Ed Bartosh

@mythi so far FPGA initcontainer was not optional. It's better to have it disabled in preprogrammed mode, but it's not mandatory at all.

@linjiach Can you provide screencast script and image building instructions for the OpenCL demo or just point us out to where to get those, please?

@linjiach I was only able to find your [demo image on the dockerhub](https://hub.docker.com/r/linjiach/fpga-opencl-centos) and [your set of repos on the github](https://github.com/linjiach?tab=repositories) However, there is no mentioning of fpga-opencl-centos there. Can...

@daixiang0 Let's discuss this with the following example: Let's say we have 8 pf devices: /dev/dlb0 - /dev/dlb7. Pods request them as dlb.intel.com/pf resources, work with them and release them...

Another way is to make one VF device and request it. This way you'll always get the same N.

@mythi would it make sense to sort [this output](https://github.com/intel/intel-device-plugins-for-kubernetes/blob/main/cmd/dlb_plugin/dlb_plugin.go#L81) to make N a bit more predictable?

> initcontainer enables 4 vfs from each pf and configures vfs (this might need to be improved, but for now, I and Mikko agreed to just set static number 4...

> What is the desirable thing to run with initcontainer image? Do we run with docker? CI doesn't have docker as far as I know. It uses either CRI-O or...

> e2e fails because pf is not available anymore after vfs are made. What decision would we need to make for this? Why it's not available? e2e test from main...

> Do you have an idea on the default value for PF and VF? I'd not use default values for them if possible. It depends on the workload requirements I...