We use Azure at work and it’s wild how much stuff is just manual labor behind the scenes on their platform.
Seems like it’s always on fire somewhere, ripe for burnout and MBOs aren’t aligned with reality from what I hear.
I remember their customer-side push for semantically merging Office and some other suites into something-or-other like 365 or Copilot, creating enormous confusion. I have difficulty believing MSFT distinguishes between handmade and machine-learning tooling in that type of environment.
The amount of times we've had to reach out to an experienced person who had transferred to another team/org about processes in our team was ridiculous.
AWS is no different. Everything is on fire all the time, but customers usually don't notice.
Above those layers is a ton of jank, e.g. Elastic Beanstalk. But my understanding is the fundamentals on AWS - compute, storage, networking - are pretty much rock-solid.
Disclosure: I work at AWS, at the core EC2 dataplane, no less. We monitor our production environment very aggressively with very tight SLAs, and our technical leadership is competent as well.
(Although I like the idea of Management BuyOut, too.)
If the docs are correct I'm sure they'll depreciate the offering and rename it to something else and leave you on an island either way.
to what extent, though? over the course of my working with azure, ive been disappointed constantly, by just about every service i've used. things that should work a particular way don't, and when you try to investigate you end up in a maze of out-of-date and/or totally wrong documentation. i've started a project before, followed their precise "quick start!" guide to the letter, and find it totally non-functional and wrong.
things that DO work end getting updated (often due to financial incentives) and break, or shuddered entirely. some values that you might want to change for certain services are hard-coded (which, to their credit, you can verify from their public github repos) for ostensibly no reason. a lot of their documentation is internal only, and the public facing stuff feels perennially out-of-season.
(opinion) i feel like fundamentally microsoft has embraced ai development so wholly because they've not been able to describe their services in documentation very well, and thus rely heavily on the ai's ability to digest what scraps of truth remain and the various source code to be effective at delivering output
But this is certainly exactly what I'd expect after indiscriminately laying off half your workers, then immediately offering a buyout, to make sure any experienced people left can find the door.
If this incident, or the difficulty they're having with remediation, are somehow not caused by any of that, then it's at least a preview of what things are going to be like.
What changed is the productivity per employee: 2021-2026: +60% 2024-2026: +38%
To be honest, I also assumed your take on this at first. But after a bit of research I found quite a different reality!
The total populace did not change, because it was replaced by new "talent".
Btw does everyone see how insanely crappy Windows 11 is compared to MacOS
The irony is that I can finally get out of the Microsoft ecosystem thanks to Microsoft. The only thing that was keeping me there before is Visual Studio (which runs reliably only on Windows). But with the advent of LLMs and Visual Studio Code (which is cross platform), I haven't touched Visual Studio in months, and doubt I ever will again. LLMs should also ease the transition into a Linux ecosystem. RTFM? Nah ATFLLM.
One has to spend more on third party tools for getting MacOS to feature parity with a windows than a Win Pro license...
Linux is fine on the other hand, but you compared to a joke of an operating system.
This is just an anecdote, but I have the exact opposite experience. I'm a C++ games developer working on Windows 11 every day and other than the micro frustrations with stupid UI, the system just works. I know I can rely on it day to day. But I was also given a MacBook last year for our iOS ports and I don't understand how anyone considers it a serious operating system for work - it's full of bugs to a point where I started using AI to work around some of them. I thought maybe MacOS 27 would fix some of them and sure it did - but then it introduced a whole bunch of new ones. As a developer it's probably the worst OS you could pick - if you hate windows for whatever reason go straight to Linux, at least you have actual control over your OS compared to MacOS where basic things like spotlight are broken and there's no way to fix them other than add to the thousands of complaints on apple forums.
Honestly, for personal machines the M series is wonderful hardware, and I have young kids so no time to tune Linux (I ran it at home for around a decade or so).
But yeah, the interface has many, many bugs.
My "favourite" (new) bug is that sometimes you can't move a window from the laptop screen to another screen. It goes away if you full screen the app, but it's bizarre to me that it made it to production.
Then due to typical clueless management and lack of taste and complete buffoonery, they just fucked it all up. I'm not complaining because I think everyone is better off, but it's just amazing to have witnessed that even when they luck out on a bet, they manage to fuck it up. Copilot started as a coding assistant, and before it could breath they destroyed the brand by shoving the name into literally everything without thinking or taste. Ignored the coding part, because that's lame, and instead focused on half assed Windows or Office apps and integrations that no one will ever use. Once claude code dominated the scene, they realized the major fuckup and tried to catch up, but it's too little too late.
Again, not complaining really.
I actually know quite a few (non-tech) people that love to use Copilot in those apps. So I'm not sure it's as big of a fuck-up as you think it is (or as I certainly used to think it is).
On profitable products though; sure as a general concept, but the rate of return, opportunity cost and market competition factor into the profitability equation too. You could go waste your time building a desktop app for a chatgpt clone and make some "profit" in the same time your competitor threw similar resources at making a claude code and that makes 100x your profit. That would mean you actually lost despite the profit. The opportunity cost isn't only in building the wrong thing, but also in the human resources you allocated to the task that deprived other areas. I see no world where the 6 dozen copilots we got from Microsoft have any good ROI beyond diluting a brand and making them more of a laughing stock.
We need to rewire ourselves to think about things that are beneficial to society as a whole. Not "profitable." Profit just means more money going from the many to the few. That's what's gotten us into this awful situation.
So fortunately, there's not really an incompatibility there.
Windows does always cater to backwards companility and making IT departments happy, so that explains some of the lagging technology and systems. I can't believe NTFS is still the default file systems. That's generations behind Apple's APFS.
The stock has popped, and that's what Wall Street cares about, but I could see a world where this chronic underinvestment and mismanagement basically implodes the entire company in the coming years.
New Xbox to cost $1200-$1800.
Gamepass limited to 15 hours a month for Ultimate.
New Xbox cost is not confirmed.
It works perfectly on a fast home connection over ethernet, mind blowing how responsive it is and load times are non existent.
But as far as I can tell most people subscribe to game pass anyway? Why buy the hardware at all when I can just cloud game in the same subscription. .
Here, Series S is £400-£430 for 512GB or £500-£520 for 1TB.
Xbox Series X is £600-£620 for digital 1TB or £50 more for the disc version. That's much more than I paid for mine a year after launch (£450 I think), but luckily still nowhere near 1700 Euros.
Indeed you’re seeing the consequences of their past decade+ of management
A quick google search turned up these global sales numbers: PS3 87.40 million, PS4 117.20 million, PS5 96.44 million. Xbox 360 85.73 million, Xbox One 57.96 million, Xbox Series X/S 35.68 million.
It seems to be back in stock now, but at a cost of about 1000 Euros for a Series/S I'm not really feeling enticed even if GTA 6 is around the corner. They probably really shot themselves in the foot by not having units out a year ago and now fewer will buy due to pricing.
I think Sony is shooting themselves in the foot with the digital only consoles too - unless they maintain a very attractive set of exclusives at all times, people will just choose whichever digital subscription is best or cheapest for them instead of something tied to a particular console.
They need to sell consoles to get people to become subscribers.
Nothing to do with today’s outage though.
Xbox Game Pass was an attractive offer for consumers when MS was getting people onboard and the price was so heavily subsidized that it could never have been a sustainable model. Once they tried to start bumping the price into something profitable people immediately said "okay well that's clearly not worth it, I'll just cancel."
They attempted to make it $30/month last year before backtracking. At $360/year, even factoring in the new $70 price of AAA new releases and ignoring you can wait a few months to buy them cheaper, you had to be buying 5 new release games per year at launch just to get close to breaking even with the subscription price. 6 before you save money. So...
> It's a really attractive offering for people who don't play games all that much (that is, more than 2 hours a day)
No, it's absolutely not worth it for these people. Even at the current prices they backtracked to ($280/year if you want to play new releases on day 1), if you're a person who doesn't game that much, it's cheaper to just buy a few new games per year. If you're not a frequent gamer, why would you of all people need access to a massive library of games?
I personally have a backlog of 45 games I want to play written down in a note. And those are just the ones I wrote down. Most of these games you can get for peanuts. I just played through Fallout New Vegas for my first time. It regularly goes on sale for $0.99, or $5 if you want all the DLC. I got 90 hours of playtime out of that.
The idea any casual gamers should be paying for a subscription service to a Netflix-style game library is nuts to me. These aren't movies and tv shows you'll be flying through every evening, they're big games that can take weeks to play even if you're putting in a few hours per day.
I see what you did there.
Fabric is a weird one, as it's sortof a GUI over spark, but also PowerBI and lots of weird services.
It feels like a "data plaform" for people who've never used one before and only have the vaguest idea of what it means.
That being said, Spark is pretty good and we're running a lot of workflows on it (and probably building more on top, I suspect that I'll be much angrier about it in approximately 6 months or so).
Ui is clunky, they are terribly slow.
The only reason companies use Entra is it appears cheaper but eventually you pay as much as competitors. Okta is a way better IAM solution if you need one. Although it’s very pricey.
Yeah, and additionally if all your counterparties are on it, you can externalise your internal tools onto them relatively easily.
This mostly doesn't work, but it's a beautiful dream for executives.
I agree. All the execs need to do is tell Copilot to "fix this so that it never happens again. Make no mistakes" and run it on bypass permissions. Tada!
Maybe they can use azure hosted openai models? Or are they down as well?
Just was a perfect opportunity to showcase astra and fable and other models
If this continues the company will see a lot of customers leaving. And they were already in a bind since Europe is finally realizing that its dependence on American tech could be used against them.
They're not corrupt (in a lot of cases), mostly just ignorant. Like, they don't understand the tech, and microsoft is safe, they're already using it so handing over large swathes of their infra to them seems reasonable.
MS have wonderful sales people, and this is why they do well with governments and large enterprises.
There's also a co-ordination problem, in that all of your counterparties (like your citizens, for instance) are familiar with MS and therefore it costs a bunch locally to move, for little short-term reward.
This is actually a place where (good) regulation could derisk a lot of these changes, but you'd need a committed group of politicians who can convince the civil servants to keep at it. It'll happen, but slowly and in starts over the next 5-15 years.
If they transfer this to sovereign providers it would cost Big Tech serious money. So they're quietly lobbying to slow down the speed at which European nations are moving away from them.
Yeah, this is definitely really important to them (not important enough for them to comply with the laws, but whatevs).
This absolutely predates AI.
I think this is the logical conclusion of using copilot to write your stuff for the last two years. Nobody knows what it does and more and more bugs are being solved by just pointing the AI back to it again. Combine with the fact that so many people have been laid off and replaced with copilot I'm not surprised that the few who remain aren't quite sure how to solve it