Ren'Py Persistent Object Size Limit Configuration for Low-Traffic Workloads

Answer the question first

Ren'Py does not expose a configuration flag or built‑in API that caps the size of the persistent object, nor does it automatically prune old entries or compress the pickle file. The engine serializes persistent to persistent.py using Python’s pickle module, and that file is written uncompressed by default.

What can be done?

• Manual pruning – Add logic to clear or restructure data when it exceeds a threshold.
• Playthrough counter – Track how many times the player has started a new playthrough and trigger cleanup when the counter passes a limit.
• Custom serialization hook – Override the default persistent.save() routine (or edit the block in persistent.py) to compress the pickle payload with zlib, gzip, or a similar library.

Practical steps for this case

1. Track playthroughs

   init python:
       if not persistent.playthrough_count:
           persistent.playthrough_count = 0
   
   label start:
       $ persistent.playthrough_count += 1
       # ... game logic ...
   

2. Prune old data

   init python:
       def prune_persistent(max_entries=100):
           # Example: keep only the most recent 100 keys
           keys = list(persistent.__dict__.keys())
           if len(keys) > max_entries:
               for k in keys[:-max_entries]:
                   del persistent[k]
   

   $ prune_persistent()
3. Compress the pickle output

   init python:
       import zlib, pickle
       def persistent_save():
           # Serialize normally
           data = pickle.dumps(persistent.__dict__)
           # Compress
           compressed = zlib.compress(data)
           # Write to file
           with open(os.path.join(persistent_dir, "persistent.py"), "wb") as f:
               f.write(compressed)
   
   # Hook into Ren'Py’s save routine
   init python:
       persistent.save = persistent_save
   

4. Verify
   • Run the game, inspect persistent.py size before and after adding the compression hook.
   • Check that the game loads correctly after a save/restore cycle.
   • Measure load times on the target low‑traffic device to ensure compression overhead is acceptable.

How to confirm the approach works

• Open the save directory and list the file size of persistent.py.
• Use a small Python script to decompress the file and unpickle it, verifying that all expected keys are present.
• Run the game with the persistent_save hook enabled and record the time taken to load a save.

Missing diagnostic detail

Which Ren'Py version are you using? Some newer releases expose persistent.save() as a public method, while older ones require editing persistent.py directly. Knowing the exact version will help tailor the override correctly.