The Linux community has a hard time attracting talent precisely because the people in charge have essentially zero skill in interpersonal relationships
You don't need interpersonal skills as long as you have good communication, existing talent, and a highly desirable product. Coddling the riff-raff will kill Linux faster than scaring them away.
Linus is too busy to hold everyone's hands to make them feel special. Maybe she has untapped potential that could revolutionize Linux, but more than likely, she's just a plebeian trying to see what it's like to work with the elite. Some of it is also a culture issue. Linus comes from a culture where you don't sugar coat everything. If she doesn't like it, she can fork Linux or make her own kernel. Extremely skilled people tend to be rough around the edges.
GPLv3 is not compatible with nearly any other license. If your program is not GPLv3, you won't be able to use the reference code. you are free to reinvent the wheel and compare its output against the reference and hope it works well. Most programs in use are not GPL and FLIF will probaly never get supported until someone remakes it under a new license like BSD, Apache. or CDDL.
Rule of thumb. If your license requires a lawyer to review it, it's not free.
But it does prevent you from using any of the original GPL code. Some of your code so happens to look similar? Prove you didn't use GPL or get sued! Can't prove it?
Users never look at the source code, only developers do, and GPL is anti-developer. Kind of like city folk saying "screw farmers, they shouldn't get subsidies! If they want Internet, they should come to the city". Just wait until you have no more farmers.
"Honor Society" was full of functional idiots that so happened to do very well regurgitating information. Most of them had virtually no ability to apply their knowledge. They couldn't find their way of a paper bag without instructions, and even then, they could tell you the instructions but had no ability to comprehend what they meant.
"Fake" is the most widely used social skill. Being fake takes too much effort, introverts don't like to waste mental capacity because people are insecure about themselves. My #1 important thing for finding a wife was someone I could just be myself around. If I can't be myself, it's too taxing.
That's a step in the right direction, but it still doesn't measure quality very much. Design of the code is also very important. How easy is the code to maintain, extend, refactor, re-use? Does it interact in negative ways with other code? What happens when you push it to its limits? What scaling does it have? What are its failure cases and how does it handle them? How intuitive is the code? Is it written in a way that makes it difficult to use it incorrectly while being simple to use correctly or in new ways that could also be correct?
Not where I work. They let me do my thing on my own, then I bring everyone up to speed once I feel the product is ready for use. Of course the bread and butter products are programmed in groups, but the special snowflake products that just need to work, those are done by solo or small self-forming and self-organizing groups.
Most of my projects get deployed to production and last 3-5 years before fixes or changes need to be done. The group projects tend to have weekly bug fixes and strange corner cases that can take days to debug.
The biggest issue is the bus factor. When one person designs and implements a project, no one else knows how it works. The icing on the cake is that the projects rarely need to be fixed, making it difficult to pair-program bug fixes, and they tend to be critical infrastructure where time is very important.
I've been learning a lot about making cleaner code. Because many of my projects go years between looking at the code, I get to read my own code and realize how horrible some code practices are for reading. Learning by failure is a great way to learn if you can afford it.
So they've figured out how to measure how well something works. This is fine for products where the outcome can be measured with natural numbers. What about products where they can have negative value even when they do work? If I improperly write code that works well when it works, but fails in unknown ways, then my product has a negative value when compared to another product the solves the same problems, but in a way where it fails in predictable ways and can be quickly fixed.
"Friendship" in the same sense as talking to a stranger while waiting in line and calling them a "friend". A friend isn't someone that you know, it's someone you can depend upon.
I am an introverted person and I do well on my own, but I also like having some contact with other people. When I was younger, spending time away from other people allowed me to learn more about myself. As time went on, I started to reach the limits of what I could learn about myself alone and I needed to be around people to find out more about me.
Being around people is a large energy drain for me, but I do require some interaction to be optimal.
According to their site, it's not backwards compatible. yay.. You need to patch your network state to make it work. That's the whole issue with why we need IPv6, because we can't afford to patch anything. And no one is going to make a hardware router that will handle that horrible datastructure. Splitting routing information around the packet.
Yes, the idea works, but only if you can patch a lot of stuff and no hardware is going to support it, only software. It's very limited in usage.
On average, NAT reduces security. It is a leaky abstraction with no standard implementation and is know to have many bugs for any given implementation. NAT many times reduces security because many of these bugs can allow external attackers to effectively by-pass the stateful firewall.
NAT works just fine. Then why are so many forums filled with questions of why their games don't work where more than one person on their network attempts to play? Maybe because only once device can claim a port.
"DNS" works on my network without a DNS server. Has worked this way for at least a decade. I can even result my printer by name without a DNS server. They have had local subnet multicast P2P DNS protocols for a long while now.
Slashdot Mirror
The Linux community has a hard time attracting talent precisely because the people in charge have essentially zero skill in interpersonal relationships
You don't need interpersonal skills as long as you have good communication, existing talent, and a highly desirable product. Coddling the riff-raff will kill Linux faster than scaring them away.
I've had people talk to me like this. After I slept on it, I thanked them for their honesty. I'll NEVER make that same mistake again.
Linus is too busy to hold everyone's hands to make them feel special. Maybe she has untapped potential that could revolutionize Linux, but more than likely, she's just a plebeian trying to see what it's like to work with the elite. Some of it is also a culture issue. Linus comes from a culture where you don't sugar coat everything. If she doesn't like it, she can fork Linux or make her own kernel. Extremely skilled people tend to be rough around the edges.
Both perspectives are equally important, so the side that complains the most needs to grow up.
I've heard of O3 and UV water treatment. I hear they are very good at killing pretty much living thing.
GPLv3 is not compatible with nearly any other license. If your program is not GPLv3, you won't be able to use the reference code. you are free to reinvent the wheel and compare its output against the reference and hope it works well. Most programs in use are not GPL and FLIF will probaly never get supported until someone remakes it under a new license like BSD, Apache. or CDDL.
Rule of thumb. If your license requires a lawyer to review it, it's not free.
But it does prevent you from using any of the original GPL code. Some of your code so happens to look similar? Prove you didn't use GPL or get sued! Can't prove it?
Users never look at the source code, only developers do, and GPL is anti-developer. Kind of like city folk saying "screw farmers, they shouldn't get subsidies! If they want Internet, they should come to the city". Just wait until you have no more farmers.
I'm only 4 minutes in, but this is a cool video so far.
In some situations, standing out is bad. In other situations, it means you're irreplaceable.
"Honor Society" was full of functional idiots that so happened to do very well regurgitating information. Most of them had virtually no ability to apply their knowledge. They couldn't find their way of a paper bag without instructions, and even then, they could tell you the instructions but had no ability to comprehend what they meant.
/. has a selection bias towards introverts.
"Fake" is the most widely used social skill. Being fake takes too much effort, introverts don't like to waste mental capacity because people are insecure about themselves. My #1 important thing for finding a wife was someone I could just be myself around. If I can't be myself, it's too taxing.
Depends on your definition of "cheaper". Few bean counters seem to care about opportunity costs.
That's a step in the right direction, but it still doesn't measure quality very much. Design of the code is also very important. How easy is the code to maintain, extend, refactor, re-use? Does it interact in negative ways with other code? What happens when you push it to its limits? What scaling does it have? What are its failure cases and how does it handle them? How intuitive is the code? Is it written in a way that makes it difficult to use it incorrectly while being simple to use correctly or in new ways that could also be correct?
Not where I work. They let me do my thing on my own, then I bring everyone up to speed once I feel the product is ready for use. Of course the bread and butter products are programmed in groups, but the special snowflake products that just need to work, those are done by solo or small self-forming and self-organizing groups.
Most of my projects get deployed to production and last 3-5 years before fixes or changes need to be done. The group projects tend to have weekly bug fixes and strange corner cases that can take days to debug.
The biggest issue is the bus factor. When one person designs and implements a project, no one else knows how it works. The icing on the cake is that the projects rarely need to be fixed, making it difficult to pair-program bug fixes, and they tend to be critical infrastructure where time is very important.
I've been learning a lot about making cleaner code. Because many of my projects go years between looking at the code, I get to read my own code and realize how horrible some code practices are for reading. Learning by failure is a great way to learn if you can afford it.
So they've figured out how to measure how well something works. This is fine for products where the outcome can be measured with natural numbers. What about products where they can have negative value even when they do work? If I improperly write code that works well when it works, but fails in unknown ways, then my product has a negative value when compared to another product the solves the same problems, but in a way where it fails in predictable ways and can be quickly fixed.
"Friendship" in the same sense as talking to a stranger while waiting in line and calling them a "friend". A friend isn't someone that you know, it's someone you can depend upon.
I am an introverted person and I do well on my own, but I also like having some contact with other people. When I was younger, spending time away from other people allowed me to learn more about myself. As time went on, I started to reach the limits of what I could learn about myself alone and I needed to be around people to find out more about me.
Being around people is a large energy drain for me, but I do require some interaction to be optimal.
According to their site, it's not backwards compatible. yay.. You need to patch your network state to make it work. That's the whole issue with why we need IPv6, because we can't afford to patch anything. And no one is going to make a hardware router that will handle that horrible datastructure. Splitting routing information around the packet.
Yes, the idea works, but only if you can patch a lot of stuff and no hardware is going to support it, only software. It's very limited in usage.
On average, NAT reduces security. It is a leaky abstraction with no standard implementation and is know to have many bugs for any given implementation. NAT many times reduces security because many of these bugs can allow external attackers to effectively by-pass the stateful firewall.
1ms?! That's horrible to local host
Source address is 10.255.255.18; using ICMP echo-request, ID=94cf
Pinging 10.255.255.1 [10.255.255.1]
with 32 bytes data (60 bytes IP):
From 10.255.255.1: bytes=60 seq=0001 TTL=64 ID=e63f time=0.096ms
From 10.255.255.1: bytes=60 seq=0002 TTL=64 ID=2122 time=0.094ms
From 10.255.255.1: bytes=60 seq=0003 TTL=64 ID=c2c0 time=0.065ms
From 10.255.255.1: bytes=60 seq=0004 TTL=64 ID=2122 time=0.094ms
Source address is 127.0.0.1; using ICMP echo-request, ID=3cfa
Pinging 127.0.0.1 [127.0.0.1]
with 32 bytes data (60 bytes IP):
From 127.0.0.1: bytes=60 seq=0001 TTL=128 ID=3e58 time=0.008ms
From 127.0.0.1: bytes=60 seq=0002 TTL=128 ID=3e5a time=0.007ms
From 127.0.0.1: bytes=60 seq=0003 TTL=128 ID=3e5c time=0.008ms
From 127.0.0.1: bytes=60 seq=0004 TTL=128 ID=3e5e time=0.007ms
NAT works just fine. Then why are so many forums filled with questions of why their games don't work where more than one person on their network attempts to play? Maybe because only once device can claim a port.
"DNS" works on my network without a DNS server. Has worked this way for at least a decade. I can even result my printer by name without a DNS server. They have had local subnet multicast P2P DNS protocols for a long while now.
They loosened the rules on IPv6 addresses and allow multihoming via more than one ISP.