>yes, I have a database

>yes, I have a database

A Conspiracy Theorist Is Talking Shirt $21.68

DMT Has Friends For Me Shirt $21.68

A Conspiracy Theorist Is Talking Shirt $21.68

  1. 2 months ago
    Anonymous

    Now store passwords in it.

    • 2 months ago
      Anonymous

      just encrypt the password first
      whats the problem

      • 2 months ago
        Anonymous

        >encrypting passwords
        ngmi

        • 2 months ago
          Anonymous

          who cares?? noone is gonna break out a quantum computer and hack your passwords

      • 2 months ago
        Anonymous

        hash*

    • 2 months ago
      Anonymous

      Literally just hash them and store them. It's not any less secure than storing them in a real database.

  2. 2 months ago
    Anonymous

    >now convince me that it is indeed worse than those "real" "databases".

  3. 2 months ago
    Anonymous

    >no indexing
    >no concurrent operations
    you're fricked kiddo

    • 2 months ago
      Anonymous

      entire banking system probably runs on some vbscript linear searches

      • 2 months ago
        Anonymous

        It runs on DB2 on the mainframe you dumb zooomer Black person

      • 2 months ago
        Anonymous

        Nah, everything banking and big industry (airlines, payroll, e.t.c.) runs on virtualised big iron mainframes, shit is often still written in COBOL. The old mainframes were purpose built for that kind of number crunching over general purpose computing. Had instructions specifically for tabulation, queries, sorting and other similar operations in silicon. Dunno about the newer mainframe architectures from IBM, but I imagine they’re still a similar sort of design.

    • 2 months ago
      Anonymous

      honestly, nonissues. I created a new small data software at work recently and just implemented indexing adjacent to the, effectively deserialized "json" data. all writes just get written out by one thread and fsync'd.

      it works fine.

    • 2 months ago
      Anonymous

      >copy
      >write
      >then merge

      • 2 months ago
        Anonymous

        Two concurrent writes overwrite each others change

        • 2 months ago
          Anonymous

          you can do as others say. NeDB is a good example of how you can do it concurrently. basically shove partials into it and compact it at random times or startup.
          last write wins. not like you can have concurrent updates to the same "data" as in row, or whatever anyhow without some kind of race. in compact, last write wins for that data, obviously.

    • 2 months ago
      Anonymous

      >>no concurrent operations
      >you're fricked kiddo
      Concurrent operations can be supported depending on how you manage file access and concurrency.
      Boomertard.

      • 2 months ago
        Anonymous

        >how you manage file access and concurrency.
        sqlite already does this. why are you reinventing the wheel?

        • 2 months ago
          Anonymous

          >why are you reinventing the wheel?

          Sqlite- released August of 2000
          Json- released 2001.

          Frick you're moronic.

          • 2 months ago
            Anonymous

            I said sqlite because that's the best and lightest available today but relational databases with acid transactions have existed since the 1980s.

          • 2 months ago
            Anonymous

            >acid transactions
            sounds painful

  4. 2 months ago
    Anonymous

    holy based.
    most storage devices are fast enough and you have enough memory a single json file is probably good enough DB for like 500+ MiB datasets.

    • 2 months ago
      Anonymous

      I have personally seen systems that eventually grew their main json "db" to 10s of GB, and its read from and modified every other minute minute or so

      • 2 months ago
        Anonymous

        then you switch out the storage layer with something better. I've done this as well too.
        usually my next step is dropping in rocksdb which gets you going til storage for your server becomes a blocker.
        sadly you lose a lot of convenience, but that's what it is.

    • 2 months ago
      Anonymous

      >memmove 500mb of dumped json just to add one character

      • 2 months ago
        Anonymous

        nah, you write out partials and then stitch them together on compaction.
        it's really not that big of a deal.

  5. 2 months ago
    Anonymous

    guess it's better than an excel sheet

  6. 2 months ago
    Anonymous

    protobuf is superior

  7. 2 months ago
    Anonymous

    For me, it's SleekDB.
    https://sleekdb.github.io/

  8. 2 months ago
    Anonymous

    why not just sqlite

  9. 2 months ago
    Anonymous

    You don't need that much. Just use .csv.

  10. 2 months ago
    Anonymous

    i work in enterprise, trying to imagine the absolute wreckage that would have ensued if those system where build upon noSql causes enough terror to keep me awake for today. Thanks OP, will use the time gained to walk the dog or read a book

    • 2 months ago
      Anonymous

      Funny thing, the first place I saw Redis in production was a bank.

Your email address will not be published. Required fields are marked *