Migration of DB2 for z/OS systems to a new release is a very important task that requires careful planning and systematic execution, which takes time. Yet pureXML in DB2 9 for z/OS is very attractive to application developers, but it requires DB2 9 NFM. If you have a need to start an application development using pureXML and cannot seem to wait for the DB2 systems to get to DB2 9 NFM, you may want to consider the following option (which to many is not news): developing the new application using DB2 Express-C using the pureXML features and IBM Data Studio on a Windows or Linux platform (both are no charge). You may get in sync with the migration to DB2 9 NFM for application system testing, performance testing, and deployment. Here is some more info.
You also need to download the right version of DB2 Express-C for your platform. If you have a Windows platform, look at this page to determine if your Windows is 32-bit or 64-bit. If the properties of your computer does not say 64-bit, then it's likely to be 32-bit.
You also need an IBM ID to download for license agreement. Getting an IBM ID is technically very easy. Use IBM Installation Manager to download and install, which is faster. After download, you need to unzip it and run the setup.exe or something like that. It should be fine to use the typical installation type. Extremely easy :-).
To develop using SQL/XML, Java, stored procedures, and pureQuery, you can start with the Data Studio Developer no-charge features. After you install DB2 Express-C and launch it, you can download Data Studio Developer and install it. The link is listed on the DB2 Express-C download page also. Again, it's straight-forward.
Using Data Studio Developer to explore SQL/XML queries
This is my Windows XP experience. Launch the Data Studio Developer, you will see the SAMPLE database already listed in the Data Source Explorer, if you launched DB2. Right click on the SAMPLE DB, and choose "connect". Use your Windows username and password to connect. Once it's connected, you can look into "Schemas", and under your username, find "Tables". There are a few tables in the SAMPLE DB that contains an XML column: CUSTOMER, PRODUCT, and PURCHASEORDER that contains some sample data. You can list the table content by right click on the tablename -> "Data" -> "Returns all rows" or "Sample content", to display table content.
From SAMPLE DB name, right click on it -> "New SQL Script", you can copy and paste an SQL/XML query into the window. Then Choose main menu "Run" -> "Run SQL". The status and result will show. Try this query:
SELECT *FROM PURCHASEORDERWHERE XMLEXISTS('declare default element namespace "http://posample.org"; $po/PurchaseOrder[@Status="Unshipped"]' PASSING PORDER AS "po")
Remember, in order to move the queries and applications to z/OS easily, you need to stick to the syntax and scope that is supported by z/OS pureXML (and also by LUW). One of the key detailed differences is that in LUW, you must use a variable, such as $po shown above, as the starting context item for an XPath path expression, which is optional for z/OS. This is consistent for XMLQUERY, XMLEXISTS, and XMLTABLE functions.
Once you see the XML column data, you can click on the icon, and an XML editor window will open.
You could open a new Data Project, from main menu "File" -> "New" -> "Data development Project". Say you created Project1. Within Data Project Explorer, under Project1, right clock on "SQL Scripts" -> "New" -> "SQL or XQuery Script", you can get a SQL Script window also.
If you need to have a Java project, you need to create a project with the corresponding type: from main menu "File" -> "New" -> "Project", a dialogue window will open...
I hope you will like the combination of Data Studio with pureXML, if you haven't tried yet. There are much more features to explore in IBM Data Studio.