Hello. IIUC lib2geom is a work-in-progress. However it is not clear to me how much of Inkscape is ported to lib2geom. (I think it is mentioned that a large part of libnr being yet needs to be ported.)

However I am eager to use some of the better features of lib2geom, especially the following one mentioned here:

Path algebra:
Computations such as offset curves can be written with their mathematical definition and still get a bounded error, efficient curve. (preliminary trials indicate offset done this way out performs the method used in Inkscape)

Are there nightlies or such for Inkscape which include these features so I can test them?

Thank you!

jamadagni wrote:IIUC lib2geom is a work-in-progress. However it is not clear to me how much of Inkscape is ported to lib2geom.

Inkscape (stable and trunk) currently uses a private (AFAIU partial) copy of lib2geom: changes to that private copy in inkscape's source base are backported to the main lib2geom source (or done there first preferably).

Upstream changes in lib2geom are merged into Inkscape trunk from time to time (not on a scheduled basis).

jamadagni wrote:(I think it is mentioned that a large part of libnr being yet needs to be ported.)

In current trunk, libnr is already gone (aka converted to 2geom) - the internal library had been removed completely in r10589 last fall.

As mentioned in the comments of the blueprint, a bigger task still not even started is converting livarot to 2geom (livarot is used for example in the tweak tool or for the (boolean) path operations).

jamadagni wrote:However I am eager to use some of the better features of lib2geom, especially the following one mentioned here:

Path algebra:
Computations such as offset curves can be written with their mathematical definition and still get a bounded error, efficient curve. (preliminary trials indicate offset done this way out performs the method used in Inkscape)

Are there nightlies or such for Inkscape which include these features so I can test them?

AFAIK Inkscape's path operations like inset/outset/offset are based on livarot and even the latest nightlies won't use 2geom for those computations.

jamadagni wrote:here:

Path algebra:
Computations such as offset curves can be written with their mathematical definition and still get a bounded error, efficient curve. (preliminary trials indicate offset done this way out performs the method used in Inkscape)

Hello and thanks for your detailed response. As I have quoted above, the lib2geom page says that it does Path to Outline better than Inkscape (which means Livarot, as per this info). How does one test it? I mean, if I have to dig up the internals of my SVG file and get the path coords and feed them into lib2geom in some other format, I'll pass and wait for actual intergration into Inkscape to be done. If there is some other way of getting my Inkscape SVGs with paths converted into outlines using the promoted new better algorithm of lib2geom, I'd like to test/use it!