This is the correct take. QA was under attack the entire Agile(TM) cycle we just exited. Now with agentic coding, the only thing keeping it on the rails is TDD. Either you write the tests or you write the spec.
You are also responsible for the output. Welcome to the New Age. Management is just as clueless as they have ever been (some more than others) and yet most of them lack the intelligence to know exactly how it all works. Hell, there’s still plenty of engineers that don’t know how it all works.
Eventually, when you come to understanding or you reach that “enlightenment” stage, no corporate BS will penetrate you and you’ll forever see past their shenanigans. At this stage though you’ll be a grey beard and be unemployable. So they cut out anyone who knows the BS to bring in folks who believe the BS so they can continue shipping BS.
Another elephant in the room is the widespread impact of neverending COVID re-infections of people who don't wear N95s in public (most people). Vaccines don't prevent transmission, which leaves everyone open to acquiring long COVID. Long COVID is very likely underdiagnosed due to widespread ignorance, and not helping matters is that 40% of infections are asymptomatic during the acute phase.
Long COVID can include issues with memory and risk taking.
I don't think software is the only field impacted by this, but it's undoubtedly one of them considering how few people take proper precautions via regular N95 usage.
COVID in general is also heavily politicized (as seen in the silent downvotes), and the back-to-normal agenda was forced by capital interests, including tactics like mass disinformation campaigns.
It's very simple; If you work 10x faster, you need to release 1/10th the bugs as before for the user's experience to have the same level of bugginess. And we're not incentivized for this; we're still exploring how to ship 10x faster and not exploring how to meet our users' needs 10x better (often conflating the two!)
And I mean, go tell your leadership "Those projects that used to take us a full quarter? Now we can do them in 6 weeks." You don't get the rest of the quarter to stabilize your codebase. Now you jam 2 releases into the quarter.
I thought about writing a comedy, set in the far future, where everything looks beautiful but nothing works right and it's just accepted (almost unnoticed) as a running gag throughout the film. Now I wonder how prescient that would be.
Yeah, something like that but more utopian and closer to Jacques Tati's Playtime.
And I think that's what would make it funny: the affordance of the efficiency in design creating that expectation of a correctness in delivery which is slightly subverted.
ALWAYS for basically any app or cloud service based software. I run into multiple bugs ALL the time.
In my day job as a network engineer, I see some of the wildest one off bugs in embedded firmware and networking devices that I honestly can only explain as possible CME interference xD xD
But with all honesty the `make make make` culture jozfar mentioned + companies pushing out only 80% completed products at best is probably what we are seeing.
I keep joking about how 2025-2026 we have some of the most amazing tech I have ever imagined being real, yet the hugest amount of just purely broken software all around that most just never seems to fully work the way its supposed to for one reason or another!
Other times I just start to think I am a bug magnet xD but glad to hear someone else is seeing the same thing
My Linux experience gets more polished every year. Last serious issue I remember is a AMD GPU kernel crash around 2022(?).
I feel like Android has always had its glitches. I can live with the current set, but each update has me nervous about the new set of bugs that will appear.
Linux is indeed getting better, but all the commercial software (and open source projects that happen to have AGENTS.md (cough Vicinae cough)) keep surprising me with random bugs
I think software has always been terribly buggy. Back in the day, things would crash (including your entire OS in many cases) all the time. These days, it's more often smaller, stupid UX issues, but that's been the situation for a few years. I've seen no difference over the past six months.
On the one hand, surface level bugs do appear to be more common across the board. Like you almost wonder if anyone has ever test-driven the software even once before hitting deploy.
Operating systems and critical software does appear to be significantly more fault tolerant these days though. If you go back to something like Windows 95, it would BSOD with a frequency that is much higher than anything we're accustomed to these years.
I guess things keep getting more features and complexity but all of it is 80% done because it's "shippable" and updates are cheap.
This is most visibile for me on Windows/MacOS and complex web apps (e.g. GitHub and GitLab, including consoles of hyperscalers), where 80% of "normal" things work, then you need the last 20% and it's always not working as documented, only half-working, or just outright broken, and you need to find "temporary" workarounds that stay in place for years.
I feel this is being amplified by AI: tests come last (if at all) and are still written by LLMs, nobody really looks at them anymore, green pipeline checkmarks mean less than they did in the past.
We've always had bugs. Good chance that it's largely confirmation bias - you're familiar with how software development has changed in the past 6 months, and especially if you don't like it, any time you see a bug or a problem, you'll attribute it to that.
My running joke for the last couple years is "Nobody knows how to write software anymore" (even about a bug in my code). The problem is almost certainly complexity. Apps are like a house of cards, and we keep adding to the stack with new features, external dependencies, etc. My apps 30 years ago used one language, one library, one process, one thread and shipped on one platform. Now it uses Java, JavaScript, CSS, HTML, SQL and has a dozen linked libraries, talks to a couple servers, has dozens of threads, and runs on desktop, browser, tablet, phone and car console. It's a wonder anything works at all.
Also there is trouble with back end IT systems. Like I had some trouble with my homeowner's insurance which was initially my fault [1] but my insurance agent was gracious and helped resolve it and my wife paid her right in her office. Then they do some software update and in the process cancel my policy and I get a letter from my lawyer [2] wanting to collect $82.54
[1] although if I was trying to design a spell to make people not pay their bill I might try sending them 20 letters that say THIS IS NOT A BILL before they get the bill
QA got it so bad in the current LLM-fueled carnival of stupid. Also, what would be categorized in the times of yore as a "lamer" has incredible free tools at their disposal, so vulnerabilities are a lot more common. Add to that the fact that every goddamn thing is connected to the internet and you got a recipe for disaster.
Even if you only hit 1/100 times... if you can shot 10000 times, you're still in the green.
Yeah substantial degradation in really basic functionality in the last 3-4 years.
I mean, new features I don’t use and useless UI changes have been added at a high rate over the same time period, but holy Christ, it’s pretty bad when some really basic functionality bugs are getting shipped by big names all the time. Like adobe PDF reader is… an absolute hot pile of garbage.
Yes. My observation is that FAANG culture persisting the myth that testing is slow or writing clean/consistent code is slow (conveniently playing to the dev's ego, but I digress). As a result, a lot of bugs get shipped, inevitably triggering a rework cycle. But what does not get measured, does not get reported. As a result devs look busier than ever, great for their reviews, while quality dives.
A turnaround in the industry would be actually capturing the rework cycle into costs.
There has been a steady decline for decades. Release early, release often. One example: Apple used to have a thing called Gold Master release when updating software was not as immediate. The GM programs met a high bar for quality. Not so much anymore… the attitude is we can just release a new version later this week. There also used to be a thing called backwards compatibility… now no one thinks twice about breaking code on older platforms.
I've been using windows for the past 5 years or so. I was using Linux before that. I used to complain about things breaking in Linux but the situation in Windows is way worse than it ever was in Linux. A couple of days ago my co-worker turned on his computer and windows has locked him out somehow. He had to disable secure boot, recover his bitlocker passphrase from his microsoft account, and only then could he get past the boot screen. The theory is that his laptop turned off while updating. I'm thinking, why doesn't the update abort when it realizes it's not plugged in or on low battery. He said it had never happened to him before and I have also never seen it.
The other side I notice is with google products. Barring android, GCP, and the less than yearly occassional outage, I would never see a bug in google products. Now I notice them often in drive, sheets, gmail, and specially meet. The other day I joined a meeting and all participants couldn't log in. NEVER have I seen such a bad production bug from google.
My two Win 11 machines have been pretty solid for a while, but my Ubuntu Linux machine has had a lot of trouble with software updates.
When I built the machine about 12 years ago [1] it allocated a /boot partition that was big enough then but not big enough but as Ubuntu grew the /boot was too small to keep as many versions as Ubuntu wanted so kernel upgrades needed manual intervention. And for that machine it is a hassle because it runs headless and if it doesn't come up with the Ethernet working I have to pull it out of it's home and set up a workstation.
More recently I rebuilt / completely (kept the ZFS array) and this time Ubuntu didn't make a /boot so the files live on / but now when it does software updates it often fails to install the realtek driver modules so it comes up without the Ethernet and I have to take it upstairs to rebuild it. Copilot offered me a script to tell if the modules are busted after a kernel upgrade so I can do the the manual fix before the reboot now.
I guess I can't complain too much, I mean this $150 machine that I upgraded with 32GB of RAM and a NVIDIA GPU is going strong, though now that I think about it the second ZFS array I built for it is past warranty even if it only half full.
[1] I know because I named it after my favorite vixen from a video game I was playing back then
Software quality, i.e lack of bugs, from the average user's perspective, is a hygiene factor. High quality does not get rewarded, and the instant it crosses below the threshold of "good enough", users will punish you.
The eternal question for management seems to be "what can we get away with?".
A dev with quality problems in their output? Can em'.
A dev with exceptional quality in their output? Can em'.
A dev with just the right corners cut? Keep 'em, but no raise.
Is this regime right or fair? I don't think so, but it is the incentives that drive all of this.
I believe the only solution to this, at least in the for-profit software world, is to try to convince everyone to pay for quality. If the money is there, then maybe management will reward devs who care about quality user experiences.
It's similar to the ElectronJS desktop app problem. Sure the desktop apps use a lot more system resources, but it's doubtful those apps would have been created in the first place if it wasn't for the speedup of development allowed by ElectronJS.
In this case, sure software is buggier, but there's also a ton more software+features that wouldn't have existed before AI coding tools.
Electron apps are selected for things that are harmful to your user experience.
That is, if you know how to make Electron apps you also know how to make web applications and web applications are generally superior and preferred by users. No install, no bloated runtime, "just works", say it again "JUST WORKS!"
Any time somebody makes an Electron app there is a reason why they make an Electron app and it usually is a bad reason. For instance, Electron apps can live in your tray and make your computer harder to use by hiding behind twisty little icons that all look alike, popping up and covering other things that live in the tray, etc. Defying the usual procedures for starting and stopping apps, finding and moving windows, etc.
And of course the whole point of living in the tray is running all the time, sucking RAM all the time, sucking CPU all the time, sucking GPU all the time, sucking network bandwidth, etc. In short, sucking.
I mostly feel this on Microsoft stuff. O365 browser stuff was always buggy, but last month I had to fight to zoom in on something in PowerPoint. Pagination in Azure ML randomly broke, it took them a month to fix it.
Atlassian is also a bug factory, but I did not expect anything better of them. Each UI update they did to Confluence and Jira just made everything more confusing in the last 10 years.
Holy smokes is the browser versions of PPT, OneNote, word etc buggy, I could not believe that was an actual released product. Random page crashes, refresh loops, etc, all for the crime of using firefox on a mac. The bottom of their test matrix sure, but not exactly netscape on windows XP.
I don't know about just the past 6 months, but past 10-20 years, sure? I think software today has to handle much more varied and complicated I/O at a much bigger scale than it did in the past. Website UI/UX has to accommodate not only numerous browsers, but numerous devices with different viewing spaces. Think of how much broke when Python was rejiggered to not implicitly typecast unicode data. Think about all the legacy databases that can't handle spaces or various non-alphanumeric characters.
The userbase for software in the early 2000 was built for a much, much smaller audience than it is today. And I'm not sure the huge improvements in testing/debugging and general software development would totally mitigate that, in the same way modern military defense systems is still way behind modern offensive weaponry.
I use cheap android devices. My current one is a cheap Motorola and before samsung and Chinese vendors. I've never seen bugs like I have encountere since the last update (2 months ago).
Unless I reboot the device everyday, I can't make a call. The button is pressed but never proceeds to ring, doesn't even show an error message. The play service complains that I am logged out but shows my account info in the login panel!! Rebooting is the only way, android is like windows currently.
I am just scared to update anything these days. And not just android apps, Even rsync gave me a minor scare (fortunate my backup jobs are on debian stable).
I had one encounter with an obviously vibe coded website last year. Everything looked great, but nothing worked. Had to deal with the company over email.
Things might be better now, but that was a clear warning of what future may hold.
It's definitely been buggier for a while now. It's already been an issue for a decade or so due to the "release minimum viable product, fix problems with patches/updates model" (especially in the world of video games), but the rise of AI has only made the issue more apparent. Microsoft is the biggest example here (with Windows 11 seemingly screwing up something massive in every update), but other companies like Google and Meta haven't been as reliable either.
"weak computers" is the funniest part. The computer I'm typing this on *right now* is orders of magnitude faster then the Athlon XP processor that i was using in 2003, but it doesn't feel much faster. I think that hardware advancement is just slower then software bloat...
Not only buggier, but less usable overall. We can not blame everything on AI. For example events discovery platforms and ticket platforms were way better decade ago, you could search by popularity, friends going to the same event and there wasn’t as much fragmentation. That part you could blame on Cambridge Analytica scandal and late capitalism love of algorithmic feeds
Websites and mobile apps, definitely. Every time a typical proprietary mobile app updates I wonder what they will break. They always do break something or remove a feature I want.
Desktop apps, not in general. Some work better, like KDE's DE has fixed some irritants.
It's pretty hard to separate this from confirmation bias. But I've had to disable the YouTube app and switch to mobile web because it would just buffer nonstop on everything; never had issues like with it like this before.
uThermal has been documenting in shorts the myriad game breaking bugs introduced into StarCraft 2 in recent patches. The game was super stable for 15 years and now you can load 8 sieged tanks into a medivac at a time. Anyway I am seeing the same thing as you in many different arenas
Yes, I initially jumped from Android to iPhone back in 2019 or so because I felt like the experience was much more stable and it just worked, smooth, no issues, kinda boring but in a good adult way.
Now every interaction makes me want to throw the phone out the window.
If it’s not just blank, the keyboard doesn’t come up, the navbar in Safari sometimes is there, sometimes it’s a blank rectangle, etc. “ask your developer to update the apps to run well in iOS 26” is such a garbage response. We didn’t need to break everything just to have some glass simulations that make everything slower and add no value.
Google Home: holy shit dude. It was a mostly working thing. Absolutely unusable with Gemini. Will get the transcript right 15% of the time. Can’t even ask for the weather.
Roku: Netflix playing in the background but I’m still in the main menu.
iOS is definitely buggier. It used to be a silky smooth experience but it's degraded significantly in the last year or so. Safari on iOS, in particular, is very bad. Lots of problems in Adobe Creative Cloud on Win 11 (primarily Photoshop, InDesign, Illustrator) in recent months too.
This might sound like a rant, but it's more an attempt to outline the breadth of the problem by way of many examples. Software quality sucks right now, enough so that I plan my life around it. Payment processing is regularly broken? I'll maintain a couple cards from different providers. Login systems are always broken? If you require a login then I'm going to weigh the quality of your service against not just its physical dollar cost but also the amount of my time you're about to waste if I sign up. Etc.
Enterprise broke something and has been unable to text or email me receipts and reservation information for months. No trouble; I'll look up my information manually on their site when I want to modify a reservation. Great, I've found that information, surely I can modify it? Nope; that's broken too, and it asks me to call and doesn't provide the number I should call. So...I'll look up their customer support number, surely that isn't broken? Nope, somebody deployed a version of the site where I physically see the handlebars and variable name which should have referenced the phone number rather than its actual value, and they seemingly didn't notice or weren't able to roll back the change for at least a day.
And that's not to pick on them (especially because they're otherwise a good company IME), but to highlight a series of technical fuckups which never would have flown in past years, and they're not alone. My HSA just made it impossible to log in without buying a new phone, every single time I've tried to set up automatic payments for Visa they've had a different bug on their website preventing that, WalMart supposedly allows buying things for people in other states (I was buying my sister tires), but some broken AI will reverse the transaction hours later. eBay and Visa are even better on that front; their customer support seemingly aren't able to override the broken fraud-detection AI, so there are some Visa purchases I can't make, and every time I log in to eBay my account credentials are forcibly reset a day later.
Beyond the straight, obvious incompetence manifest in 3+ major bugs in nearly every website or service I interact with, there are also stranger bugs (no "strangest" bugs in recent memory). One of my coworkers has a reliable network blip every 20s (my memory is fuzzy, it might have been 10s or 15s when I timed it, but it's incredibly consistent), only hitting their outbound traffic, and it's at or upstream of the modem. One site I visit has some sort of broken filtering from my mobile carrier, but only for their login page, so I need to use a VPN to log in, but then the rest of the site has VPN detection, so I need to leave the VPN to access the rest. A little while ago, I stopped being able to accept calls while my wireless hotspot is active. Slack automatically moved precisely one of my DMs from the section I had assigned to it to a different section. Discord and Target both had an automatic retry on a failed telemetry implementation without backoff which caused you to immediately violate their rate limit by just visiting the site.
And so on. Major companies are writing broken software, missing the bugs in review, missing them in testing, don't have monitoring or alerting on core metrics to be able to roll back when their deploys suck (or just can't roll back), usually have no way to contact them to inform them of the mistake, and sometimes choose not to fix bugs as big as "many of our customers can't pay us" in days, sometimes ever.
Yeah man, I think this exactly captures what I've been feeling - every time I need to get something done, I consistently hit some experience breaking bug. And you're totally right, I've actually just come to expect it and work around it...
Do you think in the future apps that are stable and "just work" will be a moat? Feels like the direction we're trending in.
I thought about proposing that idea myself and decided against it. Consumer behavior is a bit hard to predict, and it's a little unclear to me how well "just works" will sell. I like to believe that's viable, but I don't really know.
Why is this silently downvoted? The OP didn't give any details about their devices or specific bugs. I think my conclusion might be correct, given that the corporations spend less on the user convenience/UX as a part of the enshittification trend. Also confirmed by this comment here: https://news.ycombinator.com/item?id=48920698#48924033.
I feel like no one has said the elephant in the room.
QA teams were fired/never hired in the first place (put onto the Devs/support/customer to report and test)
Management want features and selling not Lovability and polish. We are just hitting an all time of make make make.
There's a much larger elephant in the room
At this point there's a room in the elephant.
Ooh ohh Ohh I think I know what it is!
This is the correct take. QA was under attack the entire Agile(TM) cycle we just exited. Now with agentic coding, the only thing keeping it on the rails is TDD. Either you write the tests or you write the spec.
You are also responsible for the output. Welcome to the New Age. Management is just as clueless as they have ever been (some more than others) and yet most of them lack the intelligence to know exactly how it all works. Hell, there’s still plenty of engineers that don’t know how it all works.
Eventually, when you come to understanding or you reach that “enlightenment” stage, no corporate BS will penetrate you and you’ll forever see past their shenanigans. At this stage though you’ll be a grey beard and be unemployable. So they cut out anyone who knows the BS to bring in folks who believe the BS so they can continue shipping BS.
I agree with most of what you said, but this...
> At this stage though you’ll be a grey beard and be unemployable.
isn't as true as people think. I'm a graybeard and remain very much in demand, as do the majority of the graybeards I know.
YMMV but the vast majority of greybeards I know are unemployed or are doing other things than software engineering now.
Another elephant in the room is the widespread impact of neverending COVID re-infections of people who don't wear N95s in public (most people). Vaccines don't prevent transmission, which leaves everyone open to acquiring long COVID. Long COVID is very likely underdiagnosed due to widespread ignorance, and not helping matters is that 40% of infections are asymptomatic during the acute phase.
Long COVID can include issues with memory and risk taking.
https://www.camh.ca/en/camh-news-and-stories/rsch-new-study-...
I don't think software is the only field impacted by this, but it's undoubtedly one of them considering how few people take proper precautions via regular N95 usage.
COVID in general is also heavily politicized (as seen in the silent downvotes), and the back-to-normal agenda was forced by capital interests, including tactics like mass disinformation campaigns.
Sir, this is a Wendy’s.
Absolutely yes 100, I've been saying this for like a year now! Even things that were good and usable before are getting really weird quirks.
It's very simple; If you work 10x faster, you need to release 1/10th the bugs as before for the user's experience to have the same level of bugginess. And we're not incentivized for this; we're still exploring how to ship 10x faster and not exploring how to meet our users' needs 10x better (often conflating the two!)
And I mean, go tell your leadership "Those projects that used to take us a full quarter? Now we can do them in 6 weeks." You don't get the rest of the quarter to stabilize your codebase. Now you jam 2 releases into the quarter.
I keep joking the same joke every time Space Marine 2 crashes my entire Linux desktop.
> with current state of the software, I imagine that "tech priest" is imminently going to become a real job
no really praising Omnissiah sometimes helps me keep both SM2 and the browser open
> every time Space Marine 2 crashes my entire Linux desktop.
Sounds like a good situation for the litany of percussive maintenance.
I thought about writing a comedy, set in the far future, where everything looks beautiful but nothing works right and it's just accepted (almost unnoticed) as a running gag throughout the film. Now I wonder how prescient that would be.
The Film Brazil, but the environment is designed by Apple.
Yeah, something like that but more utopian and closer to Jacques Tati's Playtime.
And I think that's what would make it funny: the affordance of the efficiency in design creating that expectation of a correctness in delivery which is slightly subverted.
That would be a nice one.
ALWAYS for basically any app or cloud service based software. I run into multiple bugs ALL the time.
In my day job as a network engineer, I see some of the wildest one off bugs in embedded firmware and networking devices that I honestly can only explain as possible CME interference xD xD
But with all honesty the `make make make` culture jozfar mentioned + companies pushing out only 80% completed products at best is probably what we are seeing.
I keep joking about how 2025-2026 we have some of the most amazing tech I have ever imagined being real, yet the hugest amount of just purely broken software all around that most just never seems to fully work the way its supposed to for one reason or another!
Other times I just start to think I am a bug magnet xD but glad to hear someone else is seeing the same thing
I haven't notice any significant changes.
My Linux experience gets more polished every year. Last serious issue I remember is a AMD GPU kernel crash around 2022(?).
I feel like Android has always had its glitches. I can live with the current set, but each update has me nervous about the new set of bugs that will appear.
Linux is indeed getting better, but all the commercial software (and open source projects that happen to have AGENTS.md (cough Vicinae cough)) keep surprising me with random bugs
I think software has always been terribly buggy. Back in the day, things would crash (including your entire OS in many cases) all the time. These days, it's more often smaller, stupid UX issues, but that's been the situation for a few years. I've seen no difference over the past six months.
Yeah in Windows XP time if my computer didn't crash several times a week, it was odd.
Ironically Windows XP was the first "consumer" (i.e. not "workstation") Windows version to not crash several times a day.
In the 98/ME era, it was pretty common to reboot every couple hours or the system would randomly do that for you.
I think there are two sides to this.
On the one hand, surface level bugs do appear to be more common across the board. Like you almost wonder if anyone has ever test-driven the software even once before hitting deploy.
Operating systems and critical software does appear to be significantly more fault tolerant these days though. If you go back to something like Windows 95, it would BSOD with a frequency that is much higher than anything we're accustomed to these years.
I guess things keep getting more features and complexity but all of it is 80% done because it's "shippable" and updates are cheap.
This is most visibile for me on Windows/MacOS and complex web apps (e.g. GitHub and GitLab, including consoles of hyperscalers), where 80% of "normal" things work, then you need the last 20% and it's always not working as documented, only half-working, or just outright broken, and you need to find "temporary" workarounds that stay in place for years.
I feel this is being amplified by AI: tests come last (if at all) and are still written by LLMs, nobody really looks at them anymore, green pipeline checkmarks mean less than they did in the past.
We've always had bugs. Good chance that it's largely confirmation bias - you're familiar with how software development has changed in the past 6 months, and especially if you don't like it, any time you see a bug or a problem, you'll attribute it to that.
I've had quite a few non-tech friends ask me similar questions to OP.
My running joke for the last couple years is "Nobody knows how to write software anymore" (even about a bug in my code). The problem is almost certainly complexity. Apps are like a house of cards, and we keep adding to the stack with new features, external dependencies, etc. My apps 30 years ago used one language, one library, one process, one thread and shipped on one platform. Now it uses Java, JavaScript, CSS, HTML, SQL and has a dozen linked libraries, talks to a couple servers, has dozens of threads, and runs on desktop, browser, tablet, phone and car console. It's a wonder anything works at all.
You're not going crazy. This has been the trend for a long time now.
Also there is trouble with back end IT systems. Like I had some trouble with my homeowner's insurance which was initially my fault [1] but my insurance agent was gracious and helped resolve it and my wife paid her right in her office. Then they do some software update and in the process cancel my policy and I get a letter from my lawyer [2] wanting to collect $82.54
[1] although if I was trying to design a spell to make people not pay their bill I might try sending them 20 letters that say THIS IS NOT A BILL before they get the bill
[2] who is also their lawyer
INCREDIBLY buggier, no comparison.
QA got it so bad in the current LLM-fueled carnival of stupid. Also, what would be categorized in the times of yore as a "lamer" has incredible free tools at their disposal, so vulnerabilities are a lot more common. Add to that the fact that every goddamn thing is connected to the internet and you got a recipe for disaster.
Even if you only hit 1/100 times... if you can shot 10000 times, you're still in the green.
Yeah substantial degradation in really basic functionality in the last 3-4 years.
I mean, new features I don’t use and useless UI changes have been added at a high rate over the same time period, but holy Christ, it’s pretty bad when some really basic functionality bugs are getting shipped by big names all the time. Like adobe PDF reader is… an absolute hot pile of garbage.
Yes. My observation is that FAANG culture persisting the myth that testing is slow or writing clean/consistent code is slow (conveniently playing to the dev's ego, but I digress). As a result, a lot of bugs get shipped, inevitably triggering a rework cycle. But what does not get measured, does not get reported. As a result devs look busier than ever, great for their reviews, while quality dives.
A turnaround in the industry would be actually capturing the rework cycle into costs.
There has been a steady decline for decades. Release early, release often. One example: Apple used to have a thing called Gold Master release when updating software was not as immediate. The GM programs met a high bar for quality. Not so much anymore… the attitude is we can just release a new version later this week. There also used to be a thing called backwards compatibility… now no one thinks twice about breaking code on older platforms.
I feel it in two ways.
I've been using windows for the past 5 years or so. I was using Linux before that. I used to complain about things breaking in Linux but the situation in Windows is way worse than it ever was in Linux. A couple of days ago my co-worker turned on his computer and windows has locked him out somehow. He had to disable secure boot, recover his bitlocker passphrase from his microsoft account, and only then could he get past the boot screen. The theory is that his laptop turned off while updating. I'm thinking, why doesn't the update abort when it realizes it's not plugged in or on low battery. He said it had never happened to him before and I have also never seen it.
The other side I notice is with google products. Barring android, GCP, and the less than yearly occassional outage, I would never see a bug in google products. Now I notice them often in drive, sheets, gmail, and specially meet. The other day I joined a meeting and all participants couldn't log in. NEVER have I seen such a bad production bug from google.
Profits over wonderful products I guess
My two Win 11 machines have been pretty solid for a while, but my Ubuntu Linux machine has had a lot of trouble with software updates.
When I built the machine about 12 years ago [1] it allocated a /boot partition that was big enough then but not big enough but as Ubuntu grew the /boot was too small to keep as many versions as Ubuntu wanted so kernel upgrades needed manual intervention. And for that machine it is a hassle because it runs headless and if it doesn't come up with the Ethernet working I have to pull it out of it's home and set up a workstation.
More recently I rebuilt / completely (kept the ZFS array) and this time Ubuntu didn't make a /boot so the files live on / but now when it does software updates it often fails to install the realtek driver modules so it comes up without the Ethernet and I have to take it upstairs to rebuild it. Copilot offered me a script to tell if the modules are busted after a kernel upgrade so I can do the the manual fix before the reboot now.
I guess I can't complain too much, I mean this $150 machine that I upgraded with 32GB of RAM and a NVIDIA GPU is going strong, though now that I think about it the second ZFS array I built for it is past warranty even if it only half full.
[1] I know because I named it after my favorite vixen from a video game I was playing back then
Software quality, i.e lack of bugs, from the average user's perspective, is a hygiene factor. High quality does not get rewarded, and the instant it crosses below the threshold of "good enough", users will punish you.
The eternal question for management seems to be "what can we get away with?".
A dev with quality problems in their output? Can em'. A dev with exceptional quality in their output? Can em'. A dev with just the right corners cut? Keep 'em, but no raise.
Is this regime right or fair? I don't think so, but it is the incentives that drive all of this.
I believe the only solution to this, at least in the for-profit software world, is to try to convince everyone to pay for quality. If the money is there, then maybe management will reward devs who care about quality user experiences.
It's similar to the ElectronJS desktop app problem. Sure the desktop apps use a lot more system resources, but it's doubtful those apps would have been created in the first place if it wasn't for the speedup of development allowed by ElectronJS.
In this case, sure software is buggier, but there's also a ton more software+features that wouldn't have existed before AI coding tools.
Electron apps are selected for things that are harmful to your user experience.
That is, if you know how to make Electron apps you also know how to make web applications and web applications are generally superior and preferred by users. No install, no bloated runtime, "just works", say it again "JUST WORKS!"
Any time somebody makes an Electron app there is a reason why they make an Electron app and it usually is a bad reason. For instance, Electron apps can live in your tray and make your computer harder to use by hiding behind twisty little icons that all look alike, popping up and covering other things that live in the tray, etc. Defying the usual procedures for starting and stopping apps, finding and moving windows, etc.
And of course the whole point of living in the tray is running all the time, sucking RAM all the time, sucking CPU all the time, sucking GPU all the time, sucking network bandwidth, etc. In short, sucking.
I mostly feel this on Microsoft stuff. O365 browser stuff was always buggy, but last month I had to fight to zoom in on something in PowerPoint. Pagination in Azure ML randomly broke, it took them a month to fix it.
Atlassian is also a bug factory, but I did not expect anything better of them. Each UI update they did to Confluence and Jira just made everything more confusing in the last 10 years.
Holy smokes is the browser versions of PPT, OneNote, word etc buggy, I could not believe that was an actual released product. Random page crashes, refresh loops, etc, all for the crime of using firefox on a mac. The bottom of their test matrix sure, but not exactly netscape on windows XP.
I don't know about just the past 6 months, but past 10-20 years, sure? I think software today has to handle much more varied and complicated I/O at a much bigger scale than it did in the past. Website UI/UX has to accommodate not only numerous browsers, but numerous devices with different viewing spaces. Think of how much broke when Python was rejiggered to not implicitly typecast unicode data. Think about all the legacy databases that can't handle spaces or various non-alphanumeric characters.
The userbase for software in the early 2000 was built for a much, much smaller audience than it is today. And I'm not sure the huge improvements in testing/debugging and general software development would totally mitigate that, in the same way modern military defense systems is still way behind modern offensive weaponry.
I use cheap android devices. My current one is a cheap Motorola and before samsung and Chinese vendors. I've never seen bugs like I have encountere since the last update (2 months ago).
Unless I reboot the device everyday, I can't make a call. The button is pressed but never proceeds to ring, doesn't even show an error message. The play service complains that I am logged out but shows my account info in the login panel!! Rebooting is the only way, android is like windows currently.
I am just scared to update anything these days. And not just android apps, Even rsync gave me a minor scare (fortunate my backup jobs are on debian stable).
I had one encounter with an obviously vibe coded website last year. Everything looked great, but nothing worked. Had to deal with the company over email.
Things might be better now, but that was a clear warning of what future may hold.
Someone should come up with an index to track this objectively.
Otherwise, its all anecdotes and speculation.
Some random ideas:
1. Measure trends in HTTP status codes over time
2. Measure github issues created over time
3. Measure complaints on forums over time
4. Google trends
It's definitely been buggier for a while now. It's already been an issue for a decade or so due to the "release minimum viable product, fix problems with patches/updates model" (especially in the world of video games), but the rise of AI has only made the issue more apparent. Microsoft is the biggest example here (with Windows 11 seemingly screwing up something massive in every update), but other companies like Google and Meta haven't been as reliable either.
Sure, but I would have agreed with you 10 years ago too. We were in the throes of making really fat assets that weak computers couldn't load quickly.
"weak computers" is the funniest part. The computer I'm typing this on *right now* is orders of magnitude faster then the Athlon XP processor that i was using in 2003, but it doesn't feel much faster. I think that hardware advancement is just slower then software bloat...
Not only buggier, but less usable overall. We can not blame everything on AI. For example events discovery platforms and ticket platforms were way better decade ago, you could search by popularity, friends going to the same event and there wasn’t as much fragmentation. That part you could blame on Cambridge Analytica scandal and late capitalism love of algorithmic feeds
Websites and mobile apps, definitely. Every time a typical proprietary mobile app updates I wonder what they will break. They always do break something or remove a feature I want.
Desktop apps, not in general. Some work better, like KDE's DE has fixed some irritants.
It's pretty hard to separate this from confirmation bias. But I've had to disable the YouTube app and switch to mobile web because it would just buffer nonstop on everything; never had issues like with it like this before.
I think gstatic.com was down yesterday or something because a ton of websites just stopped working.
Spotify on Android got an order of magnitude slower recently.
These are just the examples that come to mind that I've dealt with from the last 48 hours.
uThermal has been documenting in shorts the myriad game breaking bugs introduced into StarCraft 2 in recent patches. The game was super stable for 15 years and now you can load 8 sieged tanks into a medivac at a time. Anyway I am seeing the same thing as you in many different arenas
What software are you using? I can't say I've noticed it's worse than before. Software has always been very buggy.
100% - We are busy building bridges whilst walking on them, with no time to look back and inspect the structure.
Vibe coding + AI psychosis so there isn't reality testing before it gets shipped.
On linux it is the opposite, I have never been more happy with my desktop experience.
feature velocity >> feature stability.
Core libs/software have never been better, eg I have had much better luck with stuff like ffmpeg and virsh than ever before.
If it has a UI and targeted at consumers though? Bug city.
I would say yes, but not particularly more in the last 6 months.
Yes, I initially jumped from Android to iPhone back in 2019 or so because I felt like the experience was much more stable and it just worked, smooth, no issues, kinda boring but in a good adult way. Now every interaction makes me want to throw the phone out the window. If it’s not just blank, the keyboard doesn’t come up, the navbar in Safari sometimes is there, sometimes it’s a blank rectangle, etc. “ask your developer to update the apps to run well in iOS 26” is such a garbage response. We didn’t need to break everything just to have some glass simulations that make everything slower and add no value.
Google Home: holy shit dude. It was a mostly working thing. Absolutely unusable with Gemini. Will get the transcript right 15% of the time. Can’t even ask for the weather.
Roku: Netflix playing in the background but I’m still in the main menu.
I can go on
It correlates directly with AI adoption.
I think this is the natural path of things...
This will sound rude, but.. Obviously
Related last week:
Notes on Software Quality
https://news.ycombinator.com/item?id=48821441
Related from last October:
Ask HN: Why is software quality collapsing?
https://news.ycombinator.com/item?id=45474346
The great software quality collapse or, how we normalized catastrophe
https://news.ycombinator.com/item?id=45528347
Google Ai studio: pegs a core
Opencode: pegs a core
Paypal: pegs a core
Chatgpt: pegs a core
100%
iOS is definitely buggier. It used to be a silky smooth experience but it's degraded significantly in the last year or so. Safari on iOS, in particular, is very bad. Lots of problems in Adobe Creative Cloud on Win 11 (primarily Photoshop, InDesign, Illustrator) in recent months too.
This might sound like a rant, but it's more an attempt to outline the breadth of the problem by way of many examples. Software quality sucks right now, enough so that I plan my life around it. Payment processing is regularly broken? I'll maintain a couple cards from different providers. Login systems are always broken? If you require a login then I'm going to weigh the quality of your service against not just its physical dollar cost but also the amount of my time you're about to waste if I sign up. Etc.
Enterprise broke something and has been unable to text or email me receipts and reservation information for months. No trouble; I'll look up my information manually on their site when I want to modify a reservation. Great, I've found that information, surely I can modify it? Nope; that's broken too, and it asks me to call and doesn't provide the number I should call. So...I'll look up their customer support number, surely that isn't broken? Nope, somebody deployed a version of the site where I physically see the handlebars and variable name which should have referenced the phone number rather than its actual value, and they seemingly didn't notice or weren't able to roll back the change for at least a day.
And that's not to pick on them (especially because they're otherwise a good company IME), but to highlight a series of technical fuckups which never would have flown in past years, and they're not alone. My HSA just made it impossible to log in without buying a new phone, every single time I've tried to set up automatic payments for Visa they've had a different bug on their website preventing that, WalMart supposedly allows buying things for people in other states (I was buying my sister tires), but some broken AI will reverse the transaction hours later. eBay and Visa are even better on that front; their customer support seemingly aren't able to override the broken fraud-detection AI, so there are some Visa purchases I can't make, and every time I log in to eBay my account credentials are forcibly reset a day later.
Beyond the straight, obvious incompetence manifest in 3+ major bugs in nearly every website or service I interact with, there are also stranger bugs (no "strangest" bugs in recent memory). One of my coworkers has a reliable network blip every 20s (my memory is fuzzy, it might have been 10s or 15s when I timed it, but it's incredibly consistent), only hitting their outbound traffic, and it's at or upstream of the modem. One site I visit has some sort of broken filtering from my mobile carrier, but only for their login page, so I need to use a VPN to log in, but then the rest of the site has VPN detection, so I need to leave the VPN to access the rest. A little while ago, I stopped being able to accept calls while my wireless hotspot is active. Slack automatically moved precisely one of my DMs from the section I had assigned to it to a different section. Discord and Target both had an automatic retry on a failed telemetry implementation without backoff which caused you to immediately violate their rate limit by just visiting the site.
And so on. Major companies are writing broken software, missing the bugs in review, missing them in testing, don't have monitoring or alerting on core metrics to be able to roll back when their deploys suck (or just can't roll back), usually have no way to contact them to inform them of the mistake, and sometimes choose not to fix bugs as big as "many of our customers can't pay us" in days, sometimes ever.
Yeah man, I think this exactly captures what I've been feeling - every time I need to get something done, I consistently hit some experience breaking bug. And you're totally right, I've actually just come to expect it and work around it...
Do you think in the future apps that are stable and "just work" will be a moat? Feels like the direction we're trending in.
I thought about proposing that idea myself and decided against it. Consumer behavior is a bit hard to predict, and it's a little unclear to me how well "just works" will sell. I like to believe that's viable, but I don't really know.
Honestly no, I have not noticed this.
If you're ever asking "Is it just me?", the answer is "no".
It seems you are talking about the enshittification, which only affects megacorp-baked propriprietary software and services. It does not affect, e.g, GNU/Linux: https://news.ycombinator.com/item?id=41277484 and https://pluralistic.net/2025/02/26/ursula-franklin/
Why is this silently downvoted? The OP didn't give any details about their devices or specific bugs. I think my conclusion might be correct, given that the corporations spend less on the user convenience/UX as a part of the enshittification trend. Also confirmed by this comment here: https://news.ycombinator.com/item?id=48920698#48924033.