On Thu, 23 Oct 2003, Toad wrote:
It IS called unstable you know.
Dude. That's pretty dense.
Was doing this sudden, forced fork discussed? I mean, it seems like most people here agree that it's a good idea, so what would have been the harm in throwing out the idea on devl and giving it a day for discussion?
It's like you were worried about people objecting, so you did it in stealth so that you'd rip all those nodes out of The Freenet and put them to work in this fork freenet.
No, it's like he was worried about not having a good way of testing NGR. He's a programmer. He and probably most of the other programmers on the project felt like this was the best way to go to get the job done. I'm astonished that someone not involved with the coding would try to micromanage the coders.
You (Toad) and whoever few also agreed to this beforehand have basically taken control of everyone's node without permission, inflicting, in the process, irreversible damage (removal of the routing table).
I understand that this is the unstable branch we're talking about. But, "It IS called unstable you know" works if someone's node crashes. That's what unstable means. This forced fork doesn't have a damned thing to do with the node being unstable. Running an "unstable" node shouldn't mean giving control of it to a small core of developers. And if it does, it should have been discussed beforehand so that people could (implicitly or explicitly) agree to it, or forego the upgrading. We *are* talking about a group of people generally keen on privacy issues.
Gee, did Toad take control of your node really? You didn't have to download freenet-latest-unstable.jar, and you didn't have to overwrite your freenet.jar. You did all that on UNSTABLE without a clue as to what the changes were. How hard would it have been to look at the CVS list? I would expect anyone so keen on privacy issues would at least do that! In CVS, Toad wrote:
FORK UNSTABLE NETWORK. Protocol version increases to 1.47. The purpose of this: it is likely that NGRouting and CPAlgoRT are not really compatible in their present implementations. Hopefully we can get a decent sized test network together.
30 minutes later, he posted more explanation along with instructions to DEVEL.
So, Toad certainly didn't *force* anyone to compromise their privacy. To get on the new network, you'd have to do it intentionally: download unstable.ref and overwrite your seednodes.ref. Furthermore, I'm not sure how a network fork relates to privacy issues.
Todd Walton wrote (continued):
Unleashing NGR on the network and hosing everything was a mistake. There's no other network out there of Freenet's style with Freenet's size. Managing this network is still largely virgin territory. We can forgive ourselves for screwing that up. But that should be all the more reason to exercise better care in the future, which, apparently, is not what happened here.
Please. In the future can we be given sufficient forewarning if things are going to significantly change?
Sure thing. For the future, I'm instructing the developers to mark significant changes to the code by labeling it "freenet-latest-unstable". Toad, get to it!
> And if someone doesn't like the way
things have gone, can you at least give them a fair explanation, and withhold the disingenuousness?
It ain't disingenuous. It's the truth. And Toad did give an explanation in DEVEL. Toad said:
Of course you may wish instead to revert to stable, or to run two nodes.
The purpose of the fork is so that we can have a network exclusively consisting of NGRouting nodes; with enough nodes we should be able to test NGRouting without problems with the large number of stable nodes on the old network which run the old routing algorithm, which we believe not to be very compatible. This network will eventually become the basis of Freenet 0.6, we hope.
and later he said:
Furthermore the main motivation in all this is to get something good enough to be merged to stable. We can't test NGRouting properly on the hybrid stable-unstable network. This is for the benefit of the list and not matthias.
Maybe you skipped that?
-Martin
_______________________________________________ Devl mailing list [EMAIL PROTECTED] http://dodo.freenetproject.org/cgi-bin/mailman/listinfo/devl
