Account Options>

  1. Sign in
Issues 2008-11-19
 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
18950DEFECTtriageP3garyog77 and gfortran not detected on windows
4
consensus2.xp3'+toolchain
5
6
Bill2.xp3"+toolchain"I agree with Greg.
7
Brandon
8
David
9
Garyagree w/ Greg
10
Greg2.xp3 +toolchainI think all toolchain issues should be consolidated as 2.x p3 (currently they are scattered as 2.x and future).
11
Jim
12
Ken
13
Sohail
14
Steven2.xp3'+toolchainagree w/Greg
15
16
21270ENHANCEMENTtriageP4issues@sconsos x linker supports -rpath in xcode 3 (os x 10.5)
17
consensus2.xp3'+toolchain
18
19
Bill2.xp3"+toolchain"Agree with Greg.
20
Brandon
21
David
22
Garyagree
23
Greg??? +toolchain?I'm not happy with any of the options, but determining the OS X version (which is something that really ought to be done during platform configuration) is better than other options. Maybe add it to the toolchain work, since that will revamp how platforms are set up?
24
Jim
25
Ken
26
Sohail
27
Steven2.xp3'+toolchain
28
29
22480DEFECTtriageP4issues@sconsmsvc.py throws stack due to REG_MULTI_SZ Environment
30
consensus
31
32
BillresearchAsk the user to try 1.1.0 and see if the issue is still there?
33
Brandon
34
David
35
Garyfixed by vs_revampvs_revamp invalidates this. Can't really ask random users to try that yet though. Make a checkpoint of it for testing maybe?
36
Gregno clue
37
Jim
38
Ken
39
Sohail
40
Stevenresearchcournapei think this gets sidestepped by the vs_revamp branch, and using the .bat file(s) instead of the registry
41
42
22500ENHANCEMENTtriageP4issues@sconsEnhance Variable() so it can use environment variables too.
43
consensus
44
45
BillI'd disagree with putting this in configure. Make can pick up env variables if set. That said, it should be possible to enforce that the environment won't affect a command line argument. So perhaps another keyword would be better and/or a flag on Variable(,check_env=boolean) would be the best solution? Default value would be False?
46
Brandon
47
David
48
Gary2.xp4Simplest API I can think of would be to have a function that did getattr(os.environ, NAME, DEFAULTVAL) and users could use that to set the default value, but that wouldn't be able to store whether the value came from the environ, default, or set explicitly.
49
Greg???On the one hand, I'd like to see the ability to probe shell variables for information, but on the other hand, I'd like to see it done as part of a general revamp/rewrite of Configure,
50
Jim
51
Ken
52
Sohail
53
Steven2.xp3???I'm moderately in favor of this, but not strongly enough to get excited about coding it myself. Right now letting a user override from either the command line or the environment is tedious and involves cut-and-paste code. Given how common an idiom it is, an easier way to do it would make sense. I don't have a better alternative than some sort of check_env= argument, although I'd be fine if someone else came up with a different API, too...
54
55
22510DEFECTtriageP4issues@sconsscons: *** 'NoneType' object has no attribute 'append'
56
consensus
57
58
BillGreg - I though it was policy that no patch could be applied without a testcase (either new, or fixing one that failed). Unless that policy has changed, then my request for such is not "rediculous monomania"
59
Brandon
60
David
61
GaryBill's right, but we don't necessarily have to get the testcase from the user; if we can see what's going on we can make one ourselves.
62
GregWithout studying it in depth, his analysis seems correct, Bill's ridiculous monomainia (OK, how about excessive fixation?) about a test cast not withstanding. His fix isn't complete (the flag isn't initialized anywhere and there are other places where it should be tested), but the alternative seems to be a semaphore.
63
Jim
64
Ken
65
Sohail
66
Steven1.x?p3????Seems right. I agree about needing a test case, although I hate it when the failure is non-deterministic and you're reduced to "run it N times..." That might be where we are with this. I could see putting my name on it.
67
68
22520DEFECTtriageP4issues@sconsfloating point --load-average
69
consensusanytimeP4
70
71
Billanytimep4sounds reasonable to fix this.
72
Brandon
73
David
74
Garyanytimep4should fix.
75
Greg1.xp4who?Humpf. I didn't know that we detected those options and generated a message for them. But he's right that it should be a float. Does the option processor support float args?
76
Jim
77
Ken
78
Sohail
79
Stevenanytimep4Should be trivial to fix. Yes, optparse supports 'float' as a type. Bill? Greg?
80
81
22530DEFECTtriageP4issues@sconsStrange escaped character output by scons to stdout
82
consensusINVALID
83
84
BillSteven - This looks like the readline issue I ran into which you looked at a while back. Has the fix been backed out by another patch?
85
Brandon
86
David
87
Gary
88
GreginvalidI'll bet that readline is trying to figure out the terminal type by sending out a couple of control characters and seeing what comes back. Probably varies by the style of curses support. In any event, not a SCons problem.
89
Jim
90
Ken
91
Sohail
92
StevenINVALIDThe user already updated it (8 November) saying the system was doing it and the report was invalid. I closed it.
93
94
22540PATCHtriageP4stevenknightTests fail on Windows when file path to python contains \t
95
consensus1.xstevenknight
96
97
Bill1.xp4Patch seems reasonable.
98
Brandon
99
David
100
GarySounds like Steven's working with him already.
Loading...
 
 
 
Sheet 1