After a chat with David and getting my head wrapped more around backwards compatibility, I also agree that the changes in the patch are far too strict. It's much more important to preserve backwards compatibility than to strictly conform to the RFC.

I've updated the patch to allow for (what should be) anything that was previously allowed as header name/value pairs minus carriage returns not immediately followed by a tab or space (obs-fold: CRLF 1*( SP / HTAB )). This change fixes the reported issue but should not otherwise affect backwards compatibility.

Additionally, even though line folding is deprecated by RFC 7230, I don't think it's necessarily an issue to support line folding until proven to be a problem in practice. With the current implementation, users have the ability to conform to the target server/proxy requirements, based on errors (if obs-fold isn't transparently dealt with as suggested) yielded by each as defined in the RFC. In light of that, I don't think that it's even worthwhile to start deprecating multi-parameter putheader at this point (but I'm open to argument on that one).

One note on the deprecation is that if we deprecate multi-parameter, we should also add a warning if an embedded line fold is detected in a single headers value.