library-of-basic-processes-(lbp) is really good,

https://support.industry.siemens.com/cs/de/en/view/109749508

I am trying to use with my project, but why panels use tags use inout with DB, while wincc pro uses FB instance?

In my experience, if  HMI tag use inout, there is a risk the write tag will fail, if tag hmi wrote during the  execution of FB, writing tag will not succeed. Or siemens HMI uses some technical to write tag before program execution?

Any help or advice is appreciated! I have been in trouble in this for long time.

Stay safe!

update:

some info in 

Why is data of the HMI system or the web server sometimes overwritten in the S7-1500?

 https://support.industry.siemens.com/cs/de/en/view/109478253

• If there is an exchange between non-optimized and optimized data, the following problem might occur when using InOut parameters:
  
  Example
  A DB tag with standard access is parameterized at the InOut parameter of a block with optimized block access.
  In this case the data is not transferred as a pointer but as a copy. The copy of the operand is written to the operand again when the block ends, even if the copy has not been changed in the block. In this way the write access instances of the HMI system during runtime of the block are lost.
  
  Note
  Depending on the transfer area and the data type of the parameter, STEP 7 (TIA Portal) transfers the data internally either as pointer or copy. Detailed information on this topic is available in the programming guidelines in Entry ID90885040

Remedy
For the system properties described above we recommend the following remedies:

• In parallel processes you should avoid write access to the same tag at the same time.

• Create a global interface data block with two separate data structures (possibly using PLC data types), for HMI and PLC access, for example.
• In one data structure you make write access only from Process 1, HMI, for example.
• In the other data structure you make write access only from Process 2, the PLC, for example.
• You can generally make read access from both processes to both data structures.
• In this way you prevent parallel processes from mutually overwriting the tags.

• When accessing blocks, avoid cross-overs between optimized and non-optimized areas. With S7-1200/S7-1500, newly created blocks have the attribute "Optimized block access". S7-300/S7-400 blocks do not have optimized block access. If you use these blocks for the S7-1500, we recommend that you set the block access to "optimized". You can change the block access attribute in the block Properties in the "Attributes" tab.

This may help, for s7-1500 it is really new, but I do not know if this works for S7-1200 or not....

Good luck!