Module: ppl/ppl Branch: master Commit: 0fd5be627f295dd9745620188a894774c1ea95da URL: http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commit;h=0fd5be627f295... Author: Roberto Bagnara <bagnara@cs.unipr.it> Date: Wed Apr 1 10:40:10 2009 +0200 New items for PPL 0.11 added. Other items scheduled for PPL 0.11. --- TODO | 38 +++++++++++++++++++++++++------------- 1 files changed, 25 insertions(+), 13 deletions(-) diff --git a/TODO b/TODO index 7095ba7..d35ffb1 100644 --- a/TODO +++ b/TODO @@ -5,17 +5,21 @@ Enhancements for PPL 0.10.1 or later versions - Fix all the FIXME(0.10.1). -Enhancements for PPL 0.11 or later versions -=========================================== +Enhancements for PPL 0.11 +========================= +- Add the domain of circular linear progressions. +- Finish the domain of intervals. +- Supplement the "preemption" methods of the library allowing for + deterministic mechanisms (e.g., decrementing a counter and throwing + when it reaches zero). This will probably require changing the + current interface based on `abandon_expensive_computations'. - Get rid of ugly assign functions defined in anonymous namespace of - MIP_Problem.cc replacing it with a more general purpose approach. + MIP_Problem.cc replacing it with a more general-purpose approach. - When the PPL has been configured with CPPFLAGS="-DPPL_ARM_CAN_CONTROL_FPU=1" make sure the library initialization procedure checks that the FPU can indeed be controlled, failing if that is not the case. -- Consider switching to MLGMPIDL for the OCaml interface - (see https://bugzilla.redhat.com/show_bug.cgi?id=491712). - Reconsider the datatype for the units parameter in the Watchdog constructors: should we prefer an unsigned? - Rename set_rational_sqrt_precision_parameter to @@ -23,6 +27,22 @@ Enhancements for PPL 0.11 or later versions - Intervals are best instantiated with checked numbers with particular policies: review all the interfaced boxes, augment the testsuite, and update the documentation. +- DO_TEST_F should not cause the output "failed tests" to appear, + but "tests unexpectedly succeeded". +- Reduce the number of FIXMEs to no more than 20. +- Switch to the latest versions of Autoconf and Automake (in beta test + at the time of writing): they provide features we have been looking + for for years. +- Use the OCaml-Autoconf macros: https://forge.ocamlcore.org/frs/?group_id=69 + (see also + http://rwmj.wordpress.com/2009/03/31/using-autoconf-for-ocaml-projects/). + + +Enhancements for PPL 0.11 or later versions +=========================================== + +- Consider switching to MLGMPIDL for the OCaml interface + (see https://bugzilla.redhat.com/show_bug.cgi?id=491712). - Improve the OCaml interface by supporting dynamic loading. - Enhance the support for systems not providing support for the IEEE inexact flag. @@ -33,10 +53,6 @@ Enhancements for PPL 0.11 or later versions - Make the library thread-safe. - Find a way to let the GMP and checked-int versions of the PPL coexist so as to allow using them in the same application. -- Supplement the "preemption" methods of the library allowing for - deterministic mechanisms (e.g., decrementing a counter and throwing - when it reaches zero). This will probably require changing the - current interface based on `abandon_expensive_computations'. - Provide methods computing approximations of the integer convex hull of semantic domains for the case where we are only interested in integral solutions (as is the case, e.g., in many program analyses). Then @@ -76,8 +92,6 @@ Enhancements for PPL 0.11 or later versions or we are otherwise 100% sure that the changed argument and the argument that is read afterward are not aliases (in the latter case an assert(&to != &z) shall be added. -- DO_TEST_F should not cause the output "failed tests" to appear, - but "tests unexpectedly succeeded". - Improve the handling of the --enable-instantiations configure option so that "closure by subdomain" is automatically implemented (e.g., if Pointset_Powerset(X) is specified, X should be automatically @@ -89,8 +103,6 @@ Enhancements for PPL 0.11 or later versions - Consider whether or not our interface for method time_elapse_assign() is the one needed by users: in particular, see if it should take as an argument a generator system, instead of an abstract element. -- Finish the domain of intervals. -- Reduce the number of FIXMEs to no more than 20. - Provide a generic implementation for the `widening with tokens'. - In the OK() methods, instead of having all those #ifndef NDEBUG, it is probably worthwhile to use a suitable defined `barf' stream