[Imps] Opera canvas bug?

Ian Hickson ian at hixie.ch
Fri May 11 11:00:45 PDT 2007

On Fri, 11 May 2007, Kristof Zelechovski wrote:
> But if the script applied to the test causes the browser to display an 
> error message box with a big bang to raise the dead (which it does not 
> do as a rule but it does it in my script debug mode), it should rather 
> be avoided.

I highly recommend removing modal UI for script errors. Many of my tests 
have script errors in browsers that fail the test; I'm not going to add a 
test before everything I use just in case.

In this particular case it's a non-issue. You'd run one test, find it 
doesn't work in the browser that doesn't support <canvas>, and not test 
the other <canvas> tests.

> I would be so much happier if I got the FAIL text only.

So disable the modal alert. :-)

> I have a habit of testing the results returned unless I am sure the test 
> can only fail because of a bug in my own code, which is not the case 
> here because you depend on the browser's implementation; it is not a big 
> problem and it makes life easier.

The tests here can only fail because of a "bug" in the browser -- namely, 
that the browser doesn't support <canvas>.

> Example: if(anElement = getElement()) { anElement.doSomething() }

I'm not going to do that at every point in the code, though, and I see no 
reason to only do it for the getContext() call.

Ian Hickson               U+1047E                )\._.,--....,'``.    fL
http://ln.hixie.ch/       U+263A                /,   _.. \   _\  ;`._ ,.
Things that are impossible just take longer.   `._.-(,_..'--(,_..'`-.;.'

More information about the Implementors mailing list