HEIF is like BPG: a single-frame silent movie encoded using HEVC's keyframe encoder. Thus it depends on the same patents as HEVC, making it unusable in free software unless you can afford to move all your users out of countries where these patents are valid.
Or things like Fractal Image Compression where you effectively gain artificial, but perceptible, quality by trading for time and storage space during both an analysis phase and decompression.
The patent thicket around fractal transform compression allowed for more research into DCT, wavelet, and MDCT compression, and this research allowed these paradigms to overtake fractal transform compression.
The blurry cover photo in question is not high res; it is 50 by 38 pixels and 1.5 kB, compared to the full-size cover photo that is 640 by 480 pixels and 72.3 kB. I admit I was wrong about it being an inline data: URI; I had remembered that technique from a faster paint tutorial and assumed it was being applied here as well. But even blurred photos compress fairly well in JPEG because most of their energy is concentrated in low-order DCT terms.
Including a low-resolution image directly in the HTML using a data: URI has two purposes.
Respecting viewers on capped plans
If a server sends an image that the user never scrolls to, the data transmission is wasted. If a server sends an image whose resolution exceeds that of the viewer's display device, the data transmission is wasted. When viewers are on cellular or satellite Internet connections with a usage allowance of 10 GB per month or less, wasted data transmission costs these viewers real money.
Making the site respond faster
First meaningful paint is when all the layout and text are in place above the fold, and things won't move around as more resources load. A site with an earlier first meaningful paint feels faster to viewers. So a site might optimize for an earlier first meaningful paint up by loading basic styles and fonts before images.
At work we throw out perfectly good laptops because they are out of warranty
I tend to keep a laptop longer than that. I replace the battery once it no longer holds a charge, but once the replacement battery no longer holds a charge, I have considered that the time to replace the laptop with one that probably has longer runtime out of the box. Is it practical to expect every PC user to, say, learn how to replace lithium ion cells in a laptop battery pack?
My favorite example is when he said that Disney's Little Mermaid is superior to Disney's Frozen because one comes from classical mythology and one was written by SJWs. They were both written by Hans Christian Anderson.
But does Disney's The Little Mermaid preserve the themes of H. C. Andersen's original story more closely than Frozen keeps the themes of "The Snow Queen"?
The one drawback I can see happens when the quiz covers things presented not in Slashdot's summary but in a paywalled featured article. Not everyone wants to have to spend a nontrivial amount of money for a subscription to NYTimes, LATimes, Washington Post, Wired, or WSJ just to be able to comment on things that appear in the summary or one of the alternate sources.
There are plenty of differences among the revisions of the Super NES chipset. The most obvious from the program's point of view is a bug fix in the DMA controller between CPU version 1 and CPU version 2, and some games reportedly have to slow down somewhat on launch-window consoles to avoid triggering the bug.
But the last revision to the chipset was the "1CHIP", which appeared in the last full-size Super NES consoles as well as the smaller New-Style Super NES (SNS-101). The 1CHIP has the cleanest analog video output, but certain aspects of its behavior are about as different from the common console ("2/1/3", or CPU version 2, PPU1 version 1, PPU2 version 3) as the PlayStation game support in the PlayStation 2 is from a PlayStation. This affects games like Air Strike Patrol. So I'd say that if the low-level emulation is as least as close to a 2/1/3 as the 1CHIP is, it's Good Enough(tm).
You still have a point, however, about what happens to the design once Analogue goes under or if something happens to Kevin Horton.
If I want old games to look good on modern hardware, all I need to do is select the proper options in that emulator
But then you still have to
1. buy a Retrode to get the ROM image from your cartridge, 2. add support to the emulator for whatever coprocessor the cartridge might have (if any, and if it isn't something common like DSP-1, CX4, GSU, or SA1), and 3. deal with greater input lag through a general-purpose PC operating system than this FPGA console would have.
I've talked to Kevin Horton in #nesdev on EFnet, and he explained that the upscaling uses a circular buffer of pixels that adds less than 2 milliseconds of lag.
I see this particular item differently: a stopped-clock moment in which President Trump helped save the world from the United States. After the United States left the Trans-Pacific Partnership, it passed without the harmful copyright provisions on which USTR insisted at the behest of the music and film industry associations of America. How is helping to contain the MAFIAA's push to spread copyright maximalism a bad thing?
It IS a credit reporting system because data may be shared across retailers. If this is the case you absolutely have the right to view and send a written request for review al
You're not APK; I can tell because your writing style doesn't match. But I'll quickly answer why his DNS blocklist solution (whether installed locally or through Pi-hole) isn't quite a complete solution by itself:
Sometimes the publisher itself serves this crap.
A DNS blocklist works when a third-party script displays the popup. But if the same site (e.g. files.slashdot.org serves both things essential to the website's operation (such as style sheets and images) and the popup script, trying to block it will either throw out the baby with the bathwater or send you back to the Netscape 1 web.
Furthermore, the syntax of his preferred blocklist format requires listing each individual hostname to block, not all names in a domain. if the third-party script comes from a random subdomain with a dozen or more hex digits that gets resolved by a wildcard in the DNS zone, this sort of blocklist can't handle all possibilities. The Sandstorm application suite already uses random subdomains for session separation.
Sometimes I submit a support request that a website mistakenly detected the tracking protection built into Firefox as an ad blocker. I tell them that I see ads hosted by the publisher,* such as those on Daring Fireball and those on Read the Docs, and sometimes I click ads hosted by the publisher. But I don't blindly accept scripts that allow third parties to insert arbitrary proprietary scripts that track my "click-stream" from one website to another in order to build an interest profile and try to sell me things I just bought. If a site's ad script requires such tracking in order to run, the site needs to fall back to publisher-hosted ads. Even if publisher-hosted ads have a lower CPM than interest-based ads based on tracking, it's still more than the zero that a site gets if I leave after hitting its adblock wall.
* In the web advertising market, a "publisher" is the operator of a website on which advertisements appear.
Alternatively, they don't. The user configures user-specific style overrides in their UA and the UA applies them depending on sensor output.
With what UI control in each major browser does the user so configure it to apply one of "multiple CSS style sheets, one of them being the default" during high ambient light and the other during low?
For reaching users on Windows desktop, Windows UWP, X11/Linux, Android, macOS, and iOS, which "web application loader/RTS" is superior to the HTTPS + HTML + CSS + JavaScript stack in your opinion?
but in tablets and phones you may want to change the white background to a warm white background if the light is lower, so it does not hurt the human eye or require user adjusting the brightness level.
That would be at the level of the operating system, not the browser, and certainly not an individual web application.
Also, interactive sites and specially WebRTC may want to disable touch input if the proximity sensor flags in
That would also be at the level of the operating system.
Slashdot Mirror
Firefox comes with a basic PDF viewer. So does Google Chrome (and Chromium since third quarter 2014), though Mozilla PDF.js is also available from Chrome Web Store.
Or are Chromium and Firefox also a "bug-fest"?
HEIF is like BPG: a single-frame silent movie encoded using HEVC's keyframe encoder. Thus it depends on the same patents as HEVC, making it unusable in free software unless you can afford to move all your users out of countries where these patents are valid.
Or things like Fractal Image Compression where you effectively gain artificial, but perceptible, quality by trading for time and storage space during both an analysis phase and decompression.
Not only time and storage space but also money, as Iterated Systems demanded substantial royalties for the use of techniques covered by its broad patents. And good luck moving all your users from countries where patents like these are valid to countries where they are not.
The patent thicket around fractal transform compression allowed for more research into DCT, wavelet, and MDCT compression, and this research allowed these paradigms to overtake fractal transform compression.
I think the intended use case is to read documentation on the client device through which you are remotely accessing the server.
The blurry cover photo in question is not high res; it is 50 by 38 pixels and 1.5 kB, compared to the full-size cover photo that is 640 by 480 pixels and 72.3 kB. I admit I was wrong about it being an inline data: URI; I had remembered that technique from a faster paint tutorial and assumed it was being applied here as well. But even blurred photos compress fairly well in JPEG because most of their energy is concentrated in low-order DCT terms.
Including a low-resolution image directly in the HTML using a data: URI has two purposes.
Respecting viewers on capped plans If a server sends an image that the user never scrolls to, the data transmission is wasted. If a server sends an image whose resolution exceeds that of the viewer's display device, the data transmission is wasted. When viewers are on cellular or satellite Internet connections with a usage allowance of 10 GB per month or less, wasted data transmission costs these viewers real money. Making the site respond faster First meaningful paint is when all the layout and text are in place above the fold, and things won't move around as more resources load. A site with an earlier first meaningful paint feels faster to viewers. So a site might optimize for an earlier first meaningful paint up by loading basic styles and fonts before images.At work we throw out perfectly good laptops because they are out of warranty
I tend to keep a laptop longer than that. I replace the battery once it no longer holds a charge, but once the replacement battery no longer holds a charge, I have considered that the time to replace the laptop with one that probably has longer runtime out of the box. Is it practical to expect every PC user to, say, learn how to replace lithium ion cells in a laptop battery pack?
My favorite example is when he said that Disney's Little Mermaid is superior to Disney's Frozen because one comes from classical mythology and one was written by SJWs. They were both written by Hans Christian Anderson.
But does Disney's The Little Mermaid preserve the themes of H. C. Andersen's original story more closely than Frozen keeps the themes of "The Snow Queen"?
The one drawback I can see happens when the quiz covers things presented not in Slashdot's summary but in a paywalled featured article. Not everyone wants to have to spend a nontrivial amount of money for a subscription to NYTimes, LATimes, Washington Post, Wired, or WSJ just to be able to comment on things that appear in the summary or one of the alternate sources.
There are plenty of differences among the revisions of the Super NES chipset. The most obvious from the program's point of view is a bug fix in the DMA controller between CPU version 1 and CPU version 2, and some games reportedly have to slow down somewhat on launch-window consoles to avoid triggering the bug.
But the last revision to the chipset was the "1CHIP", which appeared in the last full-size Super NES consoles as well as the smaller New-Style Super NES (SNS-101). The 1CHIP has the cleanest analog video output, but certain aspects of its behavior are about as different from the common console ("2/1/3", or CPU version 2, PPU1 version 1, PPU2 version 3) as the PlayStation game support in the PlayStation 2 is from a PlayStation. This affects games like Air Strike Patrol. So I'd say that if the low-level emulation is as least as close to a 2/1/3 as the 1CHIP is, it's Good Enough(tm).
You still have a point, however, about what happens to the design once Analogue goes under or if something happens to Kevin Horton.
If I want old games to look good on modern hardware, all I need to do is select the proper options in that emulator
But then you still have to
1. buy a Retrode to get the ROM image from your cartridge,
2. add support to the emulator for whatever coprocessor the cartridge might have (if any, and if it isn't something common like DSP-1, CX4, GSU, or SA1), and
3. deal with greater input lag through a general-purpose PC operating system than this FPGA console would have.
I've talked to Kevin Horton in #nesdev on EFnet, and he explained that the upscaling uses a circular buffer of pixels that adds less than 2 milliseconds of lag.
3. He took us out of TPP
These are all bad things he did.
I see this particular item differently: a stopped-clock moment in which President Trump helped save the world from the United States. After the United States left the Trans-Pacific Partnership, it passed without the harmful copyright provisions on which USTR insisted at the behest of the music and film industry associations of America. How is helping to contain the MAFIAA's push to spread copyright maximalism a bad thing?
There have been small hippie communes that were communist but not authoritarian.
How many of these hippie communes survived for long once the population surpassed 150, the size of one monkeysphere?
There just isn't any point in a toys only store. Even without the internet toys can and will be sold anywhere.
Such as where, particularly outside major cities? Toys R Us had a far larger selection than, say, Walmart.
It IS a credit reporting system because data may be shared across retailers. If this is the case you absolutely have the right to view and send a written request for review al
According The Retail Equation's page to request a Return Activity Report, retailers using the service are supposed to give a transaction ID when they refuse a return.
What examples of "left bias" have you found on Wikipedia that are unsupported by sources that have earned a reputation for fact-checking? They might be in need of bringing them in line with Wikipedia's point of view policy. Or is Wikipedia's guideline for determining "reputation for fact-checking" itself applied in a manner that shows a systemic bias?
You're not APK; I can tell because your writing style doesn't match. But I'll quickly answer why his DNS blocklist solution (whether installed locally or through Pi-hole) isn't quite a complete solution by itself:
Sometimes the publisher itself serves this crap.
A DNS blocklist works when a third-party script displays the popup. But if the same site (e.g. files.slashdot.org serves both things essential to the website's operation (such as style sheets and images) and the popup script, trying to block it will either throw out the baby with the bathwater or send you back to the Netscape 1 web.
Furthermore, the syntax of his preferred blocklist format requires listing each individual hostname to block, not all names in a domain. if the third-party script comes from a random subdomain with a dozen or more hex digits that gets resolved by a wildcard in the DNS zone, this sort of blocklist can't handle all possibilities. The Sandstorm application suite already uses random subdomains for session separation.
Sometimes I submit a support request that a website mistakenly detected the tracking protection built into Firefox as an ad blocker. I tell them that I see ads hosted by the publisher,* such as those on Daring Fireball and those on Read the Docs, and sometimes I click ads hosted by the publisher. But I don't blindly accept scripts that allow third parties to insert arbitrary proprietary scripts that track my "click-stream" from one website to another in order to build an interest profile and try to sell me things I just bought. If a site's ad script requires such tracking in order to run, the site needs to fall back to publisher-hosted ads. Even if publisher-hosted ads have a lower CPM than interest-based ads based on tracking, it's still more than the zero that a site gets if I leave after hitting its adblock wall.
* In the web advertising market, a "publisher" is the operator of a website on which advertisements appear.
The headline says 59 is "biggest", while the summary says 59 builds on Quantum (that is, 57), and Quantum is "biggest".
multiple CSS style sheets, one of them being the default
https://www.w3.org/TR/html401/...
Alternatively, they don't. The user configures user-specific style overrides in their UA and the UA applies them depending on sensor output.
With what UI control in each major browser does the user so configure it to apply one of "multiple CSS style sheets, one of them being the default" during high ambient light and the other during low?
When a computer maker and OS brand starts to set limits on what a CPU and GPU can be used for after buying a computer?
That ship sailed in 1985 with the lockout chips in the Nintendo Entertainment System and Atari 7800 ProSystem.
For reaching users on Windows desktop, Windows UWP, X11/Linux, Android, macOS, and iOS, which "web application loader/RTS" is superior to the HTTPS + HTML + CSS + JavaScript stack in your opinion?
For testing a Linux GUI app that I'm developing while I'm stuck on a laptop whose hardware Linux and X.Org do not support well.
When was it that ignoring the existence of the [<noscript> tag] is the best practice???
Since "progressive enhancement". See "Shouldn't we use <noscript> element?" by Jitendra Vyas et al.
but in tablets and phones you may want to change the white background to a warm white background if the light is lower, so it does not hurt the human eye or require user adjusting the brightness level.
That would be at the level of the operating system, not the browser, and certainly not an individual web application.
Also, interactive sites and specially WebRTC may want to disable touch input if the proximity sensor flags in
That would also be at the level of the operating system.