This version (2017/05/27 13:44) is a draft.
Approvals: 0/1

[00:26:04] <AlexLehm> Hi Julien, I think there is a bug in the AddressResolver code that only occurs on some linux versions or so, which breaks the tests sometimes

[00:26:16] <AlexLehm> or rather it breaks the tests in some config, but works in some

[14:08:07] <AlexLehm> temporalfox: should I copy the netty branch project on Jenkins to a new name like vertx-core-netty-branch so that if you push something to the branch it builds automatically on the eclipse fork?

[15:28:07] <temporalfox> AlexLehm hi

[15:28:10] <temporalfox> I got a blue on this branch https://vertx.ci.cloudbees.com/job/alexlehm-vert.x3-core-netty-4.1.4/15/

[15:28:34] <temporalfox> have you checked the localhost issue on windows ?

[16:00:31] <AlexLehm> the windows localhost issue is fixed, I could add a unit test for that though

[16:01:56] <AlexLehm> but the fix in netty is ok

[17:27:44] <AlexLehm> temporalfox: i have created a unit test for the localhost windows issue

[17:29:37] <temporalfox> AlexLehm I thought you fixed it in netty

[17:30:19] <AlexLehm> yes, i have implemented a unit test in vert.x that checks that the issue is fixed (in netty)

[17:30:29] <AlexLehm> no change of the main/* code in vert.x for it

[17:31:42] <temporalfox> can I merge the 4.1.4.Final branch in master ?

[17:33:43] <AlexLehm> yes please

[17:33:51] <temporalfox> cool

[17:36:31] <temporalfox> ok will do

[17:37:44] <temporalfox> it's pushed

[17:41:28] <temporalfox> so I can close

[17:41:28] <temporalfox> https://github.com/eclipse/vert.x/issues/1524#issuecomment-236211395

[17:43:47] <AlexLehm> if you think we need a unit test for the issue (regardless of netty fix), we could use the pr, otherwise you can close the issue

[17:44:21] <temporalfox> what would the test do ?

[17:44:25] <temporalfox> replicate the netty test ?

[17:45:28] <AlexLehm> well, its testing the functionality with “our” resolver code, regardless how its done by netty

[17:46:20] <AlexLehm> while it not strictly necessary since we didn't fix anything in the code, i think you can never have too many tests

[17:46:28] <AlexLehm> if the functionality was broken at some point

[17:47:56] <AlexLehm> up to you to decide if you like the test of coruse

[17:47:58] <AlexLehm> course

[17:51:22] <temporalfox> no we can make a test

[17:51:27] <temporalfox> what does it test ?

[17:51:37] <temporalfox> for instance, I've implemneted ndots=0 in vertx

[17:51:42] <temporalfox> then ported the code and test to netty

[17:51:46] <temporalfox> but I keep the tests in vertx

[17:52:22] <AlexLehm> it resolves localhost, LOCALHOST and LocalHost and checks that the resolution works and the result has the hostname “localhost”

[17:53:55] <AlexLehm> which is 2 issues in the netty resolver before, one was that it was not case-insensitive before and that is was not returning localhost as hostname when evaluating the InetAddress object that was returned

[17:58:45] <temporalfox> ok

[17:58:49] <temporalfox> can you make test for these ?

[18:00:54] <AlexLehm> i did, that is the pr for the issue 1524

[18:02:16] <AlexLehm> https://github.com/eclipse/vert.x/pull/1543

[18:11:06] <AlexLehm> should I merge the test to master and create a new pr?