«

I would like to automatically resize a window to its content. I was thinking
of using an outer table to detect the size of my content, then use
window.resizeTo( ).

The problem is I'm not sure how I can determine the size of what I think is
referred to as the 'chrome' - the window borders, toolbars, etc.

Is there a way to determine the height and width of the chrome so that I can
supply the correct parameters to the resizeTo method?

Peter wrote:
> I would like to automatically resize a window to its
> content. I was thinking of using an outer table to
> detect the size of my content, then use window.resizeTo( ).
>
> The problem is I'm not sure how I can determine the size
> of what I think is referred to as the 'chrome' - the
> window borders, toolbars, etc.
>
> Is there a way to determine the height and width of the
> chrome so that I can supply the correct parameters to the
> resizeTo method?

The dimension of the top and left chrome can be determined on browsers
that expose the screen offsets of their window to scripting, given a
mouse input event. By reading the mouse's screen co-ordinates, and its
viewport co-ordinates and using the difference to give the screen
co-ordinates of the viewport. The difference between the screen
co-ordinates of the window and the screen co-ordinates of the viewport
give the dimensions of the chrome at the top of the window and the
chrome to the left.

The screen co-ordinates of the window are not universally available in
browser object models (and in framesets the values don't necessarily
have the same meaning).

The other proposed possible strategy for determining the chrome
dimensions in total is to measure the viewport dimensions, re-size the
window to the original viewport dimensions and then re-measure the
viewport dimensions. The difference being the total dimensions of the
chrome (left+right and top+bottom), assuming that the second set of
viewport dimensions didn't drop to zero, which would invalidate the
results.

Obviously re-sizing the window a couple of times is not going to look
that good to the user. And these days the ability to re-size a browser
window is subject to user veto, so it will often not work at all.

A better approach is probably to adopt a fluid HTML design that adapts
to whatever size the user's window happens to be.

Richard.

Thank you Richard for the thorough response. That helps a lot. As far as
user experience goes, I agree with your opinion, but in this case, it is for
a private web application and the need is for the browser to conform to the
size of the content. Since I have only one browser to target, this should be
doable.

Thanks again,
Peter

Richard Cornford wrote:
> Peter wrote:
>> I would like to automatically resize a window to its
>> content. I was thinking of using an outer table to
>> detect the size of my content, then use window.resizeTo( ).
>>
>> The problem is I'm not sure how I can determine the size
>> of what I think is referred to as the 'chrome' - the
>> window borders, toolbars, etc.
>>
>> Is there a way to determine the height and width of the
>> chrome so that I can supply the correct parameters to the
>> resizeTo method?
>
> The dimension of the top and left chrome can be determined on browsers
> that expose the screen offsets of their window to scripting, given a
> mouse input event. By reading the mouse's screen co-ordinates, and its
> viewport co-ordinates and using the difference to give the screen
> co-ordinates of the viewport. The difference between the screen
> co-ordinates of the window and the screen co-ordinates of the viewport
> give the dimensions of the chrome at the top of the window and the
> chrome to the left.
>
> The screen co-ordinates of the window are not universally available in
> browser object models (and in framesets the values don't necessarily
> have the same meaning).
>
> The other proposed possible strategy for determining the chrome
> dimensions in total is to measure the viewport dimensions, re-size the
> window to the original viewport dimensions and then re-measure the
> viewport dimensions. The difference being the total dimensions of the
> chrome (left+right and top+bottom), assuming that the second set of
> viewport dimensions didn't drop to zero, which would invalidate the
> results.
>
> Obviously re-sizing the window a couple of times is not going to look
> that good to the user. And these days the ability to re-size a browser
> window is subject to user veto, so it will often not work at all.
>
> A better approach is probably to adopt a fluid HTML design that adapts
> to whatever size the user's window happens to be.
>
> Richard.