There may be a compromise that bahamut almost has implemented already. We could use the oper port method to create a port that masks connections that come through it, and have our standard connection system that functions as is. Once the host has been processed during the connection, everything ends up the same from there. If someone wants to remove their hostmask, they could go through the process of disconnection/reconnection without it having to be coded for.
They want hostmasking? Connect on 6667
They don't? Connect on 7000
It may be a way to implement the feature with the least amount of possibilities of abuse, or other failures.
On Wed, Oct 21, 2009 at 6:13 PM, Trevor Talbot <quension@mac.com> wrote:
On Oct 21, 2009, at 10:06 AM, Michael Reynolds wrote:
On Wed, Oct 21, 2009 at 3:53 AM, Trevor Talbot <quension@mac.com> wrote:
Why is the ability to opt out of hostmasking in realtime considered so desirable that it is worth the violation of the host-immutability assumption and all of the associated ill effects?
To be perfectly blunt, users don't really care about standards and protocols. What they care about is hiding their host, and the amount of work necessary to accomplish that goal. Undernet and Quakenet gets away with registration based via redundant services.
That's not what I was getting at.
In many of the comments in this thread, I'm seeing the idea that:
A) users want the ability to *avoid* hostmasking
B) it is important that they toggle it in realtime, without disconnecting
I'm asking why users want to avoid hostmasking, why it is important that it happen in real time, and why this is so useful that it outweighs the disadvantages of supporting it.
If toggling hostmasking in realtime is not required, then servers do not need to implement quit/reconnect simulations and other workarounds, and various small problems do not ripple out to annoy client maintainers, script writers and end users.
If disabling hostmasking is not required, then servers do not need to implement matching for real hosts, basic users never have to question whether they are protected, technically minded users don't have the mental baggage of a parallel host namespace they may or may not see, and channel ops and harassment victims don't have to deal with Yet Another Way abusers can evade their bans and ignores, however rare it may be.
Applying non-optional fixed hostmasking on connect seems to reduce complexity across the board. What are the advantages to increasing it?
-- Quension
_______________________________________________
DALnet-src mailing list
DALnet-src@lists.dal.net
https://lists.dal.net/mailman/listinfo/dalnet-src