back to quick_and_dirty, which is still practically important since the scheduler does not jump over subproofs;
authorwenzelm
Sat Nov 13 12:32:21 2010 +0100 (2010-11-13 ago)
changeset 405218896bd93488e
parent 40520 77a7b0a7d4b1
child 40522 37b79d789d91
back to quick_and_dirty, which is still practically important since the scheduler does not jump over subproofs;
src/Pure/System/isabelle_process.ML
     1.1 --- a/src/Pure/System/isabelle_process.ML	Sat Nov 13 11:41:02 2010 +0100
     1.2 +++ b/src/Pure/System/isabelle_process.ML	Sat Nov 13 12:32:21 2010 +0100
     1.3 @@ -168,7 +168,7 @@
     1.4      val _ = OS.Process.sleep (seconds 0.5);  (*yield to raw ML toplevel*)
     1.5      val _ = Output.raw_stdout Symbol.STX;
     1.6  
     1.7 -    val _ = quick_and_dirty := false;
     1.8 +    val _ = quick_and_dirty := true;
     1.9      val _ = Goal.parallel_proofs := 0;
    1.10      val _ = Context.set_thread_data NONE;
    1.11      val _ = Unsynchronized.change print_mode