ontologically impaired

  • 0 Posts
  • 15 Comments
Joined 1 year ago
cake
Cake day: August 1st, 2025

help-circle


  • Actually, in the Matrix ecosystem it’s going quite ok I’d say since so many of the large users are European governments. Various parties are funding the research and development of additional features which are then fed back into the Matrix spec or become important FOSS components in the ecosystem. It’s not perfect yet, but it works for the most part. So there is money, bit is usually does not flow into Matrix foundation membership but into the standard or the software.


  • Yea Synapse is very resource hungry. Element built a much more performative enterprise server implementation, but though it’s open source it comes with a proprietary license, which kind of defeats the point when it comes to digital sovereignty in the public sector. From what I heard they unfortunately concluded that this is the only way to prevent organizations from freeriding, but it still sucks for the ecosystem. I’m working on one of the larger European Matrix public sector projects and we’ve been trying to mobilize funds for upstream contributions for some time, but it’s not easy to convince the policy and finance people to approve sustainable contributions to an open source ecosystem; they want to see procurements where it’s 100% clear what you get for your money. Also, public sector communication use cases frequently have high user count, low message volume and a high number of federated instances; so the current Synapse implementation has mostly been workable as you can feasibly scale horizontally. So there is currently not really that much financial incentive and pressure from the public sector to improve the efficiency of Synapse.




  • I know and what I’m sayin is that this is by design. It’s not only about competence, competence can be bought. It’s structurally difficult because various different ministries and independent bureaucracies would need to decide to cooperate and do tedious work to combine data sets like this. It’s not a magical antidote but a structural impedance that is supposed to make abuse more difficult, obviously not impossible.

    And BundID does not entirely change to his. BundID does not assign you a unique identifier either. You get a temporary process-based ID that is generated as you apply for a government service and that is forwarded throughout the process. BundID itself only has a basic user name that you pick yourself and that is not used or stored anywhere outside it. The actual authentication is done via the eID, Elster certificate, etc. which all don’t leave cryptographic traces.

    Don’t get me wrong, this system is still kind of broken. BundID still bundles important information in a central location and urgently needs modernization. And it’s authentication amounts to ‘trust me bro’ from the POV of any other government service. But it does not introduce a global unique identifier either.


  • Depends how much you trust the state. After the Nazi period, the German state is actually not allowed to keep persistent identifiers on people. You can only identify any. given citizen through a domain ID that may not be shared or used beyond a specific domain (e.g tax ID) or a combination of changeable attributes (address, name, etc.).

    In practice this can be annoying, but it’s also a useful precaution to prevent the government form easily pooling and linking information on citizens across domains and levels of state.



  • Haha I work in German government IT and have a colleague who likes to point to Denmark as an example for how good things could be here. I’ll show him this next time he needs to put down the rose glasses.

    That said, it’s the same in Germany, only that citizen data is kept with municipalities, so we have ca. 11.000 different such shitty systems built decades ago instead of one - which makes it less attractive and damaging to compromise any given one, I guess.

    At least, however, we don’t have a citizen ID such as the one you describe, that could be used to identify people uniquely.


  • “I don’t understand the added value here,” one official wrote on an internal staff message board, seen by POLITICO. Another described the tool as “absolute shit.”

    A third official said that, judging matters on how well Element Pro works compared to Teams, the U.S. would “instantly win the war” if it ever decided to switch off its technology. All the officials were granted anonymity to speak freely.

    How is this a credible and justified critique and why is it worthy of a Politico article?

    The first two should never have gone to print, these are unqualified personal options. The third one somehow makes the claim that, because Element does not functionally behave like teams, the US would somehow be able to switch it off more easily in a war scenario…? That isn’t even coherent logic on the surface. And none of the three statements make sense when you consider that the tool is supposed to improve data security and operational resilience.

    They should not have quoted random user feedback. The critique in this article, which has somehow made it into the headline, has about the same substance as a random angry Google or Yelp review.




  • People just drink less beer, period. People now have jobs that are less compatible with being slightly daydrunk, people don’t spend their evenings in bars but on their smartphone, the adverse health effects of alcohol are widely known, etc. Some old breweries in Germany have pivoted to alcohol free alternatives such as Spezi or malt beer. But small breweries that did not make such a change early enough likely suffer a lot under this market pressure.