3 ms·

In my opinion, this is ridiculous. The failure of this argument is here: > Therefore, increasing your consumption of computing resources increases your contrib

In my opinion, this is ridiculous. The failure of this argument is here: > Therefore, increasing your consumption of computing resources increases your contribution to the suffering of others. Yes, this is technically true. I think there's an easy way to trump this: The amount by which your extra compute cycles increase global warming is much less than both of (the opportunity cost of spending your time tweaking your code instead of some other method of fighting global warming) and (the amount that some single-digit percentage of your salary could contribute to fighting global warming). At some point, you will be punished for prioritizing performance-tuning code over producing business value. If you use your dollars to fight against global warming in some way, this is a net negative. Likewise, if you do this off the clock, I believe there are other volunteer actions you could take that would decrease carbon emissions more than spending that time optimizing code.

CppCoder 6y ago
You have to put this into perspective. I am working on an MMO gane and not only can good code half the server amount but just as well the internet traffic, which in turn reduces mobil traffic consumption on the mobile device. Just like the optimized client (game) saves power on the mobile device. And if you scale this to thousands of servers and millions of players, I am sure it time invested is well worth it.
- RugnirViking 6y ago
  Yes, but if it's worthwhile, people will already do it for other reasons, like the ones you listed (decreased cooling costs, decreased server costs, decreased bandwidth costs)
bluGill 6y ago
Even worse, most programs are not run very often. The CO2 you spend working on making it more efficient (all the test runs) is more than the total CO2 you save in many cases. Thus making your first working draft of code (working meaning bug free enough for your purposes) uses less CO2 over the lifetime of your code than you would spend working on making it faster There are exceptions. There is a reason why people spend years staring at micro-benchmarks of sort to save one cycle: they can save tons of CO2 every years. However for most of us sort is just a function in our library and we shouldn't think about it. Your typical programmer thinking about sort will waste far more CO2 due to a worse algorithm than someone else already put into the standard library.
phkahler 6y ago
Dont forget there is a presumed value in writing the code in the first place. Sometimes having inefficient code is better than having no code. It's very difficult to compare these kinds of things, so common practice is to assign dollar values to things that are infungible. Perhaps the author needs to advocate for higher costs of carbon emissions rather than try to argue about the morality of such an abstract thing.
- pmiller2 6y ago
  Yes, and, sometimes code does not make the world a better place, or makes the world less better than those other things one could be doing to save the world that the GP alludes to.

CppCoder 6y ago

You have to put this into perspective. I am working on an MMO gane and not only can good code half the server amount but just as well the internet traffic, which in turn reduces mobil traffic consumption on the mobile device. Just like the optimized client (game) saves power on the mobile device. And if you scale this to thousands of servers and millions of players, I am sure it time invested is well worth it.

RugnirViking 6y ago
Yes, but if it's worthwhile, people will already do it for other reasons, like the ones you listed (decreased cooling costs, decreased server costs, decreased bandwidth costs)

bluGill 6y ago

Even worse, most programs are not run very often. The CO2 you spend working on making it more efficient (all the test runs) is more than the total CO2 you save in many cases. Thus making your first working draft of code (working meaning bug free enough for your purposes) uses less CO2 over the lifetime of your code than you would spend working on making it faster There are exceptions. There is a reason why people spend years staring at micro-benchmarks of sort to save one cycle: they can save tons of CO2 every years. However for most of us sort is just a function in our library and we shouldn't think about it. Your typical programmer thinking about sort will waste far more CO2 due to a worse algorithm than someone else already put into the standard library.

phkahler 6y ago

Dont forget there is a presumed value in writing the code in the first place. Sometimes having inefficient code is better than having no code. It's very difficult to compare these kinds of things, so common practice is to assign dollar values to things that are infungible. Perhaps the author needs to advocate for higher costs of carbon emissions rather than try to argue about the morality of such an abstract thing.

pmiller2 6y ago
Yes, and, sometimes code does not make the world a better place, or makes the world less better than those other things one could be doing to save the world that the GP alludes to.