5 ms·

>> The web uses files extensively. > Not these days. > Almost all web content is dynamically generated from one kind of indexed data store or another. Which

>> The web uses files extensively. > Not these days. > Almost all web content is dynamically generated from one kind of indexed data store or another. Which are generated from files. >> Any page is a file. > This isn’t true. But even if it were, I doubt you’d convince rektide that this would meet his definition of ‘files’. You are correct here, but every web page is generated from a file or bunch of files. And there is still static websites.

zepto 5y ago
> Which are generated from files. No they are not. >> Any page is a file. >>> This isn’t true. But even if it were, I doubt you’d convince rektide that this would meet his definition of ‘files’. >> You are correct here, but every web page is generated from a file or bunch of files. This is simply untrue. Hacker News for example is made of web pages which are not generated from a file or bunch of files. > And there is still static websites. So what? Nobody said that no websites are based on files. Only that most web pages are not.
- ghoward 5y ago
  >> Which are generated from files. > No they are not. An indexed data store is still some file or files on a filesystem. Databases are not exempt from creating files unless they are working with raw disk. > This is simply untrue. Hacker News for example is made of web pages which are not generated from a file or bunch of files. Where is the information stored for Hacker News? I suspect it's in a database that stores its data in files. Every HN page is generated from data, which is stored in files. In general, pages are generated from data which is stored in files. >> And there is still static websites. > So what? Nobody said that no websites are based on files. Only that most web pages are not. "Most" is doing a lot of heavy lifting. Is it "most" as in, most visited? Or most pages by volume?
  - zepto 5y ago
    > In general, pages are generated from data which is stored in files. So what? Nobody is argue that computers don’t store data in files. Only that web pages are not files. >> And there is still static websites. > So what? Nobody said that no websites are based on files. Only that most web pages are not. "Most" is doing a lot of heavy lifting. Is it "most" as in, most visited? Or most pages by volume? It doesn’t matter. It’s true of both.
    - ghoward 5y ago
      > So what? Nobody is argue that computers don’t store data in files. Only that web pages are not files. You conceded that all data (modulo a little) is stored in files, so you seem to be arguing for hiding that fact from users. To me, hiding the fact that files exist from users smacks of the same sort of elitism that argues against Right to Repair. Even if a user does not know how to handle their data, they should have the ability to get a third party who can do it for them. Locking down user data is, in my mind, one of the unethical things our industry does. > It doesn’t matter. It’s true of both. Honestly, almost all websites either use JavaScript or CSS, which is stored in files, so even if a website isn't static, they almost all still use static files. Also, I would argue that even a page generated from a single PHP file qualifies as a webpage being a file.
      - grandpa_on_lawn 5y ago
        All computers use transistors. Should we expose them to the user? Should we expose to the user the system libraries used by the OS? Should we expose the source code of the OS? Because the OS is also made from a bunch of files.
        ghoward 5y ago
        Did the user create the OS, libraries, or the transistors? By the user's actions, their files were created. They own that data. We should expose it to them. The other things, they probably did not create. We don't need to expose that. Nice whataboutism, but there is a difference. Edit: Also, yes, we should give the user access to the OS and system libraries. They own the computer, and we should allow them to do with it what they wish.
        rektide 5y ago
        Yes on all accounts. For users that want such. QEMU exposes the transistors. There are even better OpenFPGA grade simulation tools that we are making to even more detailedly model the transistors. Many chips have very extensive debug interfaces we can use to expose the inner workings of systems. Users should have access. We should absolutely expose the system libraries. On Android this has been an essential tool for many users to improve their daily existence. It's a huge point of controversy that Google has hired the Magisk developer who did just that, and that now the library is giving up some of it's former skills to help users manipulate system libraries. https://9to5google.com/2021/08/29/magisk-android-changes-google/ https://9to5google.com/2021/08/29/magisk-android-changes-goo... Yes, we should definitely expose the source code. I'd love it if more programs were actively debuggable. Debian for example makes it fairly easy to download debug packages, with all the debugging symbols. The kernel should be accessible. Treating users like end users is not cool & keeps humanity from getting good, from learning & understanding the world. Even if few do take up the option, the option to explore the world about us is something computing ought value.
  - a1369209993 5y ago
    > Databases are not exempt from creating files unless they are working with raw disk. There's no "unless"; /dev/sda is a file (as is eg /qemu/vm1234/disk/sda).
    - ghoward 5y ago
      Good catch. I'll take the L for being too cautious on that one.

zepto 5y ago

> Which are generated from files. No they are not. >> Any page is a file. >>> This isn’t true. But even if it were, I doubt you’d convince rektide that this would meet his definition of ‘files’. >> You are correct here, but every web page is generated from a file or bunch of files. This is simply untrue. Hacker News for example is made of web pages which are not generated from a file or bunch of files. > And there is still static websites. So what? Nobody said that no websites are based on files. Only that most web pages are not.

ghoward 5y ago
>> Which are generated from files. > No they are not. An indexed data store is still some file or files on a filesystem. Databases are not exempt from creating files unless they are working with raw disk. > This is simply untrue. Hacker News for example is made of web pages which are not generated from a file or bunch of files. Where is the information stored for Hacker News? I suspect it's in a database that stores its data in files. Every HN page is generated from data, which is stored in files. In general, pages are generated from data which is stored in files. >> And there is still static websites. > So what? Nobody said that no websites are based on files. Only that most web pages are not. "Most" is doing a lot of heavy lifting. Is it "most" as in, most visited? Or most pages by volume?
- zepto 5y ago
  > In general, pages are generated from data which is stored in files. So what? Nobody is argue that computers don’t store data in files. Only that web pages are not files. >> And there is still static websites. > So what? Nobody said that no websites are based on files. Only that most web pages are not. "Most" is doing a lot of heavy lifting. Is it "most" as in, most visited? Or most pages by volume? It doesn’t matter. It’s true of both.
  - ghoward 5y ago
    > So what? Nobody is argue that computers don’t store data in files. Only that web pages are not files. You conceded that all data (modulo a little) is stored in files, so you seem to be arguing for hiding that fact from users. To me, hiding the fact that files exist from users smacks of the same sort of elitism that argues against Right to Repair. Even if a user does not know how to handle their data, they should have the ability to get a third party who can do it for them. Locking down user data is, in my mind, one of the unethical things our industry does. > It doesn’t matter. It’s true of both. Honestly, almost all websites either use JavaScript or CSS, which is stored in files, so even if a website isn't static, they almost all still use static files. Also, I would argue that even a page generated from a single PHP file qualifies as a webpage being a file.
    - grandpa_on_lawn 5y ago
      All computers use transistors. Should we expose them to the user? Should we expose to the user the system libraries used by the OS? Should we expose the source code of the OS? Because the OS is also made from a bunch of files.
      - ghoward 5y ago
        Did the user create the OS, libraries, or the transistors? By the user's actions, their files were created. They own that data. We should expose it to them. The other things, they probably did not create. We don't need to expose that. Nice whataboutism, but there is a difference. Edit: Also, yes, we should give the user access to the OS and system libraries. They own the computer, and we should allow them to do with it what they wish.
      - rektide 5y ago
        Yes on all accounts. For users that want such. QEMU exposes the transistors. There are even better OpenFPGA grade simulation tools that we are making to even more detailedly model the transistors. Many chips have very extensive debug interfaces we can use to expose the inner workings of systems. Users should have access. We should absolutely expose the system libraries. On Android this has been an essential tool for many users to improve their daily existence. It's a huge point of controversy that Google has hired the Magisk developer who did just that, and that now the library is giving up some of it's former skills to help users manipulate system libraries. https://9to5google.com/2021/08/29/magisk-android-changes-google/ https://9to5google.com/2021/08/29/magisk-android-changes-goo... Yes, we should definitely expose the source code. I'd love it if more programs were actively debuggable. Debian for example makes it fairly easy to download debug packages, with all the debugging symbols. The kernel should be accessible. Treating users like end users is not cool & keeps humanity from getting good, from learning & understanding the world. Even if few do take up the option, the option to explore the world about us is something computing ought value.
- a1369209993 5y ago
  > Databases are not exempt from creating files unless they are working with raw disk. There's no "unless"; /dev/sda is a file (as is eg /qemu/vm1234/disk/sda).
  - ghoward 5y ago
    Good catch. I'll take the L for being too cautious on that one.