Jacques Zwar

> I think that adding base64 support for the knot vectors would be very good. We are planning to implement something similar to the `Geometry` tag to write solutions as...

@filiatra As discussed, I looked into the structure date is currently written to file. Adding a flag to gsFileData will not be sufficient. Streaming a file into `gsFileData` calls the...

**Stumbled upon another bug**. Non-consecutive functions for boundary conditions resulted in segmentation faults, as the vector containing the function expression pointers was initialized using the number of functions in the...

Added a verbose options for boundary condition output: ``` * Dirichlet : 12 * Neumann : 6 * * Summary * * Type Patch Side Unknown Components Function * Dirichlet...

I implemented a crude version of the Levenberg-Marquardt-method, based on [this](https://www.igpm.rwth-aachen.de/Numa/NumaMB/SS10/LevMarq.pdf). It significantly reduces the number of failures to: `[11 21 61 62 81 82 88]` with the same example....

> I implemented a crude version of the Levenberg-Marquardt-method, based on [this](https://www.igpm.rwth-aachen.de/Numa/NumaMB/SS10/LevMarq.pdf). It significantly reduces the number of failures to: `[11 21 61 62 81 82 88]` with the same...

Yes. BID starts at `1`. Hence, `1` is default. Should you have more than one boundary, it should only return the first one, though.

It seems that the function `eval` in `make_showable` keeps track of the results. I think that the key words are just passt without checks to the vedo interface. Maybe we...

I know that `c` is applicable to points and meshes as well, which is exactly why I think we should rename the spline color options `c`, `alpha`, etc. to unique...

Did changing BsplineBasis to BSpline fix the issue? Then I will change the output accordingly