It has always been incredibly sad to me that the German ID card (Personalausweis) has an RFID chip inside with trust zones, certificates, authorization features, and much more and just never had been used. Like at all except for getting cigarettes at vending machines.
12 years after the first RFID Personalausweis had been issued it is only possible to register your car in some cities. Maybe there are other minor uses but it's negligible.
It's a very cool technology with a certificate authority and cryptographically secured claims for various things (proving you are over 18 without revealing your DOB, only giving out the name and address, authenticating as a German citizen, pseudonymity with separate identities for each service you use etc.). All functionality is also available for use over the internet.
The German Wikipedia has a good overview: https://de.m.wikipedia.org/wiki/Personalausweis_(Deutschland...
Some Herr Doktor probably followed all the best practices to implement "trust zones, certificates, authorization features, and much more" in the ID, doing their job really well. But actually changing the processes to use those features is not anyone's job, and might actually eliminate a lot of jobs, so it never happened.
They might be slow, complicated, budgeted terribly, unbelievably incompetent by standards of typical for-profit mega corporation, but a lot of those projects work at first try and works for decades, in the end.
SLS capsule came back in one piece on first try. That German ID system probably works too. And that’s great.
They rolled that out together with finger printing.
People value their privacy here and this was overstepping too many boundaries.
Those features have also never been explained to the average Michel here. Even IT interested people are not aware or understand the good things about it.
I really like the development that has gone into the e-ID. They even have thought out a safe way to update your PIN (https://www.pin-ruecksetzbrief-bestellen.de/)! The biggest drawback of all is the lack of any marketing, IMHO.
> Das Gesetz hat sogar eine eigene Website mit einem Dashboard,
Right now most platforms don't do ID validation because users hate sharing their details. By making it more privacy-safe more platforms will do it because the barrier is lower. I really hate that, I think the internet should remain anonymous. So I can pick whatever nick and even have multiple.
For the unfamiliar, with identity-based encryption, the recipient's public key is a function of the key authority's public key and some "identity", such as a national ID number or email address. Their private key is a one-way function of their identity and the key authority's private key. So, the recipient needs to ask the key authority one time to generate their private key for them, but there's only one public key to distribute. For the whole system, the sender can calculate the recipient's public key. The private key isn't even necessarily calculated before the sender has sent their message! It's very convenient and flexible!
Of course, the downside is that the private key is deterministic and can always be re-generated by the key authority, so it's fundamentally vulnerable to attack by the key authority. Also, some of the underlying math is less well studied than standard ECDHE/DHE/RSA, so we're less confident about vulnerabilities lurking just under the surface.
> With IRMA it is easy to log in and make yourself known, by disclosing only relevant attributes of yourself. For instance, in order to watch a certain movie online, you prove that you are older than 16, and nothing else.
That’s not “cool stuff”.
But yes, in general, we’re SO CLOSE…then you have to go do Anmeldung with a paper form in person
That stuff honestly improved quite a bit in the recent years. Most of these services are just not advertised or integrated enough so far.
Yes there’s plenty corruption and disastrous bets (ISDN…), but let’s not pretend the situation isn’t intentionally created.
I mean I would also make them passports but I think that is impossible.
The IDs are actually heavily unified nowadays https://en.m.wikipedia.org/wiki/National_identity_cards_in_t...
Usually I'll argue for market efficiency over other concerns; but in this case the Europeans are on to something with the GDPR. The role of government is to make this sort of personalised identification hard - not to enable it. The end game is going to be hard times and mass discrimination against minorities.
Despite Germans have made bad experiences with that, the idea survived the 3rd Reich.
The idea that as a human being you need "governmental documents" to identify is an authoritarian core value that is fundamentally against individual freedom.
Jews were forced to always carry their "Kennkarte" with them.
Please do not tell me about positive use cases that are based on the naive idea that "the government is the good guys".
When electronic IDs are not rolled back everywhere then democracy and individual freedom will be lost for a very long time.
It helps a lot that public agencies can now offer a so called IT Zulage of a few hundred euros to 1000 per months that brings salaries on par with the private sector. In my team, this worked wonders and we managed to get some really good people.
On the other hand, the task is enormous, we were discussing last week that if we had double the man power, we would still have the same workload, because we push back on a lot of things. We have about 70 projects that we wrote and maintain and a backlog of another 12 waiting to be started.
BWI has the same problem, I’ve been approached multiple times by them for this project, which from my knowledge is being intensely worked since many years.
Meanwhile, other countries have offered a web portal for years with a digital version of the Patsientenakte and all prescriptions in one place. Works. Not in Germany though.
> On the other hand, the task is enormous, we were discussing last week that if we had double the man power, we would still have the same workload, because we push back on a lot of things. We have about 70 projects that we wrote and maintain and a backlog of another 12 waiting to be started.
I rest my case.
Are you looking for more people?
The jank was incredible and just using them you could feel the spaghetti code, incompetence and age. My advice would be to stay away as far as possible. As a user and as a developer.
I'm based in Germany and I share the sentiment that things have been a bit backwards here in terms of a widespread reluctance to let go of paper based administration. This was awkward 14 years ago when I moved here and at this point it's just beyond pathetic. But things are changing. Germans are well aware that people outside of Germany are noticing how far behind they are and are shaking their heads at those silly naive German paper fetishists. So, there's a lot of domestic pressure to actually start fixing this. The covid crisis in the last few years forced a lot of Germans to do things with their phone that until then were completely unheard off in this country. Like paying for stuff or proving that they didn't have covid. That used to be a thing where paper and rubber stamps were the only acceptable solution.
So, I look at this as something that can change quite rapidly after not having changed much at all for decades. The will and money are there and Germans are starting to remember that they can actually get some stuff done when they put their mind to it. We're also seeing this with the current energy crisis. That crisis has unlocked budgets all over the public sector. And "digitalization" (as it is referred to here) is part of those budgets. Germans love efficiency and people have been pointing out that they haven't been very efficient. Which is embarrassing and annoying. So, they are fixing it now. There are now countless of bureaucrats tasked with actually showing some results for the inflated budgets they've been given. We're talking hundreds of billions of euros here. It's not all going to be spend wisely but some of it will yield results.
Nobody forgot that, just that past successes are in no way indication of future successes.
Otherwise SV would have been in Germany/UK instead of California. But that hasn't happened.
Same how in the late '80s to early '90s everyone was saying that Japan's tech sector and economy would completely overtake the US's and yet that hasn't happened but the reverse happened. From then on US tech sector steamrolled everything. Will that last? Maybe, maybe not.
Not much consolance for the German people, who still have to deal with a lot of paper administration but a happy accident nonetheless.
All German productivity will end and even German language itself will be replaced by grunts and shrugs.
In the end, I got rid of my DMG Mori machine with its Siemens control and replaced it with a Taiwanese machine that functions reliably.
I think that's exactly why Matrix might be a good fit, as the technical federation aligns well with the pre-existing social federation. I'm really optimistic for that project!
But don't worry, if German health services doing something right is triggering your "the end is nigh!" response: As far as I know, the rollout for patients is still a long way coming and they still don't even have a date set for video chat (right now a cottage industry of anyone involved in HC doing their own WebRTC thing).
Maybe this changes too in the future?
E-Rezept was supposed to launch in 2022 but has been postponed until mid 2023. Some regions already tested it. It didn't work out well, so some regions dropped out of the testing phase. I'm pretty sure it won't work well at launch and we will have to rely on printed prescriptions for quite some time until all pharmacies and doctors use the new system.
Currently was supposed to be in a pilot phase in two regions, but both of them cancelled it due to privacy concerns: https://www1.wdr.de/nachrichten/erezept-kelber-medizin-westf...
The health industry is very weird from top to bottom. True for most countries, but Germany certainly adds a few cherries on top. Or at least massively diluted cherry essences…
Just look at the authentication of the E-Rezept (electronic prescription) service: https://github.com/gematik/api-erp/blob/master/docs/authenti... This is supposed to be standard OpenID.
I fully expect the matrix protocol to suffer the same treatment under the hand of the Gematik.
If you want to know how things end up such a chaos take a look at the definitions of the payload data: https://github.com/gematik/api-erp/blob/master/docs/erp_fhir...
6 different sets of definitions by 5 different regulating bodies, with the organizing company Gematik GmbH owned by 9 different stakeholders: https://www.gematik.de/ueber-uns/struktur
I wonder how kosher it was.
A friend working at a big radiology attempted to manage that since there are issues with FAX systems since ISDN technology has been boxed and the E-Arztbrief would have been a good solution. But when he started out comparing their database, he found awful problems. For example there are whole names in the surname field or names of a Doctor's office. You can't properly search and even if you do, you are never sure you got the right one.
Moreover, Threema's server is closed-source and so completely proprietary - and you could argue that Signal's server is often closed-source too, given years occasionally go by without public code releases.
This is the rationale.
As for Threema, true enough as it's useless without a server. But again, federation isn't a necessary condition for being open.
Correct me if I am wrong, but as far as I understand you can't make any changes to the Signal client, compile it yourself, and connect to the Signal network. You have to use the binaries from the app store.
The server is open source technically, but it's not federated. They have also not published updates in the past for months while deploying them on the server (probably to prevent people from finding out that they were testing some feature).
One of the bright lights on horizon is that the Bundeswehr opted for a open-source, federated, multi-platform and secure messaging framework. Instead of some proprietary, closed-source piece of crap from a Big-IT vendor which make same depending in a negative way.
That being said, god save the EU, since these walking tax-money black hole is now leading the whole EU.
Why do I emphasize "seem". Well there have been several German initiatives for using open source, but non of them stuck very well. Munich's going Linux comes to mind, but there were others. And I'm afraid that this may be another such "attempt", while I hope it this time different as their national security is a at stake.
Telling everyone to communicate with GPG-encrypted emails has shown to be too hard on users, who then simply use one of the many less-secure channels. You have to do something, or you know they --the US mostly (WhatsApp, Twitter, GMail/Chat) -- will listen along with everything.
How is that different when it comes to Matrix/Elements vs proprietary apps? Maybe this time there's not so much lobbying and more "user just choosing a different communication channel" than they are told to use (as it's UX is so much worse).
I wouldn't call it a "Russian" system. Just consider where Durov currently resides and has his wealth.
Letting some random company operate your army's IT infrastructure: what could possibly go wrong?
[1] https://esut.de/2020/05/meldungen/cyber-it/20897/digitales-g...
The French utilize matrix for military operations as well. This isn’t “some random company”.
And aside how do you think that anywhere in the world software for the army is created. It is called military industrial complex for a reason.
"Do what we say, just don't do what we do", as the old adage goes. How painful.
While cards are certainly convenient, they have failed me at very inopportune moments. I’ve also recently witnessed how someone could not book a ticket for a ferry in one of the mostly cashless European states - cash wasn’t an option and they didn’t have a card. This was at the official counter at the harbor.
A few month ago, card terminals of a widely used type failed hard in Germany, only cash payment was possible.
Being able to do some purchases anonymously is also a good thing - even if it’s only my wife’s birthday present.
I prefer a society where cash is an option for all (in-person) transactions. And preserving that requires exercising the use of cash.
Encrypted secure communication with (and within) the government, or my medical provider is entirely orthogonal to that.
> A few month ago, card terminals of a widely used type failed hard in Germany, only cash payment was possible.
This exactly is part of my point.
> or my medical provider is entirely orthogonal to that.
I prefer a medical provider that does a good job & shares my data, rather than incompetent medical staff that adhere to privacy policies. I expect my doctor to be a good doctor, not a good data policy keeper.
The basic problem with messaging and voice/video comm applications is that clients are not interoperable. It is easy to think that: we've had CUSeeMe, IRC, ICU, AOL Instant Messenger, Tivejo, MSN Messenger, I think more than 10 kinds of Google Chat, Facebook Messenger, Skype, Zoom, Paltalk, Yahoo Messenger, Signal, Telegram, Go2Meeting, Discord, WhatsApp, WeChat, etc.
The average person would be hard pressed to tell the difference between these applications, a cynic would say "Facebook Messenger is no different from AOL Instance|MSN|Yahoo messenger except it is integrated with Facebook". The average person doesn't question that chat programs don't interoperate but because they don't we see a pattern of "try out the new shiny, it's just as good as the old cruddy was back in the day", the new application rides high for a while, then it rots and it is it the new old cruddy before long. The one constant is that you may need to install 10 chat applications to talk to everybody you talk to.
As it is, two-sided markets let applications coast and generally rot without losing market share until things get catastrophically bad. If chat applications interoperated there would be a robust market for better applications and better servers and you'd see developers of old apps to have a reason to keep them working over time and more chances for new apps to get established.
[0]: https://en.wikipedia.org/wiki/XMPP#Non-native_deployments
Meanwhile, BWI is helping fund the work needed to address clientside controlled room membership (https://github.com/matrix-org/matrix-spec-proposals/pull/391...) as highlighted in your paper, as well as TOFU... and they're also funding work to provide MLS as an option for E2EE in Matrix too[2].
Unsure why you're talking about the unexploitable IND-CCA break :)
[1] https://matrix.org/blog/2022/05/16/independent-public-audit-...
[2] https://www.golem.de/news/bwmessenger-vom-messenger-der-bund...
PS: I talked about the seemingly unexploitable IND-CCA vulnerability because it means Matrix can't give you some security guarantees: It should be fine - we don't have an exploit, only a vulnerability - but it is not clear how to reason to arrive at "there cannot be an exploit". If you care about security guarantees, you care about it.
Good news that BWI is funding a Matrix implementation of the multi-vendor IETF standard MLS group messaging E2EE protocol.
The (translated to English) linked reference doesn't mention MLS, is it correct?
Goodbye Microsoft or Slack -specific chat services. Welcome them to compete with their Matrix client-apps.
And hey, we're in the Matrix finally.
And for that matter, the SMTP app that I don't hate.
Signal is in this list. Isn't this false? The server and clients are here: https://github.com/signalapp
I would say it's a bit disingenuous to put it in the same list as Teams, Slack and WhatsApp though.
It's worth noting that if Mattermost adopted Matrix, like Rocket.Chat has[1][2], the vast majority of these limitations would fall away :)
[1] https://www.rocket.chat/press-releases/rocket-chat-leverages...
[2] https://matrix.org/blog/2022/05/30/welcoming-rocket-chat-to-...
Remember "Tchap" (https://www.tchap.fr/), the French Gov messenger system based on Matrix ? ;-)
I still occasionally get rooms or spaces borked, and that frequency increases if E2EE is enabled.
The current server implementation is not svelte in the least, but that's a problem that's being solved with new server implementations that are already 90% of the way there (look-up Dendrite and Conduit if you haven't heard of them).
So yes, visions where always there but the implementation was indeed always a story
I hope this means more development/funding/documentation of the project :)
I used Element in the past and Matrix is a clusterfuck.
Python server slow, Go server not feature complete. Channels available uninteresting, mostly cryptocurrency. A few porn channels, that's it.
I wish it wasn't so. If anything Matrix is a replacement for IRC, absolutely not email.
Then, I am absolutely NOT installing a Bundes-anything on any of my devices. I can't trust a state that has multiple state Trojans.
If your ask whether a consulting agency earned something between 10 and 100 million on that decision you are most likely right. But I guess overall it was probably the CCC influence on the politics and the population. The club and their members are much more influential than you would expect.
And "Chatkontrolle", i.e. client-side inspection and surveillance of every message, for the unwashed masses.
But honestly all much better than the NSA listening into German state traffic. The big fives are not friends just allies.
Hmm
Never feel alone again!
It must be a very hard slog to get there whilst also upholding your ideals, so kudos to you!
But it requires some degree of technical expertise on the ground to weave together solutions, instead of just buying the Microsoft package with AD and Office.