Specifying the number of repetitions for each condition within and across trials

I am building an adapted version of a primed Go/NoGo with feedback. The experiment requires two types of trials. Trial 1 includes a square in the middle of the screen, the square is followed by a triangle. The position of the triangle has two conditions - positioned in the middle or positioned on the right. The participants are instructed to respond differently depending on the position of the triangle - press ‘right’ if the triangle is in the middle, do not press anything if the triangle is on the right. After each response participants receive feedback (“Correct”, “Incorrect”). Trial 2 includes a circle in the middle of the screen. The circle is followed by a triangle with two conditions - positioned in the middle and positioned on the left. Participants are instructed to press ‘left’ following the presentation of a triangle at all times if the preceding shape was a circle. Again feedback is provided following each response. I hope this makes sense!
I have managed to get my experiment working fine up to this stage using the builder and coder views. The conditions and feedback are displayed and saved correctly.
The structure of the experiment is as follows - loop: trial1(squares) + feedback, loop: trial2(circles) + feedback. I have also put a loop around the two trials to indicate that this will be the first block. Block 2 will have an identical setting but different rules for responding.
Question 1
The next step is for me to vary the amount of time each condition is presented within a trial. So in trial1 I would like to have the triangle appear in the middle 75% of the time and appear on the right 25% of the time. This will need to be similar in trial2. I cannot find any guidance on how to do this.
Question 2
I need to make sure that half of the trials within the whole block are trial1(squares) and half are trial2(circles). However, I need to have them presented in a random order. Not sure how to make this happen either.
Below: the structure of the experiment in builder view and the conditions file


I have managed to get the issue sorted.
Question 1:
I added two additional copies of condition one. Therefore for the square there were now 4 conditions, 3 of them the same indicating that the triangle should be in the middle and one condition indicating that the triangle should be on the right. I did the same for the ‘circle’ stimulus. Therefore there was a split of 75% and 25% per position and only two conditions in the end.
Question 2
I understand that randomisation works best within a single loop but this was not easy to achieve with the shapes component as you cannot simply reference the conditions to the file.
I have created images of the shapes and referenced them in the conditions file. Therefore the shape of the first stimuli can also be driven from the conditions file rather than from the two different types of trials. There is now only one loop that drives the whole block. All is working very well!

Below is the builder view and how it looks now and the conditions file: