Update comment obsoleted by Thomas's latest fixes.

tglsfdc · tglsfdc · commit 6358e654ca3e · 2000-03-15T23:42:14.000Z
diff --git a/src/backend/utils/adt/ruleutils.c b/src/backend/utils/adt/ruleutils.c
@@ -3,7 +3,7 @@
  *			  out of its tuple
  *
  * IDENTIFICATION
- *	  $Header: /cvsroot/pgsql/src/backend/utils/adt/ruleutils.c,v 1.45 2000/03/14 23:06:37 thomas Exp $
+ *	  $Header: /cvsroot/pgsql/src/backend/utils/adt/ruleutils.c,v 1.46 2000/03/15 23:42:14 tgl Exp $
  *
  *	  This software is copyrighted by Jan Wieck - Hamburg.
  *
@@ -998,18 +998,9 @@ get_select_query_def(Query *query, deparse_context *context)
 								 inherit_marker(rte));
 				/*
 				 * NOTE: SQL92 says you can't write column aliases unless
-				 * you write a table alias --- but the table alias could
-				 * be spelled the same as the table's real name.  This
-				 * logic is therefore all wet: it should go something like
-				 * IF we-need-to-dump-column-aliases OR relname != refname
-				 * THEN print refname;
-				 * IF we-need-to-dump-column-aliases
-				 * THEN print column alias list.
-				 * But currently we can't tell whether we need to dump
-				 * column aliases or not... without that, this clearly
-				 * backwards logic seems the best short-term approach.
-				 * Since we don't really support SQL joins yet, dropping
-				 * the list of column aliases doesn't hurt anything...
+				 * you write a table alias --- so, if there's an alias list,
+				 * make sure we emit a table alias even if it's the same as
+				 * the table's real name.
 				 */
 				if ((rte->ref != NULL)
 					&& ((strcmp(rte->relname, rte->ref->relname) != 0)
