-
Notifications
You must be signed in to change notification settings - Fork 0
/
bugs.txt
67 lines (40 loc) · 2.19 KB
/
bugs.txt
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
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
20090607 20:16
I've implemented some bug fixes that have accumulated over the past few months. Most of the bugs were identified by the Steven Boada or me. Jihoon,
a user, recently pointed out a bug in the calculation of particle mass, but that fix is on going.
The list of bugs is extensive, but here goes:
Added a copy tag for dealing with copied galaxies.
Added epsabs and epsrel parameters to the header file for integration tolerances.
Removed threading.
Fixed a bug in calling the dispersion routine and the dispersion grid.
Added a "load_gadget2_galaxy" routine for reading gadget snapshots, (best fix yet! :-P).
Fixed a bug in the mass assignment in the creation of gadget snapshots.
Fixed a bug in the Makefile.
REVERTED to the 85% estimate of the disk scale length using the fitting formula in MMW98. The new routine was underestimating the scale length.
The following files were editted:
Makefile
starscream_init.c
starscream_io.c
starscream_vel.c
Jay
-----
20081022 10:44
This update implements the proper iterative solution for finding the disk scale length instead of using the 15% approximation, [MMW,98].
This update also adjusts the bounds on the integrator to work with certain parameters.
Jay
-----
20081016 21:19
Steven Boada found a bug where the code tried to look beyond the edge of the dispersion grid and caused a segmentation fault.
Several other bugs have been noticed and fixed by the author.
The most notable change in the release is the removal of user-controlled random numbers. Starscream manages those on its own now.
Jay
-----
20080902 12:58
Dowon, a user, just identified a bug where the code compiles but crashes with a segmentation fault while running the test case.
This bug can be fixed by changing the size of the "green" and "potential" arrays in starscream_init.c to "double" from "double *" or by
downloading a new version of the code.
Jay
-----
20080826 16:09
Just identified a bug when trying to create disks for large m_d, j_d, and lambda where the GSL qags integration routine fails to converge.
This can be fixed by changing the tolerance in the calls to the QAGS routines from 0.0 to something like 1.0E-20 or downloading a new version of the code.
Jay