View Issue Details

IDProjectCategoryView StatusLast Update
0003153unrealircdpublic2013-01-09 10:48
Reportersyzop Assigned Tosyzop  
PrioritynormalSeverityfeatureReproducibilityalways
Status closedResolutionno change required 
Product Version3.3-alpha0 
Summary0003153: Consider implementing &Channels
Description*CONSIDER* implementing &Channels (local channels) in Unreal33*. Check how much work it is, how it can be done nicely without much mess (macros/subfunctions).

Only if it's reasonable, we could implement it. Otherwise, we could decide once and for all not to implement local channels.
Additional InformationBasically, any channel command needs to take into account that anything from &channels should not broadcasted to the rest.
Additionally, any code like... if (*target == '#') { ...is a channel....} else { ...is a person....} needs to be changed as well...

A good way to start on this bug would be to check out /HELPOP ?USERCMDS, check out which are channel commands and check a couple (or all) of them to see how it can be done.

Again, adding some macros (or subfunctions) could be a good idea to keep things clean (like an IsChannel() macro, just to name one)

This bug was set to PRIVATE to avoid idiot non-technical discussion.
3rd party modules

Relationships

has duplicate 0003908 closed Suggestion for '&' channels 

Activities

stskeeps

2006-12-20 15:02

reporter   ~0012862

.. or implement #channel:servermask?

syzop

2006-12-20 15:05

administrator   ~0012863

I (and pretty much all users) prefer &channel, really :P

syzop

2007-04-28 07:45

administrator   ~0013906

One thing.. think about messages from/to &#blah
&#blah can mean: a message to chanadmins (&) and higher to channel #blah
or
&#blah can mean: a normal message to channel &#blah

one solution would be to disallow # as the 2nd character of a &channel (so no &#something channels), this fixes all the confusion. I don't see any better alternative that's as simple and effective as that.

aquanight

2007-04-28 10:49

reporter   ~0013915

[quote]one solution would be to disallow # as the 2nd character of a &channel (so no &#something channels), this fixes all the confusion. I don't see any better alternative that's as simple and effective as that.[/quote]

We'd also have to disallow & as 2nd character. Else &&blah could mean msg to channel "&&blah" or message to >=admins of "&blah".

Minor side note: would implementing this mean we could make can_override possible for local opers (but they'd only have it on &channels, not #channels)? Or a seperate flag, can_local_override for that purpose? Or just not screw with override at all :p ?

Shining Phoenix

2007-04-28 17:11

reporter   ~0013917

Last edited: 2007-04-28 17:25

Or we could still use channel masks, BUT:
- A mask cannot contain wildcards
- A masked channel exists on all U:Lined servers as well. So you could have a channel #blah:someserver.somenetwork.tld, which is registered, services do work on it, and StatServ sits in there \o/
- Globalops and higher are not kept out of channels by channel masks

aquanight

2007-04-28 17:37

reporter   ~0013918

Ugh. No. I'd rather complicate stuff with local channels than complicate it with channel masks.

syzop

2007-04-29 09:01

administrator   ~0013922

right aqua

WolfSage

2007-04-29 15:09

reporter   ~0013925

What about services?

syzop

2007-04-29 15:42

administrator   ~0013926

services have nothing to do with &chans, they'll never see them. this is intended behavior.

WolfSage

2007-04-29 17:56

reporter   ~0013931

I just wanted to make sure of that =)

vonitsanet

2007-04-29 18:04

reporter   ~0013933

I was thinking a way for global opers (or netadmins?) to join on remote servers local channels.
Is this possible?

Stealth

2007-04-29 21:38

reporter   ~0013940

Should the first joined get +q? It'd make sense

aquanight

2007-04-29 22:57

reporter   ~0013941

>I was thinking a way for global opers (or netadmins?) to join on remote servers local channels.
>Is this possible?

No. Local channels exist only on the server they are created on. They are not visible on any other server. Hence *LOCAL*.

[quote]Should the first joined get +q? It'd make sense[/quote]

No it does not. In fact, I see no reason why local channels should be created any differently than nonlocal (apart from not announcing them to other servers). +q is meant to go with services. Services cannot manage local channels. So local channels can never have +q set short of module or operator intervention. It would, however, not be difficult to implement module(s) for ircd-side local channel registration similar to ChanServ. Or a bot that uses joinpartsno + operoverride.

On that note, one thing that does need to be addressed is +L: it makes sense to allow &channel to +L #channel or &channel, but absolutely no sense in +L &channel from #channel (as the linked channel would not be the same on all servers that see the link). Granted this could just be left to users realizing that such a thing is useless.

driew

2010-04-15 09:18

reporter   ~0016067

[quote]On that note, one thing that does need to be addressed is +L: it makes sense to allow &channel to +L #channel or &channel, but absolutely no sense in +L &channel from #channel (as the linked channel would not be the same on all servers that see the link). Granted this could just be left to users realizing that such a thing is useless.[/quote]

I would suggest that both +L &channel from a #channel, and +L #channel from a &channel should be allowed.
I could see potential usages for both. No reason to restrict it really.
Just like how I see +u being completely useless. But someone has discovered a good use for it.

syzop

2013-01-09 10:48

administrator   ~0017334

This feature is really too much work for too few people who need it.

Issue History

Date Modified Username Field Change
2006-12-20 06:15 syzop New Issue
2006-12-20 06:16 syzop Additional Information Updated
2006-12-20 15:02 stskeeps Note Added: 0012862
2006-12-20 15:05 syzop Note Added: 0012863
2007-04-27 03:15 stskeeps View Status private => public
2007-04-27 03:15 stskeeps Status new => acknowledged
2007-04-28 07:45 syzop Note Added: 0013906
2007-04-28 10:49 aquanight Note Added: 0013915
2007-04-28 17:11 Shining Phoenix Note Added: 0013917
2007-04-28 17:12 Shining Phoenix Note Edited: 0013917
2007-04-28 17:25 Shining Phoenix Note Edited: 0013917
2007-04-28 17:37 aquanight Note Added: 0013918
2007-04-29 09:01 syzop Note Added: 0013922
2007-04-29 15:09 WolfSage Note Added: 0013925
2007-04-29 15:42 syzop Note Added: 0013926
2007-04-29 17:56 WolfSage Note Added: 0013931
2007-04-29 18:04 vonitsanet Note Added: 0013933
2007-04-29 21:38 Stealth Note Added: 0013940
2007-04-29 22:57 aquanight Note Added: 0013941
2010-04-15 09:18 driew Note Added: 0016067
2010-05-24 14:16 syzop Relationship added has duplicate 0003908
2013-01-09 10:48 syzop Note Added: 0017334
2013-01-09 10:48 syzop Status acknowledged => closed
2013-01-09 10:48 syzop Assigned To => syzop
2013-01-09 10:48 syzop Resolution open => no change required