I would like to implement the http2-wrapped Web transport protocol on node.js . (https://datatracker.ietf.org/doc/draft-ietf-webtrans-http2/)
As node.js sits on top of nghttp2, it also depends on nghttp2 capabilities.
In principle, it is an easy task to do this, as node.js and nghttp2 support already the extended connect settings frame.
However, the protocol needs also SETTINGS_WEBTRANSPORT_MAX_SESSIONS set on the server and client side to indicate support of the protocol. Section 9.1 also defines other settings, but these can also be set using the capsule protocol, which is independent of nghttp2 and node.js and can be made in javascript. (One thing which is puzzling, is that the id in the draft is 32-bit as it should be only 16-bit. I ask the work group about this).
So I would like to ask you, if you may consider adding this setting, even if it is still in draft status?
As I have the impression, that more protocols may come along (in connection to the capsule protocol and extended connect), that require additional settings. I am wondering, if adding the setting as additional parameters to the API is a good idea, or if a more generic API is eventually needed.
I have also asked the node.js people about the issue:
nodejs/node#48962 (comment)
I would like to implement the http2-wrapped Web transport protocol on node.js . (https://datatracker.ietf.org/doc/draft-ietf-webtrans-http2/)
As node.js sits on top of nghttp2, it also depends on nghttp2 capabilities.
In principle, it is an easy task to do this, as node.js and nghttp2 support already the extended connect settings frame.
However, the protocol needs also SETTINGS_WEBTRANSPORT_MAX_SESSIONS set on the server and client side to indicate support of the protocol. Section 9.1 also defines other settings, but these can also be set using the capsule protocol, which is independent of nghttp2 and node.js and can be made in javascript. (One thing which is puzzling, is that the id in the draft is 32-bit as it should be only 16-bit. I ask the work group about this).
So I would like to ask you, if you may consider adding this setting, even if it is still in draft status?
As I have the impression, that more protocols may come along (in connection to the capsule protocol and extended connect), that require additional settings. I am wondering, if adding the setting as additional parameters to the API is a good idea, or if a more generic API is eventually needed.
I have also asked the node.js people about the issue:
nodejs/node#48962 (comment)