Re: Confused about diagram page sizing

[Cyrille Chepelov <cyrille chepelov org>]

Le Mon, Mar 18, 2002, à 10:01:39AM -0800, Rick Horowitz a écrit:

> 1. It would appear from dia's doc that the blue lines
> on a new diagram constitute page boundaries. Why are
> the pages so small on the screen? Are these really
> page boundaries?

EPS output ignores page boundaries. Dia units are centimetres, and you can
check that by default, you indeed have 21 cm between vertical page break
lines minus the left and right margins 
(or the equivalent metric measure for the width of whatever paper format you
happen to be using).

The PS, not EPS renderer will honour page breaks. You can specify a scaling
factor in the print properties dialog.

The specific scaling factor between dia's centimetres and actual size is
more or less an arbitrary constant (perhaps .5 if X knows the actual dpi
value of your screen)


> 2. Why are the pages at the top of the screen view
> (after scrolling to the top) only half pages? I find
> this confusing. Should I draw into these half pages,
> or not?

You can. You'll be able to scroll upwards using the scrolling tool, or when
there is an object high enough there. The only "strange" thing is that the
the space in this "half"-page and above is with negative Y coordinates, but
this is about all (there is absolutely no problem with drawing in Y- space).

> 3. Why does the screen view grow ever larger?  For
> example, I drew a diagram into a page (within a
> blue-lined box) and then copy/pasted the diagram into
> the upper left hand corner (full page) in the screen,
> only to find that the screen view had grown larger and
> that my copied diagram was no longer in the upper left
> corner. What exactly is going on?

This has been fixed in CVS.

> 4. If I were to print from a diagram screen, would I
> get a printed page for each "page" on the dia diagram
> that I've drawn?  How can I get rid of all these
> excess empty dia pages?

By writing some logic to eliminate empty pages. For the moment, we have no
way of determining whether a page is empty or useful (but it shouldn't be
too difficult to implement that functionality using each object's bounding
box. For not really rectangular objects, this would still spew out an
occasional empty page, but woud still be better than what we have now.)

Thanks for the comments !

	-- Cyrille

-- 
Grumpf.