5 Oct 2024 |
outfoxxed | can do | 08:00:27 |
K900 | @Nick Cao wanted 6.8 in 24.11 | 08:00:33 |
K900 | I'm not confident but the KDE rooms are also saying 6.7.3 breaks a bunch of shit | 08:00:49 |
K900 | So maybe we go for that | 08:00:52 |
outfoxxed | In reply to @k900:0upti.me @Nick Cao wanted 6.8 in 24.11 ideally this can get in there too ill try to get an isolated environment with 6.8 to test in | 08:01:19 |
outfoxxed | since I don't feel like rebuilding for 10 hours | 08:01:35 |
K900 | The problem with getting 6.8 in 24.11 is that it'll be our last staging cycle before freeze | 08:02:09 |
K900 | And the same cycle will have the Darwin stdenv rewrite | 08:02:17 |
outfoxxed | so its going to be a massive rebuild and we likely wont have time to get the patch in even if 6.8 can get in? | 08:02:39 |
K900 | The patch is fine to get in after freeze probably | 08:02:56 |
outfoxxed | stdenv rewrite means rebuilding every package on darwin probably, right? | 08:03:03 |
K900 | Happens every staging cycle anyway | 08:03:10 |
K900 | It's more about how long fixing the regressions will take | 08:03:22 |
K900 | And how that affects the timeline for future staging cycles | 08:03:33 |
outfoxxed | whats the freeze date? if the patch is ready before then is there a chance it could be applied during the regression test period or should i just wait | 08:04:15 |
K900 | Basically what I'm worried about is that the cycle being too long and then we don't have enough time to dogfood 6.8 | 08:05:04 |
K900 | To be confident | 08:05:06 |
K900 | * Basically what I'm worried about is the cycle being too long and then we don't have enough time to dogfood 6.8 | 08:05:14 |
outfoxxed | yeah I've seen a fair amount of regressions for "minor" version bumps that it might be worth pushing off to the next release | 08:06:08 |
K900 | And ideally I'd like to get 6.8 and the patch in two separate cycles so we don't have to play "was it this change or that change" when something epxplodes | 08:06:20 |
K900 | * And ideally I'd like to get 6.8 and the patch in two separate cycles so we don't have to play "was it this change or that change" when something explodes | 08:06:23 |
K900 | And the timeline is tight | 08:06:38 |
outfoxxed | personally I don't need to get this in before the release, since at least most of my users are on unstable so I'll go with whatever | 08:07:10 |
K900 | Anyway I guess we'll have to see how this cycle goes | 08:07:44 |
K900 | And then we can see what we do about the next one | 08:07:51 |
outfoxxed | if we're doing 6.7 in 24.11 maybe it makes sense to merge the patch anyway for all the kde users, and rebase it for 6.8 when thats ready? | 08:08:06 |
K900 | Because right now this cycle is going like this | 08:08:09 |
K900 | Download image.png | 08:08:11 |
outfoxxed | In reply to @outfoxxed:outfoxxed.me if we're doing 6.7 in 24.11 maybe it makes sense to merge the patch anyway for all the kde users, and rebase it for 6.8 when thats ready? if not ill wait till 6.8 to fix the nits and rebase | 08:09:01 |
K900 | The problem is we don't know if we're doing 6.7 or 6.8 before we try 6.8 | 08:09:35 |