View Issue Details

IDProjectCategoryView StatusLast Update
0001365unrealircdpublic2003-11-26 19:53
ReporterAngryWolf Assigned Tocodemastr 
PrioritynormalSeverityfeatureReproducibilityalways
Status resolvedResolutionfixed 
PlatformLinuxOSSuSEOS Version8.0
Product Version3.2-beta18 
Summary0001365: hooktype modification requests
Description1. /botmotd doesn't work remotely (because of "NOTICE AUTH"). Could this be fixed? [syz: **fixed**]

2. HOOKTYPE_REHASHFLAG is very useful, but to send back more informative responses (who used the flag, was it remotely or not), I need cptr and sptr as parameters in the callback function. Could you please add them?
TagsNo tags attached.
3rd party modules

Activities

AngryWolf

2003-11-15 15:51

reporter   ~0004037

Oh, and let me complement the bugreport with another idea, a feature request for the new +f channel mode: if I type /mode +f [3j#i1]:10, it works great, but we know "i" is the default setting. Would be there any problems with allowing the users to set simply [3j1]:10, for example?

syzop

2003-11-15 18:56

administrator   ~0004040

[@+f] yes, this won't be changed.. it is that way to make it easy and backward compatible.

AngryWolf

2003-11-15 19:07

reporter   ~0004041

Ok, than forget my previous bugnote. :-)

AngryWolf

2003-11-17 06:40

reporter   ~0004049

I know I'm getting to ask too much, but I think it worth to try another suggestion.

3. Do you think it would be a good idea to allow modules to change the text of part messages? I mean to change the return value of the callback function for HOOKTYPE_LOCAL_JOIN to char *. I can give you an example why I need this feature, if my suggestion is not enough.

syzop

2003-11-20 01:39

administrator   ~0004067

botmotd issue fixed in .1983 :).

Haven't looked at the hooks, but both requests would break current modules so I would have to ask code/sts and I'm not in the mood to break modules this time (maybe next release).

codemastr

2003-11-23 23:17

reporter   ~0004085

I'm not sure about the part message thing yet, but I'll agree with the HOOKTYPE_REHASHFLAG. That's really something that should have been in there from the beginning, but some how I didn't think of it. And, as far as I know, AngryWolf is the only one who is currently using this hook, so if he doesn't care about it breaking his modules, then I don't either :).

Btw, when you said "HOOKTYPE_LOCAL_JOIN to char *" I assume you really meant to say HOOKTYPE_LOCAL_PART? What I'm thinking is, perhaps the best way to handle this is with a seperate hook, HOOKTYPE_PRE_LOCAL_PART, that would work just like the other _PRE_ hooks, it triggers before the actual part and HOOKTYPE_LOCAL_PART triggers when the user actually parts. That would prevent it from breaking existing modules.

AngryWolf

2003-11-24 07:53

reporter   ~0004088

Yes, I meant to say HOOKTYPE_LOCAL_PART, sorry. Also, I agree with both adding a HOOKTYPE_PRE_LOCAL_PART and breaking my modules. :-)

syzop

2003-11-24 14:46

administrator   ~0004090

I use this hook too in my Monitor module ;p.
It broke already from b17->b18 so I thought it would be nice if it wouldnt at b18->b19... If you think it's worth it, then go for it. I think my current module uses 'return 0;' so that's like return NULL... which presumably will mean "strip comment"/"no part reason" so at least it won't crash ;).

AngryWolf

2003-11-24 18:58

reporter   ~0004097

Last edited: 2003-11-24 19:03

This hook = HOOKTYPE_LOCAL_PART? If yes, it's not worth to break your module (coz it would change part messages), in my opinion, the _PRE_ hook should be implemented instead. And eventually in the future, if you plan to break a bunch of things in current modules, I know, you, coders, will increment the module version number that UnrealIRCd will support at least, so it will be a peg to hang on cleaning up the code of hooktypes... you already made a great work (I mean the module API), keep it up. :-)

edited on: 11-24-03 19:03

codemastr

2003-11-24 21:29

reporter   ~0004105

The HOOKTYPE_REHASHFLAG syntax has been changed to:
callback(aClient *cptr, aClient *sptr, char *theflag)
as of .1994

AngryWolf

2003-11-25 09:34

reporter   ~0004108

Great, thanks!

AngryWolf

2003-11-25 09:56

reporter   ~0004109

Or... oops, it's still not perfect. Now I can see different RunHooks for the same hooktype:

1. RunHook3(HOOKTYPE_REHASHFLAG, cptr, sptr, parv[1]);
2. RunHook(HOOKTYPE_REHASHFLAG, parv[1]);

:-(

codemastr

2003-11-26 19:53

reporter   ~0004128

HOOKTYPE_PRE_LOCAL_PART was added in .1995:
char *callback(aClient *sptr, aChannel *chptr, char *comment);

and the rehashflag stuff should work now.

Issue History

Date Modified Username Field Change
2003-11-14 19:54 AngryWolf New Issue
2003-11-15 15:51 AngryWolf Note Added: 0004037
2003-11-15 18:56 syzop Note Added: 0004040
2003-11-15 19:07 AngryWolf Note Added: 0004041
2003-11-17 06:40 AngryWolf Note Added: 0004049
2003-11-20 01:39 syzop Note Added: 0004067
2003-11-20 15:57 syzop Severity minor => feature
2003-11-20 15:57 syzop Summary /botmotd and hooktype rehashflag => hooktype modification requests
2003-11-20 15:57 syzop Description Updated
2003-11-23 23:17 codemastr Note Added: 0004085
2003-11-24 07:53 AngryWolf Note Added: 0004088
2003-11-24 14:46 syzop Note Added: 0004090
2003-11-24 18:58 AngryWolf Note Added: 0004097
2003-11-24 19:03 AngryWolf Note Edited: 0004097
2003-11-24 21:29 codemastr Note Added: 0004105
2003-11-25 09:34 AngryWolf Note Added: 0004108
2003-11-25 09:56 AngryWolf Note Added: 0004109
2003-11-26 19:53 codemastr Status new => resolved
2003-11-26 19:53 codemastr Resolution open => fixed
2003-11-26 19:53 codemastr Assigned To => codemastr
2003-11-26 19:53 codemastr Note Added: 0004128