[thelist] Re: Relational DB Help

rudy r937 at interlog.com
Sat Feb 23 16:55:01 CST 2002


>> I've been told repeatedly that we don't want to store the page content
in
>> the db table itself, that it's better to store a link to the article in
the
>> db table instead. This way, not only do you not weigh down the db with
these
>
>Forgive me but that's a crock of...

from the point of view of database performance, i'd have to agree

there's really no problem storing content in text/memo fields


>The idea is not to store the html in the db if possible, just the text.
>It makes no odds how much text.

true, but this brings up an interesting point

to store the "content" you'll probably want to store *at least* <p> tags,
if not <img> and <hx> and so on

it's kinda hard to separate the html

the way to split it is to have the pages produced by templates which
"include" the content (i.e. the navigation etc. is in the template) and
then the included content is *mostly* text with some simple html

the search function would then be able to find relevant pages, but it would
be a database search, and not something like atomz which indexes static
files

make sense?


rudy




More information about the thelist mailing list