I don't use QoS, I use an FairQ + AQM. Dropped packets is not just normal for TCP, it's a requirement. Dropping a packet early before bufferbloat kicks in reduces latency is reduces packetloss by signalling to the sender to backoff before the buffer fills and bloat plus burst loss occurs.
That's only true if the sender is DDOSing you or you literally don't have enough bandwidth. I QoS my incoming just fine by dropping packets which signals the sender to back off. And it's not just any packets getting dropped, it's statically biased to almost always be packets from the fattest flows. I can have my connection loaded to 99% of its provisioned rate and get 0 dropped gaming packets and less than 1ms of jitter.
They're conflating "talent" and "skilled labor". You don't train talent and you don't teach talent, you find talent.
I'm of two opinions about communication.
The practical side of me says communications is important. The ideal side of me says if the audience can't understand it, either you don't understand it or the audience is not smart enough in whatever domain that it matters. That said, I have issue with many in my profession using technical jargon because they use it incorrectly. They use it more like an ambiguous buzzword than a technical word with specific meaning. But again, this really just means the person talking doesn't understand.
In my experience, communication is more like a soft skill of using simple enough words that the receiver thinks they understand just enough. Just a "feel good" kind of skill. It's like describing a cube to someone who doesn't understand 3D shapes. In the end, you just tell them it a bunch of connected lines and they feel better about themselves because they know what a line is.
Working more than 40 hours or so makes you much more prone to mistakes. It is literally unprofessional to allow yourself to get into such a state. Learning when to say "no" is required to be a professional. Time management is critical to businesses and you can't do planning if you don't have good estimations. Back-pressure from people saying "no" is critical to the proper operations of a business.
Forced freedom is not freedom and the logical dissonance required to willingly conflate entitled access to other people's work and freedom is blind zealotry. With BSD, you always have access to the original, just no guarantee about access to derivatives. Which is "better" is subjective, but BSD is objectively more free.
The licenses affect different people differently, but BSD is still more free. Free isn't about what you can do, it's about what someone says you can't or have to do. The more someone controls you, the less free you are. The main argument isn't about which is more free, but which ideology is "better" as a whole. BSD is objectively more free than GPL, but GPL is subjectively better than BSD by enforcing certain subsets of freedom at the expense of others deemed less important.
Spinning disk has about as many sudden deaths are SSDs, the only different is spinners have an additional set of failures that give warning. In other words. If you had a harddrive that never died from mechanical issues, its failure rate would be very similar to an SSD.
Wear leveling is much less effective without TRIM. TRIM reduces write amplification by letting the wear leveling algorithm know when some data is no longer referenced.
I find it difficult to tell how good or bad someone else is in my profession without hanging out with them and talking on related subjects over several days. Not very practical for interviews. I cannot fault someone for not being good at picking out another skilled workers.
Making sure equipment and what-not is kept up-to-date is part of the responsibility of the department. If they deem that they want a new scoreboard in order to do what they need to do, then that's their decision. Their decisions will only be questioned if they fail to deliver positive results. And depends on what you mean by "publicly funded". If the money is from private sources, like student tuition and private donations, then it's not public money. Funding is not in one large pool. All of the money is earmarked and tracked.
Most people's problem solving ability is just trying the same tactic over and over again until it just happens to succeed in one instance. Just as important is to know if a problem has actually been solved, and not just swept under the carpet.
The whole "try random things until the symptom goes away" strategy that seems so prevalent. I r hard werking pragammer that putz n ovrtyme! I swear, semantics just whooshes over 99% of people's heads. These two variables have the same name, they must mean the same thing!
You have no idea how much I have to deal with this. My team leader, half jokingly, told me to stop reading other people's code because I keep finding problems and we have too many deliverables to have to deal with all of the bugs I find. A recent bug was when I was adding some new features to a project. I noticed that the database relationships seemed "strange". After a few hours of digging, I found a bug that has been resulting in corrupted data for nearly 20 years. The code in question had been worked on many tens of times over these 20 years, including several refactors and re-writes. Yet somehow every single programmer who ever touched it never noticed this exact same bug and every singe time recreated the same bug.
I didn't even see a bug. I just had a gut feeling that the inferred abstract semantics of the data was not intuitively clear to me, and I decided to go look at the code to validate. Any of the dozens of programmers who wrote and or reviewed the code could have done the same, yet no one did. This kind of situation happens to me all of them. Like several times a week. My team leader says I have super human attention to detail, but I really need to stop "breaking" "stable" code. My problem is I think abstractly about everything. I need to fully understand the semantics of everything, otherwise my mind just has grid lock with all of the possibility. I have a high attention to detail because I'm forced to. I have ADD, and the only way to keep my concentration on track is to have absolute confidence in my understanding of the code, otherwise my mind wanders, trying to find all edge cases.
I already have a connection fee. It pretty much has never changed for my entire life. It has technically changed, but at less rate than inflation. The grid is maintained by a company independent of who generates the power and gets paid entirely by the connection fee. Some energy companies even have negative margins for power sold. They're encouraged to reduce grid consumption.
Our energy companies make most of their profit from connection fees, energy consulting and energy services, but not almost nothing from energy sold. They're loving house level renewables because of the up-tick in demand for consulting and services.
I think battery banks for load shifting will probably be a better investment for areas where it doesn't get crazy hot. I live in the Midwest with $0.10/kwh flat rate or $0.05/offpeak and $0.15/onpeak. And I calculated that it would take about 8 years to pay off a $5,000 simple "just plug it into a general power outlet" programmable load shifting battery pack that even doubles as a UPS for all devices on the same breaker.
I only recently learned about these and they're pretty sweet. They come out of the box with a preconfigured charge timeframe, idle timeframe, and discharge timeframe, along with a rate of charge and discharge. You can configure these values yourself. It also has an option to monitor some energy site that has "current rates" for energy where it applies. Then the device and charge and discharge based on dynamic pricing. You can set high and low dynamic charging levels. It also works as a UPS by triggering your breaker when an outage is detected, then continues to supply power to the circuit. From what I've read, you don't need any professional installation at all, just plug it into any outlet in your house. You can have more than one in your house. Not sure if on the same circuit. I assume you can because it's meant to be brain-dead easy to use safely without any knowledge of your house.
What has a 10 year life? Not solar panels, they're much much higher. A recent study on individual solar cells, not entire panels, is that around 90% of used solar cells from 20+ years ago are still above 90% of their original fresh from the factory rating, and a substantial portion are above 95% of their original rating. What does drop over time is the entire panel for many panels, but not all. All it takes is one cell in the entire panel to make the output low and one low panel in a group to make the entire group low. But it's easy to swap out panels, less so individual cells. Even at the panel level, the 80/20 rule still applies. About 80% of panels are perfectly fine even after 20 years. The trend towards smaller panels with many micro-inverts is making this issue less of an issue.
The problem I've seen others with community college is almost none of any credits for their major transferred and only GDRs did. They got stuck doing only GDRs for the first 1.5 years at their CC, but then were forced to try packing 4 years of their major into 2-3 years assuming they wanted to finish in 4, which may not be possible depending on prerequisite classes.
It's actually better than that for value. The $40mil saved was based on what the system recognized. The system can only track power in much coarser grained chunks, like minutes. The batter bank is providing power down into the millisecond ranges. If the grid could track at a finer resolution, it's already paid itself off.
And the bank is vastly overbuilt. It was designed to maintain its rated capacity for 10 years, then they added another 10% raw capacity for good measure. And the 10 years is based on the assumption of 100% cycle every day for 10 years. It's being cycled much less than that. The batteries should be good for several decades assuming no other major deterioration issues.
You can do that, but that process is going to be like 10% efficient. Even hydrogen fuel cells are only about 30% efficient fuel-to-wheel, and they have a 95% efficient motor. All electric vehicles are about 70% fuel-to-wheel efficient.
One of the reputable tech questions YouTube channels did a comparison between gas and electric in several different ways. Using the standard USA grid distribution of renewable power sources, the break even for throwing out your gas car and all sources to build a brand new electric vehicle is about 4 years.
It's not that simple. Most of our state wide athletic programs bring in more money than the spend, creating a net profit regardless how ridiculous the amounts we pay for it, and ear-marked 3rd party donations from some very wealthy alumni. Why'd they spend $1mil on some fancy score board? Because someone donated $800k and the new board has increased paid spectator attendance enough to compensate for the $200k we had to cover.
You're conflating "highly educated" with "highly trained/skilled". Or as the head of the Computer Science department told us freshmen, "We're here to teach you how to learn for yourselves. We will present you with common and novel problems and expect you to figure out how to solve them on your own. You will have plenty of time in the real world to learn practical skills that will quickly fall out of style to the next computing fad, but what you learn here will last a life time."
Quite a bit of teaching and homework was done in pseduo-code and tests were open-book with internet access, but not open neighbor. The teacher didn't grade your ability to code, they graded your ability to reason. I do wish they spend time on coding in the sense of writing clean code and refactoring.
And if they used appropriate curses? I find mild cursing to soften the blow. If someone told me I was a "fucking idiot", to me that just means I done messed up and I need to learn from my mistake. If someone said I made an "absurd mistake", I would take that to mean that I am an incompetent person and will never be decent in their eyes.
You can beat around the bush all you want, but direct insults are less insulting that indirect ones.
Slashdot Mirror
I don't use QoS, I use an FairQ + AQM. Dropped packets is not just normal for TCP, it's a requirement. Dropping a packet early before bufferbloat kicks in reduces latency is reduces packetloss by signalling to the sender to backoff before the buffer fills and bloat plus burst loss occurs.
That's only true if the sender is DDOSing you or you literally don't have enough bandwidth. I QoS my incoming just fine by dropping packets which signals the sender to back off. And it's not just any packets getting dropped, it's statically biased to almost always be packets from the fattest flows. I can have my connection loaded to 99% of its provisioned rate and get 0 dropped gaming packets and less than 1ms of jitter.
They're conflating "talent" and "skilled labor". You don't train talent and you don't teach talent, you find talent.
I'm of two opinions about communication.
The practical side of me says communications is important. The ideal side of me says if the audience can't understand it, either you don't understand it or the audience is not smart enough in whatever domain that it matters. That said, I have issue with many in my profession using technical jargon because they use it incorrectly. They use it more like an ambiguous buzzword than a technical word with specific meaning. But again, this really just means the person talking doesn't understand.
In my experience, communication is more like a soft skill of using simple enough words that the receiver thinks they understand just enough. Just a "feel good" kind of skill. It's like describing a cube to someone who doesn't understand 3D shapes. In the end, you just tell them it a bunch of connected lines and they feel better about themselves because they know what a line is.
Working more than 40 hours or so makes you much more prone to mistakes. It is literally unprofessional to allow yourself to get into such a state. Learning when to say "no" is required to be a professional. Time management is critical to businesses and you can't do planning if you don't have good estimations. Back-pressure from people saying "no" is critical to the proper operations of a business.
Forced freedom is not freedom and the logical dissonance required to willingly conflate entitled access to other people's work and freedom is blind zealotry. With BSD, you always have access to the original, just no guarantee about access to derivatives. Which is "better" is subjective, but BSD is objectively more free.
The licenses affect different people differently, but BSD is still more free. Free isn't about what you can do, it's about what someone says you can't or have to do. The more someone controls you, the less free you are. The main argument isn't about which is more free, but which ideology is "better" as a whole. BSD is objectively more free than GPL, but GPL is subjectively better than BSD by enforcing certain subsets of freedom at the expense of others deemed less important.
because they're only making one video, aimed at billions of people
Ahh yes. The one-size-fits-all that fits no one. The "the average airman didn't exist" paradox.
Spinning disk has about as many sudden deaths are SSDs, the only different is spinners have an additional set of failures that give warning. In other words. If you had a harddrive that never died from mechanical issues, its failure rate would be very similar to an SSD.
Wear leveling is much less effective without TRIM. TRIM reduces write amplification by letting the wear leveling algorithm know when some data is no longer referenced.
I find it difficult to tell how good or bad someone else is in my profession without hanging out with them and talking on related subjects over several days. Not very practical for interviews. I cannot fault someone for not being good at picking out another skilled workers.
Making sure equipment and what-not is kept up-to-date is part of the responsibility of the department. If they deem that they want a new scoreboard in order to do what they need to do, then that's their decision. Their decisions will only be questioned if they fail to deliver positive results. And depends on what you mean by "publicly funded". If the money is from private sources, like student tuition and private donations, then it's not public money. Funding is not in one large pool. All of the money is earmarked and tracked.
Most people's problem solving ability is just trying the same tactic over and over again until it just happens to succeed in one instance. Just as important is to know if a problem has actually been solved, and not just swept under the carpet.
The whole "try random things until the symptom goes away" strategy that seems so prevalent. I r hard werking pragammer that putz n ovrtyme! I swear, semantics just whooshes over 99% of people's heads. These two variables have the same name, they must mean the same thing!
You have no idea how much I have to deal with this. My team leader, half jokingly, told me to stop reading other people's code because I keep finding problems and we have too many deliverables to have to deal with all of the bugs I find. A recent bug was when I was adding some new features to a project. I noticed that the database relationships seemed "strange". After a few hours of digging, I found a bug that has been resulting in corrupted data for nearly 20 years. The code in question had been worked on many tens of times over these 20 years, including several refactors and re-writes. Yet somehow every single programmer who ever touched it never noticed this exact same bug and every singe time recreated the same bug.
I didn't even see a bug. I just had a gut feeling that the inferred abstract semantics of the data was not intuitively clear to me, and I decided to go look at the code to validate. Any of the dozens of programmers who wrote and or reviewed the code could have done the same, yet no one did. This kind of situation happens to me all of them. Like several times a week. My team leader says I have super human attention to detail, but I really need to stop "breaking" "stable" code. My problem is I think abstractly about everything. I need to fully understand the semantics of everything, otherwise my mind just has grid lock with all of the possibility. I have a high attention to detail because I'm forced to. I have ADD, and the only way to keep my concentration on track is to have absolute confidence in my understanding of the code, otherwise my mind wanders, trying to find all edge cases.
This is why natural languages are horrible for communicating technical information.
I already have a connection fee. It pretty much has never changed for my entire life. It has technically changed, but at less rate than inflation. The grid is maintained by a company independent of who generates the power and gets paid entirely by the connection fee. Some energy companies even have negative margins for power sold. They're encouraged to reduce grid consumption.
Our energy companies make most of their profit from connection fees, energy consulting and energy services, but not almost nothing from energy sold. They're loving house level renewables because of the up-tick in demand for consulting and services.
I think battery banks for load shifting will probably be a better investment for areas where it doesn't get crazy hot. I live in the Midwest with $0.10/kwh flat rate or $0.05/offpeak and $0.15/onpeak. And I calculated that it would take about 8 years to pay off a $5,000 simple "just plug it into a general power outlet" programmable load shifting battery pack that even doubles as a UPS for all devices on the same breaker.
I only recently learned about these and they're pretty sweet. They come out of the box with a preconfigured charge timeframe, idle timeframe, and discharge timeframe, along with a rate of charge and discharge. You can configure these values yourself. It also has an option to monitor some energy site that has "current rates" for energy where it applies. Then the device and charge and discharge based on dynamic pricing. You can set high and low dynamic charging levels. It also works as a UPS by triggering your breaker when an outage is detected, then continues to supply power to the circuit. From what I've read, you don't need any professional installation at all, just plug it into any outlet in your house. You can have more than one in your house. Not sure if on the same circuit. I assume you can because it's meant to be brain-dead easy to use safely without any knowledge of your house.
What has a 10 year life? Not solar panels, they're much much higher. A recent study on individual solar cells, not entire panels, is that around 90% of used solar cells from 20+ years ago are still above 90% of their original fresh from the factory rating, and a substantial portion are above 95% of their original rating. What does drop over time is the entire panel for many panels, but not all. All it takes is one cell in the entire panel to make the output low and one low panel in a group to make the entire group low. But it's easy to swap out panels, less so individual cells. Even at the panel level, the 80/20 rule still applies. About 80% of panels are perfectly fine even after 20 years. The trend towards smaller panels with many micro-inverts is making this issue less of an issue.
Chronic exposure to wood dust is a very serious thing.
The problem I've seen others with community college is almost none of any credits for their major transferred and only GDRs did. They got stuck doing only GDRs for the first 1.5 years at their CC, but then were forced to try packing 4 years of their major into 2-3 years assuming they wanted to finish in 4, which may not be possible depending on prerequisite classes.
It's actually better than that for value. The $40mil saved was based on what the system recognized. The system can only track power in much coarser grained chunks, like minutes. The batter bank is providing power down into the millisecond ranges. If the grid could track at a finer resolution, it's already paid itself off.
And the bank is vastly overbuilt. It was designed to maintain its rated capacity for 10 years, then they added another 10% raw capacity for good measure. And the 10 years is based on the assumption of 100% cycle every day for 10 years. It's being cycled much less than that. The batteries should be good for several decades assuming no other major deterioration issues.
You can do that, but that process is going to be like 10% efficient. Even hydrogen fuel cells are only about 30% efficient fuel-to-wheel, and they have a 95% efficient motor. All electric vehicles are about 70% fuel-to-wheel efficient.
One of the reputable tech questions YouTube channels did a comparison between gas and electric in several different ways. Using the standard USA grid distribution of renewable power sources, the break even for throwing out your gas car and all sources to build a brand new electric vehicle is about 4 years.
"did I remember to plug in" and "is my car charged" can both be handled by smart phones.
It's not that simple. Most of our state wide athletic programs bring in more money than the spend, creating a net profit regardless how ridiculous the amounts we pay for it, and ear-marked 3rd party donations from some very wealthy alumni. Why'd they spend $1mil on some fancy score board? Because someone donated $800k and the new board has increased paid spectator attendance enough to compensate for the $200k we had to cover.
You're conflating "highly educated" with "highly trained/skilled". Or as the head of the Computer Science department told us freshmen, "We're here to teach you how to learn for yourselves. We will present you with common and novel problems and expect you to figure out how to solve them on your own. You will have plenty of time in the real world to learn practical skills that will quickly fall out of style to the next computing fad, but what you learn here will last a life time."
Quite a bit of teaching and homework was done in pseduo-code and tests were open-book with internet access, but not open neighbor. The teacher didn't grade your ability to code, they graded your ability to reason. I do wish they spend time on coding in the sense of writing clean code and refactoring.
And if they used appropriate curses? I find mild cursing to soften the blow. If someone told me I was a "fucking idiot", to me that just means I done messed up and I need to learn from my mistake. If someone said I made an "absurd mistake", I would take that to mean that I am an incompetent person and will never be decent in their eyes.
You can beat around the bush all you want, but direct insults are less insulting that indirect ones.