Re: plsxml setup yields errors with checkdeps.sql

"Gabriel Gonzalez" <no-spam_at_no-spam.com> wrote in message news:3e4aadf6$1_2_at_binarykiller.newsgroups.com...
> > Oracle8i Enterprise Edition Release 8.1.7.0.0 - Production
>
> Two things:
>
> 1) The latest version of 817 is 8174, so first suggestion is to apply the
> latest patch...
>
> 2) Also, if you upgraded the DB or applied a patch, did you make sure that
> everything compiled/upgraded correctly? Maybe some of your Oracle DB
> objects are out of sync (ie, you have version 1 of a certain package, but
> version 3 of another)
>
> Also,
>
> There are many ways of tying Oracle and XML. It all depends heavily on
what
> you are trying to do, and how. So if you want more help, let us know what
> the structure of the existing system is.
>
>
>
>
> -----------== Posted via Newsfeed.Com - Uncensored Usenet News
==----------
> http://www.newsfeed.com The #1 Newsgroup Service in the World!
> -----= Over 100,000 Newsgroups - Unlimited Fast Downloads - 19 Servers
=-----

Gabriel,

Thanks for the quick response. I will indeed patch the server. The install is a clean one and I do not think that we have upgraded the server from a prior version. Further I did not notice any compilation errors when installing nor when loading the plsxml cartridge.

The existing application is written with a VB front end connecting to VB COM+ objects via HTTP and ASP. The middle tier is broken into two components a Business logic layer and a Data Access layer. The original back end was written on MSSQL Server 2000. The MSSQL Backend makes extensive use of the FOR XML EXPLICIT Construct. All of the data from the database is streamed out of the back end with an ADO stream. The middle tier is fairly light weight and basically does little data manipulation. A basic interaction goes like this... The front end builds an XML Document and wraps in a checksum and request tag. This request is posted via HTTP to an asp page on the middle tier web server. The asp page instantiates objects verifies the Check sum for authenticity. Assuming the data is valid the Request tag is stripped from the document and the rest of the data is passed to the backend via the data access dll. This layer is basically a black box pass through. It will establish a connection to the database via an .udl file. It will then read the information passed from the business logic layer so it knows what procedure to call and pass in the incoming XML document as the procedures only parameter. The procedures then parse the XML Document using sp_xml_prepapredocument and go about doing the tasks specified in that procedure. Each procedure that receives XML is basically a wrapper procedure that will then call off to worker procedures. Each worker procedure is responsible for returning a portion of the completed XML Document. For example. To get system start up information. we select the opening tag into the XML stream in this case <Root> We then select a <Parameters> Tag if needed and fill in the Parameters Info. After that we Select a <Data> tag into the XML stream. Once the foundation is in place we will make a call to a worker proc that contains some of the start up information. So the first procedure will return the system options for the program In a <SystemOptions> tag. Then another worker is called that returns information regarding user defined datasets in a <Datasets> Tag. This continues until all of the information is included and sent to the ADO stream at which time we select the closing tags into the stream and the procedure terminates. This approach has one big benefit. It allows us to break the XML Documents up into components. So If I need to get the system options I can use the same procedure that is used in the System Start up procedure. Same goes for datasets the user can see. If he needs a list of available datasets for him to work in I call the same worker proc from a different wrapper and the XML always comes out the same. Once the stream gets back to the data access layer it is passed back to the business logic layer that adds a response tag and a new checksum tag for the returning information. This information is then passed back to the application via HTTP as a response from the asp page called by the application.

Our goal is to be able to rewrite the backend on Oracle and swap out the MSSQLDA.dll (Data Access dll) with an ORADA.dll and keep the rest of the application exactly the same as it currently is

Unfortunately we are having trouble achieving the same level of control over the resulting XML document as was provided by FOR XML EXPLICIT in MSSQL 2K. This is why we are looking into plsxml as DBXML.Query looks like it might do the trick if it performs its task in a reasonable time and manner.

IF you need further information or code samples from the existing system I'll try my best to get the information to you. Any thoughts you may have on this would be a great help.

Regards,
Bill Received on Wed Feb 12 2003 - 17:01:28 CST