Closed Bug 98694 Opened 23 years ago Closed 6 years ago

:hover pseudo class does not affect table column elements

Categories

(Core :: CSS Parsing and Computation, enhancement, P4)

enhancement

Tracking

()

RESOLVED WONTFIX
Future

People

(Reporter: sinchi, Assigned: glazou)

References

()

Details

(Keywords: css2, helpwanted, testcase)

Attachments

(2 files)

When :hover pseudoclas is defined to entire column of table, this column must be highlighted wher mouse is hover it, but didn't. See demo.
*** Bug 98695 has been marked as a duplicate of this bug. ***
Reporter I can't view the provided url to verify this report. Can you please double check that it is indeed correct and working. Also take a look at bug 5693. I would guess that your report is a dupe of that bug 5693. Thanks.
Summary: CSS2 pseudoclass :hover do not affect to <COL> element. → CSS2 pseudoclass :hover does not affect to <COL> element.
Sample URL (http://quaqua.chat.ru/bugzilla/col.html) is good, I have tested it just now. The bug #5693 is resembling, but it isn't same as this bug. tr:hover still works (buggy, but works), but col:hover works nowise.
Attached file Demonstration (HTML source code) (deleted) —
This is more related to bug 915 than bug 5693, although both, and probably more related to bug 915, would need to be fixed to get this to work. (i.e., a full fix for bug 915 and bug 5693 probably require some hacking in the event system too.) At least that's what I'd guess it would take to fix this. Giving this to glazman, owner of bug 915.
Assignee: dbaron → glazman
Status: UNCONFIRMED → NEW
Depends on: col-align-inherit, 5693
Ever confirmed: true
Summary: CSS2 pseudoclass :hover does not affect to <COL> element. → CSS2 :hover pseudoclass does not affect table column elements
Adding css2 keyword.
Keywords: css2
milstone bulk change
Target Milestone: --- → mozilla1.0.1
Actually, I'm not sure this is even a bug. If :hover applies to the element being hovered and all its content tree parents (which is our model -- see bug 5693), then it's not a bug.
Replying to 8: The main question is: "Does <td> element have dual inheritance relating <tr> and <col>?". It's being discussed in #915, and my humble opinion is: "Yes, it does, because logically table is 2D matrix with two peer dimensions". Maybe, it's very difficult to embody in program code, but on a theoretical plane, I think, that's it.
The real question, I think, is whether :hover is hierarchical from the top-most node under the pointer (which in the case of a table is typically the cell contents), or if it is hierarchical from all the nodes geometrically under the pointer, which I think is a bad idea since it gives misleading feedback about what a click will affect (a click only affects the top-most node).
Reconfirmed using FizzillaCFM/2002071208. Setting All/All. Is this a duplicate of the column style resolution bug?
OS: Windows 2000 → All
Hardware: PC → All
No. It's related to: * bug 5693, except the eventual fix decided that doing :hover in a way that would allow this to work would be incorrect. See bug 5693 comment 7, bug 5693 comment 8, bug 5693 comment 9, bug 5693 comment 36, etc. * bug 4510.
(I had said in the past that it was related to bug 915, but that's wrong for the testcase, since the testcase sets 'background-color'. However, if you interpret the bug as being more general than just the testcase, then it could be related to bug 915.)
Depends on: 4510
So... what _should_ we be doing here? I personally feel that there is no reasonable way that hovering a <td> can make the <col> hovered if we're doing hierarchical :hover (which we are).
Keywords: qawanted
No longer depends on: col-align-inherit
retargeting
Target Milestone: mozilla1.0.1 → Future
Target Milestone: Future → mozilla1.0.1
Keywords: testcase
Severity: normal → enhancement
Keywords: helpwanted
Summary: CSS2 :hover pseudoclass does not affect table column elements → :hover pseudo class does not affect table column elements
Target Milestone: mozilla1.0.1 → Future
I ran into this problem too in January 2005 ..... The hover effect would be very useful on large (time)tables to allow a crosshair to better find a location. From my understanding <col> should work the same as a <tr>. So a <colgroup> is the same as a <tbody>.
Btw.: http://www.w3.org/TR/html4/struct/tables.html#h-11.2.4 11.2.4 Column groups: the COLGROUP and COL elements Column groups allow authors to create structural divisions within a table. Authors may highlight this structure through style sheets or HTML attributes.
Still waiting...
Feel free to propose a :hover model that makes sense while allowing what this bug asks for. The current model (:hover applied to topmost node under pointer and all its ancestors) doesn't allow hover on columns, since they have no descendants.
And note that if :hover does apply to table column elements, it has to be based on the semantic table structure, not the CSS display types, which means this basically depends on bug 915.
Depends on: col-align-inherit
QA Contact: ian → style-system
This probably depends more closely to bug 76481 too. Because of how the event is going upstream from the element under the mouse pointer, <col> and <colgroup> don't get a look in (because despite a genuine need for a faux DOM to cross link rows and columns, one is not in any specs and thus does not exist). I don't know how sane a solution it would be to change the concept of what is under the mouse pointer to include a extra depth dimension in the sense of layering (eg. a div behind another div) for triggering on multiple elements before passing the event up the DOM. In the CSS spec (which is relevant more so here than the structure itself, being :hover after all), the table layout is defined as layers that effect other elements despite the missing links in the DOM structure. So it's expected that it not be a limitation of the event handling.
Priority: -- → P4
Still an issue with the attached testcases on: Mozilla/5.0 (X11; Linux i686; rv:25.0) Gecko/20130728 Firefox/25.0 I haven't seen this supported on other browsers either. Considering this and the above comments, shouldn't it be a wontfix?
Keywords: qawanted
Just like bug 915, it is indeed a wontfix
Status: NEW → RESOLVED
Closed: 6 years ago
Resolution: --- → WONTFIX
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Creator:
Created:
Updated:
Size: