GIMP Version: 2.6.8
Operating System: Linux
OS Version: Kubuntu 10.04
GIMP Experience: Advanced Level

---

1) When you drag the "handles" on a control points, you can force them to be aligned and symmetrical by depressing Shift when you drag. But is there a key combo to have them aligned but not symmetrical?

2) If you mark create a path with straight lines with a few plain clicks, you can later obtain Bezier curve handles on them by Ctrl-click. But is there a way to predict which handle you will get (ie, if you need to drag to the right or the left...)

3) I sometimes end up with two points that are linked (they move together if I drag one of them). How do I get there (in case it's useful someday)?

When clicking an anchor point with the Ctrl key down, my experience is that the first handle goes which ever way you move it (while still holding down the Ctrl key). The other handle (while still holding down the Ctrl key) you can move in another direction.

Of course, if you want them straightened, you use the Shift key, which you already know.

I hope I understood your question correctly.

1) Not to my knowledge.

2) It always seems to be the "from" (or "into") control point which is accessed first. That is, given a polygonal path from A to B to C, clicking on B while holding the CTRL key will select B's handle which directs "from A" towards B.

3) You can select multiple anchors by holding down SHIFT while clicking on an anchor. All anchors in such a selected group move in concert (along with their associated control handles).

Hmm. This is not what I get... if I click three points form left to right, and then Control-drag the middle point, I seem to always get the "back" handle:

Dragging right:


Dragging left:


And if I click the points right to left, I of course get the opposite behavior.

Strange.

Too bad. But at least it's not my oversight

This seems coherent with what I get in well controlled conditions. Unfortunately Gimp path directionality isn't always preserved during path processing...

That explains it, thanks.

I am certainly not disputing this claim; but if you are aware of a specific scenario which may corrupt the directionality of paths, I would appreciate if you let me know.

Speaking of - anyone know of a script to reverse path direction? I could have used one the other day and couldn't locate one in a quick search.

-Rob A>

Reverse path or reverse only some stroke?

I believe the following code works (though I have not performed extensive testing).

Hey Saulgoode, are you planning to add this to your scripts site?

Just done something equivalent (in Python of course, and without the stroke selection). A catch is that this doesn't preserve the order of the strokes, and this may be important (stroking the path with a gradient, for instance). The code would have to copy all strokes, reversing some in the process (and even then I have the hazy memory that you get them in the reverse order, at least with the python API...)

reverse-strokes.py added to my SourceForge collection. You can select which strokes to reverse by surrounding them in a selection (actually the script only tests that at least one end point is in the selection).