Account Options>

  1. Sign in
Issues 2009-12-01
 Share
The version of the browser you are using is no longer supported. Please upgrade to a supported browser.Dismiss

 
$
%
123
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
|
 
Still loading...
ABCDEFG
1
IDVotesType/MemberMilestonePriOwnerSummary/Comments
2
3
24460DEFECTtriageP4issues@sconsVariables instance bug
4
consensus: not resolved last time
5
6
Billincluded in 2450 belowAgree with Greg
7
Brandon
8
David
9
Garyincluded in 2450 belowGreg: assume you mean 2450. (((Yes, I corrected it; dyslexics of the world, untie!)))
10
Gregincluded in 2450 belowI have to admit I don't know what the code is supposed to be doing, but Steven wrote it, so I'll defer to him. Variables.instance isn't used anywhere else as far as I can find, so the code doesn't seem to have any effect. See also 2450 below, which removes this line as a side-effect.
11
Jim
12
Ken
13
Sohail
14
Steventhe intent is to treat it like a singleton, but i don't think it actually works to do that. if the tests pass and the line's unneccessary, remove it
15
16
24850DEFECTtriageP2issues@sconsSCons does not rebuild executables when static libraries dependencies change
17
consensus
18
19
Bill
20
Brandon
21
David
22
GaryresearchgaryoI looked at this and think there is something actually wrong, although (as I noted in the bug and Greg has more info) there are various problems with the testcase. I propose to work w/ the OP to clean up the testcase and see if it still happens.
23
Greginvalid?First, there's no master SConstruct. Second, the SConstruct in prog doesn't invoke the SConstruct in llib, so it doesn't know how to build libhello.a. Third, even with all that fixed, it still doesn't rebuild hello.o when hello.h is changed, either from the top level or within prog. Damifino why. It may indeed be a bug, but the test case is pretty bogus.
24
Jim
25
Ken
26
Sohail
27
Stevenresearchgaryochanges to hello.c should definitely trigger rebuild, changes to hello.h only if CPPPATH is specified. sounds like there might be a problem here, garyo investigating sounds right
28
29
24690PATCHtriageP4issues@sconsAdded basic support for the LDC D compiler on Linux
30
consensus: 2.x p3 +Easy
31
32
Bill2.xp3Don't know alot about D, but if this is supporting a different compiler, should it be a different tool rather than modifying an existing tool?
33
Brandon
34
David
35
Gary2.xp3Agree w/ Steven
36
Greg2.xp3 +EasyTraditionally, for marginal tools like this one, we just apply any patches suggested by the community. I see no reason not to do that here. The patch and test need some minor work (instead of multiple Appends, only need one with a list; test should work on non-Linux as well), but I say add it soon.
37
Jim
38
Ken
39
Sohail
40
Steven2.xp3seems reasonable. a separate tool might be more "correct" but the differences in behavior are small enough that keeping it here seems okay to me
41
42
24700PATCHtriageP4issues@sconsDir.alter_target method failure with build_dir enabled
43
consensus
44
45
BillAgree with Greg. Is the user concerned with the functioning of non-public API and has submitted a patch to fix that?
46
Brandon
47
David
48
Garyresearchask OPWould like to see the user-level problem, like Bill & Greg, but I think his analysis is correct that there's a logic error in that code.
49
GregWhat user-level error does this cause?
50
Jim
51
Ken
52
Sohail
53
Stevenresearchask OPagree w/everyone else. this is likely a bug, but we need a user-level specification of the error
54
55
24710DEFECTtriageP4issues@sconsAllow "" and <> search paths to be different in CPPScanner
56
consensus
57
58
Billfuturep3Perhaps each compiler can specify a string which is substed to evaluate the search path in the proper order for each?
59
Brandon
60
David
61
Garyfuturep3Agree w/ Greg and Bill. We've made it this long w/o it (though once you point it out it's kind of obvious).
62
GregNeeds to be done, but I have no idea about when.
63
Jim
64
Ken
65
Sohail
66
Steveninvaliddevil's advocate: is this a real problem? do any real compilers actually implement radically different search paths for <> and "" ? in practice, i haven't found significant deviation from the behavior of both searching the -I options with "" always searching the same directory as the file first. tangible example?
67
68
24720ENHANCEMENTtriageP4issues@sconsadd check for existence of variable
69
consensus
70
71
Bill +sconf_revamp?
72
Brandon
73
David
74
Garyinvalidp4 +EasyI closed this, it's just SConf.CheckDeclaration (I think?)
75
Greg +sconf_revamp?
76
Jim
77
Ken
78
Sohail
79
Stevenp4 +Easy? +sconf_revampMight be better to just add it and get it off the list...?
80
81
24730ENHANCEMENTtriageP4issues@sconsadd test for size of variable
82
consensus
83
84
Billinvalid
85
Brandon
86
David
87
GaryinvalidI closed it.
88
GreginvalidAlready there.
89
Jim
90
Ken
91
Sohail
92
Steveninvalid
93
94
24740DEFECTtriageP4issues@sconsAlwaysBuild does not work on Install targets
95
consensus
96
97
BillCould it be caused by the Command having no source?
98
Brandon
99
David
100
Garyresearchno ideas here
Loading...
 
 
 
Sheet 1