Just a sample of the Echomail archive
[ << oldest | < older | list | newer > | newest >> ]
|  Message 738  |
|  Nicholas Boel to Björn Felten  |
|  MakeNL v3.4.1 Release  |
|  29 Dec 13 10:43:52  |
 Hello Björn, On 29 Dec 13 15:14, Björn Felten wrote to Ward Dossche: WD>> It has to be enabled "top down" and not "bottom up". Even the WD>> developpers are not certain what will happen if it is done WD>> bottom-up, I've read the whole segment could be dropped. BF> Chicken! BF> Why not live dangerous. Enable it and see what happens. It wouldn't even be considered living dangerous. We *NEED* him to enable it before anyone else can even try. Nothing on his end would be dropped, and actually everything would work on his end. If ZC1 doesn't have it enabled, though.. once a modified Z2 segment made it's way over here, it would get screwed up. We basically need ALL ZCs to enable it at once. Then all RCs, then all NCs for absolutely no interruption in the transition of the nodelist. This will most likely not happen since there's RCs and NCs that sometimes don't respond for weeks at a time. But who cares? We've gone a few days with a hiccup somewhere upstream before, and when things are restored it works again. ZCs should enable it, and then any RCs that see that the ZCs enabled it should enable theirs, and so on with NCs. There shouldn't even be any hiccups in the creation of the nodelist, because when RCs and NCs finally SEE that there has been a change, their current segments will still work, yet they have the ability to change to the new way of listing unpublished numbers. The only way it will break is if an NC makes the changes before their RC, or RC changes before their ZC. So it's on the ZCs to make the move before anyone else can do anything, and here we wait. :) BF> After all, most serious mail movers don't rely entirely on the BF> nodelist any more. Almost everything will work just fine even with the BF> entire zone 2 removed from the list. BF> At least the Fidoweb will easily survive one week without any BF> problems at all... Kind of.. Even if there was one messed up nodelist (which I don't see happening as long as the top down methodology happens as it should), it's not going to blow the whole world up (remember when Roy tried it? It errored at his RCs system and that was the end of it. His netseg didn't make the nodelist for one week.. whoop-dee-do! It was fixed shortly thereafter), and even then, there is an entire week to fix it! Regards, Nick --- GoldED+/LNX 1.1.5-b20130910 * Origin: Dark Sorrow | darksorrow.us (1:154/701) |
[ << oldest | < older | list | newer > | newest >> ]