Re: [css3-multicol] How to avoid scrolling up for next column

On Sun, Aug 22, 2010 at 10:43:57AM -0700, Brad Kemper wrote:
> 
> On Aug 22, 2010, at 1:40 AM, Peter Moulder wrote:
> 
> > Regarding column width, one could of course specify column width in ems,
> > but we found that it's best if there are an integer number of columns
> > per viewport: after all, there's pretty much no value for just a partial column
> > to be visible, so rather than show N½ columns you might as well show N or N+1.
> 
> It seems to me that for horizontal scrolling, it might be good to have a
> partial column showing, as a visual indication that there is more there. This
> is especially helpful on a device like the iPad, where there is no scroll
> indicator visible until after you start scrolling, and where scrolling an
> area like that will end up showing partial columns anyway until you get to
> the end (where, thus, the perfect alignment of the right side of the column
> is an indication that you've reached the end).

If "partial column visible" is being used as an indication of whether further
scrolling is possible, then I think much the same principle applies: you want
to make sure that the column width is such that a bit of the next column is
visible, and you might as well make it so that only a bit is visible (and avoid
having the column be almost all visible but just narrow enough that the word
"a" or the plural "s" is hidden on some line).

That said, I'd have thought that you'd want a better indicator than partially
showing whatever's in the next column: I agree that it's a very intuitive
indication, but I'm concerned that it would be distracting if you see part of
an image or just the first word or so of each line of the next column.  (It's
also unreliable, in that how much of an indication it is depends on the content
of that next column: if it contains only diagrams, then the visible part might
not clearly indicate that there's more; though I wouldn't expect this to be a
common problem.)  However, this isn't something I've looked into.

The user interface we used had an extra indication of where you are in the
document, that we added in response to a problem we found in the pilot study:
we showed a panner / thumbnail view / pager / miniature view of the whole
document, highlighting the screen's current view within that document.  I was a
bit uncomfortable adding this experimental user interface element to the study,
but judging from what we saw in the (small) pilot study compared to the (still
not particularly large) main study, I'd say that this was an improvement for
our software and setting.  As to whether or not such a widget would be useful
in other software, I think part of the reason it worked in our case is that a
multicolumn view wasn't just something one encountered on one in a thousand web
pages, it was used in half of the documents that the participants saw.  I don't
know how well horizontal scrolling and multi-column can work in general web
pages, but I think it can work in more specialized contexts such as e-book
readers, where it's feasible for the user rather than the author to decide
whether to use multiple columns and whether to use multiple individual pages or
continuous horizontal scrolling of a single very wide page.


> > Something else we've found when playing with multi-column layouts (but which we
> > haven't looked into carefully) is that the ideal space between columns is
> > actually sub-linear in font size: i.e. if you make the font size k% larger then
> > the gap between columns should be larger too but by less than k%.  I haven't
> 
> I'm sorry, but I didn't understand that last sentence, and didn't find
> anything in the study about gap size,

Yes, that wasn't part of the study, sorry I wasn't clear enough about that.
(I tried to convey it with the "when playing with" and "haven't looked into
carefully", but I agree on re-reading that it still isn't clear.)

What I mean is that the right column gap width for 15px text seems to be
somewhere strictly between 1 and 1.5 times the right column gap width for 10px
text.  The implication is that neither px nor em is ideal for specifying
column gap width if users are free to change font size.

In any case, my impression is that the current recommendation of 1em seems too
small unless a column rule is also used.

As I say, this isn't something I've looked at methodically, it's just based on
my own judgement of what column gap width looks good for a given font size, and
even then I haven't spent long trying different font sizes and column gap
widths.

I thought this worth posting to www-style because it seems to have
ramifications for the 'column-gap' property, at least for the recommendation
of what to do for the value 'normal'.

pjrm.

Received on Sunday, 22 August 2010 23:18:50 UTC