[freenet-support] Re: Old Freenet

2004-01-11 Thread Garb


-Original Message-
 Message: 2
 Date: Sat, 10 Jan 2004 19:29:38 +
 From: Toad [EMAIL PROTECTED]
 Subject: Re: [freenet-support] Old Freenet
 To: [EMAIL PROTECTED]
 Message-ID: [EMAIL PROTECTED]
 Content-Type: text/plain; charset=us-ascii

 On Sat, Jan 10, 2004 at 06:04:17PM +, Andrew Dickson wrote:
 Let's face it, freenet was working brilliantly this
 time last year... Content was slow, but it was
 accessible. At the moment freenet is dead in the
 water.

 The unstable network on the other hand is working pretty well. We should
 be merging the code in the next week or so, however the unstable network
 *is* rather small (100 nodes or so). Anyone who is willing to provide
 bug reports and upgrade their nodes daily should try the unstable branch
 however.
 
 Why is the stable network not reverting back to the
 older code until these new bugs are ironed out?


Actually I have wondered a great deal about this myself. Why did Freenet
break in the first place and - assuming that it happened through an error
introduced by an update about a year ago - why wasn't the stable version
reverted to the last working version, so we could have avoided paralyzing
the entire net for a year? Even if the fix is allmost ready for the big
launch, the other issues would be interesting to know about, just for
historical reasons.

Regards,
Jesper

___
Support mailing list
[EMAIL PROTECTED]
http://news.gmane.org/gmane.network.freenet.support


Re: [freenet-support] Re: Old Freenet

2004-01-11 Thread Jay Oliveri
On Sunday 11 January 2004 07:40 am, Garb wrote:
  On Sat, Jan 10, 2004 at 06:04:17PM +, Andrew Dickson wrote:
  Let's face it, freenet was working brilliantly this
  time last year... Content was slow, but it was
  accessible. At the moment freenet is dead in the
  water.
 
  The unstable network on the other hand is working pretty well. We
  should be merging the code in the next week or so, however the unstable
  network *is* rather small (100 nodes or so). Anyone who is willing to
  provide bug reports and upgrade their nodes daily should try the
  unstable branch however.
 
  Why is the stable network not reverting back to the
  older code until these new bugs are ironed out?

 Actually I have wondered a great deal about this myself. Why did Freenet
 break in the first place and - assuming that it happened through an error
 introduced by an update about a year ago - why wasn't the stable version
 reverted to the last working version, so we could have avoided paralyzing
 the entire net for a year? Even if the fix is allmost ready for the big
 launch, the other issues would be interesting to know about, just for
 historical reasons.

The last working version didn't work very well, especially on a larger 
scale.  There were overloading problems for nodes, and routing problems 
that next-gen routing were going to address that were made worse.

Added to this, the project recently agreed to have a stable as well as 
unstable network.  Before that, the network had stable and unstable nodes 
all attempting to work together.  This made improvements in the node 
difficult to measure when older nodes are part of the network.

Backing off to an ancient build also removes any and all improvements and 
bugfixes made all over the codebase, including some FCP related that I was 
interested in.

More than likely, things will stabilize again over time (perhaps shorter 
than everyone thinks) and we will have two separate networks and hopefully 
less complaining and whining on freenet-dev.  Overall things are looking 
better in the unstable network, and there's talk of merging some changes 
into the stable branch, but all this takes time and effort.  Keep in mind 
it's still 0.5, which implies it's not ready for 1.0 release.

-- 
Jay Oliveri  In the land of the blind,
GnuPG ID: 0x5AA5DD54  the one-eyed man is king.
FCPTools Maintainer
www.sf.net/users/joliveri

___
Support mailing list
[EMAIL PROTECTED]
http://news.gmane.org/gmane.network.freenet.support