Pathologically Eclectic Rubbish Lister | |
PerlMonks |
Re^5: Conditional many to many relationships with Class::DBIby cbraga (Pilgrim) |
on Nov 01, 2004 at 16:05 UTC ( [id://404333]=note: print w/replies, xml ) | Need Help?? |
My point is that as long as MySQL ignores the SQL standard and changes my data behind my back it's only good for blogs and nothing else.
And I disagree about your point on overhead and reducing database load by putting the "versioned data" on the SQL statement. As long as I'm specifying ... AND VERSION=xxx ... it's on the SQL statement, no need to have a new SQL dialect. Furthermore that's a case where the performance bottleneck is the database (for large databases), not the application or versioning code, so writing it in C won't make it noticeably faster. Of course you're right about only having to "solve the problem once", but that's easy. Write the version control code as stored procedures. In PostgreSQL and Oracle you can even do that in Perl. But, oh, you're using MySql, sorry. ESC[78;89;13p ESC[110;121;13p
In Section
Seekers of Perl Wisdom
|
|