[Webkit-unassigned] [Bug 79128] Fieldset unexpectedly stretches to minimum intrinsic width.

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=79128





--- Comment #9 from Julien Chaffraix <jchaffraix at webkit.org>  2012-03-18 21:06:41 PST ---
(From update of attachment 132455)
View in context: https://bugs.webkit.org/attachment.cgi?id=132455&action=review

> Source/WebCore/ChangeLog:9
> +        in css before stretching to minimum intrinsic width.

Your explanation about us matching IE and not FF sounds reasonable to be put here in your ChangeLog (note also the FF acknowledged the broken behavior).

> Source/WebCore/rendering/RenderFieldset.cpp:201
> +    if (style()->width().isSpecified())

After this change, there is no guarantee a web author cannot manually shrink a fieldset below its intrinsic size. There is some warnings on the Mozilla bug about the potential compatibility risk about that: https://bugzilla.mozilla.org/show_bug.cgi?id=504622

I don't know fieldsets enough to know how big of a risk this is.

> Source/WebCore/rendering/RenderFieldset.h:45
> +    virtual bool stretchesToMinIntrinsicLogicalWidth() const; 

Extra-space at the end of the line.

> LayoutTests/fast/forms/fieldset-width-nostretch-ifspecified.html:5
> +                width: 50%;

Length::isSpecified means 3 values:
* percent length
* fixed length
* calculated length

I would like to see a test for at least the 2 first.

> LayoutTests/fast/forms/fieldset-width-nostretch-ifspecified.html:19
> +            <div class="constrain">
> +                <h1>div</h1>
> +                <p>This is a div with width set to 50%</p>
> +            </div>
> +            <fieldset class="constrain">
> +                <h1>fieldset</h1>
> +                <p>This is a fieldset with width set to 50%</p>
> +            </fieldset>

This looks like a good candidate for a ref tests as the <div> and the <fieldset> should behave the same.

-- 
Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.