Tue, 26 Feb 2013 13:05:48 +0100signal work_available should be sufficient to initiate daisy-chained workers, and lead to separate broadcast work_finished eventually -- NB: broadcasting all worker threads tends to burn parallel CPU cycles;
wenzelm [Tue, 26 Feb 2013 13:05:48 +0100] rev 52418
signal work_available should be sufficient to initiate daisy-chained workers, and lead to separate broadcast work_finished eventually -- NB: broadcasting all worker threads tends to burn parallel CPU cycles;

Tue, 26 Feb 2013 12:50:52 +0100less eventful shutdown: merely wait for scheduler to terminate;
wenzelm [Tue, 26 Feb 2013 12:50:52 +0100] rev 52417
less eventful shutdown: merely wait for scheduler to terminate;

Tue, 26 Feb 2013 12:46:47 +0100tuned messages;
wenzelm [Tue, 26 Feb 2013 12:46:47 +0100] rev 52416
tuned messages;

Tue, 26 Feb 2013 11:57:19 +0100tuned;
wenzelm [Tue, 26 Feb 2013 11:57:19 +0100] rev 52415
tuned;

Mon, 25 Feb 2013 20:55:48 +0100merged;
wenzelm [Mon, 25 Feb 2013 20:55:48 +0100] rev 52414
merged;

Mon, 25 Feb 2013 17:47:32 +0100more explicit Goal.shutdown_futures;
wenzelm [Mon, 25 Feb 2013 17:47:32 +0100] rev 52413
more explicit Goal.shutdown_futures;

Mon, 25 Feb 2013 13:31:02 +0100reconsider 'export_code' as "thy_decl" command due to its global side-effect on the file-system;
wenzelm [Mon, 25 Feb 2013 13:31:02 +0100] rev 52412
reconsider 'export_code' as "thy_decl" command due to its global side-effect on the file-system;

Mon, 25 Feb 2013 13:29:19 +0100discontinued pointless command category "thy_schematic_goal" -- this is checked dynamically;
wenzelm [Mon, 25 Feb 2013 13:29:19 +0100] rev 52411
discontinued pointless command category "thy_schematic_goal" -- this is checked dynamically;

Mon, 25 Feb 2013 12:52:27 +0100clarified Toplevel.element_result: scheduling policies happen here;
wenzelm [Mon, 25 Feb 2013 12:52:27 +0100] rev 52410
clarified Toplevel.element_result: scheduling policies happen here;
tuned;

Mon, 25 Feb 2013 12:17:50 +0100prefer stateless 'ML_val' for tests;
wenzelm [Mon, 25 Feb 2013 12:17:50 +0100] rev 52409
prefer stateless 'ML_val' for tests;