| 18 Apr 2025 |
 hellwolf | ditto. what's blocking it from merging? | 15:42:17 |
 maralorn | I can say what's blocking me from merging, which is holidays. But I can take a look later. | 15:43:34 |
| hellwolf | oh, enjoy holidays! no hurry really, just wanted some visibility | 15:45:10 |
| hellwolf | * oh, enjoy holidays! I am in no hurry, just appreciate some visibility. | 15:45:44 |
| maralorn | I think we are basically finished. | 15:48:20 |
| maralorn | There are a lot of build errors related to static builds. I can’t really tell which of those are problematic regressions. | 15:51:22 |
| maralorn | sterni: With the new flow, is there anything I need to worry about? I can just merge into staging at any time, right? I don’t even need to merge anything before-hand because I don’t have to worry about busting any caches? | 15:53:58 |
 sterni (he/him) | you can just merge it straight in | 15:54:17 |
| sterni (he/him) | in fact you should NEVER merge staging into haskell-updates because that will bust our cache which means we'll have a terrible time in the following week | 15:54:44 |
| maralorn | Okay, good to know. | 15:55:27 |
| sterni (he/him) | i think it'll be either me or malte sitting down and sinking 1-2 days into it probably if we want to do a decent job | 15:55:53 |
| maralorn | Really, why so long? | 15:56:15 |
| sterni (he/him) | there's always annoying stuff | 15:56:29 |
| maralorn | At this point it’s resolving the merge conflict, checking eval, marking broken and hitting merge, right?^^ | 15:56:49 |
| sterni (he/him) | also you always need to factor in compilation times for some things | 15:57:00 |
| maralorn | True | 15:57:07 |
| sterni (he/him) | I mean if you take that approach | 15:57:10 |
| sterni (he/him) | I'd still want to look through the still failing tab, there is probably still a lot of relevant things in there | 15:57:25 |
| sterni (he/him) | though I think given that HLS is done, it's maybe not so bad anymore | 15:57:47 |
| sterni (he/him) | I lost my mind finishing that because you could just not get anything done | 15:57:58 |
| hellwolf | Iteratively, perhaps we could do it in next batch?
I think I could commit to make HLS/Hlint/stylishhaske/etc. for GHC912 always work since I always keep using the latest :D, which I have done so for this batch of PR. | 15:59:56 |
| hellwolf | Maybe there is some transportable snippets of code to the GHC98, too. I haven't had a look. | 16:02:12 |
 emily | it should probably go in next cycle since that's the last one before the freeze | 16:02:34 |
| maralorn | emily: how do you mean? We really want to hit 25.05 with this. Would it be fine to merge to staging rn? When is staging going to staging-next? | 16:03:34 |
| maralorn | Or is staging already to late and we have to merge to staging-next to reach 25.05? | 16:04:14 |
| maralorn | That feels wrong to me. If it is relevant it should have a maintainer. How do you find "relevant" packages? | 16:09:06 |
| emily | yes, of course it should hit 25.05 :) | 16:10:30 |
| emily | the 23rd is the nominal breaking change deadline and I assume there are breaking changes here | 16:10:40 |
| emily | I think landing in staging by then should be fine, but also: if -next hasn't started yet by then, then merging to staging after that feels fine too (because after all we would triage issues at the same time anyway) | 16:11:22 |
| emily | I don't know what the status of -next is | 16:11:46 |
