Anton Bachin
Anton Bachin
This diagram from the libuv docs sketches the logical structure of libuv. `Lwt_unix` has basically the same structure.  *(source: http://docs.libuv.org/en/v1.x/_images/architecture.png)* However, libuv is maintained by more people, and has...
This issue is about looking into storing backtraces explicitly in rejected promises and explicitly extending them during rejection propagation by `bind`, etc. If this can work, it amounts to a...
Some CPU-bound Lwt programs call `Lwt.pause ()` a lot, in order to yield to any potential I/O that may be pending. However, doing this results in Lwt calling `select`/`kevent`/`epoll` at...
Found by @bluddy in https://github.com/inhabitedtype/httpaf/pull/54#issuecomment-398740281: > OK, I redid the test with the single yield () modification above. For the record, it involves turning line 1608 in `lwt_io.ml` to >...
See [this build log](https://ci.ocaml.org/log/saved/docker-build-66ea7db121ce3c5be6be0869f7fa7f9b/a54ba5361cfd967b4d7c3cdc5fb0bfe79dd3f780).
reasoin-promise needs to stay in Reason syntax for a while. However, as people switch to newer versions of BuckleScript, it should probably switch to ReScript syntax.
From Discord: > nitpicking, you could save a line by binding `login` directly as a reason-promise: > > ```reason > external login: unit => Promise.Js.t(authResponse, Js.Promise.error) = "login" > ```...
...it should have been failing [here](https://github.com/aantron/promise/commit/681d58b879ff6a9ef5b86d61dfd26e389bdb760e#diff-a8d77ef5aad99ea3e0ca9ba334be6a95L280). It was possible to change it to a check that the promise is resolved without changing the underlying code (the commit changed only the...