Hi

I have had a long standing (at least since version 0.5.1) difficulty with point label text alignment.

(Currently using 5.2.7)

It seems that if rotate = 0 and the alignment of the scrap in the map editor matches (more or less) the alignment of the scrap in pdf outputs, then commands such as  -align right or -align tl are honoured.

If however rotate = 70 (for example) or the scrap alignment in the map editor is significantly different to the alignment of the scrap in the pdf outputs then the –align parameters are incorrectly (in my view) applied.

The effect of this is to require the user to align scraps and xvi files to match the finished pdf drawing output if text alignment is to be readily controllable.  It also makes it nearly impossible to produce a scrap that generates nice labels where it is desired to produce sets of pdf outputs with various orientations.

In the attachment I have put together a dataset that starts to demonstrate characteristics of the problem.

I have superposed survey station points and point labels so that the alignment can be observed – a clear pattern is observable.

You can see that the alignment is based on the orientation in the map editor, not the pdf output.

I have added some text that has orientation and applied some distortion to the scrap by misaligning the survey stations, so you can see (another reason)  why ‘orientation’ of point labels is not a good idea – scrap distortions misalign the text a little.

While it may seem trivial in this example, on a real map with labels (perhaps at the end of arrows) it tends to make a real mess that can only be resolved by iteration and can not usually be resolved if two or more distinct ‘rotations’ are required without changing all the point labels positions.  Not a feasible proposition.

Has anyone else noticed or had difficulty with this behaviour?

Is it likely that it can be altered?

Regards

Bruce