Re: [css-position] sticky only along one axis

On Mar 18, 2014, at 12:37 PM, Viatcheslav Ostapenko <sl.ostapenko@samsung.com> wrote:
> 
> On Mon, 2014-03-17 at 14:46 -0700, Brad Kemper wrote:
> 
>>>>> It would be useful but it would be very difficult to spec and
>>>>> implement and would probably create more trouble than it's worth.
>>>> 
>>>> How about this:
>>> 
>>>> Positioning on table-column and table-column-group items affect their
>>>> corresponding table-cell items as though those table-cell items were
>>>> selected in the same rule. Table-cell items spanning more than one
>>>> column (as with HTML's 'colspan') are only selected by
>>>> table-column/table-column-group when they originate in a cell starting
>>>> in the column(s) of the selected table-column/table-column-group.
>>> 
>>> 
>>> I would agree with Robert. It is difficult to implement and will be used
>>> less often than sticky table headers, for example.
>>> Sticky columns are useful for wide tables, which are less common than
>>> tall tables.
>>> Taking into account that sticky columns could be easily workaround by
>>> assigning position:sticky to every corresponding table cell, I would
>>> suggest to disallow position:sticky for table columns.
>> 
>> What makes it hard? Don't TDs already look to COLs for style resolution?
> 
> Let's say this way:
> Implementation of sticky positioning for table columns would bring extra
> complexity, but it will be rarely used. 

Why assume if would be rare? Headings on the left or right side of tables are not that rare. 

When designing for mobile, having a non-scrolling row header column could be pretty useful. Or having a column of action-oriented icons or checkboxes that hugs the right side as you scroll the rows to the right to see the other columns. 

By the way, it isn't just sticky positioning that is discriminating against columns but not rows. It is other positioning in the draft too. 

> Taking into account that there
> is simple workaround I don't think it's worth the trouble.

You could say the same thing about table rows. I don't think we should sacrifice useful consistency just because it requires a little extra effort.

Received on Tuesday, 18 March 2014 23:36:53 UTC