everything, also pageshow. and i was looking for .bind( 'totally-finished-doing-nothing-anymore' , scroll_down() )

[yaffol]

I'm also using lab.js to aysnchronously load in jquery mobile and other javascript assets. It's proving difficult to find an event that can be bound to, to handle things like custom extra markup, which I previously bound to pageBeforeCreate.
When loading with lab.js, it seems that the bindings are not available, at the point at which they are called - which is synchronously after the loading of jquery mobile (lab.js with a .wait() method, which is meant to only execute the next script after the previous one has 'completed') - but I'm not sure this is actually the case in reality.

Having a 'framework has definitely finished' event would be really great :)
!

I've ended up binding to window.load - I think that a 'framework ready' event would be the answer to this.

[Ruffio]

Can this three year old high priority issue be closed?

[arschmitz]

I'm actually going to keep this one open there has been some recent internal discussion about this and it may actually be addressed in 1.5 assigning this to myself

[apsdehal]

@arschmitz Any updates on this? We have a "mobileinit" event.

[arschmitz]

yes lets close it there has not been any one asking if they do we can reconsidder