View Issue Details

IDProjectCategoryView StatusLast Update
0004172unrealircdpublic2015-08-08 17:05
Reporterkatsklaw Assigned Tosyzop  
PrioritynormalSeverityfeatureReproducibilityN/A
Status closedResolutionfixed 
Summary0004172: Add ability to hide certain usermodes in LUSERS output.
DescriptionAs discussed in forum post: http://forums.unrealircd.com/viewtopic.php?f=6&t=484

The purpose if this feature is to add the ability to hide certain users, IE Services pseudo-clients. By hiding user modes, it's then possible to hide other users as well +B or +H, +N, +A users as well.
3rd party modules

Activities

Anikwa

2013-02-15 21:04

reporter   ~0017419

Should the mode-to-hide be configurable or static?

syzop

2013-03-02 19:04

administrator   ~0017458

tech: if you do this, and carefully code the clients++ / clients-- based on MODE's etc etc.. which has a lot of potential for errors (we had issues with this for several years with +H with negative user counts!!)
I mean, even if you do that, then you still would have to recalculate the user count figure if the configuration value changed, as you have to recalculate who should be hidden (from the user count)...

Other than that, I would like to hear if other people support this idea. There should be some real need for this, before we consider implementing it.

katsklaw

2013-03-03 01:13

reporter   ~0017459

An alternative would be to forgo the umode hiding and just hide IsUline() clients. That way at least services/stats are hidden.

Perhaps that's a bit easier to implement since we can just ignore propagated user counts upon link/burst. I don't think a bump in UnrealProtocol would be needed either if the uplink server simply didn't update the user count when a U:Lined server linked.

nenolod

2013-05-06 06:08

reporter   ~0017487

I think the only rationale for this is to stop people from gaming the SearchIRC and netsplit stats bots.

If someone submits a patch for this, I will look at it, but I have no interest in coding it.

syzop

2015-08-08 17:05

administrator   ~0018632

I do understand the rationale but I think the effort, change of failure (vs) CPU usage etc makes this less desired. Unless someone has some very bright idea..

Issue History

Date Modified Username Field Change
2013-02-14 21:45 katsklaw New Issue
2013-02-15 21:04 Anikwa Note Added: 0017419
2013-03-02 19:04 syzop Note Added: 0017458
2013-03-03 01:13 katsklaw Note Added: 0017459
2013-05-06 06:08 nenolod Note Added: 0017487
2015-08-08 17:05 syzop Note Added: 0018632
2015-08-08 17:05 syzop Status new => closed
2015-08-08 17:05 syzop Assigned To => syzop
2015-08-08 17:05 syzop Resolution open => fixed