Here's the summary of the IRC meeting.



Place: #openvpn-meeting on irc.freenode.net
Date: Thursday 23rd May 2019
Time: 20:00 CEST (18:00 UTC)

Planned meeting topics for this meeting were here:


Your local meeting time is easy to check from services such as



cron2, dazo, lev, mattock and ordex participated in this meeting.


Discussed tap-windows6. The two-machine tests are now working on a basic
level on EC2 Windows instances. Testing on physical computers is known
to work, so in the worst case somebody (mattock) will have to setup some
physical Windows Server 2019 (evaluation) instances and run HLK on bare

Post-meeting note: there is also the possibility that we could avoid the
HLK tests that target physical network adapters (NDISTest) by modifying
driver parameters (ifType, PhysicalMediaType):



Discussed "constants.h: make driver not halt on suspend" PR:


Agreed that the PR makes sense, but it needs proper testing before merge.


Full chatlog attached.

Samuli Seppänen
Community Manager
OpenVPN Technologies, Inc

irc freenode net: mattock
(21:00:55) cron2: YES
(21:01:00) cron2: made it
(21:01:17) dazo: \o/ :)
(21:03:17) mattock: great!
(21:04:08) ordex: hi
(21:04:09) lev__: hello
(21:04:19) ordex: cron2: still in iceland? :)
(21:04:34) cron2: yep
(21:06:11) mattock: https://community.openvpn.net/openvpn/wiki/Topics-2019-05-23
(21:06:13) vpnHelper: Title: Topics-2019-05-23 – OpenVPN Community (at 
(21:06:18) mattock: "the usual stuff" it seems
(21:06:44) mattock: tap-windows6: I was able to launch two-machine tests which 
did not fail immediately and miserably
(21:06:49) mattock: so things are moving forward
(21:07:12) cron2: good :-) - but didn't you want to have the tests finished 
last week already?
(21:08:00) mattock: "want" yes :D
(21:08:41) mattock: "could" no
(21:10:14) mattock: anyways, I will put the HLK client and support machine to 
run through all the tests next
(21:10:17) dazo: But .... I'll admit I haven't paid too attention to the 
details in this testing .... but .... a) what is breaking now?  b) who will fix 
what?  c) what else needs to be run to complete?
(21:10:54) mattock: on a basic level there should not be anything that needs 
fixing on HLK
(21:11:02) mattock: the issues may go away on a rerun
(21:11:06) cron2: a) hopefully nothing, b) either us or Stephen (depending on 
what broke)  c) after the test, you get a piece of data, and submit that to 
microsoft to get the keys to the kingdom
(21:11:35) mattock: worst case: I need to setup HLK on real computers 
(fortunately I now do have a boatload of laptops I can probably use for that)
(21:11:44) mattock: not my own, but probably available
(21:11:57) dazo: is the "two machine test" step the last thing we need to 
complete to finally close this topic?
(21:11:58) mattock: "test on physical machines" is known to work
(21:12:12) mattock: yes I think so
(21:12:23) mattock: two-machine tests are working a basic level
(21:13:02) mattock: any issues that may arise could be related to 
virtualization and/or rozmansi's "virtual tap-windows6 patch"
(21:13:02) dazo: have we updated our own docs on all the various steps etc?
(21:13:12) mattock: yes of course, there are too many steps to remember them
(21:13:18) mattock: HLKTesting in the wiki
(21:13:27) mattock: plus I have automated the HLK environment setup
(21:13:30) dazo: cool! thx!  I'll catch up on that later on :)
(21:13:39) mattock: you want to start HLK testing, eh? :D
(21:14:09) ordex: :D
(21:14:13) ordex: dazo always wanted to
(21:14:17) dazo: nope, just get an idea when we might close this topic :)
(21:14:31) mattock: ah how I would love that :D
(21:14:33) dazo: and I've heard "next week" for half a year or so ;-)
(21:14:55) dazo: btw .... I believe lev__ might have a few bugfixes to 
tap-windows6 too
(21:14:58) mattock: now, it seems likely that we (company) will get somebody 
else to take the pain going forward (post-this-release)
(21:15:01) mattock: he sure does
(21:15:12) lev__: speaking of tap-windows6, are we OK to merge 
no_halt_on_suspend thing
(21:15:29) cron2: lev__: don't be so impatient
(21:15:37) cron2: if we're OK, I'll merge it
(21:16:11) mattock: also note that every change to tap-windows6 will require me 
to run all the tests again :D
(21:16:22) mattock: so let's bundle them when the tests are passing
(21:16:47) dazo: backstory to that is that we want to ship an updated 
tap-windows6 driver with a coming OpenVPN Connect .... which currently doesn't 
handle suspend/resume too well.  We have a "workaround" in the OpenVPN 3 Core 
Lib .... but it isn't the proper fix
(21:17:42) dazo: mattock1: understood .... and that might be needed in not too 
far future
(21:17:48) cron2: dazo: it might well be the right fix, but we need a bit more 
exposure and testing
(21:17:57) dazo: agreed!
(21:18:13) mattock: dazo: for openvpn connect you don't need Windows Server 
2016/2019 support, or do you?
(21:18:15) cron2: so I'm all for having a driver *with* that change out and 
tested, and I think Selva is also mostly fine
(21:18:37) dazo: mattock1: I doubt so for the few first releases
(21:19:27) mattock: dazo: then you just need cross-signed / attestation signed
(21:19:31) mattock: but enough of tap-windows6?
(21:21:24) dazo: so openvpn 2.5?
(21:21:38) cron2: seems like it
(21:22:16) dazo: how good have we been at updating 
https://community.openvpn.net/openvpn/wiki/StatusOfOpenvpn25 ?
(21:22:18) vpnHelper: Title: StatusOfOpenvpn25 – OpenVPN Community (at 
(21:22:37) ordex: not so good
(21:22:39) cron2: well
(21:22:59) cron2: we didn't actually get anything major done in the last 6 
months, so the page was updated 2 months ago - that is better than the rest
(21:23:25) ordex: we made some progress on sitnl (hurray!!)
(21:23:34) cron2: yeah, the above was a bit unfair
(21:23:53) dazo: :)
(21:23:55) cron2: I was making progress with merging that but this week's 
conference got in the way.  Will continue with merging & rambling over the 
(21:24:45) dazo: cool!  I'm planning to allocate lots of hours tomorrow for 
openvpn 2 coding and maybe even some review
(21:25:05) cron2: what is on your coding plate?
(21:27:09) dazo: fixing up the plugin examples (low prio) ... and I thought the 
struct argv patches where applied, but seems there is still some more work here 
and review plaisthos' auth-token-hmac patches
(21:27:26) cron2: (I'll be travelling for ~7 hours tomorrow - bus, 4h plane, 
train - so won't get much done)
(21:27:40) dazo: no worries
(21:28:16) cron2: dazo: review is always welcome :-) - I was just wondering 
which coding bits you had.  There's a bit of plugin stuff indeed, and argv...
(21:28:20) dazo: but if there's anything eminent I should dive into, let me
(21:28:21) dazo: know
(21:30:07) cron2: having plaisthos' patch set reviewed & ACKed would be great, 
and seems to fall squarely into your field of expertise
(21:30:22) dazo: yeah, I'm fine with reviewing that
(21:30:26) cron2: (and would sort out a number of open issues plus other 
no-longer-needed-then patches)
(21:31:16) cron2: ordex: what's keeping you busy these days?
(21:32:09) ordex: I was busy with all kind of offices here in ITaly in the past 
two weeks and now I am catching up on my corp backlog
(21:32:16) ordex: it has all been a bit unfortunate :/
(21:32:28) ordex: but I am trying to get back to "get stuff done"
(21:32:54) ordex: I'll try to force myself to stop everything and so some 
community work tomorrow too
(21:32:57) ordex: *do
(21:33:14) ordex: otherwise there will always be "this something else that 
needs to be done"
(21:33:16) ordex: :S
(21:33:22) dazo: hehe ... yeah
(21:33:46) dazo: but if you get a chance (you have the corp blessing) to do 
community work, that's all good!
(21:33:59) ordex: yap
(21:34:36) cron2: great :)
(21:34:46) cron2: now if we can revive syzzer...
(21:35:53) ordex: :D
(21:36:21) cron2: so... anything else on 2.5?  (I'm not going to do review 
today - leaving for a nice social evening soonish and have my head elsewhere)
(21:37:00) ***dazo got another meeting in 25 min as well
(21:37:10) ordex: +1 for the social event ;P
(21:38:03) cron2: (as a side note, with kids around, december is already really 
stressful, with lots of "pre-christmas" activities - so this is why I can't do 
any december dates)
(21:39:03) mattock: ok call it day?
(21:39:13) ordex: yeah, I think most of us will be in that situation, but I 
wanted to propose them for democracy sake :)
(21:39:52) cron2: +1 to both :)
(21:40:07) ***cron2 holds his thumbs for the HLK tests  ;)
(21:40:07) dazo: yeah, December is horrendous for anything but family
(21:40:16) ordex: and food
(21:40:17) ordex: :D
(21:40:51) cron2: so.  *wave*
(21:40:59) dazo: o/
(21:41:22) ordex: <o
(21:41:50) mattock: bye!

Attachment: signature.asc
Description: OpenPGP digital signature

Openvpn-devel mailing list

Reply via email to