The style of a button for changing page layout doesn't happen. This is done through Javascript and something is obviously going on since the button is changing as the cursor moves over it.

Was tested using the 5/17/99 build of NGLayout for a Windows platform

hi, reboughner. nglayout bugs usually should get filed under the "DOM level 1" or "DOM level 0" component. thanks!!!

Eric, I believe the style change notification is going through. The change doesn't seem to be propogated to the widget, though.

Using the July 26 build for Linux, this seems to be working. I think this should be WORKSFORME, but can anyone else confirm? This bug report isn't too clear. I simplified the attachment and got rid of the extra bits to try and test this a bit better, and asides from a minor button rendering bug, it works for me.

I've retried this example with the M8 build and it still doesn't work. Obviously, something is going on because I see the button quivver but its style or the lettering on it doesn't change. It works fine in IE 5.0

Oops, my mistake. Everything works as of the July 30th build on windows 98 except that the button does not change color. It should toggle to yellow from black onMouseOver.

Moving to M16. Result is aesthetically incorrect but nonfatal and pollmann's got lots of worse bugs to fix for now.

Bulk moving [testcase] code to new testcase keyword. Sorry for the spam!

Adding beta1 keyword

PDT wants to know more details on why this is a beta1 stopper

Moving to M14 per beta1 keyword.

This already almost works. The problem is that you get btn.style.color which was set to "#000000" and we return "rgb(0,0,0)". If you compare that string format rather than "#nnnnnn", it works. (Of course for black and yellow you could just use the color name.) Attaching a test case.

I am glad that the example works with the proper coding. Just two comments: 1. I rarely use the rgb(x,x,x) form for the color since I prefer the hex format. I just didn't think to try it. 2. For your information, IE does return the hex format and that is the reason I used that form.

Since rgb(0,0,0) is a valid, canonical format for returning the color attribute, I'm going to close this issue as invalid. It is part of the DOM spec that attributes may be returned in a format different than that specified. For more information on this see the following URL: http://www.w3.org/TR/DOM-Level-2/css.html#CSS-CSS2Properties Thanks!

Per pollmann's comments, marking Verified. PDT-