e5a054a02e7a80af675a99ba146c27465837f1d8
markd
  Thu Dec 21 10:13:23 2023 -0800
more work on running tests in parallel

diff --git src/hg/checkTableCoords/tests/makefile src/hg/checkTableCoords/tests/makefile
index b721fdb..a45c419 100644
--- src/hg/checkTableCoords/tests/makefile
+++ src/hg/checkTableCoords/tests/makefile
@@ -27,22 +27,22 @@
 	for TEST in ${BEDS} ; do \
 	  ${hgLoadBed} test $${TEST}Good input/$${TEST}Good.bed -verbose=0 ; \
 	  ${hgLoadBed} -noStrict test $${TEST}Bad  input/$${TEST}Bad.bed -verbose=0 ; \
 	done
 	for TEST in ${PSLS} ; do \
 	 ${hgLoadPsl} -noHistory test -table=$${TEST}Good input/$${TEST}Good.psl -verbose=0 ; \
 	 ${hgLoadPsl} -noHistory test -table=$${TEST}Bad  input/$${TEST}Bad.psl -verbose=0 ; \
 	done
 
 mkdirs:
 	@mkdir -p output
 
 clean:
 	rm -fr output
 	rm -f *.tab
-	${hgsql} test -e 'drop table chromInfo;' > /dev/null 2>&1; true
+	${hgsql} test -e 'drop table if exists chromInfo;'
 
 
 # angie I removed -verbose=2 from the Bad test because it is unstable.
 # if somebody adds some extra statements in libraries and utilities with verbose,
 # then the output changes and is not stable and leads to false failures.  
 # You could use a different flag though perhaps instead of -verbose=2?