No they didn't destroy all the extensions and many of the popular ones are long since back up and running. Noscript for example.
Yeah, after being reprogrammed. It's pretty obvious that the popular extensions are worth all that effort and even get special treatment by Mozilla, but that didn't do any good for the various other extensions I used to use where the developers said, "Sorry, it's too much work to redo everything."
Some kind of transition plan or emulation would have been nice. But, hey, if people insist that they did it for security reasons, just switching to an entirely new ecosystem is okay.
It's also worth nothing that much of their campaign revolved around "making the web faster" or something like that. They didn't even advertise it as a web browser.
When all this fancy Javascript nonsense really started picking up steam years ago, it was common to get only a blank page unless you used a really popular browser. Why? The only way to get fancy animated special effects was to render a blank page and then fade/animate stuff into view. Thus, you avoided the oh-so-dreaded Flash of Unstyled Content (v2.0). Of course, this meant that the 10% of the population not using one of the Big Three browsers would get no content at all. Hey, no big deal! It's not like people actually needed to access "Healthcare.org" or anything.
Yeah, I remember when IE ruled the roost and web developers were screaming their heads off about standards compliance, just because a table column was misaligned by a pixel or two (but the page as a whole still worked fine). Once Google took over and became the darling of designers, nobody gave a toss about standards, even if it meant web pages wouldn't display AT ALL if you dared to use something with less than double-digit market share. We still have frameworks that detect browsers by names like "Firefox" instead of "Gecko", and inject all kinds of stupid hacks even if the page doesn't use those features.
Hell, just a few days ago I looked at the source of the Crystal-Lang homepage, and they get their jQuery library directly from a 3rd-party site rather than hosting the file locally. The designers of a programming language have a web site that doesn't use even remotely sane coding techniques. Shocker.
If we want to keep Google from taking over the Internet, first tell the hyperactive UX people to stop using every shiny toy Chrome has to offer.
Their documentation is organized like a slideshow. Pass.
Aside from Hello World, their other simple example shows how to set up an HTTP server. Everything in the 7-line example is obvious and explained in detail, except for the weird part "do |context|". You have to sift through considerable amounts of documentation to figure out what the hell that does. Why gloss over the only interesting/confusing thing in the example? Pass.
Here we have some people developing a programming language, and their web page directly downloads jQuery from a 3rd-party site, rather than hosting the library locally. LOL... pass.
While we're at it, could we also have a mechanism to override auto-updating? It sucks when a developer sells his extension, and then everything auto-updates to the all-new system without appropriate disclosure. One of many reasons I don't want ANYTHING to auto-update anymore.
Pre-rendering a page when you move a mouse cursor over a tab isn't prediction, it's reaction.
Using so-called "idle" resources to render a page in the background takes time away from the browser UI. In this situation, the most likely outcome is that there will be a slight pause if you actually do click on the tab after mousing over it (and Firefox has had TONS of problems with random pauses over the last decade, especially with regards to garbage collection). I perfectly understand what you're saying, as I've studied CPU design, but despite all the hype and promises, these techniques never actually result in a net benefit in the real world. Hardware people get caches and prediction. From what I've seen, software people don't.
UIs were responsive in the days of Win2K and arguably XP, but that all went to hell when people started using all this turbo superprefetch nonsense, and used that as an excuse to justify tons of new bloat. Take a look at Win10. It does preventive maintenance all day long using "idle time" and it's slow as a dog and completely unpredictable, even if you have an SSD. The same has been happening to Linux for a while (at least with the distros I've tried).
When they actually make UIs more responsive, then I'll be impressed. Mozilla has been crooning about both speed and responsiveness for a decade... and I haven't seen it. Nobody else has, either, which is exactly why they've lost so much market share.
We already had tabbed browsing for the desktop. It was called the Taskbar. Then Microsoft tried to copy the OSX Dock principles of combining multiple functions under one button and they broke everything.
Maybe they should focus on making Explorer more like Total Commander or Directory Opus. I can't believe how difficult it still is to copy files back and forth, forcing you to shuffle windows around or have one full-screen window open at a time. Oh, and bring back Quick Launch while you're at it.
If you mouse-over a tab and it starts rendering something that's not even visible, that requires system resources and can slow down the whole system. That is NOT queuing things in a reasonable way, especially if garbage collection kicks in, yo.
I was rather amused to read the RISC-V processor whitepaper, and apparently they've designed the processor to (optionally) support 128-bit memory address space. They figure it'll be necessary for supercomputers to support spaces that big by 2030 or something.
Yeah, it's pretty clear that things like text messages, e-mail, and chat never worked properly before the multi-billion dollar international empires got into the business.
A lot of people should realize that the text messages themselves are not expensive to host. It doesn't costs a billion dollars a year to host Tweets. It's the bloated management, advertisement infrastructure, and AI data mining experts and analytic that cost all that money. Think about how a typical web page might be 10MB per page view. Is that all content, or is that flashy pop-up ads and data mining scripting?
Would you really be happier if it were just a single lump sum with no additional information?
That's pretty much how price quotes work. The itemized fees don't show up in the ads when you're shopping around.
We'll file your federal and state taxes for FREE*. (*If you qualify, but we won't tell you up-front if you do! Actual price disclosed after you give us your financial info.)
I've always been convinced the final straw will be the Deathswitch. Someone will figure out how to trigger a popular device's killswitch remotely, either via a hack or social engineering attack, and thousands, if not millions of devices will be bricked en masse overnight.
That MIGHT get peoples' attention to finally take security and privacy seriously. Maybe. For a little while.
Nothing in that article suggests "minimal" other than the usual marketing BS. In fact, it works exactly as I expected it to, which is far from minimal.
Exactly. These silly acceleration features are exactly what cause responsiveness issues, not solve them. Every good UI programmer is supposed to know this.
Yeah, but it was pretty standard treatment for everyone, not just me. The bottom performers get the axe, but everyone gets warnings about their performance just to kick us in the ass (hence, all the secrecy about how the numbers are calculated). In general, that's the way warehouse jobs are, especially when they fully convert to computerized voice picking. BTW, I didn't work for Amazon -- I worked for a huge medical distributor.
I was just smart enough to save/invest most of my money, so I was in a financial situation where I could afford to quit when things got really bad. For others desperate for a job, they just put up with the abuse as best they can.
My experience is that people never survive the "improvement plan". If they were too incompetent to do their job before, that is not going to change.
As someone who used to work 15 hours a day in a warehouse, I take great offense to calling people incompetent if they cannot meet quotas. Nobody can meet quota in places like that, and since quotas are "adjusted" by computers for a variety of hidden variables we were not allowed to know, it's impossible to know why you weren't meeting quota at all. Even the managers didn't know how our numbers were calculated.
Eventually they came around and told me I was below quota, and they wanted me to sign all kinds of papers enrolling me into some improvement plan. Instead, I gave them a three week notice. If I was incompetent, they sure didn't show it, since they tried really hard to keep me working there.
Unlike steel, aluminum cracks with repeated flexing over time, and is totally unsuitable for laptop cases. Plastic, contrary to popular belief, is way better (as long as you use a high quality plastic).
The problem is that if you close all but one tab/window and point the last window to "about:blank", memory doesn't come back. The browser isn't managing memory usage per page/tab/window, but collectively. Whether it's the Javascript engine or how the web pages use the heap is irrelevant. The bottom line is that the "leaks" stay with the browser until you restart the whole works. This has remained true even after the Quantum update that supposedly gives each tab/window its own process. I have no idea how the Javascript engine works, but apparently it doesn't work very well, and hasn't for over a decade.
I remember when Hyper Threading was first introduced, and I noted that a large number of my games (Viper Racing and Dungeon Keeper being notable examples), would run at half speed. I don't mean the framerates were halved, I mean the actual games themselves ran at half speed. It was like a built-in slo-mo feature.
Some applications, like Media Player also suffered timing issues. Disabling Hyper Threading always fixed everything. This went on for years. When the ability to disable HT in the BIOS was removed, I solved the problem by getting Media Player to play a blank media file in the background while playing my games. Somehow, this fixed the HT timing issues.
Needless to say, I never held HT with high regard. Oh, and AMD's SMT implementation didn't have this issue, which surprised me greatly when I switched to AMD down the line.
Slashdot Mirror
No they didn't destroy all the extensions and many of the popular ones are long since back up and running. Noscript for example.
Yeah, after being reprogrammed. It's pretty obvious that the popular extensions are worth all that effort and even get special treatment by Mozilla, but that didn't do any good for the various other extensions I used to use where the developers said, "Sorry, it's too much work to redo everything."
Some kind of transition plan or emulation would have been nice. But, hey, if people insist that they did it for security reasons, just switching to an entirely new ecosystem is okay.
It's also worth nothing that much of their campaign revolved around "making the web faster" or something like that. They didn't even advertise it as a web browser.
Web developers deserve a lot of the blame, here.
When all this fancy Javascript nonsense really started picking up steam years ago, it was common to get only a blank page unless you used a really popular browser. Why? The only way to get fancy animated special effects was to render a blank page and then fade/animate stuff into view. Thus, you avoided the oh-so-dreaded Flash of Unstyled Content (v2.0). Of course, this meant that the 10% of the population not using one of the Big Three browsers would get no content at all. Hey, no big deal! It's not like people actually needed to access "Healthcare.org" or anything.
Yeah, I remember when IE ruled the roost and web developers were screaming their heads off about standards compliance, just because a table column was misaligned by a pixel or two (but the page as a whole still worked fine). Once Google took over and became the darling of designers, nobody gave a toss about standards, even if it meant web pages wouldn't display AT ALL if you dared to use something with less than double-digit market share. We still have frameworks that detect browsers by names like "Firefox" instead of "Gecko", and inject all kinds of stupid hacks even if the page doesn't use those features.
Hell, just a few days ago I looked at the source of the Crystal-Lang homepage, and they get their jQuery library directly from a 3rd-party site rather than hosting the file locally. The designers of a programming language have a web site that doesn't use even remotely sane coding techniques. Shocker.
If we want to keep Google from taking over the Internet, first tell the hyperactive UX people to stop using every shiny toy Chrome has to offer.
Good. Saturn is pretty, but the red storm gives Jupiter that extra bit of intrigue!
Let me put it this way...
Their documentation is organized like a slideshow. Pass.
Aside from Hello World, their other simple example shows how to set up an HTTP server. Everything in the 7-line example is obvious and explained in detail, except for the weird part "do |context|". You have to sift through considerable amounts of documentation to figure out what the hell that does. Why gloss over the only interesting/confusing thing in the example? Pass.
Here we have some people developing a programming language, and their web page directly downloads jQuery from a 3rd-party site, rather than hosting the library locally. LOL... pass.
While we're at it, could we also have a mechanism to override auto-updating? It sucks when a developer sells his extension, and then everything auto-updates to the all-new system without appropriate disclosure. One of many reasons I don't want ANYTHING to auto-update anymore.
Pre-rendering a page when you move a mouse cursor over a tab isn't prediction, it's reaction.
Using so-called "idle" resources to render a page in the background takes time away from the browser UI. In this situation, the most likely outcome is that there will be a slight pause if you actually do click on the tab after mousing over it (and Firefox has had TONS of problems with random pauses over the last decade, especially with regards to garbage collection). I perfectly understand what you're saying, as I've studied CPU design, but despite all the hype and promises, these techniques never actually result in a net benefit in the real world. Hardware people get caches and prediction. From what I've seen, software people don't.
UIs were responsive in the days of Win2K and arguably XP, but that all went to hell when people started using all this turbo superprefetch nonsense, and used that as an excuse to justify tons of new bloat. Take a look at Win10. It does preventive maintenance all day long using "idle time" and it's slow as a dog and completely unpredictable, even if you have an SSD. The same has been happening to Linux for a while (at least with the distros I've tried).
When they actually make UIs more responsive, then I'll be impressed. Mozilla has been crooning about both speed and responsiveness for a decade... and I haven't seen it. Nobody else has, either, which is exactly why they've lost so much market share.
Even better, this outage highlights the very technical problem that routing was meant to prevent. It literally defeats the point of the Internet.
And the Lord said, go forth and consolidate!
We already had tabbed browsing for the desktop. It was called the Taskbar. Then Microsoft tried to copy the OSX Dock principles of combining multiple functions under one button and they broke everything.
Maybe they should focus on making Explorer more like Total Commander or Directory Opus. I can't believe how difficult it still is to copy files back and forth, forcing you to shuffle windows around or have one full-screen window open at a time. Oh, and bring back Quick Launch while you're at it.
If you mouse-over a tab and it starts rendering something that's not even visible, that requires system resources and can slow down the whole system. That is NOT queuing things in a reasonable way, especially if garbage collection kicks in, yo.
If I read the spec correctly, they were indeed talking about address space, not bus width or SIMD.
RISC-V is clever in some ways, but pretty bonkers in others.
I was rather amused to read the RISC-V processor whitepaper, and apparently they've designed the processor to (optionally) support 128-bit memory address space. They figure it'll be necessary for supercomputers to support spaces that big by 2030 or something.
Yeah, it's pretty clear that things like text messages, e-mail, and chat never worked properly before the multi-billion dollar international empires got into the business.
A lot of people should realize that the text messages themselves are not expensive to host. It doesn't costs a billion dollars a year to host Tweets. It's the bloated management, advertisement infrastructure, and AI data mining experts and analytic that cost all that money. Think about how a typical web page might be 10MB per page view. Is that all content, or is that flashy pop-up ads and data mining scripting?
Would you really be happier if it were just a single lump sum with no additional information?
That's pretty much how price quotes work. The itemized fees don't show up in the ads when you're shopping around.
We'll file your federal and state taxes for FREE*. (*If you qualify, but we won't tell you up-front if you do! Actual price disclosed after you give us your financial info.)
I've always been convinced the final straw will be the Deathswitch. Someone will figure out how to trigger a popular device's killswitch remotely, either via a hack or social engineering attack, and thousands, if not millions of devices will be bricked en masse overnight.
That MIGHT get peoples' attention to finally take security and privacy seriously. Maybe. For a little while.
Nothing in that article suggests "minimal" other than the usual marketing BS. In fact, it works exactly as I expected it to, which is far from minimal.
Exactly. These silly acceleration features are exactly what cause responsiveness issues, not solve them. Every good UI programmer is supposed to know this.
Yeah, but it was pretty standard treatment for everyone, not just me. The bottom performers get the axe, but everyone gets warnings about their performance just to kick us in the ass (hence, all the secrecy about how the numbers are calculated). In general, that's the way warehouse jobs are, especially when they fully convert to computerized voice picking. BTW, I didn't work for Amazon -- I worked for a huge medical distributor.
I was just smart enough to save/invest most of my money, so I was in a financial situation where I could afford to quit when things got really bad. For others desperate for a job, they just put up with the abuse as best they can.
My experience is that people never survive the "improvement plan". If they were too incompetent to do their job before, that is not going to change.
As someone who used to work 15 hours a day in a warehouse, I take great offense to calling people incompetent if they cannot meet quotas. Nobody can meet quota in places like that, and since quotas are "adjusted" by computers for a variety of hidden variables we were not allowed to know, it's impossible to know why you weren't meeting quota at all. Even the managers didn't know how our numbers were calculated.
Eventually they came around and told me I was below quota, and they wanted me to sign all kinds of papers enrolling me into some improvement plan. Instead, I gave them a three week notice. If I was incompetent, they sure didn't show it, since they tried really hard to keep me working there.
Business ethics in a nutshell.
Seriously. It's all about pissing off the fewest people.
Unlike steel, aluminum cracks with repeated flexing over time, and is totally unsuitable for laptop cases. Plastic, contrary to popular belief, is way better (as long as you use a high quality plastic).
The problem is that if you close all but one tab/window and point the last window to "about:blank", memory doesn't come back. The browser isn't managing memory usage per page/tab/window, but collectively. Whether it's the Javascript engine or how the web pages use the heap is irrelevant. The bottom line is that the "leaks" stay with the browser until you restart the whole works. This has remained true even after the Quantum update that supposedly gives each tab/window its own process. I have no idea how the Javascript engine works, but apparently it doesn't work very well, and hasn't for over a decade.
Yes, it's a significant problem.
I remember when Hyper Threading was first introduced, and I noted that a large number of my games (Viper Racing and Dungeon Keeper being notable examples), would run at half speed. I don't mean the framerates were halved, I mean the actual games themselves ran at half speed. It was like a built-in slo-mo feature.
Some applications, like Media Player also suffered timing issues. Disabling Hyper Threading always fixed everything. This went on for years. When the ability to disable HT in the BIOS was removed, I solved the problem by getting Media Player to play a blank media file in the background while playing my games. Somehow, this fixed the HT timing issues.
Needless to say, I never held HT with high regard. Oh, and AMD's SMT implementation didn't have this issue, which surprised me greatly when I switched to AMD down the line.
True, but GOG is mostly a seller of Steam keys these days (you know, much like today's retail boxes).
I remember when they sold only downloads from their own site.
And yet, developers were still able to deliver games that were fun to play before all this analytics nonsense.
Games weren't necessarily more fun back in the day, but I certainly appreciated and enjoyed them (and the Internet in general) a lot more.