Conversation
Signed-off-by: George Jenkins <[email protected]>
Signed-off-by: George Jenkins <[email protected]>
|
Thank you @gjenkins8! Letting the CI do its thing - but as for some feedback on the implementation, I think we should move this into something like a Would you be open to that? |
Hey, thanks for the feedback. That makes sense to me, I can move. The big potential hiccup, is I am using private methods in the pdk (root) package. Let me see what I can do / will potentially need. ie. I might need to move some of the things in the pdk package into an |
Signed-off-by: George Jenkins <[email protected]>
|
a743206 is pretty significant. I moved memory related functionality into an As you may imagine, a lot of code movement was required for this. Ultimately, I think it might be worth splitting this PR into smaller PRs (refactors vs feature addition), but I've leave this up to reviewer(s) to decide if needed. edit: I could also imagine there could be different ways to slice and dice these changes e.g. all host / wasmimport functions could be in an " |
|
Unscientific datapoints: this line increases the compiled wasm module size from 2.3Mib to 6.6Mib (big go, GOOS=wasip1): and these three lines increase the compiled module size from 2.3Mib to 9.1Mib: So it seems like a worthy goal have to |
evacchi
left a comment
There was a problem hiding this comment.
from what I can tell (I am not super-familiar with this codebase) the changes make sense, I'll refer to @mhmd-azeez and/or @nilslice for approval :)
Co-authored-by: Edoardo Vacchi <[email protected]>
mhmd-azeez
left a comment
There was a problem hiding this comment.
This unlocks a lot of REST client libraries, thank you very much!
|
will get this out in the next release! |
Create a
HTTPTransportthat "round trips" go native http request/responses via Extism's internal HTTP request/response mechanism