getting " Cannot read property 'type' of undefined" after my code returns?

[fearless_fool]

I'm getting the following error (show with stack trace) after my code completes:

1. Uncaught TypeError: Cannot read property 'type' of undefined
1. Vcmain.js:7
2. (anonymous function)
3. Ve.mainmain.js:40
4. (anonymous function)main.js:25
5. bmain.js:11
6. Se.(anonymous function).kcmain.js:25
7. Mmain.js:9
8. Se.(anonymous function).kcmain.js:25
9. Re.(anonymous function).Kdmain.js:24
10. Se.(anonymous function).kcmain.js:25
11. Wemain.js:25
12. (anonymous function)%7Bcommon,util,controls,places_impl%7D.js:1

Here's my js code (pretty straightforward):

$(function() {

    var loc = new google.maps.LatLng($('#map_canvas').attr('lat'), $('#map_canvas').attr('lng'));

    var maxZoomService = new google.maps.MaxZoomService();

    maxZoomService.getMaxZoomAtLatLng(loc, function(response) {

        // zoom to maxZoom - 1 if we can determine it, 14 otherwise

        var zoom = (response.status == google.maps.MaxZoomStatus.OK) ? (response.zoom - 1) : 14;

        var mapOptions = {

            center: loc,

            zoom: zoom,

            mapTypeId: google.maps.MapTypeId.SATELLITE};

        var myMap = new google.maps.Map(document.getElementById("map_canvas"), mapOptions);

    });

});

What makes this difficult to debug is that the TypeError happens deep inside the google.maps.__gjsload__ code, long after my code has returned.  I can't figure out what state my code is creating that is causing the google library code to choke.

Any ideas, or suggestions on how to debug?

[Barry Hunter]

Not sure how you attach the attributes, to #map_canvas, but are they numbers?

The API expects you to be sending numbers (not strings) when creating
a LatLng object.

try wrapping them in parseFloat.

If that guess doesnt help suggest supplying a link to your map.

> --
> You received this message because you are subscribed to the Google Groups
> "Google Maps JavaScript API v3" group.
> To view this discussion on the web visit
> https://groups.google.com/d/msg/google-maps-js-api-v3/-/NDVqO_jXaDcJ.
> To post to this group, send email to google-map...@googlegroups.com.
> To unsubscribe from this group, send email to
> google-maps-js-a...@googlegroups.com.
> For more options, visit this group at
> http://groups.google.com/group/google-maps-js-api-v3?hl=en.

[fearless_fool]

@barryhunter:

Good suggestion, but it appears that I am passing floats (and not strings).  But I realized that I was loading the //maps/googleapis.com/maps script twice on the same page -- fixing that makes the problem -- er -- different.

Now I'm seeing the script work the first time and fail thereafter.  More specifically, the async callback to 

    maxZoomService.getMaxZoomAtLatLng()

gets called the first time the page is loaded.  But if I reload the page, the callback never happens (at least, it doesn't trigger a breakpoint in Chrome's debugger), and I get a 

Uncaught TypeError: Cannot read property 'value' of null

error from within

  %7Bcommon,util,controls,places_impl,maxzoom%7D.js:4

after my function returns.  Curiouser and curiouser.

[Barry Hunter]

On Tue, Jan 31, 2012 at 12:25 AM, fearless_fool <rdp...@gmail.com> wrote:
>  But I realized that I was loading the //maps/googleapis.com/maps script
> twice on the same page

we have NO POSSIBLE WAY to know you did that. Unless you have another
thread, but its kind of rude to split a discussion without maintaining
a link between the threads.

That's pretty much why a link is always asked for.

(I suppose unless we find a way to hack into your computer, so we can
see your screen. But that is a bit OTT for a help forum)

[Robert Poor]

Barry!

On Tue, Jan 31, 2012 at 04:23, Barry Hunter <barryb...@gmail.com> wrote:
> we have NO POSSIBLE WAY to know you did that. Unless you have another
> thread, but its kind of rude to split a discussion without maintaining
> a link between the threads.

Oy! Ouch. I intended no rudeness. I'd verified that I was passing
floats, as you'd suggested. Then I noticed that I was loading the
google scripts twice and thought that might make a difference --
loading the script once didn't make the problem go away, it just made
it different. So I reported on that.

> That's pretty much why a link is always asked for.

Agreed. It's a reasonable first step to say "I've observed problem X
-- has anyone else observed anything similar?" before posting a link.
Since we've gone that route, a link might be helpful. See below.

> (I suppose unless we find a way to hack into your computer, so we can
> see your screen. But that is a bit OTT for a help forum)

I'm honored that you're so invested in helping. A link is
problematic: if you're asking for a standalone / minimum case app,
that would take a LONG time to extract and test. The entire app is a
large Ruby on Rails project, being developed for a client and isn't
public. We could set you up with access to the app -- PM me if you
want to go this route.

Thanks!

- ff