644392 - Flash ExternalInterface methods no longer accept JS arrays, only Strings (FF 4.0)

Status: RESOLVED INCOMPLETE

(Core :: JavaScript Engine, defect)

Description

[Matt Bishop]

User-Agent:       Mozilla/5.0 (Macintosh; U; Intel Mac OS X 10.6; en-US; rv:1.9.2.16) Gecko/20110319 Firefox/3.6.16
Build Identifier: Firefox 4.0

A SWF file can expose methods for Javascript to call. These are called ExternalInterfaces.  One can pass javascript types, including arrays and arrays of arrays, into these ExternalInterface methods.

In every browser I have tried, including Firefox 3.6.15, Chrome 10, Safari 5.0.4, and IE8, this functionality works as expected.  However, Firefox 4.0 will only accept string values.  It will not accept arrays, as other browsers do.

Reproducible: Always

Steps to Reproduce:
1. Log into carbonexploration.cloud-spring.com
2. Add some items to your profile by dragging them from the left palette into the middle of the screen.  Add several different types of items.
3. Click on the "Explore" tab and see a large SWF load up.

Actual Results:  
You don't see graphs, though you do see text labels.  The label text is passed in via ExternalInterface calls as Strings.  The data that drives the graphs is passed in as 2D arrays of Strings.

Expected Results:  
You should see some horizontal bar graphs with your carbon footprint data.

It should have rendered graphs.  I have ascertained that the data for the graphs is indeed making it to the Javascript the pushes the array data into the SWF via Javascript.

Comment 1

[Ed Morley [:emorley]]

Would you be able to try and find the regression range:
http://harthur.github.com/mozregression/

Comment 2

[Matt Bishop]

I have confirmed this bug on FF4.0 for WinXP too.

I will work on regressing it.

Comment 3

[Matt Bishop]

Also OS X 10.6.

Comment 4

[Ed Morley [:emorley]]

Matt, any luck with the regression range?

Does it still occur using 4.0.1 or perhaps using the latest nightly?

Comment 5

[Thomas Ahlblom]

Matt, have you found any regression range yet?

Comment 6

[Paul Townsend]

I am still seeing this issue on FF 5.0.1

Comment 7

[Vlad [QA]]

Any news about this issue?

Comment 8

[(mostly gone) XtC4UaLL [:xtc4uall]]

Confirmed against Mozilla/5.0 (Windows NT 5.1; rv:8.0a1) Gecko/20110809 Firefox/8.0a1 ID:20110809030751

WFM against:
Opera/9.80 (Windows NT 5.1; U; Edition Next; en) Presto/2.9.186 Version/12.00
Mozilla/5.0 (Windows NT 5.1) AppleWebKit/535.1 (KHTML, like Gecko) Chrome/15.0.847.0 Safari/535.1

Will do some Regrange Hunting later on ...

Comment 9

[(mostly gone) XtC4UaLL [:xtc4uall]]

Last good nightly: 2011-01-04
First bad nightly: 2011-01-05

Pushlog:
http://hg.mozilla.org/mozilla-central/pushloghtml?fromchange=e3b1c75995ac&tochange=d2bd42931b03

Comment 10

[(mostly gone) XtC4UaLL [:xtc4uall]]

Against TM repo:

Last good nightly: 2010-12-30
First bad nightly: 2010-12-31

Pushlog:
http://hg.mozilla.org/tracemonkey/pushloghtml?fromchange=7afbb62b190b&tochange=8dc385ffb820

Comment 11

[Benjamin Smedberg]

Is it possible to get a contained testcase for this that doesn't involve a login screen? The TM regression range doesn't have any obvious suspects, and I've verified that at least in basic tests it is possible to pass JS arrays through the NPAPI boundary still.

Comment 12

[(mostly gone) XtC4UaLL [:xtc4uall]]

FWIW, I created a Testaccount on the Site what exposes the Issue:
L: Bug644392
P: mozilla

Comment 13

[u279076]

@bsmedberg is the account info provided by comment 12 sufficient?

Comment 14

[Manuela Muntean [Away]]

The URL provided in the description isn't available anymore. I've tried both on the latest Nightly (20130111030906) and Chrome.

Is there another URL that I can test with?

Comment 15

[Matt Bishop]

The site I found it with is no longer active, sorry.

Comment 16

[u279076]

Unfortunately, I don't think there's much QA can do without a reproducible website or minimized testcase. Please re-add qawanted if/when this can be provided.

Comment 17

[Ciprian Muresan [:cmuresan], Formerly of Ecosystem QA - please needinfo @mcoman for things regarding the team]

Hi,

Based on comment 14 and comment 15 I will mark this issue as Resolved-Incomplete. If you find an URL where this issue is reproducible please feel free to reopen this bug.

Thanks,
Cipri