← Feed

Posts

• Post #3024181

  @krismicinski@types.pl Yeah, I think I got the gist of all of this. Thanks! I'm finding a lot of the datalog literature that focuses on program transformations to be needlessly complicated compared to simple dataflow graph transformations (more common in the DB literature, it seems). Regardless, thanks for the perspective. I've context switched back to debugging annoying assembly and bootstrapping issues, so already forgot half of what I know about datalog.

• Post #3024180

  I don't know of any research that the pl community has done to make ebpf programming actually good. There's a massively deployed jit/verified execution environment that is integral to most modern infrastructure, and programming it sucks. Seems like low-hanging fruit. Am I missing existing work?

• Post #3024179

  How does one find a weekend to un-jankify their emacs configuration? I crap you not, I have a ton of cruft in there from before useful versions of use-package and melpa exist. Now it is Frankenstein monstered in with more modern approaches. Sigh. If only time grew on trees.