User-Agent: Build Identifier: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.6) Gecko/20040113 The bottom panel of the referenced web site is styled in CSS. It is positioned via the following CSS properties: bottom: 0px; left: 0px; right: 0px; position: absolute; margin: 0px 0px 0px 0px; This worked fine in Internet Explorer 6 and Mozilla 1.5; however behaves rather interestingly in Mozilla 1.6, oftentimes positioning the panel below where it ought to be (very bottom of page). Reproducible: Always Steps to Reproduce: 1. Visit the web site.

Regarding the .css file http://justinarthur.deepstructure.info/index.css ------------------------------------------------ Setting left and right offset values to an abs. pos. element can not be honored simultanously for sure. .footpane { width: 100%; bottom: 0px; left: 0px; right: 0px; position: absolute; /* If the goal is to position the div at the bottom of the viewport, then position:fixed is the way to go; see bug 105286 */ Declaring width: 100%; height: 31px; border: 2px solid Gray; will create an overflow since such width value does not take into consideration the padding, border and margin value. So, the element uses 100% of the available width as given by its immediate parent (which is the Initial C.B.) and adds borders around it. Such overflow will bring an horiz. scrollbar. font-size: xx-small; /* xx-small is very small font-size value and can be overriden by user settings in Edit/Preferences.../Appearance/Fonts/Minimum font size. Now, if overridden, then the text will wrap and the height of the cell will be greater than 31px, again bringing another issue */ I examined carefully the .css file http://justinarthur.deepstructure.info/index.css and I am convinced that at least 9 css declarations are pointless or irrelevant (and should have been removed) while 3 are contradictory or wrong or causing issues. One point I was not able to understand is what this declaration (found in html.css) table {-moz-box-sizing: border-box;} does exactly. The .css file http://justinarthur.deepstructure.info/index.css is over 80 lines and can not serve to enlighten, to corner precisely a bug. A reduced testcase would be necessary here. In any case, I don't understand how or why one should link this bug to bug 232838: there is no explanation, no demonstration, no reasoning given to that effect. Scrollbar or scroll is not even mentioned anywhere in the summary or description or actual results of this bugfile. FWIW, I set/changed position to fixed and removed the 2px gray borders around the footpane table and everything behave accordingly: the footpane div positioned itself at the bottom of the viewport. Mozilla 1.7alpha build 2004021508 under XP Pro SP1 here.

Points taken. It is quite possible that a bug in Mozilla 1.5 and IE6 actually allows this to render as the author had intended.

No, there's a real bug here, and it has something to do with not repositioning elements relative to the bottom when the scrollbars are added.

Note that comment 1 does explain why there's horizontal overflow (although it's worth noting that specifying both 'right' and 'left' can make sense when 'width' is 'auto' (unspecified)). However, it doesn't explain why we sometimes position the absolutely positioned element above the scrollbar and sometimes under it.

(In reply to comment #4) > (...) However, it doesn't explain why we sometimes position > the absolutely positioned element above the scrollbar and sometimes under it. In comment #1, I should have been saying (...) uses 100% of the available width as given by its **offsetParent** or **nearest containing positioned element** (which is the Initial C.B.)(...) instead of **immediate parent** I wish you could direct me to some document explaining what do/are these declarations: *|*::-moz-viewport-scroll { overflow: auto; } and table {-moz-box-sizing: border-box;} As far as I can say right now, this bugfile seems like a DUPLICATE of bug 211562.

In this bug, zooming makes the problem go away, but in bug 211562 it does not. Please don't confuse bugs with unrelated issues.

I hope the HTML I just attached serves as a cleaner test case. Sorry about putting you through the mess of my personal web site in progress before. :) Waiting for confirmation, and affected operating system clarification.

Isn't bug 234393 actually a duplicate of bug 211562? The last line in the testcase is really obscured. There is no vertical scrollbar. But you can scroll the page using the arrow keys. That is interesting. I can confirm this with Firebird 20040128. It was working in Moz 1.5.

Win 98/2k.

Changes to bug 211562 descriptions have rendered this bug a dupe of it.

*** This bug has been marked as a duplicate of 211562 ***