User talk:ClueBot Commons/Archives/2016/January

This is an archive of past discussions with User:ClueBot Commons. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.

Stop your ClueBot NG. The article solved is with reference now. Greets — Preceding unsigned comment added by 91.59.24.188 (talk • contribs)

ClueBot NG hasn't reverted anything because it doesn't have references, that's not the way that the robot works. The robot works on probability and there is something in your edit that has set it off as being possible vandalism. In all fairness, if I came across that edit, as I know nothing about the topic of the article, I would have reverted it as well. If it is not vandalism, then please follow the instructions at the top of this talk page to report it as a false positive. I would then suggest that you make a request on the talk page of the article for your edit to be added by an established editor.--5 albert square (talk) 22:10, 1 January 2016 (UTC)

Thats the way to chase authors. I say goodby to wikipedia. So long, and thanks for all the fish. — Preceding unsigned comment added by 91.59.24.188 (talk) 04:04, 2 January 2016 (UTC)

Goodbye then, when someone was quite nicely trying to help you. —k6ka 🍁 (Talk · Contributions) 22:01, 2 January 2016 (UTC)

So the archiving at Talk:List of multi-level marketing companies got screwed up because a template wasn't being subst'd properly. This, and a little cluetardiness on my part, led to a bunch of stuff being archived into unhelpful places. So I created archives for years from 2007 to 2014, and moved the Archive/2015/August to Archive/2015. My question is: will I need to edit the Master Detailed Indices by hand, or will ClueBot find my revised archive pages and redo them by itself? I'm willing to do the handwork if I have to, but I don't want to do something which will get done in time without me. Argyriou (talk) 17:15, 29 December 2015 (UTC)

@Cobi:--5 albert square (talk) 15:25, 3 January 2016 (UTC)

{{atop}} I see you have reverted my changes. I had made an honest mistake with the year of release of "Valentine's Day is a Lie" but my addition to the table was NOT vandalism and I even included a reference. Seelamviraj (talk) 00:10, 3 January 2016 (UTC)

Please read the notices at the top of this page. General Ization ^Talk 00:14, 3 January 2016 (UTC)

{{atop}} Cluebot III archived Wikipedia:WikiProject Resource Exchange/Resource Request less than two hours after marked as resolved. For other users references, It should be removed at least 24 hours ago.―― Phoenix7777 (talk) 11:18, 4 January 2016 (UTC)

If you check the configuration on the page, {{resolved}} is listed as an "archivenow" keyword. So the bot archived it as soon as it saw it. -- Cobi^(t|c|b) 12:03, 4 January 2016 (UTC)

{{atop}} Couldn't find this in the documentation or FAQ list; apologies if I missed it. Does Cluebot NG operate on all pages, or just article space? If there's a page it doesn't watch (such as one in the Wikipedia space), is there a category to add the page to or some other way to request the bot's attention? (My curiousity is specifically about the possibility of having Cluebot attend to the Wikipedia:Reference desks.) —Steve Summit (talk) 10:31, 6 January 2016 (UTC)

@Scs: No, ClueBot NG doesn't look at pages outside the article namespace. The old ClueBot did this, but as ClueBot NG uses an artificial neural network and not a list of simple heuristics, having it operate in other namespaces would require an entirely different database of edits (Which it doesn't have). I would support bringing the old ClueBot back though, just so it can monitor those other pages. (And despite popular belief, adding a page to User:ClueBot NG/AngryOptin does not get CBNG to monitor it if it isn't in the article namespace) —k6ka 🍁 (Talk · Contributions) 11:53, 6 January 2016 (UTC)

@K6ka:@Scs: Hi K6ka and scs, when I add pages to AngryOptin, when in page edit mode (Editing User:ClueBot NG/AngryOptin), the page starts off with (quotes):

"--

 -- This is the ClueBot Angry Optin list."

(...) "-- For main namespace articles:

 -- * [ [Pagename]] - reason. 
 --
 -- For pages in other namespaces:
 -- * [ [Namespace:Pagename]] - reason. 
 --
 -- Please note that pages outside of the main namespace will *also* need to be listed at User:ClueBot/Optin."

So even if I do that, it still will not work for pages outside Article space? Silly thing here is, the text refers to [ [User:Cluebot/Optin]], not to [ [User:Cluebot NG/Optin]] what might make a difference? I'm confused now, as both Cluebot and Cluebot NG have this Optin page? Poepkop (talk) 14:04, 6 January 2016 (UTC).

@Poepkop: Even when listed on both pages, ClueBot NG still doesn't check or revert edits made outside of the main namespace. That text is either outdated or misleading. (By the way, you can "escape" wikimarkup by surrounding the text with <nowiki></nowiki> tags. —k6ka 🍁 (Talk · Contributions) 15:00, 6 January 2016 (UTC)

@K6ka: Okay, I see, yes, misleading by its possible outdatedness. That is a pity, actually. Thanks for the nowiki, I had this problem before elsewhere, so that will solve it, good to know. Poepkop (talk) 15:27, 6 January 2016 (UTC).

Let me make a few things clearer:

• ClueBot NG does check if it is on the Optin list if it is outside the main namespace
• Unless you actually know what you are doing, please, PLEASE don't add pages outside the Article space to the Optin list: The bot will not work correctly on pages that do not *look* like articles. For example, User:Cobi/Cobi looks enough like an article (but is still a stub), but User:Cobi does not at all, nor does any page that has discussion on it.
• AngryOptin will cause the bot to ignore standard anti-edit-warring code to edit-war with other editors on that page. Be very careful what you list there.
• If you need to list a non-article on AngryOptin, it won't make any difference unless you also list it on Optin. Even then, this is even more of a weird case.
• The proper links to the relevant pages are: User:ClueBot NG/Optin and User:ClueBot NG/AngryOptin.

I hope this clears things up. -- Cobi^(t|c|b) 16:56, 6 January 2016 (UTC)

A very simplistic way to look at it is the bot checks each diff that it gets and tries to see if the new version looks more like an article than the last version. Since things that don't look like articles will fail this test most of the time, the bot will not function correctly. So the key to adding anything to Optin is to make sure that the articles there look like articles. Most of the current things on there do not, and should be removed.

As for the AngryOptin, there are very few cases where this should be needed for articles, and I really cannot see a good use for it for non-articles. -- Cobi^(t|c|b) 17:02, 6 January 2016 (UTC)

Thanks Cobi, that clarifies a surprising lot about those Optin pages. I think it's worth putting a warning there letting editors know about this. —k6ka 🍁 (Talk · Contributions) 17:42, 6 January 2016 (UTC)

There is quite a large and severe edit warning above the edit box for the Optin page. -- Cobi^(t|c|b) 21:23, 6 January 2016 (UTC)

I notice that ClueBot III calculates the key for archiving to something other than a subpage using MD5. Lowercase sigmabot III also did this until recently. However, MD5 has known security flaws. Earlier this month, Σ, as part of another change, updated Lowercase sigmabot III to use the much stronger SHA256 instead of MD5. Can the same be done here when the maintainers get a chance? jcgoble3 (talk) 01:27, 14 January 2016 (UTC)

Be aware that he will escalate this critical security issue to BAG as needed. →Σσς. (Sigma) 01:55, 14 January 2016 (UTC)

Why are you speaking for him? SQL^{Query me!} 01:59, 14 January 2016 (UTC)

Simply a heads up based on prior observations, as he previously approached me with the same request. Would communication not be accelerated, were the full conditions laid out at the start? →Σσς. (Sigma) 02:04, 14 January 2016 (UTC)

Just looks strange is all I guess. Up to sigma already, eh? SQL^{Query me!} 02:19, 14 January 2016 (UTC)

I'll get to it at some point. However, it is primarily in place to prevent against accidental misconfiguration targetting other pages. As this is rarely a useful thing to do, it supports keys to allow it to do so. Collision attacks against MD5 are getting much easier, but the effort needed, combined with the limited size of user-provided input (the source page name) to target a specific victim page, along with the fact that the bot reports what page it is archiving from, and the revision history, allow for very easy reversion of such a vandalistic edit, followed by blocking. Furthermore, since the source page would likely be a senseless page title, if the source page were deleted and creation of that page were blocked, then they would have to generate a new collision.

It just seems like a lot of effort to go through to edit a page by proxy that is at most semi-protected, for very little gain, and is very easy for any user to fix, and any admin to block the user abusing the bot. But, yes, like I said, I will get to it, but I don't consider it anything more than a low security risk, especially since pre-image attacks are still very difficult against MD5. -- Cobi^(t|c|b) 02:55, 14 January 2016 (UTC)

Glad to see you're still around, Cobi! I agree that it really isn't a critical security issue any more than "the encyclopedia anyone can edit" is. SQL^{Query me!} 02:59, 14 January 2016 (UTC)

@SQL: I agree entirely. I did my best to convey in my own addition the urgency that was present in the request that I received. Whether it is more correctly read as sarcasm is not something I consciously considered in the writing of that.

Best, →Σσς. (Sigma) 03:08, 14 January 2016 (UTC)

To clarify Sigma's comments, I recently approached him privately regarding a more serious issue with his archive bot. That specific issue is not present in ClueBot III. The issue of upgrading from MD5 was a secondary issue that I suggested he change at the same time since he would have to generate new keys anyway. I do not see MD5 an urgent matter. BAG was only brought in on Sigma's bot because of the first, more serious bug; the matter of MD5 would not be worth raising the matter with BAG. It's simply a suggestion, not something serious enough for me to actively pursue like the other issue. jcgoble3 (talk) 03:51, 14 January 2016 (UTC)

The current available data shows that the number of sectarian deaths of Shias is approximately 250 a year in a country of 180 million people as compared to most other developed nations such as the UK which have approximately 52,528 hate crimes a year shows that it is not a rampant problem. Most of the hate-crimes in the UK, however are not homicide and most crimes are race related and not religious or anti-semitic hate crimes, in the UK there were over 600 incidents of violence related hate crimes and relativised to Pakistan's populous it would be 1800 incidents. However, most of these incidents did not include homicide. [1] Recently there has been a major decrease in attacks on Shias, it has decreased in Baluchistan and over the whole country as a result of National Action Plan and Zarb e Azb. http://tribune.com.pk/story/986698/good-performance-balochistan-witnesses-50-decline-in-sectarian-violence/ Also there has been a major decrease of sectarian violence all over Pakistan by approximately 50% since 2010 to 2015 http://www.satp.org/satporgtp/countries/pakistan/database/sect-killing.htm the marked increase in 2015 was due to the blowback of targeting militants. Most major or high profile sectarian incidents are condemned directly by the Government itself such as the violence against a Christian couple was directly condemned by the Prime Minister, the attacks on Agha Khani Shias was also directly condemned by the Prime Minister. [well known, citations can be found easily] Furthermore the perpetrators who attacked the Christian couple were arrested and put on trial. [citations can be found easily] Developments to curb sectarian violence has been taken actively by the current Government and empirical data shows a marked decrease in sectarian violence in Pakistan as well. [citations are available online] Politically Pakistan has tried to remain neutral and pacify both Saudi Arabia and Iran [citations are available online] which is also what the public opinion of the country holds [Dawn Poll] as over 80% of Pakistanis in an online poll want to remain neutral. [the poll is available at Dawn.com and then view results, I also have a screenshot] Another issue is that during the reign of the Zardari regime who was Shia, there was in fact a marked increase in sectarian violence towards Shias by approximately 60% from 2008 to 2013 http://www.satp.org/satporgtp/countries/pakistan/database/sect-killing.htm, whilst during the reign of the Nawaz regime who is Sunni, there has been a marked decrease of 49.5% http://www.satp.org/satporgtp/countries/pakistan/database/sect-killing.htm

Pakistan is also the only Sunni-majority country that has elected a Shia for Prime Minister or had a Shia Head of State. Not only that it has had over three different Shia Heads of Governments or Head of State and elected a Shia-led party to power over 4 times and that too included a Shia woman, coincidently the first muslim female Prime Minister in the whole world. [Fact, you can find citations anywhere] No other Sunni-majority country in the world has done this so far.

Whatever ID is required, go check it, seriously. — Preceding unsigned comment added by 144.32.240.10 (talk • contribs)

Hi IP,

Please read the boxes at the top of this page. ClueBot NG is not a human, it is a robot, therefore it cannot check your edit or any of the links provided.

The robot has not reverted you because of the current data available, that is not the way it works. It works on probability and there is something in your edit that has triggered the robot and made it revert this as possible vandalism. For example if you have used words such as "hate", "violence" etc then that may trigger it as they are commonly used vandal phrases. However that does not mean that your edit is vandalism and from what I can see it is not. I would suggest that you post on the talk page of the article and suggest your changes to the article there. If the information provided belongs in the article, an established editor can then make the edit.

I hope this explains things.--5 albert square (talk) 00:43, 19 January 2016 (UTC)

When I first started on Wiki I did a lot of counter vandalism, and was very familiar with the tiered notice system. However, now that I do a lot of work over at the Account Creation team and the helpdesk, I see people who are well intentioned, but sometimes misguided new users join the project. Sometimes their first encounter with wikipedia is ClueBot. Sometimes its a false positive, but more often than not, when ClueBot does hit them, it is technically correct, however the automated Level-1 warnings can be harsh for well-intention first time users - scaring them off from contributing. They were effectively bitten. My proposal would be for ClueBot to be adjusted to determine if this editor just started editing within the last 24 hours and has no current warnings, use a softer welcome/warning, and then hold notices for perhaps 30-60 minutes before starting in with the standard warning escalations. Sometimes all it takes is that first template to kick start a 'CV' race against a well intention editor by well intentioned CV user. Sometimes it takes new users some time to notice and understand talk page warnings. But end up in an edit war with a CV or BOT and end up scared off. However if the editor has made prior contributions (more than 24 hours ago) or has exiting warnings, then operate as normal. I think by adding this bit of 'new user' buffer will help keep some new contributors around. Let real people handle some of the CV work for these new users, and let the bot handle more egregious and persistant vandals. Tiggerjay (talk) 18:52, 19 January 2016 (UTC)

{{sofixit|Special:Prefixindex/User:ClueBot NG/Warnings/V}} -- Cobi^(t|c|b) 00:26, 20 January 2016 (UTC)

What Tiggerjay is suggesting here isn't a simple template change. Tiggerjay is suggesting that the bot employ different tactics and use different messages depending on how new a user is. That's not something that can be changed by anyone except the bot maintainers, because it would require new code. (For the record, I have no opinion on the proposed change; I'm just pointing out why it isn't a "sofixit" case.) jcgoble3 (talk) 02:33, 20 January 2016 (UTC)

Thanks jcgoble3. I was scratching my head over the sofixit comment, yes I am proposing primarily a code change. I have slightly revised the section title. Tiggerjay (talk) 02:38, 20 January 2016 (UTC)

Ah. Is there a specific reason why we can't just make level 1 softer in general? It probably wouldn't be too hard to select a different warning depending on edit count or something like that, but I suppose the main question is, why not just add the softer language to CBNG's level 1? -- Cobi^(t|c|b) 04:17, 20 January 2016 (UTC)

Human vandal fighters scare newbies off with their warnings too. It's not uncommon to see a new user add some unsourced information to a WP:BLP, reinsert it when other edits revert it, and don't see the warnings on their talk page until it is too late. It has earned me some personal attacks from otherwise well-intentioned editors, saying how I'm a pretentious asshole for threatening to block them and how they were going to "tell on me". —k6ka 🍁 (Talk · Contributions) 18:23, 20 January 2016 (UTC)

See User talk:Jimbo Wales/Archive 201#Auto-archiving on this page broken? – Wbm1058 (talk) 21:12, 21 January 2016 (UTC)

Also broken on my talk page... ~ Matthewrbowker ^{Drop me a note} 21:21, 21 January 2016 (UTC)

Original discussion has been archived, and resumes on User talk:Jimbo Wales. Wbm1058 (talk) 17:02, 25 January 2016 (UTC)

@Wbm1058: @Matthewrbowker: I've had an email from Rich, Cobi is waiting on Labs access, then he can go through the bot code.--5 albert square (talk) 17:52, 31 January 2016 (UTC)

I've modified Template:BLP to remove the direct wikilink to Wikipedia:Biographies of living persons/Noticeboard. Hoping that fixes this particular case. It may take several days (or weeks?) to work all of the transclusions (currently 772,900) through the job queue. However the bot should probably be modified to skip over over-linked things like this that make for unacceptably inefficient processing. Wbm1058 (talk) 18:08, 31 January 2016 (UTC)

@5 albert square: Thank you for the update. ~ Matthewrbowker ^{Drop me a note} 21:41, 31 January 2016 (UTC)

The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

---

This edit request to User:ClueBot III has been answered. Set the |answered= parameter to no to reactivate your request.

Please add the following warning to the top of the User:ClueBot III page:

Cluebot III is currently (11:09, 27 January 2016 (UTC)) underperforming when confronted with large collections of archive pages for the same talk page (around 200 archive pages for the same talk page seems to be a maximum that should not be exceeded). The issue is discussed at User talk:Jimbo Wales#Conspiracy Unveiled. Please don't remove this warning before all related issues are resolved.

--Francis Schonken (talk) 11:08, 27 January 2016 (UTC)

 Not done The bot operator @Cobi: or their designee can do this; reminder that no bots are obligated to even be running. — xaosflux ^Talk 21:29, 27 January 2016 (UTC)

The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

Well as it may be, that it is a coincidence? I clicked on the article, and clicked cancel because it was vandalism, and already wrote that the change has been canceled. Then shook post warning, just turned away, already written :D. This is not the issue here observation.--Lukaslt13 --Talk 17:09, 27 January 2016 (UTC)Lukaslt13

@Lukaslt13: I'm confused by your query. I cannot see that this bot has had any interaction with you.--5 albert square (talk) 23:55, 27 January 2016 (UTC)

Is it just me or is review.cluebot.org/ actually being redirected to this? Seems like someone hijacked the domain for promotional purposes. -- ChamithN (talk) 14:20, 30 January 2016 (UTC)

Thanks, I have emailed Cobi and Rich :)--5 albert square (talk) 19:49, 30 January 2016 (UTC)

Rich has emailed me, looks like cluebot.org expired and has now been registered by another company.--5 albert square (talk) 16:43, 31 January 2016 (UTC)

If the bot owners don't plan to restore it, I think it's best to remove the link from User:ClueBot NG/FalsePositives, rather than keeping it tagged as dead. -- ChamithN (talk) 18:46, 31 January 2016 (UTC)

Error, mr. Bot — Preceding unsigned comment added by Simon Frazer (talk • contribs) 02:38, 31 January 2016 (UTC)

Indeed an error. I reported it to the false positive interface. jcgoble3 (talk) 03:47, 31 January 2016 (UTC)

Just a question. Why is ClueBot III down and when will it be running again? I was wondering why some of my talk page posts weren't being archived. Class455fan1 (talk) 19:23, 31 January 2016 (UTC)

I think it's a coding issue. I had an email from @Rich Smith: earlier to say that @Cobi: is awaiting Labs access so he can go through the code.--5 albert square (talk) 20:02, 31 January 2016 (UTC)