Ian Stapleton Cordasco

:+1: LGTM

So, to be clear, in general Nagle's algorithm is disabled for non-proxy traffic. The socket options are kind of confusing. I think we want the algorithm enabled though for proxies....

Also, I'm not certain that at the level urllib3 is operating we can guarantee that headers are all in one packet. Also, we're still supporting 3.8, right? So the fix...

So if Retry ever gets the ability to receive the prior request and response I believe it could be overloaded to behave like Requests Authentication handling does. Some of the...

> @sigmavirus24 ok, good point on not wanting a glut of 3rd party development, but having a solid framework for development and clear documentation and/or 1 or 2 examples would...

> A little bit of background: the server is similar to this demo from its manufacturer: http://netio-4c.netio-products.com/ > Could it be possible that the server is using the particular cipher...

:heart: Thank you @alexwlchan ! I don't have the time with the kiddo and hopefully you can claim the bounty (since I couldn't even if I got the PR over...

By the way, https://github.com/requests/toolbelt/pull/316 is a reasonable addition to the enclosed but will complicate this a smidge

Please don't use screenshots. They're not searchable and folks who rely on screen readers cannot interpret them. Please instead copy and paste the plain text of the images if you...

I'm also worried about how this is handled in when handling redirects in the retry logic (e.g., server returns a `Location: //v:h` or `Location: //host:port//v:h` but I haven't investigated that