Projet #Sankey du #shiftproject

On peut faire des copies de ses sélections...

#gazaeffetdeserre
#greenhouse
#transitionecologique
#co2 #methane #NH2

Le site : https://sankey.theshiftproject.org/

Bonjour, je viens de découvrir cet outil du @shiftProject #shiftproject : #Sankey

Un graphique qui détaille des émissions des gaz à effet de serre par année et par différentes répartitions.

C'est bluffant... enfin, après avoir bien étudié le truc :)

Je recommande

#gazaeffetdeserre
#greenhouse
#transitionecologique
#co2 #methane #NH2

https://sankey.theshiftproject.org/

Une sélection d'indicateurs locaux inspirés par les recommandations du #shiftproject #roscoff #alim
https://territoiresaufutur.org/commune/29239

Vous allez à l'univershifté des Shifters du #shiftproject les 17/18 juin mais vous avez un peu de mal à faire vos choix ?

Une interface à été bricolée vite fait pour aider: https://vuejs.experimentslabs.com/univershifte-helper/

Il y a certainement quelques bugs ;)

There are alternatives to SMS 2FA

This is a follow-up to: Let's talk SMS 2FA and its shortcomings in the context of an energy crisis (and electronic components shortage)

3- FIDO2 Physical keys
This is where you insert a special USB stick in your computer and you touch a button on that USB stick. Or, for a phone, you touch the key to the back of the phone.

The user needs their security key within reach at all times, perhaps in their wallet or on their keychain (meaning here the real-world wallet and keychain)... Don't overdo it, within walking distance is fine.

Like TOTP's above, this requires the clock of the user's device to be reasonably synchronized with a universal clock.
It also has an upfront monetary cost (buying the USB stick), usually a few dozen EUR, but there are no subscriptions for the end user or for the service provider. The physical key is only required of the end user, not the service provider doing the authentication.

Also in the same way as TOTP's, no network component is involved at setup or for any subsequent use. The service you are using doesn't need to use any third parties. After they set it up, there is barely any operational or computational overhead, and no additional bills.

The biggest drawback is what happens when the key gets lost:

• Does the user have a spare key they set up on the same accounts?
• What happens then, do they buy a new key to replace the one that was lost, do they cross their fingers and hope not to lose their last hey, or do they disable 2FA?

Since these are the same principles as with house keys, the exact same precautions, it doesn't seem like a big stretch that this should have high potential.

Beyond this, these keys offer a number of additional protections compared to the other methods:

• Verification of the service every time the button is tapped, meaning that the key will only work with services it was set up on (protection against a variety of attacks, lookalikes, phishing etc)
• The key can't be fooled by a service name using typography tricks to pose as a legitimate service
• Tapping the button is required, which means it can't be automated and it avoids involuntary grants
• The prompt only appears if the program is in focus and after the key is inserted, protecting against the notification fatigue mentioned above
• Losing a key is merely a temporary and a minor financial inconvenience, but not usually a security issue
• Even in the case of theft, the lost key can be disabled remotely
• These keys have no mechanical parts and they require no maintenance, they can last years
• They have no installation process and they can be used instantly with any computer or phone (provided the right USB port is available, or NFC chip)
• There is no 6 digit code to share, and is therefore the best defense to date against account takeovers

Depending on the make and model of the key:

• They are often tamper proof, in which case their memory can't be analyzed and the secrets they hold can't be revealed
• They often have no network capability at all, leaving them unable to leak any kind of information to anything more than a few centimeters away, and making them impervious to connectivity issues
• They have no battery, using either the USB or some kind of magnetic field to get the power they need, making them more reliable
• Some are crush resistant, waterproof and other attributes like this

Some models also have Bluetooth and/or WiFi capabilities, but that requires a battery and also makes the setup more complex. Perhaps this provides enough additional convenience that it appeals to some people. However, it seems to me that the authentication landscape is already complex enough.

While all other forms of 2FA are vulnerable because they rely on a secret that can be intercepted or shared, these keys are physical. For a bad actor, taking over your account means expending considerably more time and resources, as well as getting to close geographical proximity, all of which significantly raises their risk. Getting into your account is most likely not worth the effort.

I think plugging a well-known USB stick in a computer, or touching the key to the back of a phone as if for a contactless transaction, is the most reliable and most secure way to go at the moment.

(The end)

#shiftproject

There are alternatives to SMS 2FA

This is a follow-up to: Let's talk SMS 2FA and its shortcomings in the context of an energy crisis (and electronic components shortage)

2- Time codes in an application
Also called TOTP for "time-based one-time password".
This is where you have an app on your computer or phone. At 2FA setup, the service displays a barcode that you scan with your app, thereafter producing a different 6-digit code every 30 seconds. When the service needs to verify you, they just ask for whatever code is displayed by your app at that moment.

The way this works is by applying a mathematical formula that derives the barcode (the "seed") using the current time. No network component is involved at setup or for any subsequent use. The service you are using doesn't need to use any third parties, they only need a program just like you need an app. After they set it up, there is barely any operational or computational overhead, and no additional bills; and the same is true for the end user.

This tech has been around for decades already. It used to take the form of a small plastic device called a token, with a tiny screen big enough to display a code that would change every half hour or so. These were provided by large companies to access company resources from remote locations. Another version, more recently, is being built in some credit card offerings in place of the previously fixed 3 digit number on the back.

The modern version we are considering here doesn't require a physical component or a license from a big security company. It is merely a secret code, a kind of password if you will, that is produced by open source software and is stored in an app on the end user's phone.

Drawbacks include the need for the user to back up their seeds, so they can set up their seeds on a new device in case theirs gets lost. It isn't always clear that the phone saves this list over time. The user might not know whether or not this is the case, and the phone policy might change at any time without alerting the user. Using a dedicated app like a password manager fixes this issue, but at some monetary cost (subscription) and the caveat that this subscription operator has the ability to see the user's seeds (trust issue).

Another drawback is the reliance on a near-perfect clock. If your phone is more than 30 seconds out of sync from the international clock, or if the service is, then the formula produces incoherent results and the authentication can't work. Even 10 seconds can be an issue. Thankfully, nowadays clocks are mostly synced all the time.

And finally, and most evident, these TOTP codes can be shared. Someone could convince you to tell them your code. In fact, bad actors have been known to attempt this, even going so far as to automate it. With the right dialogue cues, end users are vulnerable to this.

(To be continued)

#shiftproject

#shiftproject

There are alternatives to SMS 2FA

This is a follow-up to: Let's talk SMS 2FA and its shortcomings in the context of an energy crisis (and electronic components shortage)

3- FIDO2 Physical keys
This is where you insert a special USB stick in your computer and you touch a button on that USB stick. Or, for a phone, you touch the key to the back of the phone.

The user needs their security key within reach at all times, perhaps in their wallet or on their keychain (meaning here the real-world wallet and keychain)... Don't overdo it, within walking distance is fine.

Like TOTP's above, this requires the clock of the user's device to be reasonably synchronized with a universal clock.
It also has an upfront monetary cost (buying the USB stick), usually a few dozen EUR, but there are no subscriptions for the end user or for the service provider. The physical key is only required of the end user, not the service provider doing the authentication.

Also in the same way as TOTP's, no network component is involved at setup or for any subsequent use. The service you are using doesn't need to use any third parties. After they set it up, there is barely any operational or computational overhead, and no additional bills.

The biggest drawback is what happens when the key gets lost:

• Does the user have a spare key they set up on the same accounts?
• What happens then, do they buy a new key to replace the one that was lost, do they cross their fingers and hope not to lose their last hey, or do they disable 2FA?

Since these are the same principles as with house keys, the exact same precautions, it doesn't seem like a big stretch that this should have high potential.

Beyond this, these keys offer a number of additional protections compared to the other methods:

• Verification of the service every time the button is tapped, meaning that the key will only work with services it was set up on (protection against a variety of attacks, lookalikes, phishing etc)
• The key can't be fooled by a service name using typography tricks to pose as a legitimate service
• Tapping the button is required, which means it can't be automated and it avoids involuntary grants
• The prompt only appears if the program is in focus and after the key is inserted, protecting against the notification fatigue mentioned above
• Losing a key is merely a temporary and a minor financial inconvenience, but not usually a security issue
• Even in the case of theft, the lost key can be disabled remotely
• These keys have no mechanical parts and they require no maintenance, they can last years
• They have no installation process and they can be used instantly with any computer or phone (provided the right USB port is available, or NFC chip)
• There is no 6 digit code to share, and is therefore the best defense to date against account takeovers

Depending on the make and model of the key:

• They are often tamper proof, in which case their memory can't be analyzed and the secrets they hold can't be revealed
• They often have no network capability at all, leaving them unable to leak any kind of information to anything more than a few centimeters away, and making them impervious to connectivity issues
• They have no battery, using either the USB or some kind of magnetic field to get the power they need, making them more reliable
• Some are crush resistant, waterproof and other attributes like this

Some models also have Bluetooth and/or WiFi capabilities, but that requires a battery and also makes the setup more complex. Perhaps this provides enough additional convenience that it appeals to some people. However, it seems to me that the authentication landscape is already complex enough.

While all other forms of 2FA are vulnerable because they rely on a secret that can be intercepted or shared, these keys are physical. For a bad actor, taking over your account means expending considerably more time and resources, as well as getting to close geographical proximity, all of which significantly raises their risk. Getting into your account is most likely not worth the effort.

I think plugging a well-known USB stick in a computer, or touching the key to the back of a phone as if for a contactless transaction, is the most reliable and most secure way to go at the moment.

(The end)

#shiftproject

There are alternatives to SMS 2FA.

This is a follow-up to: Let's talk SMS 2FA and its shortcomings in the context of an energy crisis (and electronic components shortage)

It depends on every service to implement them, so they aren't ubiquitous yet, but they are available to users and they often avoid all that networking infrastructure.

1- Push notifications
This is where they send a notification to another device that is already logged in (trusted device).

Mostly implemented by the biggest like Apple and Google, when you do an action with one device that isn't authenticated, they send a notification to another device that is already known to that service. Very convenient for the end user, this doesn't rely on cell connectivity or cell towers, but still relies on a message crossing the planet several times over the internet. However, in contrast with SMS, the flow the user is engaged in already requires the same network, so the need for a reliable worldwide network infrastructure doesn't increase.

Still, since this involves sending messages across a network, there is a monetary cost for the service provider. The message doesn't rely on specialized hardware infrastructure like cell towers, just the regular internet, so the running cost could be lower than for SMS. Perhaps this cost is sometimes included in a wider offering, or a variety of commercial obligations. In any case, this comes with ties for the service provider and with most of the same drawbacks we described for SMS: availability, maintenance, compatibility etc.

There is also the concern that the service you are using might not use secure methods to send the notification. There is some degree of incentive to that effect but no guarantee. Unlike SMS, there isn't a legacy of infrastructure to contend with, hardware or otherwise, since push notifications are a fairly recent protocol, compared with SMS. It's a toss-up, which means the potential for abuse isn't zero.

But mostly, this requires several devices to work. We don't all own several devices that are capable of receiving push notifications, or have several of them with us at every moment. It is also useless when logged out everywhere and we need to log back in.

I personally fear that this tech is insidiously dangerous. We forget it is there, how we set it up, and we probably also ignore all emails reminding us to verify our settings; so when the day comes that we need it, it was already broken by configuration drift. I think it is only acceptable if the user also sets up an alternate 2FA method along this one, which in my opinion drives usability considerably down. Incidentally, usability was perhaps the best argument in favor of this tech.

Additionally, the fact that such a notification relies on the presence of a network that, coincidentally, is quite probably the same as what the user is already using, affects negatively the security of the entire endeavor as far as "second" factor goes... Indeed, the point of a second factor is that it uses a different method of doing things.
(This also applies to my advice about SMS over WiFi, to some extent)

And finally, some bad actors have been known to trigger this notification many times in a row, so many times that the user gets annoyed and they finally accept the prompt on behalf of the bad actor (just to stop the phone from buzzing), thereby defeating the entire purpose of the notification.

(To be continued)

#shiftproject

Let's go into the security side of things.

This is a follow-up to: Let's talk SMS 2FA and its shortcomings in the context of an energy crisis (and electronic components shortage)

For the value proposition of the SMS network to work (deliver texts whenever, wherever & in a timely manner), every part of it needs tight integration with every cell carrier on the whole planet. All carriers over the world agreed on a common protocol.

Since every country regulates its own requirements for mobile carriers in its borders, they all have varied requirements and incentives. This means that upgrading anything, protocol wise, has been reported to be a challenge. Interoperability is a necessity, therefore any changes are a detriment or can't happen.

Some consequences, without the recipient or the sender knowing about it:

• Any carrier asked to forward a message can see that message in clear text
• Any carrier along the chain can alter the text before forwarding it to the next hop
• Carriers can send copies to anyone, keep a backup indefinitely in their database etc

In some cases, the message can even be redirected to a different recipient altogether.

(These have all been reported to happen)

This is a systemic problem in the mobile carrier network. Neither senders nor recipients can do anything about it. By design, the messages aren't encrypted. There is risk of abuse.

Here's another issue: a phone number change.

Say you enable SMS 2FA on services X, Y and Z. Time passes and one day you change your phone contract, perhaps willingly, perhaps due to an emergency or any number of reasons. In any case, do you go back to X, Y and Z to do the 2FA change procedure? Do you remember that you need to do it before the old contract expires? Do you even have the time, if it was an emergency? Perhaps the next time you remember is because of a 2FA prompt on these services, but now the message gets sent to a phone number that is no longer yours. Do you ask a stranger to forward you the 2FA security code? Was the old number perhaps not attributed to anyone yet?

As hard as I thought about this recently, there is probably still a service out there where I set up SMS 2FA years ago and I have forgotten about it. The next time I change my phone number, that service will be lost to me.

Other minor failings inherent to SMS transport:

• Connectivity: the phone might not have cellular connection, perhaps located in a blind spot or the building is shielded, etc
• Battery: the phone might be out
• Credits: the sender failed to credit their account with their own
• Third party: the service that is billing the service you are directly using, has an outage, went out of business, made a technical change that's incompatible with your direct service, etc, delaying their capacity to send SMS by anything from hours to weeks

Here are a few things we can do to help drive down the necessity for cell towers (and therefore network hardware production, cement, etc), as individuals:

1- Enable our mobile devices to favor WiFi over the cell network, for texts and calls. This doesn't work everywhere (mostly at home, friends and family, possibly at work) but it already represents a huge portion of the time we spend anywhere.

It's just one checkbox in our phone settings.

2- Disable cell data when we aren't using it, especially when we are using a more stable network (WiFi). The phone doesn't need to keep track of nearby cell towers when the local WiFi meets all our requirements. If the towers don't get pinged, their load decreases, they don't reach capacity, there is reduced need to build hardware redundancy, less electronics being produced, less strain on the energy requirements, less strain on the rare metals industry, less hydrocarbures required overall.

This is slightly more involved since we need to tap a button whenever we get home, but it's definitely doable.

#shiftproject

Let's talk SMS 2FA and its shortcomings in the context of an energy crisis (and electronic components shortage)

[TL;DR] It is mostly that our laziness drives networking infrastructure expansion. Alternatives like TOTP and physical keys don't require that infrastructure and are also better at security.

First off, let me just say that SMS 2FA is a piece of tech that is amazingly accessible: practically everyone who knows how to use a phone understands how to read a text they received via SMS. We receive a 6-digit code, it appears as a notification, we copy the code. Job done. Sometimes the application can even read the text without involving the user.

SMS 2FA represents incredible user experience & decent security improvements for little inconvenience.

But.

There are a number of failings and shortcomings. I'll start with the monetary cost for the service provider.

Service providers (think any website or app here) are most likely using a mobile carrier who bills them for their SMS usage, or perhaps they went through the extraordinary step of interfacing with the existing carriers. Either way, there are costs (& energy usage) for their use of the global SMS network. It's not unlike our own individual SMS bill, the more we use it the more we pay, until we just opt in for the unlimited contract (but it's still there).
=> For the smaller services, this is expensive and sometimes prohibitively so. The bill is also dependant on events out of their control: how often their users reset their passwords.

Then there is the number of hops.

The way the SMS network works, any time a text is sent, it goes through a number of intermediaries before reaching its recipient. You could think of it as snail mail going from post office to post office until reaching the recipient's mail box. In the easiest cases, there are probably 4 hops involved in the transfer, and that's assuming the service and the recipient are using the same carrier, are in geographical proximity to one another, are both available at the same time (phone isn't turned off or otherwise indisposed), etc. Geographical distance means going through different routers (electronic infrastructure), possibly using other carriers as gateways when crossing borders and such, etc. Availability means the transfer might need to be reiterated several times until the recipient acknowledges delivery, and also that an automated message goes all the way back from the recipient to the original sender, using much the same infrastructure as the original message.

The issue here is not so much the size of each message. That's tiny. Rather, it's the sheer amount of physical electronic infrastructure we need all over the planet to guarantee delivery of every text message within minutes, sometimes within seconds, because these 2FA codes are time sensitive. They expire fast, often within an hour. We don't want to context switch, any delay makes us lose our train of thought. The login flow must be seamless or the service might lose on the conversion rates. Take your pick.

There are stories about how cell networks used to be overloaded at year's end parties. In years gone by, sometimes SMS would take days to arrive. Well, the infrastructure was improved and it doesn't happen as often.

The issue is also not that each text message involves many cell towers. That's not true. It involves them only at both ends (start and receive) and most of the way is handled by computers connected through the regular internet. Actually, one way for consumers to help cut down on cell tower buildup is by enabling their internet service provider (WiFi) to take over their cell connectivity so that the wired internet is preferred to cell towers.

And suddenly, with routers all over the world, for a text sent to the other side of the planet, merely 4 machines don't quite cut it. The tiny amount of data per text requires numerous energy impulses all over the world, as well as data storage, electrical redundancy, spare machines and various safeties meant to avoid data loss.

That's just for a 6-digit code, and we haven't yet gotten to the parts where it fails the user in miserable ways.

(To be continued)

#shiftproject

Is there work being done on policies to mandate that every application should quantify its network usage (locally of course) over time, possibly with recommendations to its user?

As in, the user goes to the appropriate screen in their favorite app, and they can see that they've used X amount of data over mobile, Y amount over WiFi etc. Something sensible and actionable.

For example, perhaps the app could suggest that an HD feed was wasteful in that time period because the hardware doesn't support it, with a quantified value of how many co2 emissions would have been avoided by going with a lower definition (or a % of savings), and a button to do just that.

#shiftproject

Before that, I noticed that my phone would occasionally switch between mobile data and wifi, depending where I was sitting at the time

Whenever it started losing wifi signal strength, the phone would try switching to another wifi or to my mobile carrier (whichever was more convenient, I guess)

Which means the phone is constantly assessing signal strength and reliability, not only for the current signal receiver but also for other nearby signals, in case it needs to switch to keep its user happy at all times

This is what I'm adressing with this change: I'm removing the constant mobile carrier scan and the test of signal strength with their antenna, whether I'm currently on wifi or not

In the end, I probably want to use mobile carrier signal/antenna maybe three hours on the days that I commute to work, and zero on the days that I work from home

And all I have to do is tap a notification when I leave or get back to my home (especially the latter, but not necessarily the former)

That's perhaps 9 hours per week instead of 24/7, so about a 95% drop?

#shiftproject

Things I did recently:

• added an automation on my phone to disable the mobile data antenna while I am at home (using WiFi instead);
• added the reverse automation to disable the WiFi antenna while on the go;
• trying to keep the mobile data antenna disabled when I am not using the phone (that's most of the day): calls and texts can still go through, but no distractions during the day.

The goal is to reduce the stress my usage is causing on the infrastructure, and (with my peers doing the same) reduce its energy consumption & maintenance frequency

#shiftproject

@Aeffondrement #shiftproject, #lowtechnation, #resiliencemontagne pour ma part

Plus je repense à ton msg @proclamator et plus je me dis qu’on ce fait avoir…ambassadeur de Green got depuis qqs semaines, le discours startup nation me saoule de plus en plus…et surtout, quand ça sera une banque qui pèse, ok ils investissent dans le vert mais en enrichissant qui ? Qqs privilégiés qui auront misé sur l’affaire au début ? Je pense que je vais clôturer mon compte chez eux…

https://piaille.fr/@Viandpowder/109386318015096371
#banque #ecologie #shiftproject #greengot

Quand Jean Marc Jankovici parle aux financiers https://youtu.be/nKjkdQCG-KE #récession #carbone4 #shiftproject

Le #ShiftProject vient de publier 3 nouvelles vidéos sur la #résilience face au #changementClimatique et aux perturbations qui l'accompagnent.

Une pour la #France urbaine, une pour la France rurale et une pour la France d' #outreMer :
https://www.youtube.com/watch?v=c66zoOpw9Hg
https://www.youtube.com/watch?v=Uirfccy1uRw
https://www.youtube.com/watch?v=Hho6Lws44Rc

Source:

[1] "Lean ICT -- Towards Digital Sobriety" (SHIFT Project, March 2019): https://theshiftproject.org/wp-content/uploads/2019/03/Lean-ICT-Report_The-Shift-Project_2019.pdf

#DigitalSobriety #LeanICT #SHIFTProject

Le #shiftproject rappelle qu'il fait 5kg de #protéines végétales pour produire 1kg de protéines animales. #actforclimate