Of course it helps to have a DBA around as there is a whole other tier of code management needed around procedures. Not least publicising that they are available to use!!

Speed may be less of an issue than the convenience of a solid set of calls that can be made for data. That said the nastiness of doing a SQL transform of a table using cursors falls far behind the relative ease of doing the same thing in PHP.

I’m a bigger fan of views though and don’t think enough use is made of them. Especially when moving one database’s schema to another or in developing new applications. One of the great things is that they can prevent duplication of data whilst giving new structures – brilliant for prototyping as well.

On balance though keeping away from mySQL stored procedures seems like a good idea. The support is nowhere near as good as that for SQL Server, but they are very different products.

Just seen your blog on SPs. I would also add

versioning – SPs are quite to hard to put in the same repository as the source code – finding who changed what and when can be tricky.

Performance problems – SPs give a performance boost because they store a pre-compiled query plan – They need to be recompiled periodically to become aware of new indexes or data distribution changes

Did you put any glue between SVN and bugz?

Be good to get hold of it if you have ..

Ta

N

Also, would this work on both windows and non windows world?