GitLab

[PATCH #189] This arbitrary choosen value was obviously due to the already closed bug #189. We do not need this obscure stuff anymore, as everything is working fine.

[PATCH #190] This commit closes #190, as in WValueSet the minimum was initialized with MAX_DOUBLE which lead to the invalid data. Instead we may use std::numeric_limits< T > and every thing is fine.

[CHANGE] version string changed to conform better to debians version string policy which is widely used by other systems too.

[CHANGE] extended version string and corresponding parser in cmake to allow arbitrary tags in the version string.

[ADD] Fiber Stipple module, now also uses grid offsets in order to calculate texture coordinates properly.

[ADD] Now improved coupling from coronal slice (still present, but shader are almost independt. Also bounding box is now taken from probability data instead of WGEColormapper. Implemented threshold, where probabilities below are discarded. Fixed several typos as well as focal point computation which had a few bugs. Now everything should work again.

[CHANGE] FiberStipple module shaders adopted to new module design. Basically many things, such as focal points of the stipples are now calculated on GPU instead of CPU

[FIX #188] the property group widget now adds and removes properties on the fly if the get added/removed in the associated group. But there still is a problem with the QLayout which then somehow resizes itself. Which looks ugly.

[FIX #124] issue fixed. the reason was a re-used callback for each node registered to the colormapper. the nodemaskcallback of the navslices was then applied to all nodes in the colormnapper because addUpdateCallback nests them (enqueue).