Implication: if you want a Macbook Neo this year, maybe wait until September—unless you expect the coming supply chain shock to hit Apple, too. Which is not impossible if TSMC can't meet their chip delivery dates.

If that happens, prices will shoot up and scarcity economics will take over, so buy now and be prepared to run it for the next decade.

/2

I don't need a Macbook Neo. My current "cheap mobile writing machine" niche is filled by a M3 Macbook Air. But my newest iPad and iPhone are two generation old Pro models.

I suspect … no new iPad for me in 2026/27, but *maybe* a spec-bumped Neo in Sept/Oct, which will cover a lot of the iPad use cases (and is cheaper).

Then aim to run it forever.

(I'm overdue for a complete re-think of what I do with my herd of computers, and therefore what I actually need.)

/3 (end)

@cstross

There is an entire world beyond the limits of the Apple ecosystem.

I don't know what your needs are, but maybe it is time to test the free software waters. If you start replacing your current applications with open source ones, the change to Linux can be pretty seamless.

And the sense of freedom is priceless.

But to each his own.

@cstross @Tubemeister

I saw that coming. A real pity.

I'm a bit surprised a writer has such a hard dependency with MacOS, but things happen and we can't always choose. Tell George R.R. Martin.

About the distro, I'm fully on the systemd and wayland wagon from some years ago, and I happen to love both.

Wayland solved the annoying tearing issues I had with XWindows, and systemd made my boot times an order of magnitude faster. Last time I had a compatibility issue was about two years ago.

I understand that some tinkery oriented people with a vocation of sysadmins can have issues with them, but as a mere user, there is no chance I am even testing any distro without both in the near future.

The good thing about having choices is we can choose.

@jgg @cstross It's probably possible to use a different tool if the absolute need arises, but converting a primary process like that tends to not be cheap, if only in mental energy. (Though probably in reduced output too which readily translates to actual money.)

As for systemd, I don't mind the core system, the unit files work pretty well from the user end. It's all the extra shit like resolved I can do without, and it's run by a plonker with what seems like an "it works on my laptop" approach