Closed mcampbell closed 1 year ago
I'm leaving this as-is in case the author wants to fix the example, but swapping out google in favor of ...
response = internet.get "https://duckduckgo.com/?q=#{topic}"
"worked", in that it didn't fail, but I'm getting 302 response codes. Probably that's on DDG's side.
Not sure if you want to close this or not.
Hmm, that's odd. I'll take a look.
This is a regression, I'll need to investigate it.
There was an off-by-one error in padding handling. Padding is a pretty obscure feature of HTTP/2, but it was used by Google's search results and was corrupting the response data, which was then failing to decompress.
Okay, I released an update to protocol-http2
v0.15.1 which fixes this issue. Can you please check it? Thanks!
Success!
~/dev/async-test via 💎 v3.1.2
at 09:14 ❯ ruby ./multiple-requests.rb
Found python: 59 times.
Found rust: 73 times.
Found ruby: 59 times.
Not sure what your closing protocol is; I can close the ticket if you like. =D
Thanks!
Thanks for reporting back!
Is this user error? The first example works fine, but the second one using a
barrier
fails with the following error. It's a copy/paste of the example.The same happens with ruby 3.2.0 (also under rbenv)
Gemfile.lock