On Thu Jan 26 13:41:45 2023 +0000, Jacek Caban wrote:
It's better now, but it's still not great to break the assumption that all script hosts are on the list. For example, in theory, script host initialization may lead to recursive creation of the script host, potentially leading to creating multiple script hosts for the same language. While such case is not very interesting, you could avoid the whole thing by keeping this part unchanged and counting active script hosts in `init_script_engine`.
Oh, just noticed in that case I'll have to populate the script host's "script" later, after it succeeds initialization, since it's what I'm using to check. (well, I guess script state would also work, but I don't know if it's reliable enough since it depends on what the script engine reports)