Archive 1

←

Archive 5

Archive 6

Archive 7

Archive 8

Archive 9

Archive 10

Anyone else seeing a significant decrease in positive vandalism detections from ClueBot NG (I just mainly use ClueBot, but the others are included as well) in the past day or so? Think there's a cause for this? Theopolisme _TALK 18:20, 28 June 2012 (UTC)

Yep. I noticed it too.

And its not just us... Check out the leaderboard. Morning277 started 3 days ago and has had a hit rate of 4%! (plus AGF-revert rate of 7%). On the other hand, Jmajeremy has had a hit rate of 43% in 1 day.

The toolserver reverts graph looks pretty unchanged. I would expect it to go massively down if there was a problem with the algorithm or with someone doing lots of false "innocent" classifications. (Ignore the drop at the end, that is for today, there are fewer reverts because the day isn't done).

My current theory is that we just have periods when there is a lot of STiki activity and the hit rate is low during these periods, especially once the easy-to-detect vandalism has been reverted. This may actually be in large part down to Morning277, who has managed 2,040 classifications in 3 days... More classifications than the number of reverts by all users in that time! (1,339 according to the graph)

I don't really know how Morning achieved this... it is roughly 1 every 2 minutes... but presumably Morning wasn't using STiki 24 hours a day... so Morning must be working very fast... and presumably Jmajeremy used STiki during a rare period while Morning was asleep!

Yaris678 (talk) 07:18, 29 June 2012 (UTC)

Teehee! Ah, those young'uns.. :D Theopolisme _TALK 15:41, 29 June 2012 (UTC)

That's how we doooz it! The only way to describe it is "Kid In A Candy Store." After using Huggle for so long, STiki is so easy for me. Sorry if it is causing others grief. --Morning277 (talk) 17:20, 29 June 2012 (UTC)

I generally agree with all of the above commentary. Shortly I will work on the "queue exhaustion" indicator functionality. This way, at least, no one wastes their time classifying edits if one or more editors have recently done a humongous amount of work. The fact that STiki is searching for a finite quantity of vandalism instances has interesting ramifications. The more popular STiki becomes, the less useful it appears -- and vice versa! Thanks, West.andrew.g (talk) 13:45, 30 June 2012 (UTC)

We're currently at wikidefcon level 4. benzband (talk) 14:54, 30 June 2012 (UTC)

Hmmm... I would like to retract some of my above analysis because it appears that the reverts graph hasn't been updated since early on the 22nd of June. I should have noticed that one!

The conclusions my still be correct but I would be interested to see the graph when it does get updated. It may show that STiki is now doing a lot more reverts that Huggle. The Wikidefcon thingy is based on Huggle reverts... So the hit rate in STiki may actually be a more accurate reflection of current vandalism levels... although the revert rate in Huggle is still useful as it is independent of the STiki software.

Yaris678 (talk) 20:44, 1 July 2012 (UTC)

@Morning227, It's definitely not causing me grief. I am pleased that STiki is getting so much use... and more importantly that the encylopedia is being protected while still being open to anyone to edit. I guess if I have any concern at all it is over how much attention you can give each diff if you are working so quickly. For example, what would you do if you saw someone adding a "fact" that may or may not be true/verifiable?

Yaris678 (talk) 21:06, 1 July 2012 (UTC)

This is the easiest way that I can explain it.....If you are familiar with Huggle, you know that there are many features available that are not in STiki. In fact, Huggle can do at least 10 times more than STiki can. It is pre-loaded with template messages for warning editors, welcoming newbies, and any other type of notification that you want to give. It also has numerous filters so you can narrow down just about any type of edit that you want to look for (I will get into the reason why I mention this a little later). For vandalism, Huggle finds "new" edits that are being patrolled by just about everyone who is looking for vandalism. As such, vandalism is usually caught immediately, the editor warned, and the vandalism reverted. With so many people looking for "new" vandalism, you have to be quick in order to make an edit or you will be wasting your time (as you will get beat to the edit if you do not act quickly). I have used Huggle so much that I am very quick with identifying vandalism and edits that need to be reverted. Now, the reason why I mention that Huggle can do more than STiki is because although you can do more globally with Huggle, STiki is THE BEST tool to use to fight vandalism as the tool already has filtered what would take you a while to do in Huggle. In addition, it finds vandalism that slips through the cracks. Basically, it is the part of Huggle that I use on a regular basis without having to worry about all of the other Huggle features. It is simple which is what is good. You do not have to use it for weeks or go through extensive training to be able to understand all of the features. Now with Huggle, if you do not recognize something immediately as vandalism, it is best to just skip the edit and move on so that you can get to actual vandalism that you recognize. With STiki, you can take your time as the identified vandalism is already hours old and no one else is paying attention. You can take the time to determine if it is vandalism or not. With that in mind, I am finding more "good faith edit" reverts than I do vandalism with STiki. And, while I am able to take my time, I pass on anything that I think is borderline or a topic that I am unfamiliar with and just let anything that does not seem like vandalism, spam, or bad editing to go as innocent. Based on my experience with Huggle (which will definitely train your eyes to recognize spam, vandalism, and test edits), I am able to navigate through STiki rather quickly. As time goes on, I feel like it will slow down a little, but for now I am enjoying being able to move through so quickly. I know this is a long story, but it is the best way that I can explain it. --Morning277 (talk) 15:44, 6 July 2012 (UTC)

Thank you for your story. I found it very interesting. I was also pleased to see how you appreciate the differences with STiki. Keep up the good work! Yaris678 (talk) 18:21, 6 July 2012 (UTC)

Just had a random thought. If you like filtering edits in Huggle, you may be interested in Wikipedia talk:STiki/Feature development. This lays out the things that the STiki algorithm uses to detect vandalism. Yaris678 (talk) 20:29, 8 July 2012 (UTC)

@West.andrew.g, This makes me wonder if it may be worth adding additional queues. I often check articles a few days after I have edited them with STiki and often come across vandalism that has not been caught by STiki or ClueBot NG. I think it would be worth having a "revist queue" that looks for edits to articles (or by editors) that have previously had reverts through STiki. It could rank the edits according to the previous STiki hit rate for that editor or article. If this excluded edits by established editors I bet it would have a decent hit rate.

Yaris678 (talk) 21:06, 1 July 2012 (UTC)

Hi! I am a CVUA instructor and use STiki myself, and I was wondering if you guys could enable STiki on my student, User: Sesamevoila's account. Thank you, Electriccatfish2 (talk) 20:21, 4 July 2012 (UTC).

 Done West.andrew.g (talk) 20:37, 4 July 2012 (UTC)

Thanks! Electriccatfish2 (talk) 20:39, 4 July 2012 (UTC)

Thanks a lot Sesamevoila (talk) 17:08, 5 July 2012 (UTC)

Hello! You have an awesome job of creating and improving STiki. While at it, I was wondering whether there is a way the s/ware can be customised so that a CVU member can revert edits arising from Vandalism from editors in a certain country or IP block.. Or is this a feature that needs some time to prepare and code?

Thank you. ^SUwanja _{Talk to Me. Email Me.} 17:27, 8 July 2012 (UTC)

I imagine that is theoretically possible for IP edits, although it would take effort to code.

Why are you interested in such a feature?

Yaris678 (talk) 11:32, 12 July 2012 (UTC)

My interest in this feature was to counter malicious edits from Kenya.Around here in Kenya, people and more so the academics do not consider Wikipedia a credible source of information and some even go further to vandalize Wikipedia just to prove that malicious edits stay live at Wikipedia for quite sometime.The global South is growing in terms of internet penetration and Wikipedia's reliability is in question which is apparently worrying to me. I know Wikipedia is not yet complete but it is a great Work in progress and probably the largest of all WIP projetcs, in fact we are 100M articles to get anywhere close to the sum of all human knowledge. See [1]

1. http://en.wikipedia.org/wiki/User:Emijrp/All_human_knowledge Uwanja _{Talk to Me. Email Me.} 03:51, 13 July 2012 (UTC)

Greetings. While I respect the note of stephenwanjau the fact nonetheless remains that malicious behavior is more frequent in certain IP ranges, whether they map to certain geographical ranges, educational institutions, or otherwise. The STiki tool does not explicitly allow filtering by certain country or IP range. Indeed, this is performed *implicitly* as bad regions/blocks are more likely to have their edits appear atop the "STiki (metadata)" quueue.

However, should you have reason to investigate a particular IP block (there are plenty of legitimate reasons), I encourage you to investigate my WikiAudit tool. Although less sexy and user-friendly than STiki, it can be used to summarize recent (per a timestamp argument) contributions from a particular IP range. This could indeed be an effective means of vandalism patrol and I'd be interested to hear your observations of using it in such a matter. Thanks, West.andrew.g (talk) 04:58, 13 July 2012 (UTC)

This happened today. I was asked to judge my own revert by STiki. It was quite hillarious, but shouldn't this be avoided? Thanks! :D Anir1uph (talk) 09:48, 12 July 2012 (UTC)

I know of similar incidents in the past. Two questions:

1. What queue were you on?
2. Did it look like the queue was suffering from queue exhaustion? i.e. was there a very low hit rate?

Of course, it makes no sense to ask you to check your own edit... It's undertandable that your edits will be checked when we have queue exhaustion, but logically the server should send that edit to a different user. Of course, it is quite a rare occurrence so I guess it just depends on how easy it would be for Andrew to implement such a feature.

Yaris678 (talk) 11:30, 12 July 2012 (UTC)

Happened to me once last week, too. -- Gareth Griffith-Jones (talk) 11:36, 12 July 2012 (UTC)

1. it generally happens when i Change the queue. so a revert made by me in Queue1 is later shown to me when i am checking Queue2.
2. I am not sure but this happens for reverts having substantial content removal or content addition.--DBigXray 11:40, 12 July 2012 (UTC)

I was constantly using CLue-Bot NG. And there didnt seem to be any queue exhaustion, as i was getting a good revert rate. I didnt change the queue nor was the revert substantial. Come to think of it, i should have classified it as vandalism, not as a good faith edit. :) Anir1uph (talk) 11:47, 12 July 2012 (UTC)

... and I was constantly using the alternative STiki queue, and was not experiencing any exhaustion either. -- Gareth Griffith-Jones (talk) 13:15, 12 July 2012 (UTC)

Update I must add that a few times i am presented my own reverts and a lot of times i am presented with reverts made by User:Gareth Griffith-Jones using STiki. so what it appears is the diffs are presented irrespective of the person who has made it if the edits qualify in the queue algorithm. --DBigXray 13:34, 12 July 2012 (UTC)

Yes, and vice versa – on going to The history I have over the past few weeks quite often found your name as the previous editor. I take it that on those occasions, you are using the Clue-Bot NG queue, whilst I always use the STiki queue. -- Gareth Griffith-Jones (talk) 15:28, 12 July 2012 (UTC)

What do you mean by The history? Do you have a record of the edits you have looked at? Or a you looking at contributions to the wiki? Are you saying you have reverted DBigXray several times? Yaris678 (talk) 15:44, 12 July 2012 (UTC)

Not at all. Neither the first, nor the second, nor the third. The history appertaining to the article being judged. -- Gareth Griffith-Jones (talk) 16:53, 12 July 2012 (UTC)

I get it. You have, over the last few weeks, been servered edits and then gone to the history of the article to find that the edit was by DBigXray. You do know that you can see the name of the user in the STiki GUI, right? (of course, it is may still be worth checking the history for vandalism a few edits down... is it just that you always do that so there is no need to look at the user's name?). Yaris678 (talk) 17:57, 12 July 2012 (UTC)

Correct. Also, after identifing the nature of the revision, and clicking on, for example, [Vandalism (Und..] and the red message "unknown error check page history" has appeared in the LAST REVERT box. I have, this evening had one – Indian cuisine – and again the reason was that DBigXray had already reverted the vandalism.

-- Gareth Griffith-Jones (talk) 19:27, 12 July 2012 (UTC)

I'm getting similar messages as well. I remember a discussion from a while ago... is work underway for queue exhaustion notification? Theopolisme _TALK 19:39, 12 July 2012 (UTC)

• I too have been asked to judge on Gareth Griffith-Jones's STiki reverts, and that is useless as he invariably makes correct reverts. :D
• I am often presented with the red message Unknown error check page history. In all such cases, i go to the page and manually revert the vandalism. And in none of those cases, has the vandalism already been reverted. For cases when the vandalism has already been reverted, i see a separate red message: Beaten to a revert? (or something close to that). Anir1uph (talk) 04:25, 13 July 2012 (UTC)

Regarding the latter of these two comments, this is not the first time I've heard of this (check the archives). Can you deduce any pattern when this occurs? Is it a new page? Does the page history have 2+ editors? I tried to troubleshoot this with a user in the past using debugging code and we weren't able to deduce anything on the client side. That is, errors can also happen outside the control of STiki; think about cases of massive replication lag or when the database is failing altogether during software upgrades. I am not trying to push blame elsewhere, but I encourage users who are experiencing such problems to launch STiki from the terminal/command-prompt and submit to me any output that occurs in fail cases. Thanks, West.andrew.g (talk) 04:41, 13 July 2012 (UTC)

So, far i have found no such pattern of its occurrence. I think, maybe it happens when (as u say) database fails, or when the user's internet connection speed fluctuates.

(edit conflict) Hello all. First off, if any of you are attending Wikimania in Washington D.C. right now; I am there as well and would love to talk to some STiki users (send me an email directly to expedite that process). Regarding the problem described above, all I can assure is that a user using the "STiki (metadata)" queue should not see their own reverts or those of another user using that same queue. In general, usage of the "metadata" queue should reduce but not entirely eliminate such instances. I prefer to filter third party queues (CBNG, WikiTrust) as little as possible, which occasionally gives rise to strange behavior. However, we should not discount the possibility that a STiki user is abusing the tool and some queue is detecting that fact (albeit unlikely given our new user reqs.).

To draw attention to earlier discussion, my focus moving forward is on developing an optimized anti-vandalism algorithm that comprehensively outperforms these problematic issues, rather than addressing such minor issues in a programmatic fashion. Its not my goal to incrementally correct the errors of CBNG et al.

Although you should feel free to notify the maintainers of those algorithms in a polite way. Such logic errors might be leading to false positives in the autonomous operation of such tools (speaking about CBNG in particular).

I will report that the "queue exhaustion" functionality for the STiki GUI has been implemented locally and is currently being tested on my own machine. Along those lines, I note that queue exhaustion has been at an *extreme high* recently. I don't know whether to interpret it is a positive or negative, but user(s) have been going through 5000+ classifications daily. This has been reducing revert rates to ~20% (whereas at 1k-2k daily classifications, it would be closer to 50%). No doubt, this may be bringing odder cases to the attention of STIki users and frustrating the user base. Thanks, West.andrew.g (talk) 04:31, 13 July 2012 (UTC)

I am eagerly waiting for the new release! Anir1uph (talk) 01:48, 14 July 2012 (UTC)

While everyone is already aware and this is probably unneeded repetition, just putting this out there. Theopolisme _TALK 15:37, 13 July 2012 (UTC)

The fact some queues allow this to happen is known. Actually, I'm more interested in the way the tool rendered visually in that screenshot. I understand your resolution is probably large and the text size is small; but its quite different from the screenshots on the WP:STiki main page (how it looks on my own machine). In particular, the "classification panel" looks out of whack as the back button is large and the classification buttons seem to overrun the sidebar. Anyone else have this? On your own machines, does STiki appear more similar to the screenshot of Theopolisme or those on the main project page? Thanks, West.andrew.g (talk) 03:46, 14 July 2012 (UTC)

Mine looks like the one on the STiki main page. Perhaps Theopolisme is running it on a Mac? (I am using windows 7, Java version 7, update 5). Anir1uph (talk) 04:14, 14 July 2012 (UTC)

yeah User:Theopolisme seems to be running it on a Mac. He can increase the text size from the Options--DBigXray 09:04, 14 July 2012 (UTC)

I am using Windows XP, Java version 7, update 5, Firefox 13.0.1 (and always the STiki queue) – and STiki looks like the one on the STiki main page. -- Gareth Griffith-Jones (talk) 08:01, 14 July 2012 (UTC)

Yep, I'm on a Mac running OS X Lion + whichever version of Java is the latest. I'd actually never looked at the screengrabs before - but yep, this is my STiki little home. ;0 Theopolisme _TALK 13:28, 14 July 2012 (UTC)

I'm sick of seeing the same Anti-Vandalism barnstar even for 5000 or 10000 reverts.Hereby I present this rather inperfect version which I think can be made better collectively.Regards ≫TheStrike Σagle≪ 14:01, 16 July 2012 (UTC) With Barnstar template

		The STiki Barnstar of Merit
		Put your message here. ≫TheStrike Σagle≪ 14:27, 16 July 2012 (UTC)

• That is great. I liked it a lot and appreciate SE's efforts. — TheSpecialUser ^(TSU) 14:33, 16 July 2012 (UTC)
• Indeed, I also like the design. With some users ascending the ranks rather quickly as of late, it does look a bit redundant when their talk pages become filled with 2 or 3 nearly identical messages from the STiki team. I am fine with re-using the same message, though. Someone should codify this in a page similar to WP:STiki/milestone_template (perhaps WP:STiki/milestone_template_alt) and likewise update Wikipedia:STiki/milestones once we develop a distribution policy (i.e., which barnstar, when?). Thanks for your efforts SE. West.andrew.g (talk) 14:53, 16 July 2012 (UTC)
• (edit conflict) Theopolisme likes this. But when you said barnstar, do you mean something you put onto someone's talk or something you display along with userboxes and such? Because if it's a userbox (kinda like the Editor things... you know, the registered, experienced, journeyman, what have you), then would it make sense to have stripes or something denoting how many reverts have been made w/ STiki? I.E., no stripes=1 revert, 1 stripe=1000, 2=2000, 3=5000, 4=10000? Just an idea. Theopolisme _TALK 14:54, 16 July 2012 (UTC)
  • Fine ideas here (I like the bronze = 1k, silver = 5k, gold = 10k). We should keep the welcome template for the first use; as it contains different information than the "awards". Question, what happens for milestones >10k? Maybe at this point we revert to using the "anti-vandalism + STiki recognition" template? Considering this probably won't happen too often, and if someone does collect 25k, 50k, 100k rewards than the barnstars are unlikely to be in close proximity on the individual's user page. Thanks, West.andrew.g (talk) 15:23, 16 July 2012 (UTC)

This may sound ridiculous but maybe we can give them the same Gold BOM with MASTERED watermarked in it.I tried in my computer and it seemed fine. ≫TheStrike Σagle≪ 15:30, 16 July 2012 (UTC)

• I feel a sort of lack of balance in the design. I think the image should be widened to occupy more part of the box. Its just a design issue, as i completely support the idea and the initiative. Anir1uph (talk) 14:58, 16 July 2012 (UTC)
• Agree with Anir, Perhaps we can develop something like WP:Service awards. In this image the size of the barnstar can by increased and the colour can be Bronze, Silver and Gold .--DBigXray 15:05, 16 July 2012 (UTC)

 Doing... I'm already in the process of creating Silver and Gold.I'm unable to decide the colour of Bronze though.Help! I've had the idea of service awards since many days.I support it and will try my best to propose some.:) ≫TheStrike Σagle≪ 15:13, 16 July 2012 (UTC)

Umm.. Why don't you just take a best guess at the color bronze should be? Some shade of brown seems appropriate? Thanks, West.andrew.g (talk) 16:52, 16 July 2012 (UTC)

Yeah The dark colour doesn't look good, its too dark. Please choose the bronze from WP:Barnstars. --DBigXray 17:05, 16 July 2012 (UTC)

Agree and  Done ≫TheStrike Σagle≪ 17:20, 16 July 2012 (UTC)

 Done

• I guess we must stick to the the usual Stiki Antivandalism Barnstar for 1000 reverts, for 5k, 10k, 25k Bronze , silver and Gold can be used. For higher reverts some changes need to be done on the ribbon and may be add some glow to the barnstar. Also In the current form. I guess it would look better if we can trim the blank portion of the ribbon a bit so that the size of the Barnstar can be increased. As of now due to its small size it appears like The Minor Barnstar--DBigXray 16:58, 16 July 2012 (UTC)

I agree with DBigXray, although it is number of classifications, not number of reverts. I have one idea to add. See below.

number of classifications	message
1	Welcome message
1k	STiki&Antivandalism barnstar
5k	Bronze medal
10k	Silver medal
25k	Gold medal
X	Personal message from West.andrew.g

Where X is the number of classifications by West.andrew.g at the time.

Part of me thinks that Andrew should just write whatever comes to him at the time. But perhaps it should be something in the style of a telegram from the Queen (or President, as it is in the US). Or maybe we could develop a template that looks like a parchment with a wax seal. We can add in lots of formality and pomp and circumstance but obviously leave a space in the template for Andrew to add his own message. Yaris678 (talk) 22:53, 16 July 2012 (UTC)

I am fine with this proposal. Anyone want to template these up in the same style as our "STiki + Anti-vandalism" template? Thanks, West.andrew.g (talk) 00:32, 17 July 2012 (UTC)

Agree--DBigXray 11:47, 17 July 2012 (UTC)

On a side note, going by the current (16 July 2012) leader-board statistics, Allens would take 334 days to have more edits (STiki classifications) than West.andrew.g. A very long time! :D Anir1uph | talk | contrib 01:03, 17 July 2012 (UTC)

So we have the best part of a year to come up with a good template for the personal message. Excellent... Unless fraggle81 keeps up the current rate, in which case we only have 105 days! Yaris678 (talk) 11:41, 17 July 2012 (UTC)

LOL. and recently I was shocked it was the first time I had given the welcome template and the barnstar at the same occasion. Stiki users are indeed getting very active nowadays--DBigXray 11:47, 17 July 2012 (UTC)

In terms of 5k to 25k, check out Wikipedia:STiki/barnstar of merit. All improvements and suggestions welcome. Yaris678 (talk) 12:45, 17 July 2012 (UTC)

Good work on that Yaris, I would only suggest some monor modifications in the image as I have suggested above.--DBigXray 13:05, 17 July 2012 (UTC)

Yes Yaris, thanks for codifying. Especially impressive that you managed to combine bronze/silver/gold into a single template. I probably would have gone the naive route and made three template pages -- but I am a big fan of minimalism. DBigXray: Be bold! Make the changes to your liking and let's see what everyone else thinks.

Moreover, we should post the new system guidelines to the nightly updated "milestones" page so it is clear to anyone who needs to handle that task. Many thanks to DBig for handling this for most of its existence. Indeed, usage has been very high lately; but at my last status update (I email myself reports 4 times daily) I noticed things were a bit more "normal": i.e., 1000 classifications in the last 24 hours with around a 40% revert rate. Thanks, West.andrew.g (talk) 14:14, 17 July 2012 (UTC)

Thanks guys. In terms of the size... perhaps its worth considering the case of Template:The Barnstar of National Merit. This has one version similar to ours and a bigger "alt" version. Which version do people prefer? If we go for the bigger one there is a lot of space for the message... although I suppose we can tweak the formatting. Maybe fix the width of the box that the message appears in. Decisions, decisions. Yaris678 (talk) 14:17, 17 July 2012 (UTC)

Well I will suggest something this is in between these two images in terms of the barnstar Size. The alt version is one helluva Gigantic barnstar, perhaps a variety of that can be used for >= 25k classifications. --DBigXray 14:54, 17 July 2012 (UTC)

I'l like to have a wider image. I really don't know how to do that, but i'd prefer the image to be ~ twice its current width (keeping the height of the box fixed), so it occupies a larger % of area of the box. Thanks! Anir1uph | talk | contrib 20:43, 17 July 2012 (UTC)

Images modified. Changed to official STiki logo, and trimmed the vertical height of the blue area, in the hopes this would help us to gain some image width without sacrificing additional height. Thanks, West.andrew.g (talk) 18:33, 18 July 2012 (UTC)

DBigXray likes this. great, this one looks much better, We are good to go now, We will now be using them to reward STiki users. --DBigXray 18:43, 18 July 2012 (UTC)

Thanks! Looks much better. :) Anir1uph | talk | contrib 19:07, 18 July 2012 (UTC)

@DBig: I am going to get out of the way so we don't edit conflict. One comment: Might it be preferable just to document the individual templates on their respective pages (rather than establishing the usage guidelines there and pointing to other templates? Aside from maybe a "see also..."). Indeed, someone besides ourselves may wish to use these templates (say, to reward a STiki user for a particular revert). They might be a bit confused with all this specific usage information. Why not consolidate our award pattern documentation at WP:STiki/milestones and let the template pages standalone? West.andrew.g (talk) 19:16, 18 July 2012 (UTC)

Apologies for (edit conflict) :P (I am done now please go ahead) agreed putting it at one page , the milestone page seems to be a better idea.--DBigXray 19:24, 18 July 2012 (UTC)

I am done now, as well, in case I made any mistakes. Thanks, West.andrew.g (talk) 20:19, 18 July 2012 (UTC)

Proposal Withdrawn - Strike Eagle
The following discussion has been closed. Please do not modify it.
Here's my proposal-Can be modified by other users :) Name of Service Award No.Of Edits Rookie user of STiki 1-100 Intermediate(or similar word 200-500 Advanced User of STK 500-1000 Experienced User 1000-3000 Veteran User 3000-5000 Senior User 5000-10000 Master User 10000-25000 Grandmaster user 25000-50000 Ultimate User 50000-eee ≫TheStrike Σagle≪ 16:52, 16 July 2012 (UTC)  Disagree when i said WP:Service awards i was not saying we need a service award such as this. What i meant was our barnstars can take clue from the design of service awards. This system will be just too much, in my opinion.--DBigXray 17:01, 16 July 2012 (UTC) I agree with DbigXray. IMO, we use STiki for removing vandalism, not really for any awards. The existing system of Barnstars seems good. This would put too much importance on milestones. Not that i don't appreciate the effort being put into this by StrikeEagle. Anir1uph (talk) 17:24, 16 July 2012 (UTC) Proposal Withdrawn ≫TheStrike Σagle≪ 17:27, 16 July 2012 (UTC)

Hi! I usually use STiki, but when I tried to log in, it kept on saying "Log-In Failed". Can someone please help me out? Thanks, Electriccatfish2 (talk) 23:30, 16 July 2012 (UTC).

I can seem to log-in and make edits. Anir1uph | talk | contrib 00:02, 17 July 2012 (UTC)

Yep... I just logged out and in again and it works fine for me. Two random thoughts:

1. Are you using the latest version of STiki?
2. Have you recently changed your firewall?

Yaris678 (talk) 00:03, 17 July 2012 (UTC)

STiki server is up and operating. Is CAPS-LOCK on? Are you using STiki from the same connection you typically do? West.andrew.g (talk) 00:34, 17 July 2012 (UTC)

We'll assume this issue is sorted out unless Electriccatfish posts to the contrary. I'll post a talkback to his page. Thanks, West.andrew.g (talk) 14:06, 17 July 2012 (UTC)

for Context refer Wikipedia talk:Counter-Vandalism Unit/Academy#CVUA instructors need to keep themselves updated on Antivandal tools --DBigXray 12:53, 17 July 2012 (UTC)

Hi, just a quick note back to you about the posting at CVUA and on my student's talk page about the fact that STiki combines all the edits into a single diff. Thanks for pointing that out. However, STiki still gives the warning that you're looking at "1 intermediate revisions not shown; rollback situation Metadata reflects only most recent edit in chain", which indicates that you're not looking at all the edits. I have taken a look and it seems as though you're correct, so just wanted to point out that there is an inconsistency which may lead to confusion, as it obviously did with me :-). Again, thanks for pointing it out. No talkback needed, I'll check here. Vertium ^{When all is said} _{and done} 12:44, 17 July 2012 (UTC)

Thanks Vertium for your message. The Diff screen in fact combines all the revisions into one and the metadata such as time of edit etc is for the latest revision. You may have a point here about the Stiki warning that is displayed and the need to clarify it in a better (simpler/less technical) way for all STiki users. lets see what others have to say about it, regards--DBigXray 12:53, 17 July 2012 (UTC)

Would it be better if the message said "The diff browser shows a combined diff for 2 edits by the same user. The edit properties box shows information for the last of these edits. If instructed, STiki will revert all 2 edits." Yaris678 (talk) 13:58, 17 July 2012 (UTC)

Agreed, the message could be more clear. However, I just duplicated the message Mediawiki interface displays when a diff spans more than one revision. I liked Yaris' suggestion and will make this change to be pushed in the next release, unless there are any objections? Thanks, West.andrew.g (talk) 14:04, 17 July 2012 (UTC)

Agree Definitely support the clearer, friendlier message. Theopolisme _TALK 14:50, 17 July 2012 (UTC)

Agree Much better. Instead of including a number, it might say "The diff browser shows a combined diff for multiple edits by the same user. The edit properties shows information for the last of these edits. If instructed, STiki will revert all the included edits". My thinking is that there may be more than 2 edits in the span. Of course, if you can use a variable to count the edits in the span and replace that in the message, then the less quantifiable language wouldn't be needed. In any case, I support the clearer language. Many thanks. Vertium ^{When all is said} _{and done} 20:05, 17 July 2012 (UTC)

Yep. I was assuming that the number would be a variable. Like how the current message says "1 intermediate revisions". Yaris678 (talk) 10:51, 18 July 2012 (UTC)

Agree The note by Yaris definitely is more user friendly. A minor modification of the above can be "The diff browser shows a combined diff for 2 edits by the same user. The edit properties box shows information for the last most recent of these edits. If instructed, STiki will revert all 2 edits." This statement clarifies the second sentence which is the real cause of confusion (if any) --DBigXray 20:28, 17 July 2012 (UTC)

If people think that "most recent" is easier to understand than "last" then let's go with that. Yaris678 (talk) 10:51, 18 July 2012 (UTC)

Agree (Now that my question has been answered) -- Gareth Griffith-Jones (talk) 07:40, 18 July 2012 (UTC)

Note that I went ahead and archived alot of earlier discussion in order to get this page cleaned up. Thanks to StrikeEagle for his initial design and Yaris for the initial template. I made some visual and template revisions, which can be reviewed at Wikipedia:STiki/barnstar of merit. Moreover, our new distribution policy has been codified at Wikipedia:STiki/milestones. Feel free to comment here as needed. West.andrew.g (talk) 20:35, 18 July 2012 (UTC)

So you didn't like the number of edits being determined automatically by the choice of metal? I suppose this way does give more flexibility. Yaris678 (talk) 09:13, 19 July 2012 (UTC)

Hope no offense was taken, but I was just going for a little flexibility and the ability to change policy down the road. Even though we tend to control distribution, exclusively, if someone wanted to do a modification its one less little bit of template hacking for them to figure out. On that thought, maybe platinum or Kryptonite should be the metals used for higher levels of achievement (though "coloring" maybe prove challenging). My alternative thought was stacking of stars: i.e., "a double gold barnstar." Thanks, West.andrew.g (talk) 19:20, 24 July 2012 (UTC)

Have I gone crazy, or did the infobox width expand non-trivially? I see a recent edit to Template:Infobox_software and the switch to "translation available". Maybe it just changed some wrapping threshold? Anyone else notice this? See also GIMP, which has considerably more data yet remains thinner. Thanks, West.andrew.g (talk) 16:42, 24 July 2012 (UTC)

As I am seeing the same expansion so I assume that so far both of us have managed to keep our sanity --DBigXray 19:22, 24 July 2012 (UTC)

Hi! I changed my password yesterday and now when I log in, it says "log-in failed". Can someone please fix this? Thanks, Electric Catfish 22:53, 25 July 2012 (UTC).

Possibly related, I had trouble logging in yesterday with my usual user name and password. It resolved itself after about 10 minutes and lots of cursing. Caps lock was off and spelling was correct. Fraggle81 (talk) 23:13, 25 July 2012 (UTC)

Keep me updated, but it's not a STiki issue it doesn't seem (at least, I had no issue at login running the latest code). Moreover, the STiki server seems to be in fine condition (you wouldn't even get the GUI open if it wasn't). I would suggest: (1) check CAPS, (2) check to make sure no leading/trailing spaces, (3) see if you can successfully logout and login on the browser Wikipedia interface. The fact both of you have seen this lately, though, is quite weird. Thanks, West.andrew.g (talk) 23:58, 25 July 2012 (UTC)

Now it works. Electric Catfish 00:45, 26 July 2012 (UTC)

It is in connection of this and this that I pledge to make a request here for using STiki. Thanks. VIVEK RAI :  Friend?  05:41, 25 July 2012 (UTC)

Sure no one is in a hurry, we will wait until you are ready with your antivandal reverts. Let me assure you that the result will be purely based on the appropriateness of your reverts. nothing else. In my personal opinion, as long as an editor does not edit disruptively and follows WP:VANDAL and WP:NOTVANDAL there is no reason to disallow him from using STiki.

On a side note no one on this page WT:STiki is a wikipedia admin. so we will not be digging into your past. cheers--DBigXray 13:58, 25 July 2012 (UTC)

I probably would have granted immediately if the CUVA instructor showed up. However, the request for just a couple of simple reverts was very reasonable and could have been completed in less time than all the back and forth above. To that end, we'll wait and see what Vivek wants to do and ask him to show a couple of reverts.

In the general case, a few manual reverts and warning placements would probably be a good idea before using any assistant (Twinkle included). I'm not super familiar with CUVA but aren't their "lessons" one completes? Could we make some hard rule about lesson completion and STiki access? I too would like some evidence of the ability to tell between vandalism and good-faith. Thanks, West.andrew.g (talk) 16:37, 25 July 2012 (UTC)

@Andrew Normally the WP:CVUA instructors ask their students to read the required Policy pages, ask them few policy related and vandalism-situation questions to test their understanding.They also ask them to make some manual reverts from WP:RCP and Special:RecentChanges. After which the candidate is cleared for the next level that is using automated script such as STiki. So i guess this process is safe enough to be trusted. On the other hand the fact that Vivek was shying away from making reverts, led me to suspect that he is not familiar with these things.--DBigXray 16:52, 25 July 2012 (UTC)

Thank you for your concern. I am Vivek Rai's instructor at WP:CVUA. I can tell you a few things. I am a new instructor there, and he is my first student. I have only had one 35-minute instructional session with Vivek so far, and I can tell you that he shows good knowledge and a positive attitude, especially towards encouraging people to make constructive edits. However, I have only touched the surface of what I need to teach him for graduation in the Academy. We discussed how to identify various types of vandalism and I helped him with a few case studies, and his answers were satisfactory. However, I did not have the chance to ask him to make a few reverts, I did not discuss with him about WP:NOTVAND, and we have not discussed anything about WP:NPP or using WP:TW as a tool for reverts. He seems to be proficient enough so far at using Twinkle to add tags and for WP:CSD, I don't think that we will have much trouble learning the rest of its functions. But I would say right now, he needs a little more work, and I would say that after a few more sessions, he will be ready to use STiki. Elizium23 (talk) 19:47, 25 July 2012 (UTC)

Thank you for your response. If you request here, your student's approval will be immediate. Thanks, West.andrew.g (talk) 06:45, 4 August 2012 (UTC)

Could a developer merge all of my STiki edits as Tomtomn00 to Thine Antique Pen on the leaderboard? Thanks! ⇒TAP 13:51, 25 July 2012 (UTC)

Accepted per [1] will be done shortly, cheers--DBigXray 14:04, 25 July 2012 (UTC)

Thanks! ⇒TAP 14:05, 25 July 2012 (UTC)

 Done Will be reflected on next leaderboard update. Thanks, West.andrew.g (talk) 16:41, 25 July 2012 (UTC)

Thank you! Thine Antique Pen 16:56, 25 July 2012 (UTC)

Is there a requirement for STiki now, as there wasn't when I started using it? I seem to not be able to use the tool on my public account, hence I'm asking if there is a requirement here. :P Thine Antique Pen 17:12, 25 July 2012 (UTC)

Yes please refer Wikipedia:STiki#Using_STiki.--DBigXray 17:25, 25 July 2012 (UTC)

Yes, we will grant that account explicit access. Have any of the other the other individuals with STiki account access tried out that functionality yet? If not, why don't you give it a whirl in this case (and if there is some issue, I am around). Thanks, West.andrew.g (talk) 18:54, 25 July 2012 (UTC)

Not yet,I tried once but failed as the commands to be used were not clear, will give it a try later on as I am busy. --DBigXray 12:26, 26 July 2012 (UTC)

Fair enough, if you do need support when you become available, just let me know. Thanks, West.andrew.g (talk) 06:53, 4 August 2012 (UTC)

I have been served a null diff by STiki and this time I double checked: I am definitely using the "STiki (metadata)" queue - Yaris678 (talk) 22:36, 3 August 2012 (UTC)

Isnt that normal and expected, if the IP has self reverted (probably doing edit tests on the article.) check out article history. Or do you want to say something else, that i missed ? --DBigXray 22:42, 3 August 2012 (UTC)

I remember a number of these being reported when the combined diff was first brought in... and I thought that Andrew made it so that these were dismissed before they were presented to the user... but I could be wrong.

If I am right then it means it didn't work, for some reason, in this case.

Yaris678 (talk) 23:12, 3 August 2012 (UTC)

Hmmm... maybe it wasn't fixed. All I found in the archive was Wikipedia talk:STiki/TalkArchive04#Blank diff? Yaris678 (talk) 23:22, 3 August 2012 (UTC)

You mean this shouldn't happen? I get these quite a lot and have done since I started using STiki. Fraggle81 (talk) 03:25, 4 August 2012 (UTC)

Likewise, Fraggle. -- Gareth Griffith-Jones (talk) 05:25, 4 August 2012 (UTC)

Agreed. It is silly to pop these to end users and in my recent experience I have been seeing several of these, as well. Should be straightforward to detect for all queues. Added to the bug/feature table as T#016. Will be included in next release. Thanks, West.andrew.g (talk) 06:55, 4 August 2012 (UTC)

Cool. Thanks. Yaris678 (talk) 19:10, 4 August 2012 (UTC)

Instead of the manual archiving by West.andrew.g, why don't we setup up archiving by a bot for the STiki talk page? That will preserve the edits histories, and various versions, and add a search bar (just for convenience). Thanks! Anir1uph | talk | contrib 10:49, 19 July 2012 (UTC)

Support, were you thinking ClueBot or Misza? Theopolisme _TALK 21:02, 23 July 2012 (UTC)

Either. DOn't know the difference b/w them, tho i use Misza on my talkpage. Anir1uph | talk | contrib 17:08, 24 July 2012 (UTC)

Not much experience here, either, though via my cooperation with ClueBot-NG I have noticed increasing discussion on their commons about ClueBot III (the archiving bot, I think) being down or having other issue. FWIW, it seems their anti-vandal efforts have largely been in auto-pilot lately, too, without too much development (... real life happens). One thing we definitely need is support for sticky posts. Thanks, West.andrew.g (talk) 19:13, 24 July 2012 (UTC)

Well for sticky posts there exists stuffs like {{DNAU}} (for misza) and {{User:ClueBot III/DoNotArchiveUntil|unix timestamp here}} . Also there are commands for quick archiving of particular thread if its closed and no longer needed. I guess using a bot would reduce the unnecessary overhead of manual archiving and keep the Stiki talk page regularly trimmed. (I use Cluebot III on my talk page by the way. )--DBigXray 19:20, 24 July 2012 (UTC)

Well, we can then start using Cluebot here, as it seems that the STiki representatives are more used to that. Anir1uph | talk | contrib 19:52, 24 July 2012 (UTC)

Andrew, Shall we go ahead ? I guess 7 or 10 days delay prior to archiving is enough.--DBigXray 11:50, 27 July 2012 (UTC)

I will set up archiving with ClueBot now, if no one has any objections. Objections, anyone? Theopolisme _TALK 03:53, 4 August 2012 (UTC)

That's fine. I don't know all the parameters, but try to keep the organization as similar as possible to what I've been doing manually in the past. I've been trying to keep 25 threads/archive. Other than that, do what you think is best. Thanks, West.andrew.g (talk) 06:43, 4 August 2012 (UTC)

Another thought, the "bug/feature table" will be need to made sticky (we don't want it archived). Thanks, West.andrew.g (talk) 06:46, 4 August 2012 (UTC)

Will do. Starting now. Theopolisme _TALK 06:47, 4 August 2012 (UTC)

 Done - I will wait and make sure the bot does the right thing, but otherwise we should be good. Theopolisme _TALK 07:14, 4 August 2012 (UTC)

Did minor tweaking from Wikipedia_talk:STiki/Archive0%(counter)d to Wikipedia_talk:STiki/Archive %(counter)d so that we get Wikipedia_talk:STiki/Archive 8 next time. Also I have changed the template style to auto with a search box. Search box can be very helpful for new editors who would want to look at answers to frequently asked questions already answered earlier. and Auto will reduce the need to add links here manually.--DBigXray 14:47, 4 August 2012 (UTC)

When I originally did this I made it auto, but I think Andrew thought it looked better with manual, per this diff. Do you know of a way to modify the auto archive box's naming/display conventions? Would it be hard to create a special Template:Archive box/STiki? Theopolisme _TALK 15:18, 4 August 2012 (UTC)

Yeah I saw that, Apparently Andrew prefers a "slimmer" Archive box (with a single column). While I have no preference for the width of the archive box, but for keeping the search box, i guess this width is sufficient enough, any furthur trimming of the width will make the search box harder to type. lets wait for Andrew's opinion--DBigXray 15:24, 4 August 2012 (UTC)

While I do like a slim box, my initial attempts were more about trying to fix some template alignment yuckiness I was seeing on my machine. Somewhere along the line that got fixed and I think things look fine now (and understand we need some width for a search box). I'm happy. Now let's just confirm that it works. How often does the bot come along and look for content to be archived? West.andrew.g (talk) 17:24, 4 August 2012 (UTC)

P.S. Would someone be willing to setup the equivalent on my personal talk page? Thanks, West.andrew.g (talk) 17:29, 4 August 2012 (UTC)

P.P.S Is there a reason only Archive #5 and #6 got the archive navigation bit by DBig? I don't care either way, but we should be consistent. West.andrew.g (talk) 17:34, 4 August 2012 (UTC)

The First archiving is generally done after 2 days, probably to give us time to decide the parameters . Also as requested the Feature request thread will be kept until year 2022 and the Change log thread will remain on top until 31 December 2012, as we have sneaked the {{DNAU}} template. I'll add the navigation bit to other pages as well :) --DBigXray 17:39, 4 August 2012 (UTC)

The bot has archived 3 old threads from here. And everything has worked as expected. I have also added an Archive index That will be written shortly.--DBigXray 13:26, 5 August 2012 (UTC)

I'm having the same problem as I had before with not being able to log in. Everything is correct but STiki says Log-in Failed. Fraggle81 (talk) 22:47, 5 August 2012 (UTC)

Ok, I managed to login with nothing changed, I made my first classification and STiki automatically closed. I've restarted and it seems fine now. Fraggle81 (talk) 22:53, 5 August 2012 (UTC)

In order to save (at least some of) these from appearing again and again, would it make sense to do a section on WP:STiki about "Can't Log In? Yaddaydaada"... ? Theopolisme _TALK 22:58, 5 August 2012 (UTC)

We need to figure out what is going on, first. Are any other tool users reporting things like this? Maybe it has something to due with a pure API login? This seems strange, though, because I'd imagine that the browser front-end is just an interface to these API calls. The fact it self-heals is extremely confusing and might be worth lodging a Village Pump (technical) post. Thanks, West.andrew.g (talk) 23:49, 5 August 2012 (UTC)

Hello, I am a CVUA instructor and my student, Activism1234, would like to learn how to use the tool but is unable to because he does not fully qualify yet. I will watch all of his edits with the tool and will take full responsibility if anything happens. Can a expection be made? -- Cheers, Riley Huntley _talk No talkback needed; I'll temporarily watch here. 02:15, 13 August 2012 (UTC)

I'm on Windows at the moment (... PowerPoint is the *only* reason this happens). When I get back on the Linux side of things, I will fulfill this request (within the next couple hours) and I will acknowledged here. Try to ensure your student understands WP:VANDALISM and WP:NOTVANDALISM before turning him/her loose. Thanks, West.andrew.g (talk) 05:10, 13 August 2012 (UTC)

 Done West.andrew.g (talk) 06:38, 13 August 2012 (UTC)