Wikipedia:Bots/Requests for approval/DavidWSBot 1

The following discussion is an archived debate. Please do not modify it. Subsequent comments should be made in a new section. The result of the discussion was

Approved.

DavidWSBot 1

Operator: _David ^WS _(contribs)

Automatic or Manually Assisted: Automatic

Programming Language(s): PHP

Function Summary: Notifies opted-in sysops when a user they have blocked requests an unblock.

Edit period(s) (e.g. Continuous, daily, one time run): Continuous, a 1.5 minute break after each check.

Already has a bot flag (Y/N): N

Function Details: Finds all users in the category Requests for unblock and retrieves the sysop that blocked the user. If the sysop is opted-in then it notifies them using a template I created.

Discussion

What if the same user requests an unblock multiple times? MaxSem^{(Han shot first!)} 16:37, 26 November 2008 (UTC)[reply]

Good point you make there. I'll add some code so it doesn't notify more than once. Thanks for bringing that up :) _David ^WS _(contribs) 16:48, 26 November 2008 (UTC)[reply]

Okay, I added code to make sure it's not a duplicate notification. _David ^WS _(contribs) 16:52, 26 November 2008 (UTC)[reply]

The main concern I have is that this could be abusable. If its opt-in, it would remove most of the risk, but I can imagine situations like:

1. Troll creates account with offensive username

2. Admin blocks

3. Troll requests unblock, causing the bot to paste the offensive username on the admin's talk page.

Not a major abuse, but its possible. It may be worth it to parse the block reason and not actually show the name if its a username block, or have some regexes for "bad usernames."

Also, can the bot handle rangeblocks, autoblocks, and hardblocks of an IP affecting a registered user? And for the "duplicate notifications," is there a time limit so that requests that are really separate (long time between requests, admin blocks the same user twice, etc.) still do a notification for both. Mr.Z-man 17:21, 26 November 2008 (UTC)[reply]

I could regex match the username to make sure it's not offensive, and essentially just censor it on the admin page, or not show it at all. I don't have anything set up for rangeblocks, autoblocks, or hardblocks, although I will likely add that later. I'm going to take a look now, as that shouldn't be hard to implement. However, I'd rather wait and request that as a seperate task later. For the duplicate notifications, I can have it check how long ago the template was added to the page. Presently, the sysop will need to remove the template before it will notify again. _David ^WS _(contribs) 17:25, 26 November 2008 (UTC)[reply]

I'm going to write a PCRE to match offensive terms. Also, the autoblocks should be easy, it's just a separate category. I can have the bot load users listed there too. Again, however, I would rather wait until a test to make sure it works properly with normal blocks (as it should, I've tested it somewhat locally). I'll work on that regex. _David ^WS _(contribs) 17:29, 26 November 2008 (UTC)[reply]

I see no problems with offensive usernames. The blocking admin had already seen them while blocking, even without the bot they can receive messages from good-faith users with username mentioned. At most, we could tweak the message template to look like "the user you blocked is requesting an unblock [[User talk:$1|here]]". MaxSem^{(Han shot first!)} 17:33, 26 November 2008 (UTC)[reply]

Okay, then I'll forget the censorship thing. I do agree with you. I'll work on something that checks the timestamp of the template if it is already there, and if it is over two hours old, add a new one. _David ^WS _(contribs) 17:48, 26 November 2008 (UTC)[reply]

I have decided upon, along with the notifcation template, adding an HTML comment in the form , and then extract the timestamp with that. I have already written code to do so. If the difference between now and that timestamp is 21600 seconds (6 hours), then it will post the notification again. _David ^WS _(contribs) 18:10, 26 November 2008 (UTC)[reply]

As it is only an opt-in system and the code looks good in my opinion,

Approved for trial (30 edits). Please provide a link to the relevant contributions and/or diffs when the trial is complete. Please report back here with your results. I will opt-in myself as an additional user. — E ↗ 00:03, 6 December 2008 (UTC)[reply]

Thanks. I'll let you know. _David ^WS _(contribs) 00:21, 6 December 2008 (UTC)[reply]

It was rough at first but I worked out all the kinks. Still has trial edits left, but it now notifies right and the matching of timestamp html comments works well. I had to fix a lot of bugs though. _David ^WS _(contribs) 16:51, 6 December 2008 (UTC)[reply]

(indent) Very nice work David. Remember, fixing errors and improvements are what a bot operator does for a living. You seem to have ironed out all the errors and the latest edit looks excellent. Let me know when your trial edits are complete. — E ↗ 03:49, 7 December 2008 (UTC)[reply]

Well, technically, it is now over the trial limit due to some edits I made with the account a while back to give it autoconfirmed status, so I figured I'd turn it off and let you know. (I just added something so it notifies me of what it has done every twelve hours). Anyway, 31 edits, so, yeah. _David ^WS _(contribs) 18:58, 14 December 2008 (UTC)[reply]

Trial complete. _David ^WS _(contribs) 20:00, 14 December 2008 (UTC)[reply]

Approved. Edits look fine, you may need to fix the notifications to yourself (it added 5 over a period of 6 minutes, which is obviously incorrect). — E ↗ 05:11, 15 December 2008 (UTC)[reply]

The above discussion is preserved as an archive of the debate. Please do not modify it. Subsequent comments should be made in a new section.