Hi Claes,

Thanks for the code. We'll update the source.

A quick update on our progress on upgrading the latest gSoap version 2.8.36. Quite a few compile errors are due to type definition changes within the newer gSoap versions. For example, shown below are a few compiler errors in opc_utl.cpp where subroutine: bool opc_requestoptions_to_mask() and others complain about a wrong type, where a boolean is expected. Drilling deeper into gSoap, a comparison of old versus new reveals s0__RequestOptions definition and others are different from the existing ones. These changes ripple to various locations within Proview code. Upgrading gSoap requires wholesale changes to various subroutines to fix all compatible issues in newer gSoap versions.



Because of the involved effort to upgrade, we are faced with several options: 1) Spend the time to upgrade Proview software to the latest gSoap version with all modification required, 2) upgrade Proview to the latest compatible version of gSoap (we have not looked to see how far back to go) or 3) attempt to find the root cause of the ReturnValuesOnReply = "false" bug, because We don't know if the existing version of gSoap is actually the culprit or not. Upgrading gSoap seemed like a reasonable place to start.

For increased debugging insight, pwre.pl was modified to log more detailed information. Also this week, I have been introduced to an amazing software debug tool called ValGrind: valgrind.org/ This software is now loaded onto our test server and logging every memory transaction. It is quite revealing and apparently a de facto standard for C++ debugging. Highly recommended. You may be surprised at what is uncovered.

This effort is on-going and will continue to update progress. Your comments and suggestions are greatly appreciated.

Regards,
Ron

Hi Claes,

A few years back, you suggested a fix for a negative time offset error from the OPC Server. Back then, we fixed this error in v5.5 source code. Recently, we have been preparing to move to v6.1.1 and discovered the invalid time offset error remains. Again, we have fixed the code in v6.1.1. Respectively, we recommend this patch is included in future releases for those across the pond.

Many thanks!

/Ron

Hi Ron,

I'm sorry the fix wasn't inserted. I'll make sure it's inserted in the V6.1.3 release.

/Claes