Node Graph UI discussion

14 May 2007 - 8:56am
8 years ago
7 replies
2824 reads
Forrest Maready

Has anyone here designed or is designing a node graph/dependency graph
UI? I'd love to have some discussions off list about your experience and
bounce some ideas off you.
One point of contention- vertical flow is more appropriate for the
metaphor, but it makes naming channels difficult. Is there any consensus
about whether writing things vertically:


as oppose to "Channel" rotated 90 degrees clockwise is easier to read?



14 May 2007 - 9:10am
Jack L. Moffett


It is a long-standing rule of typography that text should not be
stacked. Stacked text is much more difficult to read than rotated
text. Letters are designed specifically to be placed side-by-side.


14 May 2007 - 9:46am
Shep McKee

Not just a long-standing rule of typography - but also supported by
empirical evidence:
"horizontal text is fastest, followed by both rotated conditions,
followed by marquee [stacked]"
"marquee text took almost twice as long to read as the standard
horizontal text."

Reading Vertical Text: Rotated Vs. Marquee.

On May 14, 2007, at 10:10 AM, Jack Moffett wrote:
> It is a long-standing rule of typography that text should not be
> stacked.

14 May 2007 - 10:06am
Jeff Howard

A good example of this is the type on book spines. They're meant to
be read vertically and you'll almost never see stacked type.

As to the difference between clockwise and counterclockwise rotation,
American audiences are more familiar with text rotated clockwise 90
degrees, but that doesn't necessarily hold true for other countries.
I was taught that British titles orient the text the other way to
optimize for reading sideways on a shelf. (American titles opted for
spine text that's easier to read when the book is resting face-up on
a table.)

// jeff

14 May 2007 - 12:41pm
leo.frishberg a...

In a recent design for such rotated text, we determined that both
rotations were appropriate for on-screen reading, depending on which way
the user would have to turn their head. So, if rotated text appeared on
the right side of the screen, the natural direction to rotate it would
be counter-clockwise; when on the left, clockwise.

Any literature supporting our empirical "swag" would be appreciated.


14 May 2007 - 1:02pm
Jack L. Moffett


That's interesting. I thought you were going to say the exact
opposite. Putting text on the left and rotating it counter-clockwise
would allow for a slight tilt of the head that would line the eyes up
enough for left-to-right (bottom-to-top) reading, and vice versa. It
seems slightly more awkward the other way, like I have to tilt my
head farther.

In any case, I doubt that the difference would be all that
significant. Unless it is more than a few words (in which case it
shouldn't be rotated), it can be read easily without tilting your
head. The tilt is more a natural reaction than a necessity.

Of course, I can't back up any of this with empirical evidence. There
goes my gut again. ;)


14 May 2007 - 5:14pm
Vijay Venkatraman

Hmm, yea, I would have thought that Leo would say the exact opposite too i.e.
text on the right side of the screen would be rotated clockwise, while that
on the left side would be rotated counter-clockwise. Leo, maybe this is an
inadvertent mistake in your post?

- Vijay

14 May 2007 - 6:45pm
leo.frishberg a...

I sit corrected.

In spite of my best efforts to tell my right from my left and counter
clockwise from clockwise I screwed it up.

Right hand side text rotated clockwise / left hand side text rotated
counter clockwise.

Thanks all.


Syndicate content Get the feed