Vladimir Panteleev

> I think the relevant question to ask is: "should the client of this allocator call `deallocate`?" > > Since we're using DbI, the presence of `deallocate` signals to clients...

> Granted, but I'm not talking about other methods, I'm talking about `deallocate`. And a container _must_ call `deallocate` if it is present—to do otherwise would be a bug. Okay....

Here is a proposal for a way forward: - Go ahead and remove `deallocate` and `reallocate` from `GCAllocator` - Add a `ManagedMemoryAllocator` OSLT with the current unsafe behavior of `GCAllocator`...

> I'm not opposed on principle, but I don't see any reason to do this now, since it doesn't block anything. Either way, it's a separate PR. Agreed, just making...

I would say that the root of the issue, really, is that D doesn't have syntax for an empty AA *literal*. For arrays, we have `[]`, and we can add...

> Jan 27 15:26:55 [main(VpEu) WRN] GET https://api.github.com/repos/dlang/dmd/issues/13575/comments failed; Not Found 404. Race condition in GitHub API.

Bump, currently I use Lawndesk, but it would be cool to have it in an actively maintained open source launcher.

> Simply change the "Swipe Up" gesture to not launch the app drawer. This is only meaningful if bundled together with a guarantee that all apps are reachable from the...

OK, I found that `(wrong-type-argument consp nil)` is caused by something unrelated to `straight.el` (related to `undo-tree` which advises killing buffers). However, the `(wrong-type-argument sequencep consp)` error looks like a...

@dakotaKat Thanks for making and publishing this. I noticed that issues were disabled on this repository, so I'm guessing you're not interested in maintaining this project to keep up with...