[thelist] Mac/IE layer positioning issue

JTocher janice at discoverysystems.com
Mon Oct 1 15:24:05 CDT 2001


Bill and other layer gurus...

Having a hard time even getting a *very* simple layer to work cross
platform/cross browser.

See the following:

http://averra.com/mstest/testing.htm

The dot should fall in the middle of the square - have tried adding tables,
removing tables, adding a 'layer' tag for Netscape changing the 'div' tag to
a span and nothing seems to give the cross platform/browser results.

Am beginning to think common code is a lost cause and that will have to use
two different stylesheets. Although, Bill, you mentioned that you were able
to get your page to work across all browsers and all platforms - howja do
it???


thx...
Janice


on 10/1/01 2:06 PM, you wrote:

> Just getting back to my Email now...
> 
> Actually, I think this was the "well-known" (to everyone but me, anyway!)
> Mac bug. It was coming from 0,0 of the table, not the page.
> 
> The function was being called from the first cell in a nested table. As a
> result, the script dropped divs at absolute positions, and the position was
> counted from 0,0 of the NESTED table in which the function was called.
> 
> The solution was to push the function call up to the first line of the body.
> Then it counted from 0,0 of the page as expected. Lesson learned.
> 
> Took me forever to figure this out, only because it was only occurring only
> in IE5 for Mac, and no others. I was not aware of this bug.
> 
> BH
> 
> 
>>> Never mind - figured it out. IE on the Mac was the only browser that was
>>> counting pixels from the point where the function was called
>> (i.e. inside
>>> the table) instead of from 0,0 of the window. I did not know this would
>>> happen, but to fix it, I moved the function call to the first
>> line of the
>>> body and it then worked just fine.
>> 
>> ??? Could you please explain more? There is a well-known IE5Mac bug that
>> causes absolutely positioned elements placed in a table to be
>> misplaced. I
>> think (but have not tested) that their position is counted from the 0 0
>> coordinate of the table, not of the document, as should happen with
>> position: absolute
> 
> 
> ---------------------------------------
> For unsubscribe and other options, including
> the Tip Harvester and archive of TheList go to:
> http://lists.evolt.org Workers of the Web, evolt ! 






More information about the thelist mailing list