Unfortunately, this time around the majority of AI build up are GPUs that are likely difficult to accomodate in a random build.

If you want a GPU for graphics, well, many of them don’t even have video ports.

If your use case doesn’t need those, well, you might not be able to reasonably power and cool the sorts of chips that are being bought up.

The latest wrinkle is that a lot of that overbuying is likely to go towards Grace Blackwell, which is a standalone unit. Ironically despite being a product built around a GPU but needing a video port, their video port is driven by a non-nvidia chip.

It is a shock, but at least they received their money without being left holding the bag. They have a committed backlog over a year long, they seem to be avoiding manufacturing more than they have already sold …

I had some files that i knew had duplicates, but didn’t exactly match and while the filenames were not identical, you could tell by looking if they were the same.

Would have been very tedious to do all of them, LLM was able to identify a “good enough” number of duplicates and only made a few mistakes. Greatly sped up the manual work required to clean up the collection.

But that’s so far from most advertised scenarios and not compelling from a “make lots of money” perspective.

CPI isn’t perfect, but it’s more indicative of real world concerns of average person. I don’t care about gold, but I do care about goods accounted for in the CPI.

Just because a commodity is roughly fixed in amount and availability not as easily manipulated by making up numbers does not mean it has some magical absolute “value”. Value is always subjective and heavily based on context.

Offer me a bar of gold for everything in my pantry today? Sure thing, I know I can get a lot more money and refill my pantry. If you made that same offer, but people are disinterested in gold? No, because ultimately that bar is useless to me without the wider context of people willing to part with significant resources in exchange for the gold.

Modern monetary policy is as much about psychology as it is about math. Gaming the system so that, largely, people have a number that behaves predictably to general goods and services over time. In the great depression, the relative value of gold in real terms fluctuated wildly, and society exploded in no small part due to people’s feelings relative to the numbers despite the fundamentals not changing much between “things are going great” and “everything is ruined”.

While gold has intrinsic usage, the value is driven more largely by speculation.

Forget gold to dollar for a moment, but think gold to groceries. In 1985 an ounce of gold would be good for maybe 5 typical trips to the grocery store. Today it’d be about 12-15.

The concept of objective, absolute value is generally flawed, but especially gold is pretty distant from even a vague attempt at it.

Both KDE and Gnome have a comparable set of default keyboard shortcuts.

The difference is if you are in KDE, you have easier ability to adjust to what you want with a lot more available shortcut actions, while in Gnome you generally are expected to live with the choices of the devs.

Which I find to be a weird stance.

Gnome also believes that a window must have control over its own titlebar to draw it as it sees fit while simultaneously declaring it must not have control over a tray icon.

Also funny that Gnome seems to have objected to KDE proposal and wrote their own even though they seem to say point blank that while they are dictating how all the other DEs will do it, they themselves will be ignoring it. Why get in the business of a protocol you don’t even want to implement in the first place…

Mainly because gnome is harder to ignore than a lot of other opinionated DEs.

It’s been the default target for fedora and red hat, and like other choices rh makes, it propagates throughout the broader ecosystem.

Even if you ignore them, they dictate how Linux desktops are broadly allowed to work by largely asserting authority over FreeDesktop and by extension Wayland.

One of these is that they absolutely hate the concept of server side decorations, as a result even as they begrudgingly allowed it as a Wayland protocol, they insisted that it must not be mandatory and they are allowed to ignore it. This means applications that do not care about their decorations otherwise now must care about their decorations. As a user, the consequence is that any GTK application you might use is likely to just pop out as a gnome looking window among a bunch of otherwise consistent windows.

I did that for a while, but ultimately got too frustrated due to a few things:

• Having to explore extensions in the first place for fairly basic functionally that I would have expected gnome to naturally implement
• Every update would break extensions, and I’d have to wait for the extension author to update it, and about a fourth of the time the extension was abandoned and I would just have to go without or search for a similar.
• At their best I always felt a lot of the extensions were having to settle for a lesser experience due to limitations of the extension mechanism.

Ultimately, I found that experience I was trying to get to that gnome never would normally support and would evaporate on updates and never be quite right anyway was just a natural featureset of Plasma. So I just do that and haven’t had to sweat updates nearly as much as when I kept trying to make a go of it with gnome shell. I kept giving it a shot because of my gnome 2 experience (admittedly augmented by compiz), but gnome 3/mutter have not been a good fit for me.