| 28 May 2022 |
 hexa | nvm | 01:02:27 |
|  Zhaofeng Li joined the room. | 02:44:28 |
| 9 Jun 2022 |
|  @v0id:nltrix.net joined the room. | 22:39:14 |
| 10 Jun 2022 |
| hexa | the acme test is currently blocked by mypy typechecking | 14:40:00 |
| hexa | diff --git a/nixos/tests/acme.nix b/nixos/tests/acme.nix
index 2dd06a50f40..467623fafc1 100644
--- a/nixos/tests/acme.nix
+++ b/nixos/tests/acme.nix
@@ -578,7 +578,7 @@ in {
webserver.wait_for_unit(f"acme-finished-{test_domain}.target")
wait_for_server()
check_connection(client, test_domain)
- rc, _ = client.execute(
+ rc, __ = client.execute(
f"openssl s_client -CAfile /tmp/ca.crt -connect {test_alias}:443"
" </dev/null 2>/dev/null | openssl x509 -noout -text"
f" | grep DNS: | grep {test_alias}"
| 14:41:11 |
| hexa | because _ is used some lines earlier as a result for a string | 14:41:43 |
| hexa | building '/nix/store/c3cisgffip8i49b32ijvwrr02d6ssyvl-nixos-test-driver-acme.drv'...
testScriptWithTypes:299: error: Incompatible types in assignment (expression
has type "str", variable has type "int")
rc, _ = client.execute(
^
Found 1 error in 1 file (checked 1 source file)
| 14:41:54 |
| hexa | m1cr0man^ | 14:54:41 |
 m1cr0man | Oh grand I'll get on that tonight! | 14:55:36 |
| hexa | the patch above works, just not sure if that is the nice way 😄 | 14:56:03 |
| hexa | tbh _ should be treated differently | 14:56:15 |
| m1cr0man | Yeah I'm surprised mypy is complaining about it, pretty sure _ is generally regarded as a discard assignment | 14:57:49 |
| hexa | we could rc = client.execute()[0] | 14:58:26 |
| 27 Jun 2022 |
|  @tinybronca:sibnsk.net joined the room. | 08:21:01 |
| @tinybronca:sibnsk.net | 08:22:48 |
| 12 Jul 2022 |
|  @pederbs:pvv.ntnu.no joined the room. | 22:39:02 |
| 26 Jul 2022 |
| @tinybronca:sibnsk.net changed their display name from tinybronca to tailrec. | 14:46:33 |
| @tinybronca:sibnsk.net changed their display name from tailrec to tinybronca. | 16:01:31 |
| 1 Aug 2022 |
|  @better_sleeping:converser.eu joined the room. | 09:13:07 |
| @better_sleeping:converser.eu left the room. | 09:13:24 |
| 5 Aug 2022 |
 raitobezarius | Hey there, I ran into an unexpected edge case, I have some SSL cert containing multiple aliases, it is still "alive" but when I remove one of the server name, it still tries to generate a SSL cert for this multiple alias even if the domain declaration set do not contain this particular domain I removed | 17:56:47 |
| raitobezarius | Does it ring a bell? | 17:56:56 |
| raitobezarius | I see some domainhash.txt so I guess I should bust it ahem | 18:01:06 |
| raitobezarius | hm, just deleted the lego state, worked fine | 18:03:14 |
| 10 Aug 2022 |
| m1cr0man | Yes it does actually | 18:23:18 |
| m1cr0man | Oh that was a while ago 😅 lemme re-read this a couple times | 18:23:34 |
| raitobezarius | I fixed it by cheating but I do think the issue still exist :P | 18:23:49 |
| m1cr0man | yeah so the domainhash is updated when you add or remove domains. When you remove a domain you can't (as in lego/ACME doesn't let you) remove just that domain from the cert and keep it, it needs to be regenerated. So I'm pretty confident this is expected behaviour, even if it seems a bit odd initially | 18:44:51 |
| m1cr0man | https://github.com/NixOS/nixpkgs/commit/152736d39eeee7ff91274cb3cfe506b4611a37ac thanks roberth for doing that. I have had the most hectic two months of my life 😅 I'm only getting to sit down and look at the acme module again now | 18:46:40 |
| hexa | urgh, yeah | 20:37:58 |
