The internet is responsible for around 5% of global carbon emissions. It's difficult to measure, and generally not visible to the public.
The solution however is fairly simple — use less data.
Fast, accessible websites usually produce less CO2 and we already know how to do that. Now we just need to get web developers to do it. This will require:
- educating developers and designers - smart defaults for web developers - fostering competition - making the invisible cost of the internet visible
This is a reletively new dicipline — so anything you have to add to the mix will have a large impact.
Right, the internet (not necessarily the web)
> The solution however is fairly simple — use less data
Agreed.
> Fast, accessible websites usually produce less CO2
That's a drop in the bucket. If you look at what's _actually_ using data on the internet, it's streaming services (netflix, zoom, spotify). Fast, accessible websites have plenty of benefits, but if you want to actually reduce the amount of waste, far more effective solutions are things like lowering your video quality slightly (backgrounding futurama? Drop it to 480p.)
At the end of the day, the real culprit is the infrastructure, and emergence of tech giants who don't have environmental sustainability anywhere in their mission.
But, in what sense in video streaming not part of the web?
Sustainable web design is just as applicable to streaming as e-commerce or a blog.
It’s also not at all the same as turning off a lightbulb.
It’s more like one action resulting in turning off millions of other peoples light bulbs.
One of the points he made was that it's very difficult to sell sustainable design, and much easier to just tell clients it's about performance.
His other observation was that clients care a lot if you tell them their competitor is doing environmentalism better.
How much of that is high quality video streaming? How much could be reduced by forbiddin resolutions over 1080p for example?
This is not about "hunger to consume more data". Actual data doesn't take much. Needlessly inflated codebases and excessive video fidelity do.
Most of the people could use a decades-old computer and a <1Mbit/s connection to get and send all the information they might ever be interested in if we didn't insist on using huge JavaScript libraries, fancy design and HD+ video we lived perfectly well without just a decade ago. The problem is we seemingly need ever increasing everything to keep the economy growing.
Regular people don't ask for this stuff. It is presented to them as a gift, and they take the gift because of this or that specific reason. Don't blame regular people; they aren't the ones actively trying to keep the growth machine rolling.
The computer I'm writing this on is nearly a decade old and performs perfectly well on modern websites. Granted, I'm not going to try using a < 1Mb/s connection.
That being said, I agree that things like client-side (and server-side) scripting and video decoding (and encoding) reflects a disinterest in environmental concerns. I will disagree about the quantity of data playing a negligible role, since serving it may not be very energy hungry but building out infrastructure is. We can satisfy our needs with a lot less if we used all of our resources more effectively.
What's that, bloated news site? You don't want a bill of 10 USD per page served? Cut your bloat.
This also has the effect of shutting out people with expensive data, or slow networks. Who i'd guess are the same people most effected by the climate crisis.
I am a big fan of sustainability, but this seems a little bit much.
A simple half-way solution to this is just to support a dark mode. People generally prefer it anyway.
Only on OLED screens, which still don't have a that impressive market share.
Anyone remember blackle.com?
It rated as not mobile friendly (even if I have separate media queries for mobile/tablet).
It ranked negatively on MozRank (how well linked (?) it is from other websites, which I don't understand -- neither does it explains -- how it relates to eco friendliness). At least the main site itself is penalized by the same MozRank criteria.
Seems like a tool to generate leads to Mightybyte consulting services (who developed the tool, a founding organization behind sustainable web design).
About the manifesto I have mixed feelings, to me it feels rather "wishy washy".
Wish them luck with the project.
It also has the same bug I encounter often on full bleed designs where there's extra padding/margin that makes a horizontal scrollbar visible. Which reminds me that even experienced designers still don't have a good grasp on CSS /s
MightyBytes and WholegrainDigital, seem to be the leaders in this space.
Which is why I'll be sending them a long, annoying list of things I think can be improved.
---
the idea of the MozLink thing is that you reduce the number of clicks it takes to find your website — and so reducing the number of requests made to servers.
It's not a very good metric.
And then they send you on a merry goose chase through various blog sites if you want to understand what MozLink is, and how you can improve it. How ironic.
I think the only reasonable step "up" (or down, specs wise) would be acceptable performance on $30 KaiOS devices, to maximize accessibility to visitors from developing countries, but that's about it?
> Efficient
> The products and services we provide will use the least amount of energy and material resources possible.
Ha! Let's see how much impact this has on the current web industry.
Development velocity vs efficiency.
I've yet to see a shop avoid Tower of Babel stacks that increase the former, at the expense of the latter. Unless it's a backwards-compatible drop-in upgrade. And even then, it doesn't bubble high on the priority list.
[0] Obligatory https://m.youtube.com/watch?v=BwozkbmjzC4
As for backwards-compatible drop-in upgrades. They do exists, and people do use them. Making them easier would go a long way.
An example would be migrating a wordpress site to be statically generated — there are plugins and services that do that.
If you're going to pitch slower and more expensive development to a team, it needs a lot better hook than this.
... Unless we're talking about Exxon, who will probably adopt this tomorrow to show how seriously they take climate change.
Even then, reliable data and analysis may produce differing conclusions based upon shifting priorities. Today's discussions are tied to greenhouse gasses. In prior decades, environmental discussions were tied to: deforestation and habitat loss, groundwater pollution, air pollution with respect to public health and acid rain, the ozone hole, etc..
Even if we could agree on priorities, it is difficult to assess the true impact because technological change brings about behavioural change. The automobile may be cleaner than the horse and buggy. On the other hand, more people walked or used mass transit before the automobile. Streaming a movie may be cleaner than watching a DVD, but is it cleaner when each member of a household is viewing a different movie on a portable device in contrast to watching a broadcast together on the family TV.
I mean, hey, there are things we can do besides picketing amazon for building spaceships instead of solar arrays.
Lot's of hosting companies including all the large clouds have marketing pages about their sustainability efforts but only an obscure, expense, and inaccessible Azure service can provide CO2 pollution estimations from ones usage...
As for the web, imho there's mostly one hard solution being static web-sites being aggressively cached in a broadcast tree (as in real offline-modes, website downloading). Just like high-speed personal transportation vs collective fixed-schedule medium-speed transportation. I'd guess internet consumption is not much dependent on distance, but speed and duplication of the transfers, that's probably a thing.
In this kind of realm i'd be curious to see the life-cycle energy cost comparison of radio-broadcast numeric TV vs cabled broadcast. Perhaps we could use numeric radio-broadcast as a distribution channel for many-to-one websites like news and newsletter-type of big social media profiles. I hope to one day say "oh it's 4:03 i'll tune my receiver to get the local blog updates".
# Can you delete my test results?
If you have submitted a test and do not want the results public, then we have a simple policy:
[…]
If you have tested a public url we will not delete the results.
IEA claims datacenters account for 1% of electricity demand and data networks are another 1%:
IEA (2020), Data Centres and Data Transmission Networks, IEA, Paris https://www.iea.org/reports/data-centres-and-data-transmissi...
Now, those figures are _electricity_ not _carbon emissions_ - so perhaps the 3.8% figure includes e.g. datacenter construction emissions, or emissions related to producing servers and networking equipment?
How do they measure this? Do they have the electricity consumption and supply of each datacenter around the globe? Two DCz in Poland, one using coal energy, or another one using only renewable energies; the first one has latest gen servers and networking equipment; the second one 15 year old stuff. Which one uses less CO2?
There can be lots of differences depending on location and specificities.
