Graphic memory doesn't just survive logouts, it survives soft-reboots as well (e.g. the login screen in Linux likes to show garbage from the previous boot). To clear the memory you have to switch of the computer completely.
I'm running 14.04 happily with a bunch of PPAs. I have gcc5 PPA, I have the libreoffice fresh PPA and one or two others. It works perfectly fine and there is no wrecking of the monolithic dependency tree, and regular updated happen just fine.
How cute. You have three PPAs and call that proof that the system works? Try that again at the scale of the Apple or Android stores where you have to deal with a million apps from third parties.
PPAs plug into the monolithic dependency tree, there is absolutely zero tooling in place to ensure that they don't break things. The only reason why things don't break often is because people work very hard behind the scenes to keep things running. The monolithic dependency tree is essentially the Linux version of Window's "DLL Hell".
It's not trivial, no, but it's by no means impossible and not even that hard. I have compilers installed from 10.04, 12.04, 14.04 and the gcc5 PPA installed on this machine, all using apt. The former two are in two different chroot environments. That is good practice.
Yes, because your OS is garbage. Chroots and virtual machines are a workaround for crappy OSs that are incapable of giving you reproducible and verifiable behavior.
No, because the system allows and heck even encourages such things.
Yes, in much the same way as a broken car encourages repair.
This is why it's not a walled garden.
There is one monolithic dependency tree. Your distri controls it.
Take a simple everyday example: You want to install software "foobar", so you do "apt-get install foobar". Awesome, takes three seconds and you are done. That's how it should be. But now "foobar" release a brand new version and you want to use it right now, but distri won't have it for another release for six month. So what do you do now? Wait for somebody to build a PPA? Grab the sources of the.deb and patch them up to the newest version?./configure && make directly from source? All of this is possible, all of this works. All of this is also a thousand times more complicated then "apt-get install foobar". Just like a jailbreak. Your distris won't stop you from leaving it's walled garden, but it won't give you much of a helping hand either.
The AMD Open Source Driver on Linux do the same thing. It's not really a new or spectacular bug, graphics cards and drivers have done that stuff for quite a long while. Once there was also a fun bug that would make large texts in Firefox 'bleed' into the desktop background image, so it wasn't just showing old content, but actively manipulating content of another application.
You can add new repositories from other people (PPAs) you can create your own repositories (and share them with others), you can make your own dpkgs inside the main tree OR to install dependency free in/opt. You can side load and have multiple installed side by side.
Yes I can and in doing so I will wreak havoc to the monolithic dependency tree. There is no way to do a simple task such as installing two different version of the same package via apt. The system isn't build with that flexibility in mind.
I have build scripts which create temporary chroot ubuntu installs of various versions (with caching!) so I can get fully automatic repeatable builds of a package I distribute.
Yes and that is the problem. You had to literally abandon your main OS and reinstall a new one to get dependable behavior. That is a failure of the OS, not a feature.
Or grab the source, and configure && make && make install in a sandbox/VM,
Again, you are working around the system. The ability to completely by pass the package manager and doing things the manual way is not an argument for the package manager being good, it's the very reason why it's crap.
Oh. Today I learned apt-add-repository and the AUR don't exist.
Those fuck around with the monolithic dependency tree and cause trouble all the time. The package management system has no means to keep third party apps in a separate namespace.
1) Being legally allowed to make a change or redistribute something 2) Having the realistic possibility to make use of freedoms granted in 1)
Number 2) is why the GPL exist, as the pure legality to make a change isn't enough, you also need access to source code and build files to make it practical and those are required by the GPL. The problem arises when the build files don't work, when the dependencies don't match and stuff just doesn't work. That to me is a clear violation of the principles behind 2). If I get the source code and am unable to reproduce the binary then something is very obviously wrong.
Are you for real? Open source platforms have long since abandoned "make install", with one or two exceptions.
Virtually every single piece of Open Source software out there uses "make install". Your distribution wraps the result of that up in a package which you can cleanly install and uninstall, but that doesn't solve the fundamental problems with the concept of "make install" (i.e. single global name space into which all packages go, no way to install two different version of the same package, no relocation, etc.). It's a fundamentally flawed approach.
The result was (and largely still is) a clean, virus free, simple installation framework.
Yes and there is nothing wrong with reaching that goal, the problem is that they did it at the expensive of the users freedom. Distributions still have no good way of dealing with software that isn't provided by them.
No it isn't. You can side-load software using any other repository system if you like (see e.g. bedrock linux).
Yes, I am free to not use the repository system that doesn't work, just like on iPhone or any other proprietary system. If I work around what the system offers then I gain more freedom. That's not a good thing, that's the problem. The repository system should be the very thing that gains me freedom, not the thing that gets in the way of it.
They provide complete sources.
Which are nearly worthless as the system doesn't provide a clean way for me to modify them and reintroduce them into the system. The next time I ran a upgrade my modification will be overwriting by the distribution. And when I put packages on hold or increase the epoch version I cut myself of from security updates and such. Repositories are not build for user modification, they are build around a central authority that provides everything and decides everything. The user can comply or switch to another distri that suffers from much the same problems.
Yes and I am also free to jailbreak an iPhone. Just because I can do it, doesn't mean that the system is welcoming to third party software. The monolithic dependency trees used in the Free Software world are worse then what you see in the proprietary world. With Windows, as messy at it is, installing third party software isn't difficult and Microsoft goes to great length to ensure backward and forward compatibility to OS released a decade ago. It's not always works out of course, but it's a lot better then Linux world where every distribution is incompatible to everything else and even to past and future versions of itself.
I don't have to worry about apt-get install doing dangerous stuff.
Yes, but it does so by being a walled garden. The distri decides what goes into the repository, what version of the program and so on. As a user I can just consume what they provide me, my ability to change or object what they do is extremely limited and any more complex change will break the monolithic dependency tree. It's a system that violates everything Free Software should stands for, it works, but it doesn't give much freedom at all.
Not disagreeing on the overall point, one problem however with the Free Software world is that those fights often go into the wrong direction. Take the app model on iOS and Android, on one side it's a framework that is used to provide a walled garden, on the other side the clean isolation of apps from each other drastically reduces the potential for abuse and in turn provides a flourishing ecosystem of applications. On your average Free Software system by contrast you always have to worry about a 'make install' wreaking havoc, as there is no limit to what it can do. If you run a program you hand your system over to that program, you are no longer in control, you have to trust the software to do the right thing. This goes wrong all the time without even malicious intent, a wrong library here or there and stuff will break.
On the OLPC, which had a similar isolation model for apps, there was this idea of having a 'source' button on the keyboard, you could press it and it automatically jump you to the source code. You could modify things without wreaking havoc and you could share your modified apps with your friends with a click of a button. It never got fully implemented from what I understand, but as a concept it's still an incredible interesting idea for how Free Software should work. A system bulid from the ground up to encourage modification instead of a system where modifications are complicated, time consuming and dangerous. If people here of a new cool app on iOS and Android, they can try it with a few button presses. On a Free Software system I head over to Github, see a long list of required dependencies and essentially give up at that point, as more often then not, it's just not worth the effort.
Instead of fighting against stuff, Free Software would do well focusing more on building the better system. A system where all the software is under a Free license should make it easier to share stuff, but instead it's often harder. All the freedom the license give doesn't really trickle down to the user.
This is just one example, there are plenty other areas, like the fight against Javascript. Javascript is not the enemy, it should be seen as a building block for the semantic web, as it allows to decouple the data from the presentation, something that HTML claimed to do, but completely failed to do in reality.
Last time I checked about 5% of Steam gamers had a PC meeting the recommended specs from Oculus. That was a few month ago, so it might be a bit more by now, but I am pretty sure that it's still a tiny minority of gamers, and far less when it comes to regular PC users.
They made some weird deal with Microsoft that got them the gamepads really cheap or at least that's what they say. It's kind of weird, as they also include another small remote-like controller in the box as well. The gamepad feels completely redundant and unnecessary. And with Oculus Touch yet another controller will be released in a few month.
Or, guy with unfinished product has no idea of cost but needs to get investors to keep giving him money so he can finish building it.
They already shipped the DK1 ($300) and DK2 ($350) headsets at that point, so they should have had a reasonable good idea how much it would cost. We are not talking about a Kickstarter gone crazy and unexpected cost overruns, they deliberately build the thing outside of the originally intended price range, since they already knew how to build one that fits the target.
This graph seems to indicate that the problem of lack of women in computer science correlated pretty directly with the rise of the home computer and video games in the mid 80's.
What is wrong with discussing a "gender empathy gap"
What has that to do with Open Source? I have written Open Source code for 15+ years and never seen any gender issues pop up ever. Nobody cares what gender you are. You are judged by your patches and the usefulness of your contributions and nothing else. When conflicts happens, they happen for technical reasons or religious reasons (vi vs Emacs), not for gender reasons. And if somebody is really unhappy with how an Open Source project is going, they fork it or spend their time on a creation of their own. There is no gender check when you want to create a Github repo. You click a few buttons and start to submit code. Maybe somebody will find that code useful or maybe not. It's your repo and you can do whatever you want with it. You do not need anybodies approval to write Open Source code.
The biggest difference is that ZFS has full knowledge of the state of the file system, rsync on the other side doesn't, it's stateless, it has to start from zero each time and regather the information on each and every run on both sides, which is a really slow and potentially error prone process (i.e. when files change while rsync runs). ZFS knows what's going on in the filesystem and its snapshots the filesystem at a single point in time, so it thus it can be be far quicker and won't produce inconsistencies in the transmitted data. Tracking renames would speed some things up a little, but to match ZFS it would need a way to get the information about what changed in the filesystem from the filesystem itself and at the moment such functionality isn't available.
The same as what happened with the rest of the project, it kind of stalled. There was a lot of development after Milestone1, but that never converged into a finished release and then staled, developers left, new joined, etc., the release now is essentially a snapshot of that incomplete thing with a bit of polish. Nolok is still around and maybe one day I or somebody else will implement it, but there really isn't enough coherence in the development at the moment for that. Here is an animation test from seven years ago.
You included the.git tree in your count, so you are counting everything twice. For the rest you can blame git as well, it stores and downloads the complete history, so everytime a music file gets updated a little, you end up with a duplicate music file in the project history, same with the graphics. Git isn't very good at dealing with changing binary files and none of the half dozen workarounds (annex, large file, etc.) are worth the effort to save a hundred megabytes.
The bloatiness is the result of a lack of cleanup, the project accumulated lots of graphics over the years, lots of that isn't really used or critical, but might break some random user contributed level if removed. So old graphics stay in for fear of breaking stuff. This game is also very "community driven", meaning everybody can contribute some crappy thing and that then ends up in the source tree, rejection of bad contribution doesn't really happen, unless they break something major.
Biggest chunk of the size is however simply the music, which comes along in very high quality compression, there is no reason you couldn't up the compression ratio and make the package half the size. Some audiophile might notice, I didn't.
There is of course also the plain old reason that tech advances, 1280x720 true color graphics take a lot more space then 320x200 256 color graphics. Same with midi music vs ogg. Also the game uses plain PNGs everywhere, you could of course use compressed textures or if you really wanted to JPEGs for some graphics, but that isn't really worth the effort and makes modifications harder.
Not that big, you only download a 50GB game from Steam like once a month, probably a lot less. HD streaming on the other side you can have running for hours each day (average American watches like 4h TV a day).
News about ad networks doing evil tracking and such comes up all the time, but I have to wonder what all that data is any good for? I don't use ad block and yet all the ads I see tend to be either be pretty unimpressive and untargeted. They might know that I am male and they might be clever enough to show video games advertisement on a video game site, but that's about it. In a lot of ways the targeting makes the ads actually worse, as they end up covering a far narrower range of products and lead to a lot of repetition. Youtube can be notoriously bad at this, showing you the same ad 20 times in a row. There are also some ads that are extremely personalized, but they are unimpressive in the other direction, Amazon ads for example will just show me products that I just watched on their site. So won't show me anything new, just stuff I am already familiar with.
Never seen ads that actually extrapolate my behavior and interest and end up recommending me a product that I would actually be interested in buying. I found good old untargeted advertisement on TV or magazines far more useful for that, as that showed a lot of products that I didn't even know existed.
I think it's not so much that it isn't fast enough, but that you end up with a big box with a big spinning mirror. What he wants to do is to take the same technology and replace the spinning mirror with a DLP-style micromirror, thus you could do the same thing much smaller, cheaper and with less power, which in turn would allow you to have more of those devices for better tracking.
Overclocking processors generally voids the warranty and in general overclocking isn't something that appears in marketing materials for most or all processors.
Actually the box of my AMD FX-8320 reads:
* Unlocked out of the box * Achieve ultimate power with up to 5.0Ghz and 8 cores * Overclock your system for supercharged performance
The joke is when you read the included papers inside of the box:
Use only the heatsink/fan provided. Use of any other heatsink/fan will void warranty
Differences when compiled can be shrugged off as different compiler versions.
Yep, and it's not just different compilers, time stamps, compile order on parallel builds, the order of files in the filesystems, install path, compile flags, etc. will all change the resulting binary. Reproducible builds just hasn't been a thing in Free Software community and only very recently did Debian start work on ensuring that their binaries are byte-for-byte reproducible, but that's of course just Debian, we are still far far away from having reproducible builds be the default way how Free Software binaries are distributed.
Slashdot Mirror
Graphic memory doesn't just survive logouts, it survives soft-reboots as well (e.g. the login screen in Linux likes to show garbage from the previous boot). To clear the memory you have to switch of the computer completely.
I'm running 14.04 happily with a bunch of PPAs. I have gcc5 PPA, I have the libreoffice fresh PPA and one or two others. It works perfectly fine and there is no wrecking of the monolithic dependency tree, and regular updated happen just fine.
How cute. You have three PPAs and call that proof that the system works? Try that again at the scale of the Apple or Android stores where you have to deal with a million apps from third parties.
PPAs plug into the monolithic dependency tree, there is absolutely zero tooling in place to ensure that they don't break things. The only reason why things don't break often is because people work very hard behind the scenes to keep things running. The monolithic dependency tree is essentially the Linux version of Window's "DLL Hell".
It's not trivial, no, but it's by no means impossible and not even that hard. I have compilers installed from 10.04, 12.04, 14.04 and the gcc5 PPA installed on this machine, all using apt. The former two are in two different chroot environments. That is good practice.
Yes, because your OS is garbage. Chroots and virtual machines are a workaround for crappy OSs that are incapable of giving you reproducible and verifiable behavior.
No, because the system allows and heck even encourages such things.
Yes, in much the same way as a broken car encourages repair.
This is why it's not a walled garden.
There is one monolithic dependency tree. Your distri controls it.
Take a simple everyday example: You want to install software "foobar", so you do "apt-get install foobar". Awesome, takes three seconds and you are done. That's how it should be. But now "foobar" release a brand new version and you want to use it right now, but distri won't have it for another release for six month. So what do you do now? Wait for somebody to build a PPA? Grab the sources of the .deb and patch them up to the newest version? ./configure && make directly from source? All of this is possible, all of this works. All of this is also a thousand times more complicated then "apt-get install foobar". Just like a jailbreak. Your distris won't stop you from leaving it's walled garden, but it won't give you much of a helping hand either.
The AMD Open Source Driver on Linux do the same thing. It's not really a new or spectacular bug, graphics cards and drivers have done that stuff for quite a long while. Once there was also a fun bug that would make large texts in Firefox 'bleed' into the desktop background image, so it wasn't just showing old content, but actively manipulating content of another application.
You can add new repositories from other people (PPAs) you can create your own repositories (and share them with others), you can make your own dpkgs inside the main tree OR to install dependency free in /opt. You can side load and have multiple installed side by side.
Yes I can and in doing so I will wreak havoc to the monolithic dependency tree. There is no way to do a simple task such as installing two different version of the same package via apt. The system isn't build with that flexibility in mind.
I have build scripts which create temporary chroot ubuntu installs of various versions (with caching!) so I can get fully automatic repeatable builds of a package I distribute.
Yes and that is the problem. You had to literally abandon your main OS and reinstall a new one to get dependable behavior. That is a failure of the OS, not a feature.
Or grab the source, and configure && make && make install in a sandbox/VM,
Again, you are working around the system. The ability to completely by pass the package manager and doing things the manual way is not an argument for the package manager being good, it's the very reason why it's crap.
Oh. Today I learned apt-add-repository and the AUR don't exist.
Those fuck around with the monolithic dependency tree and cause trouble all the time. The package management system has no means to keep third party apps in a separate namespace.
There are two components to software freedom:
1) Being legally allowed to make a change or redistribute something
2) Having the realistic possibility to make use of freedoms granted in 1)
Number 2) is why the GPL exist, as the pure legality to make a change isn't enough, you also need access to source code and build files to make it practical and those are required by the GPL. The problem arises when the build files don't work, when the dependencies don't match and stuff just doesn't work. That to me is a clear violation of the principles behind 2). If I get the source code and am unable to reproduce the binary then something is very obviously wrong.
Are you for real? Open source platforms have long since abandoned "make install", with one or two exceptions.
Virtually every single piece of Open Source software out there uses "make install". Your distribution wraps the result of that up in a package which you can cleanly install and uninstall, but that doesn't solve the fundamental problems with the concept of "make install" (i.e. single global name space into which all packages go, no way to install two different version of the same package, no relocation, etc.). It's a fundamentally flawed approach.
The result was (and largely still is) a clean, virus free, simple installation framework.
Yes and there is nothing wrong with reaching that goal, the problem is that they did it at the expensive of the users freedom. Distributions still have no good way of dealing with software that isn't provided by them.
No it isn't. You can side-load software using any other repository system if you like (see e.g. bedrock linux).
Yes, I am free to not use the repository system that doesn't work, just like on iPhone or any other proprietary system. If I work around what the system offers then I gain more freedom. That's not a good thing, that's the problem. The repository system should be the very thing that gains me freedom, not the thing that gets in the way of it.
They provide complete sources.
Which are nearly worthless as the system doesn't provide a clean way for me to modify them and reintroduce them into the system. The next time I ran a upgrade my modification will be overwriting by the distribution. And when I put packages on hold or increase the epoch version I cut myself of from security updates and such. Repositories are not build for user modification, they are build around a central authority that provides everything and decides everything. The user can comply or switch to another distri that suffers from much the same problems.
Yes and I am also free to jailbreak an iPhone. Just because I can do it, doesn't mean that the system is welcoming to third party software. The monolithic dependency trees used in the Free Software world are worse then what you see in the proprietary world. With Windows, as messy at it is, installing third party software isn't difficult and Microsoft goes to great length to ensure backward and forward compatibility to OS released a decade ago. It's not always works out of course, but it's a lot better then Linux world where every distribution is incompatible to everything else and even to past and future versions of itself.
I don't have to worry about apt-get install doing dangerous stuff.
Yes, but it does so by being a walled garden. The distri decides what goes into the repository, what version of the program and so on. As a user I can just consume what they provide me, my ability to change or object what they do is extremely limited and any more complex change will break the monolithic dependency tree. It's a system that violates everything Free Software should stands for, it works, but it doesn't give much freedom at all.
Not disagreeing on the overall point, one problem however with the Free Software world is that those fights often go into the wrong direction. Take the app model on iOS and Android, on one side it's a framework that is used to provide a walled garden, on the other side the clean isolation of apps from each other drastically reduces the potential for abuse and in turn provides a flourishing ecosystem of applications. On your average Free Software system by contrast you always have to worry about a 'make install' wreaking havoc, as there is no limit to what it can do. If you run a program you hand your system over to that program, you are no longer in control, you have to trust the software to do the right thing. This goes wrong all the time without even malicious intent, a wrong library here or there and stuff will break.
On the OLPC, which had a similar isolation model for apps, there was this idea of having a 'source' button on the keyboard, you could press it and it automatically jump you to the source code. You could modify things without wreaking havoc and you could share your modified apps with your friends with a click of a button. It never got fully implemented from what I understand, but as a concept it's still an incredible interesting idea for how Free Software should work. A system bulid from the ground up to encourage modification instead of a system where modifications are complicated, time consuming and dangerous. If people here of a new cool app on iOS and Android, they can try it with a few button presses. On a Free Software system I head over to Github, see a long list of required dependencies and essentially give up at that point, as more often then not, it's just not worth the effort.
Instead of fighting against stuff, Free Software would do well focusing more on building the better system. A system where all the software is under a Free license should make it easier to share stuff, but instead it's often harder. All the freedom the license give doesn't really trickle down to the user.
This is just one example, there are plenty other areas, like the fight against Javascript. Javascript is not the enemy, it should be seen as a building block for the semantic web, as it allows to decouple the data from the presentation, something that HTML claimed to do, but completely failed to do in reality.
Last time I checked about 5% of Steam gamers had a PC meeting the recommended specs from Oculus. That was a few month ago, so it might be a bit more by now, but I am pretty sure that it's still a tiny minority of gamers, and far less when it comes to regular PC users.
They made some weird deal with Microsoft that got them the gamepads really cheap or at least that's what they say. It's kind of weird, as they also include another small remote-like controller in the box as well. The gamepad feels completely redundant and unnecessary. And with Oculus Touch yet another controller will be released in a few month.
Or, guy with unfinished product has no idea of cost but needs to get investors to keep giving him money so he can finish building it.
They already shipped the DK1 ($300) and DK2 ($350) headsets at that point, so they should have had a reasonable good idea how much it would cost. We are not talking about a Kickstarter gone crazy and unexpected cost overruns, they deliberately build the thing outside of the originally intended price range, since they already knew how to build one that fits the target.
This graph seems to indicate that the problem of lack of women in computer science correlated pretty directly with the rise of the home computer and video games in the mid 80's.
What is wrong with discussing a "gender empathy gap"
What has that to do with Open Source? I have written Open Source code for 15+ years and never seen any gender issues pop up ever. Nobody cares what gender you are. You are judged by your patches and the usefulness of your contributions and nothing else. When conflicts happens, they happen for technical reasons or religious reasons (vi vs Emacs), not for gender reasons. And if somebody is really unhappy with how an Open Source project is going, they fork it or spend their time on a creation of their own. There is no gender check when you want to create a Github repo. You click a few buttons and start to submit code. Maybe somebody will find that code useful or maybe not. It's your repo and you can do whatever you want with it. You do not need anybodies approval to write Open Source code.
The biggest difference is that ZFS has full knowledge of the state of the file system, rsync on the other side doesn't, it's stateless, it has to start from zero each time and regather the information on each and every run on both sides, which is a really slow and potentially error prone process (i.e. when files change while rsync runs). ZFS knows what's going on in the filesystem and its snapshots the filesystem at a single point in time, so it thus it can be be far quicker and won't produce inconsistencies in the transmitted data. Tracking renames would speed some things up a little, but to match ZFS it would need a way to get the information about what changed in the filesystem from the filesystem itself and at the moment such functionality isn't available.
Secondly: What the hell happened to Nolok?
The same as what happened with the rest of the project, it kind of stalled. There was a lot of development after Milestone1, but that never converged into a finished release and then staled, developers left, new joined, etc., the release now is essentially a snapshot of that incomplete thing with a bit of polish. Nolok is still around and maybe one day I or somebody else will implement it, but there really isn't enough coherence in the development at the moment for that. Here is an animation test from seven years ago.
You included the .git tree in your count, so you are counting everything twice. For the rest you can blame git as well, it stores and downloads the complete history, so everytime a music file gets updated a little, you end up with a duplicate music file in the project history, same with the graphics. Git isn't very good at dealing with changing binary files and none of the half dozen workarounds (annex, large file, etc.) are worth the effort to save a hundred megabytes.
The bloatiness is the result of a lack of cleanup, the project accumulated lots of graphics over the years, lots of that isn't really used or critical, but might break some random user contributed level if removed. So old graphics stay in for fear of breaking stuff. This game is also very "community driven", meaning everybody can contribute some crappy thing and that then ends up in the source tree, rejection of bad contribution doesn't really happen, unless they break something major.
Biggest chunk of the size is however simply the music, which comes along in very high quality compression, there is no reason you couldn't up the compression ratio and make the package half the size. Some audiophile might notice, I didn't.
There is of course also the plain old reason that tech advances, 1280x720 true color graphics take a lot more space then 320x200 256 color graphics. Same with midi music vs ogg. Also the game uses plain PNGs everywhere, you could of course use compressed textures or if you really wanted to JPEGs for some graphics, but that isn't really worth the effort and makes modifications harder.
Not that big, you only download a 50GB game from Steam like once a month, probably a lot less. HD streaming on the other side you can have running for hours each day (average American watches like 4h TV a day).
News about ad networks doing evil tracking and such comes up all the time, but I have to wonder what all that data is any good for? I don't use ad block and yet all the ads I see tend to be either be pretty unimpressive and untargeted. They might know that I am male and they might be clever enough to show video games advertisement on a video game site, but that's about it. In a lot of ways the targeting makes the ads actually worse, as they end up covering a far narrower range of products and lead to a lot of repetition. Youtube can be notoriously bad at this, showing you the same ad 20 times in a row. There are also some ads that are extremely personalized, but they are unimpressive in the other direction, Amazon ads for example will just show me products that I just watched on their site. So won't show me anything new, just stuff I am already familiar with.
Never seen ads that actually extrapolate my behavior and interest and end up recommending me a product that I would actually be interested in buying. I found good old untargeted advertisement on TV or magazines far more useful for that, as that showed a lot of products that I didn't even know existed.
I think it's not so much that it isn't fast enough, but that you end up with a big box with a big spinning mirror. What he wants to do is to take the same technology and replace the spinning mirror with a DLP-style micromirror, thus you could do the same thing much smaller, cheaper and with less power, which in turn would allow you to have more of those devices for better tracking.
Overclocking processors generally voids the warranty and in general overclocking isn't something that appears in marketing materials for most or all processors.
Actually the box of my AMD FX-8320 reads:
The joke is when you read the included papers inside of the box:
Differences when compiled can be shrugged off as different compiler versions.
Yep, and it's not just different compilers, time stamps, compile order on parallel builds, the order of files in the filesystems, install path, compile flags, etc. will all change the resulting binary. Reproducible builds just hasn't been a thing in Free Software community and only very recently did Debian start work on ensuring that their binaries are byte-for-byte reproducible, but that's of course just Debian, we are still far far away from having reproducible builds be the default way how Free Software binaries are distributed.
Can you use a retro reflector to kill the laser before the laser kills the mirror?