[e-lang] Vat figure
e-lang at zesty.ca
Sat Jun 18 12:32:11 EDT 2005
On Sat, 18 Jun 2005, Mark Miller wrote:
> I agree that we need a figure that shows the single-vat situation before
> showing the multi-vat one. However, I don't think it's worth the paper
> real-estate (much more precious than web-page real estate) to split the first
> into 1a and 1b. In your figure, when I turn on only "Queue Outline", "Queue
> Label", "Stack Labels", "Objects", "Queue Frames", and "Frames", the result
> looks like a good combined illustration.
Using only these layers makes the boundaries between the three parts
of the vat difficult or impossible to distinguish and largely cancels
the usefulness of the figure. If there are no fills, it is essential
to compare 1a and 1b so you can tell which parts of the box the labels
refer to. If you combine both parts into one figure, the fills are
essential for the same reason. The clearest thing to do is to include
the fills and both figures. Failing that, a single figure with fills
is better than two figures with no fills.
I'm curious what the others on this list think. The figure reduced
to black-and-white can be seen at
The figure as I'd originally envisioned it can be seen at
> What is "Bounding Box" for?
"Bounding Box" ensures a fixed bounding box for both versions of the
> I turned off "Queue Fill" and "Fills" because
> Springer prefers straight black-and-white. (Had I stuck with my original
> drawing, I would have removed all gray scale as well.)
It seems more important that the figures be comprehensible than that
they meet some arbitrary stone-age conditions. Does Springer have an
absolute requirement for black and white only? I'd find that hard to
believe, since I've submitted in Springer format before, and I can't
imagine that all photographs and screen shots are forbidden.
> Flipping the arrows is fine. I like the idea of having the big figure read
> To ease the transition from the small figure to the big one,
> you may want to try illustrating the single-vat version of the same example
> within the small figure.
I'd rather not. The whole point of the doing the small figure is to
> > I would like to relabel the entities in the big figure as well, which
> > would mean also relabeling them in the text. Instead of calling them
> > SH, AM, and XL, I would like to use concrete names such as Account,
> > Manager, and Display respectively. If feasible I would even like to
> > use method names "setBalance" and "balanceChanged" instead of "setStatus"
> > and "statusChanged". What do you think?
> I would prefer that you not do these relabellings.
Can you elaborate on this? Do you believe that "SH", "AM", and "XL" are
easier to understand than "Account", "Manager", and "Display"? Or is it
just the method relabellings that you would prefer to avoid?
More information about the e-lang