Discussion:
Using Quicksilver?? You had better read this...
(too old to reply)
Nomen Nescio
2019-01-25 23:46:53 UTC
Permalink
re QUICKSILVER (latest version)

Summary: QS's "RANDOM" generation of remailers is anything but
random.


I noticed some oddities about the Auto / Random remailers that
QUICKSILVER (latest) generated in chains (i.e., *,*,*,* ... ).
Some remailers hardly ever showed up.

So I decided on a proper test.

These are the stats that were being used:

Last update: Fri 25 Jan 2019 19:55:00 GMT
mixmaster history latency uptime
--------------------------------------------
austria ************ 22:00 100.00%
banana ********** 23:30 100.00%
ciambella ************ 14:00 100.00%
congeries *****+* 27:00 100.00%
fotonl1 ************ 28:00 100.00%
haph ************ 13:00 100.00%
inwtx *#**#* 8:00 100.00%
kreti ************ 28:30 100.00%
paranoia ************ 26:00 100.00%
redjohn ************ 8:30 100.00%
roance ************ 23:30 100.00%
devurandom ++++++++++++ 1:46:00 99.72%
dizum ************ 16:00 99.66%
freierede ++++++++++++ 1:32:00 99.64%
kroken ++++++++*+++ 1:19:30 99.09%
eurovibes ++++++++++++ 2:01:00 98.56%

All the above 16 remailers qualify for selection per mix.cfg.
Frell, frell2, lambton, slow do not. No remailers are manually
barred. No chains are stated to be broken.

I had QUICKSILVER auto-generate 1,120 "random" chains of total
length 10, with frell2 as the Exit remailer for all chains. It
was done by encrypting a 180KB text file using
Chain: *,*,*,*,*,*,*,*,*,frell2;copies=10;
and doing this 14 times.

I did not use longer chains because that could affect the auto-
generation choices (QUICKSILVER's limit to chain length is 20
if all the keys are 1Kb and down to 10 if all are long).

Frell2 accepts mix input from all 16 remailers so doesn't skew
the last, 9th, entry in each chain's "random" selection.

So I got 1,120 * 9 "random" remailers chosen, excluding frell2.

In the 1,120 chains, there were zero occurrences of the
following pairs in any order:
(freierede devurandom)
(haph ciambella)
which is as expected (same machine)

I discarded 8 random chains plus 8 remailer-instances to get
to exactly 10,000.

10,000 / 16 = 625

So, I'd expect, with statistical variation, if the selection was truly random, each of the 16 remailers to occur on average 625 times, and the >99% confidence interval is about (570, 680).

But what I got was very different. I've divided the results
into 3 groups.

REMAILER N DIFF FROM EXPECTED (625)

VERY UNPOPULAR
0 haph 167 -73%
1 banana 205 -67%
2 freierede 224 -64%

SOMEWHAT UNPOPULAR
3 devurandom 550 -12%
4 ciambella 587 -6%

POPULAR
5 fotonl1 732 17%
6 inwtx 742 19%
7 kreti 747 20%
8 paranoia 749 20%
9 roance 749 20%
A redjohn 754 21%
B eurovibes 755 21%
C congeries 757 21%
D kroken 758 21%
E dizum 760 22%
F austria 764 22%

TOTAL 10,000 0%


Waddafook is going on ??

I repeated the study with a fresh install of QUICKSILVER and
with shorter "random" chains and frell as exit.

The results were essentially the same - same three groups of
remailers.

banana, haph, freierede - very unpopular, equally.
ciambella, devurandom - somewhat unpopular, equally.
(the rest) - popular, equally


Here's the raw data (1,120 remailer chains, abbreviated as
above i.e. 0=haph):
FADC9FADE
F1346A43B
46879D831
F69C8E416
A3FC931E7
169C6E19F
F38CFE53D
6D548D9CB
B3D48A6DB
4EFA971CD
3AD6A7CED
591CA4B7F
6DC369E69
45AEB13E6
BF74EAC63
4517E471C
B67A97461
4974A7FEB
FE7F94F73
F4DBEA968
1EFB3C851
53E1CEDA3
C58AFDBFD
8DC6D41AF
B94654FED
45F1C8534
654B7394B
F9BC7DE9D
18D78C5E4
C7FA4EDA9
DC63CF9ED
1C79E4FB1
EDF753CF6
68D139E38
186EDB71D
DA36C7857
718568BEA
8B75D834B
F76193BF8
FA15316EF
B437FE17E
EA57B6CE4
65BD39E13
F5865F946
A83F138E4
9A8B43F7B
AF375CAF7
38A6B85CF
EC7B413AD
8FE7F59FA
FE1AD719D
CA1F73F7C
E5B83F1D7
469EC1DA8
3EA8F9BAD
643E87E53
E31A89C85
4EF3516D5
D9CBFB3D4
8A6DB4EFA
971CD3AD6
A7CED591C
A4B7F6DC3
F5AEB13D6
E5865D946
A83FE1384
9CB3D947C
A93DBE31F
F5843E517
75F671FC9
831E6FC76
78BC4DF68
3B5AFDAB3
159F5B84D
A58359CFE
46387EB63
CBA835876
35681DBFC
3678D5975
4E63FD6CF
DA3B59F63
E7F41F948
71DC54B1A
6E53C7A19
1C7831846
E5D6A7CDB
EA8D3CE7D
F38FBEA78
C5EA9F7C6
70BFED68E
62CA60E50
C8E6A52BD
EA6C0A7DC
AF02BA820
72BC9DA6D
675BE08C7
FC85D065A
5876DA0FC
E7859A7BE
67B6FE5B9
6D70FC7BE
95C2ADCBD
EF87E8025
26D56CA7D
ADEF27DB7
2C759C7B2
AFC9A78E6
5AEFDE69F
F82F0EC69
7FC7E27D2
67DF6A958
8DB8EDBE9
DE08FC0F7
720A6F0AF
EB826BCD9
2690C28A2
985AF20E2
C9FEA5DB9
7AFDB70A5
F76C85C69
CED02C5F2
D2E76DB2A
CB5D7BD7A
EA26076AE
D798CFA62
795D7EA9D
B8E058C7F
5802CB79E
75AD8625F
7CFE0A296
850CF82C9
8A096D2C7
E68ED2A02
F7C9E0B6D
7EC0967AE
EC8A98F59
EDA60AD7F
8975D2059
E56025D2F
95F0A65A7
06F2C65FB
BAD7FEAF6
F2BC80FEB
5C76295B6
9B0EA75B2
FE0C72AF2
80752890D
FE2AD905C
69859F07E
E6726C0F9
6EFBC9F6C
72B86B52A
095BA6E96
A82B570E8
5798DA70B
A5BAEB586
0F27DFEB8
D627FE067
9728C52AF
C9D0F9DE0
BA509B578
B79C72D9F
FE9AB60C5
2EF9A02AD
6FC6EF79B
A92AF602D
6AE56D9C8
D58F5EC29
1A4EC8DB5
4BFCA84FE
C4F9AF967
897B5E689
CF7EF9E54
ADC6D875D
9DC2A8CBE
C71245C7B
9C51CD268
6A169B271
DB1482A5D
281B968A1
EDC61DBFD
9DF74DA8B
47F8CA8CD
9246D24A1
E8B42EC64
62DC75F41
5F1ED1956
6DC9DE5B7
E75F269FB
8E7A8245E
64DF1965F
BA2C129D6
C7FA2D651
E6145BC2F
2C14AE4C1
CA425E289
1AD71DF85
BE1AB16E7
2B8F2D78D
E527EB251
D215F687F
96E54B21B
B9FD4B5E4
768BE27C8
85A2E984B
FE5BAC6AC
97A679F84
5BF2B1AFB
49A2B6219
D857967E9
12A6529BA
89A791B96
574F981A7
D97B9FB68
B8AC281A7
BD6A76158
1C9E8FC4A
4CAD74C51
5FD2FDE29
D6814578B
128D2C918
5AC1792D1
C17DC47B6
7EB759C2D
6751724A9
45E2F81EF
6ADB526C7
6425C7B6A
157BC1A28
A75FDEC65
4B61DA71D
9D64F62CF
A54F74286
A41A57DFA
75496DE56
F8EF6C96A
174F7C647
86471CEA6
6C96ABD67
FD185265D
69FC2E1C7
F76AC97CE
F1894F7A4
2617D4B9E
456E8C18B
6AFB54B7D
FAB96FE1A
2CE7C18A9
DF5ECF5DB
6896086CD
8AEDBC09C
C6A5EF75A
E963F79AB
6EB7CF9DB
CA6B57B3E
7A90B37FD
39C8ACE39
6D083975D
73D53A0B6
F693CF893
5C0EB7D87
7638EDFAC
5AEF7CD93
9E89E0B59
89DBAEFDC
B05F35D65
A0795AD3A
0F3CEB370
F308AC50A
F85AF0CDF
CE360B8F9
5F89C0A9C
9C5D68390
56C0AF756
97DA63C5F
BC96D03AE
89067E8C9
783E5DB70
85C75E78A
03B0C37F3
BEC8BE38C
8B0A86D9F
658E59AEF
E973ABCA7
8BE73853F
BFE9AFD87
597F938C5
589DAEDCB
6AFB0650F
D9C38F507
E3765EC9E
DEA89CA50
DE35EF07F
F06DE5D85
86EC7F3C8
CA9BC0E69
FA70BEC39
A60B7836C
8BE58D703
3D69FD9ED
65BDEF80F
F8D7039BA
9C0F509C0
CE35E87DB
C6938D93A
CA8EC0D56
DAB9CB98F
06B9C53DB
CA6738CBD
89C875AF7
87CBDCE60
E8A30FA86
39D30EA5E
C6FBC76C8
6AF037897
A7EAF06BC
7AEC0EBAF
6A3B58FC8
EF937EAB8
507B6A098
C7DF0E936
83BA8E608
AF039DC07
5BD067A60
5A6875E0C
3A907A95D
7FC7E0367
E9BDE87FA
47F94B6C7
473F93E54
48DF3594F
CAF39C645
6BC943EAB
59C76CB98
AB439DCF7
378FAEBFD
6AF8C34EF
3CA637DE4
3A6C985D4
D9AF9E53E
8AEB4E638
9D7E8C659
639A347FE
35B69875F
86E7FA4E8
D4E564E36
B354B639D
AD3C9EB3D
ED5F98B93
435A65CB9
EFC539AC7
AC8AE7D83
4EC89EA34
DF6AF87C6
DC58D9BA7
8F73B54D3
47C57BE98
58B4869E5
598FA7B47
39786D869
9AF948D5F
38D4A986E
DB3EB48A7
57DEB6E39
C4DA3FD94
793784D5E
B5935ECD9
34D95E8AC
A89F5AE74
BC9F4CE8D
56934CA39
5FB37BF5A
BF4E6D975
C6345F645
DB5CA76F4
95D49DCB7
A83B8D3C6
3B7CD67AE
C8FC947FE
634584EB5
BAF8A7EA4
F8E3DC49D
7A6379BF6
CF7E9D5AE
83BAF975B
9C84FA643
B78379BF7
BA74BC54E
B476598CE
CEF4ECD4C
DC7EB378B
6C57FE7D4
8E6BC4DBC
A3EC37E4F
96DB3A4DF
46CF6476A
4EB7DC6F9
5FA96876F
5396FE45C
DAE43CDBA
AFC58BA8E
B3AB698B4
CFE5385FC
7BD5E34EB
7B396F475
8C7896CE8
73DF856D3
8F7BD97BA
C2782ABEF
B5E9D097E
D7F5E7DEC
70DA8C690
F0E5D6C2A
0A568FCE8
7C2786B70
7BF9C59D8
2FD275D09
20BA902A6
598AC9762
6DBED08EA
BA5DAE8F6
DCF89DE58
AE9570A98
75C9B05BF
07CBD592E
2E6A57C62
7AFC2FA82
F0E50FCB2
2E6C5D8C9
2F607E628
85E956FBC
728F75F75
2B6C782BD
DB89C6A89
E08EAD07D
CD07F27BE
7B58BEF8E
0B6CA80ED
2EAF0DC2D
9AC769BE6
F85E2C58D
ECB0528F7
0AD9E0F2C
DC065F08B
9FAC78A9C
C0F27AB0E
9BC08F952
F08A05E8B
20D7A5CD7
ACFA28FA8
79A0758E5
90F80E7B9
20CFDCF25
B657CB9DC
86BD927EA
F90FDCE72
E50A5C6DE
C2E57B09D
9FE2BED69
28E90BD9F
029F2DB7A
78F9D026A
CD6EFB2F9
7EBF72A59
70DC062E8
D6B078BC9
B5E9BF9C8
7DEABCAE8
97AD7E8CF
69D2EDF05
E7B59CF96
A02B907E5
826F9EB27
6BA0D72C0
9CE209DE8
8FEA79C28
0BFADFEA9
F75980C80
897B25EBC
C726DE76F
D96F27CD9
ADEFCD8AF
6759EF6DB
F7DFEDF6E
56BED5BD2
C5680A9F8
CD520CE6D
E829E569A
384E57DF9
F31E8F5C3
1D837B94B
EB5FE978B
5E8F38BA9
61534E693
A4D56ACF3
48618367E
3D8E3C5EA
FB98A7B8A
6C738D34B
1F9D71FA7
FAB36A598
57BCFBCA9
394EAF687
7356FA6E8
B6DAE7B8C
FE4567FB9
48BA584CE
CF8B94C1A
5CAB58DF9
579F178A6
E6DE4AB96
ECF8E5DF1
4D97DB869
89614F6B3
6A7C45F19
35C9BD93D
EF518F65E
E4DEBA87A
9C7AE1FAC
D1E5C9AB9
48D3F7EA1
863E47CD6
871FDC9E5
D3B835BC4
4C7DF4D1B
9F3CA91A4
1E43DF8AF
A781BF4DA
9FAD438E1
ECB9F461D
17AB8FA53
4DF85B983
E14D697E8
D84B85A6F
FC5B4C9EA
16F518968
DE3AED35A
A56E365CF
BED63C9ED
C5AF3159A
B8F5C9457
A54C15DAC
68F94C546
39BA31CBA
DCE7B6C3B
D6376D4B1
D65C9DA67
4BAE76B79
C37865A9B
475F3BED7
DEF9CDFEA
DCAF18E53
C1DA4DC5F
C94D167E8
469519EF1
D36C56C48
4A6FD5BA4
83F8A9BF3
AF4C75CBF
BFD1873FE
8C6EC16E1
3B64B5F19
BEF1A431F
6C568ECF4
8EDF1C317
B58E3A694
4B6FE6A85
EFBD183FA
4CD893F8A
7958D5BD7
95F65DC93
B6F47ACFB
A89AF7A3B
438E6D5A8
A96DFA56B
FACBE73F8
FB6FBE7D9
85F68457F
584E8B7A4
34BECD748
3E89CB94B
5DB5EBA8D
7856BE69F
A3953F7ED
D37F54856
5DA7593BD
948A38DEC
CE87F835E
CAEDF36C9
DF39B897D
FB938E9B8
A8CD975FC
3F65BE8DB
A9B5D647A
7A9F6789A
85479359C
D6874BC6F
8DB453BCE
B94EC3FA4
5CE4593C8
583FE4D64
BDF694E35
ABFE4CFBD
387AD56D8
AC4A5FE73
5BF4BE6C4
5C689C3B6
DABDE8B94
F483F63BE
9C7E49AE4
78A46A3C9
CF6CEDF79
CF6A9CDB4
8A95ABD8E
89FA4FD7A
8ADCFA5D7
3DAE65948
9C859BEC4
FE395A3B8
AF9DABD63
89BEA357F
BE7D3EC5E
DFB59FECD
FC89F7649
4A6EF5C6A
7E6CF6D97
AD65DB9F4
47B45AD3B
539E6A43D
B6D94D7B9
D3489C89A
6B543E956
497B643A6
4FD854AB5
C6FB75DF7
6CB6A7BD9
DE3D46AB5
E45FCD45A
53CF5BDA6
49B374B5F
76D9E54B5
F67957B46
3F7BE4367
7CEBC93EB
853E9CE5C
679EAC54A
EC437657D
6FEA64756
4BD367CF9
B57BCD7E8
A75E8AB47
6E3A4F37A
E4693CAF8
FC9B694F9
948FE6F86
BFA6B7E37
437E54C38
A8C749F74
75EB6DF9C
637D8CA3F
E97B6E7B4
675F8E5F4
BD3EA3C6B
FCA493F79
E8A74A7DA
E73CA8B97
95AB3E96B
DE83FE7DE
CE576AF5E
B5DA5F7C6
BECF8CB68
78463DB45
F4A7D6BC6
836FBD6C4
9D594739B
AECD648FE
7657B3F75
FE68F7DFB
7BF95FB43
CD8FB4C83
E37F96CDA
B839654C9
A6798CBD5
6E93563FC
ECB38FB8C
9F3753ECD
4EB986573
B4368C3EF
86BDF56DC
D46ED5A89
8AB3DFABC
C6DF8495C
A7436758D
CAF5ADBEF
E3AE9D36D
5E8B56AF7
EDC4D36CA
E7598DB47
E9D83DCF9
D3F8DA4E9
47C3F84F5
4E3BDA63C
6BEF9E3F6
3C6738C47
DCB5FB95B
B6475CF85
E89374AB5
4EAD46AF4
5AFC6D8F5
9FA574DCE
83947BF8E
F68C7DA49
963B796B4
AB6CB6A4F
EC36DB65E
3C6FB9A4B
F5B3A4B39
EAB75E8B5
6CB5F83C6
4FEA4EFC8
D6EDA6D86
EC8E47EB6
475E9AD58
F45E7A687
5C74F9B3A
85F746D96
69BC8AE9D
ED3BE734D
5E4A3DBE7
9D6AED95B
85E643854
FB43ADB84
A84CB73E6
46F87FBDF
CEF86AC73
FA9F846CD
F59D49ECB
ACBA43E95
7E964A54B
FB45698F7
35DF8934B
85DB67DE4
CFA5D9745
48E76A3F6
735A976D4
B74AB975C
AC76A5B8E
CF69BAC39
54C5897EA
AC7D8B6DC
35A3CEBDF
3A89FC6D9
D45F4D74C
C3A8CD893
639AC6AF7
98A967F5B
AED8B785A
437CBA937
E4A968CB9
3B8C56495
657F8BE4D
8B9FED8B7
A95B4FB3D
D63B73C47
ED9B84BAD
ADB3D5879
EB75CEB98
7ADCFB5AB
5E4A9B634
FDE8A46F9
84FB368C3
CB7EDFC86
7BA975F78
F3683CB6E
A864F8B74
8AD8A756A
36A75CA6F
6AD3CA95A
3947DA3E5
AF9A3E7A4
69EF835DE
F7E39F864
6D8A64BAE
E67DEAFCD
6CA3F86F5
3485F8BD8
78A36C4BA
CB36597FD
76DC58BD4
D39EA746E
86975EB56
564DBC7A8
AC8E976FA
D68A4FA69
87398B7CB
AD4BAE687
768C765F8
56EB3A9DC
BCE45C9FD
C43C54C79
3AFB478E7
4AF54D85B
E3DF543C5
5E87536F3
583AFDC38
C37BF56CF
583F73DBC
67469AF64
75FC58D5C
4C9B58ED9
3AC58D69A
3B73E78DC
C8DA6EC6F
68FD39E35
67D9CBE85
E683CAD45
AB57ED64D
39859B3DF
3F759C8A3
DA6EA684A
B9FE76FC3
86E9CBA95
A65D7AF59
F9749C795
ACD96E7D6
384E89AB7
98C5943E6
5DC948FE9
8473DA736
DCA7DF79B
74F8C4FE3
4FDB8356A
F98C43C59
9BE8F7B39
8496A8DAC
4BC5B3645
ABF93EBDF
473A69BF9
8D564F9BC
8578CE89D
3C945BAF4
96B9D8BDC
95D7BCA68
E496EDA93
B8FDB63BA
A7E4BF385
4D39E847A
6E5C6D9B5
964CDB5EC
357FC4A85
3645C6DEB
67AF8B9A4
E7845B376
DC4B9AC9E
45AC736C8
3D94A96AE
5A8C6EB6A
EC459B8E6
E5A84CA96
E4AE64D69
9BF7B4C5F
F63E64C8E
5E6BD56EF
95A6D375B
B745BA3EF
473BD69FA
9DF395ED4
48F7EAC63
AE94584C8
94C67DC54
5F9DC4F98
45EB8AC64
CE5BC4794
39DB58D73
59B65983A
9EF9BD86C
3BE546D7A
F9EF5D8CD
BD3FBC94D
ACF53F9A7
A5E43BE6A
DE6B4A8EB
D65FE3D93
3895CA9B6
3C74A9BA4
9FBDEF57E
3AE5D3CAF
F47DA6435
8C6528B6D
D2AF4C6EC
28E475C4E
76DFAC48A
E627DF5AD
D8274E9A7
E57F8B4DB
C52D9F2A7
45B9A8F6E
8C5BE569D
B76A9D8E2
829E2486D
57BE2BD59
FEDA9F875
A2E975698
A45DB9E2C
E7BECF9BA
487FA869B
AE97E6CF8
A45A4BC8B
5B862FB98
B852672EA
7C82C97DF
5CE6F9E8C
B2CEF68BF
DFBA48FEB
456E4827E
BC248E7A9
DF6D8BADB
2E42C5B25
45CB8FC7B
8D526EC27
DF24CFDC8
F2B49C7B8
FE92B5F27
7B6C9A2DF
FA24BC8B9
45E9AD4EF
B65BF29F8
D5976DC97
462E8CDFC
B29DEFC9A
692D7645A
CA48E7BE9
6EA26FBAD
6F894568C
62E8BC628
4EBA2E87A
BD6B95A74
C9ED5BA69
7CE57FC59
AE846BD62
8F6D5A429
58F5EF5A4
C4A8B469D
B956BA64B
A5D9A8F6B
F824AB2DB
A568B2DEA
C4869AED7
7B842CE6B
5AFB5964C
72B46DABE
A426EB8FB
67C28D42A
C957DE4F8
DBC57F9DF
9AB84C25E
9A5DE75A8
278BCDB7D
E678E52BD
8DF8CD7BD
45D975D8E
EB4D87485
52D65DF5E
A45D287FC
9B2CF248B
26D4CE97B
8C726EF9E
F26D9EF56
A594569EB
C43ADEC7A
9E834EB4F
8FA3164A7
5DBC8B7E8
B5A1EF56D
36C5ADC9A
83F9C71A3
9769C1A3E
C1BA78493
DB5638D36
ECF6B17E3
CD5EF6C86
7DBADE51E
EDA9FBC93
3E7B3A8DB
8963CB49F
AEB548FA3
7EFDEC187
DE6DE7A5F
38E1941D9
C591BD183
DB3A47346
543B6AD34
89A15FB1E
8FACDA467
98BC7B51A
39AB31CD1
F7B476C94
DC51E698B
A1396CB31
16EBC984D
D3F5C839B
18FD5CFE9
349DB7AFD
6AE76153C
164BD6E5C
4BDEC941B
F9D1E7DBF
A43F493D7
D35D45AD7
F4AD831ED
9BF1C49C7
148D5C19A
A54875846
C79138694
F1A4864A6
F6357A431
A1E7CFB1E
4DC4AE453
9AFDA6D53
D5C1FD6A4
5863BD453
1ADB1DB94
D73C7EA85
1C3E47DA8
8B3978531
CB713A9F8
C871A34C8
9E83F1641
31AD1731C
AE7FE76A5
AE3CD9B8D
DA4E5D75B
BCD8A3967
4E98E9F5A
39561C784
C9DE58AED
158CB914A
316BAC957
B1C76E7F5
4A57F5865
FD98FDC69
9F1DA5BC4
E3DB8D56E
3FCB36E7A
A5DC81B4F
9C4BF786E
CE6CAB1C7
376A8639C
EC395FE4D
386E43B73
36AFCA9D7
4CDEB9F7B
8D4AC536D
5D7963A7B
58FD764CF
5C4ADC65C
9C4B39AC6
4F97CD7A3
D935BEDCB
CBD953DC9
CD96D8AE7
7E973DBE8
BECB58A3B
9CDB6456A
6C36A3FCD
39C59E45C
EB958C39B
A8CBA93A8
E978537D3
AFBEF9847
859438DE7
6FD8CB4F8
F4E8D45E9
5E47D54D7
8F37CF7BE
6ACEF3B47
3A74879ED
D6E8FD3B7
9CAEDA7FA
C9BE549E3
49DE4D9BE
A7C5A3684
79CE48DBF
D8CABD6C8
C346A8563
684EFC4A7
B57BFEB47
4EB7D3BF4
35D4B8DF9
8C794E598
BA7F47563
4B378BC7F
EFC7A95E6
C6BCAE6CE
47C65D8CF
5AFDE8A3F
75A3C957F
DB84D6C5D
73965F3C6
9BA4C384C
E6B3CF5CF
476D56394
EB873CB68
49D3E98FB
3D76386B5
3AF5945EF
3796DB96F
A9F58C54F
8F96473B8
8F7B94F37
5BA86C5D8
D8ED7F543
5FC36BC6E
385C86574
BADEF9537
AF65DCED4
43F6C5684
BE93CDB58
E65B8DC49
8FC83BC7A

Till this is resolved, I'm going back to JBN. I kick myself
for not having done this test years ago. What is Mr Christman
doing? And, why?

I suggest you repeat the experiment, you will reach identical
conclusions.
Anonymous
2019-01-26 01:09:15 UTC
Permalink
Redo the test excluding haph, banana, and freierede and see what you get.
Nomen Nescio
2019-01-26 12:07:59 UTC
Permalink
In article
Post by Anonymous
Redo the test excluding haph, banana, and freierede and see what you get.
he was too stupid to figure it out the first time.
Anonymous Remailer (austria)
2019-01-26 15:53:56 UTC
Permalink
QS per se does not select the remailers, the MIXMASTER code invoked by
QS does this.
Post by Anonymous
Redo the test excluding haph, banana, and freierede and see what you get.
Your point is that restrictions on neighbors by haph, banana and
freirede are responsible for the 'anomaly'. I considered this for a
minute, but while relevant it cannot account for such a big variation.
Do the math.
Nomen Nescio
2019-01-27 00:34:03 UTC
Permalink
Post by Anonymous Remailer (austria)
QS per se does not select the remailers, the MIXMASTER code invoked by
QS does this.
Post by Anonymous
Redo the test excluding haph, banana, and freierede and see what you get.
Your point is that restrictions on neighbors by haph, banana and
freirede are responsible for the 'anomaly'. I considered this for a
minute, but while relevant it cannot account for such a big variation.
Do the math.
Your ignorance is showing. Quicksilver has absolutely nothing
to do with it. It's all on Mixmaster.
Anonymous Remailer (austria)
2019-01-27 12:42:49 UTC
Permalink
Nomen Nescio wrote...
Post by Nomen Nescio
Your ignorance is showing. Quicksilver has absolutely nothing
to do with it. It's all on Mixmaster.
Is there some part of what I wrote

"QS per se does not select the remailers, the MIXMASTER code invoked by
QS does this."

which you cannot understand?

QS's install package includes mixmaster code, which is invoked by it at
runtime.

It is noted you have no explanation for the phenomenon, only hot air
bluster.

What knowledge is MIXMASTER using to make the selections?
Nomen Nescio
2019-01-27 18:13:51 UTC
Permalink
Post by Anonymous Remailer (austria)
Nomen Nescio wrote...
Post by Nomen Nescio
Your ignorance is showing. Quicksilver has absolutely nothing
to do with it. It's all on Mixmaster.
"QS per se does not select the remailers, the MIXMASTER code invoked by
QS does this."
MIXMASTER code
Post by Anonymous Remailer (austria)
which you cannot understand?
I understand you haven't examined any of the runtime ASCII files
notes and docs in the QSL folder.
Post by Anonymous Remailer (austria)
QS's install package includes mixmaster code, which is invoked by it at
runtime.
If it didn't it wouldn't work.
Post by Anonymous Remailer (austria)
It is noted you have no explanation for the phenomenon, only hot air
bluster.
I don't see it as a phenomenon, bug or anything else. I see it
as a pointless whining exercise over your attempt to break
program code - and failing.
Post by Anonymous Remailer (austria)
What knowledge is MIXMASTER using to make the selections?
The same parameters it always has, all based on your template
choices.

mix [-hpmdSv] [-t ***@host] [-g newsgroup] [-s subject]
[-a filename] [-n nym] [-l mix1,mix2,mix3,...] [-c num]
[***@host] [filename]

mix [-f[rfg] filename]

mix --nym-config=yournym [name=Pseudonym] [opt=nymserver
options] [--latency=hours] [--reply-chain=rem1,rem2,...]

Now run along and go play in the traffic.
Anonymous
2019-01-27 21:34:23 UTC
Permalink
Post by Nomen Nescio
Post by Anonymous Remailer (austria)
It is noted you have no explanation for the phenomenon, only
hot air bluster.
I don't see it as a phenomenon, bug or anything else. I see it
as a pointless whining exercise over your attempt to break
program code - and failing.
Now run along and go play in the traffic.
That the O.P. was confused about what QS did and what MM did is
irrelevant. The oddness remains.

In repeatable experiments, I checked it too, MIXMASTER chooses 3 fully-
qualifying remailers (banana, haph, freierede) to fill random positions
only one-third as often as would be expected, and you find it unremarkable?

Factors like chain-adjacency, exit status etc were eliminated as
explanations.

Stupid fools (like you) shouldn't be in apas or using this tech. If you're
a spook, your doing a bad job, because suspicions are heightened.
Post by Nomen Nescio
Post by Anonymous Remailer (austria)
(banana slow hsub)
(freierede devurandom)
(frell frell2)
(haph ciambella)
So what caused the "anomaly" re banana? Neither slow or hsub
are in the selection pool.
Why was freierede low, but devurandom not?
Why was haph low, but ciambella not?
And why were the low sooo low?
Good questions. Anyone has answers?

Here's one. MM is somehow NOT choosing remailers which are uncompromised,
like Zax's.

But you wouldn't understand.
Nomen Nescio
2019-01-28 07:20:59 UTC
Permalink
In article
Post by Anonymous
Post by Nomen Nescio
Post by Anonymous Remailer (austria)
It is noted you have no explanation for the phenomenon, only
hot air bluster.
I don't see it as a phenomenon, bug or anything else. I see it
as a pointless whining exercise over your attempt to break
program code - and failing.
Now run along and go play in the traffic.
That the O.P. was confused about what QS did and what MM did is
irrelevant. The oddness remains.
post the exact template you're using.
Post by Anonymous
In repeatable experiments, I checked it too, MIXMASTER chooses 3 fully-
qualifying remailers (banana, haph, freierede) to fill random positions
only one-third as often as would be expected, and you find it unremarkable?
yep. sometimes a remailer is momentarily unreachable because of
high utilization. they aren't just sitting there idle waiting
for somebody who expects fedex p1 smtp service every post.
servers get saturated.

i personally hammer the fuck out of the remailers and some of
them won't accept mail from any sources other than remailers.
most if not all tor exits are blacklisted by the middleman
remailers. i don't see that as a statistical issue because
point of submission has zero bearing once a post enters the
remailer network.
Post by Anonymous
Factors like chain-adjacency, exit status etc were eliminated as
explanations.
your testing is flawed because you can't control the progress
and end result of certain variables.

you're using the same kind of crap modeling theory as the global
warming assholes. the results are slanted towards what you want
them to be, not what they actually are.
Post by Anonymous
Stupid fools (like you) shouldn't be in apas or using this tech. If you're
a spook, your doing a bad job, because suspicions are heightened.
i'm using it because i know and trust how it works. you're
trying to figure out a way to manipulate (attack...) it and
validate your premise.
Post by Anonymous
Post by Nomen Nescio
Post by Anonymous Remailer (austria)
(banana slow hsub)
(freierede devurandom)
(frell frell2)
(haph ciambella)
So what caused the "anomaly" re banana? Neither slow or hsub
are in the selection pool.
Why was freierede low, but devurandom not?
Why was haph low, but ciambella not?
And why were the low sooo low?
Good questions. Anyone has answers?
Here's one. MM is somehow NOT choosing remailers which are uncompromised,
like Zax's.
"uncompromised" - lol.
Post by Anonymous
But you wouldn't understand.
maybe you should use a different host. try
sourgrapes.mixmin.net.

recommend you repeat your exercise in futility using qsl and
post via omnimix anon mail server. it has exceptionally good
logging - if enabled. up to you to figure out how to do that.
if you know anything at all about smtp, this should be a snap.

i'll give you one hint, don't use random with your test
parameters, you'll be waiting until next year.

read the logs.

should be enlightening.
Anonymous
2019-01-28 12:19:14 UTC
Permalink
Post by Nomen Nescio
In article
Post by Anonymous
Post by Anonymous Remailer (austria)
It is noted you have no explanation for the phenomenon
Insults removed.
Post by Nomen Nescio
Post by Anonymous
That the O.P. was confused about what QS did and what MM did is
irrelevant. The oddness remains.
post the exact template you're using.
Here is --my-- QS template:


Host: [valid onion one, irrelevant]
From: ***@anonymous.com
Chain: *,*,*,*,*,*,*,*,*,frell2;copies=10;
To: ***@anonymous.com
Subject: None

[text file of 270 Kb so as to generate 11 * 10 Mix chains per SEND]


I did 11 iterations, to get 11 * 10 * 11 * 9 remailer instances
(excluding frell2), and took first 10,000 remailers to do the stats
on, 10 times bigger sample than the O.P.'s.

SEC3 mlist etc.

mix.cfg set to exclude, on grounds of latency or reliability, only
frell, frell2, lambton, slow.

Exactly the same results 4 days later than the O.P. haph, banana,
freireide chosen only a third as often as should be expected. No
significant difference between those three. ciambella, devurandom
chosen about 10% less often than should be expected, but this could
be down to some same machine exclusions for adjacent remailers.
Post by Nomen Nescio
Post by Anonymous
In repeatable experiments, I checked it too, MIXMASTER chooses 3 fully-
qualifying remailers (banana, haph, freierede) to fill random positions
only one-third as often as would be expected, and you find it unremarkable?
yep. sometimes a remailer is momentarily unreachable because of
high utilization. they aren't just sitting there idle waiting
for somebody who expects fedex p1 smtp service every post.
servers get saturated.
i personally hammer the fuck out of the remailers and some of
them won't accept mail from any sources other than remailers.
most if not all tor exits are blacklisted by the middleman
remailers. i don't see that as a statistical issue because
point of submission has zero bearing once a post enters the
remailer network.
See next comment by me.
Post by Nomen Nescio
Post by Anonymous
Factors like chain-adjacency, exit status etc were eliminated as
explanations.
your testing is flawed because you can't control the progress
and end result of certain variables.
Irrelevant. Neither I nor the O.P. was measuring successful
propagation, only what "random" remailers were chosen by mix.
Post by Nomen Nescio
you're using the same kind of crap modeling theory as the global
warming assholes. the results are slanted towards what you want
them to be, not what they actually are.
No. I'm smart. mix and/or whatever Quicksilver has tacked on to it
(this could be some bug or trap at the QS end, overriding mix) is, at
least currently, discouraging those three or five remailers.
Post by Nomen Nescio
Post by Anonymous
Stupid fools (like you) shouldn't be in apas or using this tech. If you're
a spook, your doing a bad job, because suspicions are heightened.
i'm using it because i know and trust how it works. you're
trying to figure out a way to manipulate (attack...) it and
validate your premise.
It doesn't matter what my motives are, or the motives of others
who've now noticed the same phen.

The concerns are genuine and if you cannot see them, then you're
either stupid, a troll or a spook, or some combination of them. That
too is irrelevant.
Post by Nomen Nescio
Post by Anonymous
Post by Anonymous Remailer (austria)
(banana slow hsub)
(freierede devurandom)
(frell frell2)
(haph ciambella)
So what caused the "anomaly" re banana? Neither slow or hsub
are in the selection pool.
Why was freierede low, but devurandom not?
Why was haph low, but ciambella not?
And why were the low sooo low?
Good questions. Anyone has answers?
Still waiting. Bluster and insults and irrelevant "do this instead"
noted and ignored.
Post by Nomen Nescio
Post by Anonymous
Here's one. MM is somehow NOT choosing remailers which are uncompromised,
like Zax's.
"uncompromised" - lol.
Post by Anonymous
But you wouldn't understand.
maybe you should use a different host. try sourgrapes.mixmin.net.
recommend you repeat your exercise in futility using qsl and
post via omnimix anon mail server. it has exceptionally good
logging - if enabled. up to you to figure out how to do that.
if you know anything at all about smtp, this should be a snap.
i'll give you one hint, don't use random with your test
parameters, you'll be waiting until next year.
read the logs.
should be enlightening.
It would be irrelevant. Look up Logical Fallacies and count how many
you've committed in these last two posts.

I did the test too with JBN2 (much smaller numbers, though) which
prog calls the same Mixmaster code and there, RANDOM (and AUTO)
appear to do exactly what they say. No unpopularity of the three (or
five).

That is the trump card argument.

You are displaying the same blind fanaticism and faith as a religious
fruitcake or MMGW crackpot. I forgot to add, Jesus sucks Mohammed's
dick in Paradise, while Krsna jerks off watching, and it's getting
HOTTER and HOTTER and it's all man's fault.

Do the test yourself and or come up with an explanation.

Till then, fuck Allah and the Tooth Fairy off. Or just fuck off
yourself.
Anonymous
2019-01-28 23:16:20 UTC
Permalink
Post by Anonymous
mix.cfg set to exclude, on grounds of latency or reliability, only
frell, frell2, lambton, slow.
Exactly the same results 4 days later than the O.P. haph, banana,
freireide chosen only a third as often as should be expected. No
significant difference between those three. ciambella, devurandom
chosen about 10% less often than should be expected, but this could
be down to some same machine exclusions for adjacent remailers.
Run by the same remop(s):

- ciambella & haph
- banana & slow
- devurandom & fotonl1 & freierede
- frell & frell2

When in your test system QSL creates a starex configuration that
reflects these associations you get lower occurrence propabilities for
exactly these remailers.
Nomen Nescio
2019-01-29 01:01:20 UTC
Permalink
Post by Anonymous
mix.cfg set to exclude, on grounds of latency or reliability, only=20
frell, frell2, lambton, slow.
Exactly the same results 4 days later than the O.P. haph, banana,=20
freireide chosen only a third as often as should be expected. No=20
significant difference between those three. ciambella, devurandom=20
chosen about 10% less often than should be expected, but this could=20
be down to some same machine exclusions for adjacent remailers.
- ciambella & haph
- banana & slow
- devurandom & fotonl1 & freierede
- frell & frell2
When in your test system QSL creates a starex configuration that
reflects these associations you get lower occurrence propabilities for
exactly these remailers.
FFS, are you lazy, a troll or a moron?

Read what the OP and three others, including me, have written above.

This avenue has already been explored right in the first post.

frell, frell2 hsub and slow are NOT in the allowed list per mix.cfg.

Copying from above:--

Per SEC3, groups of remailers sharing a machine or operator:
(banana slow hsub)
(freierede devurandom)
(haph ciambella)

So what caused the "anomaly" re banana? Neither slow or hsub are in the
selection pool.

Why was freierede low, but devurandom not?

Why was haph low, but ciambella not?

And why were the low sooo low? (one-third)


Stupid morons please excuse, these discussions are not meant for you.
Nomen Nescio
2019-01-29 11:50:03 UTC
Permalink
Post by Nomen Nescio
Post by Anonymous
mix.cfg set to exclude, on grounds of latency or reliability, only=20
frell, frell2, lambton, slow.
Exactly the same results 4 days later than the O.P. haph, banana,=20
freireide chosen only a third as often as should be expected. No=20
significant difference between those three. ciambella, devurandom=20
chosen about 10% less often than should be expected, but this could=20
be down to some same machine exclusions for adjacent remailers.
- ciambella & haph
- banana & slow
- devurandom & fotonl1 & freierede
- frell & frell2
When in your test system QSL creates a starex configuration that
reflects these associations you get lower occurrence propabilities for
exactly these remailers.
FFS, are you lazy, a troll or a moron?
Read what the OP and three others, including me, have written above.
This avenue has already been explored right in the first post.
frell, frell2 hsub and slow are NOT in the allowed list per mix.cfg.
Copying from above:--
(banana slow hsub)
(freierede devurandom)
(haph ciambella)
So what caused the "anomaly" re banana? Neither slow or hsub are in the
selection pool.
Why was freierede low, but devurandom not?
Why was haph low, but ciambella not?
And why were the low sooo low? (one-third)
Stupid morons please excuse, these discussions are not meant for you.
five questions in your post. who's stupid?
Nomen Nescio
2019-01-29 17:07:03 UTC
Permalink
Post by Nomen Nescio
five questions in your post. who's stupid?
Predictable baboon-pack response when the groupthink is challenged, even a
bit.....
Sir Gaygory's Owner's Owner ๐Ÿถ็ฌ›
2019-01-29 21:14:29 UTC
Permalink
On Tue, 29 Jan 2019 12:50:03 +0100 (CET), LO AND BEHOLD; Nomen Nescio
<***@dizum.com> determined that the following was of great
importance and subsequently decided to freely share it with us in
<***@dizum.com>:

โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก In article <***@dizum.com> Nomen Nescio
โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก <***@dizum.com> wrote:
โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก In article <***@haph.org> Anonymous
โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก <***@haph.org> wrote:
โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก On Mon, 28 Jan 2019 07:19:14 -0500 (EST), Anonymous wrote:
โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก mix.cfg set to exclude, on grounds of latency or reliability, only=20
โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก frell, frell2, lambton, slow. Exactly the same results 4 days later
โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก than the O.P. haph, banana,=20 freireide chosen only a third as often
โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก as should be expected. No=20 significant difference between those three.
โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก ciambella, devurandom=20 chosen about 10% less often than should be
โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก expected, but this could=20 be down to some same machine exclusions for
โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก adjacent remailers.
โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก Run by the same remop(s): - ciambella & haph - banana & slow -
โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก devurandom & fotonl1 & freierede - frell & frell2 When in your test
โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก system QSL creates a starex configuration that reflects these
โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก associations you get lower occurrence propabilities for exactly these
โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก remailers.
โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก FFS, are you lazy, a troll or a moron? Read what the OP and three
โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก others, including me, have written above. This avenue has already been
โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก explored right in the first post. frell, frell2 hsub and slow are NOT in
โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก the allowed list per mix.cfg. Copying from above:-- Per SEC3, groups of
โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก remailers sharing a machine or operator: (banana slow hsub) (freierede
โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก devurandom) (haph ciambella) So what caused the "anomaly" re banana?
โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก Neither slow or hsub are in the selection pool. Why was freierede low,
โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก but devurandom not? Why was haph low, but ciambella not? And why were
โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก the low sooo low? (one-third) Stupid morons please excuse, these
โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก discussions are not meant for you.
โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก
โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก five questions in your post. who's stupid?
โœกโœกโœกโœกโœกโœกโœกโœกโœกโœกโœก

it's too bad that no matter what anon posting method he uses, he still comes off as a complete asshole.

ask him about his "hawt" wife. LOL
--
[THIS POAST HAS PASSED TRIMCHECKยฎ VALIDATION]

THIS SPACE FOR RENT


"Thanks to muzzies and their apologist-enablers like puppy whistle, this
seems to be the new norm in the world. It's spreading like a cancer,
and it's time we admit we're at war with pure evil. We need to put an
end to this muzzie plague, or life on Earth is going to become pure hell
everywhere. We need to get these people out of every civilized
country, and there's only one way to do it. IOW, we have to become
like them, with an emphasis on expediency over cruelty." - Checkmate (of alt.checkmate)

"Pussy Willow has just proven that Trump's crackdown on previously
unenforced immigration policies is working. We'll deal with the domestic
terrorists as needed, but we don't need to be letting the muzzie
terrorists get a foothold in our country too. One need only look at what
they're doing in Europe right now to know we're doing the right thing by
keeping them out, which is our right and our duty. - Checkmate (#1 pussy willow fan)

-

"You just made puppy whistle's sig line longer." - Janithor

-

"If I have a complaint about the (Southern Poverty) Law Center's description (of the alt-right movement), it is the phrase "heavy use of social media," which implies the alt-right is a real-world movement which uses a lot of social media. This is backwards: it is an online movement which occasionally appears in the real world. Where it gets punched." - Jason Rhode

-

"I think we should destroy every last fucking mosque in America." - "Checkmate, DoW #1" <***@The.Edge> proves for us that white males are violent in Message-ID: <***@news.altopia.com>

-

Golden Killfile, June 2005
KOTM, November 2006
Bob Allisat Memorial Hook, Line & Sinker, November 2006
Special Ops Cody Memorial Purple Heart, November 2006
Special Ops Cody Memorial Purple Heart, September 2007
Tony Sidaway Memorial "Drama Queen" Award, November 2006
Busted Urinal Award, April 2007
Order of the Holey Sockpuppet, September 2007
Barbara Woodhouse Memorial Dog Whistle, September 2006
Barbara Woodhouse Memorial Dog Whistle, April 2008
Tinfoil Sombrero, February 2007
AUK Mascot, September 2007
Putting the Awards Out of Order to Screw With the OCD Fuckheads, March 2016
Nomen Nescio
2019-01-29 12:11:18 UTC
Permalink
Post by Anonymous
Post by Anonymous
mix.cfg set to exclude, on grounds of latency or reliability, only
frell, frell2, lambton, slow.
Exactly the same results 4 days later than the O.P. haph, banana,
freireide chosen only a third as often as should be expected. No
significant difference between those three. ciambella, devurandom
chosen about 10% less often than should be expected, but this could
be down to some same machine exclusions for adjacent remailers.
- ciambella & haph
- banana & slow
- devurandom & fotonl1 & freierede
- frell & frell2
When in your test system QSL creates a starex configuration that
reflects these associations you get lower occurrence propabilities for
exactly these remailers.
lol! slap.
Nomen Nescio
2019-01-28 10:46:48 UTC
Permalink
In article
Post by Anonymous
Post by Nomen Nescio
Post by Anonymous Remailer (austria)
It is noted you have no explanation for the phenomenon, only
hot air bluster.
I don't see it as a phenomenon, bug or anything else. I see it
as a pointless whining exercise over your attempt to break
program code - and failing.
Now run along and go play in the traffic.
That the O.P. was confused about what QS did and what MM did is
irrelevant. The oddness remains.
The sender, not the agent, is responsible for any "oddness".

https://tools.ietf.org/html/draft-sassaman-mixmaster-02#section-
2.1

2.1 Message Creation
Anonymous
2019-01-28 15:23:10 UTC
Permalink
Post by Nomen Nescio
In article
Post by Anonymous
Post by Nomen Nescio
Post by Anonymous Remailer (austria)
It is noted you have no explanation for the phenomenon, only
hot air bluster.
I don't see it as a phenomenon, bug or anything else. I see it
as a pointless whining exercise over your attempt to break
program code - and failing.
Now run along and go play in the traffic.
That the O.P. was confused about what QS did and what MM did is
irrelevant. The oddness remains.
The sender, not the agent, is responsible for any "oddness".
https://tools.ietf.org/html/draft-sassaman-mixmaster-02#section-2.1
2.1 Message Creation
If I read LS's cryptic words as you have, then the O.P. was correct on
that point too, and the usual bunch of cretins who attacked her or him got
that wrong too.

APAS has gone to the dogs, they don't make anonymice like they used to.

No wonder most of us oldtimers don't post here much any more.

Introducing a panta-style hashcash system for posts in, and autocancel of
those without it, might help a little bit.
Nomen Nescio
2019-01-29 12:21:46 UTC
Permalink
In article
Post by Anonymous
Post by Nomen Nescio
In article
Post by Anonymous
Post by Nomen Nescio
Post by Anonymous Remailer (austria)
It is noted you have no explanation for the phenomenon, only
hot air bluster.
I don't see it as a phenomenon, bug or anything else. I see it
as a pointless whining exercise over your attempt to break
program code - and failing.
Now run along and go play in the traffic.
That the O.P. was confused about what QS did and what MM did is
irrelevant. The oddness remains.
The sender, not the agent, is responsible for any "oddness".
https://tools.ietf.org/html/draft-sassaman-mixmaster-02#section-2.1
2.1 Message Creation
If I read LS's cryptic words as you have, then the O.P. was correct on
that point too, and the usual bunch of cretins who attacked her or him got
that wrong too.
APAS has gone to the dogs, they don't make anonymice like they used to.
No wonder most of us oldtimers don't post here much any more.
Introducing a panta-style hashcash system for posts in, and autocancel of
those without it, might help a little bit.
You could also try using your head for something besides a knot
to keep your backbone from unraveling.

It should be trivial of you to repeat the experiment sans QSL
using only mixmaster, thus validating your claims.

https://www.quicksilvermail.net/remailers/mixmaster/

Hop to it.

https://www.quicksilvermail.net/remailers/mixmaster/manual.htm

-
Ybbx sbe n zrffntr, "Fhozvggrq hfvat Zvkznfgre fgnaqnybar, ab
pyvrag
fbsgjner."
Anonymous Remailer (austria)
2019-01-29 16:07:22 UTC
Permalink
Post by Nomen Nescio
It should be trivial of you to repeat the experiment sans QSL
using only mixmaster, thus validating your claims.
https://www.quicksilvermail.net/remailers/mixmaster/
Done, and I think one other did so too, using JBN2 as the other way to
call mixmaster.

Given the same stats, JBN2 shows no such preferences or dislikes in its
AUTO choices.

Go and see it for yourself. You want to risk compromising your anonymity,
great.

I'm manually inserting banana into all my chains until I complete a
thorough review of https://www.quicksilvermail.net/qslite/source/

I'd never touch Omnimix because it isn't open source.
Post by Nomen Nescio
Ybbx sbe n zrffntr, "Fhozvggrq hfvat Zvkznfgre fgnaqnybar, ab pyvrag fbsgjner."
Gur ercyl vf, V hfr WOA.
Nomen Nescio
2019-01-29 19:37:40 UTC
Permalink
Post by Anonymous Remailer (austria)
I'd never touch Omnimix because it isn't open source.
https://danner-net.de/om/OmniMix_IDE_2.5.0_Uno_Setup.exe
Anonymous Remailer (austria)
2019-01-27 14:05:50 UTC
Permalink
Nomen Nescio wrote...
Post by Nomen Nescio
Your ignorance is showing. Quicksilver has absolutely nothing
to do with it. It's all on Mixmaster.
Is there some part of what I wrote

"QS per se does not select the remailers, the MIXMASTER code invoked by QS
does this."

which you cannot understand?

QS's install package includes mixmaster code, which is invoked by it at
runtime.

It is noted you have no explanation for the phenomenon, only hot air bluster.

What knowledge is MIXMASTER using to make the selections?
Nomen Nescio
2019-01-27 18:33:52 UTC
Permalink
Post by Anonymous Remailer (austria)
Nomen Nescio wrote...
Post by Nomen Nescio
Your ignorance is showing. Quicksilver has absolutely nothing
to do with it. It's all on Mixmaster.
"QS per se does not select the remailers, the MIXMASTER code invoked by QS
does this."
which you cannot understand?
QS's install package includes mixmaster code, which is invoked by it at
runtime.
It is noted you have no explanation for the phenomenon, only hot air bluster.
What knowledge is MIXMASTER using to make the selections?
Whatever you place in your template, bozo.
QS-virgin
2019-01-27 20:04:43 UTC
Permalink
this is just another pile of bullbleep instigated by anti-QS jerks.
If they want Omnimix, or whatever, use it, and quit trying to defame
QS.
It's one of that AUK asshats. Same stalker who pollutes the
political groups.
Maybe maybe not but does it matter? Are the person or persons correct?

I understand its MixMaster, not QS, but I am worried too. JBN/XP uses
Mix but I looked with similar stats and it doesn't avoid haph/banana/
freierede statistically.

I am getting on well with QS Lite 1.1.4 except in one area it is
buggy,
"randomly" crashes after one deletes messages from outbox, and then
you lose all unsaved settings and work.

Anticipatory thank you for help but please not more trolling or insult
trading.
Anonymous Remailer (austria)
2019-01-26 23:38:25 UTC
Permalink
Post by Anonymous
Redo the test excluding haph, banana, and freierede and see what you get.
Did it for my own PoM. Used SEC3 stats, likely same as OP.

mixmaster history latency uptime
--------------------------------------------
banana *********** 22:30 100.00%
ciambella ************ 15:00 100.00%
devurandom ++++++++++++ 1:49:30 100.00%
fotonl1 ************ 29:00 100.00%
haph ************ 14:00 100.00%
inwtx ****#*# 5:30 100.00%
kreti ************ 28:00 100.00%
paranoia ************ 26:00 100.00%
redjohn *#********** 9:00 100.00%
roance ************ 24:30 100.00%
dizum ************ 16:00 99.78%
congeries *****+* 27:00 99.59%
austria ************ 23:00 99.45%
kroken ++++++**+++* 1:16:00 99.45%
freierede ++++++++++++ 1:41:00 98.89%
frell2 --------+-++ 5:45:00 99.13%
frell ---------+-+ 6:34:30 98.48%
eurovibes ++++++++++++ 2:02:00 98.02%

I set mix.cfg so the first 16 qualified, same as for the OP, and then manually
barred the 3 via qs.

Exit set to frell2. Chain = 9 random + frell2.

Results

roance 823
eurovibes 810
kroken 804
fotonl1 795
redjohn 794
dizum 768
congeries 766
devurandom 758
paranoia 754
austria 753
ciambella 746
inwtx 740
kreti 689
TOTAL 10000

10000/13 = 769

All in the range (-11%, +7%), ie no signficant difference. kreti is the only
outlier but only mildly.

I don't understand what happened with the OP's test (which I repeated and got
similar results to him).

Per SEC3, groups of remailers sharing a machine or operator:
(banana slow hsub)
(freierede devurandom)
(frell frell2)
(haph ciambella)

So what caused the "anomaly" re banana? Neither slow or hsub are in the
selection pool.

Why was freierede low, but devurandom not?

Why was haph low, but ciambella not?

And why were the low sooo low?
Nomen Nescio
2019-01-26 12:28:28 UTC
Permalink
Post by Nomen Nescio
re QUICKSILVER (latest version)
Waddafook is going on ??
I repeated the study with a fresh install of QUICKSILVER and
with shorter "random" chains and frell as exit.
The results were essentially the same - same three groups of
remailers.
Quicksilver doesn't generate the remailer selections you stupid
oblivious fuck.
Loading...