277099 - columnar layout in roc's blog renders badly in trunk [columns]

Status: RESOLVED INVALID

(Core :: Layout, defect)

Description

[Yusuf Goolamabbas]

User-Agent:       Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.8a6) Gecko/20050104 Firefox/1.0+
Build Identifier: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.8a6) Gecko/20050104 Firefox/1.0+

Hi, Visiting roc's blog in Firefox 1.0+ nightly I see the blog rendered in a
columnar fashion whereas visiting the url with Firefox 1.0, the blog renders as
intended

aebrahim on irc confirmed the same bad rendering with a Windows nightly build
He was using Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.8a6)
Gecko/20050101 Firefox/1.0+.
He sees the blog rendering in 4 columns at 1600x1200 whereas I see the blog
rendering as 3 columns at 1024x768 


Reproducible: Always

Steps to Reproduce:
1.Visit http://weblogs.mozillazine.org/roc/ with Firefox nightly
2. Observe the rendering
3.Visit above url with Firefox 1.0
4. Observe the rendering

Actual Results:  
Rendering is columnar in nightly build 

Expected Results:  
Rendering in nightly build should be same as Firefox 1.0

Comment 1

[Vidar Haarr (not reading bugmail)]

Looks a bit strange in Mozilla/5.0 (X11; U; Linux i686; nb-NO; rv:1.8a6)
Gecko/20050103 as well.

Moving to layout.

Comment 2

[Yusuf Goolamabbas]

moving to core/layout and asking for blocking 1.8a6

Comment 3

[Christian :Biesinger (don't email me, ping me on IRC)]

err, isn't this blog _supposed_ to render in columns, and ff 1.0 just does not
support that?

Comment 4

[Vidar Haarr (not reading bugmail)]

Yes, it is. I didn't say it shouldn't, I said it looked strange :)
I guess I should've read the initial description, I just commented from request
on IRC "if the site looked strange in your build or not".

The specific problem I saw was the image
<http://weblogs.mozillazine.org/roc/archives/BOI.jpg>, which overflows 60-70% to
the right.

Comment 5

[Yusuf Goolamabbas]

changing summary and specifically mentioning "columns" as requested by roc in 
https://bugzilla.mozilla.org/show_bug.cgi?id=251162#c54

Comment 6

[Simon Montagu :smontagu]

Attachment: Testcase with the image at the beginning of the article

(In reply to comment #4)

> The specific problem I saw was the image
> <http://weblogs.mozillazine.org/roc/archives/BOI.jpg>, which overflows 60-70%
to
> the right.

It overflows without columns also, if the viewport is narrower than ~800
pixels.

What seems more serious is that if you move the image to the beginning of the
<div class="columns">, text in subsequent columns is rendered on top of the
image, as in this testcase.

Comment 7

[Vidar Haarr (not reading bugmail)]

I just tested this in
<http://archive.mozilla.org/pub/mozilla/nightly/2004-10-30-07-trunk/>, and it
behaves - as far as I could see - just like todays build. So it isn't an
regression, at least.

Comment 8

[Boris Zbarsky [:bzbarsky]]

The behavior described in comment 4 and comment 6 is correct given the markup on
that page.

So what exactly is this bug about, if I may ask?

Comment 9

[Boris Zbarsky [:bzbarsky]]

I did find one issue with that page, which I've filed, with minimal-ish
testcase, as bug 277355.

Comment 10

[Simon Montagu :smontagu]

(In reply to comment #8)
> The behavior described in comment 4 and comment 6 is correct given the markup on
> that page.

Which behaviour described in comment 6 is correct? If you mean

> It overflows without columns also, if the viewport is narrower than ~800
> pixels.

then that's what I was saying myself. If you mean the next part

> if you move the image to the beginning of the
> <div class="columns">, text in subsequent columns is rendered on top of the
> image

then I'm very surprised to discover that that is correct behaviour. Can you explain?

Comment 11

[Boris Zbarsky [:bzbarsky]]

I mean both parts of the behavior.  The image is overflowing its parent, but the
overflowing parts should still be rendered.  The overflow setting on the image,
which I presume was meant to prevent this, is irrelevant here, since that
applies to the _children_ of the image (none in this case).

There don't seem to be any provisions for clipping overflow in columns or
otherwise handling it in the spec, but that's something to bring up with the
working group...

Comment 12

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

What bz said.