-
Notifications
You must be signed in to change notification settings - Fork 7
/
Copy pathChangeLog
44 lines (35 loc) · 2.03 KB
/
ChangeLog
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
Change log for DWSOLVER.
--------------------------------------------------------------------------------
- Version 1.2
- Released October 22, 2010
- Contact Joseph.L.Rios@nasa.gov
--------------------------------------------------------------------------------
Found and fixed a memory allocation error in the subproblem thread. Error
was only visible when there were more rows in the master problem than there
were variables in a given subproblem. None of the prior examples exhibited this
behavior.
Added command line option to enforce integral constraints in the subproblems.
This is important if experimenting with the integerization heuristic (-i flag).
Added some comments throughout.
--------------------------------------------------------------------------------
- Version 1.1
- Released October 15, 2010
- Contact Joseph.L.Rios@nasa.gov
--------------------------------------------------------------------------------
Fixed distribution package which was missing a header file. Added automatic
detection of OSX build so that the --enable-named-semaphores flag is now not
necessary when building on Mac (it will be set automatically).
--------------------------------------------------------------------------------
- Version 1.0
- Released September 8, 2010
- Contact Joseph.L.Rios@nasa.gov
--------------------------------------------------------------------------------
Initial release. Dantzig-Wolfe Solver (DWSOLVER) is a general, parallel
implementation of the Dantzig-Wolfe Decomposition algorithm built upon the
GNU Linear Programming Kit (GLPK). This is a command-line only tool. All
input and output is file-based. The user must supply a decomposed linear
program as input. For details, execute "dwsolver --help". For examples, see
the "examples" directory. Current major limitations include the necessity of
bounded subproblems, necessity of every master variable appearing in exactly
one subproblem. For a large number of problems, these are not major issues.
For other problems, there are some workarounds as described in the examples.