On Mon, Apr 16, 2001 at 08:38:09PM +0100, Adam Langley wrote:
"is it auto-generated?"
It's a script - but it has to be run manually for the moment
Why? If it is a script can't it just be run from a cron job?
"Is your intention to make it automatically upload to sourceforge"
No - because of
On Mon, Apr 16, 2001 at 05:26:00PM -0700, Mr . Bad wrote:
"AL" == Adrian Lopez [EMAIL PROTECTED] writes:
AL It seems like various people here would rather do away with
AL KSKs. I'd like these people to address how data submission
AL might be dealt with in lieu of public
This is absolutely cool - well done, although unfortunately I couldn't read
anyone else's messages - perhaps there weren't any...
Ian.
On Tue, Apr 17, 2001 at 12:43:45PM -0400, Gianni Johansson wrote:
Check this out:
On Thu, Apr 19, 2001 at 03:00:19PM -0500, Brandon wrote:
I have a couple of bugs I'd like to fix. I can probably get to that
tonight.
I am going to NY tomorrow morning, but perhaps Mr Bad could do a release
this weekend?
Ian.
PGP signature
Since many applications of Freenet take quite a bit of time (adding to a
key-index, sending messages using think-cash etc). Fortunately most of
these are not time-critical, however the problem is that current clients
to do this tend to block, so you have to leave the client running for
quite a
On Fri, Apr 20, 2001 at 11:24:33AM +1200, David McNab wrote:
I've had success with embedding the JRE into a client installer.
Sun are very clear about their requirements for JRE redistro - required file
set etc.
I'm confident I can make a fully-compliant installer with JRE built in, one
On Tue, Apr 24, 2001 at 08:42:18PM -0400, Michael Carmack wrote:
I'm personally developing a couple of projects around the Freenet
base. It's not so attractive when my imports look like:
import java.io.*;
import java.util.*;
import Freenet.contrib.*;
Why does the concept of using
On Thu, Apr 26, 2001 at 12:28:52AM +0100, Dave Hooper wrote:
My sourceforge username is stripwax (no laughing at the back...)
Thanks...
You are added.
Ian.
PGP signature
Rather than waste energy trying to update the website, I have just
created a new module called www and moved the translation-capable
website into it. I have updated the website scripts to account for this
change.
Ian.
PGP signature
Great! Firstly, you should join the [EMAIL PROTECTED] mailing list.
Secondly, grab a sourceforge username and send it to me. Thirdly, read the
docs on sourceforge about using CVS. Fourthly, take a look at the existing web
structure to see how it works (it is pretty simple, stuff goes in
This seems to be a problem related to the new auto-translating website,
although I did a rebuild and it is working again.
If it happens again, I will investigate more deeply.
Ian.
On Thu, Apr 26, 2001 at 08:22:54PM +0200, Jan-Thomas Czornack wrote:
Yes, for several hours already.
Jantho
On Fri, Apr 27, 2001 at 02:05:18AM -0700, Aaron Voisine wrote:
looking at the code, the simplest way to implement zlib compression is to
change line 39 of Freenet/client/Document.java to read:
this.data = new GZIPInputStream(source);
(you'd also have to add import java.util.zip.* near the
FCP* seems to be separating out into three layers, these are:
[* By FCP I also include the XML-RPC interface]
#1 Basic
Standard insert/request along with pk generation stuff
#2 Metadata
Intercepts metadata documents and acts accordingly
#3 Stacks
Supports stacks, these are a generic name for
On Fri, Apr 27, 2001 at 11:38:27AM -0700, Mr . Bad wrote:
2) Change InsertClient so that it will do the equivalent of
freenetmirror or PutFiles if the local file is a directory.
I actually think that we should eat our own dogfood, and replace the
FreenetInsert and
I have created a new mailing list for discussion of the next major
release, namely 0.4. The list is called piesky and you can subscribe
at:
http://lists.freenetproject.org/mailman/listinfo/piesky
The intention of this list is to allow those who are focussed on 0.4 to
work without the noise now
On Fri, Apr 27, 2001 at 01:39:09PM -0700, Mr . Bad wrote:
IC == Ian Clarke [EMAIL PROTECTED] writes:
IC The intention of this list is to allow those who are focussed
IC on 0.4 to work without the noise now found on devl (most of
IC which is actually not noise, but only relevant
On Sun, Apr 29, 2001 at 12:23:45AM -0500, Brandon wrote:
How about 0.3.10 having an FCP enhancement for bulk inserts, suitable for a
whole site, similar to Freenetmirror.
I can have something like this into the XML-RPC API by 0.3.10. Details
need to be worked out.
...and it's FCP at
On Sat, Apr 28, 2001 at 01:40:25AM -0400, Travis Bemann wrote:
#3 Stacks
Supports stacks, these are a generic name for documents which are
updated by appending a higher number each-time. Stacks may be public or
private, private stacks are SSKs and can only be added to by the SSK's
Here is a document I just finished - comments are welcome:
Freenet Client Protocol Layer #3 : Stacks
Draft
Introduction
This document is an informal description of the third layer of the
Freenet Client Protocol (FCP). It allows higher level operations
involving Freenet inserts
We could really use a Freesite HOWTO for the How to publish link on the
website, explaining how to create a Freesite using FreenetMirror or one of
the other tools.
Any takers?
Ian.
PGP signature
On Tue, May 01, 2001 at 11:42:16PM -0500, Brandon wrote:
No, you've got it all wrong. The solution to the puzzle gives you an
insertion address, and the email recipient only checks valid
addresses. So you dont have to filter any spam at all, since anything put
in those valid spots
In an effort to encourage use use of the SF bugtracker, I have added
links from the Freenet homepage.
Could people please bookmark this page:
http://sourceforge.net/tracker/?func=browsegroup_id=978atid=100978
And check it every so-often. Some bugs will be auto-assigned, but most
won't.
Ian.
On Fri, May 04, 2001 at 11:40:26AM +1200, David McNab wrote:
Ian's questions about FreeWeb are making me think about a more general
Freenet problem, which I toss out now to dev for discussion.
The problem is:
To design a translation path which will reliably render a short human
readable
On Thu, May 03, 2001 at 11:12:51PM -0400, Tavin Cole wrote:
As has been discussed, the servlet API in its 0.3 form had to be at least
temporarily discarded while developing 0.4. Before we bring back this
hulking, klunky beast, let's consider using a much simpler plug-in model.
I think that
I have also been thinking about how logging could be enhanced further.
One idea I had was to have a dump log facility, where the last 10
messages of each debugging level are outputted, and could form part of a
bug report. This would ease the requirement that someone needs to try
to recreate a
I would like to tidy up CVS.
The first thing to do is to cull unused CVS modules from CVS (via a
request to Sourceforge support).
The modules I think should be culled are as follows:
website : Now we use www
freenet : Someone checked in freenet instead of Freenet by
I haven't heard anything about the GCJ stuff that MJR was working on,
what happened to it (it was cool)?
Ian.
PGP signature
Someone told me that there would be no metadata permitted in 0.4 CHKs.
If this is true then it needs to be rethought - I can imagine no reason
why client-side metadata would not be permitted. For example, it would
prevent web browsers from displaying documents referenced by CHKs.
Is this true?
On Sat, May 05, 2001 at 05:29:31PM -0500, Brandon wrote:
I think it would better to disallow DMI in CHKs altogether. That would
make everything must simpler. You can put the DMI for small CHKs in a
redirect.
I disagree, this means that a redirect would be required just to insert
a small image
On Fri, May 04, 2001 at 04:04:26PM -0700, Aaron Voisine wrote:
Didn't someone suggest a while back putting an entire web page or web site
including images into a single tar file? The file could then be gzipped
solving the compression, prefetching, and uneven dropout issues.
Yeah, and then
On Thu, May 03, 2001 at 09:56:43AM -0400, Timm Murray wrote:
I'm not about to do any scientific servay of it, but I know
a lot of data could benifit (web pages, other text, all Freenet metadata, etc.).
OTOH, a lot of data would not benifit. So, I say make it a CLI option
to compress before
Anyone planning on going to DefCon this year? I am just wondering
whether Freenet will be represented there
Ian.
PGP signature
On Sun, May 06, 2001 at 10:36:22AM +0100, Adam Langley wrote:
On Sat, May 05, 2001 at 05:05:19PM -0700, Ian Clarke wrote:
Metadata for CHKs can be stored in a separate document.
As mentioned, this would double the time required to retrieve these
documents, and since I anticipate
The FCP doc is somewhat unclear about what happens with metadata (ie. is
it prepended to the data? what format does it take?).
Perhaps an example should be given with MIME-type metadata..
Ian.
PGP signature
On Sun, May 06, 2001 at 05:11:32PM -0500, Scott Gregory Miller wrote:
I disagree, this means that a redirect would be required just to insert
a small image into Freenet if we wanted to associated a MIME-type with
it. That would (on average) double the time required to retrieve the
image.
On Mon, May 07, 2001 at 11:12:32AM +1200, David McNab wrote:
From: Ian Clarke [EMAIL PROTECTED]
The FCP doc is somewhat unclear about what happens with metadata (ie. is
it prepended to the data? what format does it take?).
Perhaps an example should be given with MIME-type metadata..
Hmm
So right now things are a-little messy with client apps. Some talk to
the node using Freenet's encrypted protocol, some use FCP, some
are plugins for the node, and others use XML-RPC.
I think that some standardization is in-order. The Whiterose approach
is that the node itself speaks two
On Sun, May 06, 2001 at 10:05:46PM -0400, Tavin Cole wrote:
Read the list archives.
Don't you think I have anything better to do ;-)
Ok, I'll do anything for our favorite rock star..
oh dear, this is going to stick isn't it.
Yes, Ian, there's already a simple plug-in API in 0.4 that allows
On Mon, May 07, 2001 at 11:07:51PM -0500, Brandon wrote:
There is no expense in requesting an additional file if the files are
requested simultaneously, at least no expense worth mentioning.
There is if one of them has been dropped - and how will you know what
metadata file to request before
On Tue, May 08, 2001 at 11:40:14AM -0500, Brandon wrote:
Really, the only reason not to put the embed the metadata key is that it
will make CHKs longer. I don't see this as being a serious problem, or
even noticable at all.
But by that logic, why select metadata as a special case - it is
you should announce these things to the web mailing list so that the
translators can do their thing.
Ian.
On Tue, May 08, 2001 at 09:38:01PM +0200, Sebastian Sp?th wrote:
OK, here is the deal:
http://download.sourceforge.net/freenet/Freenet_setup0.3.9.1-1.exe is
out (sorry for introducing
On Tue, May 08, 2001 at 01:15:44PM -0500, Brandon wrote:
Appending the metadata key to the CHK strikes me as an inelegant
solution to a non-problem.
It has already been generally agreed that there are two problems which
have to be solved. Only one solution to these two problems has been
On Sat, May 12, 2001 at 06:44:11PM -0500, Mathew Ryden wrote:
Actually it was a deficiency in the FCP specs. GenerateSVKKeypair returns a
keypair without the last three bytes -- the keysize and keytype. This was
not accidental, although the docs are lacking in what exactly is returned.
So the
On Sat, May 12, 2001 at 09:10:40PM -0500, Mathew Ryden wrote:
The problem is that 0.3 also has the size (i guess). That means that a call
to GenerateSVKKeypair would have to take in a file so it would know the size
and encode that in correctly. It's impossible to do it in this method
cleanly,
On Sun, May 13, 2001 at 10:07:58AM -0700, Ian Clarke wrote:
I suspect that the size in question is the size of the key rather than
the size of the data.
...or maybe not...
Ian.
PGP signature
A DC or Dining Cryptographers ring is a way that a number of nodes on
a network can form a ring that works like ethernet, where anyone on the
ring can broadcast information to the others but without betraying their
identity. This can work over TCP/IP or any other communication
protocol. Scott
, Ian Clarke wrote:
A DC or Dining Cryptographers ring is a way that a number of nodes on
a network can form a ring that works like ethernet, where anyone on the
ring can broadcast information to the others but without betraying their
identity. This can work over TCP/IP or any other
On Mon, May 14, 2001 at 01:11:25PM -0700, Ian Clarke wrote:
transmission), this dramatically reduces the problem, in fact,
if the quiet data rates are sufficiently low, then they are
insignificant and you only get a 200% increase in data-flow during
busy periods
Actually, less than
On Tue, May 15, 2001 at 08:29:21PM -0400, Mark J. Roberts wrote:
On Tue, 15 May 2001, Ian Clarke wrote:
For those who think that layer #2 should be a front-end to layer #1, all
you are doing is adding unnescessary bloat. Any smart implementation of
layer #2 will interface directly
On Tue, May 15, 2001 at 07:10:04PM -0500, [EMAIL PROTECTED] wrote:
You people are yammering about layers.
'
WHERE THE FUCK IS THE DEFINITION OF THESE LAYERS!!!
Layer #1 - FCP as it is currently defined
Layer #2 - Metadata handling built into FCP
Layer #3 - Stacks
Ian.
PGP signature
On Tue, May 15, 2001 at 04:49:06PM -0700, Ian Clarke wrote:
In an attempt to integrate Whiterose into CVS,
s/cvs/the website
Ian.
PGP signature
On Tue, May 15, 2001 at 09:21:56PM -0500, Brandon wrote:
If you can't come to an agreement on the philosophy of FPC, you might
consider switching to XML-RPC, which has a different design philosophy
closer to the one which you're espousing. I think there's room for
multiple design philosophies
On Wed, May 16, 2001 at 08:59:11PM +0100, Adam Langley wrote:
I really don't think that the FCP parsing times are too great -
certainly not compared to the network lag times. If people want
stunning performance they can use FCP or FNP. The cost of maintaining
all the different interfaces (in
Theo mentioned Need To Know on p241 of O'Reilly's P2P book. This is a
joke implying that this is their claim to fame.
Sheesh.
Ian.
On Fri, May 18, 2001 at 03:43:50PM +0200, Oskar Sandberg wrote:
WTF does that mean?
On Thu, May 17, 2001 at 04:44:05PM -0700, Ian Clarke wrote:
http
Through a rather Heath Robinson combination of procmail, python, ssh,
and PHP, I have updated the donations page at:
http://freenetproject.org/index.php?page=donations
To show the current PayPal account balance, and also list recent donors
to the Freenet project (giving them the option of
Wow, this list has got pretty quite the last few days, is this because
it is exam season or something?
Ian.
PGP signature
On Sun, May 20, 2001 at 11:11:46PM -0700, John George wrote:
I have been working this weekend on python code for an XML-RPC client.
I downloaded the source tarball and used it. I am sure there is a
large amount of deltas from the source tarball so where should I look
for what is being
Taking the ethernet approach to communication
Ian.
On Wed, May 23, 2001 at 11:35:12PM +0100, Peter Barr wrote:
Hi Ian, I wonder if you remember me. My name is Peter Barr and I used
to go out with Rachel Gibson who lived in your block in first year (she
says hi by the way). I saw you
On Fri, May 25, 2001 at 05:56:32AM -0400, Tavin Cole wrote:
fred 0.4 seems to be able to function sufficiently well for FCP now.
Groovy - I can set up a inform0.4.php script for testing of 0.4 (or we
could just implement the announcement stuff).
Ian.
PGP signature
On Sat, May 26, 2001 at 12:16:56AM +1000, Emil Mikulic wrote:
We may want to provide a fast C store generator with the installers.
You may want to provide a fast (and low memory footprint) node
with the intallers, too.
Whiterose is on the way ;-)
But seriously, 0.4 has a number of
On Fri, May 25, 2001 at 10:32:19AM -0700, Mr. Bad wrote:
I'm going to start beta-testing 0.4 this wknd.
How confident are oh-point-four-ers that it's getting to be time to
merge 0.4 back from experimental into the main trunk?
Hmmm, we could do this, and move further improvements to 0.3 to
I have been thinking about a piece of Java code which, when run, would
do some test inserts and requests via FProxy, FCP, and XML-RPC. This
could be run prior to a release, or even executed automatically as part
of a daily build process, and would smoke test a freenet node. The
same code should
From the sourceforge website:
(2001-05-24) PLEASE NOTE: SourceForge project shell services are
offline pending completion of an unscheduled maintenance event. The
SourceForge team is working to ensure that shell services are restored
in as timely a manner as is possible. No ETA has yet
This guy seems to have some interesting ideas about improving Freenet
performance:
- Forwarded message from Ray Heasman [EMAIL PROTECTED] -
From: Ray Heasman [EMAIL PROTECTED]
To: [EMAIL PROTECTED]
Cc: [EMAIL PROTECTED]
Subject: Some useful (I hope) musings on freenet
Hi Ian,
I was
On Sat, May 26, 2001 at 12:27:55PM -0500, Brandon wrote:
I've been thinking on the caching problem recently and it seems to me that
it is is indeed disadvantageous to cache aggresively because files cached
on the edge of the network have a decreased probability of being found by
an
So one thing we might consider is making 0.3.9.2 incorporate the
probabilistic caching that we have been talking about. It is fairly
unanimous that such an approach is a good idea, would be quick to
implement, and it could be ages if we wait for someone to get around to
simulating it before we
Shell access to sourceforge is now working again, check your email INBOX,
you may need to change your sourceforge password.
Ian.
PGP signature
On Mon, May 28, 2001 at 12:45:26AM -0400, Tavin Cole wrote:
Perhaps, but the node would basically fail on every request, except in
the unlikely event that it receives a request for the data in which it
wants to specialise - it would be a lottery.
Nah, it could employ a drifting algorithm
Just tried out the new Freeweb. The most obvious problem seemed to be
that whenever I tried to retrieve a file that I had inserted, it simply
contained the word bar...?
At a more general level, I would really encourage the FreeWeb guys to
get rid of that whole DNS business, it complicates
http://www.ucomics.com/doonesbury/viewdb.cfm?uc_full_date=20010528uc_comic=dbuc_daction=X
Ian.
PGP signature
On Wed, May 30, 2001 at 10:10:29AM +1200, David McNab wrote:
It does create the problem that if people start passing around
www.xxx.free URLs then these can only be used with Freeweb. Freeweb
should probably just interpret http://localhost:8081/ requests through
the proxy, it isn't as
On Tue, May 29, 2001 at 07:20:16PM -0400, eigenman wrote:
These arguments are dry and outdated.
Installing a program is not a massive task.
No, but writing and maintaining a piece of software for multiple web browsers,
and multiple platforms, and persuading newbies to change their browser
On Tue, May 29, 2001 at 07:24:54PM -0400, eigenman wrote:
Ok touche. Although the dot protocol is still nice such as
xxx.yyy.zzz.free.
What? They are the same, except David's protocol prepends www. and appends
.free. How that makes things simpler is beyond me.
Ian.
PGP signature
On Wed, May 30, 2001 at 12:28:31AM +0100, Dave Hooper wrote:
What's stopping David from doing this anyway, with or without the consent of
The Freenet Corporation?
A desire to write good software.
Ian.
PGP signature
On Tue, May 29, 2001 at 08:04:38PM -0400, eigenman wrote:
You do use HTTP directly. The WWW is HTTP.
But Ian just told me that KSK@ is the simplest choice for easy naming on the
freenet protocol.
He sugested this as an equivaliant to .free naming.
If this is insecure then freenet is
On Wed, May 30, 2001 at 02:02:27PM +1200, David McNab wrote:
Have a closer look at FreeWeb.
It may not have been noticed that it's easy to use FreeWeb in a way which
complies completely with the existing standards - this was true even in the
first prototype.
I have looked closely at FreeWeb
On Wed, May 30, 2001 at 02:05:16PM +1200, David McNab wrote:
I have not heard a single example of this. If true, then it should be
considered, and the MSK protocol may even need to be modified, but it
really
doesn't mean that we should convert to your .free protocol.
Try wget, it
On Wed, May 30, 2001 at 02:09:59PM +1200, David McNab wrote:
BUT, the MSK@SSK@... is *not* spoofable. KSK@blah is spoofable. FreeWeb
allow MSK@SSK@... - it looks up the KSK every time to avoid nasty long
URLs. Freenet is secure if you use MSK@SSK@
Again... groan... FreeWeb publishes and
On Wed, May 30, 2001 at 05:47:52PM +0200, Sebastian Sp?th wrote:
Hi all,
you can subscribe to the cvs change notification lists now. It is not
spam proof or moderated yet as it has to allow posts from
@users.sourgeforge.net mails. Does anybody can help me with a matching
regex expression
On Wed, May 30, 2001 at 12:07:59PM -0400, Benjamin Coates wrote:
How is freenet not a protocol with similar standing to HTTP and FTP? I would
say that's exactly what it is...
Because it *USES* HTTP for communcation with the local node, it is not a
replacement for HTTP, and there is no need
On Thu, May 31, 2001 at 07:31:12AM -0700, Mr. Bad wrote:
JM == Jamie Morken [EMAIL PROTECTED] writes:
JM Hi all, Here is a new scientist article on freenet's
JM development. Looks like Intel will be using Freenet
JM (opensource?)
JM
On Sun, Jun 03, 2001 at 02:58:36PM +1200, David McNab wrote:
I'm willing to write a complete Freenet node configurator as a native
Windows GUI app.
Good idea, might I suggest that we define a standard format to specify
command line options, maybe in XML, which would look like this:
section
It seems to have been a long time since we had some more long-term
discussions about Freenet, and features for the future. While it is
good that we have focussed on more short-term issues (hell, we have
*users* now!), I think it is about time that a few people started
concentrating on
On Sat, Jun 02, 2001 at 11:42:06PM -0700, coderman wrote:
In the aardvark example, I am led to believe that the keys are hash values,
and that lexographic distance is determined by numeric distance between the
two. Meaning, roughly, that a key with all 'a's would have a numericaly
On Sun, Jun 03, 2001 at 03:40:58PM -0400, Tavin Cole wrote:
1) is an inevitable consequence of any improvement to Freenet
functionality,
True, but this is something that can be cleanly separated from the
requirements for implementing a Freenet node.
Which is essentially semantics.
and
On Sun, Jun 03, 2001 at 10:05:36PM +0200, Oskar Sandberg wrote:
Another reason for seperation is that you could make the resulting URIs
protocol independant, to work for people who want to share files directly
(ftp or http) as well as via Freenet. (For example one could already hack
a
On Sun, Jun 03, 2001 at 05:40:16PM -0400, Benjamin Coates wrote:
Yeah, once you put a search entry onto a popular keyword, it would stay
around, relevant or not. Even reputation filtering wouldn't help that, since
it would be applied (by necessity) at the client after requesting all the
On Sun, Jun 03, 2001 at 03:10:01PM -0700, coderman wrote:
Who would be responsible for the unrequest?
The requestor (the clients should be designed to make this easy, or
automatic in some circumstances). Crypto would ensure that you can only
unrequest your own requests.
Could this be used
On Mon, Jun 04, 2001 at 12:50:23AM -0400, Mark J. Roberts wrote:
I have no idea how the unrequest actually works. I'm under the
impression that it works per-node and not per-reply, i.e., The node who
sent this response is evil! Punish it!
Nope, an unrequest simply undoes the effect of a
I keep getting:
Network Error: Freenet.KeyException: Couldn't read MSK Mapfile:
With recent CVS on both Snarfoo and GJs - is anyone else experiencing
problems?
Ian.
PGP signature
http://freenetproject.org/index.php?page=donations
Make sure you are sitting down, and check the recent amonymous donations
near the bottom of the page.
I have emailed this guy to make sure his donation was intentional.
Ian.
PGP signature
On Wed, Jun 06, 2001 at 12:17:12AM +1200, David McNab wrote:
Perhaps this is way ancient, but in case it's not, here goes.
I just ran a simulation prog (see below) for comparing linear versus
It is an old idea, but it is great that someone is looking into an
implementation.
Actually, I
I have noticed with recent CVS that FProxy is slow to return pages, and
sometimes it hangs for ages without returning an error message. Now it
may be something todo with my HTL 100, combined with QueryRestarted
messages coming from somewhere. I am getting suspicious that
QueryRestarted messages
On Wed, Jun 06, 2001 at 10:48:10AM +0200, Oskar Sandberg wrote:
On Tue, Jun 05, 2001 at 08:44:44PM -0700, Ian Clarke wrote:
I have noticed with recent CVS that FProxy is slow to return pages, and
sometimes it hangs for ages without returning an error message. Now it
may be something todo
On Wed, Jun 06, 2001 at 12:24:52PM -0400, Gianni Johansson wrote:
Problem 1: Coarse grained lock in MapHandler.lookup()
In the current implementation every request that goes through an MSK has
to acquire the lock for the synchronized static function MapHandler.lookup().
This means that
FProxy definitely seems to be less healthy these days, GJ pointed out
some obvious potential problems. I will do some investigating
this-evening, others should too.
Ian.
- Forwarded message from z08544 [EMAIL PROTECTED] -
From: z08544 [EMAIL PROTECTED]
To: [EMAIL PROTECTED]
Subject:
More - it looks like FProxy's mapfile handling is the culprit here
again.
Ian.
- Forwarded message from z08544 [EMAIL PROTECTED] -
From: z08544 [EMAIL PROTECTED]
To: Support@Freenetproject. Org [EMAIL PROTECTED]
Subject: [freenet-support] RE: SLOW!!! Why two nodes does not learn each
Great, I think that this warrents our 0.3.9.2 release.
I am tempted to reccomend that we incorporate Oskar's proposal for
probabilistic caching of documents (higher probability closer to the
origin of the request) which should have a beneficial effect on document
longevity.
Of course, in an
On Wed, Jun 06, 2001 at 01:07:10PM -0500, thelema wrote:
I'm all for trying out things that could help the network, but I don't
like jumping in head first. Not to mention I'd imagine lots of people
aren't going to upgrade to this newest version, leaving us in a
situation with a mixed node
On Thu, Jun 07, 2001 at 12:52:37PM -0400, Tavin Cole wrote:
I have a new suggestion for this: just count the number of hits on each key
and call that P, and whenever P reaches a certain limit, delete the key.
Er - I am probably misunderstanding you, but wouldn't this create a
trivial attack
1 - 100 of 3893 matches
Mail list logo
