No, which is why you maintain just the one that functions properly without javascript.
Like I said, for many reasonably complex applications that is basically 2 separate code bases.
I think AC means make and maintain the scriptless code base, and don't start developing the script-required code base at all. If you develop only the scriptless code base, you don't have to maintain "2 separate code bases."
No, which is why you maintain just the one that functions properly without javascript.
Most people prefer the dragging map navigation that requires JavaScript to the screen-at-a-time map navigation that does not require JavaScript. If your map site operates exclusively in the screen-at-a-time mode, the majority of your users will leave your site in favor of a competitor's site that offers the dragging mode.
I don't want a general purpose interface. It causes problems like all the web security issues of the past 20 years.
Have fun creating your own apps for whatever you use.
As opposed to at least once for IE, once for Chrome, once for Safari, once for Firefox, once for iOS, once for Android, and once for UWP?
The IE 11, Chrome, Safari, and Firefox versions of a web application will share much more code than the Win32, UWP, macOS, and X11/Linux versions of a native application.[1] The Chrome version can be accessed through Chrome for Android. There's also less of a recurring fee for a web application's TLS certificate than for software publisher certificates on iOS App Store, Mac App Store, Windows Store, and Windows SmartScreen, as a web application's certificate can be domain-validated and automatically renewed without charge. Nor is there a delay of days or weeks for a third party to review bug fixes or new features.
Why bother with X11/Linux? It's user base is so small, it isn't worth the effort.
You are telling a user of X11/Linux both at home (Debian) and at work (Ubuntu) that he "isn't worth the effort."
I know that Visual Studio can compile for UWP and Windows Desktop.
Can it compile the same code base for both?
By the way, the world is on Windows 10, not 7. Get with the times.
By "once for Windows 7", I was referring to a Windows desktop application as opposed to a UWP application. A UWP application targets users of Windows 10 and Windows 10 S. A Windows desktop application targets users of Windows 10 and Windows 7. Though Windows 10 recently edged out Windows 7 in web browser usage share,[2] users of Windows 7 greatly outnumber users of Windows 10 S. Thus if a developer has the budget for only Windows desktop or UWP, he's more likely to conclude that Windows 10 S "isn't worth the effort."
Meanwhile, there are products that allow one to write the code once and compile it for multiple operating systems. There is one that plugs into Visual Studio that allows one to compile for Windows, UWP, iOS, Android, and I think MacOS.
A developer needs to buy the product and buy the publisher certificate for each platform. Even if Google can afford it, a small developer is less likely to be able to.
I see you don't know that HTTPS is a bolt-on to HTTP.
What's technically wrong with the design of this particular bolt-on?
[1] Unless it uses the Electron framework, which is a copy of Chromium hardcoded to view one website. The "native" desktop applications for Skype, Slack, and Discord all use it. I consider Electron to be bloatware. [2] Source: "Windows 10 Visits to US Government Sites Surpass Windows 7 for the First Time", dated a week ago
The "individual apps" you're talking about are browsers.
Google Earth is not a browser. Slashdot is not a browser. Twitter is not a browser. They are web applications that happen to be accessed through a browser.
That's literally the topic of the conversation you're in: Apple won't let non-Webkit browsers run on iOS.
And my workaround for the fact that "Apple won't let non-Webkit browsers run on iOS" is to make native client applications instead of web applications.
I'm very much aware of shared libraries. But I was under the impression that compilers could not perform link-time or whole-program optimization across a.so or.dll boundary. I was also under the impression that application submission policies on iOS, Windows Store (UWP), and game consoles required applications to be "self-contained", not linking to any.so or.dll files that do not ship with the operating system. (For example, see Apple's App Store Review Guidelines 2.5.2 beginning "Apps should be self-contained in their bundles".) The easiest way I know to ensure whole-program optimization and self-containment of executable code is to link the application statically, except for libraries that ship with the operating system.
With JavaScript outside Greasemonkey, it's also a good practice to combine multiple scripts into one file to reduce HTTP request round trips and the likelihood of dependency race conditions.
In a properly written contract, "the fine for breach of contract" is the insured value of the house. Intentionally allowing it to burn would allow your house insurer to collect from the nonfeasant private fire department for what amounts to arson.
trying to do some crappy workaround with arrows on each edge of the map where you click an arrow and the entire page refreshes with the map moved that direction is a stupid way to avoid Javascript.
Some anti-JavaScript hardliners here and on SoylentNews have stated that they actually prefer what you call "a stupid way to avoid Javascript." Or they would prefer to download, audit, compile, and install a native map viewer application distributed in source code form.
It appears you object to integrating audio and video features into a text based platform. Would have preferred making a separate audio and video based platform for use alongside the text based platform? If so, what would this platform have looked like?
Hangouts is one of those obviously-silly-on-the-face-of-it services for people who don't realize that users can already communicate with one another using standard protocols and therefore nobody needs another fucking "chat website."
Sometimes these FCWs, such as Hangouts, Skype, Slack, and Discord, fill perceived functionality gaps in Internet Relay Chat (IRC) and other standard chat protocols. For example, how can one set up a SIP call over IRC? Or view messages sent to you or to a group while you were offline or were using a different device? Or attach files when both sides are behind a firewall?
The use of the browser as a general purpose interface is stupid and dangerous
What general-purpose interface that isn't a web browser would you prefer? Or would you prefer to require each application's developer to make the application six times: once for Windows 7 desktop, once for UWP, once for macOS, once for iOS, once for X11/Linux, and once for Android?
https://tech.slashdot.org
People like you are using an unencrypted text based and inherently insecure technology
I guess some of the confusion about the meaning of "free market" comes from the habit of some right-libertarians to confuse it with "unregulated market" when they fail to see monopolist practice as private coercion.
This "new API, which Mozilla gave more then a year's notice of," launched without counterparts to several categories of functionality present in the old API. This was despite extension developers giving Mozilla "more then a year's notice of" the fact that these categories of functionality were missing in the new API.
Need a specific example? Let me know when the request for a way to rebind shortcuts becomes RESOLVED FIXED or even ASSIGNED. Right now, it's marked as "NEW" which means "will not be worked on by staff".
Several business applications support certain features in Chrome and not Firefox, or have Chrome extensions but not Firefox extensions.
I thought the WebExtensions transition was intended to let Chrome extension developers upload their extensions to addons.mozilla.org nearly unchanged. If a particular extension's publisher has stated its business decision to deliberately make it exclusive to Chrome Web Store, please name and shame.
Perhaps once you're done conflating George Orwell's Animal Farm ("four legs good, two legs bad^W better") with National Lampoon's Animal House ("Toga! Toga! Toga!"), you could add in Nintendo's Animal Crossing for good measure.
Ctrl+Q in Firefox for Linux is mapped to "immediately close all open HTML documents." But it's adjacent to Ctrl+W (close one document) and Ctrl+Tab (switch to next open document in the same window). Some Slashdot users claim that Firefox on some platforms instead binds quit to Ctrl+Shift+Q, but this is still adjacent to Ctrl+Shift+Tab (switch to previous open document in the same window). "Restore Previous Session" restores which documents were open, but it doesn't always restore changes made by scripts to the DOM of those documents, nor data entered into unsubmitted forms in those documents, especially if the form was added to the document by a script. Slashdot D2 comment forms are one example of this, as are comment forms on Explosm.net (the home of the webcomic Cyanide & Happiness).
There used to be an extension called "Keybinder" to disable the Ctrl+Q shortcut for quit. But as described on the README of its GitHub repository, it didn't make the transition because WebExtensions don't support anything analogous to XUL keysets. This is bug 1325692, which was marked "wontfix" for Firefox 57.
Slashdot Mirror
No, which is why you maintain just the one that functions properly without javascript.
Like I said, for many reasonably complex applications that is basically 2 separate code bases.
I think AC means make and maintain the scriptless code base, and don't start developing the script-required code base at all. If you develop only the scriptless code base, you don't have to maintain "2 separate code bases."
I'm fairly certain that rtb61 was referring to YouTube, a video host operated by Google.
No, which is why you maintain just the one that functions properly without javascript.
Most people prefer the dragging map navigation that requires JavaScript to the screen-at-a-time map navigation that does not require JavaScript. If your map site operates exclusively in the screen-at-a-time mode, the majority of your users will leave your site in favor of a competitor's site that offers the dragging mode.
Is this the second coming of Kuro5hin?
I don't want a general purpose interface. It causes problems like all the web security issues of the past 20 years.
Have fun creating your own apps for whatever you use.
As opposed to at least once for IE, once for Chrome, once for Safari, once for Firefox, once for iOS, once for Android, and once for UWP?
The IE 11, Chrome, Safari, and Firefox versions of a web application will share much more code than the Win32, UWP, macOS, and X11/Linux versions of a native application.[1] The Chrome version can be accessed through Chrome for Android. There's also less of a recurring fee for a web application's TLS certificate than for software publisher certificates on iOS App Store, Mac App Store, Windows Store, and Windows SmartScreen, as a web application's certificate can be domain-validated and automatically renewed without charge. Nor is there a delay of days or weeks for a third party to review bug fixes or new features.
Why bother with X11/Linux? It's user base is so small, it isn't worth the effort.
You are telling a user of X11/Linux both at home (Debian) and at work (Ubuntu) that he "isn't worth the effort."
I know that Visual Studio can compile for UWP and Windows Desktop.
Can it compile the same code base for both?
By the way, the world is on Windows 10, not 7. Get with the times.
By "once for Windows 7", I was referring to a Windows desktop application as opposed to a UWP application. A UWP application targets users of Windows 10 and Windows 10 S. A Windows desktop application targets users of Windows 10 and Windows 7. Though Windows 10 recently edged out Windows 7 in web browser usage share,[2] users of Windows 7 greatly outnumber users of Windows 10 S. Thus if a developer has the budget for only Windows desktop or UWP, he's more likely to conclude that Windows 10 S "isn't worth the effort."
Meanwhile, there are products that allow one to write the code once and compile it for multiple operating systems. There is one that plugs into Visual Studio that allows one to compile for Windows, UWP, iOS, Android, and I think MacOS.
A developer needs to buy the product and buy the publisher certificate for each platform. Even if Google can afford it, a small developer is less likely to be able to.
I see you don't know that HTTPS is a bolt-on to HTTP.
What's technically wrong with the design of this particular bolt-on?
[1] Unless it uses the Electron framework, which is a copy of Chromium hardcoded to view one website. The "native" desktop applications for Skype, Slack, and Discord all use it. I consider Electron to be bloatware.
[2] Source: "Windows 10 Visits to US Government Sites Surpass Windows 7 for the First Time", dated a week ago
don't port Chrome; instead, port Google Earth.
The "individual apps" you're talking about are browsers.
Google Earth is not a browser. Slashdot is not a browser. Twitter is not a browser. They are web applications that happen to be accessed through a browser.
That's literally the topic of the conversation you're in: Apple won't let non-Webkit browsers run on iOS.
And my workaround for the fact that "Apple won't let non-Webkit browsers run on iOS" is to make native client applications instead of web applications.
I'm very much aware of shared libraries. But I was under the impression that compilers could not perform link-time or whole-program optimization across a .so or .dll boundary. I was also under the impression that application submission policies on iOS, Windows Store (UWP), and game consoles required applications to be "self-contained", not linking to any .so or .dll files that do not ship with the operating system. (For example, see Apple's App Store Review Guidelines 2.5.2 beginning "Apps should be self-contained in their bundles".) The easiest way I know to ensure whole-program optimization and self-containment of executable code is to link the application statically, except for libraries that ship with the operating system.
With JavaScript outside Greasemonkey, it's also a good practice to combine multiple scripts into one file to reduce HTTP request round trips and the likelihood of dependency race conditions.
In a properly written contract, "the fine for breach of contract" is the insured value of the house. Intentionally allowing it to burn would allow your house insurer to collect from the nonfeasant private fire department for what amounts to arson.
trying to do some crappy workaround with arrows on each edge of the map where you click an arrow and the entire page refreshes with the map moved that direction is a stupid way to avoid Javascript.
Some anti-JavaScript hardliners here and on SoylentNews have stated that they actually prefer what you call "a stupid way to avoid Javascript." Or they would prefer to download, audit, compile, and install a native map viewer application distributed in source code form.
It appears you object to integrating audio and video features into a text based platform. Would have preferred making a separate audio and video based platform for use alongside the text based platform? If so, what would this platform have looked like?
What happens when you visit the URL in Chrome (which is a Safari wrapper) on an iPad?
Hangouts is one of those obviously-silly-on-the-face-of-it services for people who don't realize that users can already communicate with one another using standard protocols and therefore nobody needs another fucking "chat website."
Sometimes these FCWs, such as Hangouts, Skype, Slack, and Discord, fill perceived functionality gaps in Internet Relay Chat (IRC) and other standard chat protocols. For example, how can one set up a SIP call over IRC? Or view messages sent to you or to a group while you were offline or were using a different device? Or attach files when both sides are behind a firewall?
Some also block Firefox with messages to download Chrome.
Not if you're on Linux. They seem to only bully MS users. They know GNU/Linux users know better.
My experience differs from yours on Ubuntu 16.04 LTS, Firefox 57.0.3 (64-bit), visiting https://www.google.com/earth/
The use of the browser as a general purpose interface is stupid and dangerous
What general-purpose interface that isn't a web browser would you prefer? Or would you prefer to require each application's developer to make the application six times: once for Windows 7 desktop, once for UWP, once for macOS, once for iOS, once for X11/Linux, and once for Android?
https://tech.slashdot.org
People like you are using an unencrypted text based and inherently insecure technology
What's so unencrypted here?
Don't port the browser; instead, port the individual apps. For example, don't port Chrome; instead, port Google Earth.
I guess some of the confusion about the meaning of "free market" comes from the habit of some right-libertarians to confuse it with "unregulated market" when they fail to see monopolist practice as private coercion.
since I can't just copy the files over they don't work. I'm stuck putting everything in one ginormous file.
That's been true for years. Are native applications distributed as a pile of .o files to link together at runtime, or as a single .exe file?
You want it to pass through [the cost of a native iOS rewrite of a web application] to you in terms of more expensive product?
Yes. Offer the web version to the public without charge, and paywall the special version for (generally richer) users of iOS.
I assume the logic is that if a web platform feature is at Candidate Recommendation status at the start of beta, it's likely to be at least a Proposed Recommendation once beta ends.
This is bug 1325692, which was marked "wontfix" for Firefox 57.
Yes? That's not surprising given that Firefox 57 is already out.
Bug 1325692 was reported 2016-12-23, just shy of eleven months before the 2017-11-14 release of Firefox 57.
This "new API, which Mozilla gave more then a year's notice of," launched without counterparts to several categories of functionality present in the old API. This was despite extension developers giving Mozilla "more then a year's notice of" the fact that these categories of functionality were missing in the new API.
Need a specific example? Let me know when the request for a way to rebind shortcuts becomes RESOLVED FIXED or even ASSIGNED. Right now, it's marked as "NEW" which means "will not be worked on by staff".
rewrite it in Swift or Objective-C++
to get any kind of usefull javascript performance you need to write a jit compiler
What does "a jit compiler" have to do with an application that has been rewritten in Swift or ObjC?
Several business applications support certain features in Chrome and not Firefox, or have Chrome extensions but not Firefox extensions.
I thought the WebExtensions transition was intended to let Chrome extension developers upload their extensions to addons.mozilla.org nearly unchanged. If a particular extension's publisher has stated its business decision to deliberately make it exclusive to Chrome Web Store, please name and shame.
Perhaps once you're done conflating George Orwell's Animal Farm ("four legs good, two legs bad^W better") with National Lampoon's Animal House ("Toga! Toga! Toga!"), you could add in Nintendo's Animal Crossing for good measure.
My mileage did vary.
Ctrl+Q in Firefox for Linux is mapped to "immediately close all open HTML documents." But it's adjacent to Ctrl+W (close one document) and Ctrl+Tab (switch to next open document in the same window). Some Slashdot users claim that Firefox on some platforms instead binds quit to Ctrl+Shift+Q, but this is still adjacent to Ctrl+Shift+Tab (switch to previous open document in the same window). "Restore Previous Session" restores which documents were open, but it doesn't always restore changes made by scripts to the DOM of those documents, nor data entered into unsubmitted forms in those documents, especially if the form was added to the document by a script. Slashdot D2 comment forms are one example of this, as are comment forms on Explosm.net (the home of the webcomic Cyanide & Happiness).
There used to be an extension called "Keybinder" to disable the Ctrl+Q shortcut for quit. But as described on the README of its GitHub repository, it didn't make the transition because WebExtensions don't support anything analogous to XUL keysets. This is bug 1325692, which was marked "wontfix" for Firefox 57.