Wikipedia talk:STiki/Archive 9

This is an archive of past discussions on Wikipedia:STiki. 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.

Archive 5

←

Archive 7

→

Warning for registered users only

I usually have warnings disabled, because 95% of what I revert are IP edits and I don't see a sense in warning an IP. Most of them have only done exact the one edit I revert. So I try to activate warnings when I revert something a registered user did, but that relies on me checking the user name. If it isn't to much work, can you implement an option to only issue a warning if the user is a registered one? Or maybe give a visual feedback if the editor is register or an IP, like a red frame or something? -- Windharp (talk) 12:29, 4 September 2012 (UTC)

There is a good reason for warning IP's, it alerts other editors to previous vandalism even if the IP themselves don't read it. It also helps to ensure the swift blocking of persistent vandals through AIV. I would say if you're going to revert as vandalism you should always leave a warning. Fraggle81 (talk) 13:48, 4 September 2012 (UTC)

It's mostly IP users' contributions that are vandalism for wikipedia (you can also see ClueBot's contributions for this). You need to warn IPs. With the warning some IPs will be aware that their recent edits was not good for wikipedia and they may improve their quality of their edits. In my opinion, everyone should be warned. Torreslfchero (talk) 14:39, 4 September 2012 (UTC)

There is no reason for such a feature. All vandals must be warned. There are *many* IP vandals who commit enough vandalism instances to get blocked. Under your philosophy an IP user could commit an unlimited amount of vandalism and would never get noticed/blocked/warned! STiki has taken some care to ensure that warnings will only be issued under the appropriate circumstances. Those warnings have turned many vandals into productive editors! West.andrew.g (talk) 15:12, 4 September 2012 (UTC)

I have to agree with everyone above as my first edit was as an IP and my first edit was vandalism. When I got that warning, I decided to follow it's instructions and create an account. Now, if I didn't get warned, I probably would have continued my joking around and got blocked. Even if you don't think a warning will help, leave one -- Cheers, Riley Huntley _talk 15:55, 4 September 2012 (UTC)

I quite agree. Warnings are critical and should be given in every instance of vandalism. Thanks. Vertium ^{When all is said} _{and done} 16:21, 4 September 2012 (UTC)

Persistent vandals should definitely be warned. The case of an IP that makes a single edit is an interesting one suitable for A/B testing, similar to the A/B test on warning wording. Vandals can reform but they can also enjoy provoking. There is also the chance that the next user of the IP address will be someone totally different (Andrew has attempted to deal with that by putting a time limit on IP warnings, but it doesn't remove the possibility). The test would probably be attached to Huggle (That was how the warning test worked so the infrastructure should be in place). For vandal reverts where there was no previous edits, the test would only warn 50% of IPs, creating two sets of IPs. The test would look for differences in the number of following edits that were vandalism and the number of following edits that were not vandalism.

Perhaps an A/B test of single IPs would be a good suggestion for the vandalism studies people. They seem open to suggestion.

Or maybe we should suggest it to the editor engagement experiments people who did the warning-wording trial.

Yaris678 (talk) 16:56, 4 September 2012 (UTC)

Nothing to add to what Yaris said here, but it did give me an interesting thought. One of the drawbacks of warning IPs is that, due to DHCP, the warning message might be received by a *different* editor sometime in the future. Why not auto-clear, hat, or in some other way suppress the message/warning after some period of IP account inactivity? Is this done in any way at current? It's certainly silly to have year's old warnings sitting on un-used talk pages. However, it is not unimaginable that this could be pushed into the days/weeks scope of thinking. Thanks, West.andrew.g (talk) 17:45, 4 September 2012 (UTC)

Genius! I massively agree with that. I would love a bot that went round clearing template warnings from the talk pages of inactive IPs. Admins generally ignore "stale" warnings. If people are interested, they can always check the history of the talk page. Yaris678 (talk) 18:31, 4 September 2012 (UTC)

(edit conflict) And a name as well: The Stale Warnings Cleaner Bot (alright, I admit that might need some work). That would actually be pretty fantastic, as right now stale warnings only make it harder to see the actual new, relevant ones. Any idea how one would go about coding this, Andrew? Theo polisme 20:48, 4 September 2012 (UTC)

Coding would seem quite trivial, whether undertaken by myself or someone else with bot experience. Its the community consensus and parameter tweaking (i.e., how "old") which would probably be more troublesome. If you feel strongly about it, feel free to pursue via the typical channels (Village Pump? RfC?). West.andrew.g (talk) 20:58, 4 September 2012 (UTC)

This made me wonder about the logic of how STiki applies the various levels of warnings? Does it look at the recency of other warnings to decide which level warning to apply? I do agree that any previous warning more than a few months old is hardly worthwhile for the purpose of vandalism warnings. I presume they'd still be available as part of the talk page history should anyone need to investigate socking or any other activity related to the IP account. Great idea. Vertium ^{When all is said} _{and done} 20:46, 4 September 2012 (UTC)

Right now it focuses primarily on increasing warning levels intra-month. This is not optimal (but follows the logic of CBNG and Huggle to a large extent). There is an outstanding ticket number to improve this. One or two archives back there was also a sizeable discussion (whose title was AIV focused, I believe). Thanks, West.andrew.g (talk) 21:31, 4 September 2012 (UTC)

Andrew is refering to T#018 above and the discussion at Wikipedia talk:STiki/Archive 8#AIV reports on stale final warnings. Yaris678 (talk) 08:25, 5 September 2012 (UTC)

@Theopolisme, maybe StaleBot for short. Yaris678 (talk) 08:27, 5 September 2012 (UTC)

OK. Anyone got any suggestions about the best place to start a talk with the wider community about such a bot? As Andrew says, getting consensus on the specifics could be troublesome. How about Wikipedia:Village pump (technical)? Yaris678 (talk) 11:39, 5 September 2012 (UTC)

Wikipedia:Bot requests? Callanecc (talk • contribs • logs) 12:38, 8 September 2012 (UTC)

Thanks. I didn't know about that page. I have now posted Wikipedia:Bot requests/Archive 50#Bot to remove patently stale warnings from IP talk pages. Yaris678 (talk) 12:22, 10 September 2012 (UTC)

Appeal to use Stiki

I've been on Wikipedia for a while now and was wondering if I could use Stiki. I revert using Twinkle and even with permission will for most of my vandal reverts. Thanks ♠♥♣Shaun9876♠♥♣ ^Talk ^Email 00:31, 12 September 2012 (UTC)

Hi,

Yes, you obviously have a certain amount of experience of reverting vandalism using Twinkle. Have you tried requesting rollbacker rights at Wikipedia:Requests for permissions/Rollback? I suspect you will get them. That will allow you to use both STiki and Huggle.

Yaris678 (talk) 07:35, 12 September 2012 (UTC)

Done -- I agree with Yaris, but regardless, I've explicitly granted you rights. Thanks, West.andrew.g (talk) 17:31, 12 September 2012 (UTC)

Thanks ♠♥♣Shaun9876♠♥♣ ^Talk ^Email 22:48, 12 September 2012 (UTC)

Misuse of STiki

Good morning to all! This might be of interest. Sincerely, -- Gareth Griffith-Jones (GG-J's Talk) 10:08, 21 September 2012 (UTC)

Executive summary: Not a tool issue. Just a content-dispute where one or two edits were made with STiki. User:GoShow did make some false statements about STiki, seemingly in an effort to shift blame. Those have been contested. User:GoShow has been a reasonable STiki user, and we will allow that to continue, albeit with maybe a touch more monitoring. West.andrew.g (talk) 16:39, 21 September 2012 (UTC)

(copied from User talk:West.andrew.g) Hi, sir, it is a fascinating honor to meet you sir, Sorry about this new edit summary issue, I don't want higher manipulative users keep warning just for the benefit of me or other medium user not knowing much, or just taking advantage of me and others,to have me or them out, because I have seen often while observing other's contributions, but I value alot from STiki and so do others, as well they do mistakes, I just hope you would know that is just me in stating facts and not trying to pick a dispute, otherwise, an extravagant honor to your, sir. Sincerely,--GoShow (...............) 20:09, 21 September 2012 (UTC)

There seems to be an issue here in your ability to take constructive criticism and follow policy (and run-on sentences, but I'll set that aside). Why have you just deleted all the critical content from your talk page (using an IP address, no less)? You seem to realize these are experienced users, but for some reason you consider them manipulative? I don't get it; but I'll let those more directly involved handle the immediate situation. If any more issues arise, STiki access may be revoked. Thanks, West.andrew.g (talk) 20:26, 21 September 2012 (UTC)

No, there are some not all higher users who do use ordinary edits, but have multiple accounts to use on other edits for right reasons and other reasons, of course I am not blaming User:EatsShootsAndLeaves for two accounts either--GoShow (...............) 21:13, 21 September 2012 (UTC)

I'm done with this. My only job here is to prevent STiki abuse. For interested others, I'll note this thread seems to be ongoing in other places (e.g., User_talk:EatsShootsAndLeaves). Thanks, West.andrew.g (talk) 21:20, 21 September 2012 (UTC)

Permission

Hi there, i would love to be granted permission to use this tool. I update football statistics and other sport related stats and have helped cleared up vandalism and i thought it was time i got my own account so i would be grateful to be able to use this superb tool. Thanks, Telfordbuck (talk) 16:44, 26 September 2012 (UTC)

Registration was made this afternoon. -- Gareth Griffith-Jones (GG-J's Talk) 18:48, 26 September 2012 (UTC)

I know, i have downloaded the tool, I am looking through edits and then undoing the vandalism manually. But i feel, if granted permission, i will also be able to warn users so they cant repeat. Telfordbuck (talk) 19:10, 26 September 2012 (UTC)

You may be wondering how i was looking through without permission, the answer is that i was using one of my mates accounts to look through, luckily he trusts me!!! but i dont want to have to do this anymore, i want to be able to use my own account, one of the main reasons i got my own account Telfordbuck (talk) 19:29, 26 September 2012 (UTC)

Or would it be better if i asked again in a month or two and just carry on what im doing until i prove im dedicated? Telfordbuck (talk) 19:42, 26 September 2012 (UTC)

That would be an excellent plan. You need much more experience before being let loose with a powerful tool such as STiki. Good editing! Sincerely, -- Gareth Griffith-Jones (GG-J's Talk) 19:48, 26 September 2012 (UTC)

I'll interject here. A "month or two" of experience is not required. We want to establish two things: (1) good-faith intentions (seemingly already present), and (2) the ability to differentiate between WP:VANDALISM, WP:NOTVANDALISM, and good-faith editing. Towards achieving the second milestone: The folks at the counter-vandalism unit (WP:CVUA) can mentor you to this end. Alternatively, you can read the vandalism documentation (as well as WP:STiki) yourself and continue manual editing. Compile a list of the cases where making a vandalism/good-faith/innocent determination was difficult; what you ultimately decided; and why. If you come back with about a dozen of these and we agree with your reasoning, I'll be more inclined to approve your account. West.andrew.g (talk) 20:35, 26 September 2012 (UTC)

I'll butt in here I Agree that Telfordbuck should enroll in the CUVA and get some Basic training and then start to use STiki I walked the same path ;) Fox2k11 (talk) 23:25, 26 September 2012 (UTC)

OK, here is an example, http://en.wikipedia.org/enwiki/w/index.php?title=Travel_Sick&diff=514829940&oldid=514809754 now i would be tempted to say this is vandalism but it is possibly a good faith edit so they can purchase the dvd, just going the wrong way about it Telfordbuck (talk) 16:58, 27 September 2012 (UTC)

On the other hand , this is clearly vandalism: http://en.wikipedia.org/enwiki/w/index.php?title=Positional_notation&diff=prev&oldid=514809316 Telfordbuck (talk) 17:00,27 September 2012(UTC)

Another one is that this person is a repeat offender http://en.wikipedia.org/wiki/Special:Contributions/87.254.1.2 and using stiki would help me warn him. Telfordbuck (talk) 17:11, 27 September 2012 (UTC)

Also, this is blatant vandalism repeatedly http://en.wikipedia.org/wiki/Special:Contributions/194.81.39.200 Telfordbuck (talk) 17:11, 27 September 2012 (UTC)

This is one i wasnt too sure on : http://en.wikipedia.org/enwiki/w/index.php?title=Gabriel_Agbonlahor&diff=514830332&oldid=514828895 . In my opnion, the editor is vandalising as he has already being warned but as i cant guarantee it, i would just make a good faith revert. Telfordbuck (talk) 17:30, 27 September 2012 (UTC)

Excellent start here, as I agree with your reasoning on all examples. I'd like to see a few more cases, though. Please concentrate on the *borderline* cases. Blatant vandalism is rarely something anyone has difficulty with. Six more examples should suffice. Please also submit them all together, for the sanity of those who closely watch this page. Thanks, West.andrew.g (talk) 19:28, 27 September 2012 (UTC)

Hi there, i thought i'd send show 3 now and 3 tomorrow,:

http://en.wikipedia.org/enwiki/w/index.php?title=Healthy,_Hunger-Free_Kids_Act_of_2010&diff=prev&oldid=514862607 This is one i found difficult to decide as i don't know if it was genuine or a sarcastic one so i would leave it for another editor to authenticate http://en.wikipedia.org/enwiki/w/index.php?title=Flozell_Adams&diff=prev&oldid=514870822 This one, although many would say is blatant vandalism, i would say it is a good faith edit as the editor may think they are a good sportsperson and just went the wrong way about displaying this opinion http://en.wikipedia.org/enwiki/w/index.php?title=Bridge&diff=prev&oldid=514875753 I would say this is a good faith edit because this fact easily could be true but needs to back up information with a reference. Telfordbuck (talk) 21:08, 27 September 2012 (UTC)

Sorry i havent submitted this one with the others, http://en.wikipedia.org/enwiki/w/index.php?title=Anatomical_terms_of_motion&action=edit&undoafter=507434139&undo=514863695 This one appears to be just a revert as i believe that just because someone adds bold text, it doesn't necessarily constitute vandalism Telfordbuck (talk) 21:23, 27 September 2012 (UTC)

And these are the other two: http://en.wikipedia.org/enwiki/w/index.php?title=Snatch_squad&diff=prev&oldid=514765902 This is definitely a good faith edit but i believe it should be reverted as it was better before. And this one is the 6th: http://en.wikipedia.org/enwiki/w/index.php?title=Surkhet&diff=prev&oldid=514967756 I would also class this as good faith as it is an opinion therefore not for wikipedia, but isn't vandalism. — Preceding unsigned comment added by Telfordbuck (talk • contribs) 16:21, 28 September 2012 (UTC)

I think all the rationale here are reasonable. You now have STiki access. In cases where you are unsure, be sure to use the "Pass" option and leave the edit for a more experienced user. Thanks, West.andrew.g (talk) 13:42, 29 September 2012 (UTC)

Requesting permission to use STiki

I want to revert vandalism more efficiently using STiki. I am reverting vandalism using Twinkle. I have currently made 371 edits to the article namespace. I made mistakes in distinguishing vandalism and good-faith edits. So I do not want to make mistakes in distinguishing blatant vandalism and good faith edits by detecting vandalism. Hope you grant me permission to use STiki Sriharsh1234 10:29, 22 September 2012 (UTC)

Hi Sriharsh. Your admission that you've "made mistakes" is a bit concerning; how have these shortcomings been addressed? I see you've contacted some of the counter-vandalism academy folks as well. Are you formally associated with that process (i.e., have a mentor and undergone some of their training "lessons")? Thanks, West.andrew.g (talk) 21:09, 23 September 2012 (UTC)

I learnt from Floating Boat ho to distinguish between blatant vandalism and good faith edits. Now I know that what is vandalism and what are good faith edits. Sriharsh1234 01:59, 26 September 2012 (UTC)

mmmm? Really, not from a schoolboy! -- Gareth Griffith-Jones (GG-J's Talk) 18:52, 26 September 2012 (UTC)

I really don't care who taught you, my goal is to assess your ability to do the task at hand. I did see your lesson with Floating Boat and I interpret this a sign of good faith. However, even you must admit that an open-ended 5 question test doesn't really prove anything dramatic besides your good intentions. Moreover, auditing your recent contributions I have a few questions about recent edits:

* [1] -- Why did you classify this as vandalism?

* [2] -- Why did you give this user a level 3 deletion warning when their last warning was three years ago? Why not a friendly level-1? Isn't this a bit WP:BITE?

You do have a good base of anti-vandal work, but I also see plenty of self-reverts in your history. Before any approval decision will be made, please address the questions above. Thanks, West.andrew.g (talk) 20:22, 26 September 2012 (UTC)

* [3] I classified the edit as vandalism because the editor was telling something about him/her self and I did not see anything constructive in the edit.

* Incorrect, a simple click-through reveals "I am Weasel" is a cartoon with an existing article. This was constructive editing. West.andrew.g (talk)

* [4] I gave the editor 3rd level warning because the editor removed about 7 KB from the article and the editor has been blocked once Sriharsh1234 09:22, 27 September 2012 (UTC)

* An incorrect assessment, in my opinion. The block was issued 3 years ago! There is also a note on the talk page that indicates the IP is a proxy server. It is convention that if an account makes no bad edits for ~1 month, the warning severity is restarted. Familiarize yourself also with DHCP and its impact on the warning of IP users. West.andrew.g (talk)

I am not enthusiastic about your responses. I think you should lean on the WP:CVUA process for more lessons and experience. Alternatively, as I've done with Telford below, I'll also allow you to compile a list of approximately one dozen cases you've handled where the classification is *borderline* vandalism/good-faith. Justify your decisions, ideally by citing Wikipedia policy. Thanks, 19:22, 27 September 2012 (UTC)

Thanks for your advice! I will lean on WP:CVUA first and later request access to this tool.--Sriharsh1234 08:13, 28 September 2012 (UTC)

Andrew, I'm just letting you know that without prejudice to the Stiki request, I have declined a request from this user for the Rollback tool until they have successfully completed a CVUA course. Kudpung กุดผึ้ง (talk) 08:56, 30 September 2012 (UTC)

Will STIki allow me to rv an edit that would contravene 3RR?

Is there any logic in the program to remind or stop me from reverting an edit more than twice in a 24 hour period? If I make a few hundred RVs in a day I could easily forget whether I've encountered that page already in the last 24 hours. Thanks. Alanl (talk) 09:19, 6 October 2012 (UTC)

Technically speaking, STiki will allow you to break WP:3RR. However, note that the policy page says *vandalism reverts* are exempt from this rule. I suppose 3 "good faith" classifications might run somewhat afoul, but subjectivity abounds and I imagine most STiki "good faith" reverts could just as well be scored as vandalism.

My informal understanding is that 3RR is designed to prevent edit wars and legitimate content disputes (which rarely exhibit vandalism properties and thus STiki is unlikely to display them). STiki edits pertain to more objective matters of quality; and objective reverts seem to be outside 3RR scope. If somewhat has a reasonable qualm to pick, you'd expect a talk page message by 2RR methinks, which should heighten ones awareness for any future STiki use.

I'll ponder this for a bit, but unless someone disagrees with this reasoning then I don't see it as a high priority issue. A quick DB query suggests there is no possibility 3RR has ever been violated in STiki's history; and users like "Fraggle" were pretty epic users (1000s of edits daily). I would not be concerned. Does Huggle do any checks for this? Thanks, West.andrew.g (talk) 19:03, 6 October 2012 (UTC)

Constantly freezing

STiki has been constanly freezing just now. It can't be my laptop because it's fairly new (less than a year old). I haven't been able to perform many edits on it because of this. Thanks. Lighthead ^þ 06:35, 13 September 2012 (UTC)

I need a little more detail to track this to anything concrete. How did the problem manifest? Would the diff browser just not advance and the buttons were unresponsive? Did you try to restart? How many times failed? Is it working now? Were any error messages popped? Assuming the problem went away as silently as it came, I have a sneaking suspicion this might have to do with STiki's backup routine, which briefly locks a table (< 30 seconds?) while it dumps its content to an external drive. This won't break STiki, but you won't be able to delete/update from a table while locked (so it looks like a hang). While backups took just seconds when STiki started out, the DB is now on the order of 10s of GB. I probably just need to move the old and stable data to archival tables so the "active" portions spend less time under lock. Anyone else have problems last night? Or a problem that generally arising just after 6:00UTC? Thanks, West.andrew.g (talk) 13:38, 13 September 2012 (UTC)

The diff browser did not advance, check on that. The buttons were unresponsive, check on that. I actually didn't try to restart (I haven't been using it a long time, I'm not aware of all those tricks). It failed perhaps 5 times or more. I'm actually too busy to see if it's working now. I'll try later tonight. No error messages popped up. I had to open task manager to close it, and it didn't close right away. Hope that helps. Looking back I don't think I could have restarted it, because the menu folder buttons weren't responding either. Hopefully somebody else will let you know if they had any problems. Thanks. Also if you have any more questions, please ask. Thanks, again. Lighthead ^þ 23:21, 13 September 2012 (UTC)

I'm on mobile, so this shall be brief: I'll entend upon it later tonight (CST). But yes, Andrew- I've been experience very similar things at around 6UTC. Theo polisme 00:25, 14 September 2012 (UTC)

Added to bug/feature table as T#020. Thanks, West.andrew.g (talk) 22:25, 16 September 2012 (UTC)

I've encountering the same it starts up and after a few diff's to check on it just Freezes and i have to kill it via the task manager this occured first time after i updated my java runtime since then I'm not able to use stiki --Fox2k11 (talk) 01:29, 25 September 2012 (UTC)

* Fox2k11, Are you saying that you haven't had a chance to use STiki since then... or are you saying that STiki no longer works for you, now that you have updated your Java runtime? Yaris678 (talk) 09:39, 25 September 2012 (UTC)

* Moreover, is this a problem that only manifests around 6:00UTC each day, or this a problem that happens at all times of the day? I don't know how upgrading Java would negatively impact anything... What operating system do you use? Can you try to launch STiki from the command-line/prompt/terminal and see if any textual error messages are output? Thanks, West.andrew.g (talk) 16:06, 25 September 2012 (UTC)

the scenarios is as follows I start Stiki and when the first Diff shows up I judge over it by hitting the proper button and then stiki freezes and if I'm lucky after some seconds it continues to work or it still remains freezed and i have to kill it via the taskmanger when it continue to work after a few diff's it crashes I stumbled over this first time as mentioned when i have updated my Java runtime to the current version! I'm using Windows 7 Ultimate 32-bit w/ SP3 and Current Java version I have to say it worked before the update! currently I'm using Huggle because a proper work with stiki is currently not Give! Fox2k11 (talk) 20:27, 25 September 2012 (UTC)

Do you have the most recent version of STiki? Can you launch STiki from the command prompt (similar to the UNIX/Linux launching instructions) and see if any textual errors get output when these freezes occur? Thanks, West.andrew.g (talk) 15:32, 28 September 2012 (UTC)

Yes I have the Current Version and I started STiki from the console and the log ouput can be found Here --Fox2k11 (talk) 17:56, 28 September 2012 (UTC)

Thanks for the trace, Fox. Things seem to go south around line 275. I am curious, you above mentioned that it "periodically freezes"? Did you see any edits when you produced this trace, or did things just fail immediately? It seems to be a database connection issue, but I couldn't find a single database Java class getting loaded before the errors (thus it would seem a connection was *never* established). You shouldn't have been able to see an edit without talking to the database first. Let's start with this... Thanks, West.andrew.g (talk) 14:19, 29 September 2012 (UTC)

Well it starts up I log-in and see the first DIFF as mentioned Above I click the proper Button and that button stays "pressed" as i mentioned sometimes I'm lucky and it just works after a few seconds and then i can work with it for a sporadic amount of times (sometime 2 minutes sometimes 5 and so) if I'm not so lucky the button just stays Pressed and STiki stays Freezed and i have to terminate it via the taskmanager and thats it! I made also a screenshot for you! http://img696.imageshack.us/img696/7538/stikicap.jpg --Fox2k11 (talk) 02:52, 30 September 2012 (UTC)

Greetings and sorry for the delay. I just spent a bit reading the traces for some of your recent interactions with the program. Something that struck me as odd was that most of the sessions (including short ones) tended to end with calls to the "queue_wipe" procedure, which is typical of proper and bug-free program exits (this routine basically frees up any edits that were enqueued for you to view, so others can get them). So once a classification button freezes and you start clicking "X"/exit, those exit messages are getting through to the STiki server, indicating that the database connection hasn't been completely severed. The exit, however, fails because the program is still stuck on some thread, seemingly as the result of a database action. I do find it odd the "SQLIntegrityRestraint" exception showed, given that I have only seen that error when, i.e., one tries to insert a primary key that already exists.

So at this point I am still a bit stumped. These troubles are happening on the same network as with your previous STiki experience? Is it residential and reasonably fast/reliable? I might write up a little debug/test routine to try to delve a little deeper and have you run it for me. Thanks, West.andrew.g (talk) 16:13, 5 October 2012 (UTC)

Hi Andrew as I mentioned before I had no issues with STiki *Before* I updated ma java runtime I read about a security hole in java 6 (wich was still on my system..I'm a bummer when it comes to updates) so i downloaded the current java install file (Java 7) and the problems started as mentioned above and still exist and it seems I'm not the only one --Fox2k11 (talk) 17:18, 5 October 2012 (UTC)

Hi guys, I've recently come back to Wikipedia after a break and downloaded the new version of STiki and I'm experiencing similar problems that weren't apparent in the older versions that I was using back near April. First it froze like this: http://i.imgur.com/eJjz6.jpg - the edit did show up in my contributions. Then this happened: http://i.imgur.com/KrhzX.jpg. S D 5 18:23, 6 October 2012 (UTC)

Welcome back. Can you tell us what version of Java you are running (and your OS)? Do you recall if Java updated during your break? With multiple users now experiencing this, I'm definitely motivated to track this issue down. Thanks, West.andrew.g (talk) 18:29, 6 October 2012 (UTC)

Hi Andrew - Windows 7 64 bit, and the version of Java is 6 (update 35). No idea if it updated, though I presume it did do some minor version updates. Thanks S D 5 18:36, 6 October 2012 (UTC)

Are your experiences similar to Fox's in that the tool is basically un-useable, with sessions never surviving more than a couple minutes? Thanks, West.andrew.g (talk) 19:52, 6 October 2012 (UTC)

And if they are, could you also run from terminal and see if there is any output (and produce the console trace). I'm hopeful they'll be identical to Fox's so at least we know we are looking at a consistent problem. Thanks, West.andrew.g (talk) 20:16, 6 October 2012 (UTC)

I was only able to make a few edits (all of which froze the program briefly) before a total freeze. Typically now that I'm running from console to get the trace the program is behaving flawlessly with no freezing! S D 5 20:38, 6 October 2012 (UTC)

same happened to me when i run it from terminal this is weird... --Fox2k11 (talk) 20:58, 6 October 2012 (UTC)

Keep me updated. It could also be a temporal issue (one time or recurring). If you switch to non-terminal start-up now; I have to assume it would also be working, yes? West.andrew.g (talk) 21:26, 6 October 2012 (UTC)

For me it now does work fine on a "normal" startup, with the exception of a ~15 second freeze when changing the rev queue from Cluebot to STiki. S D 5 21:41, 6 October 2012 (UTC)

I just gave it a test run and after a Initial Freeze of a few seconds I can now work with it... maybe you where right and it was a Temporal (DB?) Glitch?! --Fox2k11 (talk) 21:48, 6 October 2012 (UTC)

Glad to see things seem to be working out for you both. Yes, the initial delay will tend to happen (though 15 seconds is quite a while; it depends on your network speed and where in the world relative to STiki's servers in Philadelphia, USA). This is because everything has to be re-fetched whenever the queue is changed or you login (e.g., to avoid those edits you previously "passed"). Hopefully after that the queuing and pre-fetch system makes for rather responsive operation. Let me know if these problems happen in the future, and if they do, try to produce a very precise time of exactly "when". Thanks, West.andrew.g (talk) 17:34, 9 October 2012 (UTC)

feature request, request for usage

hi andrew, i'm a very long-term editor on the project although have been editing only sporadically since 2006. i started going through my watchlist today and noticed quite a large number of simple vandalism or test edits (nothing too offensive, really, just cruft). i downloaded stiki and ran it and was surprised at how effective the algorithm(s) are at detecting this sort of stuff and decided it would be nice to help with the development of the project (that is, stiki, not wikipedia). i would hope that you'd be willing to "grant me special permission" or whatever to use the tool. i've some thousands of article edits (slightly > 50%) since 2004-ish. a further third are interacting with other editors trying to fix things.

the other thing i'd mention is that i tend to keep myself through tor at home because i'm a paranoiac. this means that my traffic to sites i trust goes through ssl and bypasses the proxy (tor). unfortunately i don't see a way to tell stiki to use https instead of straight http. since we're communicating a cookie and i assume the network i live upon to be hostile, it would make my life easier and me feel safer. anyways.

slick tool. (honestly, i'm a perl guy and want to not like it, but it does work well!) –alex (... aa:talk 03:34, 5 October 2012 (UTC))

Hi Alex. First off, *you* do not need any special permission to use STiki. Because you have plenty of main namespace / article edits, you automatically meet our qualification criteria. The permission requests floating around this page are from less experienced editors. This is besides the point: I see you used STiki today to view at least 5 edits (however, your edit summaries don't indicate you used STiki to undo them? did you manually revert the edit in-browser and then classify them in STiki after the fact?)

STiki does have an option to use HTTPs communication. You can find it under the "options" menu; I believe it is disabled by default. Thanks and happy editing, Thanks, West.andrew.g (talk)

so the problem is if it tries to revert over tor, it fails because of the enwiki policy of no-tor-period. hence my request to use https. i switched over to https, and it could not log in. when i run over the regular internet, no-tor, everything is fine with http/80, but https/443 fails to log in (the java just hangs). not sure what the issue is. it also doesn't seem to be making the actual edits, if i look at my edit history. ... aa:talk 19:31, 5 October 2012 (UTC)

huh, okay, so it is now. interesting. ... aa:talk 19:31, 5 October 2012 (UTC)

okay, so i guess two things are happening. one seems to be database lag. the other, and this is a guess, is that other users are seeing the same edits and reverting them before i decide and mash a button. ... aa:talk 19:46, 5 October 2012 (UTC)

Couldn't parse the language of this chain of events perfectly, here's what I can respond: (1) No, you won't be able to use Tor. (2) Did the login issues fix themselves, it works for you over both http/https now? (3) Database lag (on the Wikipedia) side is out of my control, can't do much about that. (4) Pay attention to the "last revert panel" in STiki. Threading allows the GUI to advance while communication with Wikipedia happens in the background. When a revert eventually commits, this panel will report that. If an editor beat you to the revert, the panel will also report that. Thanks, West.andrew.g (talk) 19:12, 6 October 2012 (UTC)

Nope, I can't get https to work at all. Is this confirmed working on the Mac? ... aa:talk 00:53, 7 October 2012 (UTC)

I don't know which STiki user's use HTTP vs. HTTPS or what operating systems they employ. I can confirm that it works just fine on my Linux box and I believe Window's users have also had success. The program just hangs? If you start it from the terminal/command-line, is there any output or just an infinite hang? Thanks, West.andrew.g (talk) 18:30, 9 October 2012 (UTC)

6:00-7:00+ UTC

I'm running STiki in Wineskin on a Mac, and around 6-7 the database seems to lag an exorbitant amount of time... have you already offloaded data/whatever the original plans to speed it up were to do? I have to wait 5-10 minutes to see the first diff, and after hitting a button I have to wait another 2-5 minutes to see the next diff... Cheers! Theo polisme 07:07, 11 October 2012 (UTC)

Hmmm.. I enabled a [mysqldump] (the backup command) option that does not lock tables and allows for concurrent querying while the backup occurs. It seems that maybe the backup thread is still taking priority over front-end users. I'll look into this and report back. Thanks, West.andrew.g (talk) 18:05, 11 October 2012 (UTC)

Problem should now largely be eliminated. Let me know if any further issue arises. Thanks, West.andrew.g (talk) 16:02, 16 October 2012 (UTC)

You don't need wineskin to run STiki on mac. It's java so it's native to both Mac and Windows. FXtraderFX (talk) 23:05, 17 October 2012 (UTC)

Request for approval

I am requesting permission to use STiki to help fight vandalism. I have 14,000+ edits, have reverted quite a lot of vandalism by patrolling recent changes and I have just requested rollback rights. Many thanks. – MrX 01:29, 18 October 2012 (UTC)

Hello MrX, as you have over 1000 article edits you are automatically granted permission to use STiki. Thanks Fraggle81 (talk) 01:57, 18 October 2012 (UTC)

Thanks Fraggle8! – MrX 02:02, 18 October 2012 (UTC)

Request for Approval

Hello STiki team,

I have been using Wikipedia for a while now. I switched accounts and would like to regain my STiki privileges. My old username is Jakebarrington. Thanks, FXtraderFX (talk) 23:08, 17 October 2012 (UTC)

I noticed someone (i.e., you) has been active on the Jakebarrington account today. Just for simple confirmation that it is "you", can you login to that account and make an edit/confirmation in this thread? Then, how do you want to handle the STiki statistics? Is this new account permanent (transfer stats?) or would you the like two to persist separately. Thanks, West.andrew.g (talk) 00:38, 18 October 2012 (UTC)

Thanks for the reply. Originally I was going to have Jakebarrington just be for edits and improvement and have FXtraderFX for anti-vandalism, but I have decided against that. I will just stick with Jakebarrington. Thanks for the help! Jakebarrington (talk) 13:51, 19 October 2012 (UTC)

Done -- matter closed. Congratulations on your recent barnstar and good luck in future editing. West.andrew.g (talk) 16:29, 19 October 2012 (UTC)

STiki doesn't open at all

Stiki these days refuses to open up at all, and shows up a box saying either I don't have internet, have no MySQL access or a new version of Stiki has been released. This is happening across multiple systems with different internet connections. --Rsrikanth05 (talk) 20:55, 16 October 2012 (UTC)

Greetings. I notice the message on your user/talk page which indicates recent power outages. Given that no one else is experiencing this problem (?), I suspect that could be the issue. What happens if you try to go this web address? What about [5]? Could regional ISPs also be doing some port firewalling until everything is stable and back online? Let's begin the troubleshooting with my above questions. Thanks, West.andrew.g (talk) 21:52, 16 October 2012 (UTC)

Err, the power outages is at my place of study. Right now I'm at home and I have no issues. The links provided work fine, it redirects to your University page. I'll check my Firewall settings and get back. Thanks. --Rsrikanth05 (talk) 12:07, 18 October 2012 (UTC)

Yes, the first link should have redirected to my webpage. The second one should not go there (as it attempts to connect via port 3306, the same one used by STiki's database/MySQL connection)? Do you see my webpage on the second one as well? Thanks, West.andrew.g (talk) 16:31, 19 October 2012 (UTC)

The second one says "<clip>Non-ASCII nonsense</clip>". Is this supposed to be what I see? --Rsrikanth05 (talk) 12:22, 20 October 2012 (UTC)

Yes, seeing that message would suggest traffic is probably being allowed through that port. STiki continues not to open? What OS are you running? Did it suddenly just stop working for you one day? No system changes that might have initiated this change? Thanks, West.andrew.g (talk) 15:58, 20 October 2012 (UTC)

I'm using Windows XP. I just did a Java update and I think that has solved it. It's working again now. Thank you!. :). --Rsrikanth05 (talk) 17:23, 20 October 2012 (UTC)

Done Matter closed. Glad to hear everything is fixed. Thanks, West.andrew.g (talk) 22:54, 20 October 2012 (UTC)

STiki reverting when it shouldn't

Sometimes edit the article in the wiki interface, instead of reverting the edit with the STiki interface. When I go back to the STiki interface I usually press the button that best approximates what I have done. For example, if the edit found was vandalism, but there was other vandalism, I tend to remove all the vandalism in the wiki interface and then press "vandalism" in STiki. The STiki interface should just inform me that I have been beaten to the revert (by myself) and make no edit. This has all worked fine .... until now.

Check out this edit!

STiki is reverting the edits by 194.75.169.163, to give the version by CBNG. (confirmation). This despite the fact that I had already reverted to an older version using Twinkle.

This is obviously a bug that doesn't crop up very often but I thought you should know.

Yaris678 (talk) 12:07, 19 October 2012 (UTC)

Hmmm... and this one isn't as bad but may be related. I have just done this edit and then pressed "Good Faith Revert" to see what happened. It didn't do another edit (which is good) but it did say "Undid 0 edits / unknown error / check page hist" Yaris678 (talk) 16:16, 19 October 2012 (UTC)

Okay, here is why this occurs... All actions taken using the "native rollback" functionality are marked as minor. Rules state that AGF edits must *not* be marked as minor. Thus we can never use native rollback to perform AGF. Thus, I implement "software rollback", which in theory has the same effect, but uses some more internal logic and different API functionality (and bandwidth). If we only need to "software rollback" 1 edit, we can use the basic *revert* functionality. However, if we need to "software rollback" 2+ edits, I basically go back two edits, get the entire wikitext of that version, then make a "full text edit" that replaces all content in the existing version. Whereas native rollback and revert have some nice integrity checking built in to make sure the edit chain is stable (i.e., no conflicting edits while all this was happening), the same cannot be said of the "full text" case, This is precisely the branch you went down in the first case. An explicit check will be added to remedy this.
The second case you report took us down the path of using software rollback, but simply performing a *revert*. Here the constraint caught your intermediate edit and did not allow it to commit. In this case (and for any edit attempt), the API kicks back one of three responses -> SUCCESS, NOCHANGE, or FAIL. I have been mapping the "nochange" to the "beaten to revert" message. I map the "fail" one to the message you report above of "unknown error". Because you didn't do an identity revert it didn't trip the "nochange" message. I think I just need to add a line into the "fail" message that also accounts for edit conflicts.
Done -- TL;DR = One bug fix implemented to be rolled out in the next release fixing an AGF<->intermediate-edit corner case. One message updated to better reflect what is going on behind the scenes. Thanks, West.andrew.g (talk) 17:13, 19 October 2012 (UTC)

Thanks Andrew. Presumably people will get the same issue with vandalism reverts if they don't have the rollbacker right. --Yaris678 (talk) 10:33, 20 October 2012 (UTC)

Yes, but the same fix handles/handled this as well. Thanks West.andrew.g (talk) 16:00, 20 October 2012 (UTC)

Cool. Thanks. Yaris678 (talk) 05:58, 21 October 2012 (UTC)

Leaderboard

This is a really minor issue, but the leaderboard uses a rather excessive amount of RAM and CPU loading on my computer (~400MB and a full thread for ~30 seconds respectively), which is far more than it should. Perhaps it would be prudent to cut it down a bit, so that users with less than 50 or 100 users aren't displayed? Or, similar to Wikipedia: List of Wikipedians by number of edits, list them on another page? It's not a huge issue right now, but it's only going to keep growing, and 400MB RAM usage is already a fairly large amount. Sellyme^Talk 17:29, 20 October 2012 (UTC)

I'll be frank in saying this is not a priority (nor do I like the idea (for reasons described); although I guess an alternate implementation is possible). I presume the main reason the 200k of wiki-markup blows up to ~400MB in your browser is because the table is sort-able in so many dimensions. This sort-ability is critical. I (and others) pay little attention to who is actually atop the leaderboard. Instead, I sort it by "last use" and "user since" fields to see how many users are recently active and as a means to audit new user quality. Fragmenting the table over multiple pages would destroy this functionality in all dimensions and hide away precisely those new users. What browser are you using? and are you on constrained hardware? FWIW, it loads, parses, and re-sorts at a latency that isn't even human-recognizable on my machine. Thanks, West.andrew.g (talk) 22:53, 20 October 2012 (UTC)

You're right in that the sortability is fairly important, and the large issue with the RAM usage. I'm running on an AMD A8-3870K quad-core CPU @ 3.0GHz, with 2GB of DDR3 RAM. I'm running Firefox 16.0.1, and it should be noted that the 400MB figure is the amount of excess RAM on top of Firefox's usage without the page open, not the usage of the entire browser. Perhaps users who have under a certain amount of uses and are inactive could be trimmed? That would maintain all of the usability of the list while indefinitely preventing it from becoming overly large. Sellyme^Talk 23:31, 20 October 2012 (UTC)

Maybe a good compromise would be making a sub-page of inactive STiki users, who haven't used the tool in 365 days. This would take a significant chunk out of it. The sub-page would keep the same order of users, from most edits to least amount, but it would only be for the inactive users. Regards, — Moe Epsilon 06:08, 21 October 2012 (UTC)

For those that have scalability issues, I'll ask: "How do you use the leaderboard?" Is it mainly about checking your own position? Thanks, West.andrew.g (talk) 06:43, 21 October 2012 (UTC)

I primarily use the leaderboard to compare myself against other active users, to see what my usage of the tool is representative to the majority. The vandalism and AGF reverts of inactive users of over a year ago are with extremely different data sets than what they would be now, so they aren't really comparable, and as such I pretty much ignore anyone who hasn't used the tool recently (unless I'm comparing historical data). I also use the edits/day figures to gauge what is considered a fairly decent contribution, so I can give myself a "target" to try to maintain. In the past I've often promised myself to do a ton of work on Wiki projects and the like and got bored after a day or two, so having a goal of sorts makes it a lot easier to give a consistent contribution. Sellyme^Talk 14:29, 21 October 2012 (UTC)

STiki's server down [update: fixed]

STiki's server appears to be down, as it is currently unreachable via SSH and other services. I will be heading into my office shortly to investigate the issue. Thanks, West.andrew.g (talk) 22:19, 21 October 2012 (UTC)

Done And after a trip to the office, it should be up again. Some silliness about WICD (the network manager) failing to attach to the daemon bus. Seems to be an isolated incident. Thanks, West.andrew.g (talk) 01:24, 22 October 2012 (UTC)

STiki on new computer

Hi, i have downloaded Stiki on a new computer and i am trying to open the jar file but it says windows cant open it Do you know why this is? Telfordbuck (talk) 16:18, 19 October 2012 (UTC)

Do you have Java installed? -test link-. Thanks, West.andrew.g (talk) 16:47, 19 October 2012 (UTC)

I've installed but isnt working , it says the file (java that is) is corrupt. Telfordbuck (talk) 19:20, 19 October 2012 (UTC)

As in the executable you downloaded from the Java folks is corrupt (not the STiki JAR)? If that's the case, its a little outside of my jurisdiction. Ensure you downloaded the executable for the proper OS/platform and maybe give it a second download attempt. Thanks, West.andrew.g (talk) 22:38, 20 October 2012 (UTC)

Nope still not worked. Oh well getting another laptop soon anyway :) Telfordbuck (talk) 20:58, 21 October 2012 (UTC)

Java is installed now, and its saying stiki isnt working. and coming up with the 4 reasons. Telfordbuck (talk) 21:09, 21 October 2012 (UTC)

See the below section about the STiki server being down. I believe the server might have been out of order at the time you posted this message. Give it another try and let me know if it works for you. Thanks, West.andrew.g (talk) 02:16, 22 October 2012 (UTC)

Yep working now :) Telfordbuck (talk) 10:37, 22 October 2012 (UTC)

Request for Approval

Hi there, I've been reverting vandalism for the past few days and found this tool. I think it'd be so much easier for me to revert and I'd love to use a tool such as this, even though I don't have rollback. Please consider giving me access as it'd be so much easier on my part. Thanks, Hair^Talk 00:56, 23 October 2012 (UTC)

Done Permission granted. A decent body of revert work here on en.wiki combined with Wikia and technical experience just seems to make sense in this case. Happy reverting. Thanks, West.andrew.g (talk) 18:16, 23 October 2012 (UTC)

Request for Approbation

I'd now like to use my STiki for faster vandalism reverts. Have rollback, have more than a thousand edits, but no approval yet. TruPepitoM Talk To Me 02:07, 29 October 2012 (UTC)

As you have rollback rights no approval is necessary. You should just be able to download STiki, log in and start using it. Fraggle81 (talk) 02:22, 29 October 2012 (UTC)

I would like to draw attention to the issues on TruPepitoM's talk page. This user has had rollback removed for edits using Huggle and STiki. Fraggle81 (talk) 15:08, 31 October 2012 (UTC)

Request for Approval

Hi there I would like to help fight vandalism and I try to revert vandalism when I can and I think it would make it much easier for me to fight vandalism with this tool and I'd love to use a tool such as this, even though I don't have the rollback permission/right, nor do I have at least 1000 article edits I only have 445 edits, but I do try to be on Wikipedia everyday when I can and hope to get over 1000 article edits in the next few months if possible, so please consider giving me access as it'd be so much easier on my part. Thanks. Sk8terguy27 _Talk 23:10, 30 October 2012 (UTC)

Greetings Sk8ter guy. I think you have good intentions in coming here, but you don't have a ton of article namespace edits, and of greater concern, your talk page and its archives seems dotted with prior controversy and warnings. I think it would be best if you visit and enroll at WP:CVUA whereupon after completing a couple of lessons, your account would be fast-tracked for STiki approval. Thanks, West.andrew.g (talk) 18:35, 31 October 2012 (UTC)

Request for Approval

Hi, I'm Vacation9. I would like to be able to revert vandalism easier (I already use Twinkle and have made around 200 Article Namespace vandalism/speedy deletion edits). I think STiki would be a great tool to do this and believe I am worthy of holding this tool. Please let me know your decision and be free to look at my talk page. Vacation9 (talk) 02:56, 1 November 2012 (UTC)

Done. A decent (although recent) body of anti-vandalism and other anti-damage work inclined me to approve here. Make sure you read WP:VANDALISM and WP:NOTVANDALISM. If in doubt, classify conservatively. Thanks, West.andrew.g (talk) 03:09, 1 November 2012 (UTC)

Thank you! I reviewed these pages and am ready to fight vandalism efficiently! Vacation9 (talk) 06:08, 1 November 2012 (UTC)

Wiki-Defcon

Hey, In-case anyone was wanting to use STiki to update Wiki Defcon here is a handy dandy chart to put the Stats up STIKI USERS: You'll need to do some math in order to garner the Reverts/Minute. Do the following: - In STiki, click "Rev. Queue", then "Recent usage stats." -Next, note the number of classifications (total, from all queues) from the last hour. - Multiply this number by the percentage of reverts (in parentheses). - Round to a whole number, and then ?? multiply it by ??? in order to adjust for less use of STiki vs. Huggle. - Divide your result by 60 (minutes in an hour). ♠♥♣Shaun9876♠♥♣ ^Talk 02:35, 5 November 2012 (UTC)

Hi Shaun — I think you misunderstood me at Template talk:Vandalism information...if you read the above (copied from my post at the template talk), you'll see the ?'s — we still need to determine how to adjust for Stiki<Huggle (at least as far as usage goes). For now, this is rather pointless - the key step is missing! :) —Theo polisme 00:57, 6 November 2012 (UTC)

AIV Reporting

Hi all (particularly Andrew). Any explanation for why STiki made this report? (see why it was declined)? —Theo polisme 22:18, 12 November 2012 (UTC)

Am I missing something, or did someone at AIV drop the ball here? The diff you provided goes to a link made Nov. 12 (today). The "last warning" on the user's talk page was issued on Nov. 9?? Thanks, West.andrew.g (talk) 22:50, 12 November 2012 (UTC)

Thank goodness, I was afraid STiki was losing its touch! [must trust technology...must trust technology...] :) —Theo polisme 00:18, 13 November 2012 (UTC)

Apologies. I had looked only at the time and missed the date of the last warning. Daniel Case (talk) 03:02, 13 November 2012 (UTC)

Next release?

Hi Andrew. When will the next release be? You did a bug fix a while ago but it hasn't been released yet. Without the fix, STiki users could potentially revert to the wrong version. Yaris678 (talk) 17:33, 13 November 2012 (UTC)

It is in the works. I have accumulated a couple of minor fixes that have already been implemented. I just tend to wait for some larger front-end change so end-users can have some tangible benefit that encourages them to update. Whether that "large fix" comes along soon or not remains to be seen (Thanksgiving, maybe?). Regardless, I will push an update around that time-frame -- as I have written code that enables the STiki server to be easily migrated (with my upcoming graduation, the server may no longer live at Penn, and infrastructure will permit the server to move with minimal impact to GUI users). I'd like to have this "migration enabled" version in the hands of users long before it becomes necessary. Thanks, West.andrew.g (talk) 21:18, 13 November 2012 (UTC)

Special approval

Any chance I could get Stiki approval? Fighting vandalism through recent changes is tedious. Thx. little green rosetta(talk)
central scrutinizer 20:26, 19 November 2012 (UTC)

I was a little torn on this one. You seem to have done some valuable but very narrow work on Wikipedia so far. Anti-vandalism work did not seem to be a significant part of that experience. However, I've taken this as a sign of good-faith and enabled your access. Reading WP:VANDALISM, WP:NOTVANDALISM, and WP:STiki is a must. Be cautious or you'll find yourself in quick trouble (but you seem to hang around ANI and what not, so you seem aware of the pitfalls). Thanks, West.andrew.g (talk) 21:27, 19 November 2012 (UTC)

No need to worry. I'll be using Stiki to remove stuff like "fart" edits from articles. Thanks. little green rosetta(talk)
central scrutinizer 03:06, 20 November 2012 (UTC)

Where has the time gone?

1,000 days of Stiki ... congratulations Andrew!
Cheers! -- Gareth Griffith-Jones/The Welsh Buzzard 08:12, 23 November 2012 (UTC)

CHANGELOG for the 2012-07-18 release

Nothing earth-shattering in this release. However, with no pressing work on the horizon I though it wise to go ahead and push out these few minor improvements we've accumulated lately. As always, thanks for your continued support. West.andrew.g (talk) 20:46, 18 July 2012 (UTC)

A "recent usage statistics" option has been added to the "Queue" menu. This reports basic numbers about STiki/queue use over the last 1/6 hours. As high rates of use can cause queue exhaustion and lower hit-rates, users can interpret these statistics to determine how best to allocate their STiki usage time and queue choice (T#014).
Minor changes to the "excessive pass use" and "don't template the regulars" message dialogues per discussion at Wikipedia_talk:STiki (see archives).
The message displayed in the diff browser that indicates rollback situations (i.e., 2+ edits are encapsulated by the diff and would be reverted) has been revised per Wikipedia_talk:STiki (see archives).
The default comment left with AGF reverts has been changed per a talk page request from "good faith revert..." to "reverting good faith edit(s)..." Custom AGF messages will need to be re-authored.

Approval to use STiki

Hello, I'd like to ask for approval to use STiki. The majority of my edits are anti-vandalism edits, however, I think I'd be more efficient using STiki over the default editing tools and Twinkle. Fox Wilson (talk) 02:17, 28 November 2012 (UTC)

You seem to have a good base of anti-vandalism work (with no complaints). I've approved your account. Read WP:STiki. Thanks, West.andrew.g (talk) 02:25, 28 November 2012 (UTC)

Requesting STiki permission

Hi. I would like permission to use STiki. I have 241 edits on Wikipedia, and most are vandalism reversion and warnings. I realise that 241 is somewhat under the limit of 500, but I do have 790 changes on Simple English Wikipedia that are mainly vandalism reversion. I've recently become more interested in the English Wikipedia, partly because there's so much more vandalism to revert...

Manie sans delire 00:41, 1 December 2012 (UTC)

What he says is true, and technically it isn't a limit but a reasonable guideline... §h₳un 9∞76 01:07, 1 December 2012 (UTC)

Well, technically speaking, it's she... But yes, I do agree that it's just a guideline that's set up to make sure that the user understands vandalism.

--Manie sans delire 01:20, 1 December 2012 (UTC)

STiki access has been approved for User:Maniesansdelire. Happy editing! Thanks, West.andrew.g (talk) 08:44, 1 December 2012 (UTC)

T#017

I support this request. And the last editor's comment line could be longer—it gets truncated. --Greenmaven (talk) 01:06, 1 December 2012 (UTC)

Its already been implemented and will be pushed at the next release (see below). Thanks, West.andrew.g (talk) 18:48, 2 December 2012 (UTC)

Next release?

It's been quite a few months; is a new release in the pipeline? It seems we've accumulated enough bug fixes/changes/etc...then I again, I may be misjudging things again. —Theo polisme 02:17, 2 December 2012 (UTC)

There was a just-archived thread along these lines. A new release should be pretty eminent, but if my memory serves me correctly, it is primarily bug fixes and back-end stuff -- which do little for the typical GUI user. Some of these new AIV/AGF dialogues that have been discussed are on my radar, but I haven't found the time to implement them yet. Thanks, West.andrew.g (talk) 18:52, 2 December 2012 (UTC)

The client-side thing I'd like to see rolled out is the change in the way STiki deals with good-faith reverts and reverts by people without the rollback "right". Without this change, people could be reverting to the wrong version of an article. Yaris678 (talk) 12:51, 4 December 2012 (UTC)

Nothing too impressive, but I've decided to go ahead and push the release. Thanks, West.andrew.g (talk) 17:27, 4 December 2012 (UTC)

Thanks. Yaris678 (talk) 17:36, 4 December 2012 (UTC)

Navigation through prior diffs

Sometimes Stiki appears to "skip" a diff (the screen flashes very quickly). I can't tell if it is actually skipping a diff for me to review or not. My fear is I think that sometimes Stiki might be capturing a double "I" and marking a diff innocent before I actually had a chance to review this. When this happens for G or V, its no big deal as I can look in my contribution history and manually check the edit there. But is there a way to scroll through all the diffs I have previously reviewed? Thanks. little green rosetta(talk)
central scrutinizer 16:58, 29 November 2012 (UTC)

So when you review these seemingly double-clicks for "G" or "V", is it the case that two edits were actually classified? In the "I" case, you could always use the back button to see if there was an intermediate edit you missed. Assuming there are no bugs here and this is just human perception/error, I imagine some type of viewable "action-log" might fulfill what you are looking for? Thanks, West.andrew.g (talk) 17:04, 29 November 2012 (UTC)

If you are talking about multiple edits, then I have no clue, as the screen moves by too quickly. Ah, I just noticed the back button now! The log is actually a great idea, but a hot key for the back button wouldn't hurt as well. Thanks! little green rosetta(talk)
central scrutinizer 18:10, 29 November 2012 (UTC)

Back button has hotkey "B". It can only be used if the previous edit was classified "innocent" and you can go back at most one edit. You could use this to determine if the screen "flicker" is actually an edit being skipped. I'll put the log functionality on the longer-term TODO list. Thanks, West.andrew.g (talk) 18:44, 29 November 2012 (UTC)

Addition to what Andrew said: The "back" function can also be used if an edit had not been classified and the user pressed the "pass" button. Yaris678 (talk) 12:43, 4 December 2012 (UTC)

Also, in response to the idea of an 'action log': perhaps some sort of undo (one click for both warning and reversion undo—could have some sort of basic error handling if revids didn't match up (i.e. a link to the history)) button? Somewhat like what Huggle has? Just throwing that out there. —Theo polisme 22:05, 11 December 2012 (UTC)