6 ms·

Connectivity seems like a tough nut to crack. There are many protocols out there which need to be supported, secret sauces gained by many years of experience, w

Connectivity seems like a tough nut to crack. There are many protocols out there which need to be supported, secret sauces gained by many years of experience, working around bugs of many specific devices etc. It's not about having one awesome design, it's about testing with many different devices across many different environments etc. This project seems expensive, risky, with unclear benefits.

londons_explore 2y ago
Apple being apple, they can simply support some newer modern standards and declare the rest unsupported. Mobile networks will be forced to update because no mobile network would survive an iPhone not being able to connect.
- katbyte 2y ago
  They can also keep throwing money at it until it’s solved
- avidiax 2y ago
  Apple has this sort of advantage in numerous standards. Everyone tests their Bluetooth & WiFi against Apple, and iterates until it works perfectly. So this means that Apple effectively gets the right to interpret standards in the way they want, and implement the subset of optional features that they want, and the rest of the industry will conform to them. And if something doesn't work, the user is unlikely to blame Apple. They always blame the other guy. So Apple seems to be one of the only companies that could plausibly homegrow a 5G modem, picking and choosing the parts of the protocol that they want to implement, and perhaps supporting only 5G instead of all the legacy protocols as a sort of "brave" stance. At the very least, their development of such a program could induce their chip vendors to give a price break.
  - londons_explore 2y ago
    Notable that apple did their "esim is the only option" in the USA, but in other countries they were forced to include a SIM slot because they couldn't force mobile networks to support esims. And esims are just a software/provisioning change for most mobile networks. Yet apple couldn't force it to happen. Even now, 5 years on, esims aren't supported on most mobile networks outside the USA and many models of phone don't support them.
    - shiroiushi 2y ago
      Apple has a commanding marketshare in the US and some other rich countries, but in places like India, it's a rather small player. So Apple might be able to force some things in some countries like the US, but it's not going to fly in many other places. The esim thing is a little curious though, because it seems like that should be pretty easy for carriers to support, unlike for instance only supporting 5G. Maybe there's something I don't understand about esim implementation at the carrier level. As an aside, though, esim is really a godsend when traveling. Instead of having to get a local SIM card somewhere and get that set up, there's a bunch of easy-to-use services where you just install an app on your phone and buy a temporary-use esim while you're traveling in a particular country. You can buy the esim beforehand, and then switch to it when your plane lands.
      - Nextgrid 2y ago
        eSIM is to a certain extent harder for carriers than 5G. You need to keep in mind that carriers have near-zero actual engineering capability. They're a marketing & financing operation with a very small technical skeleton crew, with most of the actual network management outsourced to their equipment vendor. 5G is relatively "simple" to support - you give a ton of money to your equipment vendor, they give you new equipment to install and it's done. The vendor happily does all the work for a fee and they can do it fine because it's not their first rodeo. eSIM on the other hand ties into the account & billing management software which is custom for each carrier (and it's a "big ball of mud" of crappy Java accumulated over 2 decades). The vendor can't help here - each carrier has their own ball of mud, so there is no economies of scale possible here. 5G is completely transparent to the "big ball of mud". The differing bits are all handled by the vendor-provided equipment, the ball of mud doesn't know/care. 5G doesn't change any fundamentals - the big ball of mud can treat it as a 2G network. You've still got phone numbers, CDRs, etc. eSIM is a fundamental change though. You go from physical SIMs that must be created & registered in advance, and then linked to an account by some privileged user (CS agent or in-store employee), to "SIMs" that can be created out of thin air and registered by the user themselves. This requires new functionality in the "big ball of mud". The security profile also changes - with physical SIMs the security is provided by the privileged user who is supposed to check the legitimacy of any SIM swap/registration request. With eSIMs this is done online by the user, so there needs to be additional security checks, fraud protection, etc that the existing system (which so far was pretty low stakes as no security-critical action could be done online) suddenly needs to be hardened to support. This would be no big deal for a dedicated engineering team at any decent tech company or even startup, but these companies are as far away from "tech/startup" as you can imagine, and no competent engineer would want to work there for the terrible pay/conditions/pervasive mediocrity that's in such companies. It's the same situation as with legacy banks still not being able to catch up to the functionality & user experience offered by modern "challenger banks" despite having orders of magnitude more resources and capital.
  - shiroiushi 2y ago
    >perhaps supporting only 5G instead of all the legacy protocols as a sort of "brave" stance. Is this even feasible? I thought 5G wasn't really usable outside urban environments because it has a lower range than 4G.
    - iggldiggl 2y ago
      Is that an inherent property of 5G? Or just a side-effect of it being predominantly deployed in high-bandwidth high-frequency settings with the consequent low range resulting from the high frequency?
      - londons_explore 2y ago
        The main benefits of 5g lie in dense urban deployments, but it can totally be deployed in huge rural cells too.
  - gruez 2y ago
    >Apple has this sort of advantage in numerous standards. Everyone tests their Bluetooth & WiFi against Apple, and iterates until it works perfectly. So this means that Apple effectively gets the right to interpret standards in the way they want, and implement the subset of optional features that they want, and the rest of the industry will conform to them. is this any different than any other major vendor? I can't imagine not testing against qualcomm (atheros) or intel's wifi implementation, for instance.
scyzoryk_xyz 2y ago
I can’t imagine that at their scale they haven’t done the math before pursuing a project. They are big enough that the benefits might also include internal know-how, negotiating advantage and long-term security

londons_explore 2y ago

Apple being apple, they can simply support some newer modern standards and declare the rest unsupported. Mobile networks will be forced to update because no mobile network would survive an iPhone not being able to connect.

katbyte 2y ago
They can also keep throwing money at it until it’s solved
avidiax 2y ago
Apple has this sort of advantage in numerous standards. Everyone tests their Bluetooth & WiFi against Apple, and iterates until it works perfectly. So this means that Apple effectively gets the right to interpret standards in the way they want, and implement the subset of optional features that they want, and the rest of the industry will conform to them. And if something doesn't work, the user is unlikely to blame Apple. They always blame the other guy. So Apple seems to be one of the only companies that could plausibly homegrow a 5G modem, picking and choosing the parts of the protocol that they want to implement, and perhaps supporting only 5G instead of all the legacy protocols as a sort of "brave" stance. At the very least, their development of such a program could induce their chip vendors to give a price break.
- londons_explore 2y ago
  Notable that apple did their "esim is the only option" in the USA, but in other countries they were forced to include a SIM slot because they couldn't force mobile networks to support esims. And esims are just a software/provisioning change for most mobile networks. Yet apple couldn't force it to happen. Even now, 5 years on, esims aren't supported on most mobile networks outside the USA and many models of phone don't support them.
  - shiroiushi 2y ago
    Apple has a commanding marketshare in the US and some other rich countries, but in places like India, it's a rather small player. So Apple might be able to force some things in some countries like the US, but it's not going to fly in many other places. The esim thing is a little curious though, because it seems like that should be pretty easy for carriers to support, unlike for instance only supporting 5G. Maybe there's something I don't understand about esim implementation at the carrier level. As an aside, though, esim is really a godsend when traveling. Instead of having to get a local SIM card somewhere and get that set up, there's a bunch of easy-to-use services where you just install an app on your phone and buy a temporary-use esim while you're traveling in a particular country. You can buy the esim beforehand, and then switch to it when your plane lands.
    - Nextgrid 2y ago
      eSIM is to a certain extent harder for carriers than 5G. You need to keep in mind that carriers have near-zero actual engineering capability. They're a marketing & financing operation with a very small technical skeleton crew, with most of the actual network management outsourced to their equipment vendor. 5G is relatively "simple" to support - you give a ton of money to your equipment vendor, they give you new equipment to install and it's done. The vendor happily does all the work for a fee and they can do it fine because it's not their first rodeo. eSIM on the other hand ties into the account & billing management software which is custom for each carrier (and it's a "big ball of mud" of crappy Java accumulated over 2 decades). The vendor can't help here - each carrier has their own ball of mud, so there is no economies of scale possible here. 5G is completely transparent to the "big ball of mud". The differing bits are all handled by the vendor-provided equipment, the ball of mud doesn't know/care. 5G doesn't change any fundamentals - the big ball of mud can treat it as a 2G network. You've still got phone numbers, CDRs, etc. eSIM is a fundamental change though. You go from physical SIMs that must be created & registered in advance, and then linked to an account by some privileged user (CS agent or in-store employee), to "SIMs" that can be created out of thin air and registered by the user themselves. This requires new functionality in the "big ball of mud". The security profile also changes - with physical SIMs the security is provided by the privileged user who is supposed to check the legitimacy of any SIM swap/registration request. With eSIMs this is done online by the user, so there needs to be additional security checks, fraud protection, etc that the existing system (which so far was pretty low stakes as no security-critical action could be done online) suddenly needs to be hardened to support. This would be no big deal for a dedicated engineering team at any decent tech company or even startup, but these companies are as far away from "tech/startup" as you can imagine, and no competent engineer would want to work there for the terrible pay/conditions/pervasive mediocrity that's in such companies. It's the same situation as with legacy banks still not being able to catch up to the functionality & user experience offered by modern "challenger banks" despite having orders of magnitude more resources and capital.
- shiroiushi 2y ago
  >perhaps supporting only 5G instead of all the legacy protocols as a sort of "brave" stance. Is this even feasible? I thought 5G wasn't really usable outside urban environments because it has a lower range than 4G.
  - iggldiggl 2y ago
    Is that an inherent property of 5G? Or just a side-effect of it being predominantly deployed in high-bandwidth high-frequency settings with the consequent low range resulting from the high frequency?
    - londons_explore 2y ago
      The main benefits of 5g lie in dense urban deployments, but it can totally be deployed in huge rural cells too.
- gruez 2y ago
  >Apple has this sort of advantage in numerous standards. Everyone tests their Bluetooth & WiFi against Apple, and iterates until it works perfectly. So this means that Apple effectively gets the right to interpret standards in the way they want, and implement the subset of optional features that they want, and the rest of the industry will conform to them. is this any different than any other major vendor? I can't imagine not testing against qualcomm (atheros) or intel's wifi implementation, for instance.

scyzoryk_xyz 2y ago

I can’t imagine that at their scale they haven’t done the math before pursuing a project. They are big enough that the benefits might also include internal know-how, negotiating advantage and long-term security