Snap intersection between guide and path

[pRototype]

I have a feature request that I hope to come true one day.

It would be very handy to have a snap function that make intersection between a guide and path being snapable.

Or to make the "Snap to path intersections" also make intersections between guide/path to be snapable.

I use Inkscape to depict physical problems, and that feature would for sure make some of the works easier.

Thanks.

[LucaDC (lucadc)]

Sincerely I'd like it too. But, to say the truth, I never asked because I fear that snapping targets would increase exponentially on my drawings (I use so many guides...) so whenever I need such intersections I just draw a segment on the guide (very easy with snapping nodes) and, after snapping to the newly created intersection, delete it (sometimes I recycle it a few times before deleting).
I find this solution quick enough for my needings.
Of course if this is going to be implemented without aggravating too much my poor Pentium IV, consider a +1 from me :)

[Diederik van Lierop (mail-diedenrezi)]

> I have a feature request that I hope to come true one day.

Sometimes that day comes sooner than you ever dared to dream of ;-)

Snapping to guide-path intersections has been implemented as of rev. #10639. I don't think that it will be much of an issue if many guides are used, because it's easily toggled of by either disabling snapping to guides or to path intersections. But then again, I don't know what LucaDC really means with "so many guides" :-). Just try it out and let me know whether you like it or not.

Snapping to intersections is still quite stressing for your CPU and you will quickly see a noticeable lag when many intersections need to be calculated. But that's a different story.

[su_v (suv-lp)]

Snapping to intersections (path-path or guide-path) only consider one intersection per segment: if one of the path segments (or guides) intersects the same segment repeatedly, only one of the intersections acts as active snap target.

Is this intentional (-> performance)?

[Diederik van Lierop (mail-diedenrezi)]

No, this is not intentional. But there's more going on here, which Nathan is helping me to debug. See

http://sourceforge.net/mailarchive/forum.php?forum_name=lib2geom-devel&max_rows=25&style=nested&viewmonth=201109&viewday=23

or

http://www.diedenrezi.nl/SingleSegmentPath.webm

Yes ~suv, this is your curve ;-)