Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!csd4.milw.wisc.edu!dogie!pscaaf@vms.macc.wisc.edu
From: pscaaf@vms.macc.wisc.edu (Ron Peterson)
Newsgroups: comp.databases
Subject: Re: Ingres is making me mad!!
Message-ID: <1229@dogie.edu>
Date: 23 Feb 89 18:55:07 GMT
Sender: news@dogie.edu
Organization: Public Service Commission of Wisconsin
Lines: 13

In article <349@daitc.daitc.mil>, jkrueger@daitc.daitc.mil (Jonathan Krueger) writes...

>In article <1949@tank.uchicago.edu>, dwayne@rover writes:
>>[he has] a questionairre 80% of which are y/n/u questions
>>[the rest] is floating point and character data types.
>>
>>Well, I hit that legendary boundary of 127 columns in a table
>>does anybody have any ideas that might help me get around
>>this (I think undocumented) limitation [without breaking into tables]???

If you will represent the response of several questions as one column using
character strings you will reduce the number of columns that you will need.
"ynuyn" would correspond to 5 questions y n u y n