Hacker News new | past | comments | ask | show | jobs | submit login
Google’s Environmental Report (environment.google)
282 points by chrisseldo on June 5, 2017 | hide | past | web | favorite | 144 comments

there's some pretty shady reporting on the summary page: https://environment.google/projects/environmental-report-201...

9.2M sqft LEED certified - they display percentages for most stats, but only an absolute for this one? going to make me dig to find their total office square footage to find out what this means?

Then the next stat, 31% LEED Platinum. oh good, a percentage. except it's not, it's a percentage of their LEED certified buildings.

or "100% landfill diversion, in 6 of our data centres". surely google has more than 6 data centres, but let's only look at the environmental stats of the ones that are doing a good job. [The stat is 84% across all their datacentres.]

"100% renewable energy". sounds good, except oh no, it's "we will achieve". What's the percentage right now? [it was 44% in 2015, the last year they report.]

[edit - skimmed the report, added in some numbers to my comment] The report itself appears to be mostly the same marketing fluff, until you get down to the "environmental data" appendix on page 66, which actually looks like a pretty solid report of their environmental footprint. And it looks like they're doing a decent job, so i'm not sure why the real numbers are buried so deeply.

More importantly, LEED is bullshit.

LEED is a private entity that you pay to get LEED certification based on your planned designs. Once you have certification, you could build literally anything else. If you buy a LEED certified building, you could convert it to be heated and powered by burning tires and endangered animals- still LEED Certified.

There is no decertification process, because that would mean spending money to reduce the number of buildings with the very-public name on it. Why would LEED ever want that?

Oh my goodness. I work in the architecture industry at a firm that does a great number of LEED projects, and there is not a shred of truth in this claim.

LEED's own website seems to say it's true: http://www.usgbc.org/help/will-my-project-be-decertified-if-...

Yes, it's true there's no LEED "decertification" process, but everything else the parent comment claimed or implied is false, including the implication here that "LEED decertificaiton" is a thing that's needed due to any sort of prevalent LEED fraud.

Let's fact check the post:

LEED is a private entity... true

that you pay to get LEED certification... true

based on your planned designs... true

Once you have certification, you could build literally anything else... true

If you buy a LEED certified building, you could convert it to be heated and powered by burning tires and endangered animals- still LEED Certified... speculative

There is no decertification process,... true

because that would mean spending money to reduce the number of buildings with the very-public name on it.... speculative

The parent is right to call me out, I'm a bit of a dickhead when I get onto topics like this. And you're also right in that some of my claims are speculative.

I don't think anyone has tried the 'burning tires and endangered animals' move quite yet. I should give it a try!

I look forward to your follow-up post.

Okay. I'm going to respond in more depth to the original nay-sayer here, but will use your reply as a jumping-off point as you've succinctly summarized it. Hope you don't mind.

> LEED is a private entity...

And? The AIA, which administers licensing exams for architects, is also a private entity. What is the actual criticism here? That the US Green Building Council (the administrator of LEED) should be a government program? That would have given them the delightful opportunity to be defunded by Trump right now...I'm good with the current situation.

> that you pay to get LEED certification...

If the implication here is that it is a pay-to-play system, that's patently false. Yes, you pay for the certification - if you demonstrate meeting the lengthy list of requirements for the certification you are seeking.

>based on your planned designs... >Once you have certification, you could build literally anything else...

This is where this so-called critique really goes off the rails. For clarity, LEED certifies architectural projects. Those outside the industry may not understand this, but the only thing architectural firms produce is plans. We call them Construction Documents. Architecture firms do not create buildings, unless they are a design-build firm (a small minority). The post I was replying to here was clearly implying that these 'plans' are just some non-binding byproduct of the process with no real impact on what gets built, which is the opposite of the truth. the plans produced by an architectural firm, not the built product, are what is subject to building code approval and are the artifact which carries all legal liability. LEED criteria direct the content of these plans. The accusation that firms are or could be putting 'LEED-y things' in their plans and then building something else is preposterous and baseless. This would be fraud and malpractice. Please provide some evidence if you're going to continue making this claim.

>If you buy a LEED certified building, you could convert it to be heated and powered by burning tires and endangered animals- still LEED Certified...

This is quite true, and admittedly is not a use-case traditional LEED is designed to address. But again, it's your onus to provide some examples of the system being abused in this way. The certification is intended to apply to the design documents and subsequent construction project; casually it is extended to a 'building', but of course buildings are things which can change all the time and even are ill-defined (you can get LEED certification for a campus development project, for example, which does not necessarily imply all buildings in that development are individually LEED-certified projects). The plaque which is provided for a LEED project and granted for use in marketing the project has the year of certification on it, and generally the certification is sought by developers or owners as a marketing effort at the time of their construction effort. I'm not generally aware of any companies purchasing LEED-certified buildings so they can use the certification for their marketing benefit, even leaving aside how faithfully they maintain the LEED-applicable elements. In fact I believe the first ever LEED-certified speculative property development (meaning a building explicitly built with the intent of selling to commercial occupants other than the developer) was just certified this past year.

>There is no decertification process,...

Again, the only reason a de-certification process would seem prudent is if the problem of people buying LEED-certified properties, "de-LEEDing" them, and continuing to promote their LEED-certification for marketing purposes, is a real thing. Just like voter fraud, you'll need to provide some evidence before I'm going to be convinced USGBC needs to start worrying about this. While they are not "de-certifying" legacy projects, one of the new components of the latest standard is a 'Performance Score' (which is currently optional but heavily encouraged), which measures exactly the kind of ongoing metrics you seem to be concerned about, and gives you a fancy digital status board to display these stats in your lobby. [https://www.leedon.io/faq.html]

>because that would mean spending money to reduce the number of buildings with the very-public name on it....

What are you suggesting to be the ulterior motives of the USGBC here? Because I was able to divine the ulterior motives of your LEED scare site in the course of like, three clicks. I'll summarize how I see it as someone who has a bit of familiarity with its history. Look, LEED is a deeply-flawed system. This is a topic discussed ad-naseum in the industry. But what's the alternative you're proposing? Because the system LEED replaced was...nothing at all. Back in the good old days, "green building" was mostly an idea made of marshmallow fluff and good intentions, with some grass on a roof here and there. Aside from the true pioneers who were generally working in obscurity, the sum total of a "green" architectural project of any significance generally was some planters and a showy geothermal system that didn't actually work, maybe some dubiously-"sustainable" material with no actual credentials used in a prominent application. Greenwashing was pretty much all that ever happened, because large clients by and large were only interested in "green" for its marketing value and not willing to spend more than the bare minimum on it. Now, I'm not saying the system we have today is too much different! But what's significant is it provides a framework that practitioners can defer to. Now when a client says "we're interested in green", we can direct them to LEED, thanks to its brand recognition, and LEED comes with rules. Rules you have to spend to meet, and rules which are influencing the products vendors put on the market. Not to say exceptions don't get made (they have a system in place to petition for them), and not to say the point system is particularly equitable, and not to say that even the most aggressive LEED certification is anywhere near where we need to be. But it's providing a framework which is pushing the industry forward in observable ways, on a mulititude of levels. And most admirably, they are aggressively (in architectural timelines) deprecating prior certifications and rolling out new, more stringent ones, including things like the ongoing performance metrics I mentioned above.

I don't know if LEED is the ultimate answer. There are certainly better standards out there, gaining traction (Green Globes, Passive House are a couple notable ones). But LEED is on the ground, with tens of thousands of certified projects, and is pushing its standards higher every year. I'm not going to sneeze at that.

Thanks for explaining this. Until this comment I hadn't understood why LEED should be expected to have a real effect.

Really? I had heard that from a few sources, and done some cursory research that seemed to confirm it.

Can you provide sources to the contrary? I'd rather admit to being wrong and then become right than go around sounding like I know shit when I don't.

(My main source is my wife, who's a civil engineer for the city approving building construction. She has strong opinions on LEED because she's had to approve buildings that met city standards but basically fucked off with their LEED promises).

Well, I guess that the burden is on you to present sources to your claim :)

As another reply said, their own website[0] says there's no decertification process: "No, projects will not be decertified for performance."

There's the fact that many LEED certified buildings are less energy efficient than other buildings, even after spending hundreds of thousands or more to get their LEED certification[1][2].

There's even a LEED Exposed website dedicated to calling out their organization[3]. Primarily, the fact that they are a non-government organization whose rules many governments say buildings must follow.

There's lots of reasons to be suspicious of LEED, and I'd love to hear counter-arguments.

[0] http://www.usgbc.org/help/will-my-project-be-decertified-if-... [1] https://www.forbes.com/sites/realspin/2014/04/30/leed-certif... [2]http://www.washingtonexaminer.com/exography-worst-of-the-wor... [3]http://www.leedexposed.com/what-is-leed/

This isn't a counter-argument, just an observation:

"LEED Exposed" is run by the Environmental Policy Alliance[1], which is "devoted to uncovering the funding and hidden agendas behind environmental activist groups."

It looks like the Environmental Policy Alliance is a shell organization run by Berman & Co.[2][3], a PR firm for everybody's favorite lineup of corporate bad guys: cigarettes, soda and fast food, and (presumably) some company that benefits from lessened pressure for environmental design.

The Forbes source you gave is written by someone at Berman & Company.

I don't want to make any particular claims about LEED or the USGBC, because I just don't know enough about them. However, the origins of the criticisms against them suggest that there's more at play than just LEED being a bad or ineffective standard.

[1]: http://environmentalpolicyalliance.org/

[2]: http://www.sourcewatch.org/index.php/Environmental_Policy_Al...

[3]: https://en.wikipedia.org/wiki/Berman_and_Company

Great points.

I think overall, I'd prefer to see proper government regulated standards. The EPA's energy star stuff has worked pretty well at enforcing home appliances meet certain criteria for efficiency.

There needs to be some bite behind these rules, and not just good intentions.

I agree; I think that government standards (and enforcement) for efficient building design would go a long way.

That being said, Berman & Co. (or any number of opaque lobbying firms) don't stop at LEED - it's just low hanging fruit. They're also running "EPA Facts" [1], "Green Decoys" [2], and so forth.

My worry is that with LEED out of the way, the EPA's replacement will become the new target. With Scott Pruitt in charge, that may be easier than ever.

[1]: https://epafacts.com/

[2]: https://www.greendecoys.com/

Who's claim? The commentator or in Google's?

The person I replied to. Was that unclear?

Thanks, it was unclear.

It seems your claim is half true.

"The idea that there is this new thing call decertification is inaccurate," says Scot Horst, USGBC's senior vice president for LEED. "The way LEED works is we have a rating system; you send us information about your project, and we certify to that. But let's say that there was someone out there who lied about the prerequisite information or unintentionally provided inaccurate information. We have always had a policy to go back and say this wasn’t what it was represented to be. That is nothing new."

"The only way you would ever lose a certification that you've already been awarded is if you didn't meet the prerequisites of the system; USGBC currently does not evaluate or monitor the ongoing operations of a building, Horst adds."


So by my reading, you could lose certification for not building to your plans.

Why am I not surprised. Funnily enough, I work in the same building as the EPA, and they advertise their LEED cert. SMH

It could be that their office's interior design and construction was certified LEED as opposed to the building itself. There are different types of LEED certification available so maybe this is the case in your example?

Big respect for Google!

Creating that site is a really cool move, and it will encourage other people to do the same.

I especially like the fact that they'll mount sensors for air quality on Street View cars. And the fact that they're using almost 100% renewable energy.

This is what I call leading by example.

> I especially like the fact that they'll mount sensors for air quality on Street View cars

There was a really good article on Ars about using this to detect natural gas leaks just a couple weeks ago:


Lots of companies have been publishing sustainability reports for years now (our first was published in 2010) -- it's not unique to Google. We're on track to hit 100% renewable by 2025. Here's our site: https://www.bloomberg.com/bcause/planet/

There is a great need to hurry up with enthusing and glamorising the most sustainable human consumables and activities and interests, by the truly effective means of our Age (mass commercial and public advertising) This also means stigmatising and depreciating many kinds of commerce and spending which have to date been marketed and defended as rightful livelihoods.

It is past time to be just tending our own very fine gardens. We need to make it difficult for each other to pollute and mis-consume, like we did with smoking.

I agree. As soon as someone shares a number, such as with diversity, it now opens up the questions of why others yet do not or will not.

The office component is such b.s. that it throws the rest of the report into doubt. While solar panels and shuttles are small gestures in the right direction, Google continues to make the wrong large decisions by putting their offices in the middle of nowhere, miles from any real place, surrounded by acres of parking. The sustainable thing would be to put their buildings in cities and adjacent to mass transit so employees can walk and ride to work from their nearby homes.

Disclaimer: I work at Google but the following comments are just based on personal observation and public information.

I think you're talking about the Mountain View campus and I agree it seems superficially to be an example of what you're talking about... I imagine with the benefit of hindsight and with modern (post 2000s :) sensibilities in mind both Google and Facebook would have tilted more heavily towards building up their SF presences back when that would have still been plausible/affordable.

However the parking restrictions are actually enforced by municipal regulation; I believe Google and many other companies (including my previous employer, VMware) would prefer to build campuses with much less parking but they are required to have a certain ratio of parking spaces to employees due to these antiquated regulations.

Meanwhile the other Google campuses are in fact located in large cities, though I believe transit access (Seattle?) is still an issue. One of the main reasons I choose to live in New York and work at the New York office is because I am a huge fan of the set up here where the office is centrally located and highly transit accessible.

I think it's more that the fact of having a giant, expensive separate campus is itself not a particularly progressive situation, regulations aside. Just a few common sense examples of the problems with this approach: massive, custom-built complexes are very hard to re-use if and when search advertising revenue declines; tax revenues go to random places that don't need them, instead of cities that could use rich companies headquartered there to better support public services; many employees still have to do business in big cities, so there is a massive amount of unnecessary travel to and from cities and international airports; etc. The issue is "campuses" as much as it has anything specific to do with parking regulations, which I assume all of these companies knew about when they made the decision to build the campuses.

Yep, I agree with you to an extent but I was also using "campuses" and "offices" interchangeably.

The Google New York offices follow exactly the model that you are suggesting.

Agreed! Here's hoping they cancel the Mountainview expansion and focus on replicating the New York approach.

Where would you locate an office for 20,000 people in Northern CA in a city? SF? You'd make housing in the city much worse, and you'd be hard pressed to find the land or permits for it.

Salesforce somehow manages to house 7000 of their employees in San Francisco, and Wells Fargo somehow seats 8000 of theirs. Nobody is suggesting that Google should move all of their people to SF, but I am suggesting they should consolidate into larger offices in the centers of principal cities: San Francisco, San Jose, and Oakland.

And most importantly, outside the Bay Area.

I don't think it's viable to encourage companies to do this unless public transport and other services like schools are improved significantly in these cities. San Francisco isn't NYC.

The "Mountain View" campus now sprawls from Sunnyvale to Redwood Shores. It's a 20-mile bike ride from one to the other. The Redwood Shores office is so unabashedly exurban that you'd have to walk 3.5 miles to the nearest public library, 3.9 miles to the nearest coffee shop, and 4.5 miles to the nearest point of public transportation. These are Google's newest offices, reflecting their current thinking on "sustainability": everybody has to drive everywhere.


There's a ton wrong with what you've just said:

- First, the Pac Shores offices that Google acquired is in Redwood City, not Redwood shores, which is further to the north. It's not a 'continuation' of their campus but an entirely new campus.

- Second, Pac Shores was built in the middle of an industrially zoned area on a port. The massive salt ponds separating Pac Shores from RWC have had different proposals to further urbanize the area but they've been repeatedly shot down. The area is developing, give it time.

- Lastly, there are bicycle, walking paths, and caltrain shuttles going into downtown RWC from Pac shores, so it's easily accessible to the public library or a coffee shop if you want to go (I recommend Bliss coffee).

What's the point?

It would not have been possible for google to put their headquarters in SF. Obtaining that much office space was difficult enough in the middle of nowhere. It would have been completely impossible to do it in SF. The transit impact would have been catastrophic, as many (if not most) employees at google HQ do not live and do not wish to live in SF. They can't all drive there, they can't all ride caltrain (which is basically at or close to max capacity already), and no one seems to want them running even more shuttles into SF.

Yes, the valley is suffering greatly due to 50+ years of horrible planning. No, google hasn't obstinately made a socially harmful decision in the presence of better alternatives.

Google's new headquarters[0] to be built in downtown London is being built parallel to the platform at King's Cross Station. It will have parking space for 686 bikes but only four parking spots for cars.

[0] http://www.marketwatch.com/story/google-submits-landscraper-...

Yes, and their latest Zurich office is also adjacent to the main station of that city. Clearly there is a schism between USA Google and Europe Google as regards land use.


No there isn't. Mountain View is the exception. All of the other big Google offices in the US (New York City, Seattle, San Francisco) are in the heart of downtown cores with great transportation. I'm in the NYC office and have an entrance into the subway from inside the office that goes to the A, C, E, and L lines. The 1, 2, and 3 lines are a block away. We have three bike rooms and hundreds of daily bike commuters (me included). I don't know of anyone who drives to work, and why would you? That's crazy to do in Manhattan.

We have been struggling to use public transport as a country for at least 5 decades. It works in some areas and doesn't work in others. Seeing some of the failures I don't ever see public transit becoming the only (or at least majority of) transit which is what is really required for the kind of sustainability you are talking about.

Wouldn't lots of Electric Vehicles be just about as sustainable as that? And it wouldn't require changing everything about how transit works in all the places public transit doesn't work.

I live in Omaha, so let me use that as an example. Omaha has about 5x the surface area of Manhattan but only about 1m people (Manhattan has about 5m I think). Getting from anywhere to anywhere takes about 20 minutes (most take more like 10), because we have a huge amount interstate style freeways in town. Rush hour adds maybe 5 minutes to that for my current commute for an upper limit for any trip of 25m.

It is a huge waste of gas that we all use our cars, but there simply isn't the population density to make buses work at scale. There aren't enough hubs of activity that would cause a subway to make sense. Even with public transit for the city we still have to deal with all the farmers near the city so we would still need parking accommodations. In order to get anything like the level of convenience cars provide there would need to be a huge amount of them, with 15 minute car rides a bus cannot compete on convenience.

But electric vehicles combined with all the wind farms we have would make it sustainable and not require changing the way a million people live their lives for no benefit. And it looks like EVs are coming in the next decade or so. Waiting just looks practical, our alternative is to rip up our streets move our buildings closer together and hope that a public transport system will make it worth while.

Public Transit is hard to sell in Nebraska the state that invented arbor day, because it is simply impractical. Try selling public transit in Texas where they actively love oil and actively hate the environment (I would too if I had to giant spiders, scorpions and solifuges that are native to Texas).

The problem is that eventually those freeways will get clogged at rush hour unless you continually expand them, which is crazy expensive if you need to rebuild dozens of bridges etc, and suffers from the laws of diminishing returns (adding a 5th lane to a 4 lane high way does not result in a 25% capacity gain).

I'd be surprised in the long run if even Omaha isn't susceptible to this, and electric vehicles don't really do much to help this (perhaps autonomous driving may allow a small %age increase but I am highly sceptical past that).

Houston and Dallas both have fast growing light rail systems, and there is the Texas Central railway project to link the cities with high speed rail so I don't think they're a perfect example - they also have a massive wind power boom going on right now.

> (perhaps autonomous driving may allow a small %age increase but I am highly sceptical past that).

I'm imagining in the future that some roads will be reserved for automated use only except in emergencies. At that point, you greatly increase the trust in individual agents and planning can be communicated ("I plan to move 3 lanes to the right in 1.5 miles, FYI"). Average speed increases, "turbulence" reduces, which improves throughput and speed consistency though doesn't help with exit bottlenecks.

That said, you're right. The law of diminishing returns continues to apply even if it kicks the can down the road a bit (i.e. adding a 5th lane results in a 10% capacity gain without SDVs, 21% gain with SDVs -- numbers are completely made up).

> The problem is that eventually those freeways will get clogged at rush hour unless you continually expand them

Only if density increases. It remains constant because Omaha (and many midwest cities) just gets bigger.

Which offices? Most of the google offices I know about are right in city center.


I work at Google in the Seattle office which is, as you say, in a city close to mass transit in a very walkable area. (My walk to work is about 25 minutes if I don't take a bus.)

Instead of accolades, many many Seattle residents complain now that we are driving up real estate prices and pushing people out of the cite because we all want to live close to where we work.

You can't win.

This seems to be a mountain view exception.

From a quick google search, it appears 20,000 of Google's 60,000 or so employees are headquartered in Mountain View, and it is widely considered the company's "headquarters." I'm not sure it's apt to think of it as an exception. And it's in the process of being rebuilt, which can be understood as a doubling down on this approach.

Perhaps. However, calichoochoo implied this at least true of one more office, as it "continues to make the wrong large decisions by putting their offices in the middle of nowhere, miles from any real place, surrounded by acres of parking". This seems to only apply to mountain view.

FWIW I agree; I can't think of a place I'd like to work less than the bay area. There's no good option for living there, let alone commuting.

So it's better to pull an Apple and completely cannibalize a city by building a giant spaceship in the middle?

For smaller offices it makes sense to have them in the middle of the city, for something as large as their Mountain View office, it'd be a huge mess inside a city for someone of that scale.

And you're forgetting the fact that the headquarter has been growing naturally over the years, so if they had indeed started in a city, they would've probably had a really hard time expanding.

How does a data center run fully renewable, do they have huge battery installations or is it one of those "credits" things where, their consumption is compensated by production equal amount of solar and wind contracts they signed up to?


> We were one of the first corporations to create large-scale, long-term contracts to buy renewable energy directly; we signed our first agreement to purchase all the electricity from a 114-megawatt wind farm in Iowa, in 2010. Today, we are the world’s largest corporate buyer of renewable power, with commitments reaching 2.6 gigawatts (2,600 megawatts) of wind and solar energy.

Still don't get how that works. A 114MW wind farm almost never produces 114MW. In fact, it's output is fairly unpredictable, and can swing from 10MW to 90MW and back to 20MW over the course of 30-40 minutes.

When a company that requires high levels of constant power "buys" that power, what are they buying, exactly? Because they're still just as reliant on nuclear/coal/natural gas as they were before they bought that power. Are they just signing up to pay a higher rate for those 114MW of power they get from the grid?

They are guaranteeing for that producer their power is purchased at x$/whatever for 20 years. Massively lowers the risk profile for the wind farm, and lets them focus on potentially expanding. Google is basically acting as a subsidizer of renewable energy here. So while yes they rely on the grid, they're a market force for increasing the percent of the power produced on the grid to be renewable.

Every MW produced by the renewable is a MW not needed to be produced by natural gas/coal

> Every MW produced by the renewable is a MW not needed to be produced by natural gas/coal

This should be expressed in units of energy, not power, i.e. MWh.

I guess that makes more sense. It's more an act of corporate goodwill on their part, helping to build a certain level of renewable energy.

Partly. But it's also an act of risk hedging. By locking in prices for 20 years, the buyer is insulated from rising prices and the generator is insulated from falling prices.

They are not "just as reliant as before", they are only reliant on gas/coal/whatever when the wind is having a dip. It is clearer when you look at the energy consumed in joules: A 114 MW datacenter consumes 114* 10^6 * 3600* 24 Joules per day. The renewable power generator that Google has a contract with commits itself to putting that much Joules into the grid every day. When the wind is out, nonrenewables fill up the slack, and when there are high winds, some of the customers that would have otherwise been supplied from coal plants get wind energy instead.

On average though, 114 MW is pulled out of the grid by the datacenter and 114 MW is put in by the renewable energy company. The rest of the grid, bith producers _and consumers_, act as a battery for when the output of the renewables is either higher or lower than the contracted amount for the datacaenter.

> "On average though, 114 MW is pulled out of the grid by the datacenter and 114 MW is put in by the renewable energy company."

That's simply not the case. Wind farms fluctuate greatly and produce on average no where near their max output.

> "they are only reliant on gas/coal/whatever when the wind is having a dip"

Which is all the time. I've worked on this problem, there isn't some constant output from wind farms.

This is at best a tangential point. If the wind farm is 114MW with a 10% capacity factor, his point is still correct.

Yes, they still rely on coal and natural gas to smooth the load, but his point was that by subsidizing the wind power, they still reduce emissions the same amount as if they didn't.

Obviously, this strategy only works when renewable penetration is fairly low, once you start having oversupplies then you no longer have 100% efficiency with this strategy.

Would it make you happy if they had written "On average though, x MW is pulled out of the grid by the datacenter and x MW is put in by the renewable energy company, where x is some value < 114 MW"? Sheesh. The point being made is pretty clear.

Presumably they pay more money for the renewable energy, overall increasing the demand for renewables and decreasing demand for coal.

The data centers don't run on renewable. They just purchase the equivalent in renewable energy. In theory it results in reduced emissions, but it provides a misleading view into how viable renewable energy is.

It's "one of those credit things." They sign power purchase agreements, typically 20-year terms with a fixed price per MWh, with generators of renewable power. Both entities are typically attached to the grid, without a direct line between them. The buyer contracts to buy an equal amount of power from the grid that the seller puts onto the grid. The actual timing of generation and consumption does not need to be aligned.

The contrast of the text of those articles is absolutely abysmal.

Does the fact that they released this on WWDC day, a high traffic tech news day, suggest there may be something in here that they're hoping gets overlooked?

They are linking to it directly from the www.google.com homepage, complete with a colorful custom Google animation. If they're trying to bury this report, they're doing a very bad job of it.

I'd imaging this is more a response to Trump pulling out of the Paris Climate Agreement than anything related to WWDC.

Hey everyone!

I'm a developer at Aclima. We are working with Google on the mobile sensing tech talked about here: https://environment.google/projects/airview/

Honestly I'm really excited about it, but of course I'm just a tiny bit biased :)

If anyone has any questions I would love to do the best I can to answer them. Cheers!

Is Google the only large company to own its own TLD?

No, see https://en.m.wikipedia.org/wiki/List_of_Internet_top-level_d... for many more examples (.amex, .bmw, .canon, ...)

Google is certainly making more public of use their TLD than many of those other brands, though.

How did Google get a TLD? I read https://www.theregister.co.uk/2014/11/26/google_turns_on_goo... Can anyone do this with enough $?

Yeah, pretty much. There was an application process to go through as well, but if it's your trademark it's pretty much yours. (Though who knows when the next round of application opens..)

I believe we own our TLD as well, but don't use it, yet - it's for brand protection. If you have a look at my comment history and deduce my employer, you might understand why with our name...

I wonder why you wanted interested commenters to look through your history. You thought you would be viewed as shilling for BBC?

When they talk about refurbished components, are they really refurbished or just "overstock" sold as refurbished? It just seems odd that there would be that many refurbished components... or maybe they don't do that many server upgrades and just replace the whole things.

There's something funny about them reporting market-based emissions in the appendix. Mostly, that they're reporting them all the way back to 2011 when this was only approved by 2015. It's not really a big deal, but it's a bit unusual.

You can read a bit about the market-based scope 2 method here:


Let me explain a bit: market-based emissions are electricity power grid emissions computed the market-based method. The method measures your emissions differently depending on whether or not you bought a green instrument. Roughly speaking, green instruments are money you pay the power company that they will invest into green sources of energy, like wind and solar.

Under the market-based method, if you bought a green instrument, you can then report lower emissions, as if all of your electricity only came from the green energy you purchased. Conversely, if an instrument was available for purchase but you didn't buy one, your emissions will look like they all came from non-green sources of energy, like coal (these are called residual emissions). Of course, this is all a bit of an accounting lie: when you use the power grid, you can't tell the power company to not send you electricity from their coal power plant. Therefore, you also have to report your location-based emissions, which is the "normal" reporting that has been used in all previous years and is still being used. Roughly speaking, the location-based method is, take the total emissions of the entire power grid you're using, and multiply that by your fractional of the power usage of the grid over all other customers. Your emissions are then computed as your proportion of all green and non-green emissions that the entire power grid produces.

The market-based method is supposed to incentivise companies to buy green instruments because they can then report lower emissions under one method. The purchase of green instruments in turn is supposed to incentivise power companies to invest in green energy. The whole thing, however, is kind of based on an accounting "trick". This is why the greenhouse gas analysts at my job aren't huge fans of the market-based method.

WRI was pressured by many industry leaders to adopt the market-based method as part of the reporting protocol. If Google is reporting market-based emissions all the way back to 2011, way before it was approved last year, I wouldn't be surprised if they were one of the companies that lobbied WRI to include the market-based method. They are also mentioned as a motivating example in that link above.

Google has their own TLD? Not sure how I feel about that. Then again, we do have a myriad of others.

I'm the lead engineer of the Google Registry, which runs .google as well as the other dozens of TLDs that we run. Feel free to ask any questions. Our site is https://www.registry.google/ and our code is released as Free/Open Source Software at https://nomulus.foo (another one of our TLDs).

> I'm the lead engineer of the Google Registry ... Feel free to ask any questions

Oooh, oooh. I got a bunch. I have no idea if you can answer any of them though.

- Why aren't you using the .google tld more? Only place I ever saw it is registry., domains. and blog..

- I was going to ask why com.google doesn't resolve, but I see you fixed that since I last checked! How about google.google now? Why aren't you guys having more fun with this? :)

- Why is Google Domains still not available in 98% of the world's countries?

- Is owning .app and .foo part of some kind of evil master plan or are you guys just buying TLDs for fun? Because they're two fantastic TLDs and making both of them available for purchase would certainly be nice.

- ... and, uh, can you guys buy .cdn and make it available for purchase, too?

Domain names are sticky. It tends not to be worth it to change the domain name for sites that have already existed for a long time. For new sites, when you're choosing a new domain name anyway, it's easier to make it be on a new TLD.

com.google was used for an April Fools joke in 2015. We redirected it to the main homepage after that ended because it was widely linked from news articles and social media, and we didn't want to create a bunch of dead links. I'd love to have more fun with this stuff, and believe me, on the registry team we've had all sorts of fun ideas for sites like google.google and others.

I can't speak for Google Domains; they're a separate team that we do not work closely with. See e.g. https://archive.icann.org/en/topics/new-gtlds/gac-board-regi...

There's no evil master plan; we're still trying to figure out the gTLD business just like all the other registry operators. Releasing them as generic strings is certainly the easiest thing that could be done, but maybe there's something cooler?

Who knows when the next round of gTLD expansion will occur. It'll probably be many years. Until then, no new gTLDs can be created. .cdn is one of the ones that, in hindsight, at least one person probably should've thought of to apply for back in 2012. And I'm sure someone will take it in the next round.

>There's no evil master plan; we're still trying to figure out the gTLD business just like all the other registry operators.

I think you meant to say, there is no cabal [0][1].

[0] https://web.archive.org/web/20061016181703/http://from.to/ca... (Cmd-A to see the hidden text) [1] https://en.wikipedia.org/wiki/There_Is_No_Cabal

Are you ever going to make .app, .dev, etc. available for public registration?

Personally, I would love to. The decisions aren't up to just me though, so I don't have any real news or plans to share.

I don't think any of his reservations have to do with the technical aspects so much as ones which you'd likely be biased [1] against; and there's no point privileging a pro-corporate counterargument by bring them up here.

[1] https://en.wikipedia.org/wiki/Selection_bias

Why doesn't https://google.google redirect to Google?

We could do that really easily, but what would the benefit be? Google Search already has a canonical domain name, google.com. Adding a less convenient redirect doesn't benefit anything, but would require us to host that site indefinitely, as once released and linked to, we would never take it down lest we'd break existing links.

It's not the initial rollout that's problematic, it's committing ourselves to indefinitely maintaining it for value that is questionable at best.

So, why?

Companies shouldn't have ownership of TLDs, so why did you register so many TLDs that have nothing to do with Google, and which you don't allow to be used freely? Profits?

This is disgusting. The fact that corporations can control TLDs in the first place is disgusting, and I'd by far prefer if every TLD would have to be controlled by cooperatives (like denic) or other kinds of nonprofits, completely independent of everything else.

Especially in the current situation, where Google is cybersquatting on entire TLDs, just because they can. And while domains are frequently removed for squatting, nothing’s done with TLDs. Open the TLD for everyone, and use it, or get rid of it. But the current situation is disgusting. (I fundamentally disagree with companies controlling any kind of infrastructure)

That's the way it's been since the earliest days of the World Wide Web. You can read the history of .com here https://icannwiki.org/.com , but the TL;DR is that it was run by a private company as early as 1991, then acquired by Network Solutions in 1993, which was acquired by Verisign in 2000 who still run it to this day. .net has a similar history. I don't see any a priori reason why profit-driven companies should be excluded, and as you point out, there are plenty of non-profit registry operators to choose from (including Public Interest Registry, the non-profit that runs .org), if you care strongly about that.

denic is a ccTLD operator, whereas what we're discussing here are all gTLDs.

My issue is that there’s corporations squatting on TLDs, which no normal person can afford.

Not only does this increase the advantages an established corporation can have over individuals or startups, it also ends up with Google literally domainsquatting entire TLDs just because they can.

And there’s no registrar above them that could simply revoke their domains for squatting.

And de is as much a ccTLD as .io – it’s not really the actual case anymore. DE is one of the largest domains overall, even counting gTLDs.

And we could simply solve this – use the power of democracy (aka, governments) to force TLD ownership to be entirely non-profit.

I don’t want to see large corporations buy massive amounts of TLDs (as google has done) and then sit on them. In addition to the antitrust issues that can appear with this.

How should a startup ever be able to compete in a reasonable timeframe with Google if we allow these corporations to grow ever further, and gain ever more power?

I can't really comment on the rest of it, for obvious reasons, but this part is not accurate:

> And de is as much a ccTLD as .io – it’s not really the actual case anymore.

.de is clearly still the ccTLD for Germany. It's not at all like the situation of .io, which is run by a British company as, effectively, a two-letter generic TLD that has nothing to do with the actual Indian Ocean territories. .de clearly has a lot to do with Germany, including being widely used there (but not anywhere else), and also being run by denic, a German company headquartered in Frankfurt. I was just in Frankfurt last week talking with a bunch of denic people, actually. They absolutely do consider it a country code TLD, proudly and unapologetically.

You are about 5 years too late in your criticism. This whole debate was had when ICANN expanded top-level domains in 2012: https://en.wikipedia.org/wiki/Generic_top-level_domain#Expan...

Not agreeing or disagreeing, but I don't see the relevance of your point.

kuschku's comment makes it sound like he was unaware of what happened in 2012, and he is fundamentally misunderstanding the current state of gTLDs, so I was informing him.

Furthermore, the ship has sailed on the debate of whether these expanded TLDs are a good thing, so it's not a productive discussion. We might as well talk about whether commerce should be banned on the internet - that was a raging debate in the early/mid 90s, and that debate is over.

> Furthermore, the ship has sailed on the debate of whether these expanded TLDs are a good thing

Why? We could simply remove all of them again.

Or we could open TLDs to be affordable for everyone.

But the current situation has literally Google domainsquatting on TLDs. That’s... insane.

> Why? We could simply remove all of them again.

No, we couldn't. That would break the Internet in a large variety of horrible ways. You can't remove entire TLDs once delegated and widely linked to. That's closing the barn door after the horse has escaped.

> You can't remove entire TLDs once delegated and widely linked to. That's closing the barn door after the horse has escaped.

Why not?

We can also remove domains once registered and widely linked to.

The NIC of .eu just removed over 10'000 due to squatting.

Why shouldn’t we just remove Google’s TLDs where it’s squatting?

TLDs are qualitatively quite different than second-level domain names.

Do you have a link to the .eu actions you're referring to?

And the simplest answer is contracts. Every operator of a new gTLD has paid ICANN a substantial amount of money, has a contract to operate the TLD, and thus has certain legal rights. The TLDs cannot be taken away without good cause. There is no language in the contracts requiring a certain number of domains be launched by a certain date, so that cannot be justification for removing the operation grant. And of course contracts cannot be unilaterally changed.

I am not a lawyer, this is my personal opinion and not my employer's, all other disclaimers, please consult your own lawyer or Wikipedia if you have questions about contract law, etc.

When people are arguing whether or not the ramifications would be significant, you can't point to ink and call it blood.

Laws can be overturned, the legality of contracts overruled, recompenses rewarded. There's nothing "irrevocable" here, aside from the dangerous precedents people try and establish and sit on ("it's always been this way!!1") hoping to "run out the clock" and get away with moves that would never have flown in a later world.

Here’s the example case: https://www.out-law.com/page-8457

But EURid also de-registers almost all other domain squatters, as they actually wrote that into the contract.

> And the simplest answer is contracts.

Contracts are simply something enforced by law. Law can be changed. If not, law is simply something enforced by society and constitutions, constitutions can be changed.

The German constitution, for example, says that the right to own property only exists as long as that property is used for public good. In any other case, it can be seized.

Seizing the TLDs, or requiring DNS servers in your country to redirect the TLDs to other operators, would be an option.

And, as drastic as this sounds, it’s preferable over the current situation. Corporate squatting is inacceptable.

Yes, that case is an example of a qualitative difference between TLDs and SLDs. .eu, as a ccTLD for the European Union, has a TLD-wide policy that all registrants must be based in the European Economic Area. Those 10K domain names were all registered by a Chinese woman who used a fake address in London to register the domain names. So they are being fought over in court not because of "squatting", but because she violated the registry policy.

Note that it is up to the registry operator to set said policy (which ICANN must approve). So that mechanism does not even exist to take away a TLD.

Good luck in getting international law passed to take away TLDs from corporations. I just don't see it happening.

> Seizing the TLDs, or requiring DNS servers in your country to redirect the TLDs to other operators, would be an option.

This would break the Internet. Domain names have to work the same everywhere. You cannot have domain names pointing to different DNS zones entirely depending on where you access them from.

> This would break the Internet. Domain names have to work the same everywhere. You cannot have domain names pointing to different DNS zones entirely depending on where you access them from.

Domain names point to different DNS zones already today, everywhere.

First of all, anycast DNS is a common example.

Second, there’s massive government censorship in many countries, including the UK, and there’s also some US ISPs that modify any and all DNS queries for some domains (noticeable by the fact that DNSSEC-enabled domains are stripped of the signature, and return a false A record there).

> Good luck in getting international law passed to take away TLDs from corporations. I just don't see it happening.

Doesn’t have to be international

> This would break the Internet.

So? Giving corporations any control over the Internet has turnt out to be the far more breaking move already.

Google, for example, has recently massively destroyed support for IDN domains in Google Chrome – and no one bat an eye, despite it leading to billions in costs for many people and companies, some of which even had used an IDN domain in their trademark, and were now unable to use it.

Removing these new TLDs is a far smaller change.

> That would break the Internet in a large variety of horrible ways

Scare talk aside, no, it wouldn't. In most cases people would be unaffected; at worst people've dealt with broken links before and will continue to do so. It would only truly "break" anything for early adopters who dove in without thinking, and knew the risks of doing so.

Ask your bosses (the ones you cite here [1] as being the ones calling the actual shots) why everything isn't already .google etc. :)

[1] https://news.ycombinator.com/item?id=14490271

The 2012 round of gTLDs aren't provisional and they aren't used by "early adopters". They are part of the Internet now. There are millions of sites using them now, and any business owner creating a new site on, e.g., .design, is simply buying a domain name through a registrar; they are not "diving in without thinking" or being "early adopters" or "taking risks".

There were many years of discussion prior to releasing the latest gTLDs. This is not a continually ongoing process; the discussion was completed, and the new gTLDs were released. That can't be gone back on, not without causing harm to millions of website owners and network operators. It's simply not ever going to happen, and trying to argue otherwise isn't a fruitful use of time. Though, by all means, your time is yours to do with what you will, and if you really want to start a movement to try to roll back all the new gTLDs, godspeed. It seems like tilting at windmills to me though.

I've already addressed why everything isn't on .google already -- domain names are sticky, and it takes a lot of work to change the domain name of an established product. It tends not to be worth it. For new webpages though, e.g. environment.google, you're starting to see it being used.

Correction: There are 27.4M domains on new gTLDs registered; full stats here: https://ntldstats.com/tld

The total economic cost of yanking all of those from the existing established sites would easily run into the billions of dollars. Those domain names are on business cards, on sides of vans, on buildings, on advertising, etc. Untold amounts of money have already been spent building up reputations for sites on these new gTLDs, much of which would be lost if the domain name were yanked. The economic damage would be massive. And of course all existing links in press, blogs, social media, etc., would all die. That's why it won't ever happen.

Google Chrome has yanked out all support for Internationalized Domain names recently.

This includes far more domains, which are not just on business cards, but even in corporate names, trademarks, on billions of products.

Yet, it was still done.

Because Google felt like it.

Again, you are purposely misrepresenting things so as to paint Google in the most negative possible light. Here is, I believe, what you are referring to: http://www.pcworld.com/article/3191539/security/phishing-att...

It's about keeping web users safe by preventing certain types of phishing attacks. Support isn't "yanked"; the affected domain names are simply rendered in punycode (which isn't nearly as bad as yanking said domain names from the Internet entirely). There are lots of very real security issues around IDNs.

And that's not a security issue?

For decades we've trained users to consider domains that look different as malware.

Several major domains that relied on IDN currently are having issues with exactly that.

Because they had the IDN even on their business card, and now users are calling their support because of malware.

We already gave one company control over a major part of internet infrastructure, they are destroying other businesses, and we want to allow their questionable management to control more of the internet?

A company that tries to secretly hide their software in other installers, to try and secretly change your default webbrowser? https://www.youtube.com/watch?v=jWx1P93nS0c&t=51s

I can only paint that in bad light that looks evil if you look behind the façade.

Honest question: why make the distinction between top level domains and secondary level domains?

Anyone can register a second level domain under .com or the other main TLDs. This used to be the few main namespaces shared by everyone, at reasonable prices, and it went just fine.

ICANN made a de-facto distinction with the obscene pricing for top-level domains. There's a new global namespace, but you can't afford to get into that one if you're not a big fish.

That seems to be what the landscape looks like nowadays.

Almost every large company applied for gTLDs for their trademarks.

There's some surprises:

* .nike * .staples * .marshalls * .pamperedchef * .tiffany * .firestone

Though I have a feeling many of these are only going to be used for internal networks.

> Though I have a feeling many of these are only going to be used for internal networks.

Why would an internal network need a publicly resolvable domain?

Its good practice to own the domain you use internally so that someone else cant register it, then proceed to do evil.

My favorite from the first round of expansion is https://icannwiki.org/.afamilycompany

Starting cost is close to $200,000.

Seems like ICANN gets a lot of money these days...

Quite a few companies have vanity TLDs. Everything from "aarp" to "zappos".

See the list here: http://data.iana.org/TLD/tlds-alpha-by-domain.txt

.Google is a easy one. I remember hearing rumors about .amazon leading to more controversy because there also happens to be the river and the region.

It's not just a rumor; ICANN did refuse to grant .amazon to Amazon because of objections from various South American countries (through which the actual Amazon River and region run). More info here: https://icannwiki.org/.amazon

Don't they own a few? .XYZ is another which Google uses for https://abc.xyz

Google doesn't own xyz: https://en.wikipedia.org/wiki/.xyz

How bizarre that the terms of use for .xyz specifically bans illegal or fraudulent defamation of Justin Bieber and Oprah Winfrey. Presumably this only covers things that were already covered by the law for everyone else, but it is still kind of baffling. I assume the two of them were involved financially with the .xyz TLD?

I think it's reactive, i.e. in response to some prior incident, not implying an ownership stake. Do you really think Justin Bieber is involved financially with the domain name registry business? I doubt it.

I don't see why not, many celebrities/rich people have broad investments, especially in tech.

That does sound more likely.

"I jus' wan' a globally unique identifier to call you by, girl"

Almost certainly in response to affiliate marketers faking their endorsements of products.

Alphabet can move anything heavily polluting outside of Google and do their environmental press from Google.

These reports have existed for years before Alphabet. The track record is clear. There isn't always some trick or catch.

(Disclosure: I work at Google.)

> There isn't always some trick or catch

For some reason this concept seems impossible for the majority of HN to comprehend.

Look at the appendix, they report three different "scopes" for emissions. yeah, they could move anything polluting to somewhere else in the alphabet umbrella, but it's still in-scope for reporting.

"Scope 1 emissions are direct emissions from sources we own or control, such as company vehicles or generators at Google’s offices and data centers. Scope 2 emissions are indirect emissions from the production of electricity we purchase to run our operations. The location-based category reflects the average carbon intensity of the grids where our operations are located and thus where our energy consumption occurs. The market-based category incorporates our procurement choices, i.e., our renewable energy purchases via contractual mechanisms like PPAs. Scope 3 emissions are indirect emissions from other sources in our value chain, such as business travel or our suppliers"

By the way, the scopes are all part of the standard WRI greenhouse gas protocol for reporting emissions. You might be interested in how the market-based method works.


I don't see how Nest Labs would be considered to be part of Google the company, or in Google's value chain. They're just part of the parent company, Alphabet.

Finally catching up with Apple.

Using Apple as a role model for environmental sustainability is pretty flawed. They produce a massive amount of consumer goods, all electronic based and most requiring batteries. The amount of material that must be extracted from the earth to create their gadgets is boggling. Their products must all be shipped, across oceans, across countries and into your home. Furthermore, because of their "cutting edge" manufacturing processes their products require an enormous amount of specialized tooling. Tooling for which all the materials must be extracted from the earth, shipped, smelted and so on. And then there's the packaging. Pounds and pounds of meticulously manufactured packaging per product. Where did all the glossy ink come from? Where does it go? What about the polystyrene? Or the mold used to make that inset?

Production of consumer goods is pretty gnarly, no matter how it's painted by the producer. Buying a few carbon credits, using a bit of solar (the materials for which must be mined) or giving your employees electric cars or using LED lights in your offices does little to offset all the damage done by all the things I mentioned above.

Except that Foxconn still isn't building Apple products with clean energy.

Apparently they will in 2018: https://www.ft.com/content/087525c4-786b-11e5-a95a-27d368e1d...

Besides the other comment, also worth staring at : https://www.forbes.com/sites/alexepstein/2016/01/08/the-trut...

Nothing in this silly post doesn't also apply to Google. Google datacenters are connected to the grid as well. It's not feasible for Apple stores in cities and shopping malls to be connected to solar or wind farms. It's still all renewable energy.

Google's datacenters are large, but they don't consume the resources needed to produce 1 billion smart phones.

Android does. Google can't claim credit for Android when it likes and disown it when there are unpleasant side effects.

The parent is claiming one has to catch up to the other. I am pointing out this it not actually right. You are claiming "they are both just as bad".

So you don't seem to disagree with me?

You posted a blog post by a Randian climate change denier who authored a book called The Moral Case for Fossil Fuels.

I'm assuming you just searched for the first article that tries to spin Apple's environmental efforts negatively and that you don't actually endorse anything he wrote.

If in fact you do, consider this: when you deposit cash in an ATM on one side of town and withdraw it from another on the other side of town, the cash may not be physically the same cash you deposited. Electrons work the same way.

<A lot of words attacking this author's character>

Do you actually want to address the point he makes, or the point i made (that we are talking about apple), or just want to character assassinate a person? Because it seems a lot like the latter.

(If you consider him to be "not even worth discussing", then please do that instead of the character assassination, because it's not interesting, useful, or contributing to the discussion)

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact