Compatibility is still an issue, because the amount of just weird that people run on Windows boxes, you can't keep up. On macOS, almost everything is ported over. Rosetta is just "give a shot if you really need something that old" safety net.

Panther Lake really closed the gap on battery life. I've seen some real life reviews and it remarkable how much better it gotten on the Intel side. Of course, there is still a gap to ARM (both Qualcomm and Apple Silicon), but it's not large enough to put up with the other hassles. Right now, the whole Windows on ARM thing is keeping Intel honest and in all fairness, that does seem to be working.

From a development standpoint, what a absolute nightmare. Cross-platform development is so hit and miss. Will a library support Windows/MSYS? It does, great. Try to build it for ARM64 windows target and more times than not, it just implodes. Heck, even ARM64-linux cross compilation isn't a given. The toolchain is much further along for sure, but it still a hassle.

And a salute for all those keeping macOS support going for those libraries out there. It's thankless stuff.

Your argument all rests on absolute average wages ($7 vs $30) and ignores that the overall purchasing power for Americans hasn't increased for decades (https://www.pewresearch.org/short-reads/2018/08/07/for-most-us-workers-real-wages-have-barely-budged-for-decades/).

A big part of this is that core essentials haven't gotten that much cheaper overall (and some are rising rapidly). Food is more expensive and energy and water costs are growing and have largely kept pace with inflation. The cost savings in other goods just is not enough to meet the increasing size of a down payment to secure a home.

Housing has been outpacing inflation for quite sometime. This is includes rent as well as home costs. As the overall cost of housing becomes a larger precents of wages, it takes longer and longer to get a house and fewer and fewer households have the ability to save enough to meet those rising costs. Especially as rent is outpacing inflation and wages as well. (https://home.treasury.gov/news/featured-stories/rent-house-prices-and-demographics).

There is the trend of more and more single family housing being acquired by private equity to then rent out that puts additional pressure on costs and inventory and again, causing housing to outpace inflation even considering amenities. (https://www.npr.org/sections/planet-money/2025/09/09/g-s1-87699/private-equity-corporate-landlords).

Garages are cheap, A/C is an affordable add on to central heating and since central heating is often demanded by code, it makes little sense not to install it.

Add things like the costs of childcare growing in costs as availability is going down. (https://www.census.gov/library/stories/2024/01/rising-child-care-cost.html)

There are ton of headwinds in this economy that makes delaying or not having a family make more and more sense for an ever growing set of our population.

I personally think a lot of these need to be addressed with programs to increase the supply of affordable housing, eliminating corporate ownership of single family houses and additional controls to prevent rent and landlord abuses.

I also think universal health care and subsidized child-care would reduce the risk of accumulating crippling debt while put more money in people's pockets and giving them to ability to shift positions and increase wages without the threat of losing critical benefits for their family.

So, yea, just a woke snowflake here. Sorry if you made it this far.

Just so long as they can accumulate all the actual money and control for resources and dole them out to us as they see fit, because, governments, who needs that hassle, right.

There are great reasons for labor reform that have zero to with AI. Four day work week, mandatory overtime pay, the list is long. But, less billionaires, can't have that.

One issue with the overall architecture (which is just statistical prediction) is that it can't really provide useful insights on why it did what it did. Which was requirement of expert systems back in the day.

Honestly, it seem like building better static analysis tools for finding these kind of problems is a better way to go overall. The tools could be more relaxed on reporting potential issues and allow more false positive versus focusing on reporting things that are certainly bugs, but still be based on a set of rules and patterns that can be used to explain the error in a consistent way.

If the study reported on how much time was spent using Windows overall versus Macs. Because that makes a big difference. If there was, say, ten times the usage of Windows than MacOS, then affects the overall uptime.

Without this, the numbers don't have the context to really make a lot of conclusion. Not to mention that this probably all self-reported numbers as well.

Nothing like being a junior developer at a company whose business relies on getting rid of roles *exactly* like yours.

Then again, there's completely useless stock options that you can pretend will pay off big time.

So, this argument all pivots on this theory of extended mind by two philosophers, a theory that (looks around) isn't widely accepted at all in their specialized field of studies.

Let's compare this to the growing body of negative effects of constant access social media on children and the related ethical concerns.

If Gen Z decides to learn how to use maps more, keep written journals and do basic math with a calculator and make friends in real life that's a overall win for everybody.

They will still have access to computing when they need to use it. Again turning computers back to their roots as a tool, not as a gateway for serving constant distraction and near-zero-quality media is fine by me.

Yes, those are the arguments that some are making about AI. Frankly, they are more credible than a CEO essentially saying that they are hurt some people don't just trust companies like them to do the right thing. With the numbers being thrown around, skepticism is required here. It's bad enough when our government throws massive money at military projects with very dubious actual value. But the potential labor disruption and further accumulation of wealth at the very top that is possible with AI, it is irresponsible not to have the highest level of oversight and regulation in this case.

Oh, and some of the people you are making unhappy are the very demographic that brought NVidia to where they are. If AI is so important, NVidia should just commit 100%. Announce you are leaving the consumer and low-end workstation GPU market entirely.

It's a trending topic that's easy to get quite a few clicks from, so I expect a lot more articles and content to continue for quite some time.

The biggest advocates of Linux don't want the "year of the Linux Desktop" and the homogeneity that it would bring. It's the choices that they have that makes the platform interesting to them.

Of course, that means most users have to make choices they don't want or even understand.

Won't create a demand for highly memory optimized versions of applications. The costs for rewriting that code in desktop specific environments is very high and the inertia behind web frameworks and applications is so large that nobody wants to move to something better, despite how bad the current frameworks can be.

Most users expect more out of all their applications. Application frameworks have to deal with monitor scaling, use very high quality font and vector rendering and stay highly responsive. They use more animations to provide a better visual experience. This means you have to cache more complex visual assets and have more memory overhead. Older X11 and Win32 applications just look bad in comparison.

This is all due to a irrational land grab for computing resources that just happens to make it more expensive to build or even own your own computer.

Developers aren't the problem here, it's the out of control economic forces that want to consolidate even more wealth and resources in fewer and fewer places on the backs of increasingly strained labor.

To potentially be in a worse place than you started. But, hey, sunk costs and all that. Honestly, what is the savings really going to be after Google cranks up your per-user costs when they have captured enough of your data so that you can't switch back?

Microsoft makes money on Windows and Office because it is good enough for the job and has a decent ROI. It's their core business. Sure, they want you on Azure. But, hey, you want on-premise? Happy to sell it to you as well.

Google Workspace could be left to languish in the dust and it wouldn't affect the bottom line that much. Bonus, all your documents are strung around on their servers.

And no, you aren't getting rid of those spreadsheets anytime soon. They can take full advantage of a local machine and they make sense to the people that use them. It costs less to just keep the machines and licenses up to date versus trying to make an application that is essentially a specific spreadsheet, but has to run on pretty expensive computing resources for no actual gain.

Being able to run a local client on a local copy of a document is a good thing.

Multiple tabs are good. Markdown support is fine. And, of course, just more marketing shoving AI into every crevice.

There is a good OS lurking around in Windows, but, it is buried under too much slop.

Here's hoping a true Pro version becomes a thing.

I use Windows as my daily driver (it is what works best for me and I came from Linux) and AI is a big distraction for them. The CEO has become obsessed with AI and how them not being leader threatens the whole company and is pushing AI where it doesn't want to be and where nobody wants it.

Eroding Windows and Office with AI slop is a much, much larger threat to actual business model than OpenAI, et al. will ever be. Microsoft is fortunate that they have the opportunity to go backwards (with a few new things and fixes) to keep market share.

Making a much less-sucky Windows Pro isn't hard. Fix backup and update image backup and restore. Undo the dumbing down of OneDrive; don't sync any folders by default, but allow sync of any folder. Allow local accounts, make OneDrive and Copilot anything opt-in via installing from the store (or winget). Get any web stuff out of the shell whenever possible. Get Control Panel replaced with a workable solution.

Way easier and cheaper than gluing in AI slop willy-nilly, I would think.

While it is easy to hop on the bandwagon, this isn't an issue that strikes at the heart of the shell. It was a set of app packages that didn't get setup in time. There's already a workaround (run a script to register the packages). A notable mistake, but not a fundamental flaw.

Microsoft management buying into way too much AI hype; that will ruin Windows and Office. They have to hope there are enough developers to keep the AI rot at arms length so they can just remove it and you know, just make things work better?

Java is adopting a lot of C# features, with varying success[1]. As such, more developers are seeing the advantages of those features so why not just use C# where it is way more refined. The cross-platform story is pretty complete. The runtime and the tooling. VS Code works. IntelliJ fans have Rider. The dotnet command tool does the things you need it to.

Some things in Java are just clunky. Building and package management can really become a nightmare. You have choices where you don't really want them.

Microsoft is doing a better job stewarding the platform than Oracle, and as unpopular as Microsoft can be, it is nothing compared to the ill will that Oracle has brought. Microsoft has unhappy customers; Oracle has paying prisoners. The main challenge Microsoft has to overcome is the older versions of the platform itself. As an example, some developers think Entity Framework only works with SQL Server. That was never the case. EF Core works swimmingly with a lot of databases, with Postgres support becoming a notable highlight.

[1] Choices like type-erasure for generics, not having auto boxing and unboxing really makes for some clunky APIs. You have things like

IntStream mapToInt()

C# has

IEnumerable<T> Select(...)

where T can be int, string, whatever. Async and await alongside IO libraries that are non-blocking by default is another win. Sure, function coloring, but it's been a feature for ten years and it just works.