Tom Thorogood

> regarding this errors: I think the est. pattern in this lib is: "bad.. ", with maybe slightly more context than just 'bad SVCB' Yeah I've gone back and forth...

I've left quite a few TODOs and comments where I've discovered incorrect or problematic behaviour. I've successfully run the CoreDNS test suite against this change now as well. Figuring out...

> Duplicate of #48897 @seankhliao I don't think this is a duplicate of #48897. That issue was seeking an `io.Reader` that always returns `io.EOF`, this issue is instead proposing an...

`crypto/tls` has let you do this since go1.18 so it doesn't seem like the worst idea (https://pkg.go.dev/crypto/tls#Conn.NetConn). Though they have the benefit of always having a TCP connection (except when...

Hmm, I'm not sure about this. It seems like we explicitly handle the provided count value being too large so this may be a breaking change. See `unpackRRslice`: https://github.com/miekg/dns/blob/0d504a67cffe8a05b3ff4057c7602b4afe0b26ef/msg.go#L663-L666 What...

> > Thanks for the unit test. > > I'll probably merge this pr first: #1507 which may or may not introduce a conflict here (FYI) > > Thank you...

Can you post the output of https://tmthrgd.github.io/mpw-js/test/?

@leemachin That error *shouldn't* have prevented the test suite from functioning, but either way, it has now been fixed (see #9). Would you mind running the test suite again for...

@jmichael2497 Thanks for the update. Is there any error message from the failure or in the debug console?

@jmichael2497 (From memory): The slowdown in Firefox is caused by missing support for the right PBKDF2 algorithm which is instead done partly in JavaScript.