August 8, 2019 at 4:21 pm
>> It's funny how you chastise people for using 1950's technology (which was actually awesome and still has seriously good application in the modern world that a lot of today's "kids" under the age of 50 don't grok) and then refer to a book written in 1834 to try to explain your stance. <<
I'm glad I got my laugh 🙂 wait until I start posting things about Euclidean geometry and go back to the original source material. I simply picked the first book that had a definition of number in it sitting on my shelf. That happens to be one of the oldest and most acceptable definitions. I think it goes back to "The Analytical Arts" a book from the 1500s that was about to introduce those newfangled Arab numerals.
>> They can be used to control fragmentation is wide clustered indexes, act a surrogate keys to protect the natural key(s) of a table from public view and to make changes of the natural keys a non-issue when it comes to relations that depend on the keys, can be used as tie breakers for unique indexes that wouldn't otherwise be unique, and a whole bunch of other things that are based on the math behind integer datatypes. <<
None of those things have anything to do with the valid data model that were trying to build and then implement in a particular product. For example, if I'm using Teradata, I don't have indexes because of that database is based on hashing. I'm talking about building a valid data model and then implementing it; you're talking about implementation first and design later.
>> While I agree with Scott Pletcher that it's stupid to add (for example) an IDENTITY column to every table without reasonable cause (especially if there's a well documented key such as you'll find in the ISO world and world of telephony) ... <<
That particular design problem is a result the mindset that has not understood the difference between the abstract model and in implementing it that physical product. They are still building sequential files with record numbers. Grant. Grant. Grant
>> For example, remember "Push Stacks" for converting Adjacency Lists to Nested Sets? <<
Oh yeah 🙂 that was straight out of a freshman data structures book. I like your postings, along with Ben Gan's and Adam's stuff. I was surprised that the XML , Xquery, etc guys did not come up with something.
Please post DDL and follow ANSI/ISO standards when asking for help.
Viewing post 31 (of 30 total)
You must be logged in to reply to this topic. Login to reply