Send Devl mailing list submissions to
devl at freenetproject.org
To subscribe or unsubscribe via the World Wide Web, visit
http://www.uprizer.com/mailman/listinfo/devl
or, via email, send a message with subject or body 'help' to
devl-request at freenetproject.org
You can reach the person managing the list at
devl-admin at freenetproject.org
When replying, please edit your Subject line so it is more specific
than "Re: Contents of Devl digest..."
Today's Topics:
1. Re: Killing Freenet (Re: [freenet-devl] Aardvark) (Tavin Cole)
2. Re: List gone quiet (Gianni Johansson)
3. Re: Killing Freenet (Re: [freenet-devl] Aardvark) (Tavin Cole)
4. Re: Aardvark (Scott G. Miller)
5. Re: Problems for splitting (Timm Murray)
6. Re: Aardvark (Chris Anderson)
7. Re: RE: [freenet-devl] Aardvark (Timm Murray)
--__--__--
Message: 1
Date: Sat, 3 Feb 2001 13:10:34 -0500
From: Tavin Cole <[email protected]>
To: devl at freenetproject.org
Subject: Re: Killing Freenet (Re: [freenet-devl] Aardvark)
Reply-To: devl at freenetproject.org
On Sat, Feb 03, 2001 at 06:39:46PM +0100, Oskar Sandberg wrote:
> On Sat, Feb 03, 2001 at 10:55:58AM -0500, Peter Todd wrote:
> < >
> > > Well, Setup.java says:
> > >
> > > public static void setParamConnectTimeout() {
> > > String id = "connectTimeout";
> > > expComment("How long to wait to connect to a host before giving up
> > > (in milliseconds)");
> > > long l = params.getlong(id,Core.defaultConnectTimeout);
> > > if (expert)
> > > l = getNumber("?",l);
> > > out.println(id+ "=" + l);
> > > }
> > >
> > > I'm assuming that the first time this is run the params object is empty...
> >
> > Most Linux users probably don't use Setup.java I know I didn't say
> > anything about it's use in my Install and Admin artical.
>
> We should make it so the node does not start unless Setup has been run. I
> intended to make the syntax as close to normal options when installing
> debian packages. I don't remember how right I got it.
Abolish sample.freenetrc and make the node refuse to start without .freenetrc
present.
--
// Tavin Cole
--__--__--
Message: 2
From: Gianni Johansson <[email protected]>
To: devl at freenetproject.org
Subject: Re: [freenet-devl] List gone quiet
Date: Sat, 3 Feb 2001 13:26:11 -0500
Reply-To: devl at freenetproject.org
On Saturday 03 February 2001 12:53, you wrote:
> On Sat, Feb 03, 2001 at 12:35:39PM -0500, Gianni Johansson wrote:
> > On Friday 02 February 2001 22:12, you wrote:
> > > > Hey, why don't we turn the filter on? I haven't heard a bug report
> > > > in
> > >
> > > ages.
> >
> > I second this.
> >
> > I haven't seen any bugs either. Even if there are some, we should get
> > the filter out into wider use (with the appropriate warnings and caveats)
> > so that we find them.
> >
> > If people have objections could they please provide specific examples of
> > known exploits. Not just generic complaints that "filtering is
> > inherently difficult and imperfect so we shouldn't have any filtering"
> > which is sort of equivalent to "I can easily open the door to your house
> > with 1kg of plastic explosives, so it's stupid for you to put a lock on
> > it."
>
> Why don't I just come up with an exploit (damn, I can't use the
> meta-refresh thing anymore, I already mentioned that) and quietly gather IP
> addresses awhile, then once I have a list of people who've accessed
> documents about, say, how to manufacture some kind of illegal drug, I'll
> just let the police announce the exploit on my behalf.
That's one way to go.
But why strain yourself.
Just leave the filtering turned off like it is right now and post the same
html page with an image from an http server on it. Suggestions:
http://www.riaa.com/graphics/Home_Top_Title.gif,
http://www.cia.gov/graphics/cianew1.jpg.
You are right that filtering is imperfect. But it sure is better that what's
in the current public release.
Please do come up with as many exploits as you can and publish them.
-- gj
--
Web page inside Freenet:
freenet:KSK at webpages/gj_jump0
--__--__--
Message: 3
Date: Sat, 3 Feb 2001 13:22:42 -0500
From: Tavin Cole <[email protected]>
To: devl at freenetproject.org
Subject: Re: Killing Freenet (Re: [freenet-devl] Aardvark)
Reply-To: devl at freenetproject.org
On Sat, Feb 03, 2001 at 06:21:44PM +0100, Sebastian Spaeth wrote:
> "Scott G. Miller" wrote:
> > > It would be interesting to see what the exception that causes the removal
> > > is. Are we failing to connect (maybe we need to increase the connect
> > > timeout, our arbitrary cutoff after 5 seconds when I believe the TCP says
> > > two minutes is somewhat shaky), are we failing to get responses back
> > > from nodes we send to, or is the authentication not working, or what?
> >
> > Like I've said many times before, its too damned harsh to remove a node
> > from the datastore for failing to connect once. Freenet should be able to
> > tolerate a node disappearing for an hour or two now and then. I really
> > think we should implement the decayed contact heuristic.
>
> I second that. I like agl's implementation which will ignore all
> references to a node for a certain time if not reachable, enlarging the
> ignore phase every time, until it is totally dropped in the end (if I
> understood his scheme correct).
I third that. We have to accept that a lot of people who might run Freenet
nodes will also be likely to shut their node down for the evening to run
an Unreal Tournament or Diablo server or some such thing.
> How about adding "backup references" when a node comes across an other
> data source reference for a key that it already knows. It could hold up
> to two or three references for a key, using the backup references only
> when the primary node is not reachable. This way it wouldn't hurt the
> routing system, I guess.
btw, guys, so far in my re-implementation of the datastore for 0.4, I am
making it *possible* to associate more than one node reference with a key.
Though, at this point, they are all equal -- there's not a primary +
backups.
Of course, I'm quite aware of the undesirability of changing how the
algorithms work without proper analysis & agreement, so if anyone feels
uncomfortable with this, I can excise it for now in favor of the old
one key => one node mappings.
However, I don't think it really changes how the algorithms work, it
just harmlessly increases the chance of successfully routing a key.
Comments?
--
// Tavin Cole
--__--__--
Message: 4
Date: Sat, 3 Feb 2001 14:05:59 -0500
To: devl at freenetproject.org
Subject: Re: [freenet-devl] Aardvark
From: "Scott G. Miller" <[email protected]>
Reply-To: devl at freenetproject.org
--xXmbgvnjoT4axfJE
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable
On Sat, Feb 03, 2001 at 01:10:48PM -0500, Chris Anderson wrote:
>=20
> On Sat, Feb 03, 2001, Tavin Cole wrote:
>=20
> > Since when do nodes get to know the plaintext of their KSKs, let
> > alone make decisions about how to treat them? That's called an
> > attack.
>=20
> Hey, it's my node... If I want to be stupid, nobody can stop me.
> Since a client and a node are not distinguishable, there is no reason
> a node can't know the plaintext of it's data. Encryption does
> degrade the speed of the local node's data by about 200X, one way to
> speed it up is to cache the plaintext or get rid of local encryption all
> together for transient nodes.
Okay, now you're just showing some obvious ignorance. The encryption does
not degrade the speed by 200x, doubtfully any more than 2x. You can't get
rid of encryption for transient nodes, since transient nodes exist in the
network as well and will be contacted. You *could* have a second listen
port for local-only addresses that didn't use encryption.
Scott
--xXmbgvnjoT4axfJE
Content-Type: application/pgp-signature
Content-Disposition: inline
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.4 (GNU/Linux)
Comment: For info see http://www.gnupg.org
iD8DBQE6fFaXr9IW4v3mHtQRArJCAJ427ccNmtjzTDqwcQvMHkqKtE5qMQCdEJXW
t5eoxCcrB1SqJU2mRO9kIUY=
=3tHX
-----END PGP SIGNATURE-----
--xXmbgvnjoT4axfJE--
--__--__--
Message: 5
From: "Timm Murray" <[email protected]>
To: <devl at freenetproject.org>
Subject: Re: [freenet-devl] Problems for splitting
Date: Sat, 3 Feb 2001 13:28:31 -0800
Reply-To: devl at freenetproject.org
I know it seems weird, but have you tried it? I have a laptop that you can
only put stuff on using floppy disks. I have tried installing many RPMs on
it, but the only ones that fail to load (with a segfault) are the ones I
split.
----- Original Message -----
From: "Mark J. Roberts" <[email protected]>
To: <devl at freenetproject.org>
Sent: Friday, February 02, 2001 4:41 PM
Subject: Re: [freenet-devl] Problems for splitting
> On Fri, 2 Feb 2001, Timm Murray wrote:
>
> > I've said this almost every time file splitting comes up, but
> > I'll say it again.
> >
> > Some types of data do not split well. For instance, RPMs (before
> > I found the light of Debian) segfault when I tried to use one
> > that I had split. I don't know why, but we should find out before
> > splitting becomes mandatory.
>
> Nonsense. Hash the recombined file and compare with that of the original.
>
>
> --
> Mark Roberts
> mjr at statesmean.com
>
>
> _______________________________________________
> Devl mailing list
> Devl at freenetproject.org
> http://www.uprizer.com/mailman/listinfo/devl
>
--__--__--
Message: 6
Date: Sat, 3 Feb 2001 14:36:41 -0500 (EST)
From: Chris Anderson <[email protected]>
To: devl at freenetproject.org
Subject: Re: [freenet-devl] Aardvark
Reply-To: devl at freenetproject.org
On Sat, 3 Feb 2001, Scott G. Miller wrote:
>
> Okay, now you're just showing some obvious ignorance. The
> encryption does not degrade the speed by 200x, doubtfully any more
> than 2x.
What are we measuring here? Local data access or network data
access? Am I allowed to remove the local node from the speed
calculation? Am I allowed to remove java's performance issues?
I'll get you some numbers if you describe the exact performance test.
--__--__--
Message: 7
Date: Sat, 3 Feb 2001 14:39:24 -0500
From: "Timm Murray" <[email protected]>
To: <devl at freenetproject.org>
Subject: Re: RE: [freenet-devl] Aardvark
Reply-To: devl at freenetproject.org
You just don't understand Oscar yet . . .
Stephen Tidey wrote on 2/1/01 1:56 am:
>Well that's a bit provocative! =20
>I haven't been in this group=20
>for long but I would of=20
>thought that you would have=20
>at least given an explanation=20
>why rather than just shout=20
>at everyone?
>
>-----Original Message-----
>From: Oskar Sandberg=20
>[mailto:md98-osa at nada.kth.
>se] Sent: 31 January 2001=20
>23:27
>To: devl at freenetproject.org
>Subject: Re: [freenet-devl]=20
>Aardvark
>
>
>
>For fucks sake people, YOU=20
>DON'T LINK TO KSKS! It's=20
>fucking nuts!
>
>On Wed, Jan 31, 2001 at=20
>05:02:03PM -0500, Benjamin=20
>Coates wrote:
>> >From Ian Clarke=20
><ian at octayne.com>
>> >I think that given that=20
>Aardvark has been down for=20
>the last few days, we
>> >should remove it from=20
>the FProxy gateway page=20
>for 0.3.7.
>> >
>> >Any suggestions for an=20
>alternative?
>> >
>> >Ian.
>>=20
>> I've been inserting a mirror=20
>of Steve's key index at=20
>KSK at KeyIndex.txt as a
>
>> trivial example of a=20
>date-based redirect.
>>=20
>> --
>> Benjamin Coates
>>=20
=0A=
Timm Murray=0A=
=0A=
-----------=0A=
Great spirits have allways encountered violent opposition from mediocre =
minds=0A=
--Albert Einstein
--__--__--
_______________________________________________
Devl mailing list
Devl at freenetproject.org
http://www.uprizer.com/mailman/listinfo/devl
End of Devl Digest
