buy a playstation and a cheap Linux boxen, no fancy hardware needed, except a good CPU and lots of ram [...] Hook those two units up to a switch able input display
You appear to recommend a PlayStation 4 game console as a substitute for a gaming GPU and Windows license on a desktop PC. So where does that leave laptop users? Does the PlayStation Vita have a good selection of games?
The workaround for Restricted Boot, which I define as UEFI Secure Boot that a PC's owner cannot reconfigure, is to buy a different make and model PC without Restricted Boot. But that fails if all close substitutes also have Restricted Boot. So which PC form factors are more likely to have Restricted Boot? Is it mostly, say, laptops smaller than 12 inches or with a detachable keyboard?
Thank you for bearing with me on this line of discussion.
To be technically correct those chromebooks that are alive and well have linux on them (under the google stuff).
Likewise, to be "the best kind of correct", Android devices run Linux, and many DVRs and "smart" TVs run Linux. Perhaps Richard Stallman was right that "Linux" is too ambiguous as the designation of a personal computing platform.
I was referring to X11/Linux because not all applications that I regularly use under Xubuntu on my Dell Inspiron mini 1012 netbook have been rewritten in JavaScript as Chrome Apps to run under Chrome OS, which is the only "Linux distribution" that a Chromebook in OS verification mode can run. In fact, with the announced deprecation of Chrome Apps in Chrome for Windows, macOS, and X11/Linux, the Chrome App platform has become less attractive to developers because they can no longer target both Chrome OS and traditional desktop and laptop platforms with one development effort.
Or do you consider it practical to buy a Chromebook, put it in developer mode, and use Crouton or similar to install an X11/Linux distribution in a chroot? In that case, the bootloader of a Chromebook makes it too easy for anyone who touches the machine to switch from developer mode back to operating system verification. It in fact encourages the user to press Space then Enter rather than allowing developer mode to boot, and because switching modes wipes the machine, it puts my uncommitted changes and the use of applications installed on the machine at risk.
So let's start from square one, as you mentioned earlier: Given apps, you choose an OS, and then you choose hardware to run that OS. The apps I want to run are a web browser, a text editor, Python with Pillow Imaging Library and IDLE, ca65, GNU Make and Coreutils, and FCEUX debugging version. All run on Xubuntu: FCEUX debugging version in Wine and the rest natively. But all also run on Windows, as MSYS provides GNU Make and Coreutils on Windows. And as far as I can tell, more 10-11 inch laptops manufactured in 2016 are warranted to run Windows than to run a version of Linux that supports GNU, X11, and Wine.
If you want Ubuntu or whatever Lenovo and I'm sure others support it on netbook sized machines
Which machines in the list you cited are netbooks that are still manufactured? Or do I need to go find a list of all Lenovo netbooks and Ctrl+F for each in the list you cited?
So when have you ever been unable to run whatever software you want in Windows? Use what ever hardware you want?
I can't use hardware that lacks a Windows driver, such as hardware that I developed myself. Developing device drivers for Windows 10 requires paying a recurring fee for an EV code signing certificate and the corporation or LLC to become eligible for said certificate.
you chose the OS to fit what you want to run and not the other way around.
And you choose the hardware to run a particular operating system. The trouble is that a lot of hardware form factors, such as laptops smaller than 12 inches, don't have an offering by an X11/Linux-specialist company such as System76.
Ironically, I can get my old windows software to work on Macs & Linux boxen via WINE easier
That's fine if you want to run old user mode software. To run anything newer or which needs a bespoke driver, such as recent versions of the client for the iTunes Store or the Fitbit client, you need genuine Windows or something else that replicates its driverBI, not just its ABI. Have you tried ReactOS, which claims to support Windows binary interfaces more thoroughly than Wine does? If so, have your experiences been useful?
"Up until recently" is a big one: ability to develop, test, and deploy drivers for low-volume hardware that you have built as an individual hobbyist. Linux and *BSD allow this; Windows doesn't because as of Windows 10, all new drivers must be signed with an EV code signing certificate, which is available only to established companies, not individuals.
Let me try to guess the Linux Foundation's reasoning: Without JavaScript, developers are more likely to make Windows-exclusive apps instead of web apps that can be run in a browser on X11/Linux.
If by "noscript.net" you meant whitelisting only those applications with a legitimate need for script, I'd agree. But if by "noscript.net" you meant that no application has a legitimate need for script, I'd have to disagree.
Would you prefer that JavaScript applications be instead served as static HTML pages styled with CSS, with all logic server-side? A web-based paint program, for example, would need to use a server-side image map instead of Canvas and operate by click-wait-click-wait-click instead of dragging. A web-based video game would have to be turn-based. And a Canvas-based vector animation viewer (which is at least better than Flash) would have to render the sequence of frames to WebM on the server side, which bloats the video by a factor of ten in my tests.
Or would you prefer that applications be instead written in the platform's preferred language and UI library? Good luck running a Mac app written in Swift and using Cocoa Touch API on your Windows PC or a Windows app written in C# and using WPF on your Mac.
Let's Encrypt is rate-limited in such a way that it's only "free" if you own a valid domain. Someone setting up a web server on a private network, such as inside a home, library, or museum, might not own a domain for that purpose.
Self-signed is OK for [Google], but not for us. Get it?
Any company can join the major web browsers' root certificate programs so long as it can afford the cost of operating a CA and hiring a third-party auditor to verify that its issuance policy is being followed. Google is such a company.
Then the problem becomes setting up the means through which the CA's root certificate is "pre-distributed over a secure sideband", such as a head of household wanting to make a private server available to visiting friends and family or a public library wanting to make a private server available to visiting patrons.
OCSP stapling means the server contacts the CA on the client's behalf and returns a cached OCSP response signed by the CA to the client. Thus the CA sees one OCSP request from the server per day as the server notices that the cached response is about to expire, as opposed to a request from each client. But in the case that Anonymous Coward #53085831 described, both the server and the client are on a LAN behind a NAT. When both the client and server have the same IPv4 address, stapling isn't quite as effective at hiding clients.
There is no K in Yamaha, maker of motorcycles and FM synthesizers. It might be what your computer's spell checker recommended, but many spell checkers fail to take into account phonics or context, instead focusing on presses of a key near the presumably intended one. A "yarmulke", a beanie worn by Jewish men, was probably meant.
What are you buying that Mac for? If you're developing iPhone / iPad apps - sure.
Or for testing web applications in Safari for macOS. Chrome is an imperfect proxy because of how much Blink and WebKit have diverged in the three and a half years since the fork.
How can they claim the overall PC market is down 16% when they've excluded a significant segment that's seen year-over-year growth from that statistic?
Because devices in said "significant segment that's seen year-over-year growth from that statistic" are less useful for PC-only applications, particularly high school programming homework.
Furthermore, Google has announced a plan to remove support for Chrome Apps from the Google Chrome browser on traditional PC operating systems (Windows, macOS, and X11/Linux). After the Chrome Apps switchoff, developers can't target Chrome and get both traditional PC and Chromebook support; they have to choose one or the other. This means a lot of developers of applications for devices with a keyboard are likely to give up on Chrome OS in favor of exclusively targeting traditional PCs.
Slashdot Mirror
buy a playstation and a cheap Linux boxen, no fancy hardware needed, except a good CPU and lots of ram [...] Hook those two units up to a switch able input display
You appear to recommend a PlayStation 4 game console as a substitute for a gaming GPU and Windows license on a desktop PC. So where does that leave laptop users? Does the PlayStation Vita have a good selection of games?
The workaround for Restricted Boot, which I define as UEFI Secure Boot that a PC's owner cannot reconfigure, is to buy a different make and model PC without Restricted Boot. But that fails if all close substitutes also have Restricted Boot. So which PC form factors are more likely to have Restricted Boot? Is it mostly, say, laptops smaller than 12 inches or with a detachable keyboard?
Thank you for bearing with me on this line of discussion.
To be technically correct those chromebooks that are alive and well have linux on them (under the google stuff).
Likewise, to be "the best kind of correct", Android devices run Linux, and many DVRs and "smart" TVs run Linux. Perhaps Richard Stallman was right that "Linux" is too ambiguous as the designation of a personal computing platform.
I was referring to X11/Linux because not all applications that I regularly use under Xubuntu on my Dell Inspiron mini 1012 netbook have been rewritten in JavaScript as Chrome Apps to run under Chrome OS, which is the only "Linux distribution" that a Chromebook in OS verification mode can run. In fact, with the announced deprecation of Chrome Apps in Chrome for Windows, macOS, and X11/Linux, the Chrome App platform has become less attractive to developers because they can no longer target both Chrome OS and traditional desktop and laptop platforms with one development effort.
Or do you consider it practical to buy a Chromebook, put it in developer mode, and use Crouton or similar to install an X11/Linux distribution in a chroot? In that case, the bootloader of a Chromebook makes it too easy for anyone who touches the machine to switch from developer mode back to operating system verification. It in fact encourages the user to press Space then Enter rather than allowing developer mode to boot, and because switching modes wipes the machine, it puts my uncommitted changes and the use of applications installed on the machine at risk.
So let's start from square one, as you mentioned earlier: Given apps, you choose an OS, and then you choose hardware to run that OS. The apps I want to run are a web browser, a text editor, Python with Pillow Imaging Library and IDLE, ca65, GNU Make and Coreutils, and FCEUX debugging version. All run on Xubuntu: FCEUX debugging version in Wine and the rest natively. But all also run on Windows, as MSYS provides GNU Make and Coreutils on Windows. And as far as I can tell, more 10-11 inch laptops manufactured in 2016 are warranted to run Windows than to run a version of Linux that supports GNU, X11, and Wine.
If you want Ubuntu or whatever Lenovo and I'm sure others support it on netbook sized machines
Which machines in the list you cited are netbooks that are still manufactured? Or do I need to go find a list of all Lenovo netbooks and Ctrl+F for each in the list you cited?
Came, past tense. Used Linux netbooks exist. New ones do not.
So when have you ever been unable to run whatever software you want in Windows? Use what ever hardware you want?
I can't use hardware that lacks a Windows driver, such as hardware that I developed myself. Developing device drivers for Windows 10 requires paying a recurring fee for an EV code signing certificate and the corporation or LLC to become eligible for said certificate.
Was it a 16-bit executable on 64-bit Windows? If so, use DOSBox and a set of Windows 3.1 install diskettes.
you chose the OS to fit what you want to run and not the other way around.
And you choose the hardware to run a particular operating system. The trouble is that a lot of hardware form factors, such as laptops smaller than 12 inches, don't have an offering by an X11/Linux-specialist company such as System76.
How does MinGW or Clang fail on Windows?
And please don't bring the GNU tools in to this -- they wouldn't exist in a meaningful form without Linux.
GCC and GNU Binutils would exist on *BSD and Solaris, and Linux probably would not have existed, were it not for the AT&T lawsuit.
WINE
APPS LISTED AS GARBAGE
Ironically, I can get my old windows software to work on Macs & Linux boxen via WINE easier
That's fine if you want to run old user mode software. To run anything newer or which needs a bespoke driver, such as recent versions of the client for the iTunes Store or the Fitbit client, you need genuine Windows or something else that replicates its driverBI, not just its ABI. Have you tried ReactOS, which claims to support Windows binary interfaces more thoroughly than Wine does? If so, have your experiences been useful?
The licensing cost for the client version of the Windows operating system is subsidized by trialware publishers. The end user does not see this cost.
"Up until recently" is a big one: ability to develop, test, and deploy drivers for low-volume hardware that you have built as an individual hobbyist. Linux and *BSD allow this; Windows doesn't because as of Windows 10, all new drivers must be signed with an EV code signing certificate, which is available only to established companies, not individuals.
How did a USB hub fail when you tried it?
Let me try to guess the Linux Foundation's reasoning: Without JavaScript, developers are more likely to make Windows-exclusive apps instead of web apps that can be run in a browser on X11/Linux.
Why do you consider it desirable to shut down the ability to make an app once and run it on Windows, macOS, X11/Linux, iOS, and Android?
If by "noscript.net" you meant whitelisting only those applications with a legitimate need for script, I'd agree. But if by "noscript.net" you meant that no application has a legitimate need for script, I'd have to disagree.
Would you prefer that JavaScript applications be instead served as static HTML pages styled with CSS, with all logic server-side? A web-based paint program, for example, would need to use a server-side image map instead of Canvas and operate by click-wait-click-wait-click instead of dragging. A web-based video game would have to be turn-based. And a Canvas-based vector animation viewer (which is at least better than Flash) would have to render the sequence of frames to WebM on the server side, which bloats the video by a factor of ten in my tests.
Or would you prefer that applications be instead written in the platform's preferred language and UI library? Good luck running a Mac app written in Swift and using Cocoa Touch API on your Windows PC or a Windows app written in C# and using WPF on your Mac.
Automated renewal is the intent. In practice, it took several months after Let's Encrypt entered public beta for some web hosting providers to let users even upload their own certificates without having to file a support ticket. (See, for example, a blog post from a month ago.) It got so bad that one passive-aggressive fellow wrote a tool to request a certificate from Let's Encrypt and automatically file a support ticket.
Let's Encrypt is rate-limited in such a way that it's only "free" if you own a valid domain. Someone setting up a web server on a private network, such as inside a home, library, or museum, might not own a domain for that purpose.
How would you recommend that developers "learn what usability and accessibility mean" without observing users?
Have you run analytics on how many potential customers you are turning away for not supporting TLS 1.2?
Self-signed is OK for [Google], but not for us. Get it?
Any company can join the major web browsers' root certificate programs so long as it can afford the cost of operating a CA and hiring a third-party auditor to verify that its issuance policy is being followed. Google is such a company.
Then the problem becomes setting up the means through which the CA's root certificate is "pre-distributed over a secure sideband", such as a head of household wanting to make a private server available to visiting friends and family or a public library wanting to make a private server available to visiting patrons.
OCSP stapling means the server contacts the CA on the client's behalf and returns a cached OCSP response signed by the CA to the client. Thus the CA sees one OCSP request from the server per day as the server notices that the cached response is about to expire, as opposed to a request from each client. But in the case that Anonymous Coward #53085831 described, both the server and the client are on a LAN behind a NAT. When both the client and server have the same IPv4 address, stapling isn't quite as effective at hiding clients.
There is no K in Yamaha, maker of motorcycles and FM synthesizers. It might be what your computer's spell checker recommended, but many spell checkers fail to take into account phonics or context, instead focusing on presses of a key near the presumably intended one. A "yarmulke", a beanie worn by Jewish men, was probably meant.
What are you buying that Mac for? If you're developing iPhone / iPad apps - sure.
Or for testing web applications in Safari for macOS. Chrome is an imperfect proxy because of how much Blink and WebKit have diverged in the three and a half years since the fork.
How can they claim the overall PC market is down 16% when they've excluded a significant segment that's seen year-over-year growth from that statistic?
Because devices in said "significant segment that's seen year-over-year growth from that statistic" are less useful for PC-only applications, particularly high school programming homework.
Furthermore, Google has announced a plan to remove support for Chrome Apps from the Google Chrome browser on traditional PC operating systems (Windows, macOS, and X11/Linux). After the Chrome Apps switchoff, developers can't target Chrome and get both traditional PC and Chromebook support; they have to choose one or the other. This means a lot of developers of applications for devices with a keyboard are likely to give up on Chrome OS in favor of exclusively targeting traditional PCs.