- ProIV Resource Centre
- → Viewing Profile: Posts: mdexter
Community Stats
- Group Members
- Active Posts 70
- Profile Views 3,967
- Member Title Advanced
- Age Age Unknown
- Birthday Birthday Unknown
-
Gender
Male
Previous Fields
-
First Name
Mark
-
Surname
Dexter
-
Nationality
USA
-
Year Started ProIV
1985
-
Highest ProIV Version Used
5.5
-
Company Working For
Dexter + Chaney
-
Companies Worked For
Dexter + Chaney
0
Neutral
User Tools
Friends
mdexter hasn't added any friends yet.
Latest Visitors
Posts I've Made
In Topic: 5.5 & Vista
09 February 2007 - 09:49 PM
We have at least two customers who report to us that they are successfully using our PRO-IV 5.5-based app with Vista. We are not officially supporting it, since PRO-IV is not. However, it may be that it works ok with some tweaking. Mark Dexter
In Topic: SQL Dynamic Query (Multiple tables)
09 February 2007 - 09:42 PM
The problem might be that PRO-IV adds some text to SQL when it is used in the selection portion of a cycle (e.g., in the cycle entry logic). For example, in SQL Server, if the target table is in LOOKUP mode, PRO-IV will add the query hint "WITH(NOLOCK)" to the SQL. This is fine as long as the SQL command only contains one table. However, if you are trying to do a sort or select on more than one table, PRO-IV doesn't add this query hint correctly, so the SQL command fails.
The only work-around that we have found is to build a view in SQL and then build a PRO-IV file spec that references the view. A typical case is where you need to sort or select on columns not in the primary table. So if you build a view that joins the tables you need, create a PRO-IV file spec that references this view, and then use that as the primary file in the cycle, it works. It's more work, but in some cases it can make a dramatic improvement in performance (e.g., in a case where otherwise you are doing a lot of read / DSEL).
A simple fix for PRO-IV would be to have an option in logic to tell PRO-IV not to alter the SQL at all (e.g., something like SQL DYNAMIC RAW). This would seem to be fairly easy for them to implement and would help us out a lot. With that, you could create any type of select statement you wanted, with as many joins as needed. The developer would just need so ensure that the column list returned by the SELECT statement matched the PRO-IV file spec for the primary file in the cycle, which is easy to do.
Hope this helps. Mark Dexter
The only work-around that we have found is to build a view in SQL and then build a PRO-IV file spec that references the view. A typical case is where you need to sort or select on columns not in the primary table. So if you build a view that joins the tables you need, create a PRO-IV file spec that references this view, and then use that as the primary file in the cycle, it works. It's more work, but in some cases it can make a dramatic improvement in performance (e.g., in a case where otherwise you are doing a lot of read / DSEL).
A simple fix for PRO-IV would be to have an option in logic to tell PRO-IV not to alter the SQL at all (e.g., something like SQL DYNAMIC RAW). This would seem to be fairly easy for them to implement and would help us out a lot. With that, you could create any type of select statement you wanted, with as many joins as needed. The developer would just need so ensure that the column list returned by the SELECT statement matched the PRO-IV file spec for the primary file in the cycle, which is easy to do.
Hope this helps. Mark Dexter
In Topic: How are you doing on Version 6
02 January 2007 - 10:29 PM
We have been trying to create a test environment with our application and V6 for several months. At the present time, we are waiting for a menu bug to be fixed before we can do any real testing. (Without this fix, the menus were exceedingly slow to the point that we cannot run our app at all.)
Before that, we were unable to gen our app in V6.
We are hoping to be able to start testing in a few weeks, but it depends on PRO-IV.
Mark
Before that, we were unable to gen our app in V6.
We are hoping to be able to start testing in a few weeks, but it depends on PRO-IV.
Mark
In Topic: Barcode Printing
03 October 2006 - 10:34 PM
Okidata dot matrix printers!!?? Do they still make these? Takes me back to the good old days. Televideo terminals, making your own 25-pin cables (I've still got the crimping tool around here somewhere ...). I feel old all of a sudden. Mark Dexter
In Topic: Oracle Identity Columns
20 July 2006 - 09:59 PM
Multi-user problem in SQL Server:
A note of caution about my previous post using the SQL Server Identity column and an INSERT trigger. We tested this in a multi-user simulation and it appears to only allow one user at a time to INSERT rows. We aren't sure whether this is because of the trigger or just a consequence of using the IDENTITY column in SQL Server. So we are going to use a different method (the NEWID() function in SQL Server). Mark Dexter
A note of caution about my previous post using the SQL Server Identity column and an INSERT trigger. We tested this in a multi-user simulation and it appears to only allow one user at a time to INSERT rows. We aren't sure whether this is because of the trigger or just a consequence of using the IDENTITY column in SQL Server. So we are going to use a different method (the NEWID() function in SQL Server). Mark Dexter
- ProIV Resource Centre
- → Viewing Profile: Posts: mdexter
- Privacy Policy