|
00:07
Geth left
00:08
Geth joined
|
|||
| Geth | rakudo/lizmat-mixed-int-num-infixes: 9a17de9efa | (Elizabeth Mattijsen)++ | src/core.c/Num.rakumod Address japhb++'s thinko catches |
00:08 | |
| lizmat | japhb++ ^^ | ||
|
00:26
librasteve_ left
|
|||
| japhb | lizmat++ | 00:59 | |
|
04:36
kjp left
04:38
kjp joined
05:35
kjp left
05:36
kjp joined
05:37
kjp left,
kjp joined
05:57
melezhik joined
|
|||
| patrickb | [Coke]: When's the blin run planned? I haven't merged all the PTY stuff yet. But now that I got write perms on the libuv fork repo I could. | 06:35 | |
|
08:16
melezhik left
11:28
librasteve_ joined
|
|||
| patrickb | [Coke]: The PTY branches are now all ready (they now rely on the temporary libuv fork, are rebased onto latest main and pass make test and make spectest). I could merge the entire lot. Can you give me a final "go ahead"? (i.e. the blin run didn't start yet) | 13:07 | |
| disbot6 | <melezhik.> patrickb: ooo what’s PTY branches ? | 13:42 | |
| <melezhik.> ooc | 13:45 | ||
| patrickb | adding PTY support to Rakudo involved changes in Rakudo, moar and the underlying third party libuv library. So I had git branches for all of them. | 14:10 | |
| disbot6 | <melezhik.> Ok. We can run tests over brownie if you want , it allows to build Rakudo source code | 14:14 | |
| <melezhik.> Is it just one Rakudo branch from which we can build Rakudo ? Or many ? | 14:15 | ||
|
14:30
librasteve_ left
|
|||
| [Coke] | I haven't started blin yet, can start it sunday late if needed. | 14:34 | |
| patrickb | Then I'll go ahead and merge it in the next few hours. Thanks! | 14:56 | |
| [Coke] | Sounds good. Thanks! | 15:04 | |
| .tell melezhik next release is basically a week from today | 15:05 | ||
| .tell melezhik if you want to try to coordinate a brownie run after patrickb's merge, also very cool. | |||
|
16:28
melezhik3 joined
16:32
melezhik joined
|
|||
| melezhik | [Coke]: absolute doable just need to point brownie to the head I guess | 16:32 | |
| I will just need free agents ) , patrickb please let me know when you merge into master so we can run brownie round and then compare with 2025.11 | 16:33 | ||
|
16:33
melezhik3 left
16:45
finanalyst joined
|
|||
| finanalyst | anyone on line use zef in a dockerfile ? I'm having a problem with `zef install --/test 'Getopt::Long'` | 16:48 | |
| some time ago I had no problem with `RUN zef ...` but now something has changed | 16:49 | ||
| seems that I broke a chained sequence of zef into separate RUNs. So the problem is not zef, but Docker image building. | 16:56 | ||
| [Coke] | can you gist the Dockerfile showing problems? | 17:54 | |
| finanalyst | [Coke] The problem is installing modules with zef. I have now discovered that the cpan repository is not enabled upon installation | 18:35 | |
| The module that is not being found is Getoption::Long | 18:36 | ||
| [Coke] | That module doesn't exist? | 18:38 | |
| Do you mean "Getopt::Long" ? | |||
| finanalyst | The difficulty is that if a docker image that only has an installed zef is run as a container, and then a terminal is opened in the container, `zef install Getoption::Long` will run without error | ||
| [Coke] | raku.land/cpan:LEONT/Getopt::Long ? | ||
| finanalyst | My bad Getopt::Long | ||
| [Coke] | Yah. No reason why that shouldn't work in a dockerfile with zef. | 18:39 | |
| Can you show the dockerfile so we can see if something else is going on? weird pathing, weird zef env vars... | |||
| finanalyst | So I looked at the zef configuration file in the docker image, and the 'enable' field for cpan is 0 | 18:40 | |
| I found that Getopt::Long is in cpan, while Getopt::Long::Grammar is in fez | 18:41 | ||
| lizmat | finanalyst: what is the REA setting in zef's config? It should find Getopt::Long there | 18:42 | |
| finanalyst | looking | ||
| yes the 'rea' repository is enable: 1 | 18:43 | ||
| lizmat | then it should install Getopt::Long from there | 18:44 | |
| finanalyst | However, when I include a `RUN zef update` command in the dockerfile, I get an error saying that the mirror can be established | 18:45 | |
| lizmat | (not ?) | 18:46 | |
| [Coke] | I can't help you if you don't share the Dockerfile. :) | ||
| finanalyst | not be established | ||
| [Coke] -> afk | 18:47 | ||
| finanalyst | [Coke] sorry. Making a gist | ||
| lizmat | finanalyst: the latest Getopt::Long on the REA is Getopt::Long:ver<0.4.2>:auth<cpan:LEONT> | 18:48 | |
|
18:51
melezhik left
|
|||
| finanalyst | <script src="gist.github.com/finanalyst/7480bea...script> | 18:53 | |
| gist.github.com/finanalyst/7480bea...adee4ef66a | |||
| FWIW The dockerfile I had developed only included Cro and it was working up until v1.0.0 of zef | 18:55 | ||
| Then I began to get install failures, and I have only needed to work out why now. | 18:56 | ||
|
19:08
melezhik_ joined
19:10
melezhik_ left
|
|||
| finanalyst | lizmat: is there a way to increase the time zef tries to get information before it fails? | 19:41 | |
| lizmat | ugexe might know, I do not offhand | ||
| finanalyst | ugexe: zef is failing to update the rea mirror after an update. I want to know if its a timing issue. Can I increase the time for an update? | 19:43 | |
| The problem occurs when setting up a docker image | |||
| ugexe | does `curl raw.githubusercontent.com/Raku/REA...META.json` run in that docker image work? | 19:55 | |
| finanalyst | ugexe: checking | 19:58 | |
| curl is not found. | 20:01 | ||
| leont | Right I still need to transfer that one to zef | 20:03 | |
| I've moved most of them, it was my last raku release to CPAN, but I just haven't needed to do any release of since February 2023 | 20:04 | ||
| finanalyst | ugexe: by installing curl in the Alpine image, zef update completed without a problem, and the module installed without a problem | 20:08 | |
| lizmat | shouldn't zef spot the lack of 'curl' ? | 20:09 | |
| patrickb | lizmat: May I kindly ask for a moar and nqp bump? | ||
| finanalyst | leont: Getopt::Long was the first problematic module, but there are several others | ||
| lizmat | patrickb: will do | ||
| patrickb | ❤️ | 20:10 | |
| lizmat | patrickb: no changes to NQP, correct ? | ||
| patrickb | no only moar | ||
| lizmat | patrickb: how should I describe the MoarVM work? "Implement PTY support" ? or something more specific? | 20:12 | |
| patrickb | that's actually pretty on spot | ||
| no PTY before, now there is :-D | 20:13 | ||
| Geth | nqp/main: 01ddba8a4d | (Elizabeth Mattijsen)++ | tools/templates/MOAR_REVISION Bump MoarVM to get PTY support implemented by patrickb++ |
||
| rakudo/lizmat-mixed-int-num-infixes: 291f0aaf84 | (Elizabeth Mattijsen)++ | tools/templates/NQP_REVISION Bump NQP to get MoarVM PTY support implemented by patrickb++ |
20:19 | ||
| lizmat | argh, wrong branch | ||
| Geth | rakudo/main: 1d6aae348b | (Elizabeth Mattijsen)++ | tools/templates/NQP_REVISION Bump NQP to get MoarVM PTY support implemented by patrickb++ |
20:20 | |
| patrickb | Thank you! Greatly appreciated. | 20:21 | |
| ugexe | if they didn't have curl installed but zef didn't complain about not having an available fetcher, then it would have been using wget | 20:22 | |
| lizmat | I guess there is no wget either then ? | 20:25 | |
| in that container, I mean? | |||
| ugexe | there has to be. not only did zef not complain that no available fetcher was found, but it is clearly fetching the fez ecosystem list | 20:26 | |
| lizmat | afk& | 20:27 | |
| disbot6 | <melezhik.> My 2 cents . On minimal vanilla alpine Linux wget is way too old by default , so it might not work as expected , but maybe this is not a cause of the issue … | 20:29 | |
| patrickb | does the error "Could not open (BUILDPLAN)". Failed to stat file: no such file or directory" during `install-core-dist.raku` ring a bell for anyone? | 20:58 | |
| [Coke] | Did we ever get a look at the Dockerfile? | 21:39 | |
| Anything else incoming before I hit blin? | |||
| patrickb: you want that win box spun up, let me know. | |||
| c: 1d6aae348 3.put | 21:40 | ||
| committable6 | [Coke], ¦1d6aae3: «3» | ||
| [Coke] | blin started | 21:43 | |
| We never had a fix for that Unicode Security issue, I don't think, so that'll still be there. | |||
| SmokeMachine | Test::Time's test should not be hanging any more | 22:13 | |
| patrickb | That PTY merge seems to have broken Rakudo on Windows. My best guess is some form of bit-rot. I'm investigating, but I'll definitely won't be able to get this fixed tonight. | 22:15 | |
| Shall we revert? | |||
| [Coke] | Do you think you can fix it this weekend? | 22:17 | |
| I'm happy to give a day or two. | |||
| I will kill the blin run, though. | 22:18 | ||
| patrickb | Maybe. But as I don't know what's wrong though it's a bit hard to tell. | 22:19 | |
| I think I found the issue. | 22:55 | ||
| That latest MoarVM commit fixes the issue. | 23:04 | ||
| That'll need another moar and nqp bump. And then (given things are all good then) merging of the respective Rakudo PR and then blin can go. | 23:05 | ||
| I'm off to bed now. lizmat can you do the bumps once more? | 23:07 | ||
| Coke: Please don't start blin until the Rakudo PR is merged. | |||
| Geth | nqp/main: 57217cbd75 | (Elizabeth Mattijsen)++ | tools/templates/MOAR_REVISION Bump MoarVM to unbreak the Windows build |
23:29 | |
| rakudo/main: 81a6e768d3 | (Elizabeth Mattijsen)++ | tools/templates/NQP_REVISION Bump NQP to unbreak the Windows build |
23:34 | ||
| rakudo/main: 6 commits pushed by (Patrick Böker)++, (Elizabeth Mattijsen)++ | 23:53 | ||
| [Coke] | restarting blin | 23:54 | |
| lizmat | [Coke]: please wait until I double checked locally whether it tests ok :-) | 23:55 | |
| probably will, but just in case | |||
| [Coke] | I'll stop it if not. | 23:56 | |
| lizmat | ok | 23:57 | |
| [Coke] | if it tests OK, I'll let it keep going. | ||
| c: 81a6e768d pi.say | |||
| committable6 | [Coke], ¦81a6e76: «3.141592653589793» | ||