Ben Boeckel

On March 28, 2019 8:05:02 PM EDT, Jefferson Amstutz wrote: >Yes, we have discussed expanding `ospLoadModule()` (which is what does >the ispc module load in `ospInit()`) to allow for a...

I think it worked before in that the implementations were likely linked directly rather than loaded at runtime. I'm going to make a patch which does the equivalent of the...

What are your thoughts on the other bullet points in my previous comments? Should there be a separate issue for that?

Ah, I didn't realize that they were meant to be linked to directly. In that case, the subdirectory probably doesn't make sense.

> that takes either an assumed full path to the module, or a set of search paths that lets the programmer choose exactly where modules are searched. I would think...

Note that the PR that closed this only adds an `--exclude` flag. Is there some way to have PyPI add this flag when verifying uploads? Or does this need a...

I'll note that OSPRay (Intel's raytracing backend) also ends up setting FTZ behavior via its rkcommon library. See . This may warrant expanding the check to other platforms beyond just...

> Such convention is packaging friendly because many package managers do not accept - in version strings. These packages should have a `build` field or something to put it. For...

> In other words, those are now defaults that could be overridden, whereas before they were always used. Yes. Our packaging setup is a bit odd in that we prefer...

One thing that might work is to make a new container, assign it to that new container, then delete the container There might be a better way, but that should...