Interesting "feature" in TIA Portal V19 Update 2 and V18 Update 4.

SCL program is not executed correctly after compiling in TIA Portal

jacek d

We are getting closer and closer to IT-like workflows. You update your tools and your code doesn't work anymore. 

You are right there! Shall we start build up "old" stock to deliver a "long term stability" service?

Did I understand correctly? --
Customers should create a support request to get the patch file and instructions? In other words, the patch is not available for public download.

qwazee

Did I understand correctly? --

Customers should create a support request to get the patch file and instructions? In other words, the patch is not available for public download.

Indeed. I was looking for the patch to add the link to the first post, then I found out that it can be only obtained directly from the Support.

It is what it is. What becomes obvious is that there's a growing need for code base automated testing.

When I was involved in maintaining our own OEM PLC and associated language compiler, we did just that.

We had a series of code files that used all code statements in all possible permutations that we could think of, which tested the compiler. To test the resultant runtime code, we had a series of test routines that counted from 1 to 10 by using all functions and control statements in ingenious ways. For example GOTO's jumping all over the place, trig functions returning expected known values, map statements, loops etc. If any functions / statements failed the count would fail.

Anytime we made any changes to the compiler or runtime sequencer, we ran the tests. This is known as regression testing. Despite this, we did have a few obscure bugs get through undetected, as with the greatest will in the world, you cannot think of every conceivable scenario.

So, I do have a degree of sympathy for the developers. What MUST happen though is HONESTY. You have to tell the world that there is a bug, what versions it affects, what possible issues it might cause (the failure mode), any workarounds, and finally when it is likely be fixed. They seem to have done that adequately in this case.

It is certainly better than the port 502 issue with the CP343 / CP443 firmware bug a few years back. They broke the firmware so badly, that the port could not even be opened if the port number was 502. Now that IS unacceptable as the points directly to inadequate QA and validation testing. No excuses for that one. They also never even mentioned it in public. I had the issue, and I only found out about this cause from an internal contact in Siemens through this forum. Even the SR team knew nothing about the issue. Trying to hide issues doesn't help anyone.

Smiffy.