[ https://issues.apache.org/jira/browse/SSHD-205?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13549082#comment-13549082 ]
Guillaume Nodet commented on SSHD-205: -------------------------------------- It's mostly historical as the client part has only been added after the server was written. I agree things could be cleaned. > move ssh-connection (RFC4254) code from sshd.client.* / sshd.server.* to > sshd.connection.* > ------------------------------------------------------------------------------------------ > > Key: SSHD-205 > URL: https://issues.apache.org/jira/browse/SSHD-205 > Project: MINA SSHD > Issue Type: Improvement > Affects Versions: 0.8.0 > Environment: n/a > Reporter: Andrew C > > In wading through the SSH sources I've been having some trouble trying to > understand how the code base implements RFC4254 (ssh-connection). For the > most part the ssh-connection protocol is symmetric - either and can initiate > any request - but the code doesn't reflect this. > As an example, SSH_MSG_CHANNEL_OPEN requests such as "direct-tcpip". Since > the protocol is symmetric, both the ssh-client and ssh-server are allowed to > initiate the request yet the code is structured to strongly (and > confusingly) imply it is always ssh-client to ssh-server. > For instance, ChannelDirectTcpip classes exist in both > ...sshd.server.channel and ...sshd.client.channel, suggesting one is > ssh-server only and the other is ssh-client only, which isn't true. > Can I suggest, long term, looking at moving all the connection code to > somewhere more neutral such as: > sshd.connection.channel.DirectTcpipChannel{Client,Server} > say. It would also make a good first step towards separating the code > implementing the ssh-connection protocol from the lower-level ssh-client and > ssh-server. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira