Re: Naming Convention for Columns

From: Todd Boss <boss_at_netcom.com>
Date: 1998/03/11
Message-ID: <bossEpo3Gq.9sx_at_netcom.com>#1/1


In article <3505ad61.266543_at_read.news.global.net.uk>, Keith Boulton <boulkenospam_at_globalnet.co.uk> wrote:
>On Tue, 10 Mar 1998 12:52:07 GMT, jeff_kalchik_at_spammenot.mw.3com.com
>(Jeff Kalchik) wrote:
>
>>
>>While I appreciate your .sig line, I absolutely disagree with your
>>naming convention. I'll fire anyone who pulls that kind of s**t
>>working for me. It creates absolutely unmaintainable systems. I've
>>got enough grief with underscores and dashes not being consistent; I
>>don't need to worry about what's in T3.C47.
>>
>
>I assumed the poster was being ironic.
>
>Having said that, I did work for an organisation once which had DB2
>naming conventions where table and column names were limited to 8
>characters. The first three characters were a system mnemonic,
>followed by a five digit number i.e. select xyz00024, xyz00057 from
>xyz11003.
>

I worked at a large organization with similar sounding conventions for databases and devices...they limited the names to 9 characters and names had to be in this format: xxxxdbyyy where xxxx = application acronym

      db implied "database" (subsititute dv for device)
      yyy = the three letter abbreviation for what the database
            actually was (for devices, substitute numbers starting
            at 001 and incrementing).

Sounds familiar. Perhaps the same sick individual worked at both organizations. If you can imagine trying to create 100 devices using this scheme, and trying not to confuse vdevnos with the device names (it was a sybase app) you can imagine my personal hell. :-) Or trying to ascertain the purpose of a database when the only useful information was a three letter abbreviation of its use.

boss Received on Wed Mar 11 1998 - 00:00:00 CET

Original text of this message