Soni L.
Soni L.
(huh, spurious CI failure?)
we don't like the idea of embedding machine-readable metadata into the `NOTE`/`TODO` tags... if we're gonna be developing the tools for this, may as well use the tag system.
oh. in which case, a separate tag would still be convenient, as it could be checked when updating the testsuite without having to sift through existing uses of `TODO`/`NOTE`.
for context: ``` $ grep -r -a ';; NOTE' test/ test/regress/regress-3.txt:;;; NOTE: bug when calling function defined with type after call test/regress/regress-25.txt:;;; NOTE: infinite loop when \0 character is in...
hmm... in which case, we should maybe explicitly move threads out of component model MVP? also yes, they use "worlds", but as far as we can tell they're nothing like...
also, we should note that we aren't asking component model to solve distributed computing - we're asking it to enable WASI to solve it. (in fact, we would rather have...
but thread.new (and wasm threads) isn't actually shared-everything. it can share memory, but not full module instances. (unless we misunderstand the threading proposal?) regardless, it's still easier to add sharing...
hmm... yes, we see. okay, shared-everything-threads looks useful for the particulars of threads, but... hmm... one of the things the current component model does is that you instantiate modules and...
we would argue the problem is the assumption of everything under a single component instance being shared. why would you want that to be true?
https://requests.readthedocs.io/en/master/api/#requests.Response.is_permanent_redirect ?