Live data from Hacker News

Ask HN: Is Software Getting Worse?

news.ycombinator.com

81–90 of 107 posts

Re: Ask HN: Is Software Getting Worse?

#81
The given examples are no so much about software getting worse but specifically web programming to do so. Here the answer is clearly yes, web programming is in a suboptimal phase at the moment with web programmers acting more like carpenters who try to cobble up a house using only IKEA flat-packed products which they jury-rig together using copious amounts of glue, packing tape and baling twine than like the artisan of old who went into the woods with an axe and built himself a log cabin. True, the log cabin was rather spartan but it it kept out the elements and lasted a long time. The IKEA mansion has all the furnishings you want and some of them even work, sometimes. Often they don't, the lights go dim when you want them bright or end up blinding you just when you dozed off in bed, the dishwasher only works when the washing machine is first switched on and off, the doorbell rings day and night except for when someone is actually at the door, etc.

Blame the enormous amount of churn in the web tool world, the chase after the latest fads, the fact that many 'web programmers' entered this field not so much because they like the intellectual challenge of solving puzzles but because it pays well, the fact that project targets keep on moving or whatever.

Commercial software is getting 'worse' in that the distinction between customer and product is disappearing. A product is that what is made by a company in order to sell it to a customer so the company gets to make money. A customer is he who buys a product from a company because he deems it fit for some purpose.

That was then but now things have changed. Final sales are making way for rental agreements, products are turning into advertising and data mining applications to be used by companies to herd and milk their customers. Customers are thus productised and sold to other companies who target them with directed advertising based on what was harvested earlier.

Free software is still mostly freed from this plague - although even there it sometimes shows up, e.g. the Shopping lens [1] which appeared in Ubuntu 12.04 and only recently disappeared was an unwelcome reminder of this phenomenon - so there is a way out.

Re: Ask HN: Is Software Getting Worse?

#82
post #80
post #79

Earlier quoted context omitted.

You genuinely believe a CPU from 2023 to be no better than one from 2012?

I really believe in that. We used to have 4 cores in 2012 plus and this is slightly different than 2 cores (hardware cryptorgaphy in 2012th processor was more noticeable for me than extra cores). Modern processors eats significantly less energy and can hold significantly more of RAM and that's all I can notice.

We used to have 4 cores in 2012 plus and this is slightly different than 2 cores

It is double actually. Also there are CPUs with 64 cores now. Also in 2012 there were CPUs with more than 4 cores.

that's all I can notice

What you can notice is not real information when speed, cores and energy use can be quantified.

Re: Ask HN: Is Software Getting Worse?

#83
post #80

Earlier quoted context omitted.

I really believe in that. We used to have 4 cores in 2012 plus and this is slightly different than 2 cores (hardware cryptorgaphy in 2012th processor was more noticeable for me than extra cores). Modern processors eats significantly less energy and can hold significantly more of RAM and that's all I can notice.

We used to have 4 cores in 2012 plus and this is slightly different than 2 cores It is double actually. Also there are CPUs with 64 cores now. Also in 2012 there were CPUs with more than 4 cores. that's all I can notice What you can notice is not real information when speed, cores and energy use can be quantified.

We also used to have 4 cores in 2009 but that was very unstable.

BTW, for modern multitasking computers, the main jump in performance comes from going from a single processor to dual processor. Than Amdahl's law neglects the difference.

> What you can notice is not real information when speed, cores and energy use can be quantified.

Single-core performance is totally the same, the number of cores does not change anything visible and energy use seems like the only outcome of 10 years spent from 2012.

Re: Ask HN: Is Software Getting Worse?

#84
post #83

Earlier quoted context omitted.

We used to have 4 cores in 2012 plus and this is slightly different than 2 cores It is double actually. Also there are CPUs with 64 cores now. Also in 2012 there were CPUs with more than 4 cores. that's all I can notice What you can notice is not real information when speed, cores and energy use can be quantified.

We also used to have 4 cores in 2009 but that was very unstable. BTW, for modern multitasking computers, the main jump in performance comes from going from a single processor to dual processor. Than Amdahl's law neglects the difference. > What you can notice is not real information when speed, cores and energy use can be quantified. Single-core performance is totally the same, the number of cores does not change anyt…

We also used to have 4 cores in 2009 but that was very unstable.

This is nonsense

Than Amdahl's law neglects the difference.

Then you don't understand "Amdahl's law". It is very basic and only about parts of software that aren't multithreaded becoming a bottleneck. This sounds like desperate pessimism.

Single-core performance is totally the same,

This is not true either and you could look at benchmarks to see it.

the number of cores does not change anything visible

Disable hyperthreading and all your cores in the bios, then run multithreaded software and say the same thing.

Re: Ask HN: Is Software Getting Worse?

#85

When was the last time you had to debug an IRQ conflict? I remember when "Plug and Play" was derisively nicknamed "Plug and Pray". Today, it all just works. When I'm assembling a computer, I don't have to set jumpers. I don't have to fiddle with making sure the boot drive is at the end of the IDE cable rather than the middle. I can just snap all the pieces together like Lego, hit the power button, and be assured that…

> When was the last time you had to debug an IRQ conflict?

Never. But in my home server I have 4, sometimes 5 RAID controllers. Some combinations of RAID/AHCI/IDE modes won't work due to lack of resources. I presume it's about I/O addresses - the boot message is not really informative compared to Device Manager. I regret not taking a picture of the error message. The system won't even pass the POST when this happens.

> I remember when "Plug and Play" was derisively nicknamed "Plug and Pray". Today, it all just works.

No, it doesn't! I just plugged in an old webcam and there's no driver. It's the exact same "Pray"... well, not really. I don't pray. These days I default to the "It's not going to work" mindset.

> When I'm assembling a computer, I don't have to set jumpers.

How is that a good thing? Overclock something, system won't boot, you have to reset the CMOS and loose all settings, including the boot order.

> I don't have to fiddle with making sure the boot drive is at the end of the IDE cable rather than the middle.

No, you could have used the jumpers, but I think you hated them too much to let them help you. I wish I had the jumpers now. Every time I insert or remove a SATA HDD from the rack I have to redo the boot order.

> I can just snap all the pieces together like Lego, hit the power button, and be assured that I'll get a bootable system (assuming, of course, that I haven't been a dolt and forgotten to plug the video card power cable in).

And then you notice that the system won't recognize the CPU without a BIOS/UEFI update, which you can't do because it won't boot unless it recognises the CPU. Then the OS is installed with IDE drivers and you can't switch to AHCI without major OS surgery.

> When was the last time you had an application blue-screen/bugcheck/kernel panic your machine? Yes, Windows still blue-screens from time to time, but over the past decade, I've found that 100% of my blue-screens have been caused by faulty drivers, rather than application code or bugs in the OS itself.

These days there's no blue screen or error. Apps just won't start (click the icon and it does nothing), opened apps suddenly close without any error (just dissappear from the screen), system suddenly reboots, or won't wake up from sleep, or an update makes the system unbootable or stuck in a boot cycle.

> This wasn't always the case. I remember, on Windows 98, there was one particular game that my brother had (I think it was Reader Rabbit), which would repeatedly and reliably blue-screen the machine when we got to a certain level. I haven't seen any errors like that in more than decade. And even the driver blue-screens are getting better. I remember not too long ago, my Windows PC's monitor blinked off, then came back. When I looked in Event Viewer, I saw that the GPU driver had crashed and had been automatically restarted. This is something that still causes kernel panics on Linux and MacOS, but Windows just shrugs it off and keeps on chugging.

True. This part is better.

> With regards to Linux, when was the last time you had to mess with xorg.conf? Wifi drivers? WPA supplicant? I remember when I had to download the Windows drivers for my wireless card, extract the binary blobs, compile NDISWrapper, and then pray that I'd set everything up correctly, before unplugging the Ethernet cable to test whether my wifi was working. Now? I browse Hacker News while Linux is installing, because wifi drivers have been part of the kernel for years.

I did most of the enumerated items this week.

> But other than comms, has software improved? I have a hard time arguing otherwise.

It did improve a little in stability, at a very very high cost of user's money, time and privacy: waiting for stuff to open, commands to get processed, buying ever faster and more expensive hardware to do basically the exact same things as 20 years ago, only slower, and not because of the 56K modem.

It's because of people always wanting the latest newest stuff that good old software gets abandoned. See: IRC, FTP, Opera Presto, websites without JS, single-user OS, also hardware: ethernet on laptops, headphone jacks on phones.

Re: Ask HN: Is Software Getting Worse?

#86
post #53

Earlier quoted context omitted.

DaVinci Resolve. Proper video editing, VFX, and colour grading software as used on stuff you've seen in the cinema and on telly. Free, if you don't want the clever "neural engine" stuff or >4k timelines. Oh, you do want that? That'll be 300 quid then, and it's yours for life. They have never even charged for the updates, either. Of course, they make their money on cameras and control surfaces.

DaVinci and Ableton are both good examples to me. I love Ableton and use it a ton, but man does it crash almost every session for me. I finally bit the bullet and bought Resolve to use some of its non free effects and have already run into a number of bugs. I’m still very grateful I have access to that level of software, but it does feel like things being 90% complete at best is the standard these days.

> I love Ableton and use it a ton, but man does it crash almost every session for me.

Is that when using a plugin window or using the Ableton UI itself?

I can only recall plugins crashing for me... Always worth checking for updates with the vendors.

Re: Ask HN: Is Software Getting Worse?

#87
Yea, ofc. There's just a lot going on all at once. And some of it is honest to god huge quality of life improvements, which impose guard rails around piles of garbage and creates reluctance to see or acknowledge very real negatives.

Re: Ask HN: Is Software Getting Worse?

#88
post #83

Earlier quoted context omitted.

We also used to have 4 cores in 2009 but that was very unstable. BTW, for modern multitasking computers, the main jump in performance comes from going from a single processor to dual processor. Than Amdahl's law neglects the difference. > What you can notice is not real information when speed, cores and energy use can be quantified. Single-core performance is totally the same, the number of cores does not change anyt…

We also used to have 4 cores in 2009 but that was very unstable. This is nonsense Than Amdahl's law neglects the difference. Then you don't understand "Amdahl's law". It is very basic and only about parts of software that aren't multithreaded becoming a bottleneck. This sounds like desperate pessimism. Single-core performance is totally the same, This is not true either and you could look at benchmarks to see it. the…

> This is nonsense

Core 2 duo Quad from 2007+ used to have 2 L2 cashes (not 4) and no L3 and task manager of Windows refused to see more than 2 cores, but for some rare tasks Quad was perfect such as playing GTA4. That is just my memories, feel free to correct me if you see me mistaken.

> Then you don't understand "Amdahl's law". It is very basic and only about parts of software that aren't multithreaded becoming a bottleneck.

Everything except of browser is not multithreaded in 2023, for example a garbage collecting of interpreted languages.

> This is not true either and you could look at benchmarks to see it.

I have a heavy single-threaded application which I need to run 24/7 and I still use Pentium 4 because no modern computer can make it 2x faster.

> Disable hyperthreading and all your cores in the bios, then run multithreaded software and say the same thing.

I mean starting from 2 cores the number of cores does not add anything visible.

Re: Ask HN: Is Software Getting Worse?

#89

Earlier quoted context omitted.

Same with Mac OS constantly asking Apple whether it's fine to start a program. Which was fun not long ago when that Apple server went down and people couldn't start programs anymore as long as the Mac was connected to the internet because Apple of course couldn't imagine their own creation being fallible.

Are you talking about gatekeeper flags or app signatures? As far as I know, both of those can be managed locally, I don’t think Apple is informed every time you try to open a process.

I have no idea how it's called but it is a very noticeable effect and also happens whenever I run a freshly compiled executable for the first time. It usually causes a delay of about 0.5-5 seconds depending on the wifi connection. Which is kind of insane, to upgrade to a literally 3 times faster machine and then it sometimes runs slower than before.

And not network-related but still annoying is Mac OS' habit of always spinning up connected harddrives each time a "save file" dialog opens, which causes that dialog to be completely unresponsive until all drives are ready. Save a file and the most expensive Mac can be bottlenecked by a single external HDD. It's just so unnecessary.

Re: Ask HN: Is Software Getting Worse?

#90
I think the issue is the incentives have shifted over the years.

A lot of software today is built to drive a specific user behavior to generate revenue instead of being designed from the start to provide customer value.

User-hostile behavior, dark patterns, etc. are all a symptom of this. Additionally time-to-market and marketing bullet points are often more important than user experience of bug-free software because many companies have found they can make the same revenue without having to focus on the customer experience.

Classic software was not as big-free as people like to think. There was lots of BAD software. The difference was that with higher prices and the inability to push updates (aside from mailing floppies) meant that the incentive was to produce software that had fewer bugs.

Post reply on HN