You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
Catalog import merges cross-referenced objects and overwrites their coordinates (Collinder 33/34 point at IC 1848) #667
Symptom: push-to on Collinder 33 sends the user to IC 1848's position. A user centered on "Cr 33" saw the Status screen read 2h51m / 60.36, while Cr 33's J2000 position is 02h59m18s / +60°24'.
Cause (release branch, 3775008): astro_data/collinder.txt has the correct position for each entry, but Cr 32, 33 and 34 all list IC1848 in the NGC/Other column. load_collinder() in catalog_imports/specialized_loaders.py passes that as an aka name, and NewCatalogObject.insert() (catalog_import_utils.py) finds the existing IC 1848 object through find_object_id() and reuses it, coordinates included, adding only the catalog_objects row. So in pifinder_objects.db, Col 32, Col 33 and Col 34 all resolve to one object (9688) at RA 42.8250 / Dec 60.4083 (02h51m18s / +60.41).
Scope: comparing all 471 Col entries in the DB against collinder.txt, 7 are more than 10' from their source position:
Entry
Cross-ref
Offset
Col 314
NGC 6222
121'
Col 34
IC 1848
71'
Col 33
IC 1848
59'
Col 208
IC 2488
43'
Col 401
(none)
40' (a different cause, worth a look)
Col 253
NGC 4230
12'
Col 191
IC 2391
10'
Possible fix direction: when an entry's aka matches an existing object whose position differs by more than a tolerance, insert it as its own object (or keep the source coordinates) instead of merging. Other loaders use the same aka-merge path, so other catalogs may have the same problem.
Fixed by the database rebuild in #697 (4d6e6d6). Col 33 is now its own sky object (with Lyn 91) at 02h59m18s +60°24′, and Col 34 is its own (with Lyn 92). Col 32 stays with IC 1848, as Collinder gives it. The evidence is in #697's "Related issues" section.
Activity
brickbots commentedon Nov 19, 2026
Fixed by the database rebuild in #697 (4d6e6d6). Col 33 is now its own sky object (with Lyn 91) at 02h59m18s +60°24′, and Col 34 is its own (with Lyn 92). Col 32 stays with IC 1848, as Collinder gives it. The evidence is in #697's "Related issues" section.