Sorry, I must have been unclear, because a few people have commented the same thing on the software. To be clear, I didn't see the software in any repository - the only place I found it was some old (2010'ish) website. If installing from a deb file you found is really a bad idea (which I disagree with, because it did end up working, and I don't know that there was another good option here), then someone should make the UI not try to do it and fail.
The Samba problem was from longer ago: I don't even remember the distribution. And yeah, I got it working in the end, but the graphical UI (which I found via some settings UI, which looked like the right thing) didn't work.
I brought both these up only as random examples of the sort of problems that I hit every time I try Linux. And also of examples of the kind of response you get when you look for help.
No, I didn't have a working version. I had no version and it was not in a repository. The only versions I found were on a random website with source and a.deb file.
In any case, you'd have no expectation of a random.exe from "some old website" working either.
You are incorrect. I absolutely would expect the equivalent software to work in Windows. To be clear, I don't think this was older than 2010 or something. Tons of software from 1998 still works on Windows computers. There's certain kinds of stuff that will break (anything using physical ports, certain frameworks), but usually it's fine.
And, to be clear, that's also kind of irrelevant. This isn't about backwards compatibility of the software - the software actually worked fine. It's just the distribution failed to install a deb file when you double clicked on it. This is baffling, because I was still able to install the package via the command line (with no special switches). I have no idea what it was doing when I double clicked on it, and it's desperately sad that nobody fixed the problem for years.
So yeah, I'm not complaining about some specific problem of backward compatibility of some software. These were just examples of problems I hit on my latest attempts, because they were freshest in mind. And that's my point: every time I try Linux I hit baffling problems using the included tools and UI. Nobody gives a crap, and all you hear when you look for answers is that you're doing it wrong.
I mean, look at the stupid/angry/unhelpful responses I've got here. It's a perfect microcosm of the "trying out Linux experience". If you hit similar problems in Windows, what you'd get is a pull-quote top-result on Google that probably fixes your problem. And, if you couldn't follow that, you'd scroll down a notch and see somewhat walking you through the problem on Youtube.
To be clear, this doesn't surprise or bother me all that much - but since the subject of this article is "why don't people switch to Linux?", I figured maybe the feedback from my repeated, honest attempts over the years, was relevant.
You know, actually, I don't often find that Windows software stops working or installing. I just installed an old game I wrote in like 1997 for DirectX 2 or something. The cheesy install wizard looked hilarious and it did a weird stutter when it started, but it worked fine.
I mean yeah, sure, I've hit problems with stuff because MS, for example, changed how security around ports works in Vista (I think? that's getting to be a while ago). Or because the developer used a framework that dead-ended like Adobe Air. But barring a specific hard stop like that, if I find some random bit of software on somebody's 2005 web page, in some oddly named folder on some University server, the Windows binary will almost certainly just work.
I have plenty of complaints with MS, but they actually do a pretty good job on backwards compatibility. If, for example, MSI files made before 2012 just stopped working when you clicked on them, MS would treat that as a problem.
The software I wanted hadn't been updated for a while, and all there was on some old website was either source or a.deb file to download. Since I just wanted to run the software (and not have a fun romp through a bunch of old dependencies), I took the deb. That whole part seemed pretty reasonable, and it seemed to be installing. And then it hung, icon lingering on the taskbar, process gone.
So I gave up and installed it from the command line. But I was curious if I'd messed up somehow, so I checked online. I found forums filled with miserable angry nerds who were mad at the question. The underlying problem has apparently existed for years. Nobody, like, doubted that the installer was broken. That seemed to just be accepted reality: of course it doesn't work. Why would you even try that? Sure software center (or whatever) might try to do something when you click on a deb file... but you can't really think that'll work, right? You are dumb. You need to learn how a real OS does things. Great.
On the other bit, connecting to a network so you can get some files is not some crazy arcane thing, and it's pretty easy to do that with Windows. Technology wise, Samba works fine for doing this job - once I gave up on the UI, it worked fine. So yeah... maybe I'm crazy, but I think it makes sense that there's a UI for configuring Samba in a Linux distro. But I don't think anyone involved cares whether it works, and it doesn't. (Or didn't, anyway, that one was actually a problem from years back).
Every once in a while I try out Linux on the desktop; the only consistent part of the experience over the years has been that it's terrible. Installing software is terrible. The package managers are terrible. It's a complete dumpster fire. A couple times ago I had mixed success with using.deb packages. Then last time, my success rate with "double clicking a.deb file" went to zero - they'd just hang on the taskbar doing nothing. The forums told me to sit and spin; apparently I should have been looking for snaps now.
The samba config tool did absolutely nothing, just silent, complete failure.
To be clear, I got everything I wanted set up via command line (and just using a container for one thing, because apparently lots of other people have given up on the problem altogether) when I got tired of fighting the crappy tools, but I was blown away by how bad it was - a huge backslide over the last 15 years.
If someone just wants a web browser terminal, they should get a Chromebook. If a non-techy user wants to do a variety of things on their home desktop computer - ie. activities that involve installing software - then Linux is a desperately bad choice.
And, to be clear, your examples are insanely irrelevant - bordering on comical. There aren't, like, throngs of non-technical Windows users who are unhappy with Microsoft's dual-boot support. Because the users don't give a crap about partitions or dual booting. Because they don't need to consider anything like that to run Excel or play a game.
I thought this sounded cool until I read some comments, and it turns out:
1. This was tried 10 years ago and it didn't succeed 2. When players perform actions in these games, there will be a delay before the server can process that action and return corresponding output. If this delay is too long, the experience will suffer.
It just shows how out of touch these companies are that neither of them considered these big red flags.
I mean, just because these computer hype things were inspired by biological neurons, unless they function in just the same way then we're insulting God by calling them that.
It also pisses me off that some arses (many of whom are total jerks to me on the aviation forums) call the side-planks on airplanes "wings", when they don't work anything like bird or insect wings.
This software might be very useful for developing computer systems to solve difficult tasks. Sure, whatever... but we can't let that blind us to what's really important here.
What's truly important here is dissecting exactly what to call these systems. Using the term "neural network" makes sense on pretty much every level, and it would allow us to communicate clearly about what type of algorithms are being used, but that term (and especially "deep neural network") make all my personal bugaboos about AI flare up. These are computers not brains, so there's no neurons, so we can't use that word.
And yeah, obviously I recognize that, for most people, Google's efforts here are exactly what people mean when they talk about AI - but that makes me angry so nobody should do it. I've decided, for no good reason, that the term AI should only be used when describing an intelligence that works just the same as a human. I have weird quasi-spiritual hangups about all this, which I think you're all obliged to respect.
PS: Also, just a reminder, Google has accomplished nothing, these systems are useless and aren't improving. All just unimpressive hype. I hate technology and change, please stop. Thanks in advance for never mentioning something like this again.
..that I will continue owning that game, at least to the extent you can own a digital object. That is the core value proposition they give me. Everything else is gravy (friend list, chat, game discovery, refunds, cloud saves, etc..).
Mostly, I just want to know that when I plunk down money on a game, it will stay in the list of games I can play for a reasonably foreseeable future. I trust Valve for that because they have some track record, and over the course of many years their terms have not significantly degraded.
I'm also just generally loath to install any new store or launcher thing. I don't want a Ubisoft account or a Microsoft store account or an Origin or Epic account. I barely tolerate having a Blizzard account.
Anyway, I don't wish these guys any ill will and I think Valve could use the competition - but I don't think the world can support too many stores over the long term.
Bitcoin absolutely does not excel at moving value from place to place. Transactions are slow and costly, and will get exponentially worse over time if usage were to increase. It also isn't particularly difficult to control, secure or private.
You're right that these are all things people imagine about Bitcoin, but none of them are true. Other cryptocurrencies do better at all these measures and still aren't particularly useful.
Slashdot has become a very sad place. This used to be the kind of tech-frontier topic that might make for interesting conversation.
Now, I can predict pretty accurately which people will honk out their same, sad, banal complaints for any given headline. I don't know what they get out of this - whether they actually feel that strongly about the subject, whether they like arguing or attention, or whether it's just a kind of reflexive contrarianism.
What's clear is that none of the more frequent posters have any actual interest in learning about the subjects their crapping on. I've tried to engage these posters a few times, and it goes nowhere. Oh well, eventually I'll break the habit of reading comments here.
These games aren't hard for a computer to play. You could write a fairly straightforward algorithm that would play them both well.
What's hard is to develop a very general learning algorithm - one that doesn't know about the task - that just happens to pass the test of being able to learn these games.
The approach here seems "cheaty". That's not to say their technique is useless (and maybe it's more generalizable than I'm giving it credit for) - but from the vague overview of the article it seems like they're effectively juicing their performance.
Lots of cryptographic algorithms are fine, or may just need longer codes. The hardest ones to replace are public-keys, where I think the front runners are lattice or error correction based (see NTRU and McEliece).
The other possibility is public key encryption dies, and we have to build some wacky network of symmetric encryption trust rings or something.
If Bitcoins suddenly became easier to mine, they'd just have to increase difficulty. It's not like there's a static amount of work to be done, and this will do it faster (which is good, as otherwise Bitcoins would rapidly lose value as computing improved). If they couldn't adjust the difficulty enough to compensate, the system would need major change.
But it would likely actually collapse for a different reason: QC could make spending other people's Bitcoins very easy - and thus make all of them worthless.
I'd like a cheaper, simpler switch that doesn't have a screen or a battery and comes with normal controllers. My kids like Nintendo games, but our Wii U has largely been a bust because of all its wacky controllers and options and gimmicks.
I'm sure there's some parts of their design that could help... but it's going to be a small-constant-factor improvement that won't make up for a gap on the AI side.
What it really adds is a bunch of complications around designing a car and making the wheels turn and what not.
I haven't heard of real problems from 3rd party iOS handset makers. Sounds like that's the way to go - they don't "play favorites" like Google.
Geez, again just sorry, man - sorry that you had to go through this, and you weren't able to put all of their software on your phone. That's awful. But at least you know for next time, stick with Apple.
Well... those would both be interesting headlines when they first became possible. And this story does represent a novel level of success, though I'd agree the headline (or summary) doesn't encapsulate why this is a novel result.
Now that they've fixed the summary, my original comment is mostly useless.
But I'm glad you were able to really drive home my side point: random people are very hazy on what automation is going to be able to do, and how many jobs it might create or destroy.
I mean, I don't know that average Joe necessarily has terribly good insight on this subject (and survey results are easy to manipulate by finding a wording that leads responses) - but the different figures in the summary are very different, and suggest very different political outcomes here.
Slashdot Mirror
Sorry, I must have been unclear, because a few people have commented the same thing on the software. To be clear, I didn't see the software in any repository - the only place I found it was some old (2010'ish) website. If installing from a deb file you found is really a bad idea (which I disagree with, because it did end up working, and I don't know that there was another good option here), then someone should make the UI not try to do it and fail.
The Samba problem was from longer ago: I don't even remember the distribution. And yeah, I got it working in the end, but the graphical UI (which I found via some settings UI, which looked like the right thing) didn't work.
I brought both these up only as random examples of the sort of problems that I hit every time I try Linux. And also of examples of the kind of response you get when you look for help.
No, I didn't have a working version. I had no version and it was not in a repository. The only versions I found were on a random website with source and a .deb file.
You are incorrect. I absolutely would expect the equivalent software to work in Windows. To be clear, I don't think this was older than 2010 or something. Tons of software from 1998 still works on Windows computers. There's certain kinds of stuff that will break (anything using physical ports, certain frameworks), but usually it's fine.
And, to be clear, that's also kind of irrelevant. This isn't about backwards compatibility of the software - the software actually worked fine. It's just the distribution failed to install a deb file when you double clicked on it. This is baffling, because I was still able to install the package via the command line (with no special switches). I have no idea what it was doing when I double clicked on it, and it's desperately sad that nobody fixed the problem for years.
So yeah, I'm not complaining about some specific problem of backward compatibility of some software. These were just examples of problems I hit on my latest attempts, because they were freshest in mind. And that's my point: every time I try Linux I hit baffling problems using the included tools and UI. Nobody gives a crap, and all you hear when you look for answers is that you're doing it wrong.
I mean, look at the stupid/angry/unhelpful responses I've got here. It's a perfect microcosm of the "trying out Linux experience". If you hit similar problems in Windows, what you'd get is a pull-quote top-result on Google that probably fixes your problem. And, if you couldn't follow that, you'd scroll down a notch and see somewhat walking you through the problem on Youtube.
To be clear, this doesn't surprise or bother me all that much - but since the subject of this article is "why don't people switch to Linux?", I figured maybe the feedback from my repeated, honest attempts over the years, was relevant.
You know, actually, I don't often find that Windows software stops working or installing. I just installed an old game I wrote in like 1997 for DirectX 2 or something. The cheesy install wizard looked hilarious and it did a weird stutter when it started, but it worked fine.
I mean yeah, sure, I've hit problems with stuff because MS, for example, changed how security around ports works in Vista (I think? that's getting to be a while ago). Or because the developer used a framework that dead-ended like Adobe Air. But barring a specific hard stop like that, if I find some random bit of software on somebody's 2005 web page, in some oddly named folder on some University server, the Windows binary will almost certainly just work.
I have plenty of complaints with MS, but they actually do a pretty good job on backwards compatibility. If, for example, MSI files made before 2012 just stopped working when you clicked on them, MS would treat that as a problem.
The software I wanted hadn't been updated for a while, and all there was on some old website was either source or a .deb file to download. Since I just wanted to run the software (and not have a fun romp through a bunch of old dependencies), I took the deb. That whole part seemed pretty reasonable, and it seemed to be installing. And then it hung, icon lingering on the taskbar, process gone.
So I gave up and installed it from the command line. But I was curious if I'd messed up somehow, so I checked online. I found forums filled with miserable angry nerds who were mad at the question. The underlying problem has apparently existed for years. Nobody, like, doubted that the installer was broken. That seemed to just be accepted reality: of course it doesn't work. Why would you even try that? Sure software center (or whatever) might try to do something when you click on a deb file... but you can't really think that'll work, right? You are dumb. You need to learn how a real OS does things. Great.
On the other bit, connecting to a network so you can get some files is not some crazy arcane thing, and it's pretty easy to do that with Windows. Technology wise, Samba works fine for doing this job - once I gave up on the UI, it worked fine. So yeah... maybe I'm crazy, but I think it makes sense that there's a UI for configuring Samba in a Linux distro. But I don't think anyone involved cares whether it works, and it doesn't. (Or didn't, anyway, that one was actually a problem from years back).
Every once in a while I try out Linux on the desktop; the only consistent part of the experience over the years has been that it's terrible. Installing software is terrible. The package managers are terrible. It's a complete dumpster fire. A couple times ago I had mixed success with using .deb packages. Then last time, my success rate with "double clicking a .deb file" went to zero - they'd just hang on the taskbar doing nothing. The forums told me to sit and spin; apparently I should have been looking for snaps now.
The samba config tool did absolutely nothing, just silent, complete failure.
To be clear, I got everything I wanted set up via command line (and just using a container for one thing, because apparently lots of other people have given up on the problem altogether) when I got tired of fighting the crappy tools, but I was blown away by how bad it was - a huge backslide over the last 15 years.
If someone just wants a web browser terminal, they should get a Chromebook. If a non-techy user wants to do a variety of things on their home desktop computer - ie. activities that involve installing software - then Linux is a desperately bad choice.
And, to be clear, your examples are insanely irrelevant - bordering on comical. There aren't, like, throngs of non-technical Windows users who are unhappy with Microsoft's dual-boot support. Because the users don't give a crap about partitions or dual booting. Because they don't need to consider anything like that to run Excel or play a game.
You bring up some very interesting points - but just as many questions.
Like, sure my home computer has a powerful gaming GPU - but is it really powerful enough to stream video?
Will I have enough RAM to stream video? Is my Windows OS advanced enough?
We live in a world of uncertainty.
I thought this sounded cool until I read some comments, and it turns out:
1. This was tried 10 years ago and it didn't succeed
2. When players perform actions in these games, there will be a delay before the server can process that action and return corresponding output. If this delay is too long, the experience will suffer.
It just shows how out of touch these companies are that neither of them considered these big red flags.
Fully agree.
I mean, just because these computer hype things were inspired by biological neurons, unless they function in just the same way then we're insulting God by calling them that.
It also pisses me off that some arses (many of whom are total jerks to me on the aviation forums) call the side-planks on airplanes "wings", when they don't work anything like bird or insect wings.
This software might be very useful for developing computer systems to solve difficult tasks. Sure, whatever... but we can't let that blind us to what's really important here.
What's truly important here is dissecting exactly what to call these systems. Using the term "neural network" makes sense on pretty much every level, and it would allow us to communicate clearly about what type of algorithms are being used, but that term (and especially "deep neural network") make all my personal bugaboos about AI flare up. These are computers not brains, so there's no neurons, so we can't use that word.
And yeah, obviously I recognize that, for most people, Google's efforts here are exactly what people mean when they talk about AI - but that makes me angry so nobody should do it. I've decided, for no good reason, that the term AI should only be used when describing an intelligence that works just the same as a human. I have weird quasi-spiritual hangups about all this, which I think you're all obliged to respect.
PS: Also, just a reminder, Google has accomplished nothing, these systems are useless and aren't improving. All just unimpressive hype. I hate technology and change, please stop. Thanks in advance for never mentioning something like this again.
..that I will continue owning that game, at least to the extent you can own a digital object. That is the core value proposition they give me. Everything else is gravy (friend list, chat, game discovery, refunds, cloud saves, etc..).
Mostly, I just want to know that when I plunk down money on a game, it will stay in the list of games I can play for a reasonably foreseeable future. I trust Valve for that because they have some track record, and over the course of many years their terms have not significantly degraded.
I'm also just generally loath to install any new store or launcher thing. I don't want a Ubisoft account or a Microsoft store account or an Origin or Epic account. I barely tolerate having a Blizzard account.
Anyway, I don't wish these guys any ill will and I think Valve could use the competition - but I don't think the world can support too many stores over the long term.
Bitcoin absolutely does not excel at moving value from place to place. Transactions are slow and costly, and will get exponentially worse over time if usage were to increase. It also isn't particularly difficult to control, secure or private.
You're right that these are all things people imagine about Bitcoin, but none of them are true. Other cryptocurrencies do better at all these measures and still aren't particularly useful.
I find the whole thing bizarre/depressing.
The normal definition of "cryptographic algorithms" includes "cryptographic hash functions", which are used by cryptocurrencies.
The crypto prefix also makes sense in the "hidden/occult/mystery" sense.
I have no faith in any of them, but I think the term is well chosen.
Slashdot has become a very sad place. This used to be the kind of tech-frontier topic that might make for interesting conversation.
Now, I can predict pretty accurately which people will honk out their same, sad, banal complaints for any given headline. I don't know what they get out of this - whether they actually feel that strongly about the subject, whether they like arguing or attention, or whether it's just a kind of reflexive contrarianism.
What's clear is that none of the more frequent posters have any actual interest in learning about the subjects their crapping on. I've tried to engage these posters a few times, and it goes nowhere. Oh well, eventually I'll break the habit of reading comments here.
No, but telling them the answers to the test can be. To be fair, from the article it's hard to tell where on that spectrum they are.
These games aren't hard for a computer to play. You could write a fairly straightforward algorithm that would play them both well.
What's hard is to develop a very general learning algorithm - one that doesn't know about the task - that just happens to pass the test of being able to learn these games.
The approach here seems "cheaty". That's not to say their technique is useless (and maybe it's more generalizable than I'm giving it credit for) - but from the vague overview of the article it seems like they're effectively juicing their performance.
Lots of cryptographic algorithms are fine, or may just need longer codes. The hardest ones to replace are public-keys, where I think the front runners are lattice or error correction based (see NTRU and McEliece).
The other possibility is public key encryption dies, and we have to build some wacky network of symmetric encryption trust rings or something.
If Bitcoins suddenly became easier to mine, they'd just have to increase difficulty. It's not like there's a static amount of work to be done, and this will do it faster (which is good, as otherwise Bitcoins would rapidly lose value as computing improved). If they couldn't adjust the difficulty enough to compensate, the system would need major change.
But it would likely actually collapse for a different reason: QC could make spending other people's Bitcoins very easy - and thus make all of them worthless.
I'd like a cheaper, simpler switch that doesn't have a screen or a battery and comes with normal controllers. My kids like Nintendo games, but our Wii U has largely been a bust because of all its wacky controllers and options and gimmicks.
I'm sure there's some parts of their design that could help... but it's going to be a small-constant-factor improvement that won't make up for a gap on the AI side.
What it really adds is a bunch of complications around designing a car and making the wheels turn and what not.
I haven't heard of real problems from 3rd party iOS handset makers. Sounds like that's the way to go - they don't "play favorites" like Google.
Geez, again just sorry, man - sorry that you had to go through this, and you weren't able to put all of their software on your phone. That's awful. But at least you know for next time, stick with Apple.
Well... those would both be interesting headlines when they first became possible. And this story does represent a novel level of success, though I'd agree the headline (or summary) doesn't encapsulate why this is a novel result.
I've never seen a wild rat in Alberta. Lots of mice and mosquitoes. Sometimes a rabbit or coyote in my back yard. No rats.
Now that they've fixed the summary, my original comment is mostly useless.
But I'm glad you were able to really drive home my side point: random people are very hazy on what automation is going to be able to do, and how many jobs it might create or destroy.
I mean, I don't know that average Joe necessarily has terribly good insight on this subject (and survey results are easy to manipulate by finding a wording that leads responses) - but the different figures in the summary are very different, and suggest very different political outcomes here.
I think you need to blow off a little more space dust there.