69355 - scrolling = no and overflow:hidden both prevent ANY scrolling

Status: RESOLVED FIXED

(Core :: CSS Parsing and Computation, defect, P1)

Description

[rvj]

In Mozilla  'scrolling=no' or 'overflow:hidden' not only hides the
scrollbars but also disables scrolling 

I would also expect overflow:hidden to hide scrollbars but still allow 
scrolling (by keyboard). I would expect to use  scrolling=no if scrolling was 
to be completely disabled.?

Is this correct? IE still allows scrolling even if scrollbars are hidden ( and 
so once did Mozilla)


BTW

Does XUL support scrollBy, ScrollTo, pageYOffset, etc. If not what 
methods/properties are available for programatically scrolling the contents of 
XUL iframes?

Comment 1

[rvj]

Attachment: example of scrollbarless scrolling

Comment 2

[rvj]

PS was advised to assign to [e-mail] but assignment was rejected

Comment 3

[Boris Zbarsky [:bzbarsky]]

The CSS2 specification says for the overflow property:

hidden
  This value indicates that the content is clipped and that no scrolling
  mechanism should be provided to view the content outside the clipping region;
  users will not have access to clipped content. The size and shape of the
  clipping region is specified by the 'clip' property.

Note the "users will not have access to clipped content" part....

So it looks like we are doing the right thing here....

Comment 4

[rvj]

So I take it that under Mozilla it is not possible to have scripted page 
scrolling without scrollbars ? 

This seems a bit contradictory - ie  autoscrolling pages where the scrollbars 
have to be displayed even though they have no use?

I think under XUL I should be able to hide the scrollbars using XBL but then 
XUL seems to lack any scripting support for scrolling (scrollBy, scrollTo, 
pageYOffset, etc).

This appears to be a shortcoming in XUL iframes?? Should I file a bug on it?

PS 

Comment 5

[Keyser Sose]

Marking INVALID.

Comment 6

[Boris Zbarsky [:bzbarsky]]

*** Bug 85909 has been marked as a duplicate of this bug. ***

Comment 7

[Hixie (not reading bugmail)]

VERIFIED, although I think the DOM should allow for this, and I'll be working on
that for DOM3CSS.

Comment 8

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

Reopening.

Comment 9

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

Taking.

Comment 10

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

Attachment: patch

Comment 11

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

We should do this per the latest revisions to CSS2.1, in particular:

http://www.w3.org/TR/2003/WD-CSS21-20030915/visuren.html#q15
http://www.w3.org/TR/2003/WD-CSS21-20030915/visufx.html#overflow

Comment 12

[Boris Zbarsky [:bzbarsky]]

David, those sections don't seem to clarify the issue from CSS2 much... in
particular, the "users will not have access to clipped content" language remains
in the description of overflow:hidden.

The patch looks technically fine to me at first glance if we do want to allow
keyboard scrolling in overflow:hidden elements.

Comment 13

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

Oops.  That's a mistake.

Comment 14

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

The change in question is clear in
http://www.w3.org/Style/Group/css2-src/diffs-rec/visufx.html#overflow --
"mechanism" was changed to "user interface".

Comment 15

[Boris Zbarsky [:bzbarsky]]

Comment on attachment 130776 [details] [diff] [review]
patch

r+sr=bzbarsky, thoguh NS_STYLE_OVERFLOW_HIDDEN is now somewhat misleadningly
named....

The link you point to is member-only, but I'm assuming that the CSS2.1 spec
will be updated as you say (I'll add it to my list of last-call comments).

Comment 16

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

URL should have been
http://www.w3.org/Style/css2-updates/WD-CSS21-20030915-diff/visufx.html#overflow

Comment 17

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

Checked in, 2003-09-16 15:09 -0700, but leaving open to fix names of constants...

Comment 18

[R.K.Aa.]

So now the scrollbar is visible in the testcase - in other words it *is* given a
"User interface". It's just slightly difficult to use because it overrides my
drags after a short while, but to some extent that can be counteracted.
Mighty weird. And definatly NOT how for instance Opera works:
In Opera there is NO visible scrollbar, the content just scrolls on its own.
(attachment 25609 [details])

Comment 19

[[e-mail]]

I think this has caused a regression in some XUL I was using:
<hbox pack="end" style="overflow: hidden;">
now doesn't align right like it used to.

Comment 20

[[e-mail]]

Attachment: Simplified test case

Normally there's arbitrary content in the hbox.

Comment 21

[[e-mail]]

OK, so I need to use -moz-hidden-unscrollable in 1.6a and hidden pre-1.6a?
Will style="overflow: hidden; overflow: -moz-hidden-unscrollable;" work?

Comment 22

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

Yes, that would work, although really, I'd think in XUL you'd normally want to
just leave 'overflow' at its default, 'visible'.

Re comment 18: It looks like something strange is going on with IFRAMEs.  I'll
look into it...

Comment 23

[[e-mail]]

Unfortunately overflow: visible; has the nasty (for me) effect of making the
overflow visible...

Comment 24

[[e-mail]]

Oh, and the new property is a bit wordy, how about -moz-crop instead?

Comment 25

[[e-mail]]

Oh, and does overflow: hidden; make xul's <scrollbox> unnecessary?

Comment 26

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

comment 18 (regression specific to IFRAMEs) is now bug 220195

Comment 27

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

Also, the values for the property are all adjectives, not verbs.  -moz-cropped
or -moz-clipped might make sense (I prefer the latter), but both those terms
have other meanings in CSS, so I think I'd rather leave it as is.

Comment 28

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

This also caused "regression" bug 219693, since our percentage height quirks
don't work across scrollframes.

Comment 29

[Boris Zbarsky [:bzbarsky]]

*** Bug 220607 has been marked as a duplicate of this bug. ***

Comment 30

[Boris Zbarsky [:bzbarsky]]

Marking various regression bugs as blocking this bug.

Comment 31

[Jesse Ruderman]

This might have caused "regression" bug 220718, "mousewheel scrolls innerHTML
when it should not".

Comment 32

[Henrik Gemal]

this also seems to have caused that a cursor doesn't show up in the input field
here:
http://gemal.dk/test/mozbug.html

in build 20030916-04 it works
in build 20030917-04 it doesn't work

should I open a new bug. Not 100% sure why this bug is still open

Comment 33

[Peter van der Woude [:Peter6]]

This bug is marked FIXED but it isn't ?
The too technical lingo is a bit hard to read too.
Should an element with overflow:hidden be scrollable with keyboard arrows/mouse
or not ?
Meaning: overflow:hidden does nothing more than just remove the scrollbars?

Comment 34

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

Attachment: rename constants (checked in 2004-08-09 18:32 -0700)

This changes the meaning of the constants for mOverflow (SCROLLBARS_NONE ->
HIDDEN and HIDDEN -> CLIP) in the style struct while leaving them the same for
scrollbar styles.  This makes the changes in nsFrameFrame and nsGfxScrollFrame
a little interesting, and I removed the comment in nsGfxScrollFrame that I
think is wrong.

I tested the testcases on this bug, bug 220195, and bug 234851 (all 16
combinations on the dynamic testcase).

Comment 35

[Robert O'Callahan (:roc) (email my personal email if necessary)]

Comment on attachment 155220 [details] [diff] [review]
rename constants (checked in 2004-08-09 18:32 -0700)

-    // This isn't quite right. The scrollframe will still be scrollable using
keys.
-    // This can happen when HTML or BODY has propagated this style to the
viewport.

I think this comment is still correct. Consider

<html>
  <body style="overflow:-moz-hidden-unscrollable;">
  ...
  </body>
</html>

We'll have created a scrollframe at the viewport, no scrollbars will be shown,
but you can still scroll using keys (I suspect).

Comment 36

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

OK, I changed it to:

    // This isn't quite right (although the value is deprecated and not
    // very important). The scrollframe will still be scrollable using
    // keys.