Al Chu

while experimenting with some things for #4273, one idea that came to me would be to add a new "attribute" to the `job-list` module that returns all data job-list data...

> This would be useful also with some of the Python API and some frontend commands which currently grab all attributes by default. We have to keep the VALID_ATTRS constant...

I'm hardly an expert of the JSON support within `sqlite`, but it appears the primary value add is the ability to filter / parse the JSON values within sql queries....

it appears from sqlite 3.9 -> 3.37.2, json support is optionally compiled in, and in 3.38 it's optionally compiled out. Further complicating this, its not clear if there's a easy...

Wanted to document a side discussion had with @grondo, we also will want a way to configure which "exec" mechanism we want to use. There should be a: - default...

For consideration, also with `libsdprocess`, however the only user of `libsdprocess` is likely to be `job-exec`, so less important there.

> We should probably sanitize the guest namespace metadata before linking it into the primary namespace in job-exec. That would include detecting cycles that a user might introduce maliciously or...

Another idea, perhaps writing a treeobj should be limited to instance owners, not guest users?

#4313 may be able to solve a few of these, which helps

after merge of #4324 this list is shorter / modified * `const char *attrs[]` list of attributes in job-list `job-list.c` * header format dictionary in `flux/job/info.py` in the Python bindings...