Hugo Award winning novel Dune has magic - the spice allows the user to see the future. Hugo Award winning novel The Left Hand of Magic has... magic. Then there's winner Harry Potter and the Goblet of Fire in 2001.
I hate Trump, but you're not helping. We just found out Bassel was executed in 2015. So if you're going to blame a President, pick Obama.
But either way it's absurd. It is the Syrian government killing a Syrian citizen - terrible, but the US government can't exactly force every other government in the world to choose its executions and pardons based on our decisions.
There's a lot of boilerplate, trivial stuff in there or commits in which two lines in one file are changed and then all of the minified Javascript is regenerated. But that's true for me and all of the people I work with, too. There are still dozens of commits with serious, non-trivial code changes. The accusation that this guy was not an open source software developer is incorrect.
He worked on the Aiki Framework. Check https://github.com/aikiframewo... - he's got over 500 commits in the history changing actual code. He was a real software developer.
That's fine for you, and fine for me. But if we want Firefox to be used by more than some percentage of the browser market greater than 1%, that's not sufficient. It's best for all of us if free software browsers are the best option even for people running all of the web's bloatware. Better for freedom, for privacy, for security.
Google stands for advertising, and the power of their advertising is enhanced by data collection and proprietary software. Apple and Microsoft stand for proprietary software. br.
As long as Firefox is fully open source, even if Mozilla loses its way and has some wacky projects, it's a thousand miles ahead of the competing browser vendors.
Again, even with Firefox 54, the current stable release, it's still very noticeably slower than Chrome. But if you download the nightly build of Firefox I think the speed improvement is fantastic.
I've been sticking with Firefox over Chrome or Chromium because I like what Mozilla stands for.
But I have to use Chrome for work, and no matter how well Firefox did in some Tom's Hardware browser shootouts in 2012 or so, and no matter what numbers they show on arewefastyet.com and so forth, Firefox consistently felt painfully slow next to Chrome. That finally changed for me with 56 nightly. I'm not sure if it's as fast as Chrome, but for the first time ever it feels close enough that the difference is not an annoyance.
The article also mentions Firefox OS. I think in 2025, when WebAssembly is a mature technology, when $25 smart phones have 4GB of RAM, and when Firefox on mobile is substantially more efficient than Firefox 56 or 57 now... then Firefox OS might be practical. In 2013, it was a great idea not ready.
I run Ubuntu Mate, Xubuntu, and Elementary OS (an Ubuntu derivative with its own desktop environment) these days. I had made the switch to Linux several times in the past and then switched back to Windows due to games. When I hit my mid 30s I switched to Linux and I haven't looked back. To be fair, it's much easier for a novice to successfully install and configure Linux in the past decade than it was in, say, 2000.
I think Canonical's mobile operating system attempts were poorly planned. It's easy for anyone to say that with the benefit of hindsight, of course. But consider that the first arguably decent, feature-complete, aesthetically adequate, somewhat secure version of Android was in the 4.x release series. So Google, a company with literally more than a hundred times the financial resources of Canonical, took almost four years to make a competitive mobile operating system and application API.
It's clear - now - that Ubuntu Touch should have been a complete Android fork, or based on Meego or WebOS. Then Canonical would have had to put polish on something with free software that had most of the groundwork laid already. Instead... here we are. Canonical used the Android kernel for device driver support, but a completely different front end layer. (While I'm ranting, I'll mention that I also loved the concept of Firefox OS. But Firefox 56 alpha on desktop is finally showing me the competitive performance I wanted to see in that browser five years ago. It's clear the Firefox developers are now focused heavily on profiling for performance bottlenecks and parallelizing everything they can. Even then, if I leave it open for a few days with a few dozen tabs memory usage creeps up. Making Firefox OS in 2025, when Firefox 75 is as fast or faster than any other browser across the board and $25 mobile phones have 4GB of RAM makes perfect sense. Making Firefox OS in 2014 with 256MB of RAM and Firefox 36 was even more foolishly ambitious than the Ubuntu Touch project.)
My only ISP option is Comcast. If I can survive with 25 Mbps connection speeds or less, the overhead of a business account vs. residential is tiny. But above that level, the price difference grows rapidly. It's common for me to have two people in the house playing Minecraft while three others stream video, so I don't think 25 Mbps will work.
We have a Tivo Bolt, and the performance of Amazon Prime and Netflix is fine on that. I don't know if I'm more tolerant of slowness than you, or if the Bolt has better performance.
You and I and some others decided free software still matters even after we could afford the proprietary options. That's fantastic. Unfortunately there aren't enough of us.
I got started with Linux in the late 1990s and would set up dual boot environments but spend most of my time in Windows. I didn't switch to spending more time in Linux until about ten years ago. I'm forty, and also fortunate enough to make more than 80k.
I spend most of my time on Linux too, and I do care about free software and freedom. But for example I have a friend that used to run the Linux User Group in a small size city in his late 20s. Now he's in his 40s, and he no longer visits - let alone organizes - any LUGs and he has a Macbook Pro and a Google Nexus 6P phone. He's the picture perfect example of this trend, and he's not the only friend I have that made a similar transition as he got older.
We seem to be the exception - and the situation is getting worse now that iPhone and even Android devices keep making it harder to install free operating systems.
However, the sad truth is that it looks like there is no business case anymore for a truly open platform based on custom-designed hardware, since people refuse to spend extra money for tweakability, freedom, and security. Despite us living in times where privacy is massively endangered.
The problem is that people with money to burn are less likely to focus on tweakability and freedom. At 15, 20, or 25 you have time but not money. At 30, 35, 40 if you're lucky you are likely to have money but not time. I know a number of guys that were Linux and free software supporters in school, but once they reached 80k or better income they just switched to buying the hottest proprietary option and went on about their day - typically a Macbook Pro and an iPhone, or Samsung Galaxy something, or Google Nexus or Pixel device.
I cared about OpenMoko in 2006, but I didn't have the money. Today I'm contemplating purchases of more devices with the Free Software Foundation "Respects Your Freedom" certification. But it's tough to get excited about spending more money for much slower hardware. And the Replicant.us completely free Android version? I love what they're trying to do, but because of the (*$&%()*%&$ proprietary firmware all the devices need it renders the devices they support all but useless.
The laptop is a toy machine I only move around the house to play music when I'm working in odd places or play movies when the kids managed - again - to drain the batteries of every other portable electronic simultaneously.
I am grateful for the offer of assistance, though. I actually have been using btrfs everywhere with/home and/opt each on separate subvolumes. I know btrfs is not as stable and mature as ZFS, but so far it hasn't burned me and I have two in-house backups of everything and one offsite backup. I'm really hoping it matures into a full ZFS competitor, not because I have anything against ZFS but because I think something GPLv2 right in the kernel makes life easier for the Linux community at large than the legal shenanigans and bickering around CDDL or whatever the ZFS license is.
That said... right now I'm managing partitions across disks manually and not using the redundancy features available in btrfs (or ZFS) volumes. I'm synchronizing everything with rsync in cron as a kind of manual equivalent to RAID 1. But that gives me two advantages. The first is that I can move a physical drive between computers and only lose the time it takes to do the physical move, there is no data loss or volume rebuild required on either side. Well, that's not totally true because I have to modify some/etc/fstab entries and rsync settings in cron on both ends. But the data is available immediately. The second advantage is that I can use ntfs-formatted drives for backups when booted into Linux. I rarely boot into Windows, but every once in a while my kids want to play a Steam game that isn't available on Linux. I lose access to my backup file server when they do, but only for a few days until they get bored and then it's back in business.
So in theory, Ethereum and a few other cryptocurrencies like Safecoin and Storj.io offer a backing value in computing resources you can buy on the mining network itself. I'm excited by this possibility, and will watch it with interest.
But today, I'd say the value of any such currency is less than 1% based on those computing resources you can buy and more than 99% another form of gambling (or if you prefer, Ponzi schemes). I haven't see any argument that the current price per computing resource unit is cost-effective vs. renting a box on DigitalOcean or using some PaaS.
I'm at least five years away from investing in any of this. And that's a best case scenario. The technology may never work the way I hope.
Thanks for taking the time to reply. Nice article - good writing, useful information.
I have an old laptop with 2 GB of RAM and a Core Duo (or Core 2 Duo, I don't remember) and a whopping 1280x800 resolution. I've got Xubuntu 16.04 on it, and it runs nicely. But it mostly collects dust because I've already got a file server and the eight or so year old battery holds a charge for between 90 and 150 seconds.:)
and don't insist on TV connections to get deals that are better than standalone internet plans.
As far as I can tell, that's a fallacy. I am only buying internet from Comcast, and they'll sell me cable for just $5 more but not less. Except of course, between added local channel fees, sports, fees, equipment rental fees, etc... it's actually $45 more. So fuck 'em.
Both parties love to promise the world when they're in the minority, and then nothing gets done when they have the majority. The Republicans spent six years promising they had way better ideas than Obamacare. The American voters gave them control of Congress and the White House, now it's time to show their cards... and it looks like they have nothing.
But don't be too hard on the Republicans. The Democrats are now playing the same game. They're going to fix broadband, and education, and Obamacare, and probably give us all unicorns and cotton candy that makes you lose weight too.
So I have a much weaker understanding of this kind of thing than I would like. My understanding is that the total bandwidth of any set of USB ports connected to a single USB "card" on the motherboard is actually the max for one port. i.e. if a particular version of USB has bandwidth X, then you can't have port 1 with throughput X as the same time port 2 has throughput X. If they both use the same "card" on the motherboard, then the total throughput between them combined is X. Then on top of that, I took the 480 Mbps of USB 2.0, and assumed that on my hypothetical laptop each of my pair of USB gigabit ethernet devices could then handle 240 Mbps. That exceeds the 200 Mbps I get from my ISP, problem solved.
But you came up with 15 MB/s. How? I see on the Wikipedia page for USB 2: "Due to bus access constraints, the effective throughput of the High Speed signaling rate is limited to 280 Mbit/s or 35 MB/s". Are you just dividing that in half and subtracting about 15% for overhead? Is that how you got your numbers?
I'm not saying you are wrong. I know that on-paper figures and real world numbers can vary wildly. I have a long experience wrestling with wireless throughput as experience. I'm just trying to understand the factors better.
This was covered up-thread, but the Raspberry Pi has a 100 Mpbs ethernet port and also shares bandwidth between ethernet and USB 2.0, so that if you're doing something data-intensive over USB 2.0 like reading a big file it will actually slow your network throughput.
So that's an acceptable solution for someone that isn't using much data, but won't work if you and your friends are watching a few different shows on Youtube or Netflix at the same time.
Thanks for the Turris Omnia suggestion. That's awesome.
Though I suspect their 802.11ac wireless drivers are proprietary. I haven't done research recently, but when I checked last year there were no fully open source 802.11ac device drivers.
I have four kids and my wife and I are cable-cutters, so we're routinely drawing a lot of bandwidth on our 200 Mbps internet connection. My home router is running DD-WRT, because it's three years old. But if I didn't have that, I think I might be able to get by with an old laptop as long as it had two USB 2.0 ports. I could plug USB gigabit ethernet devices into those. The bandwidth would be capped at what USB 2.0 can handle, but that should be well above 200 Mbps.
To be pedantic about "that didn't concern itself with efficiency": they were extremely concerned with efficiency. Google, Microsoft, IBM, Oracle, HP, Amazon, Yahoo, and others in 2005 (for example) cared very much about every dollar spent on power and every dollar spent on cooling. They bought the most efficient server hardware they could afford. That hardware was just twelve years behind the levels of efficiency we have today.
This whole problem is very much one of those places where the Free Software Foundation is right - we're being locked out of legitimate uses of the products we buy, and we're being hurt by it. Not only is it annoying and insecure, but it's also environmentally wasteful. We should be able to run secure, audited-in-the-open code on the latest efficient wireless hardware. Just like we should be able to make good use of a laptop that's seven years old and not have an environmentally destructive planned obsolescence cycle for all mainstream smart phones.
Good points, but I would counter with a few things:
1. This is not the first or only work the Firefox team has been doing to improve performance. You're writing as though this Quantum Flow project is the first performance work they've done ever, or within the past five years. It isn't. The switch to multi-threading has been underway for years.
2. Firefox's problems today are largely a result of its own success ten years ago. The biggest cause of performance slowdowns is add-ons that have inefficient code, or add-ons that use inefficient code in Firefox. The Firefox team is trying to address this - and every change they make that breaks add-on compatibility brings howls of protest from the community. I don't blame the people who are upset by the changes, either. But the result is that the very things that made Firefox a big success then is slowing it down now.
3. Conversely, if you want fastest performance try running without add-ons. That's how Firefox has been doing well in benchmarks, like the Tom's Hardware Browser Shoutouts it won.
4. I don't think the Firefox developers have been the ones telling the regular users that Firefox is fine. I think it's other Firefox fans. I've been defending Firefox for years - but I'm not a Firefox developer or contributor. And I haven't been claiming it's as fast as Chrome, just that it's fast enough.
Slashdot Mirror
Hugo Award winning novel Dune has magic - the spice allows the user to see the future. Hugo Award winning novel The Left Hand of Magic has... magic. Then there's winner Harry Potter and the Goblet of Fire in 2001.
Are you serious?
I hate Trump, but you're not helping. We just found out Bassel was executed in 2015. So if you're going to blame a President, pick Obama.
But either way it's absurd. It is the Syrian government killing a Syrian citizen - terrible, but the US government can't exactly force every other government in the world to choose its executions and pardons based on our decisions.
git clone git@github.com:aikiframework/aikiframework.git ; cd aikiframework ; git log -p --author=Bassel
There's a lot of boilerplate, trivial stuff in there or commits in which two lines in one file are changed and then all of the minified Javascript is regenerated. But that's true for me and all of the people I work with, too. There are still dozens of commits with serious, non-trivial code changes. The accusation that this guy was not an open source software developer is incorrect.
He worked on the Aiki Framework. Check https://github.com/aikiframewo... - he's got over 500 commits in the history changing actual code. He was a real software developer.
I don't know what you're mouthing off about.
That's fine for you, and fine for me. But if we want Firefox to be used by more than some percentage of the browser market greater than 1%, that's not sufficient. It's best for all of us if free software browsers are the best option even for people running all of the web's bloatware. Better for freedom, for privacy, for security.
Google stands for advertising, and the power of their advertising is enhanced by data collection and proprietary software. Apple and Microsoft stand for proprietary software.
br. As long as Firefox is fully open source, even if Mozilla loses its way and has some wacky projects, it's a thousand miles ahead of the competing browser vendors.
Again, even with Firefox 54, the current stable release, it's still very noticeably slower than Chrome. But if you download the nightly build of Firefox I think the speed improvement is fantastic.
I've been sticking with Firefox over Chrome or Chromium because I like what Mozilla stands for.
But I have to use Chrome for work, and no matter how well Firefox did in some Tom's Hardware browser shootouts in 2012 or so, and no matter what numbers they show on arewefastyet.com and so forth, Firefox consistently felt painfully slow next to Chrome. That finally changed for me with 56 nightly. I'm not sure if it's as fast as Chrome, but for the first time ever it feels close enough that the difference is not an annoyance.
The article also mentions Firefox OS. I think in 2025, when WebAssembly is a mature technology, when $25 smart phones have 4GB of RAM, and when Firefox on mobile is substantially more efficient than Firefox 56 or 57 now... then Firefox OS might be practical. In 2013, it was a great idea not ready.
I run Ubuntu Mate, Xubuntu, and Elementary OS (an Ubuntu derivative with its own desktop environment) these days. I had made the switch to Linux several times in the past and then switched back to Windows due to games. When I hit my mid 30s I switched to Linux and I haven't looked back. To be fair, it's much easier for a novice to successfully install and configure Linux in the past decade than it was in, say, 2000.
I think Canonical's mobile operating system attempts were poorly planned. It's easy for anyone to say that with the benefit of hindsight, of course. But consider that the first arguably decent, feature-complete, aesthetically adequate, somewhat secure version of Android was in the 4.x release series. So Google, a company with literally more than a hundred times the financial resources of Canonical, took almost four years to make a competitive mobile operating system and application API.
It's clear - now - that Ubuntu Touch should have been a complete Android fork, or based on Meego or WebOS. Then Canonical would have had to put polish on something with free software that had most of the groundwork laid already. Instead... here we are. Canonical used the Android kernel for device driver support, but a completely different front end layer. (While I'm ranting, I'll mention that I also loved the concept of Firefox OS. But Firefox 56 alpha on desktop is finally showing me the competitive performance I wanted to see in that browser five years ago. It's clear the Firefox developers are now focused heavily on profiling for performance bottlenecks and parallelizing everything they can. Even then, if I leave it open for a few days with a few dozen tabs memory usage creeps up. Making Firefox OS in 2025, when Firefox 75 is as fast or faster than any other browser across the board and $25 mobile phones have 4GB of RAM makes perfect sense. Making Firefox OS in 2014 with 256MB of RAM and Firefox 36 was even more foolishly ambitious than the Ubuntu Touch project.)
Seconded (sorry, I posted above or I would have used mod points).
My only ISP option is Comcast. If I can survive with 25 Mbps connection speeds or less, the overhead of a business account vs. residential is tiny. But above that level, the price difference grows rapidly. It's common for me to have two people in the house playing Minecraft while three others stream video, so I don't think 25 Mbps will work.
We have a Tivo Bolt, and the performance of Amazon Prime and Netflix is fine on that. I don't know if I'm more tolerant of slowness than you, or if the Bolt has better performance.
You and I and some others decided free software still matters even after we could afford the proprietary options. That's fantastic. Unfortunately there aren't enough of us.
I got started with Linux in the late 1990s and would set up dual boot environments but spend most of my time in Windows. I didn't switch to spending more time in Linux until about ten years ago. I'm forty, and also fortunate enough to make more than 80k.
I spend most of my time on Linux too, and I do care about free software and freedom. But for example I have a friend that used to run the Linux User Group in a small size city in his late 20s. Now he's in his 40s, and he no longer visits - let alone organizes - any LUGs and he has a Macbook Pro and a Google Nexus 6P phone. He's the picture perfect example of this trend, and he's not the only friend I have that made a similar transition as he got older.
We seem to be the exception - and the situation is getting worse now that iPhone and even Android devices keep making it harder to install free operating systems.
However, the sad truth is that it looks like there is no business case anymore for a truly open platform based on custom-designed hardware, since people refuse to spend extra money for tweakability, freedom, and security. Despite us living in times where privacy is massively endangered.
The problem is that people with money to burn are less likely to focus on tweakability and freedom. At 15, 20, or 25 you have time but not money. At 30, 35, 40 if you're lucky you are likely to have money but not time. I know a number of guys that were Linux and free software supporters in school, but once they reached 80k or better income they just switched to buying the hottest proprietary option and went on about their day - typically a Macbook Pro and an iPhone, or Samsung Galaxy something, or Google Nexus or Pixel device.
I cared about OpenMoko in 2006, but I didn't have the money. Today I'm contemplating purchases of more devices with the Free Software Foundation "Respects Your Freedom" certification. But it's tough to get excited about spending more money for much slower hardware. And the Replicant.us completely free Android version? I love what they're trying to do, but because of the (*$&%()*%&$ proprietary firmware all the devices need it renders the devices they support all but useless.
The laptop is a toy machine I only move around the house to play music when I'm working in odd places or play movies when the kids managed - again - to drain the batteries of every other portable electronic simultaneously.
/home and /opt each on separate subvolumes. I know btrfs is not as stable and mature as ZFS, but so far it hasn't burned me and I have two in-house backups of everything and one offsite backup. I'm really hoping it matures into a full ZFS competitor, not because I have anything against ZFS but because I think something GPLv2 right in the kernel makes life easier for the Linux community at large than the legal shenanigans and bickering around CDDL or whatever the ZFS license is.
/etc/fstab entries and rsync settings in cron on both ends. But the data is available immediately. The second advantage is that I can use ntfs-formatted drives for backups when booted into Linux. I rarely boot into Windows, but every once in a while my kids want to play a Steam game that isn't available on Linux. I lose access to my backup file server when they do, but only for a few days until they get bored and then it's back in business.
I am grateful for the offer of assistance, though. I actually have been using btrfs everywhere with
That said... right now I'm managing partitions across disks manually and not using the redundancy features available in btrfs (or ZFS) volumes. I'm synchronizing everything with rsync in cron as a kind of manual equivalent to RAID 1. But that gives me two advantages. The first is that I can move a physical drive between computers and only lose the time it takes to do the physical move, there is no data loss or volume rebuild required on either side. Well, that's not totally true because I have to modify some
So in theory, Ethereum and a few other cryptocurrencies like Safecoin and Storj.io offer a backing value in computing resources you can buy on the mining network itself. I'm excited by this possibility, and will watch it with interest.
But today, I'd say the value of any such currency is less than 1% based on those computing resources you can buy and more than 99% another form of gambling (or if you prefer, Ponzi schemes). I haven't see any argument that the current price per computing resource unit is cost-effective vs. renting a box on DigitalOcean or using some PaaS.
I'm at least five years away from investing in any of this. And that's a best case scenario. The technology may never work the way I hope.
Thanks for taking the time to reply. Nice article - good writing, useful information.
:)
I have an old laptop with 2 GB of RAM and a Core Duo (or Core 2 Duo, I don't remember) and a whopping 1280x800 resolution. I've got Xubuntu 16.04 on it, and it runs nicely. But it mostly collects dust because I've already got a file server and the eight or so year old battery holds a charge for between 90 and 150 seconds.
and don't insist on TV connections to get deals that are better than standalone internet plans.
As far as I can tell, that's a fallacy. I am only buying internet from Comcast, and they'll sell me cable for just $5 more but not less. Except of course, between added local channel fees, sports, fees, equipment rental fees, etc... it's actually $45 more. So fuck 'em.
Both parties love to promise the world when they're in the minority, and then nothing gets done when they have the majority. The Republicans spent six years promising they had way better ideas than Obamacare. The American voters gave them control of Congress and the White House, now it's time to show their cards... and it looks like they have nothing.
But don't be too hard on the Republicans. The Democrats are now playing the same game. They're going to fix broadband, and education, and Obamacare, and probably give us all unicorns and cotton candy that makes you lose weight too.
Just until they get into power.
So I have a much weaker understanding of this kind of thing than I would like. My understanding is that the total bandwidth of any set of USB ports connected to a single USB "card" on the motherboard is actually the max for one port. i.e. if a particular version of USB has bandwidth X, then you can't have port 1 with throughput X as the same time port 2 has throughput X. If they both use the same "card" on the motherboard, then the total throughput between them combined is X. Then on top of that, I took the 480 Mbps of USB 2.0, and assumed that on my hypothetical laptop each of my pair of USB gigabit ethernet devices could then handle 240 Mbps. That exceeds the 200 Mbps I get from my ISP, problem solved.
But you came up with 15 MB/s. How? I see on the Wikipedia page for USB 2: "Due to bus access constraints, the effective throughput of the High Speed signaling rate is limited to 280 Mbit/s or 35 MB/s". Are you just dividing that in half and subtracting about 15% for overhead? Is that how you got your numbers?
I'm not saying you are wrong. I know that on-paper figures and real world numbers can vary wildly. I have a long experience wrestling with wireless throughput as experience. I'm just trying to understand the factors better.
This was covered up-thread, but the Raspberry Pi has a 100 Mpbs ethernet port and also shares bandwidth between ethernet and USB 2.0, so that if you're doing something data-intensive over USB 2.0 like reading a big file it will actually slow your network throughput.
So that's an acceptable solution for someone that isn't using much data, but won't work if you and your friends are watching a few different shows on Youtube or Netflix at the same time.
Thanks for the Turris Omnia suggestion. That's awesome.
Though I suspect their 802.11ac wireless drivers are proprietary. I haven't done research recently, but when I checked last year there were no fully open source 802.11ac device drivers.
I have four kids and my wife and I are cable-cutters, so we're routinely drawing a lot of bandwidth on our 200 Mbps internet connection. My home router is running DD-WRT, because it's three years old. But if I didn't have that, I think I might be able to get by with an old laptop as long as it had two USB 2.0 ports. I could plug USB gigabit ethernet devices into those. The bandwidth would be capped at what USB 2.0 can handle, but that should be well above 200 Mbps.
To be pedantic about "that didn't concern itself with efficiency": they were extremely concerned with efficiency. Google, Microsoft, IBM, Oracle, HP, Amazon, Yahoo, and others in 2005 (for example) cared very much about every dollar spent on power and every dollar spent on cooling. They bought the most efficient server hardware they could afford. That hardware was just twelve years behind the levels of efficiency we have today.
This whole problem is very much one of those places where the Free Software Foundation is right - we're being locked out of legitimate uses of the products we buy, and we're being hurt by it. Not only is it annoying and insecure, but it's also environmentally wasteful. We should be able to run secure, audited-in-the-open code on the latest efficient wireless hardware. Just like we should be able to make good use of a laptop that's seven years old and not have an environmentally destructive planned obsolescence cycle for all mainstream smart phones.
Good points, but I would counter with a few things:
1. This is not the first or only work the Firefox team has been doing to improve performance. You're writing as though this Quantum Flow project is the first performance work they've done ever, or within the past five years. It isn't. The switch to multi-threading has been underway for years.
2. Firefox's problems today are largely a result of its own success ten years ago. The biggest cause of performance slowdowns is add-ons that have inefficient code, or add-ons that use inefficient code in Firefox. The Firefox team is trying to address this - and every change they make that breaks add-on compatibility brings howls of protest from the community. I don't blame the people who are upset by the changes, either. But the result is that the very things that made Firefox a big success then is slowing it down now.
3. Conversely, if you want fastest performance try running without add-ons. That's how Firefox has been doing well in benchmarks, like the Tom's Hardware Browser Shoutouts it won.
4. I don't think the Firefox developers have been the ones telling the regular users that Firefox is fine. I think it's other Firefox fans. I've been defending Firefox for years - but I'm not a Firefox developer or contributor. And I haven't been claiming it's as fast as Chrome, just that it's fast enough.