Christiano Haesbaert

> > I apologize for wasting everyone's time and radically changing it, but the whole luv bits were too insane for my taste, see details below. > > Sounds good....

> > Frankly, can't see much uses, and can only remember one time I relied on it, I've added it more for the sake of correctness/completeness, and to use it...

> I don't see why the old behaviour was wrong. `read_into` is always allowed to do short reads; retrying is handled by e.g. `read_exact`. > Then the use should only...

Another note: It's also not a coincidence that `libuv` makes sure it can never return a short-count there ``` int uv_random(uv_loop_t *loop, uv_random_t *req, void *buf, size_t buflen, unsigned int...

> Summarising this: > > 1. We provide a "secure-random" flow, which is a source of random bytes. > 2. Users should take as many bytes from it as they...

Rebased and force-pushed, no changes.

> ``` > | ___| > __| _ \ | _ \ __ \ > \__ \ ( | | ( | ) | > ____/\___/ _|\___/____/ > Solo5: Bindings...

Join the church of printf ! You're very likely reaching `solo5_app_main()` and then `caml_startup()`, I'd make sure that is the case, and then think from there, it might even be...

I think the real issue with cstruct is always in the writer front, parsing from it is usually ok. I had this issue in awa-ssh where I usually want to...

@hannesm On the DoS. I disagree, you have to know what you’re doing, there is no escape. Every network stack allocates bigger chunks than necessary for sending a byte. Normally...