Re: ISO vs ANSI

In message <d562b81c.0412012255.2cdfd949_at_posting.google.com>, Ruslan Kogan <ruslan00_at_yahoo.com> writes
>Thanks for that guys!!!
>
>The issue of ISO vs ANSI has generated a lot of discussion in our
>office. Many people have differing opinions about which coding
>standard should be used to code the 30 reports we are developing.
>
>It has been suggested that the ANSI standard be used because it is:
>- Newer
>- Easier to debug
>- More efficient
>
>However, some other people beleive that the ISO standard is easier to
>follow and debug.
>
>Also, it was mentioned that the ANSI standard has bugs....does anyone
>have more details on these? Are there any known bugs with the ISO
>standard?
>

I don't know whether the standard has bugs[1], but a specific vendor's implementation may have bugs. The bug may be visible (i.e. something breaks if you use the new syntax) or more dangerously it may be invisible and silently return the wrong results. I've not encountered either, but they do exist.

>Are there things which can be done with the ANSI standard which cannot
>be done using ISO?
>

With the (+) notation[2] you can only have one level of outer join (e.g. customers with or without orders), whereas with the ANSI syntax you can have arbitrary levels (e.g. customers with or without orders with or without payments.)
>Any views and opinions would be greatly appreciated???
>

Oracle are unlikely to withdraw (+) syntax anytime soon, but I think ANSI is the way to go.

1 Can a standard have a bug? Surely if it is in the standard it can't be a bug. ;-)

2 What ISO standard is this part of? I always thought it was Oracle proprietary syntax.

-- 
Jim Smith
Because of their persistent net abuse, I ignore mail from
these domains (among others) .yahoo.com .hotmail.com .kr .cn .tw
For an explanation see <http://www.jimsmith.demon.co.uk/spam>