one image size cap for every division, 200 GB

The published figure and the figure a submission is rejected on disagreed. config.py advertised 100 GB to everyone while a per-division dict beside it read 10 GB for small, and the organizer worker rejects on the per-division number. A small-division team could follow the published guidance, build a 30 GB image, and be rejected at scoring time. Roughly half the invited seats are small division.

One number now: 200 GB for every division, stated as the compressed registry size. The per-division dict is deleted rather than corrected, because nothing in this Space ever read it -- it was a second copy that could only drift.

The organizer-side cap has to move to 200 for every division in the same window, or the mismatch simply changes direction.

149 tests OK (register 50, submit 66, val-closed 33). The Test guide now renders 200 GB.

config.py

@@ -108,16 +108,17 @@ MAX_TEST_SUBMISSIONS_PER_TRACK = 3
 # count_prior_submissions() in the organizers' test_phase_worker.py.
 TEST_SLOT_STATUSES = ("PENDING", "FINISHED", "FAILED")
 
-# Per-division image size ceiling, enforced here at registration and again by the
-# worker from the ECR manifest before any pull. Keep in step with
-# DEFAULT_SIZE_CAPS_GB in test_phase_worker.py and participant/validate_image.sh.
-TEST_IMAGE_SIZE_CAPS_GB = {"small": 10, "large": 200}
-
-# The size we publish to participants: one number for both divisions, stated as
-# the compressed size in the registry (the number `docker push` reports), not the
-# uncompressed size on disk. Over it, a team asks for an exception rather than
-# being blocked by the form.
-TEST_IMAGE_ADVERTISED_GB = 100
+# The image size ceiling, one number for every division, stated as the compressed
+# size in the registry (the number `docker push` reports), not the uncompressed
+# size on disk. Over it, a team asks for an exception rather than being blocked
+# by the form. Nothing in this Space enforces it; the organizer worker rejects
+# from the ECR manifest before any pull.
+#
+# It used to be two numbers. A per-division dict said 10 GB for small while this
+# published 100 GB to everyone, and the dict was never actually read here, so a
+# small-division team could follow the published figure and still be rejected at
+# scoring time. One number now, with no per-division split to drift back into.
+TEST_IMAGE_ADVERTISED_GB = 200
 
 
 # A registry reference we are willing to pull from. Anchored on purpose: the

tests/test_submit_test.py

@@ -157,7 +157,9 @@ class ValidateOnlyTest(unittest.TestCase):
 
         The organizers publish ONE number for both divisions and the form does
         not refuse an over-size image, so the message must state the advertised
-        size rather than TEST_IMAGE_SIZE_CAPS_GB[division]. It must also say
+        size rather than a per-division one. That per-division dict is gone as
+        of 2026-08-08: it read 10 GB for small while this published 100 to
+        everyone, and nothing here ever read it. It must also say
         WHICH quantity that number is: the participant side measures the
         uncompressed image on disk and the organizer side reads the compressed
         size from the registry, and the two differ by enough to mislead.