forked from teuben/nemo
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathBUGLETS
94 lines (61 loc) · 3.08 KB
/
BUGLETS
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
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
* active issues are now moved to https://github.com/teuben/nemo/issues
* Some TODO's also here: https://github.com/teuben/nemo/projects/1
* this is the TODO list for the 4.2 release
- default configure gives yapp.ps, but should check if pgplot is natively there
- is rman partially broken on U18?
PolyglotMan v3.2 of $Date: 2003/07/26 19:00:48 $
- is there rman -> rst ?
- more solid examples of openmp
* Below here old stuff - some are still relevant
- gcc3: when using c++ NEMO's "string" will now collide with the
default C++ type 'string'....
-DSINGLEPREC does not work for a number of programs due to
bad prototyping of the "proc" functions. This will be fixed,
and is also needed in preparation for NEMO V5 when a C++ compiler
will be the default compiler.
- large objects (> 2GB or > 4GB), specifically snapshots, can be limited
in size of they use private calls to allocate() without making sure the
argument has been casted to (size_t)
- filestruct:
The following snippet of code stores only "1" for the value of pi,
3 times. ZENO actually does it correct.
put_set(ostr,"par");
pi = 1;
put_data(ostr, "pi", DoubleType, &pi, 0);
pi = 2;
put_data(ostr, "pi", DoubleType, &pi, 0);
pi = 3;
put_data(ostr, "pi", DoubleType, &pi, 0);
put_tes(ostr,"par");
ZENO also reads correctly, and if no more 'same name' items are present,
it will repeat the last one read. But the ZENO reader will serially
read (and remember?)
- filestruct history: long character strings are broken in silly places,
and don't port very well between tsf/rsf
- comment from Walter:
> I tried to make my own potential very much in the same way in which
> plummer.c works, except that I added a routine "potential()", beside
> "potential_double()" and "potential_float()", but identical to
> "potential_double()". It appears that adding "potential()" creates a
> segmentation fault (on returning from "potential_float()"). I don't know
> why, do you?
- mktemp() should be converted mkstemp() in stropen.c (recent gcc compilers
will complain about it, but it will work)
Problem: can't get a fileno from a FILE * (the other way around
is ok, use fileno()) so we're stuck....
- -DSINGLEPREC: scfm does not work
athan92 shows some significant force differences
testsuite nbody/init and nbody/evolve/flowcode have problems
? kawai/nbody does not seem to work, neither in IONEMO mode, nor default I/O mode
(at least on linux)
- fie: ifgt(%1,0,log(%1),-10) doesn't work,e.g. seems to work in nemoinp
but not via ccdmath/tabmath, e.g.
echo -10 | tabmath - - 'ifgt(%1,0,log(%1),%1)'
- x86_64: our configure doesn't always seem to recognize /usr/X11R6/lib64 should be used
instead of /usr/X11R6/lib ....
Examples: gaia fails, diary ok:
- cfitsio: currently broken, seems to write huge files in e.g. ccdfits
- magalie: hangs on gcc 4.1.0 ?? but this is the same gcc 4.1.0 in which
falcON mis-behaves, while on the laptop it's both ok.
- pgplot (or yapp_pgplot): if the sccreen aspect ratio is not the
classic 4:3, the plot is not square