always compress eddn data, comment rationale

plugins/eddn.py

@@ -219,7 +219,16 @@ class EDDN:
             ('message', msg['message']),
         ])
 
-        encoded, compressed = http.gzip(json.dumps(to_send))
+        # About the smallest request is going to be (newlines added for brevity):
+        # {"$schemaRef":"https://eddn.edcd.io/schemas/shipyard/2","header":{"softwareName":"E:D Market
+        # Connector Windows","softwareVersion":"5.3.0-beta4extra","uploaderID":"abcdefghijklm"},"messa
+        # ge":{"systemName":"delphi","stationName":"The Oracle","marketID":128782803,"timestamp":"xxxx
+        # -xx-xxTxx:xx:xxZ","ships":[]}}
+        #
+        # Which comes to about around 307 bytes. Lets play it safe and make our minimum 0 bytes.
+        # Which compresses everything
+
+        encoded, compressed = http.gzip(json.dumps(to_send, separators=(',', ':')), max_size=0)
         headers: None | dict[str, str] = None
         if compressed:
             headers = {'Content-Encoding': 'gzip'}

util/http.py

@@ -10,6 +10,9 @@ def gzip(data: str | bytes, max_size: int = 512, encoding='utf-8') -> tuple[byte
     """
     Compress the given data if the max size is greater than specified.
 
+    The default was chosen somewhat arbitrarily, see eddn.py for some more careful
+    work towards keeping the data almost always compressed
+
     :param data: The data to compress
     :param max_size: The max size of data, in bytes, defaults to 512
     :param encoding: The encoding to use if data is a str, defaults to 'utf-8'