summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorLamont Granquist <lamont@scriptkiddie.org>2014-10-13 13:59:10 -0700
committerLamont Granquist <lamont@scriptkiddie.org>2014-10-13 13:59:10 -0700
commitbc11758c3c5477ef000794410286ab5a39d624d9 (patch)
tree6b555a142d64f1d5783741a04781f99700bb8034
parent9479cd28a8ea5b073c7dd01d352e398b3bb497f5 (diff)
downloadffi-yajl-bc11758c3c5477ef000794410286ab5a39d624d9.tar.gz
line wrap on commentlcg/better-ext-vs-ffi-logic
-rw-r--r--lib/ffi_yajl.rb13
1 files changed, 8 insertions, 5 deletions
diff --git a/lib/ffi_yajl.rb b/lib/ffi_yajl.rb
index a3d1322..7f39189 100644
--- a/lib/ffi_yajl.rb
+++ b/lib/ffi_yajl.rb
@@ -2,12 +2,15 @@
#
# Precedence:
#
-# - The FORCE_FFI_YAJL env var takes precedence over everything else, the user theoretically knows best
+# - The FORCE_FFI_YAJL env var takes precedence over everything else, the user
+# theoretically knows best
# - Java always gets ffi because jruby only supports ffi
-# - There is a conflict between loading libyajl 1.x and 2.x in the same VM process (on a fundamental basis,
-# simply guru medidate about how the c-symbols work if you load both libs). For some reason the ffi
-# interface seems to work fine sometimes (i'm not sure how) so we fall back to that -- this is much more
-# likely to be converted into a raise than to have the warn dropped, so don't bother asking for that.
+# - There is a conflict between loading libyajl 1.x and 2.x in the same VM
+# process (on a fundamental basis, simply guru medidate about how the
+# c-symbols work if you load both libs). For some reason the ffi interface
+# seems to work fine sometimes (i'm not sure how) so we fall back to that--
+# this is much more likely to be converted into a raise than to have the warn
+# dropped, so don't bother asking for that.
# - Then we try the c-ext and rescue into ffi that fails
#
if ENV['FORCE_FFI_YAJL'] == "ext"