| IRC logs at
Set by AlexDaniel on 12 June 2018.
00:49 Kaiepi left 00:54 Kaiepi joined 07:46 patrickb joined
patrickb I have just registered the 'MoarVM' AzureCI organization and made the `raku-infra` user its owner. (That same virtual user is used for the Raku and rakudo AzureCI orgs). Is this OK or should someone from MoarVM (jnthn) take ownership? 08:18
I have just requested to install the AzureCI hooks to MoarVM/MoarVM. I'd be grateful for someone with permissions to approve. This will allow me to enable AzureCI testing for MoarVM. 08:46
nwc10 patrickb: seem useful, but I can't help. (I'm not anything official) 08:47
jnthn: I think that there is a bug in commit 295cde4fcffccf45c68d75fbaec46f11aa00b0c4
patrickb Morning! 08:48
nwc10 jnthn: I think that it's missing a check for (key_in)->body.num_graphs == key_out->body.num_graphs
and both better written using MVM_string_graphs_nocheck() 08:49
and I *think* we've never noticed because the preceding if (hashval == (out)->hh.hashv) {\ has always saved us. (which implies that the hash function is pretty good. 08:50
Geth MoarVM: patrickbkr++ created pull request #1299:
Add an AzureCI test setup
09:37 sena_kun joined 10:27 Altai-man_ joined 10:30 sena_kun left 11:44 Kaiepi left 11:48 konvertex joined 12:13 Kaiepi joined 12:28 sena_kun joined 12:30 Altai-man_ left
jnthn nwc10: Hmm, that's interesting indeed... 13:36
Also, yes, if the hash function really has covered
nwc10 that's also what I thought. "That can't be buggy. It must have been found by now..."
jnthn I've lost count of the number of times I've thought that :) 13:37
patrickb jnthn: Did you see my above messages wrt AzureCI? 13:46
jnthn patrickb: Tried to do the approval, let me know if it doesn't work out
lol, nice timing :P
patrickb ;-)
jnthn, is the infra user ok as org owner or should I make you the owner of the MoarVM azure org?
jnthn What is the infra user? 13:47
e.g. is there one person who can do stuff with that, or a group of people?
rba This is a user I use and it‘s in the password store, where Matthias has access too. 13:48
MasterDuke patrickb: btw, it looks like the azure setup doesn't have all the same moar build option configurations as the current travis/appveyor setup
jnthn OK, that sounds better than it being me.
Do I have to make some kind of azure account to see CI output once this is adopted?
patrickb No, it's all in public 13:49
jnthn Very nice
patrickb once the github hooks are in place there should be nice links from github as well
rba We learnd that in azure we can put 5 users and give them the full access.
MasterDuke can multiple people be granted the ability to restart jobs? that's one annoying thing about travis, sometimes i can restart them and other times not, and i have no idea why 13:50
patrickb MasterDuke: Not sure. I think there are no limits on users or similar, but we'll have to try to find out. 14:05
MasterDuke: Do you know the ins and outs of the travis configuration? The MoarVM one is a lot more complex than those of Rakudo and NQP. 14:06
I think I 14:11
I'll set up the AzureCI config the way I have it now and see about the other build options travis currently provides in a separate step.
Geth MoarVM: fc7922e5ee | (Patrick Böker)++ | 2 files
Add an AzureCI test setup

This setup is identically part of the rakudo and nqp repositories. The only change are the four difference are the variables at the top of `azure-pipelines.yml`:
MoarVM: 83bc2e624f | (Patrick Böker)++ (committed using GitHub Web editor) | 2 files
Merge pull request #1299 from patrickbkr/azure-ci

Add an AzureCI test setup
MoarVM: 46d2a1a9d7 | (Patrick Böker)++ | README.markdown
Add an AzureCI build badge
MasterDuke patrickb: no, i just looked at the options being passed in the two different systems and saw they had some obvious differences 14:17
patrickb MasterDuke: I'll have a look in the coming days. 14:18
MasterDuke ++patrickb
patrickb jnthn: Creating the pipeline has worked. 14:19
MasterDuke heh. could probably disable the JVM build configurations... 14:20
patrickb MasterDuke: True :-P 14:25
14:27 Altai-man_ joined 14:30 sena_kun left 14:36 lucasb joined 14:45 Altai-man_ left 14:47 sena_kun joined 15:30 patrickb left 16:27 Altai-man_ joined 16:30 sena_kun left 17:20 zakharyas joined 18:28 sena_kun joined 18:30 Altai-man_ left 20:27 Altai-man_ joined 20:30 sena_kun left 21:22 zakharyas left 22:28 sena_kun joined 22:30 Altai-man_ left 23:23 sena_kun left 23:26 lucasb left
timotimo how do we write a snippet for the *ables to run so we get the number of FSA pages allocated for each size bin after the end of "perl6 -e ''" for a whole bunch of versions, i wonder 23:50