(fwd) Re: BSD random() - any good (source included)

> At  3:20 PM 07/07/94 -0500, Jim choate wrote 69K of forwards
> (so far) that I would much rather look at in sci.math or
> sci.stat.math than in my mailbox, unrequested. Didn't we have
> a little forwarding talk recently enough for your avarage
> televangelist viewer to remember it?
 
> Kindly cut it out.

I guess cypherpunks would rather not write or even read code, but
simply gossip about usenet kooks and pedophiles.

> ___________________________________________________________________
> Jamie Lawrence                                  <jamiel@sybase.com>
> 

c@lib@n


--------------------------------------------------------------------------
To find out more about the anon service, send mail to  help@vox.hacktic.nl
Please report any problems, inappropriate use etc. to admin@vox.hacktic.nl
Direct replies to the sender of this message are -not- anonymised..<YuK>..

On Sat, 8 Jul 1994 an65@vox.hacktic.nl wrote:

> > At  3:20 PM 07/07/94 -0500, Jim choate wrote 69K of forwards
> > (so far) that I would much rather look at in sci.math or
> > sci.stat.math than in my mailbox, unrequested. Didn't we have
> > a little forwarding talk recently enough for your avarage
> > televangelist viewer to remember it?
>  
> > Kindly cut it out.
> 
> I guess cypherpunks would rather not write or even read code, but
> simply gossip about usenet kooks and pedophiles.

	[1] Simply because you don't read new software announcements 
every day doesn't mean people aren't writing code. I am, for one, but 
people have other jobs and responsities to take care of.

	[2] Less than a week after a small flame war about forwarding, 
and a suggestion that it be restricted to posting pointers to relevant 
info, Jim forwards over 65K of relevant, though specialized information 
that no-one had requested. Simply saying "There's some real informative 
posts about RNGs in sci.math, thread name "XXX"" have been helpful, 
faster, saved bandwidth and people's mailboxes, and not started up this 
flame war.

--------------------------------------------------------------------------
Michael Brandt Handler                                <grendel@netaxs.com> 
Philadelphia, PA                                    <mh7p+@andrew.cmu.edu>
Currently at CMU, Pittsburgh, PA            PGP v2.6 public key on request
Boycott Canter & Siegel                <<NSA>> 1984: We're Behind Schedule

> 
> 	[2] Less than a week after a small flame war about forwarding, 
> and a suggestion that it be restricted to posting pointers to relevant 
> info, Jim forwards over 65K of relevant, though specialized information 
> that no-one had requested. Simply saying "There's some real informative 
> posts about RNGs in sci.math, thread name "XXX"" have been helpful, 
> faster, saved bandwidth and people's mailboxes, and not started up this 
> flame war.
>
The forwards were crypto related and relevant to some of the members who
don't have anything other than e-mail accounts. How do you propose these
folks get this info?

Also how about those folks who have a low latency system and the posts
in various medium to high traffic systems gets flushed regularly. These
posts were several days old and I suspect in many systems were ready to
flush (they were on mine, I 'tripped' over them doing maintenance for
something else) to the bit bucket.

And do you seriously propose that I or any other member ask prior to 
submissions? It is really funny that in general I get a few more thanks
for such posts than self-interested rebuttals like this.

While it is true that some of my questions are off the wall, I will 
continue to ask them. If they bother then .kill me. As an aside to this
I will continue to remail articles of technical interest (what c-punks
is about last I heard anything) that I feel have a good case of being
lost. 

I would also like to ask a question on a personal (no flame intended) 
nature. Were you going to post said message about these usenet submissions?
Other than myself I see very few such re-posts from anything other than
a newsgroup w/ 'crypt' in it somehow.

The flame ware, as I understand it anyway, had to do with forwarding 
multiple copies of EFF and similar material which is minimaly related
to cyrypto and most users actively look for it. I doubt a lot of the 
users here check out sci.math, sci.chaos, sci.neural-nets, etc. 

If we are really going to continue this thread then a serious discussion
relating to c-punks and some form of submission standard needs to be agreed
upon.

On Fri, 8 Jul 1994, Jim choate wrote:

> > 	[2] Less than a week after a small flame war about forwarding, 
> > and a suggestion that it be restricted to posting pointers to relevant 
> > info, Jim forwards over 65K of relevant, though specialized information 
> > that no-one had requested. Simply saying "There's some real informative 
> > posts about RNGs in sci.math, thread name "XXX"" have been helpful, 
> > faster, saved bandwidth and people's mailboxes, and not started up this 
> > flame war.
> >
> The forwards were crypto related and relevant to some of the members who
> don't have anything other than e-mail accounts. How do you propose these
> folks get this info?

From: Jim choate <ravage@bga.com>
Subject: Good RNG generator information available.

	Hey, cpunks! For those of you who are interested in random number 
generation, there's a great deal of discussion going on over in sci.math, 
thread name "XXX". If the articles have expired on your site, or if you 
don't have news access, mail me, and I'll send them out to everyone who 
wants one later on.

-- Just an idea... This is what I was doing with the  2600 Cellular 
Telephone article, except that my list of address get scragged along with 
my mail dir.

> While it is true that some of my questions are off the wall, I will 
> continue to ask them. If they bother then .kill me. As an aside to this
> I will continue to remail articles of technical interest (what c-punks
> is about last I heard anything) that I feel have a good case of being
> lost. 

	[1] Your questions (off-the-wall or not) don't bother me.
	[2] I very rarely killfile people.
 
> I would also like to ask a question on a personal (no flame intended) 
> nature. Were you going to post said message about these usenet submissions?
> Other than myself I see very few such re-posts from anything other than
> a newsgroup w/ 'crypt' in it somehow.

	No. I don't have the time or mathematical knowledge to read 
sci.math and determine whahould be forwarded to cypherpunks or not. 

> The flame ware, as I understand it anyway, had to do with forwarding 
> multiple copies of EFF and similar material which is minimaly related
> to cyrypto and most users actively look for it. I doubt a lot of the 
> users here check out sci.math, sci.chaos, sci.neural-nets, etc. 

	[1] THe EFF forads were VERY crypto related (HR 3627), we just 
got sick of seeing 4 copies of the same article pop up.
	[2] DO you know that no-one here reads those groups? I don't, but 
considering the number of intelligent and varied people here, that's a 
dangerous assumption.

> If we are really going to continue this thread then a serious discussion
> relating to c-punks and some form of submission standard needs to be agreed
> upon. 

	Agreed.
--------------------------------------------------------------------------
Michael Brandt Handler                                <grendel@netaxs.com> 
Philadelphia, PA                                    <mh7p+@andrew.cmu.edu>
Currently at CMU, Pittsburgh, PA            PGP v2.6 public key on request
Boycott Canter & Siegel                <<NSA>> 1984: We're Behind Schedule

> From: Jim choate <ravage@bga.com>

> The forwards were crypto related and relevant to some of the members who
> don't have anything other than e-mail accounts. How do you propose these
> folks get this info?

First, how were they crypto-related?  They dealt with statistical tests
for distinguishing good crypto-weak RNGs from bad ones.  Anything that
even hopes to be a strong RNG will pass this sort of tests.

Second, if people don't have Usenet, telnet, or ftp access (any of which
could have been used to retrieve these articles), they can get accounts
that do what they want.  If you want to tide them over in the interim,
you could post a pointer to sci.math, with the comment that you would be
happy to mail a copy to anyone who can't get the articles by other means.

   Eli   ebrandt@hmc.edu

> 
> First, how were they crypto-related?  They dealt with statistical tests
> for distinguishing good crypto-weak RNGs from bad ones.  Anything that
> even hopes to be a strong RNG will pass this sort of tests.
>
Even in your rebuttal you use 'crypto-weak' implying they are crpypto
related. I would appreciate a clarification on exactly what c-punks
means by 'crypto related'....

Seems perfectly clear to me, but obviously there is a major difference of
opinion here.

> Second, if people don't have Usenet, telnet, or ftp access (any of which
> could have been used to retrieve these articles), they can get accounts
> that do what they want.  If you want to tide them over in the interim,
> you could post a pointer to sci.math, with the comment that you would be
> happy to mail a copy to anyone who can't get the articles by other means.
>
Yes, I could do that if I were so inclined. I am not. Seems to me that just
forwarding it and letting that be that is a much better way than filling up
the mailing list with bitches about what you think somebody else should do.
Bottem line is that at this point there is no clear cut 'standard' that I 
have seen agreed on. Hell, I haven't even seen any discussion over it beyond
a bunch of polemics over what people like and don't like, which are not 
in any way necessarily related to what people need or want. One of the 
biggest problems w/ the list at this point is the range of expertise that
is present. With such a broad experience base almost any post is going to
piss somebody off. The real question is whether it is worth the time to 
reply to them (in general I hold it isn't) rather than just hitting 'd'
(which I recommend if you don't like it) and forgetting (even forgoing the
obviously popular habit of acidic replies). 

If you look at this rationaly all the discussion about pedophilia, eff,
nsa policy, etc. is really not directly related to crypto and coding. If
the main theme of this list is really 'c-punks write code' then my 
forwarding of the rng discussion (w/ code) is more appropriate to the topics
at hand than any of this other stuff. A very good analogy of how this type
of logic inversion is prevelant is the move to ban guns (no, I do NOT
want any replies to this particular topic, it is for example only!!!!)
by looking at the 2nd Amendment. If taken logicaly the reason for the 
amendment is to allow individuals to protect themselves against all
threats including the government. If followed through then any move to 
ban military weapons is un-constitutional and moves to ban non-military 
weapons are constitution - obviously not the tact taken.

The bottem line is it was crypto related, was in reference to source code,
and therefore fit the charter of this group.


>    Eli   ebrandt@hmc.edu
> 
>

On Fri, 8 Jul 1994 15:05:22 -0500 (CDT)  Jim choate wrote:
--------

> Bottem line is that at this point there is no clear cut 'standard' that I 
> have seen agreed on. Hell, I haven't even seen any discussion over it beyond
> a bunch of polemics over what people like and don't like, which are not 
> in any way necessarily related to what people need or want.

The reason for this is that people can usually get a hint; there's been no
need for a formal standard, because when someone does something irritating
several people shout, and the person has (with a couple of wild exceptions)
quit the irritating behavior.

You don't seem as far around the bend as either of the wild exceptions that
come to mind (Detweiler and that weird fellow that raved for about two weeks
two months ago) so I'm guessing you wouldn't want to be lumped in with them.
But by continuing to defend what obviously irritated a half dozen people,
with plenty more like me agreeing but unwilling to join in the fray to 
compound the problem, that's what you're courting.

> 
> The reason for this is that people can usually get a hint; there's been no
> need for a formal standard, because when someone does something irritating
> several people shout, and the person has (with a couple of wild exceptions)
> quit the irritating behavior.
> 
> You don't seem as far around the bend as either of the wild exceptions that
> come to mind (Detweiler and that weird fellow that raved for about two weeks
> two months ago) so I'm guessing you wouldn't want to be lumped in with them.
> But by continuing to defend what obviously irritated a half dozen people,
> with plenty more like me agreeing but unwilling to join in the fray to 
> compound the problem, that's what you're courting.
> 
Why should a couple of dozen people who scream because they don't get 
exactly what they want dictate to a list that is 700 long?

Lump me in where you feel justified. That classification and a dollar will
by you a cheap cup of coffee.                

If there was a vociferous return on this I would be much more responsive, but
what I see is the same small group of a dozen or so bitching and bitching and
bitching and bitching....(you get the point)

What is really funny is that they bitch about crypto stuff but seem completely
happy to blather on endlessly about pedophiles, off-shore banking, and a whole
host of other topics that are completely and uterly unrelated to "c-punks
write code".

So far I have recieved no reasoned responces to these issue, a lot of opinion
and emotionaly driven wailing has shown up in my box...

C'punks,

On Fri, 8 Jul 1994, Jim choate wrote:

> . . .
> What is really funny is that they bitch about crypto stuff but seem completely
> happy to blather on endlessly about pedophiles, off-shore banking, and a whole
> host of other topics that are completely and uterly unrelated to "c-punks
> write code".

Not a flame here, but a slight difference of opinion.  The admonition,
"Cypherpunks write code," should be taken metaphorically.  I think "to
write code" means to take unilateral effective action as an individual. 
That may mean writing actual code, but it could also mean dumpster diving
at Mycrotronx and anonymously releasing the recovered information.  It
could also mean creating an offshore digital bank.  Don't get too literal
on us here.  What is important is that Cypherpunks take personal
responsibility for empowering themselves against threats to privacy. 


 S a n d y

Jim choate says:
> Why should a couple of dozen people who scream because they don't get 
> exactly what they want dictate to a list that is 700 long?

Maybe its the lack of people begging you to repost dozens of articles
on a list with several hundred members that should send you a signal.

Personally, I find the reposts irritating but not devistating.
However, I suspect that when you repost things, you are doing so with
the hope that it will be of use to people. If no one finds it of use,
you should consider whether or not you are actually doing anyone a
service.

Perry

> Even in your rebuttal you use 'crypto-weak' implying they are crpypto
> related. I would appreciate a clarification on exactly what c-punks
> means by 'crypto related'....

"Crypto-weak" as opposed to "crypto-strong", cryptographically strong.
The guy was testing BSD random(), which I think is Yet Another Linear
Congruential Generator.  If you think an LC PRNG has cryptographic
relevance, you are gravely mistaken.

What's so hard about "crypto-related"?  If you can't think of any
relationship between the articles and cryptography (technical,
social, political, whatever), don't forward them.  If they had been
about cryptoanalysis of random(), that might be relevant, though
hardly ground-breaking.

> > you could post a pointer to sci.math, with the comment that you would be
> > happy to mail a copy to anyone who can't get the articles by other means.
>
> Yes, I could do that if I were so inclined. I am not.

So I see.  You're not willing to take it upon youself to mail copies
to people who can't otherwise get them, but you're happy to inflict
irrelevant material straight out of Knuth on people who could easily
get it themselves.  What *is* your rationale here?

> The bottem line is it was crypto related, was in reference to source code,
> and therefore fit the charter of this group.

Source code, yes.  Would you like to explain its relationship to crypto?

   Eli   ebrandt@hmc.edu

> 
> "Crypto-weak" as opposed to "crypto-strong", cryptographically strong.
> The guy was testing BSD random(), which I think is Yet Another Linear
> Congruential Generator.  If you think an LC PRNG has cryptographic
> relevance, you are gravely mistaken.
>
For top of the line crypto, you are absolutely correct. For a learning
resource, I disagree.

> What's so hard about "crypto-related"?  If you can't think of any
> relationship between the articles and cryptography (technical,
> social, political, whatever), don't forward them.  If they had been
> about cryptoanalysis of random(), that might be relevant, though
> hardly ground-breaking.
>
The concept of testing a RNG's stabilty and operating characteristics
is something which is directly crypto related. Much more so than pedophilia
or alt.whatever.rainbow.......

> > Yes, I could do that if I were so inclined. I am not.
> 
> So I see.  You're not willing to take it upon youself to mail copies
> to people who can't otherwise get them, but you're happy to inflict
> irrelevant material straight out of Knuth on people who could easily
> get it themselves.  What *is* your rationale here?
> 
To help those who don't have access to this material other than through
e-mail. And no I am not going to create a list of people and manualy
setup some form of sub-list to filter this stuff according to each 
individuals interest. I don't have the interest, time, or the information
on each and every  member of c-punks to carry this out. 
> > The bottem line is it was crypto related, was in reference to source code,
> > and therefore fit the charter of this group.
> 
> Source code, yes.  Would you like to explain its relationship to crypto?
> 
RNG's are commenly used in crypto, to understand the more compicated techniques
one must learn the more mundane basics.

My main rationale is that while I see lots of people making suggestions I dont
see any of them actually carrying it out. In the whole time I have been on 
this list I have *never* seen referals to the newsgroups. I have seen a 
general trend to repost this material so that people simply have it right 
then and there instead of having to send a sub-group  roaming around looking
for it. Seems a much more labor saving system.

>    Eli   ebrandt@hmc.edu
> 
>

This list is a community. I've avoided commenting on Jim Choate's
flames and "never let go of an argument" style, but he is increasing
his denunciation of the list, so I will comment.

Jim Choate writes:


> My main rationale is that while I see lots of people making suggestions I dont
> see any of them actually carrying it out. In the whole time I have been on 
> this list I have *never* seen referals to the newsgroups. I have seen a 
> general trend to repost this material so that people simply have it right 
> then and there instead of having to send a sub-group  roaming around looking
> for it. Seems a much more labor saving system.

If you have "*never*" seen referrals to the newsgroups, then you must
be not reading much of what gets posted here. I, for one, have many
dozens of time (maybe hundreds of times, since 1992) referred to
articles in sci.crypt, talk.politics.crypto, alt.security.pgp, etc.

(Including some that I wrote for those groups.)

Anyone who claims that the newsgroups never get discussed, and then
decides that all 500 subscribers simply must see a bunch of articles
on random number generators--a topic we have discussed a dozen
times--is clearly grinding an axe. Lashing out at the list as being
full of good-for-nothings simply because of complaints about these
articles is absurd.

As for the first point, that many suggestions are made but then not
carried out, this is the nature of all discussion groups I've ever
seen. After all, we're not being *paid* to do all this. We're not
organized into teams, and so on.

And despite this, impressive progress has been made:

* Remailers. Cypherpunks remailers with new features, more sites. This
is clearly the cutting edge of remailers, more so even than Julf's
site. (Cyphepunks remailers are distributed, instantiable by almost
anyone, are adding new features, etc. Julf's site remains singular,
and has not added major features in a long time.)

* Several list members are central to the development of PGP. 

* SecureDrive, CurveEncrypt, and other crypto programs are tied to
various list members.

* Experiments with digital money have been underway...the lack of
concrete progress owes more to general problems with such things than
on lack of effort. (Pr0duct Cypher, Matt Thomlinson, others.)

* Although we can't always claim everyone as a member, such people as
Whit Diffie, Phil Zimmermann, Bruce Schneier, Matt Blaze, Phiber
Optik, and Jim Bidzos have attended our meetings. Some of them are
list subscribers, etc. The 500+ subscribers included some of the
best-known cryptologists outside the NSA. (And maybe inside?)

* The role of Cypherpunks has been manifold: practical work on
remailers, tools, digital money....education and
discussion....analysis of new protocols, etc. (For example, at today's
meeting the focus is on "swIPe," an important new system written by
John Ionannaddis (sp?), Phil Karn, etc.--I hope I got the credit right.)


I could go on, but I won't. Given that I can't recall Jim Choate being
involved in any of these projects, or giving us insightful analyses of
trends, developments, and technical details, I don't think he's in a
position to condemn the rest of the list.

People who lash out at the list, calling the list a place for people
who never do anything, are revealing their own failures of
imagination.

I can't see why they choose to remain on the list if they despise it
that much.


--Tim May


-- 
..........................................................................
Timothy C. May         | Crypto Anarchy: encryption, digital money,  
tcmay@netcom.com       | anonymous networks, digital pseudonyms, zero
408-688-5409           | knowledge, reputations, information markets, 
W.A.S.T.E.: Aptos, CA  | black markets, collapse of governments.
Higher Power: 2^859433 | Public Key: PGP and MailSafe available.
"National borders are just speed bumps on the information superhighway."

> be not reading much of what gets posted here. I, for one, have many
> dozens of time (maybe hundreds of times, since 1992) referred to
> articles in sci.crypt, talk.politics.crypto, alt.security.pgp, etc.
>
Go back in your archives and notice that this responce reitterates one
I had made earlier about not seeing such references unless they  came
from a newsgroup with crypt  in it somehow. 

> times--is clearly grinding an axe. Lashing out at the list as being
> full of good-for-nothings simply because of complaints about these
> articles is absurd.
>
I didn't lash out at anyone. I posted a set of materials I thought some
might find interesting. If folks like you had left it alone that would
have been the end of it. I would not have made any further communications
on it. However, a certain clique of c-punks seem compelled to reply to
every damn post that gets sent on there, and do it ad nauseum. If the 
traffic is too high try not replying unless it is a positive contribution.
But, because you folks apparently have nothing  better to do  we have 
managed to generate a set of list traffic that vastly exceeds the original
forwards. 

> As for the first point, that many suggestions are made but then not
> carried out, this is the nature of all discussion groups I've ever
> seen. After all, we're not being *paid* to do all this. We're not
> organized into teams, and so on.
>
Reminds me of the years I worked in a science museum and people were always
coming around with new projects for me to  do. My general responce became   to
point them to the shop and tell them to have at. I have little respect for
people who have nothing better to do than tell others how to spend their time.
If the suggestion is that great, carry it out yourself.

> involved in any of these projects, or giving us insightful analyses of
> trends, developments, and technical details, I don't think he's in a
> position to condemn the rest of the list.
>
I didn't condem the list or anyone else for that matter. 

> People who lash out at the list, calling the list a place for people
> who never do anything, are revealing their own failures of
> imagination.
>
I *NEVER* said that or anything like it. Geesh, speaking of imagination.

> I can't see why they choose to remain on the list if they despise it
> that much.
>
see the line above.
> 
> --Tim May
> 
> -- 
Tim,   

you really should  quite drinking or whatever, you are seeing things.

Jim, I agree with you.  The complaints are from those who are too lazy t
simply not read a message, who feel the urge to read anything and everything,
and seeing something they've seen elsewhere feel cheated somehow.

Yes, there are several arguements:

1. "It wastes bandwidth"

A: If I didn't think that someone on this list would enjoy it, find it of
   use or interest, I wouldn't have posted it.  While I realize that not
   everyone is like me, I also realize that it is likely that others share
   the same interests as I do, or else we wouldn't be on this list.

2. "I've seen it before"
A: Great, then you don't have to read it.

3. "You should just put pointers up to the original article"
A: The original article may be on usenet.  Not everyone has access to usenet
   though they have access to internet mail.  Not everyone who has access to
   usenet has access to the particular newsgroup the message/article was posted
   in.  Not everyone who has access to the particular newsgroup will get to
   see the particular article because some systems will have a short delete
   time and by the time they find out about the article, it will be gone.
   At this point, if they're only mildly interested, they'll give up, if they're
   really interested, they will ask for the message to be sent to them out
   of someones benevolance.  Not everyone will have copies of it, not everyone
   will bother to send the message.

   Forwarding messages to the list has the advantage that anyone who might be
   interested will be able to receive it.

4. "I dislike forwarded messages so much I'm going to bitch about them until
    you stop posting them."
A:  You're an asshole with nothing better on your hands.  If you dislike wasting
    time and bandwidth why do you post complaints about it?  Why do you waste
    your time and the time of those who want to see the article, or the time
    of those who do not wish to see your rantings?  Experience tells me that
    the bitching replies to an "offending" message usually will last for a
    long time, generating far more bandwidth waste, not contain anything
    remotely on topic, and annoy the fuck out of the whole list.  So just
    because you are annoyed and like to bitch doesn't mean you should force
    the whole list to listen to your rantings.

    Besides, even if you do botch, 99.99999% of the readers will simply ignore
    your rants, or delete them on sight without reading more than the subject.

I like the idea of having a cypherpunks-fwd subgroup.  This way the assholes
on this list who only complain won't have to subscribe to it.. I will tolerate
off topic messages and I don't have a problem with hitting the delete key.
I will tolerate seeing things I've seen before and won't bitch about them
either.  But when someone calls my messages off topic be they mine or be they
forwards, all I need to is look at all the messages on the list to find the
same ones that I'd ignore and stick the nose of the accuser in them.  This
will usually shut them up quickly and silence their ridiculous rantings.

Post and let post.  If you don't like it, don't read it.  Thems by two
electrons. :-D

> From: Jim choate <ravage@bga.com>
> The concept of testing a RNG's stabilty and operating characteristics
> is something which is directly crypto related.

This discussion is going nowhere, so I'll drop it.  I suggest that those
individuals who have only e-mail access use it to tell Jim how much they
appreciate this sort of forward.  If you get less than 12 or 13 positive
responses, perhaps you could give matters a little thought.

   Eli   ebrandt@hmc.edu