On the irc.webchat.org network, occassionally they reset the servers and this results in your bots being deopped and having to re-identify themselves.
All the chanserv/nickserv scripts I have use easily identifies the bot and asks for ops when the bot joins a channel BUT doesnt handle the event of the bot no longer having ops and needing to re-identify itself and asking for ops again.
In summary, I'm after a script that does the following:
1. bot detects it no longer has ops
2. re-identifies itself to NickServ with:
/msg NickServ IDENTIFY <password>
3. ask ChanServ for ops with:
/msg ChanServ OP #<room> <nick>
The bot will check if it needs ops every 20 minutes, but you can also specify certain channels be 'op-critical', meaning the bot won't allow itself to stay deopped at all.
(As a note, this script will never flood services, and even if the same commands are somehow sent to the eggdrop's output queue twice, there's an elongated delay between commands to services, so the netadmins won't complain.)
I just edited the script i was using,
"Chanserv.tcl 0.1 by #egghelp@efnet (KuNgFo0)" and added in an extra line to make my bots msg nickserv first to re-identify themselves and then they'll ask chanserv for ops...pretty simple and straight forward..
As for you comments about my script, here's an explanation...
The only time the bot should be deopped in an op-crit channel is if it has just logged on to the IRC server, having been disconnected (this does not include the bot's initial startup).
In this situation, it is *supposed* to call the identification procs using init-server as a trigger, but the proc is being called too early (before the bot even gets on irc) which is why its not identifying. If you were to wait until the next automatic login (which occurs at :15 :35 :55 past every hour) the bot would regain ops in the op-crit channels, and then retain them (unless cservice goes offline).
I will naturally examine the script and see if I can do something to fix this bug by the next release. If your bot, and the irc server its connected to, are stable this should never pose much of a problem ofc.
Additional fyi: when the bot DOES begin to identify, it will take about 6 seconds to regain ops in *each* channel. This is so as to avoid flooding the channels services, of which the sensitivity of the flood threshold varies from net to net. This is first and foremost a multi-network channel services interaction script, and I think it's fair to desist from flooding on any network, don't you?
I don't recall receiving any e-mail from you, perhaps botservice mail was offline at the time you sent it. Sorry I can't offer a better explanation on that.