GuildWiki talk:Bots

I don't think this is necessary, really. Only RT and Dr ishmael have authorized bots, and they are a) both sysops and b) both much more knowledgable than I am. 23:04, 30 April 2008 (UTC)
 * Even so, if GuildWiki was to get more bots, this my actaully come in handy :o) -- Shadowphoenix  23:05, 30 April 2008 (UTC)
 * I don't think this is such a bad idea. It would also be useful if botting becomes a GuildWiki fad lol. Anyway, it would be helpful to have tips on when you shold be using a bot and when its okay to clutter RC because they have to be manual editting, and I think we should have even less tolerance of non-flagged accounts than what is mentioned here. I dont know about anyone else, but RC patrol is my primary way of keeping an eye on things. Its much harder when RC gets flooded with bot tasks. Also, I think it would be useful to make it mandatory to post the scripts for bot tasks that use one just in case we need to reverse the bot actions and the owner disappears. I'm not sure if you can export the AWB settings for non-scripted tasks but having that data would be good too. &mdash;♥ Jedi ♥ Rogue ♥ 23:45, 30 April 2008 (UTC)
 * It is a draft so add your ideas if u feel they are appropraite :) -- Shadowphoenix  23:49, 30 April 2008 (UTC)

Ishy's ideas:
 * Not all bots require script+framework - mine will be running self-contained Perl scripts, unless you count the Perlwikipedia module as the framework in that case. I wasn't sure how to reword this, so I left it alone.
 * I'd still like to restrict bots to flagged bot accounts. This may just be my job training speaking, but I worry for what might happen if we allow just anyone to run bots here.  We've recently seen how easily a non-bot user can disrupt the wiki by creating a large number of articles.  Yes, it would be easy enough to have another bot revert unapproved bot edits, but preventing something before it happens is better than having to do damage control afterwards.
 * I don't like the bit about bots stopping due to talkpage edits. This means that any malicious (or even unknowing) user could disrupt the bot just by posting on its talkpage.  That leaves "the big red button" that RT and I stole from Fyrenbot, which I think would be enough considering how active our current admins are.
 * Jedi's idea for when (not) to use a bot is good. There are many tasks, such as spellchecking, image license tagging, image deletion, etc. that require some subjective judgment, and therefore could not be performed by a fully automated bot.  I'll go ahead and write up a section for this.  &mdash;Dr Ishmael [[Image:Diablo_the_chicken.gif]] 04:39, 1 May 2008 (UTC)
 * The talk page security measure is made incase something was to happen when no admins are online, it is made so that reg. users can stop the bot if it is needed; however, I can see how this could be misused, I still think it should be added. -- Shadowphoenix  04:44, 1 May 2008 (UTC)
 * I'm on 14 hours a day or so, and RT is on the rest of the time. No problem. [[Image:Felix_Omni_Signature.png]] 04:53, 1 May 2008 (UTC)