hi greg > >> It might be more of a feature request than a bug. For example, if > >> you took a star shape or circle and converted to a text frame, you > >> wouldn't necessarily expect star-shaped or circular Text > >> Distances. > > > > euh, i would... > > > > i don't see in what context, calculating the distance from the > > bounding box could be useful... > > > > The answer to that is easy, in a non-skewed rectangle. > The text distances have a limited application. I don't see the sense > in wasting the programming time to cover all instances. You have to > have an entirely different concept. For a circle or a star shape, > what is Top, Bottom, Left, and Right? > Instructing people to use a second frame or perhaps the contour line > makes a lot more sense. > > If there is some answer to it, I think it would come from changing the > mind set of the contour line. If you look at an SLA file in 1.5.x, we > now have PATH and COPATH variables, which I take to describe the shape > of the frame and contour line. A contour line will do everything Text > Distances now does and much more. One of the things to overcome is > that the contour line editing is so far hidden that you don't expect > users to find it on their own. > > Another idea: why not, instead of these text distances, be able to > shrink the frame shape while leaving the contour line alone? This is > in effect what text distances are doing.
while i agree with you, that in most cases having two (or multiple frames is a better idea, if the feature exists and it does not work as people expects, then it's a bug. and i'm not sure that -- as soon as you leave the comfort zone of the rectangular frame -- the form of the shape has any consequences on the calculation... you just have to calculate it correctly, which is not trivial. ciao a.l.e p.s.: i'm all for removing half baked features that are hard to use and not really essential to scribus... but i'm really in the minority here...
