|
|
Posted by Justin Koivisto on 02/23/06 23:58
d wrote:
> "Dale" <dale.drinkard@gmail.com> wrote in message
> news:1140703564.436036.79490@e56g2000cwe.googlegroups.com...
>> I'm not sure I'm even thinking about this the right way, but here goes:
>>
>> I have a table of users. Each one of these users may be associated
>> with none, one, or many records in another table I call a todo table.
>> table user = 'id', 'name', 'bla bla bla','todo_list'
>> table todo = 'id','title','other info'
>>
>> Suppose user 'id=1253' has on his todo list items 756,241, and 672. I
>> create a string 756,241,672 and store that string in the user's
>> todo_list field. Then when I want to display the todo items I get the
>> string with a query, bust it up into an array, iterate the array and
>> query the todo table.
>>
>> I have a gut feeling I'm making it way more complicated than need be.
>> But I can't think of any other way to do it
>
> You're nearly there. Instead of storing a comma-seperated list of IDs, use
> a third table, say "chores" or something that fits in with your data. Give
> that table 3 columns - an ID, a user ID, and a todo ID. To link a user with
> a task, insert their ID and the task's ID into that table. To get the
> user's tasks, simply join the chores table with the todo table, and search
> for a particular user ID. That's a much more efficient way of storing such
> relationships.
>
> Is that cool?
I usually use a 2-col table and make a unique index with the two fields
to prevent dups. ;) It works quite well, but it makes things just
slightly more tedious when deleting tasks or users because you now need
to look into additional tables for cleanup.
--
Justin Koivisto, ZCE - justin@koivi.com
http://koivi.com
Navigation:
[Reply to this message]
|