Hello,
I am having 2 inputs that enable 2 different outputs and i would like to be 100% sure that just one of each output will go on at a time. There is this little chance that both of these inputs will go on at the same time and i would like to avoid stalling the programm or having an error that stops at an AND block. I tried to use pulse relays and the program seems to work this way. I believe this is due to the flag i am using that stalls the input for some milliseconds so just the 1st output goes on first. But is it right though or there is a better way to do that? 

For simplicity reasons i am using 2 cursors that reset the pulse relays in the attached test file.

In my programm when an output of each goes on, after some time will reset the pulse relay automatically but i already have that implemented.

The following suggestion would be enough to avoid outputs being on at the same time.

Keep in mind that the 'output' state of the Q and M blocks are one scan delayed from the 'input' state. This can cause 'timing glitches' if you are not careful. The scanning of the code is according to the block numbers as far as I can gather.

qwazee

Keep in mind that the 'output' state of the Q and M blocks are one scan delayed from the 'input' state. 

To demonstrate the point in the quoted text - simulate the design shown below.

Your circuit in its current format does what you want from it. Below is a test circuit where I1 is connected to both branches. Counters were added to show that only one output is triggered.

By the way.. have you tested your design concept on actual hardware yet? I am curious to know if you are happy with the concept of how you planned the use of the cursors.

Thank you very much for your immediate response! 
Actually i did not test it on the hardware yet but i will do it soon, when i complete some parts that are missing.
I do not know if you are talking about these two counters that i added in the test sample but i mentioned above that these counters are added for simplicity of the test in order to avoid big blocks with much information. In my real program, i do not use counters to reset these particular inputs. It is an automated phase where i trigger the output and after sometime something else is triggered and then i reset the input reason. I believe i could have used an on delay block for my test and this would be better but i wanted something fast.

Yesterday i asked about counters in another post but i got the answer i wanted so these 2 counters on this post are different, its in another project.

Anyways, thanks again for your time and help. I will test it on the logo! and will come back to you if i cannot solve any issues.

I am aware of the flags and outputs recursions but i cannot do that. When i trigger the output it will be with a 20sec off delay. I trigger the pulse relay so i will be aware if this function is enabled and so i can stop the other one that may come from another input. When i am done with all i want, i reset the relays ! 

It would be easier if i had just to trigger on the output but unfortunately i cannot do that.

Anyway, i believe i have the solution to my problem. Thanks again for your time and help! 

Sorry ixo65, that is actually a problem. It is possible to get both outputs on. To demonstrate I have connected I1 to both branches. This simulates the occurrence where both I1 and I2 comes on in the exact same scan.

The reason is as I mentioned earlier: the 'out' state of the output / memory bit is one scan late.

Keeping the format; the following is better.

But this relies on the sequence of how the blocks are scanned / processed. 

My preference design would as seen in the image. The one output is given priority but still locks out the possibility of switching on both outputs.