Fork me on GitHub
#sci
<
2021-11-13
>
Kevin09:11:04

Could Sci be suitable for isolating users who are running code on the same JVM instance? One issue I can think of is if the users share the same multimethod, they could override each other's dispatch values

borkdude09:11:38

@kevin.van.rooijen multimethods are not defined globally but within each sci context

⁉️ 1
Kevin09:11:20

I meant multimethods that are passed through the :namepsaces key

Kevin09:11:26

e.g. from a library

borkdude09:11:05

yeah that might need some work, but you can probably work around this somehow

Kevin09:11:43

I was thinking about it as well, I noticed that you can't run .addMethod directly in Sci. which is good

borkdude09:11:47

e.g. you could override the name of the multimethod with a local multimethod definition which dispatches to the global one and you have some room to do your own thing

borkdude09:11:56

or just remove the multimethods if you don't want users to access those

Kevin09:11:39

All right that'll give me something to think about

Kevin09:11:06

Other than that though, would you see any issues with using Sci in this way? (Multiple users on the same JVM)

borkdude09:11:26

it should support this use case