1. If you are a citizen or a resident, you get an ID card to use for every public service. It's just a smart card with a government PKI.
2. The public services provide an email account that can only be used within the e-government services. The card is used for accessing those services.
3. The email service accepts either identity number or registry number of the recipient. So the recipient can be a legal entity.
4. You can and almost always do provide a forwarding address, so that you don't need to check.
5. You can't use it for other purposes. No RFC defined email address is shared with you. And it's just an internal system for official issues.
I've heard some countries issue mailboxes for citizens but I am not aware of the general use of these. Also, email services were designed to be decentralized but evolved into centralized systems, a current and unsolved problem. I am not sure about the privacy and security of government provided email services.
I still have a vague hope that the United States Postal Service could be "pivoted" into being a PKI provider and distribute physical tokens to citizens. They already have substantial procedures and infrastructure for verifying identity. There would be problems, to be sure, but I'd much rather get my ubiquitous PKI for citizens from the USPS than the banks or "tech giants".
I'd like to see the USPS expanded to become a public / municipal ISP of sorts.
If you read about the history of the institution, this is really what was intended in its constitutional incorporation. It really wasn't about physical mail per se, and you can't hold the founders accountable to something that was outside the realm of imagination at the time.
There's all sorts of information-structural things that are in the bounds of the USPS per the intent of its creation.
> I still have a vague hope that the United States Postal Service could be "pivoted" into being a PKI provider
It’s going to be an uphill battle or impossible as PKIs are too obscure for the average citizen to understand the benefits and any whiff of a federal ID card will be treated like the mark of the communist coup beast.
I always discover how Estonia is really amazing for lots of technology things. AFAIK they are by quite a margin the most advanced country in Europe when in comes to egovernment services. Moreover my (admittedly outside) impression is that they often go for technologically sound solutions not the ones which some large lobby organisation pushed for. This is particularly remarkable considering how small the country is, and in stark contrast to the mess that is egovernment services in Germany the richest country in Europe.
Probably because they're so small they're overlooked by the salespeople and lobbyists from the big corps. I imagine that helps a lot. In the UK there are plenty of smart people in Government who can and would build things in a sensible way (and sometimes they do!), but there are also legions of smooth talking salespeople who usually bend the ministers' ears more easily.
Good for our neighbors (And Hi!). Latvia is also advanced in regards to eservices
:)
We also get state issued ID card with PKI. We can access tons of services. Last I read I can buy a house, fully remotely. Including notary services via video call + all parties need to sign stuff with our ID card.
We get health results via email as an encrypted pdf, where password is given at the time when I submit samples.
Many business also use ID card to sign contracts between parties.
Bank transactions involve Smart-ID, 2FA app that I have to authorize via ID card for remote setup for any new device. (It involves generating new certificates) Smart-ID is developed by Estonia and is very convenient, secure way to authorize payments.
As of communication, no state issued email. However we usually get email notifications, for example from state tax service, that we should log in and read whatever we have to.
Estonia is the founding member of the NATO Cooperative Cyber Defence Centre of Excellence... they've been at the forefront for a long time. https://ccdcoe.org/
Size of the country is also something to consider. Population of the whole Estonia is fewer than population of a single city in other country. Area of the Estonia is also minuscule.
What works for a tiny state isn't always appropriate for a big state.
Yep - Here in Finland, just over the bay in the north, the Estonian e-prescription system is often quoted as much leaner, meaner and more functional than our own borked attempt, at a fraction of the cost.
In Italy we have a worse version of what you described.
1. An ID card you can use to access some services (carta di identità digitale)
2. Another card you can use to access healthcare related services and some other services (carta nazionale servizi)
3. SPID: your digital ID to access yet some other services, and also some of the above services. It is not released by the government but by other authorized entities such as banks, the national mail service and others. You need to pay a small fee for the verification, and sometimes an annual fee. There are different SPID levels but no one actually knows the difference between them.
4. PEC (posta elettronica certificata): a digitally signed email box you can use to send/receive documents, invoices, etc. or simply messages. Those are legally attributed to you and you can use it to talk to government agencies instead of sending registered paper mail. As SPID it is issued by an authorized third party.
We also have some smartphone apps that work as a combination of the above, and need some of the above to work.
As you can see it is a mess, a waste of tax money and we will need to waste more money in the future to make this mess work.
Nice :)
Edit: and by the way when you need something really important all the above are useless: you either need to start hopping from a public office to another (we have a lot of them) and/or go to a notary (a kind of medieval bureaucrat you pay a lot of money to sign and stamp sheets of paper)
In the Netherlands we do have an inbox from the government ("Berichteninbox" which is optional, the alternative is snailmail), it's coupled to the Digital ID system (DigiD), both are apps and webservices. You can use DigiD to access information on your pension, or healthcare insurance etc. The inbox can be (optionally) coupled to many government organizations and you receive information on taxes for example. I like the way it works, it works best if you have an Android or iOS system, but you can use it without (fully on the web).
Btw, a nice insight into email is also that it is one of the very few systems that decouples protocol from provider (Matrix and xmpp do that too, not widely adopted sadly) AND also has critical adoption (which Whatsapp also has in my country, sadly we are stuck with Meta there). We should never give up email because we will likely never get an open and free system like that back without some kind of government intervention. (Even though we all know email is a sub-optimal pile of hacks.)
Using Berichtenbox is a liability. Once you activate the thing, all sorts of (semi-)government communication goes there, but you can't forward it or download it via an open API. You have to use their smartphone app or webapp.
The notifications you can set up to a normal email address invariably only say that institution X sent you a message, but never specify the topic. That means you have to login to see if it is actually important and actionable or just something you already knew or a confirmation of something you submitted.
Even worse is this common scenario:
* Get notification that X sent something to Berichtenbox
* Login to Berichtenbox (first get mobile phone for required 2FA)
* Message says new information is available in X's web portal
* Login to X's web portal (mijn.somethingsomething.nl)
* Read totally pointless message that could even have been sent in plain email
Compare this to the postal flow:
* Get letter, read it
I think these days you can deactivate Berichtenbox and receive important information via post again, but this was not an option in the first year or so, so even experimenting with it was risky.
Importantly, though, that inbox is not an email inbox. This is what the process might look like (i.e. I've been through this):
1. You can an email in your regular email stating that there is a new message in your Berichteninbox. (No clickable link, presumably to avoid phishing.)
2. You go to mijn.overheid.nl to access your Berichteninbox. You sign in with DigID.
3. You open the mentioned message, which says a PDF with the actual letter is attached.
4. You open the PDF.
5. The PDF says you'll be able to file your tax returns a month from now.
I can't rightly say that I am able to navigate the maze of standards and acronyms associated with smart cards, but the OpenSC tools on Linux have worked for me with a couple different smart cards (Nitrokey HSM and Taglio PIVKey). There are quirks. The Taglio PIVKey can't load certificates using OpenSC, but I've always generated the certificates on the device anyway.
A question about number 4. By forwarding address, do you mean to a real email address? Denmark has a similar solution, but it can only be accessed via the website or a mobile application. The idea is that the content will almost always contain person information, so it shouldn't be allowed to be transmitted via an unencrypted channel.
Side note: Denmark has a one time pad instead of a smartcard. A smart phone app has since been added, and the one time pad will be discontinued in about a year, sadly.
I have been in Estonia for a few months and get my TRP recently. It's new to me. But I heard that it's the same. It's just a notification probably. Yet, the term "forwarding address" makes me think it can be something else. I did not get any email from there yet, so I don't know actually.
The PKI thing includes a physical ID card, a software solution called Smart-ID and a mobile solutions called Mobile ID. The software solutions are just authenticator apps that you've matched with your ID.
> If you are a citizen or a resident, you get an ID card to use for every public service. It's just a smart card with a government PKI.
This is the biggest flaw in the design. Tying the ID card to a single identity.
If you're using it with a bank, it needs to be tied to your bank account. If you're using it for physical access control at your company's building, it needs to be tied to your employee account. These are different things, and should be different things, for security.
You don't want a single system for everything. It makes the incentive to break it stronger, so it gets broken more often. It makes the consequences of it getting broken larger, so the damage when it happens multiplies. And it gets integrated into everything, so the amount of time it takes to roll out fixes increases. It's a security nightmare, and it gets polynomially worse the bigger the country is that tries to do it that way. (For reference, the GDP of Estonia is less than one third the revenue of Costco.)
No, it's solid design. It's a very simple safe primitive. You can build endless infrastructure on top of it. Similar to subkeys.
For example a lot of businesses use Smart-ID on top of that. You need to tie the smartid stuff to your PKI identity. But after that you can just use that as identity.
By public services, I meant the public services provided by the state. For instance, health insurance, family doctor application, taxes, etc.
Banks require your ID whether it's smart or not. But it's not for payment purposes but for authentication. And they are not state bodies, but private commercial entities. They are not part of the PKI ecosystem of the state.
That still has the same issue mentioned in the article: it works fine inside an organization (the organization being, in this case, the whole country), but not between different organizations. For instance, how would I, a Brazilian, send a message to someone using that system?
But is it really email as we know it? It looks more like a private message system like you find in forums and social networks.
In France, we are not as advanced as Estonia when it comes to e-government services, but we have an official identification system called "France connect", and government services have private messaging systems to communicate with them. And I think many countries have similar systems. The only difference seems to be that it is better integrated in Estonia.
I could only wish the US had something like that. Very few Congresspeople could even succinctly describe email to you, let alone express the need for a system like this. And even if they could introduce a bill, Big Tech lobbyists would instantly swoop in and proclaim the idea as a threat to national democracy, and instead try to steer the legislator to just hosting entire thing on their platform instead. I fucking hate our federal government.
There are a few issues. First, Estonia is a small country and it's relatively easier. Second, there's no legacy solution to comply with when a new feature is developed. US has both federal and local government systems, and many agencies with their own services. That creates an overhead for a new and standard[1] solution.
Adding to that, even worse, each state are also implementing their own identity solutions. Take Service NSW which is an expensive front-end built on Salesforce, with its digital drivers license. Each platform has its own digital identity system, which is just waste of taxpayer money.
I really hope that we end up moving back towards supporting open protocols.
I was heartened (and a little surprised) that Jack Dorsey recently mentioned that the draconian control of the Twitter API was the worst thing Twitter had done [1].
The corporatisation of the Internet, has undone a lot of the great work that had traditionally underpinned the network.
It feels like the slow, laborious and fundamentally equitable nature of standards ratification in the open has been seen to be at odds with the OKRs of tech businesses.
Businesses that sell and work with natural resources are starting to wake up to the idea that a degree of cooperation and inter-market regulation with peer companies can positively impact individual performance. Sustaining business is even more fundamental than making profit.
In the same sense; open protocols can help to develop rich and sustainable markets that benefit the consumer; as well as those businesses that operate in within it.
It really is about incentives. When the government and universities were the primary agents influencing the internet, open protocols were favored I presume because they incentivized the decentralization that the internet was created for.
Now private corporations are the primary agents of change, and they are driven by very different incentives. When was the last time you heard of a company based around open protocols being valued at a billion dollars?
And the money involved is just too great. I don't see how anything is going to change.
And yet none of those corporations has displaced email, despite the fact that it has become a universal cyberattack channel, with a stagnant UX that doesn't address most real-world use cases for email!
I saw a need for a safer, better, decentralized protocol for email, so I drafted one (TMTP) and implemented client & server. More at:
It feels like the slow, laborious and fundamentally equitable nature of standards ratification in the open has been seen to be at odds with the OKRs of tech businesses.
At the risk of sounding like I'm trivialising this comment (with which I completely agree), this difference in behaviours has as its root the difference between a long- vs short-term mindset.
I think the issue might be the huge amount of VC cash invested, and the need for such a player to have explosive growth to a huge valuation.
Open standards of federated systems could lead to slow sustainable growth with a spot for the original designers and pushers of the protocol. But open standards won't let you fully dominate the market, they don't allow you to leverage all the VC cash, and so they don't pay back on massive investment. Because quite a lot of the benefits are shared.
Moreover, slow growth can't compete with VC cash investment. The VC backed competition will have a better UX, more features, aggressive marketing, and in general be more developed. All because they can develop their product a lot faster because they have more money behind them.
I was heartened (and a little surprised) that Jack Dorsey recently mentioned that the draconian control of the Twitter API was the worst thing Twitter had done [1].
I wasn't, because he didn't do jack shit to change it. We hear this bullshit all the time; big actors sound off about what was wrong at their previous places, but rarely did they do anything to upset the apple cart.
I'd say (and I think this happens quite often, moreso in politics) it is at least possible you may be overestimating his power to do so, perhaps at least by the time he realized it?
Twitter wasn't his github repo, it was his gazillion dollar company that has to answer to a lot of stakeholders.
(That being said, no reason to not get on them about it.)
Email is our only reliable communication method between different organizations.
I'm still of the opinion there should be public-option internet services. Everyone deserves an e-mail address that cannot be taken away from them without a court order.
> Everyone deserves an e-mail address that cannot be taken away from them without a court order.
Not even a court order, arguably. Internet access and it's essential services like email, is arguably a human right in developed countries. Almost impossible to find employment without it.
I don't think email is the issue here but it is DNS. Email relies on DNS and unfortunately government and ISPs have too much control to take your domain or have it blocked.
also, (or: alternatively?) one that can't be/won't be blocked by the centralized services' spam filters. The biggest hurdle to running your own email server nowadays isn't the online time or the data volume or anything; it's that the existing institutions don't recognize you as part of the institutional club and block your messages...
btw, Germany did this a decade ago: giving everyone an email account with the national mail service, as an "official email." I honestly don't know anyone here who uses it.
> it's that the existing institutions don't recognize you as part of the institutional club and block your messages...
How common is this if you’ve setup DKIM, SPF, etc.? I’ve only heard about problems in that context where someone hadn’t done the basics or was trying to send from shared IPs and hit some spammer’s past reputation.
This was the idea behind the USPS originally, if you read records about it's founding. It wasn't intended to be about physical mail, but about "transmission of information" or something like that. It's actually kind of striking.
People already have accounts in national databases and there's a notification system using e-mail, sms and phone. Why not just manage the e-mail for them (and if they want - they can forward it to their private e-mail of choice).
Define internet services, or do you mean email service?
There are many decisions that impact the usability and cost of the service. Some people need high volume sending or large mailbox storage. Do you punish people for sending spam? Do you filter spam, if so, how. Do people need public terminals to access the service? Etc.
I'm not saying it should be free. Quite the opposite. It should charge the user per-e-mail on an at-cost basis. It's a utility, not a hand-out. Think post office.
Do you punish people for sending spam?
Only by making them pay for every mail they send.
Do you filter spam, if so, how.
On the receiving end. A plugin system would let people choose to subscribe to updated blocklists and filtering rules, just like modern adblocking.
Do people need public terminals to access the service?
Same way it is now. The vast majority of people have their own smart devices, and for the ones who don't there's the public library.
while I agree with the idea of emails that can't be arbitrarily shutdown, SSN-xx-HERE@citizen.gov sounds like all kinds of awful. It will either be instantly unusable or require a gov approved SPAM filter, both of which are bad. It also seems like a good vector to force a backdoor on all comms.
I think the issue (as in the what) is that people should always be able to have a fallback option for sending and receiving email that's not at the whim of Google, MS et al.
SSN-xx-HERE@citizen.gov is a how, which may or may not be a good one. For one, here in France, the SSN isn't as important as it seems to be in the US, so its being public is probably less of an issue. This approach would still be bad for spam or whatever.
Another how could be by using the same kind of naming in use elsewhere, as in name.surname.213@citizen.gov. Except that not anyone would be able to randomly open an account. You'd have to go through some kind of agency that would check your ID. This would allow them to expose a way of changing (in case its overrun by spam) or unlocking (in case of lost password) your account safely.
We have a more or less similar thing in France with bank accounts: you have an "opposable right" (as in, undeniable) to have a basic bank account. Not sure if this is a French law or an EU directive, but I think the same could work for email.
Having a government approved spam filter would be better than letting an oligopoly of five companies decide what constitutes spam.
In fact, I can't think of a single market dominated by a handful of large companies hasnt been improved by the introduction of a government competitor.
There's a reason telcos lobby hard against community broadband and that financial institutions dial back the usuriousness of their fees when the post office offers bare bones accounts.
This would also require everyone to have an email client to handle their email address though. I believe this the reason most folks have a gmail/outlook account because it's easy to set up and operate, not just because it gives them a unique-ish address?
That's where right to disconnect and anti-spam measures come in. I'm in France, and i have the right to refuse my employer contacting me outside of work hours, and they pay me if i don't. And since robocall spam is illegal, i get ~1 call every 4-5 months at most, to sell me a different internet or electricity or mobile plan, and they're obligated to respect my refusal to be contacted thereafter ( and all do).
I think the issue is that since email is more and more required to interact with the Government services, they should also provide a usable alternative. Why have your citizens rely on random foreign services from which they may be cut off because a bot somewhere is having a bad day?
In France at least, many people (mostly the elderly) are having a hard time using computers and such. Some Government agencies have dedicated personnel to help them with filling in the forms and such on dedicated computers. It could probably be easier for them if they also provided email instead of relying on a third party provider. Grandma lost her password? No biggie. If she has her ID, we can reset it for her. Good luck getting any kind of support from Google / Yahoo in such a case.
Of course, I will explicitly say that I would be very much against such a service being compulsory for the people. I just think it should exist.
Price isn't the issue, the issue is that marginalised people can be denied service outright. E.g. if you don't have a phone you can't sign up for many of those services. If you have an unusual name you may be rejected. And if you have unpopular political views you may be kicked off.
Cost aside, a solution needs to be highly available. Third party services can not guarantee your email will be available for the duration of your lifespan.
There is also the issue of data stewardship, (democratic) governments can ensure independent reviews and be held accountable for security breaches and data misuse. They could also be held liable for losses incurred by service defects.
arbitrary account shutdown is a known issue with free email. losing your gmail account without explanation and with no recourse can be an awful situation for anyone, especially for vulnerable populations. This leaves the options of forcing private orgs to maintain email addresses or have a gov email for every citizen, both of which have significant drawbacks.
* interacting with the government online (I needed to provide an email address to update my driver's license and vehicle registration)
* opening a bank account
* renting an apartment
These are important things, so we might as well have some guaranteed way to access these services. Especially because you need an email to interact with a lot of government services.
But we also all need to eat and use a toilet to live. Those seem to be provided by the market to a reasonable degree. Email is also pretty cheap and there's at least some choice among providers, though of course far fewer that food types.
I don't think this is a fair comparison. You can replace a toilet, arguably upgrade to a better one with little to no disadvantage.
Taking away an email address someone has had and is their primary point of contact for years, possibly decades is irreplaceable. Being able to create a new one isn't equal to the old one.
Not sure about elsewhere in the world, but even regular mail isn't that painful in my country. Pay a nominal fee to Australia post and you can have all mail addressed to you forwarded from your old to new address for N months (or years).
The reason there are as many food types is that you pay for food, while most people (except corporations) don't pay for email. This makes it so investments in food production can be returned without waiting for network effect / vendor lock-in to reach a significant level.
Your toilet example actually proves the point. Generally, water and wastewater services are not private (I know there's exceptions and most are going terribly wrong). So yes everyone's ability to use a toilet is somewhat government guaranteed.
Yeah but the existence of toilets is mandated by code. The existence of email isn't regulated in any way. Requiring non-commercial (<-- which is doing a lot of work here) email addresses would cause a robust market to appear overnight.
In this day and age of censorship, I feel the same about web hosting. The American government should provide their citizens with a small space of hosting to share their thoughts.
I do hope that web 3 brings a DNS service that can be bought once and owned forever that nobody can tear even from your cold dead hands. I'm not holding my breath though.
I always have this feeling that email is flawed and due for a complete overhaul or replacement - and then I think about it a little harder and I realize that it's actually really good at it's intended purpose.
Other than fiddling around the edges with security improvements, spam filtering, and a few other nice-to-haves, there's not really much that need improvement.
Some features of email that are nice:
- It's completely open standard
- I can host it myself if I want, or not.
- It is completely decentralized and roughly point-to-point, subject to email routers.
- Other than getting an email address, no other 'linkage' or prepwork with that person is required.
- My address is not tied to any other service, like a phone number. (in contrast to e.g. WhatsApp)
- It supports unsolicited communication from unsolicited sources (e.g. marketing)
- It's easy to ignore communication I don't care about. (e.g. marketing)
- Non-people are supported, like group emails/aliases (support@...)
- I can trivially attach files, subject to some practical constraints
- Email can be handled by the recipient in a wide variety of ways using different client mechanisms.
- I can front-end my email in a variety of ways, such as with a contact form.
Those are just the few I can think of off the top of my head. I'm sure there are others.
Indeed. While one can complain about this or that little detail, email is by far the best communications mechanism on (or off) the Internet.
The key part of course is that it is completely open and standardized. Nobody owns it. That is a lesson that we should learn, but is every time forgotten.
No proprietary walled garden can ever come close to the usefulness of email precisely because email is open and standard. With proprietary systems it is inevitable users are subject to the whim of the owner. Might not be able to get accounts, or be arbitrarily banned, or have the app only available on limited platforms, etc.
I've been using email since the late 80s and more importantly I've had the exact same email address since the mid 90s. It's been hosted by multiple providers and the last decade I've been hosting it myself. But always the same domain and address.
email, as ancient and flawed as it is, is a shining example of the Lindy effect in play - the future life expectancy of a technology or an idea is proportional to its current age.
> There are a huge variety of intra-organizational communication systems, to the point where pretty much every large enterprise provider seems to have one (Slack, Microsoft Teams, Discord, etc etc).
That's why I find Delta Chat piggybacking on Push-IMAP such an interesting concept: https://delta.chat
The only issue I've had in my limited use is that a Deltachat email triggers a notification on the desktop before it is moved to a Deltachat-specific folder. The solution is to configure Sieve filters to do this upon reception, but just saying that we've lost all 99% of potential users
I don’t think email is as decentralized and federated as it used to be.
In theory, email is a service that is simple enough for anyone to run themselves. Most Linux distros come with sendmail, so theoretically it should be as easy as reading the manual and exposing some ports. Spam is performed server side both at the origin and at the destination to mitigate bad actors, and because email is simple, there should be no shortage of clients to choose from.
In reality, 1/4 of all email users globally are on Gmail. Apple Mail is the most popular mail client followed by Outlook, then Gmail. SMTP and IMAP are theoretically simple, but the bellwether providers use APIs on top of these protocols that have added some functionality at the expense of restricting the proliferation of email clients. Many large companies that used to run their own email (through Exchange, Zimbra, etc) are moving to hosted Office 365 or Google Workspace. One major AWS-scale outage in Gmail or Azure will incite (and has caused) serious panic and disruption (which is great for SREs like me since we’ll continue to get paid serious money to keep all this stuff running while maintaining a healthy work-life balance, but I digress).
Furthermore, one doesn’t simply “stand up” their own email server unless they don’t care about landing in people’s spam folders.
Additionally, many companies outside of the US _do_ use WhatsApp (Facebook) for official communication. I’d posit that this trend is only accelerating.
I agree that email is fundamental technology, but I can see a future where it disappears in favor of something like federated Slack (or, worse, instant messaging centralized and controlled by the FAANG cabal with insurmountable cost-of-entry). Given the suppression of “free speech” on Twitter et al during peak COVID/peak insurrection (for valid reasons), this is slightly worrying.
The thing is, you can have your email address(es) under your own domain, and change mail hosting providers while keeping your email address(es). It's true that too few people are doing that.
Apart from that, email is not going anywhere (not going away) anytime soon as the standard medium for B2B communication. And in B2C communication as well, an email address is the one baseline you can count on everyone having. I don't see that being replaced by anything proprietary either.
Email's federation is an escape hatch. It's presence means that I can go to any provider I want to if I'm dissatisfied with my current provider. I can even run my own as a last resort. (Or first resort if that's your preferred mode of operation). Until that escape hatch disappears, which is unlikely, I will always have choice of providers.
I don't have whatsapp, or discord for that matter. I have slack for work but I don't use it externally. I will probably never have those systems for my personal communication which means that if a company wants to communicate to me they are going to have to use email, full stop. I think there is a large barrier to email ever going away. Removing it from the market would require coordination that most companies and providers will probably never want to engage in. It's a lowest common denominator that all of them will want to support to avoid their users getting silo'd into a system that is not theirs.
The biggest blunder for me is that there were usable decentralized communication options before that were popular, but because of trying to monetize user's data FAANG started to tighten their grip on any decentralized solution, and I think they succeeded. They are already trying do to email the same thing they did to XMPP and RSS.
It's objectively awful... But when you step back a tiny bit, that doesn't matter. What matters is the inter-organizational community it achieved. If these organizations wanted to continue, they could come up with a negotiation technique like in http - both ends can use whatever fancy thing they both support, but fall back unto riding dinosaurs if that's the only thing that works.
I often swing between longing for a federated protocol that can be managed by technical people and used by everyone, like XMPP, and something more P2P to reduce centralization and allow everyone to instantly "open an account" with no need for technical skills, but that still needs some kind of relays for asynchronous communication, like ssb. None are technically perfect (although I really like the simplicity and extensibility of XMPP) but in the end what matters is not that: it's about how the protocols are used, how they allow all of us to communicate, how they give more power to those who aren't already using the internet to exchange information. And that is not a technical problem
That's only because running TCP/IP over a 1200 bits per seconds connection was close to impossible, but Fidonet had > 40 thousands nodes connect by 1990, so decentralization was already a thing back then.
When modems became fast enough to handle a TCP/IP connection it was ~1994 and by then Internet was already (relatively) cheap and available.
E-Mail is one of the last remaining federated systems on the Internet, but I doubt it will survive long as the large players slowly sabotage it. I think already more than 90 % of all e-mails are delivered by three or four large companies, which is a trend that will continue.
But HTTP is client/server whereas e-mail is server/server (or client/server/server/client). Small independent server operators are at the mercy of the large companies as those can just stop processing their e-mails (which they already often do). That's different for HTTP (though gatekeeping happens there too via discoverability and other mechanisms).
A lot of businesses host their own email, if not on Linux then using Microsoft Exchange (see e.g. [1]). While that is being somewhat decreased by the cloud trend, I don't see it going away, as those businesses generally like keeping their independence.
Readit News
1. If you are a citizen or a resident, you get an ID card to use for every public service. It's just a smart card with a government PKI.
2. The public services provide an email account that can only be used within the e-government services. The card is used for accessing those services.
3. The email service accepts either identity number or registry number of the recipient. So the recipient can be a legal entity.
4. You can and almost always do provide a forwarding address, so that you don't need to check.
5. You can't use it for other purposes. No RFC defined email address is shared with you. And it's just an internal system for official issues.
I've heard some countries issue mailboxes for citizens but I am not aware of the general use of these. Also, email services were designed to be decentralized but evolved into centralized systems, a current and unsolved problem. I am not sure about the privacy and security of government provided email services.
If you read about the history of the institution, this is really what was intended in its constitutional incorporation. It really wasn't about physical mail per se, and you can't hold the founders accountable to something that was outside the realm of imagination at the time.
There's all sorts of information-structural things that are in the bounds of the USPS per the intent of its creation.
Many services would want to use your PKI token as identification, we would likely give up a lot of privacy because of its existence/ease-of-use.
It’s going to be an uphill battle or impossible as PKIs are too obscure for the average citizen to understand the benefits and any whiff of a federal ID card will be treated like the mark of the communist coup beast.
We also get state issued ID card with PKI. We can access tons of services. Last I read I can buy a house, fully remotely. Including notary services via video call + all parties need to sign stuff with our ID card.
We get health results via email as an encrypted pdf, where password is given at the time when I submit samples.
Many business also use ID card to sign contracts between parties.
Bank transactions involve Smart-ID, 2FA app that I have to authorize via ID card for remote setup for any new device. (It involves generating new certificates) Smart-ID is developed by Estonia and is very convenient, secure way to authorize payments.
As of communication, no state issued email. However we usually get email notifications, for example from state tax service, that we should log in and read whatever we have to.
https://www.bbc.com/news/39655415
What works for a tiny state isn't always appropriate for a big state.
1. An ID card you can use to access some services (carta di identità digitale)
2. Another card you can use to access healthcare related services and some other services (carta nazionale servizi)
3. SPID: your digital ID to access yet some other services, and also some of the above services. It is not released by the government but by other authorized entities such as banks, the national mail service and others. You need to pay a small fee for the verification, and sometimes an annual fee. There are different SPID levels but no one actually knows the difference between them.
4. PEC (posta elettronica certificata): a digitally signed email box you can use to send/receive documents, invoices, etc. or simply messages. Those are legally attributed to you and you can use it to talk to government agencies instead of sending registered paper mail. As SPID it is issued by an authorized third party.
We also have some smartphone apps that work as a combination of the above, and need some of the above to work.
As you can see it is a mess, a waste of tax money and we will need to waste more money in the future to make this mess work.
Nice :)
Edit: and by the way when you need something really important all the above are useless: you either need to start hopping from a public office to another (we have a lot of them) and/or go to a notary (a kind of medieval bureaucrat you pay a lot of money to sign and stamp sheets of paper)
Btw, a nice insight into email is also that it is one of the very few systems that decouples protocol from provider (Matrix and xmpp do that too, not widely adopted sadly) AND also has critical adoption (which Whatsapp also has in my country, sadly we are stuck with Meta there). We should never give up email because we will likely never get an open and free system like that back without some kind of government intervention. (Even though we all know email is a sub-optimal pile of hacks.)
The notifications you can set up to a normal email address invariably only say that institution X sent you a message, but never specify the topic. That means you have to login to see if it is actually important and actionable or just something you already knew or a confirmation of something you submitted.
Even worse is this common scenario:
* Get notification that X sent something to Berichtenbox
* Login to Berichtenbox (first get mobile phone for required 2FA)
* Message says new information is available in X's web portal
* Login to X's web portal (mijn.somethingsomething.nl)
* Read totally pointless message that could even have been sent in plain email
Compare this to the postal flow:
* Get letter, read it
I think these days you can deactivate Berichtenbox and receive important information via post again, but this was not an option in the first year or so, so even experimenting with it was risky.
1. You can an email in your regular email stating that there is a new message in your Berichteninbox. (No clickable link, presumably to avoid phishing.)
2. You go to mijn.overheid.nl to access your Berichteninbox. You sign in with DigID.
3. You open the mentioned message, which says a PDF with the actual letter is attached.
4. You open the PDF.
5. The PDF says you'll be able to file your tax returns a month from now.
[1] https://www.id.ee/en/article/install-id-software/
Side note: Denmark has a one time pad instead of a smartcard. A smart phone app has since been added, and the one time pad will be discontinued in about a year, sadly.
The PKI thing includes a physical ID card, a software solution called Smart-ID and a mobile solutions called Mobile ID. The software solutions are just authenticator apps that you've matched with your ID.
This is the biggest flaw in the design. Tying the ID card to a single identity.
If you're using it with a bank, it needs to be tied to your bank account. If you're using it for physical access control at your company's building, it needs to be tied to your employee account. These are different things, and should be different things, for security.
You don't want a single system for everything. It makes the incentive to break it stronger, so it gets broken more often. It makes the consequences of it getting broken larger, so the damage when it happens multiplies. And it gets integrated into everything, so the amount of time it takes to roll out fixes increases. It's a security nightmare, and it gets polynomially worse the bigger the country is that tries to do it that way. (For reference, the GDP of Estonia is less than one third the revenue of Costco.)
No, it's solid design. It's a very simple safe primitive. You can build endless infrastructure on top of it. Similar to subkeys.
For example a lot of businesses use Smart-ID on top of that. You need to tie the smartid stuff to your PKI identity. But after that you can just use that as identity.
https://www.smart-id.com/
Banks require your ID whether it's smart or not. But it's not for payment purposes but for authentication. And they are not state bodies, but private commercial entities. They are not part of the PKI ecosystem of the state.
> You can't use it for other purposes. No RFC defined email address is shared with you
This is not entirely true. You get both:
* idcode@eesti.ee can only be used by government senders.
* you also get first.last.uniqueid@eesti.ee which works as a regular email address.
In France, we are not as advanced as Estonia when it comes to e-government services, but we have an official identification system called "France connect", and government services have private messaging systems to communicate with them. And I think many countries have similar systems. The only difference seems to be that it is better integrated in Estonia.
[1] https://xkcd.com/927/
I was heartened (and a little surprised) that Jack Dorsey recently mentioned that the draconian control of the Twitter API was the worst thing Twitter had done [1].
The corporatisation of the Internet, has undone a lot of the great work that had traditionally underpinned the network.
It feels like the slow, laborious and fundamentally equitable nature of standards ratification in the open has been seen to be at odds with the OKRs of tech businesses.
Businesses that sell and work with natural resources are starting to wake up to the idea that a degree of cooperation and inter-market regulation with peer companies can positively impact individual performance. Sustaining business is even more fundamental than making profit.
In the same sense; open protocols can help to develop rich and sustainable markets that benefit the consumer; as well as those businesses that operate in within it.
[1] https://www.revyuh.com/news/software/developers/twitters-fou...
Now private corporations are the primary agents of change, and they are driven by very different incentives. When was the last time you heard of a company based around open protocols being valued at a billion dollars?
And the money involved is just too great. I don't see how anything is going to change.
I saw a need for a safer, better, decentralized protocol for email, so I drafted one (TMTP) and implemented client & server. More at:
https://mnmnotmail.org/ & https://twitter.com/mnmnotmail
Related protocol projects in development include:
https://mathmesh.com/
https://en.wikipedia.org/wiki/Dark_Mail_Alliance
I appreciate how the tide turned, but societies appetite changes over time; and the fact is, open protocols are not anti-profit, or anti-business.
At the risk of sounding like I'm trivialising this comment (with which I completely agree), this difference in behaviours has as its root the difference between a long- vs short-term mindset.
Open standards of federated systems could lead to slow sustainable growth with a spot for the original designers and pushers of the protocol. But open standards won't let you fully dominate the market, they don't allow you to leverage all the VC cash, and so they don't pay back on massive investment. Because quite a lot of the benefits are shared.
Moreover, slow growth can't compete with VC cash investment. The VC backed competition will have a better UX, more features, aggressive marketing, and in general be more developed. All because they can develop their product a lot faster because they have more money behind them.
I wasn't, because he didn't do jack shit to change it. We hear this bullshit all the time; big actors sound off about what was wrong at their previous places, but rarely did they do anything to upset the apple cart.
Twitter wasn't his github repo, it was his gazillion dollar company that has to answer to a lot of stakeholders.
(That being said, no reason to not get on them about it.)
I'm still of the opinion there should be public-option internet services. Everyone deserves an e-mail address that cannot be taken away from them without a court order.
Not even a court order, arguably. Internet access and it's essential services like email, is arguably a human right in developed countries. Almost impossible to find employment without it.
And yet, it can be taken from you with a court order.
Like for phone numbers (at least here you can migrate the whole number, even with ndc)
The state could give out an emailadress like a social security number and you just use that as an alias and can choose whatever provider you want.
And for these emailadresses the providers would be obliged to take you. (Like for mandatory insurances. We have them where I live)
btw, Germany did this a decade ago: giving everyone an email account with the national mail service, as an "official email." I honestly don't know anyone here who uses it.
How common is this if you’ve setup DKIM, SPF, etc.? I’ve only heard about problems in that context where someone hadn’t done the basics or was trying to send from shared IPs and hit some spammer’s past reputation.
There are many decisions that impact the usability and cost of the service. Some people need high volume sending or large mailbox storage. Do you punish people for sending spam? Do you filter spam, if so, how. Do people need public terminals to access the service? Etc.
Do you punish people for sending spam?
Only by making them pay for every mail they send.
Do you filter spam, if so, how.
On the receiving end. A plugin system would let people choose to subscribe to updated blocklists and filtering rules, just like modern adblocking.
Do people need public terminals to access the service?
Same way it is now. The vast majority of people have their own smart devices, and for the ones who don't there's the public library.
SSN-xx-HERE@citizen.gov is a how, which may or may not be a good one. For one, here in France, the SSN isn't as important as it seems to be in the US, so its being public is probably less of an issue. This approach would still be bad for spam or whatever.
Another how could be by using the same kind of naming in use elsewhere, as in name.surname.213@citizen.gov. Except that not anyone would be able to randomly open an account. You'd have to go through some kind of agency that would check your ID. This would allow them to expose a way of changing (in case its overrun by spam) or unlocking (in case of lost password) your account safely.
We have a more or less similar thing in France with bank accounts: you have an "opposable right" (as in, undeniable) to have a basic bank account. Not sure if this is a French law or an EU directive, but I think the same could work for email.
In fact, I can't think of a single market dominated by a handful of large companies hasnt been improved by the introduction of a government competitor.
There's a reason telcos lobby hard against community broadband and that financial institutions dial back the usuriousness of their fees when the post office offers bare bones accounts.
In France at least, many people (mostly the elderly) are having a hard time using computers and such. Some Government agencies have dedicated personnel to help them with filling in the forms and such on dedicated computers. It could probably be easier for them if they also provided email instead of relying on a third party provider. Grandma lost her password? No biggie. If she has her ID, we can reset it for her. Good luck getting any kind of support from Google / Yahoo in such a case.
Of course, I will explicitly say that I would be very much against such a service being compulsory for the people. I just think it should exist.
There is also the issue of data stewardship, (democratic) governments can ensure independent reviews and be held accountable for security breaches and data misuse. They could also be held liable for losses incurred by service defects.
Why?
* applying for jobs
* getting covid tests/vaccines
* buying virtually anything online
* interacting with the government online (I needed to provide an email address to update my driver's license and vehicle registration)
* opening a bank account
* renting an apartment
These are important things, so we might as well have some guaranteed way to access these services. Especially because you need an email to interact with a lot of government services.
Taking away an email address someone has had and is their primary point of contact for years, possibly decades is irreplaceable. Being able to create a new one isn't equal to the old one.
Not sure about elsewhere in the world, but even regular mail isn't that painful in my country. Pay a nominal fee to Australia post and you can have all mail addressed to you forwarded from your old to new address for N months (or years).
... why? What are you basing this on legally/morally other than your own want?
Other than fiddling around the edges with security improvements, spam filtering, and a few other nice-to-haves, there's not really much that need improvement.
Some features of email that are nice:
- It's completely open standard
- I can host it myself if I want, or not.
- It is completely decentralized and roughly point-to-point, subject to email routers.
- Other than getting an email address, no other 'linkage' or prepwork with that person is required.
- My address is not tied to any other service, like a phone number. (in contrast to e.g. WhatsApp)
- It supports unsolicited communication from unsolicited sources (e.g. marketing)
- It's easy to ignore communication I don't care about. (e.g. marketing)
- Non-people are supported, like group emails/aliases (support@...)
- I can trivially attach files, subject to some practical constraints
- Email can be handled by the recipient in a wide variety of ways using different client mechanisms.
- I can front-end my email in a variety of ways, such as with a contact form.
Those are just the few I can think of off the top of my head. I'm sure there are others.
The key part of course is that it is completely open and standardized. Nobody owns it. That is a lesson that we should learn, but is every time forgotten.
No proprietary walled garden can ever come close to the usefulness of email precisely because email is open and standard. With proprietary systems it is inevitable users are subject to the whim of the owner. Might not be able to get accounts, or be arbitrarily banned, or have the app only available on limited platforms, etc.
I've been using email since the late 80s and more importantly I've had the exact same email address since the mid 90s. It's been hosted by multiple providers and the last decade I've been hosting it myself. But always the same domain and address.
No proprietary system can ever compete.
https://en.wikipedia.org/wiki/Lindy_effect
Any replacement will have to keep the above in mind because there's no test like the test of time.
And it has the security improvements and others as well (see features of e.g. Discord or WhatsApp).
Anyways, I don't think I still use email for its intended purpose anyways. It mainly became something to tie accounts to and to 2fa
- It is designed well for medium-length content, say a few paragraphs or so per message.
- It works well, and is mostly understood to be used for asynchronous communication.
- Easily and usefully searchable.
- Captures state/context well.
- Threaded
That's why I find Delta Chat piggybacking on Push-IMAP such an interesting concept: https://delta.chat
Here's my ASK HN: https://news.ycombinator.com/item?id=22854641
In theory, email is a service that is simple enough for anyone to run themselves. Most Linux distros come with sendmail, so theoretically it should be as easy as reading the manual and exposing some ports. Spam is performed server side both at the origin and at the destination to mitigate bad actors, and because email is simple, there should be no shortage of clients to choose from.
In reality, 1/4 of all email users globally are on Gmail. Apple Mail is the most popular mail client followed by Outlook, then Gmail. SMTP and IMAP are theoretically simple, but the bellwether providers use APIs on top of these protocols that have added some functionality at the expense of restricting the proliferation of email clients. Many large companies that used to run their own email (through Exchange, Zimbra, etc) are moving to hosted Office 365 or Google Workspace. One major AWS-scale outage in Gmail or Azure will incite (and has caused) serious panic and disruption (which is great for SREs like me since we’ll continue to get paid serious money to keep all this stuff running while maintaining a healthy work-life balance, but I digress).
Furthermore, one doesn’t simply “stand up” their own email server unless they don’t care about landing in people’s spam folders.
Additionally, many companies outside of the US _do_ use WhatsApp (Facebook) for official communication. I’d posit that this trend is only accelerating.
I agree that email is fundamental technology, but I can see a future where it disappears in favor of something like federated Slack (or, worse, instant messaging centralized and controlled by the FAANG cabal with insurmountable cost-of-entry). Given the suppression of “free speech” on Twitter et al during peak COVID/peak insurrection (for valid reasons), this is slightly worrying.
Apart from that, email is not going anywhere (not going away) anytime soon as the standard medium for B2B communication. And in B2C communication as well, an email address is the one baseline you can count on everyone having. I don't see that being replaced by anything proprietary either.
I don't have whatsapp, or discord for that matter. I have slack for work but I don't use it externally. I will probably never have those systems for my personal communication which means that if a company wants to communicate to me they are going to have to use email, full stop. I think there is a large barrier to email ever going away. Removing it from the market would require coordination that most companies and providers will probably never want to engage in. It's a lowest common denominator that all of them will want to support to avoid their users getting silo'd into a system that is not theirs.
"perfect is the enemy of good"
What is old is new again.
When modems became fast enough to handle a TCP/IP connection it was ~1994 and by then Internet was already (relatively) cheap and available.
Just like HTTP/1.1 can't be deprecated because too much infrastructure depends on it.
These protocols are simple and as complexity fails we all need to go back to them!
[1] https://news.ycombinator.com/item?id=26362178