If so there is a design flaw on the display, it uses 5V to 3.3V buffers but the I2C pullups R2 and R3 are on the wrong side of the buffer to have any effect...
Add a 4K7 pullup to 5V on the Arduino SDA line and another on the SCK line.
Report back if this works as these displays also suffer from bad resets at power up. (Reset R+C time constant is to quick when the supply rail takes more than a few ms to establish)
Run an I2C scanner (Google it) to see if the Arduino can see the display on the I2C bus.
One of the displays shown in post #2 turned up in the post today so I have had a look at why it does not work.
The I2C address is 0x3C but the OLED board does not acknowledge the I2C transfers.
The problem is that the board has been incorrectly fitted with uni-directional buffers and so will never work correctly. This is a design/manufacturing error!
My solution is to take the buffers U3 and U4 off the board and add wire links to replace the buffers. There are pullups on the display board (R2 and R3) already so these now do their job and extra resistors are not needed.
This picture shows the hacked area of the board:
Not a job for the faint hearted or those without a fine tip on their soldering iron!
After the hack the display works fine and shows up on an I2C scanner.
If you cannot do this hack then the board is scrap. The only other work-arounds I could suggest is to hack the Wire library to ignore the absence of the acknowledge or arrange it so another I2C function does the Ack for the display.
So my advice is do not buy these particular 128x32 I2C displays (see post #2)!