)]}'
{"/PATCHSET_LEVEL":[{"author":{"_account_id":1000003,"name":"plaisthos","display_name":"Arne Schwabe","email":"arne-openvpn@rfc2549.org","username":"plaisthos"},"change_message_id":"221da2d2202e8c0bc9c55f459ff76ab050dd0ba9","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"0c08a4c1_5990cda0","updated":"2026-07-01 11:36:26.000000000","message":"I think you need to add a paragraph or something to the man page that explain the behaviour that this command will create tun or tap depending on whether openvpn thinks dco is working and mention disable-dco to disable this.\n\nRelated question: do we properly figure out if a pre-existing dco tun mode is in the right mode (p2p vs m2mp) or do we just fail miserable?","commit_id":"f0a65c3eb248ff25a418ac3403cd603bd5364782"},{"author":{"_account_id":1000041,"name":"ralf_lici","display_name":"Ralf Lici","email":"ralf@mandelbit.com","username":"ralf_lici"},"change_message_id":"5172148b036cba2bba3a01a22dee3229aad43c3c","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"6d987b6e_996c1dc6","in_reply_to":"0c08a4c1_5990cda0","updated":"2026-07-01 12:05:39.000000000","message":"I was talking about this behavior change with Antonio, and we were thinking about adding an optional `dco` arg to `--mktun` so current users don’t see any change at all. New users could explicitly ask for a DCO iface. OTOH this would’ve been a nice excuse to push more people toward DCO :) WDYT?\n\nOn your tun‑mode question: yes, https://gerrit.openvpn.net/c/openvpn/+/1734 probes the pre‑existing DCO iface and checks its mode. In fact, 1734 is a dependency for this patch.","commit_id":"f0a65c3eb248ff25a418ac3403cd603bd5364782"},{"author":{"_account_id":1000041,"name":"ralf_lici","display_name":"Ralf Lici","email":"ralf@mandelbit.com","username":"ralf_lici"},"change_message_id":"db9d78190285964d9eccff31a44ada47e18e6060","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"29b8d60f_4b0208c5","in_reply_to":"35b95280_d7341a17","updated":"2026-07-01 15:56:10.000000000","message":"I see a few potential issues with this approach:  \n`--mktun dco-p2p --mode server`,  \n`--mktun dco-server --mode p2p`,  \n`--mktun dco-* --disable-dco`,  \n`--mktun tun --dev tap` (which would be the only way to create a TAP device, but it’s not very intuitive).  \n\nAll of these cases can be handled, but they would add extra edge‑case logic in `options.c`, and the resulting behavior might not be very clear to users.\n\nThe main question is whether we accept that, in 2.8, `--mktun` will default to DCO when available. If yes, then adding new options or arguments doesn’t seem necessary, since the existing flags already cover the needed combinations. If not, I’d avoid encoding `--mode` inside `--mktun`.","commit_id":"f0a65c3eb248ff25a418ac3403cd603bd5364782"},{"author":{"_account_id":1000003,"name":"plaisthos","display_name":"Arne Schwabe","email":"arne-openvpn@rfc2549.org","username":"plaisthos"},"change_message_id":"a905676c38c5423e54526bf5933d0f7deb5978ba","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"35b95280_d7341a17","in_reply_to":"6d987b6e_996c1dc6","updated":"2026-07-01 12:51:30.000000000","message":"you could make the argument to mktun optional. Like have --mktun [dco-p2p|dco-server|tun] and if it is not there you get DCO if possible","commit_id":"f0a65c3eb248ff25a418ac3403cd603bd5364782"}]}