Conversation

Notices

1. lnxw48a1 (lnxw48a1@nu.federati.net)'s status on Tuesday, 25-Aug-2020 16:44:21 EDT lnxw48a1
   It’s been a while since I played with #Cassandra. It was pretty enjoyable, but it is hard to wrap one’s mind around schema-less databases.
   
   Remember when you first learned to use random-access files in a programming language? What did you learn? That you had to firmly set a length for each record and which contents which at what locations of each record.
   
   So #MongoDB, Cassandra, and many other databases which use JSON are a big mental challenge just because that makes them so different.
   1. musicman (musicman@nu.federati.net)'s status on Tuesday, 25-Aug-2020 17:59:15 EDT musicman

      in reply to
      you can have schemas in Cassandra (that has been the case since at least 2011)
      1. lnxw48a1 (lnxw48a1@nu.federati.net)'s status on Tuesday, 25-Aug-2020 18:06:06 EDT lnxw48a1

         in reply to
         @musicman Has it really been that long? When I tried it (and I did enjoy it, if it wasn't obvious), it had no way to make one record's contents [i.e., datatypes, lengths, field order, or even fields used] match another record's contents.
         1. musicman (musicman@nu.federati.net)'s status on Tuesday, 25-Aug-2020 19:43:57 EDT musicman

            in reply to
            just because it existed, doesn't mean that's the version you were using...or that it was obvious how to implement