According to this site, Tildes uses as much carbon in a year as a tree (whatever that metric means) absorbs in a year. Should we start a yearly tradition of planting a tree to offset our carbon?...
According to this site, Tildes uses as much carbon in a year as a tree (whatever that metric means) absorbs in a year. Should we start a yearly tradition of planting a tree to offset our carbon? It sounds cheap enough to be doable, and like a fun tradition!
Looks as though they're relying on the Green Web Foundation for the energy source data - it seems like a reasonable enough heuristic to allow them to get something online (perfect enemy of good,...
Looks as though they're relying on the Green Web Foundation for the energy source data - it seems like a reasonable enough heuristic to allow them to get something online (perfect enemy of good, etc.), but it looks as though it's only going to show providers who've explicitly been registered in that database.
Also, while I absolutely like what Low Tech are doing with the solar server and the point they're making, I can't help but take issue slightly with the framing:
Quite a few web hosting companies claim that their servers are running on renewable energy. However, even when they actually generate solar power on-site, and do not merely “offset” fossil fuel power use by planting trees or the like, their websites are always online.
The internet's global, and a surprising number of data centres are in Iceland. Why Iceland? Because it's got a surplus of cheap electricity and essentially free cooling. Why's that electricity so cheap? Because it's >99% renewable, and both geothermal and hydroelectric lend themselves well to single big installations (compared to wind or PV solar, which scale per turbine or panel) which may as well be constructed with plenty of extra capacity on hand.
Less than 100% reliability is essential for the sustainability of an off-the-grid solar system, because above a certain threshold the fossil fuel energy used for producing and replacing the batteries is higher than the fossil fuel energy saved by the solar panels.
There are always gravity batteries, thermal batteries, and other non-fuel-intensive methods of energy storage. More generally, economies of scale are a good thing: a combination of redundant renewable sources is a great way to ride out the bumps in any one, and a lake sized gravity battery provides grid-scale consistency.
I know I'm nitpicking a bit, but the message they're sending leans a little further than I'd like towards suggesting that a renewable suggestion is necessarily worse than a non-renewable one, which seems a shame.
Let's add a bit of context to that, shall we? "Hurrah! This website is cleaner than 92% of websites tested" "This website appears to be running on sustainable energy"
Let's add a bit of context to that, shall we?
"Hurrah! This website is cleaner than 92% of websites tested"
"This website appears to be running on sustainable energy"
Note that the 10.000 views is a constant value for every site measured; you'd have to know the viewcounts to estimate a more accurate value. Reddit sure as hell has a huge multiple of that 10k.
Note that the 10.000 views is a constant value for every site measured; you'd have to know the viewcounts to estimate a more accurate value. Reddit sure as hell has a huge multiple of that 10k.
For those interested in the methodology, they have this page. It is necessarily rudimentary: in particular, I expect that the amount of data transferred is not a great estimate of how much of...
For those interested in the methodology, they have this page. It is necessarily rudimentary: in particular, I expect that the amount of data transferred is not a great estimate of how much of energy a page request uses in proportion to the total energy used by data centers: different sites, with different amounts of server-side processing being done, could presumably use vastly different amounts of energy. I expect it would be reasonably possible for a website operator with sufficient control over their servers to make a far more accurate estimate, and it might be interesting to encourage that information to be published.
More broadly, however, I have to wonder whether these types of estimates are potentially misleading in suggesting that people should care about these CO2 emissions, and potentially focus on reducing them, without giving a good sense of how these emissions fit into the global emissions landscape.
For example, using their 12kg/year estimate for Tildes, and looking around electricitymap.org, just the state of Queensland in Australia has electricity production that emits around 100,000kg/minute. With a population of around 5,000,000, if you accept per-capita emission as a reasonable comparison in this case, one person there is responsible in 10 hours for as much CO2 emissions, just from electricity production, as Tildes emits all year per their estimates. Similarly, taking a single, even extremely short, plane flight will emit more CO2 than 12kg, as will driving more than around 80km in an average petrol car (per carbonfootprint.com's 147g/km).
It's useful to think about reducing the CO2 emissions of data centers, but it's also important to understand where these numbers fit into wider CO2 emissions. It would actually probably be useful to develop a curriculum explaining relative CO2 emissions from a wide variety of sources, so that, in seeing direct emissions numbers, people could have a good sense of what those numbers mean.
I'm curious as to how that number is determined. I input a site of mine that's basically a static page with an image, and I got about 3 trees per 10,000 views. It seems like page size is one of...
I'm curious as to how that number is determined. I input a site of mine that's basically a static page with an image, and I got about 3 trees per 10,000 views. It seems like page size is one of the factors that's measured, and not anything like "work done on the backend to load this page", since they probably don't have ways of measuring that.
Also, how do JS-heavy sites compare?
Edit: Yeah, page size is a big factor, and only the size of the landing page, apparently. I just input a domain that has a decent amount of images on it, runs on Rails, and is overall very inefficient, but has a simple landing page. I'm down to 1 tree per 10k views for some reason.
You can imagine that they are not trying to provide you with a scientific measurement. It's just an amusing way of getting some perspective on the useless amount of code we're running in a browser.
You can imagine that they are not trying to provide you with a scientific measurement. It's just an amusing way of getting some perspective on the useless amount of code we're running in a browser.
According to this site, Tildes uses as much carbon in a year as a tree (whatever that metric means) absorbs in a year. Should we start a yearly tradition of planting a tree to offset our carbon? It sounds cheap enough to be doable, and like a fun tradition!
Looks as though they're relying on the Green Web Foundation for the energy source data - it seems like a reasonable enough heuristic to allow them to get something online (perfect enemy of good, etc.), but it looks as though it's only going to show providers who've explicitly been registered in that database.
Also, while I absolutely like what Low Tech are doing with the solar server and the point they're making, I can't help but take issue slightly with the framing:
The internet's global, and a surprising number of data centres are in Iceland. Why Iceland? Because it's got a surplus of cheap electricity and essentially free cooling. Why's that electricity so cheap? Because it's >99% renewable, and both geothermal and hydroelectric lend themselves well to single big installations (compared to wind or PV solar, which scale per turbine or panel) which may as well be constructed with plenty of extra capacity on hand.
There are always gravity batteries, thermal batteries, and other non-fuel-intensive methods of energy storage. More generally, economies of scale are a good thing: a combination of redundant renewable sources is a great way to ride out the bumps in any one, and a lake sized gravity battery provides grid-scale consistency.
I know I'm nitpicking a bit, but the message they're sending leans a little further than I'd like towards suggesting that a renewable suggestion is necessarily worse than a non-renewable one, which seems a shame.
My tildes competitor is pretty good too: littr.me.
I'm working on getting it to that Hacker News level. :)
Let's add a bit of context to that, shall we?
"Hurrah! This website is cleaner than 92% of websites tested"
"This website appears to be running on sustainable energy"
Note that the 10.000 views is a constant value for every site measured; you'd have to know the viewcounts to estimate a more accurate value. Reddit sure as hell has a huge multiple of that 10k.
For those interested in the methodology, they have this page. It is necessarily rudimentary: in particular, I expect that the amount of data transferred is not a great estimate of how much of energy a page request uses in proportion to the total energy used by data centers: different sites, with different amounts of server-side processing being done, could presumably use vastly different amounts of energy. I expect it would be reasonably possible for a website operator with sufficient control over their servers to make a far more accurate estimate, and it might be interesting to encourage that information to be published.
More broadly, however, I have to wonder whether these types of estimates are potentially misleading in suggesting that people should care about these CO2 emissions, and potentially focus on reducing them, without giving a good sense of how these emissions fit into the global emissions landscape.
For example, using their 12kg/year estimate for Tildes, and looking around electricitymap.org, just the state of Queensland in Australia has electricity production that emits around 100,000kg/minute. With a population of around 5,000,000, if you accept per-capita emission as a reasonable comparison in this case, one person there is responsible in 10 hours for as much CO2 emissions, just from electricity production, as Tildes emits all year per their estimates. Similarly, taking a single, even extremely short, plane flight will emit more CO2 than 12kg, as will driving more than around 80km in an average petrol car (per carbonfootprint.com's 147g/km).
It's useful to think about reducing the CO2 emissions of data centers, but it's also important to understand where these numbers fit into wider CO2 emissions. It would actually probably be useful to develop a curriculum explaining relative CO2 emissions from a wide variety of sources, so that, in seeing direct emissions numbers, people could have a good sense of what those numbers mean.
I'm curious as to how that number is determined. I input a site of mine that's basically a static page with an image, and I got about 3 trees per 10,000 views. It seems like page size is one of the factors that's measured, and not anything like "work done on the backend to load this page", since they probably don't have ways of measuring that.
Also, how do JS-heavy sites compare?
Edit: Yeah, page size is a big factor, and only the size of the landing page, apparently. I just input a domain that has a decent amount of images on it, runs on Rails, and is overall very inefficient, but has a simple landing page. I'm down to 1 tree per 10k views for some reason.
You can imagine that they are not trying to provide you with a scientific measurement. It's just an amusing way of getting some perspective on the useless amount of code we're running in a browser.