The following discussion is an archived debate. Please do not modify it. To request review of this BRFA, please start a new section at WT:BRFA. The result of the discussion was Request Expired.

Operator: Cyberpower678 (talk · contribs · SUL · edit count · logs · page moves · block log · rights log · ANI search)

Time filed: 07:19, Thursday May 31, 2012 (UTC)

Automatic, Supervised, or Manual: Automatic (will be monitored at first for any glitches)

Programming language(s): PHP

Source code available: No

Function overview: Notifies subscribed users in this list to vandalism levels when a set threshold is reached.

Links to relevant discussions (where appropriate): Wikipedia_talk:Counter-Vandalism_Unit#Task_Force

Edit period(s): Reads the ((Vandalism information)) template every minute. Edit period depends on Vandalism levels.

Estimated number of pages affected: Number of users subscribed in the list

Exclusion compliant (Yes/No): Yes

Already has a bot flag (Yes/No): No

Function details: During my Wikibreak, I was summoned to the task force to create a notifications bot for users so they don't have to proactively check the vandalism information template. This is what this task does. It reads ((Vandalism information)) and notifies the users who signed up with the minimum DefCon setting met. It only notifies people in this list.

Discussion[edit]

There is still question about how often the bot should notify a user. It checks it every minute and sends notifications out every minute which is very annoying. Some suggestions would be appreciated.—cyberpower ^Chat_Offline 07:19, 31 May 2012 (UTC)[reply]

• I can't really see the point. Personally, I just check the vandal info template before I start vandal fighting, then maybe again half an hour later. It never gets updated very often anyway. Who wants to be disturbed just to have a bot tell them the defcon has changed? Does it tell them even if it hasn't changed? I think a better thing would be just to tell them when it does change, rather than every minute or whatever. But maybe set it so it only does it when the user's status is set to 'online', so their talkpage isn't nothing but these reports. This is all totally pointless, somebody could just put the template on their watchlist and see when it changes then! I object to the bot. Rcsprinter (converse) 16:16, 1 June 2012 (UTC)[reply]

  I am doing this because it was asked of me to make a bot. Only users who subscribe to notifications will receive them and they set the minimum DefCon that they would like to receive notifications for. No it will not warn them repeatedly for the same level.—cyberpower ^Chat_{Limited Access} 17:54, 1 June 2012 (UTC)[reply]

• Oppose: Given his history, this user should be required to submit the source code along with any bot requests until he has a trustworthy track record of running and maintaining a bot. As for the bot itself, not only is this a pointless task (the template was created so interested editors could easily monitor the levels), but the whole idea of a vandalism task force is waste of Wikipedia time. If you want to "fight" vandalism, you just open up your watchlist or recent changes and start checking edits. So I'd be against wasting server resources on up-to-the-minute vandalism level checks from a bot, a) for something already handled in a by-request template and b) for the purpose of turning Wikipedia into a video game. — Bility (talk) 17:12, 1 June 2012 (UTC)[reply]

  What? Have I proven to not be able to run bots? I am running approved tasks across several wikis including this one and have maintained and updated when needed. I fail to see where you are trying to go with this.—cyberpower ^Chat_{Limited Access} 17:54, 1 June 2012 (UTC)[reply]

  I'm referring to Cyberbot I and the events surrounding it. Do you currently run a bot on en.wikipedia? What is the bot's name? Where I am going with this is that concerns were brought to light in the Cyberbot I request that you hadn't fully understood the code you were modifying and/or weren't taking the time to make sure it was correct. I would be more comfortable with a second pair of eyes on your code until you have a bit of history running/maintaining a bot. We do not need to prove you are unable to run a bot, you must prove you're able to run a bot. As far as other wikis go, they all have their own guidelines and standards for giving out bot flags, so I'm only interested in your actions on this wiki. — Bility (talk) 18:32, 1 June 2012 (UTC)[reply]

  Yes. User:Cyberbot I runs on several wikis including this one.—cyberpower ^Chat_Online 19:39, 1 June 2012 (UTC)[reply]

  Okay, that seems fine then. — Bility (talk) 21:06, 1 June 2012 (UTC)[reply]

  To add on, so that we all understand the purpose of this bot, let me try and get to the benefits this bot will provide. First, because it isn't clearly stated on this request or the instructions at the subscriber list, I am assuming this bot will "send notifications" by posting to users' talk pages. Okay. Users will either be online or offline. In the latter case, when they log in they will have a message of the last time the vandalism level passed their threshold, in which case they need to look at the template to get the most current information. Or will the bot rewrite messages to update the latest time or add messages for when the level falls back under the threshold? If so, will there be a series of messages on someone's talk page when the level flips back and forth between 2 and 3 for instance?

  Now let's say the user is online and actively doing something so they would see a talk page notification. If they are currently "fighting" vandalism, then they check their talk page and go right back to what they were doing, correct? In which case it was a pointless notification. If instead they are just browsing or doing some other type of editing, they will check their talk page and decide whether to stop what they were doing to go handle increased vandalism. I think that covers all the scenarios?

  So in conclusion, this bot will make requests to the server every minute so that users who are online and not currently fighting vandalism are aware of the elevated levels and can then maybe go deal with it. And how many people will this potentially be supporting? There are under 100 on this list and currently just three on the subscriber list. Is this a good use for a bot? — Bility (talk) 18:57, 1 June 2012 (UTC)[reply]

  The list was just created and nobody yet knows about it. It takes time for something to gain popularity. There are other methods of getting the users attention. A template that is changed by the bot for instance is another method and will trigger the new messages bar. The point of this bot is to let dedicated vandal fighters know when vandalism levels are elevated and should join in fighting. It's not mandatory and it allows each user to set their own minimum DefCon level.—cyberpower ^Chat_Online 19:39, 1 June 2012 (UTC)[reply]

  How will changing a template trigger the new messages bar? My point about there being less than 100 people in the Task Force is that is the theoretical limit even at full popularity, yes? The existing vandalism information template already "let['s] dedicated vandal fighters know when vandalism levels are elevated", so my question is whether it's worthwhile to let a bot ping the server 1,440 times a day for the additional benefit of giving a small number of people the new messages bar. I can't speak to Wikipedia's server load, so if once-a-minute requests won't affect things then I don't have a real objection, I just think this task isn't important enough to allow impacting the server. I do think small bot tasks like this can eventually add up to a material difference though, so we should limit approvals to tasks with high benefit for the project. — Bility (talk) 21:06, 1 June 2012 (UTC)[reply]

  It may be a small task, but pulling page content from Wikipedia takes less resources than making an edit. There is practically no resource gobble up when it reads the DefCon. The thing is, if dedicated vandal fighters respond more promptly by getting a notification and thereby increasing vandalism control, it is thereby an effective task in a way that it keeps Wikipedia clean. Again the list may be small, but the benefits can be big. I recommend having this bot approved for trial. If it turns out it's not worth the effort, the task can still be killed. There is a shutoff switch for easy control.—cyberpower ^Chat_Online 21:35, 1 June 2012 (UTC)[reply]

  You're recommending it for trial by submitting the BRFA and should wait until some BAG members have had their say before pushing it. There are probably another load of things they'd like to say. And also, how can you get some people to subscribe just for a trial? Rcsprinter (converse) 21:41, 1 June 2012 (UTC)[reply]

  That's not what I meant. I meant that putting it through a trial to see if it makes noticeable change by alerting those who respond to notifications better than by proactively checking the template. It's BAG's choice if and when to approve it.—cyberpower ^Chat_Online 22:07, 1 June 2012 (UTC)[reply]

• Can I please see/review the source code?
• There needs to be some kind of rate limiting to avoid spamming the crap out of people
• It'd like to see more discussion about this. Three comment on Wikipedia_talk:Counter-Vandalism_Unit isn't nearly enough (particuarly as one of the comment is "I think it would be impractical to send a message to each user's talk page each time vandalism is abnormally high")

To be clear, if you can show us, that users acutally want this bot, and that it will be useful; then we can probably go ahead with it. However at the moment, I don't think that is the case. --Chris 03:06, 2 June 2012 (UTC)[reply]

• People would have to first know that this exists. In that case it would need to be advertised.—cyberpower ^Chat_Offline 12:18, 2 June 2012 (UTC)[reply]

  Indeed. Please advertise it (Wikipedia:Village pump (proposals) is probably the best place, although any widely read page/noticeboard will do). What about my other two questions? --Chris 09:14, 8 June 2012 (UTC)[reply]

  Sorry for not getting back to you earlier. I've been incredibly busy. To answer your other two questions, I'm not really willing to post my source publicly yet but can e-Mail you the script that will run the bot. I've run simulations and the bot performs normally however, I'm not ready to give you the code yet because until I have the rate limiter installed to avoid spamming it's not ready for testing yet. I was hoping to get input here but I just came up with the idea that every user should choose for themselves by setting that when signing up. Does that sound like a good idea?—cyberpower ^Chat_{Limited Access} 10:15, 8 June 2012 (UTC)[reply]

I have to agree with Rcsprinter's concerns over the usefulness (or rather, lack thereof) of this bot. I'm not entirely sure why this needs to be a bot—at all. I mean, if the goal is to spawn the "you've got messages" box (or something equally intrusive) whenever the vandalism level is high, wouldn't this be better as a javascript-based user script? Or a gadget? All it takes is an API query for the current content of ((Vandalism information)), then a regex to see what it is, and then making a fluorescent notification DIV. You can even cache the result using a cookie so that it only asks the API every 'x' minutes. --slakr^\ talk / 02:24, 21 June 2012 (UTC)[reply]

You may be right, I've gotten back to working on this. I have added a few parameters to code and will start to find out if the community will support this.—cyberpower ^Chat_Online 20:46, 21 June 2012 (UTC)[reply]

Surely you should have got consensus before you requested approval? Rcsprinter (talk) 15:06, 22 June 2012 (UTC)[reply]

I am preparing an RfC at WP:CVU to attract primarily those involved in CVU to comment as it primarily concerns them. Other editors will also be invited to comment and support. It should be up by tomorrow.—cyberpower ^Chat_{Limited Access} 16:52, 3 July 2012 (UTC)[reply]

Sorry for the excruciatingly long delay. I have started up the RfC here.—cyberpower ^Chat_Online 00:32, 5 July 2012 (UTC)[reply]

Why is the source code unable to be released? --MZMcBride (talk) 17:47, 16 July 2012 (UTC)[reply]

At this point I can't release it yet as it's not fully complete yet with all the implementations. I will release it when I am certain is ready for testing on Wikipedia.—cyberpower ^Chat_Online 20:25, 16 July 2012 (UTC)[reply]

• Questions I have a few questions about this bot:

1. If the bot notifies a user while they are offline, does it provide them subsequent notifications when the level changes or does it stop after the first?

   The purpose of this bot is to notify them of elevated vandalism. This can be while they're offline or online. The goal being it triggers the orange bar and as a result sends an email notification about new messages. This will lure currently offline users into potentially vandal fighting. As per popular request, a schedule will be added to allow the user to set notification times. When vandalism levels change it will only notify them when the notification rate is set to 0.

2. What is the point of a notification when the user is offline?

   Talk page pings trigger email notifications. An offline user is more likely to see that.

3. How does the bot know when a user comes online and sees the message? Have you considered making some sort of status system a requirement for users who wish to use the bot?

   The bot will have a rate limiter and schedule service installed. User can then freely tell the bot when to notify them.

4. What will happen if someone breaks the formatting of your notifications list? Will the bot auto-shutoff or recognize the error?

   The bot will most likely auto-shutoff if the list can't be read. If it's and entry error, it will ignore the entry. If the parameters are not in the order specified though, the bot may screw up.

5. What will happen if a vandal adds 1000+ names randomly to the notifications list? What will prevent users from getting spammed who do not wish to use it? Would a better approach be to use a template on the user's userpage that puts them in a specific category for each level? ie ((vandal notes|2))?

   I have considered this and I believe a better approach will be to make it a template to optin with. Either way, the bot is ((bots)) compliant so if users get spammed, they can deny further notifications using that.

Thanks.--v/r - TP 14:35, 23 July 2012 (UTC)[reply]

I have answered the questions above. I am making the requested modifications and am letting you know that I haven't forgotten about this.—cyberpower ^Chat_Offline 14:40, 16 August 2012 (UTC)[reply]

((OperatorAssistanceNeeded|D)) – How are the modifications going? When they are complete, I, like Chris G, would like a copy of the source code. I understand you not wanting to post it publicly; I myself hate to release anything until it's "perfect". But given that this may be a contentious task, it would be desirable to iron out any technical problem before it is approved for trial, if it is approved for trial. Thanks, — madman 04:58, 25 August 2012 (UTC)[reply]

I got sucked into real life sooner than I thought I would. There is so much going on right now that I barely have time to come up with the mods. Rest assured though, I am working on it.—cyberpower ^Chat_Offline 12:00, 25 August 2012 (UTC)[reply]

• I think this should be marked as expired until Cyberpower is ready to continue with the development. It's just sitting here using space on the WP:BRFA page. Rcsprinter (constabulary) @ 16:47, 27 September 2012 (UTC)[reply]

  Agreed. I will reopen once its completed. Is it ok if I reopened this one once I'm ready?—cyberpower ^Chat_{Limited Access}(Now using HTML5) 17:07, 27 September 2012 (UTC)[reply]

  Of course, normal practice. Rcsprinter (natter) @ 17:08, 27 September 2012 (UTC)[reply]

  Per the above, Request Expired.. — madman 19:53, 27 September 2012 (UTC)[reply]

The above discussion is preserved as an archive of the debate. Please do not modify it. To request review of this BRFA, please start a new section at WT:BRFA.