On Example 0, the long line ("This line demonstrates the :first-line pseudo element") is too wide and ends up getting drawn behind the scrollbar; apparently the scrollbar width isn't getting passed to layout to decrease the canvas width accordingly.

Added me to cc list. i will look into this.

i think this is somewhat bogus. the device context *does* pass back scrollbar sizes, but they may not be entirely accurate (so layout may disagree with the area actually occupied by the scrollbar).

Inserting Milestone info.

Setting all current Open/Normal to M4.

does this happen under win32? I don't see any good reason why this is happening.

cpratt, could you check this problems with latest build on Win32, Mac and Linux. Letus know what you find. Thanks!

I'm not seeing this problem on example 0 in win32, however there is a similar sounding problem on http://www.bath.ac.uk/%7Epy8ieh/internet/eviltests/htmlbodyrendering6.html ...where the inner scroll bars are being added *after* the inner width is decided, and so horizontal scrolling is required when it shouldn't be.

BTW, I am pretty sure this bug I am seeing (reported 02/24/99 13:30 above) is not a widget set bug but a layout bug.

Waiting for a stable seamonkey build to test this against; bear with me...

i know the scrollbar width is passed to the layout, so this is most definatly a layout bug...

Bug also occurs on Win98 with the first scrollable DIV of http://www.bath.ac.uk/%7Epy8ieh/internet/eviltests/floatgrowth.html Changing OS from Linux to All. I believe this is probably a layout bug too. We should probably remove the [PP] radar, as this appears to be cross platform.

m5

m6

Kevin, Since this bug is now an XP thing, can you take a quick look ? It might be bogus. thanks. Reassign to kmcclusk@netscape.com

I don't see any problem in 5-19-99 10:00am build on WIN32. with Example 0. http://www.bath.ac.uk/%7Epy8ieh/internet/eviltests/htmlbodyrendering6.html, does display horizontal scrollbars on WIN32, so I am leaving this bug open. I am changing to component to Layout, and reassigning the bug to a Layout person Rick, I am reassigning to you because it looks like a layout problem.

Removing [PP] radar as per comments above.

Hey Chris. Can you please tell me if you still see this on our tier1 platforms?

Tested using 5/19 Apprunner on Tier 1 platforms (Win32, Win98, Win NT4.0 - Win NT5.0 not available). In the resource:/res/samples/test0.html test, the long line is NOT drawn beyond the scrollbar - all platforms tested are okay In the http://www.bath.ac.uk/%7Epy8ieh/internet/eviltests/htmlbodyrendering6.html test, there are scroll bars as indicated by the overflow: scroll value but they are disabled as they should be because the BODY is sufficiently contained within the HTML element - all platforms tested are okay WORKSFORME.

Agreed WORKSFORME on Win98. What about other platforms?

Looks like this problem no longer exists on Linux.

Marking WORKSFORME since it appears to now work on every platform tested with every test page tested.