Lijun Zhu

Can you try [this temporary fix](https://github.com/lijun99/isce2/commit/af55155cd8778f783508dff9af3ed2442f8f2ea5)? Basically, it reduces the batch size to avoid the out of memory issue. Let me know whether it works so that I can decide...

A recent update [51af91d](https://github.com/isce-framework/isce2/pull/722/commits/51af91d211ca6be7b2cf3c645bda3f5b1eaadeb2) fixes the issue of gpu memory leak when repeating the cuAmpcor processes within the same python script.

Thanks, @rtburns-jpl ! Looks good to me. I understand the changes will make interfacing with python easier. BTW, maybe we should abandon the Makefile, and use a CMakeLists.txt.standalone instead, such...

"--gpuid 2" is to choose the gpu device with id=2 if you have multiple gpus. Please use "nvidia-smi" to check the available gpus on your computer. If you only have...

I have merged all the pycuampcor updates to [a separate repo](https://github.com/lijun99/cuAmpcor), which can be embedded as a submodule in isce2/isce3: here is [an example](https://github.com/lijun99/isce2/commit/975021568c35c5c9cdfdd8b73a0b970e78ea6acc). @rtburns-jpl , please let me know...

openmotif is not well maintained under conda. Use the system provided libraries instead. conda uninstall openmotif ldconfig -p | grep libXm Ensure something like /lib/x86_64-linux-gnu/libXm.so exists. otherwise, install via #...

With scons, you are supposed to specify the openmotif path in your `SConfigISCE` file as, taking conda as an example, MOTIFLIBPATH=$CONDA_PREFIX/lib X11LIBPATH=$CONDA_PREFIX/lib MOTIFINCPATH=$CONDA_PREFIX/include X11INCPATH=$CONDA_PREFIX/include