search

HTTPArchive / almanac.httparchive.org

HTTP Archive's annual "State of the Web" report made by the web community
https://almanac.httparchive.org
Apache License 2.0
601 stars 164 forks source link

Sustainability 2022 #2910

Closed rviscomi closed 1 year ago

rviscomi commented 2 years ago

Sustainability 2022

If you're interested in contributing to the Sustainability chapter of the 2022 Web Almanac, please reply to this issue and indicate which role or roles best fit your interest and availability: author, reviewer, analyst, and/or editor.

Content team

Lead Authors Reviewers Analysts Editors Coordinator
@ldevernay @ldevernay @gerrymcgovernireland @timfrick @mrchrisadams @cqueern @Djohn12 @fershad @camcash17 @4upz - @tunetheweb
Expand for more information about each role 👀 - The **[content team lead](https://github.com/HTTPArchive/almanac.httparchive.org/wiki/Content-Team-Leads'-Guide)** is the chapter owner and responsible for setting the scope of the chapter and managing contributors' day-to-day progress. - **[Authors](https://github.com/HTTPArchive/almanac.httparchive.org/wiki/Authors'-Guide)** are subject matter experts and lead the content direction for each chapter. Chapters typically have one or two authors. Authors are responsible for planning the outline of the chapter, analyzing stats and trends, and writing the annual report. - **[Reviewers](https://github.com/HTTPArchive/almanac.httparchive.org/wiki/Reviewers'-Guide)** are also subject matter experts and assist authors with technical reviews during the planning, analyzing, and writing phases. - **[Analysts](https://github.com/HTTPArchive/almanac.httparchive.org/wiki/Analysts'-Guide)** are responsible for researching the stats and trends used throughout the Almanac. Analysts work closely with authors and reviewers during the planning phase to give direction on the types of stats that are possible from the dataset, and during the analyzing/writing phases to ensure that the stats are used correctly. - **[Editors](https://github.com/HTTPArchive/almanac.httparchive.org/wiki/Editors'-Guide)** are technical writers who have a penchant for both technical and non-technical content correctness. Editors have a mastery of the English language and work closely with authors to help wordsmith content and ensure that everything fits together as a cohesive unit. - The **[section coordinator](https://github.com/HTTPArchive/almanac.httparchive.org/wiki/Section-Leads'-Guide)** is the overall owner for all chapters within a section like "User Experience" or "Page Content" and helps to keep each chapter on schedule. _Note: The time commitment for each role varies by the chapter's scope and complexity as well as the number of contributors._ For an overview of how the roles work together at each phase of the project, see the [Chapter Lifecycle](https://github.com/HTTPArchive/almanac.httparchive.org/wiki/Chapter-Lifecycle) doc.

Milestone checklist

0. Form the content team

1. Plan content

2. Gather data

3. Validate results

4. Draft content

5. Publication

Chapter resources

Refer to these 2022 Sustainability resources throughout the content creation process:

📄 Google Docs for outlining and drafting content 🔍 SQL files for committing the queries used during analysis 📊 Google Sheets for saving the results of queries 📝 Markdown file for publishing content and managing public metadata 💬 #web-almanac-sustainability on Slack for team coordination

rviscomi commented 2 years ago

@cqueern @gerrymcgovernireland you've expressed an interest in a sustainability chapter before. This chapter could use coauthors, reviewers, and analysts if you still have the interest and time!

@ldevernay has stepped up to author the chapter 🎉

tunetheweb commented 2 years ago

@fershad also was interested in such a chapter.

fershad commented 2 years ago

@tunetheweb @rviscomi I'd love to put my hand up as an analyst, though will need guidance when it comes to querying the dataset.

@mrchrisadams would you also be interested in contributing?

gerrymcgovernireland commented 2 years ago

Yes, definitely interested if there’s any way I can help

From: Rick Viscomi @.> Sent: Thursday 14 April 2022 04:41 To: HTTPArchive/almanac.httparchive.org @.> Cc: gerrymcgovernireland @.>; Mention @.> Subject: Re: [HTTPArchive/almanac.httparchive.org] Sustainability 2022 (Issue #2910)

@cqueern https://github.com/cqueern @gerrymcgovernireland https://github.com/gerrymcgovernireland you've expressed an interest in a sustainability chapter before. This chapter could use coauthors, reviewers, and analysts if you still have the interest and time!

@ldevernay https://github.com/ldevernay has stepped up to author the chapter 🎉

— Reply to this email directly, view it on GitHub https://github.com/HTTPArchive/almanac.httparchive.org/issues/2910#issuecomment-1098678347 , or unsubscribe https://github.com/notifications/unsubscribe-auth/AQGRE3ZFULNGNUUXK3M5LE3VE6HVJANCNFSM5TMP2RQA . You are receiving this because you were mentioned. https://github.com/notifications/beacon/AQGRE37NPZDAONV53A2JD5TVE6HVJA5CNFSM5TMP2RQKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOIF6IASY.gif Message ID: @. @.> >

ldevernay commented 2 years ago

This enthusiasm is awesome! I will look into the outline to offer some details today.

I'm really eager to work with you all!

mrchrisadams commented 2 years ago

Hi folks.

I'd be happy to contribute as a reviewer at least, given those timelines. I'll ask internally to see if there's anyone who might be able to contribute in other roles.

cqueern commented 2 years ago

@cqueern @gerrymcgovernireland you've expressed an interest in a sustainability chapter before. This chapter could use coauthors, reviewers, and analysts if you still have the interest and time!

I'd love to support as a Reviewer. Thanks. Looking forward to working with everyone.

bkardell commented 2 years ago

I'm curious what this will cover?

ldevernay commented 2 years ago

In my opinion, the point would be to start with insights on the environmental footprint of digital then focus on sustainability applied to websites. Which tools to use to monitor/measure, where to find best practices (and how to prioritize them). Based on this, analysis of the main proxy metrics for environmental impacts : page weight and number of request (most often use to calculate environmental indicators through various models), explore how to use proxy metrics (such as Core Web Vitals) and finally which other metrics could used to check on sustainability best practices (image format, inclusion of third-party, minification, compression, cache, lazy-loading, etc).

The idea would also be to highlight the links to other topics (Performance, Accessibility, Privacy, etc) and briefly explain how to go further (design reviews, measuring other metrics, setting an environmental budget, etc).

Also : state of the art regarding repositories of best practices and tools to calculate the environmental impact (and where to go from here).

gerrymcgovernireland commented 2 years ago

These are all great points to cover. Some possible added suggestions:

  1. If there was some way we could identify page ‘waste’ that could be useful. This might include unused CSS, JS, poor coding that is slowing things down and thus wasting energy, poorly optimized images, etc. We could give good tips on reducing such waste.
  2. In the analysis I’ve been doing, over 90% of the energy impacts are either in the creation of the page or during its use. In popular sites, the vast majority of energy occurs during the page use.
  3. There have been interesting studies on the energy impact of various computer languages, with JavaScript being found to be 4.5 times more energy intense than C, for example. I wonder might there be a way to create and energy hierarchy for JavaScript, CSS, HTML, etc. In this way we could guide designers and developers to use the least energy intense option if given a choice. I wonder are there ways to identify design elements that are typically done in JS, for example, that could be done in, say, CSS
  4. This is probably out of scope, but typically 80% of CO2 and other negative impacts are caused during the manufacture of the digital devices and network architecture. I think if we wanted to deliver a true and total figure of the environmental impact of the Web, we should consider some accounting for the devices.

Tim Frick of Mightybytes and Tom Greenwood of Wholegrain Digital are two people who huge pioneering work in this space. Perhaps it would be worth reaching out to them.

From: LaurentDev @.> Sent: Thursday 14 April 2022 15:27 To: HTTPArchive/almanac.httparchive.org @.> Cc: gerrymcgovernireland @.>; Mention @.> Subject: Re: [HTTPArchive/almanac.httparchive.org] Sustainability 2022 (Issue #2910)

In my opinion, the point would be to start with insights on the environmental footprint of digital then focus on sustainability applied to websites. Which tools to use to monitor/measure, where to find best practices (and how to prioritize them). Based on this, analysis of the main proxy metrics for environmental impacts : page weight and number of request (most often use to calculate environmental indicators through various models), explore how to use proxy metrics (such as Core Web Vitals) and finally which other metrics could used to check on sustainability best practices (image format, inclusion of third-party, minification, compression, cache, lazy-loading, etc).

The idea would also be to highlight the links to other topics (Performance, Accessibility, Privacy, etc) and briefly explain how to go further (design reviews, measuring other metrics, setting an environmental budget, etc).

Also : state of the art regarding repositories of best practices and tools to calculate the environmental impact (and where to go from here).

— Reply to this email directly, view it on GitHub https://github.com/HTTPArchive/almanac.httparchive.org/issues/2910#issuecomment-1099243158 , or unsubscribe https://github.com/notifications/unsubscribe-auth/AQGRE3Y652DV7FUFNV4Q2CTVFATKHANCNFSM5TMP2RQA . You are receiving this because you were mentioned. https://github.com/notifications/beacon/AQGRE36HBI5F7TM65UN3LJ3VFATKHA5CNFSM5TMP2RQKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOIGCR5FQ.gif Message ID: @. @.> >

ldevernay commented 2 years ago

Great points @gerrymcgovernireland ! Here is the complete outline :

Introduction Environmental impact of digital Understanding the impact Main figures, resources Studies : https://www.greenit.fr/environmental-footprint-of-the-digital-world/ https://www.greenit.fr/wp-content/uploads/2021/12/EU-Study-LCA-7-DEC-EN.pdf

Most of the impact comes from user devices, mostly because their fabrication is very impactful. What we can do about that is to reduce the impact of digital services (and change the way we think about digital as being immaterial and environmentally friendly by default). Bonus : Insights from Gerry McGovern (World Wide Waste)

Evaluating the environmental impact of websites Tools (EcoIndex, website carbon, etc) : https://marmelab.com/blog/2022/04/05/greenframe-compare.html => Understanding the choice of metrics + what’s to expect from tools in the coming years

Reducing the impact Coupling best practices and measuring (via tools). Repository of best practices (115 best practices, INR, etc). Books and websites (sustainablewebdesign.org, etc)

Analysis by Metrics Monitoring the impact Page Weight Number of requests (and their distribution)

Checking for the usage of sustainability best practices on the web Cache Image Optimization => Lazy-loading, responsiveness, format, quality JS/CSS :Compression/minification, Unused code Video : preload, etc. Animations? "Those pages people will never open"? Third parties Bonus : web performance VS sustainability. Where they meet, where they differ Conclusion

This covers most of your points, I think.

On best practices, we should add "limit/avoid animations" with a focus on avoiding JS in favor of CSS for "essential" animations.

ldevernay commented 2 years ago

I sent a message to Tom yesterday and another to Tim this morning, hope they will have some time for this!

On the impact being mostly on devices, I totally agree and that's what the studies mentioned on my previous comments illustrate. This is why, in the company where I currently work, we also gather metrics on real devices. On smartphones, this is a good way to measure the depletion of the battery, which allows us, through an environmental model, to evaluate the global environmental impact of a digital service and the share that is due to device manufacturing. As of today, this is still (unfortunately) a proprietary methodology and model. More details here : https://greenspector.com/en/environmental-footprint-methodology/

timfrick commented 2 years ago

Hi All,

Just getting caught up here. Look forward to collaborating with you all. Per some of the points brought up earlier in this thread:

For the Sustainable Web Design model, we based our consumer device use numbers mostly on the Andrae study, which estimates that at 52% of the system with repeat visitors using 25% and loading 2% of data: https://sustainablewebdesign.org/calculating-digital-emissions/

I believe this is reflected in the Green Web Foundation's CO2.js as well, correct @mrchrisadams?

Per @gerrymcgovernireland's suggestion to tag waste, we're doing this in the new version of Ecograder by identifying how much of a page's emissions impact is due to uncompressed images, unused code, animations, etc. Happy to discuss details if anyone's interested. Our goal is to launch that a week from today, though work on this will be ongoing.

ldevernay commented 2 years ago

@timfrick : great to have you onboard!

This new feature from Ecograder sounds awesome! I would love to have more info on this.

I somehow missed the article on calculating digital emissions, I will read this.

ldevernay commented 2 years ago

Dear contributors, here is the document containing the outline (WIP) : https://docs.google.com/document/d/1g1ACWRTAzTlcaKKODNASLkXe4zquF-XR5oHh0GfayP8/edit#heading=h.orr5h1m9v3cm You should request edit access since we will be working on this document to write this chapter.

cqueern commented 2 years ago

Hello All.

Most of the impact comes from user devices, mostly because their fabrication is very impactful.

Are we suggesting an analysis and discussion of the sustainability or environmental impacts of hardware used to browse the web? I don't believe that would be in scope as the Archive focuses on how websites are built. We don't collect and have data about the fabrication of user devices (or even user agent data) for us to comment on.

But perhaps I misunderstand your suggestion.

ldevernay commented 2 years ago

This should not be the aim of the analysis (other studies do that pretty well) but we should keep this fact in mind because it is a huge part of the reason why digital services should be as sustainable as possible.

The reasoning here is :

This makes Web Almanac the best place to study if websites today are sustainable and show how to reduce their environmental impact with sobriety and efficiency considerations.

gerrymcgovernireland commented 2 years ago

That’s great to hear about the messages to Tom and Tim

Sounds fascinating that you’re gathering this data on the devices. I read the article from your link. It’s very impressive. I know Caleb has pointed out correctly that the device is out of scope for this project, but I would love to chat with you more about it as a side conversation, if you had time.

From: LaurentDev @.> Sent: Friday 15 April 2022 08:30 To: HTTPArchive/almanac.httparchive.org @.> Cc: gerrymcgovernireland @.>; Mention @.> Subject: Re: [HTTPArchive/almanac.httparchive.org] Sustainability 2022 (Issue #2910)

I sent a message to Tom yesterday and another to Tim this morning, hope they will have some time for this!

On the impact being mostly on devices, I totally agree and that's what the studies mentioned on my previous comments illustrate. This is why, in the company where I currently work, we also gather metrics on real devices. On smartphones, this is a good way to measure the depletion of the battery, which allows us, through an environmental model, to evaluate the global environmental impact of a digital service and the share that is due to device manufacturing. As of today, this is still (unfortunately) a proprietary methodology and model. More details here : https://greenspector.com/en/environmental-footprint-methodology/

— Reply to this email directly, https://github.com/HTTPArchive/almanac.httparchive.org/issues/2910#issuecomment-1099928218 view it on GitHub, or https://github.com/notifications/unsubscribe-auth/AQGRE36R7X7GXA2JECDEEL3VFELI3ANCNFSM5TMP2RQA unsubscribe. You are receiving this because you were mentioned. https://github.com/notifications/beacon/AQGRE36AIBGY3JI35EQ4ADLVFELI3A5CNFSM5TMP2RQKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOIGHZFGQ.gif Message ID: < @.> @.>

gerrymcgovernireland commented 2 years ago

These are good contextual points. I think we can note that heavier and more processing intense pages often contribute to forcing people to upgrade their devices sooner than they would if the pages were lighter and less processing intense. I think this would be good context, without having to go into any research or analysis of specific device impacts.

From: LaurentDev @.> Sent: Friday 15 April 2022 14:50 To: HTTPArchive/almanac.httparchive.org @.> Cc: gerrymcgovernireland @.>; Mention @.> Subject: Re: [HTTPArchive/almanac.httparchive.org] Sustainability 2022 (Issue #2910)

This should not be the aim of the analysis (other studies do that pretty well) but we should keep this fact in mind because it is a huge part of the reason why digital services should be as sustainable as possible.

The reasoning here is :

This makes Web Almanac the best place to study if websites today are sustainable and show how to reduce their environmental impact with sobriety and efficiency considerations.

— Reply to this email directly, view it on GitHub https://github.com/HTTPArchive/almanac.httparchive.org/issues/2910#issuecomment-1100121483 , or unsubscribe https://github.com/notifications/unsubscribe-auth/AQGRE32FW5GVRGSTDNUCOOTVFFXYFANCNFSM5TMP2RQA . You are receiving this because you were mentioned. https://github.com/notifications/beacon/AQGRE32NV5K4W63H36AR6I3VFFXYFA5CNFSM5TMP2RQKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOIGJILCY.gif Message ID: @. @.> >

gerrymcgovernireland commented 2 years ago

Great to see you involved, Tim!

And it would be great to hear more about how you’re identify page waste

From: LaurentDev @.> Sent: Friday 15 April 2022 13:24 To: HTTPArchive/almanac.httparchive.org @.> Cc: gerrymcgovernireland @.>; Mention @.> Subject: Re: [HTTPArchive/almanac.httparchive.org] Sustainability 2022 (Issue #2910)

@timfrick https://github.com/timfrick : great to have you onboard!

This new feature from Ecograder sounds awesome! I would love to have more info on this.

I somehow missed the article on calculating digital emissions, I will read this.

— Reply to this email directly, view it on GitHub https://github.com/HTTPArchive/almanac.httparchive.org/issues/2910#issuecomment-1100074261 , or unsubscribe https://github.com/notifications/unsubscribe-auth/AQGRE37CWTKWZ5BPB3YMRHDVFFNUHANCNFSM5TMP2RQA . You are receiving this because you were mentioned. https://github.com/notifications/beacon/AQGRE3354FTK2DWANURAFMDVFFNUHA5CNFSM5TMP2RQKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOIGI42FI.gif Message ID: @. @.> >

timfrick commented 2 years ago

Happy to chat about that at any time, @gerrymcgovernireland

fershad commented 2 years ago

For the Analysis by Metrics section of this chapter, I feel we could also look at aligning with the CMS, eCommerce, and Jamstack chapters to examine how the different (top??) platforms in those categories perform in terms of sustainability. Could look at median size of pages, and also hosting options (can self-host? does managed hosting use known green web hosts?).

rviscomi commented 2 years ago

FYI we've created a #web-almanac-sustainability channel on Slack if you'd like a higher bandwidth place to brainstorm content.

I'd also encourage everyone to request edit access to the planning doc and iterate on the chapter outline there.

radum commented 2 years ago

This should not be the aim of the analysis (other studies do that pretty well) but we should keep this fact in mind because it is a huge part of the reason why digital services should be as sustainable as possible.

The reasoning here is :

  • Most of the environmental impact of digital comes from manufacturing our devices
  • This gets worse because people around the world own lots of devices and change them too often
  • Why do we change our devices? Most of the time, because they lag
  • How do we prevent that? By making digital service more sober and efficient => sustainability

This makes Web Almanac the best place to study if websites today are sustainable and show how to reduce their environmental impact with sobriety and efficiency considerations.

I think you are making some assumptions here that have no research to back them up. People don't always change devices because of lag. I think the regional factors and many other have an important role. So IMHO this assumption is just speculation.

Making your digital service, in our particular context here the web, more sober and efficient doesn't translate into improved sustainability.

I think we need a proper definition on what sober and efficient actually means. Or what do we mean exactly by sustainability.

@timfrick and the team behind https://sustainablewebdesign.org/calculating-digital-emissions/ are able to calculate a number based on volume of data over the wire. I think this is wrong.

If you build the most performant website with LH scores in 100s and under 2s load time for 3G connections doesn't mean that is sustainable.

Sustainability is the ability to maintain or support a process continuously over time. This is not just about CO2 emissions. It is about everything. You can be performant, but if each change to the digital service requires a full rewrite, that means is not sustainable. And so on, this can be applied to the entire life cycle of a digital service.

If we are to think about CO2 emission, because this is where most people look at when they think about sustainability, I think we are focusing on the wrong thing. As I mentioned above, data volume over the wire doesn't have a direct correlation with energy consumption.

A persons router uses more or less the same amount of energy under heavy traffic vs normal / idle. Even the Anders-Andrae study says that.

From 2020 the improvement of kWh/unit/year for devices is assumed 3% as in [10]. The difference is that Wi-Fi is added to the consumer devices section. Wi-Fi is overestimated in [3] as the Wi-Fi modems electric power use is actually rather independent of handled traffic

There is also a study from Cisco that reaches the same conclusion, volume of data is not a major factor in energy consumption.

There have been articles around how Netflix is killing the planet using the same metric of volume of data over the wire, that have been debunked .

@ldevernay As you said most of the environmental impact of digital comes from manufacturing our devices among many other things.

When thinking about the actual digital service, a website, the device screen is the biggest energy consumer. Which means in a simplistic way that there are only 2 ways to fix it, improve the energy consumption of those screens or keep your users away from them.

First can't happen in a "sustainable way" due to adoption and cost, the other it is in our hands to fix. And this is where web performance is critical. Make it fast and you are one step closer. But if your UX is poor you are negating your performance improvements, as the user spends more time using that digital service to fulfil a task. Fixing the UX also is not the end of the road as there are other factors in play.

Imagine if TikTok was a web app, highly optimised with an amazing web perf grade (images and videos compressed in an ideal world). Would we call that digital service sustainable for the environment (energy consumption to be more precise)? I would say not, because spending hours with your screen ON watching videos drains your battery and so on.

On the same note, if I have a poor performant web site, but all my users power their devices with solar, would you still consider it as performing poorly on a sustainability index? All studies done in the past, don't factor enough how energy production looks like in the future. And when they do they need to guess what will happen. For the first time, wind power eclipsed both coal and nuclear in the U.S for 1 day in March. In the UK is happening more often.

IMHO Sustainability should not be a chapter in itself for web almanc. Perhaps it can be included into various chapters in that particular context.

radum commented 2 years ago

Another example where I think that volume of data over the wire is wrong

@gerrymcgovernireland posted this a while back https://twitter.com/gerrymcgovern/status/1494225017839697920

An analysis of top 1 million websites found 21 million tracking cookies belonging to 1200 companies. Every month, 197 trillion cookies move across this deceitful network, resulting in 11,442 monthly metric tonnes of CO2 emissions. Just cookie CO2. https://carbolytics.org

The assumption here is that if those tracking cookies disappear overnight 11442 metric tonnes of CO2 emission would be gone. This is unrealistic considering the devices that were moving those cookies behind the scenes will still be using the same amount of energy because as it was said before electric power use is actually rather independent of handled traffic.

timfrick commented 2 years ago

@radum, I appreciate reading your perspective. As you noted, it's about everything. That's one of the reasons we included Business Operations and Client & Project Ethos as categories on the Sustainable Web Design site. That doesn't encompass everything, but it's a start.

Also, we noted on the Calculating Digital Emissions post you referenced that there is not yet broad scientific consensus on the kWh/GB approach. We're definitely open to other ways to consider this. The information in that post is based on the most recent studies we could find. We're also updating it as new publicly available information comes out. If it turns out that a different approach reaches broader consensus, we're more than happy to change it.

With that being said, The Green Web Foundation's CO2.js is an attempt to standardize estimates using this approach, even if they are imperfect. Tools like Ecograder, Website Carbon, and EcoPing are all using it moving forward so that if people run a URL through each tool, there's at least consistency in the results. You have to start somewhere...and, given the ticking clock, sooner rather than later is better.

From my perspective, the majority of our clients (those who hire us to build websites and digital products) don't even know that this is a thing. Anything we can do to broaden their awareness of this topic—while also reducing emissions and improving a lot of websites in the process—is good. Tools like those mentioned above can help with this, even if they need to be revised or otherwise updated moving forward.

radum commented 2 years ago

@timfrick Thank you for your reply. I totally understand from where you are coming from in terms of clients and awareness. My main concern is that we are rushing to quantify something that is not possible to quantify, at least for now. And this becomes a marketing tool.

I have met companies and clients that want to be more sustainable with their web presence and ask for help to stick a label. In order to show improvement the tools you mentioned are starting to be more and more mainstream in showing progression, even though the reality could be quite different.

Same goes for CrUX data, I have seen people improving their CWV and validating the work with CrUX, but the reality in the field was quite different as CrUX doesn't reflect ones entire user base.

I understand the need of doing something rather than nothing, but I think this way is not it and will be hard to turn around later when something better appears. And lets be honest, quantifying CO2 emissions and energy consumption for websites is a very hard problem so I am not seeing a solution in the near or distant future. There are way too many variables in play here.

I love how https://www.sustainablewebmanifesto.com/ and part of https://sustainablewebdesign.org/ explain how the sustainable web should be. But the need to add a KPI to it IMHO negates the main goal. The calculators will be taken at face value and the core will be lost in translation. I think it is our duty to educate first that sustainability is not a score from 1 to 100.

I have helped clients with poor results in the tools we mention here to create a design system for their websites and that unlocked a whole new world of possibilities, where the entire process of creating new content didn't need to use loads of systems again and again but composition and reusability fixed that. I can argue that in a way they are more sustainable now.

Web performance (with everything that we can quantify for it) is about accessibility, what we can do to make a digital product accessible for everyone (low end devices, poor network conditions, small screens, small data plans, etc) not about CO2 emissions. We improve images size, page speed, page weight and all that to make it more accessible not more sustainable (especially around energy usage).

Thinking also about data transmission and energy consumption of the equipment that is used for transmission you can say (following the assumptions in the kWh/GB approach) that the more data comes in the more energy will be used and poses a critical problem for the entire infrastructure. But devices that need that infrastructure are coming online at almost an exponential rate and the network still holds. Cell towers are built for coverage and not for capacity and so on.

rviscomi commented 2 years ago

Just hack a kickoff meeting with @ldevernay. One idea that came up was to look into installing the Sustainability plugin for Lighthouse in HTTP Archive. Need to look into what stats that plugin unlocks and whether we're able to derive them from the metadata we already have, and any overhead it would add to the crawl.

cc @pmeenan @tunetheweb FYI

ldevernay commented 2 years ago

@radum : thanks for this insight. Regarding the fact that lagging often leads to buying a new device, I in fact don't have scientific studies or stats but this is based on the fact that digital services are getting bigger and bigger. https://www.greenit.fr/2010/05/24/logiciel-la-cle-de-l-obsolescence-programmee-du-materiel-informatique/ https://tonsky.me/blog/disenchantment/

Hence Wirth's Law : https://en.wikipedia.org/wiki/Wirth%27s_law

As of today, heavy website have an impact on the devices. You should not look only at the page weight but also HTTP requests, battery discharge speed, CPU consumption and so on. The battery alone is a great indicator because heavy websites (not only in size but also in scrolling, animations and such) deplete the battery quickly. The more you charge a battery, the more capacity it loses, which is a nuisance for most users and leading them to change said battery (best case scenario) or changing the device.

As you said, you can't estimate only CO2 because it could lead to some rebound effect (Jevons Paradox) and you should keep in mind that it is really tough (and, as of today, impossible) to map all environmental impacts. Only LCAs can do something almost exhaustive. But you can rely on emission factors from said LCAs and other studies to build an environmental model which results in multiple factors (such as water consumption, field occupation, metal consumption, etc). The priority is to set a group of proxies and standards and make sure everyone talks about the same thing so that calculations for different websites, from different persons, remain comparable. That's why there is an AFNOR norm planned to be released tomorrow and a W3C working group being created. This is also the reason why you need to combine best practices (with resources, priorities and enough info to prioritize them) and measurements (to make sure you're going in the right direction). Which often goes with monitoring and establishing an environmental budget.

I don't think you should distinguish performance / accessibility and sustainability.

In my opinion, Sustainability is a whole set of domains that intersect to design digital services in order not to harm people and the planet. You get benefits in accessibility from performance and sustainability. You also get benefits for sustainability from accessibility (making the user journey smoother results on less time spent on the website, for example). But you should also consider privacy (cookies, analytics, trackers and such have an environmental impact), security, the attention economy (dark patterns have an impact on environment but also often on performance and privacy) and ethics.

For a more precise definition, sustainability focuses on designing digital services that are both sober (focused on user needs, not abusing their attention or cognition) and efficient (this is where sustainability intersects with performance, even if both domains disagree on some topics).

@gerrymcgovernireland : I'm sending you an email, thanks!

@rviscomi : I think you should check on @mrchrisadams for more info on this plugin.

Djohn12 commented 2 years ago

Hi there,

really interested in giving a hand but not sure where to start nor if my technical background and English level best fit Reviewer or Editor role.

I see no Editor has been designated yet, let me know if you'd like me to fill in the blank :)

tunetheweb commented 2 years ago

To explain the difference between these roles:

Does that help you pick which role you'd be interested in @Djohn12 ?

pmeenan commented 2 years ago

@rviscomi @tunetheweb Looks like the Sustainability plugin mostly just calls the Green Web API. Rather than hammering on the API for every site we crawl, it would probably be better to reach out to the Green Web Foundation directly and integrate with the dataset and have them contribute to the chapter/methodology.

ldevernay commented 2 years ago

@pmeenan : I think you're right on that and having @mrchrisadams on board might help.

It might be more useful to crawl while checking some best practices, like this cli tool : https://github.com/cnumr/GreenIT-Analysis-cli/ The new version of Ecograder might prove useful too.

In the end, this could be really close to what Lighthouse does. What remains to be seen is what global metrics we can/should measure/calculate.

So, yes, as of today, we should stick to using the Green Web API to gather data for the chapter and maybe take some time to think about what could be done for sustainability in Lighthouse and/or Chrome.

@Djohn12 : great to see you here! I think you should join as a reviewer. @timfrick, @gerrymcgovernireland, @fershad and others : don't hesitate to join as authors. I think this will really be a collective work.

radum commented 2 years ago

@ldevernay thank you for your reply.

From the links you provided one is just persons opinion (valid or not, is not a scientific study, and the results vary highly based on region, population, period and so on.) The other one is from 2010 (event if it has updates to 2020), and the actual research has no correlation with CO2 emission. The fact that it sounds ridiculous that you need huge amount of disk space for an app to write text doesn't really mean anything as there was no analysis what goes into that disk space waste. So I am sorry but I don't see how this relates to this chapter.

As of today, heavy website have an impact on the devices. You should not look only at the page weight but also HTTP requests, battery discharge speed, CPU consumption and so on. The battery alone is a great indicator because heavy websites (not only in size but also in scrolling, animations and such) deplete the battery quickly. The more you charge a battery, the more capacity it loses, which is a nuisance for most users and leading them to change said battery (best case scenario) or changing the device.

You don't know that, because you can't measure it, and our sane conclusions might have an order of magnitude that is too high as it has been proven many times in the past. Just look at how Greenframe concludes that their score is way lower than other tools when considering more than just data over the wire.

As you said, you can't estimate only CO2 because it could lead to some rebound effect (Jevons Paradox) and you should keep in mind that it is really tough (and, as of today, impossible) to map all environmental impacts. Only LCAs can do something almost exhaustive.

This proves my point. As for to rely on emission factors from said LCAs, there are none (AFAIK) that do that for websites.

The priority is to set a group of proxies and standards and make sure everyone talks about the same thing so that calculations for different websites, from different persons, remain comparable.

Talking about the same thing and having a unified front around the topic is different vs enabling people and companies, without enough know how, to get instant gratification that they are making improvements on certain areas where in fact the reality is quite different. I am seeing this more and more where getting a label or seeing a score that is green and not red becomes more important than the actual thing we are trying to improve. And in order to achieve that tools and methods are cherry picked because they are easy to use, but with fundamental problems around the methodology and results.

In my opinion, Sustainability is a whole set of domains that intersect to design digital services in order not to harm people and the planet.

I agree and said the same thing. That is why I am raising my concerns around this topic for the almanac.

I don't want to turn this issue into a debate (especially a toxic one) so I will stop with this final comment.

My point is simple. I think we have a responsibility with the almanac to tell and show data that can drive change. The entire world uses this data and its analysis insights and trusts it is accurate. IMHO this chapter should tread carefully with what will show and what will tell its audience. Personal feelings and opinions should not influence the outcome.

mrchrisadams commented 2 years ago

Hi folks. I can contribute more fully next week, but in the meantime, I hope these might be useful.

Data for inducing new embedded emissions from device changes

The best paper with data for how design changes can impact hardware usage or embedded emissions is probably this paper from the ACM limits conference last year

https://limits.pubpub.org/pub/rn94aofn/release/2

In this paper, we examine how well online learning platforms function on older browser software and device hardware. We then perform an analysis over several years of data from a learning management system used at the authors’ university campus. We find that software and hardware become obsolete within roughly a four-year period, meaning students are likely to be required to purchase a new smartphone or laptop during their college careers. In reality, these “obsolete” devices are just as capable as they were when they were new.

Finding a model

I think it's worth being prepared that the first few years of tracking this you'll have some wild changes for these numbers, simply because even the academic literature that has passed peer review is all over the place, and company carbon emissions figures are changing wildly they decide to include in their system boundary.

There's a thread on linked in here with some network technicians weighing on this, with links to underlying data.

This report we did a few months back shows Facebook's reported carbon emissions varying by 3-4 fold from one year to the next, largely based on changes to how they calculate their footprint internally.

https://www.thegreenwebfoundation.org/publications/report-fog-of-enactment/#appendix

I'd argue it's better to have a models that we know are imprecise and clear about their short comings, but we are improving in the open, especially when you take this the multi year view.

Have a lovely weekend. C 👍

Djohn12 commented 2 years ago

@ldevernay @tunetheweb thanks for your feedback !

Let's tag me as a Reviewer then, I'll try to be up to the task ;)

I'll start by reading the documentation and I'll get back to you as soon as possible.

Wishing you all a nice week-end

ldevernay commented 2 years ago

@Djohn12 : I added you on this issue. Don't forget to add yourself to the google doc and to join #web-almanac-sustainability on Slack: https://docs.google.com/document/d/1g1ACWRTAzTlcaKKODNASLkXe4zquF-XR5oHh0GfayP8/edit#heading=h.orr5h1m9v3cm

rockeynebhwani commented 2 years ago

@ldevernay - Another idea for the chapter ...

Wappalzyer is tracking some solutions which are related to sustainability. Some examples are -

I am sure there are many other approaches businesses are taking and it may be possible to track many using Wappalyzer...

Maybe we should look at new category called 'Sustainability' in Wappalyzer and see what we get.

This can be an interesting data point for this chapter. Let me know what you think

ldevernay commented 2 years ago

@rockeynebhwani : wappalyzer is a great tool but it's difficult to directly link technologies such as Recommerce to web sustainability.They are great for general sustainability but that doesn't necessarily mean the websites using them are designed for web sustainability. It could be interesting to see if such tools follow best practices of sustainable design for the web.

Wappalyzer stats could be useful for the CMS/JAMSTACK section, for example (@fershad ).

BTW, it might be interesting to look at the implementation of analytics (Google Analytics) and third-party, since these types of tools have an environmental impact, some more than others (@fershad , @gerrymcgovernireland ).

gerrymcgovernireland commented 2 years ago

Indeed, analytics can be very heavy, particularly in media sites, often making up 75% of the weight. There’s often huge waste too with lots and lots of scripts hanging around for old marketing campaigns, etc. And tracking is energy intense. Here’s a good energy efficient analytics system

https://withcabin.com/

From: LaurentDev @.> Sent: Monday 25 April 2022 10:38 To: HTTPArchive/almanac.httparchive.org @.> Cc: gerrymcgovernireland @.>; Mention @.> Subject: Re: [HTTPArchive/almanac.httparchive.org] Sustainability 2022 (Issue #2910)

@rockeynebhwani https://github.com/rockeynebhwani : wappalyzer is a great tool but it's difficult to directly link technologies such as Recommerce to web sustainability.They are great for general sustainability but that doesn't necessarily mean the websites using them are designed for web sustainability. It could be interesting to see if such tools follow best practices of sustainable design for the web.

Wappalyzer stats could be useful for the CMS/JAMSTACK section, for example @.*** https://github.com/fershad ).

BTW, it might be interesting to look at the implementation of analytics (Google Analytics) and third-party, since these types of tools have an environmental impact, some more than others @.*** https://github.com/fershad , @gerrymcgovernireland https://github.com/gerrymcgovernireland ).

— Reply to this email directly, view it on GitHub https://github.com/HTTPArchive/almanac.httparchive.org/issues/2910#issuecomment-1108330471 , or unsubscribe https://github.com/notifications/unsubscribe-auth/AQGRE34GJJ3WB7XXXCOLMI3VGZRY5ANCNFSM5TMP2RQA . You are receiving this because you were mentioned. https://github.com/notifications/beacon/AQGRE332PJDAHANTGZQETT3VGZRY5A5CNFSM5TMP2RQKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOIIH4PZY.gif Message ID: @. @.> >

ldevernay commented 2 years ago

FYI, I still have on my TODO list a benchmark of various analytics solutions.

tpgreenwood commented 2 years ago

Hi, thanks for the invitation @ldevernay - I'm just putting my name forward here that I may be able to assist as an author and/or reviewer. I am very short of time at the moment so I can't promise my availability, but would like to help if I can.

ldevernay commented 2 years ago

@mrchrisadams @cqueern @Djohn12 @tpgreenwood : Don't forget to add yourself to the google doc and to join #web-almanac-sustainability on Slack: https://docs.google.com/document/d/1g1ACWRTAzTlcaKKODNASLkXe4zquF-XR5oHh0GfayP8/edit#heading=h.orr5h1m9v3cm

(for those of you who want to join as authors, I don't think it is necessary to join as a reviewer too, it is implied)

We seem to have a full team, thanks everyone! As soon as everyone has joined Slack and updated the google doc, we will work on the outline.

(I will be off for a few days starting today + business trip next week so my response time might be a little disrupted)

hanopcan commented 2 years ago

Hi folks, I'd like to offer some time as a reviewer if I'm not too late.

ldevernay commented 2 years ago

@hanopcan : this would be great! Please join the Slack channel and add your name on the google doc!

ldevernay commented 2 years ago

Hi, we're having a kickoff meeting next week to get to know each other and work on the outline. If you haven't yet, don't forget to join the slack channel and add you email address on the google doc until then, so that I can invite you!

camcash17 commented 2 years ago

If possible, @4upz and I would like to volunteer as Analysts.

tunetheweb commented 2 years ago

Excellent. Invited you to the project there.

tunetheweb commented 2 years ago

@ldevernay I think we're good to tick off Milestone 1 in the initial comment? We've a good solid outline there, and although it may change still I think we've enough to say we've reached the milestone.

ldevernay commented 2 years ago

That's true, thanks @tunetheweb ! I will take some time today or tomorrow to clean the google doc and make sure the global structure is appealing to readers and that the content for every section seems clear. We can then think about who wants to write what.