Level Extreme platform
Subscription
Corporate profile
Products & Services
Support
Legal
Français
ADO vs. SPT for MTS objects
Message
From
26/10/1999 15:43:27
 
General information
Forum:
Visual FoxPro
Category:
Internet applications
Miscellaneous
Thread ID:
00281738
Message ID:
00281757
Views:
14
>Okay, my understanding of it is that to use OLE DB, we FoxHeads need to use ADO, right? Is it worth doing that just to avoid using ODBC? The backend will be SQL Server.

No, but ADO is the accepted interface to OLE DB providers. Good questions. Why is OLE-DB prefered over ODBC?

Will you be passing RecordSets from the MiddleTier to the front end, or will you be using XML?

Will you be doing alot of data manipulatin in the middle tier once retrieving the data?

Or, will you just be running Stored Procedures on the SQL server and passing the results to the front end?

I think the implementation depends on what you are doing. For example, if you are going to be doing alot of processing of the data in the middle tiers, you don't want the over head of COM, which is what ADO is. It requires you to go throught COM plumbing for all your data access. Where as, if you use SPT, the result is in a VFP cursor which is manipulated directly by VFP, much faster.

If you have to return data, you can use XML also, you might want to prototype a fairly large dataset/processing... do it a few different ways...

1. SPT, Process VFP Cursor, Return XML (uses ODBC)

2. SPT, Process VFP Cursor, Return RS (Uses ODBC)

3. ADO, Process ADO RecordSet, Return RS (no ODBC)

4. ADO, ConvertToCursor, Process VFP Cursor, Return XML (no ODBC)

5. ADO, ConvertToCursor, Precess VFP Cursor, ConvertTOADO, Return RS (no ODBC)

Depending on what you are doing, certain of the above will have different advantages. My educated guess is, with a fiarly large dataset, and a fairly complex process, number 1 will return data the fastest, even though it uses ODBC.

But, number 3 may have a speed advantage over 2 if the client requires a RecordSet and you are only getting data, and not processing it in the middle tier. I think 2 will beat out 3, and by more the bigger the dataset and more complex the process.

But, as you say, ODBC is rulled out, scenerio 4 or 5 are all that you can choose from. I would expect 5 would be faster, since createing XML will be faster than building a RecordSet.

BOb
Previous
Next
Reply
Map
View

Click here to load this message in the networking platform