Forum:Charm logs: A nightmare to maintain

From the RuneScape Wiki, the wiki for all things RuneScape
Jump to: navigation, search
Forums: Yew Grove > Charm logs: A nightmare to maintain
Archive
This page or section is an archive.
Please do not edit the contents of this page.
This thread was archived on 16 June 2010 by Calebchiam.

Today I saw that blue charm drops for Dust Devils was way off, so I started going back one edit at a time to find where it went crazy. About 75 edits and 2 months into the past I find the obvious vandalism. Next I have to do the math to find out exactly what he added to the log so that I can remove that number without reverting all the valid additions to the log in the past 2 months.

What a pain to revert one bad edit! And it seems like half the edits to charm logs are vandalism. I'm proposing some changes to make charm logs easier for us to maintain.

  • Block IP editing of charm logs
    • There's no way to have any faith that the logs are real if the people submitting them are completely anonymous.
  • Lock the charm logs down after a certain number (2,000? 5,000?) of monsters killed
    • 27,000 logged kills seems like overkill to get a reasonable percentage of charm drops.

--Wowbagger421 01:40, March 30, 2010 (UTC)

Discussion

Support - I am aware that we are supposed to all be equil, and the idea of locking certain pages to users only is frightening, but the charm logs have been worrying me quite a bit as well. The above measures, I imagine, would cut down on this. Magic-icon.pngStelercusIlluminated Book of Balance.png 01:42, March 30, 2010 (UTC)

Strong oppose - As a stalker of RecentChanges, I can attest that only a small percentage of IP edits to charm logs are vandalism (maybe 10%). The overwhelming majority of IP edits to charm logs are constructive. Also, with many active users patrolling recent changes, it's much less likely that charm log vandalism will sneak into the wiki.

Furthermore, the wiki is supposed to be a place where everyone may edit (per this). There is no reason to protect the charm logs, since the protection policy does not cover this. Locking IPs out of the charm logs just doesn't seem right, especially when they are the ones who update them the most. --LiquidTalk 01:47, March 30, 2010 (UTC)

Comment - I was curious as to whether your 10% estimate was accurate so I just finished doing a case study. The King Black Dragon charm log. In total it has had 27 edits. Of those 27, I was able to identify 12 as vandalism by being completely impossible or just spam. There were a total of edits by IP addresses that seemed legitimate additions to the log. Every incidence of vandalism was by an IP address. --Wowbagger421 02:20, March 30, 2010 (UTC)
The King Black Dragon page probably receives more vandalism than some other pages would, since it is a boss monster. Very few people are going to keep track of the charm drops received over at least 50 kills (as suggested by the update link to the charm logs), so there will be only a few legitimate updates. I'll conduct a case study of a monster more commonly killed, like the Lesser demon. --LiquidTalk 02:39, March 30, 2010 (UTC)
Examining [[Lesser demon/Charm log]], I found 27 updates, 5 of which were vandalism. Check up on my work, since I did it relatively quickly. Most of those updates (~60%) were by IP's, and 4 of the vandalism instances were by IPs. Most of the edits by IPs are constructive. --LiquidTalk 02:45, March 30, 2010 (UTC)
That's a bit trickier to verify since lessers only drop one charm at once. But I'll try checking your work. --Wowbagger421 02:50, March 30, 2010 (UTC)
This log is a huge mess and I'm pretty sure several incidents of vandalism were never reverted. --Wowbagger421 03:11, March 30, 2010 (UTC)

Oppose - Partially per Helium, but I think cutting charms logs off at a certain number is stupid. You can never be too precise in my opinion. Vandalism is rarely gotten away with now anyways, as we have more and more users patrolling recent changes, ready to pounce on any would-be vandalizers. HaloTalk 01:55, March 30, 2010 (UTC)

Comment - This one example of very blatant vandalism lasted for two months before anyone noticed. If the vandal had been slightly more subtle, I doubt anyone would have ever caught on. --Wowbagger421 02:00, March 30, 2010 (UTC)
This is one example. Cases like this are exceedingly rare. The benefits of letting IPs update charm logs greatly outweigh any negative consequences, since if you look through the logs, mostly IPs update them. --LiquidTalk 02:01, March 30, 2010 (UTC)
Exceedingly rare? Is it just a coincidence that you just reverted a similar, month old case on the lesser demon charm log? --Wowbagger421 03:43, March 30, 2010 (UTC)

Comment At some point we are moving the logs to where only a certain bot will be adding the data after it checks it for accuracy. See here Forum:Preventing charm log vandalism--Degenret01 02:06, March 30, 2010 (UTC)

Support capping logs- i support that when it reaches 10,000 kills that it gets locked because it won't change. Hunter cape (t).png Sentra246Blue hallowe'en mask.png 04:42, March 30, 2010 (UTC)

Comment - There will hopefully be a new charm log system implemented soon through TLULbot (and I know I keep saying this and not delivering, but I should be ready to test both parts of this by the end of the week). This could be used to implement capping (which I'm Neutral on), but also would be performing validation on the data before allowing it into the logs, meaning the IP edits could be checked. It would be possible to be less flexible when it comes to IP edits. See Forum:Preventing charm log vandalism. I'm a regular user and I approve this message.  TLUL Talk - Contribs 07:47, March 30, 2010 (UTC) 

Oh pshh, Degen already linked it. I'm a regular user and I approve this message.  TLUL Talk - Contribs 07:48, March 30, 2010 (UTC) 

Comment - Just so everyone knows, if this passes, then we will be using RS:IAR to override Forum:Previously rejected proposals - READ THIS BEFORE YOU POST, RS:AGF, and RS:AEAE. This is the main reason that I'm opposed to it, since AGF and AEAE are the cornerstones of Wiki philosophy. --LiquidTalk 19:29, March 30, 2010 (UTC)

Typically "No original research" is a cornerstone of Wiki philosophy, but by having charm logs we're already throwing that out the window. Charm logs are research, not usual wiki material. Edits are supposed to be VERIFIABLE, but charm logs edits aren't. If we're going to be doing original research, we should at least be scientific about it and know where our research is coming from. We have the ability to block IP addresses from editing certain pages. I think this is one situation where we should. --Wowbagger421 23:03, March 30, 2010 (UTC)
Additionally, if we're going to apply Wiki philosophy to charm logs, I'm going to suggest they're removed completely. --Wowbagger421 23:07, March 30, 2010 (UTC)
If charm logs are completely removed, then we'll be left with lists like Gold charm (common), Green charm (uncommon), Crimson charm (rare), Blue charm (uncommon) or something of the sort. Then, we'll have users battling over the differences between semi-rare and uncommon. I'd say that on a practical scale, the status quo is better. --LiquidTalk 23:11, March 30, 2010 (UTC)
Actually, "No original research" is a cornerstone of WikiPEDIA philosophy, and do I have to point to pretty much every article on this wiki which is based on pure original research (all of them) and RS:NOT#Wikipedia? Hello71 00:36, March 31, 2010 (UTC)

Neutral on Blocking, Support on capping - I support 10-30k capping for the logs.  Ranged-icon.png Zap0i TalkRune scimitar.png  23:14, March 30, 2010 (UTC)

{{RFC}} --LiquidTalk 01:51, April 23, 2010 (UTC)

Oppose - Okay, I'll clarify my position on this. I don't like it. I think that limiting data is pointless, and blocking IPs who can make good edits is against wiki philosophy. Once I figure out those statistics formulas, and get off of my butt and finish coding them into TLULbot, validation will likely be implemented instead of blocking IPs. In my view, this is a far better way to go, and more compliant with what I think a wiki should be. If my opinion is unclear, I'll try to rephrase it tomorrow when I'm not so tired, because I honestly can't tell if this makes sense now. I'm a regular user and I approve this message.  TLUL Talk - Contribs 06:43, April 23, 2010 (UTC) 

Oppose - Would this edit not be allowed with this proposal? What would have been your reaction if it was an IP instead of a user? For a long time, that log showed a fixed rate instead of the correct kill count for many weeks. Some logs like the Ankou charm drop rate recieved very few edits and most can be attributed to disinformation. I am against preventing IP's from editing these logs (many good faith edits are centered on them) and a cap is unwise to me. Say if the log had 9,987 kills. You need at least 50 to update it when you set the cap at 10,000. Are you going to undo the edit? No. Ryan PM 14:55, April 23, 2010 (UTC)

Propose pragmatic approach: I think it's unrealistic to expect the charm logs to be removed, and also think that it's completely unfair to expect the community to have to support them. I think we should then put them in a sort of "limbo" where we basically say "use at your own risk." The information is unverifiable, and should never be considered correct. I suggest we update Template:Charm data to include "Charm logs contain unverifiable data contributed by a variety of users. Use at your own risk.". Endasil (Talk) @  18:10, April 23, 2010 (UTC)

P.S., I think a lot of incorrect data could be avoided by making a separate interface for updating these pages (through Javascript or something). Then, if an update to the data introduces proportions that significantly differ from the established trend, a dubious flag could also be added. Just a thought for an interesting project I might try my hand at. Endasil (Talk) @  18:14, April 23, 2010 (UTC)
See here for a similar proposed solution. I'm a regular user and I approve this message.  TLUL Talk - Contribs 02:46, April 24, 2010 (UTC) 

A better solution... Charm edits should be in some sort of (+blue +crimson +green +gold +none) form. With that, there should be a way for admins to view all charm edits like this (+5 blue +20 crimson +30 green +90 gold +200 none) and selectively remove those charm edits, while at the same time recalculating new charm totals (or, if this is too much processing, have a bot do it once in a while)--Agamemnus 21:59, April 23, 2010 (UTC)

Beat me to it! :) I've always thought this was the better way to go. Even with no malintent, the current system is an invitation to math errors. Qeltar 23:22, April 23, 2010 (UTC)
Also see here, where the proposed solution would mean very little addition is required of users, provided that the bot runs often enough. I'm a regular user and I approve this message.  TLUL Talk - Contribs 02:46, April 24, 2010 (UTC) 
Well, that would only prevent some types of vandalism, but "we'll take what we can get"??...--Agamemnus 07:13, April 24, 2010 (UTC)
Even better so: why don't list all additions that differ from the current total by a certain percentage? The listing should contain % for each charm type, and it would then be easy to delete the entry. Of course there could be a minimum registered total before this becomes active. Delapaco

Comment- a thing about blocking ip adresses. if alot of people edit the charm logs and dont have account we should encourage these people to make accounts and gain more credit for their actions. if somebody has spent the time to kill 1000 lesser demons and has counted the charms then we should be giving them the credit they deserve. few vandals take the time to create and account and by telling peolpe that they need an account to edit charm logs the people making constructive edits will make account and be given credit for their actions [[Special:Contributions/85.211.155.141|

Saradomin sword.png Shadow-fox

12:31, April 25, 2010 (UTC)]] forgot to log in

Comment - Most comment vandalism can really be caught using common sense; for instance, if someone says that they got 2,000 charms total, but only killed 400, then it is most likely vandalism. For the other stuff, there is the TLULbot. Or rather will be, silly lazy TLUL :P Ajraddatz Talk 15:46, April 27, 2010 (UTC)

Strongly oppose both blocking IPs and capping the kill count - Per all:

  • Blocking IPs: As Liquidhelium said. We'd be breaking multiple policies, being unfair to good faith registered users, and this proposal has been rejected five times already. Might I add, how hard is it to revert an edit and place a warning on the editor's talk page? I know there is 99% of the time someone watching the RecentChanges, and even if there is not, I've noticed that some of the anti-vandals here actually look back through the recent changes hours or days to see if any vandalism was missed. Charm logs are as much of a nightmare to maintain as any other mainspace page; e.g. what's the difference between adding a few profanities to Bucket and adding a few numbers to [[Greater demon/Charm log]]?
  • Capping the killcount: What if I were to kill 9,999 dust devils, then as I killed my last one for the charm log, it was locked down? For one, I would be extremely annoyed, two, those 10,000 kills would probably alter the percentages greatly to be more accurate. You can never, ever, ever be too accurate. Quest point cape.pngLil Diriz 77 Talk Summoning-icon.png 20:59, April 27, 2010 (UTC)

Support - The greater demon log also has this problem. I wanted to propose this for a while sig1.png Spam me w/ lolcatsPottery statuette detail.png Is <insert name here> awesome? I don't know, let me check... 21:03, April 27, 2010 (UTC)

Comment+Question to supporters - Do you guys realize that this proposal goes against the very concept of a wiki? A wiki is supposed to be a place where anyone can contribute and edit. Why are we suddenly proposing a draconian crackdown on I.P. editing? If you look through my contribs to the user talk namespace, I'm pretty sure you'll find more instances of the Test 2 and Test 3 templates (for mainspace vandalism) being used, than the Charm log 2 template being used. The fact that this proposal even has the amount of supporters that it does saddens me. --LiquidTalk 21:11, April 27, 2010 (UTC)

Exactly. Quest point cape.pngLil Diriz 77 Talk Summoning-icon.png 21:13, April 27, 2010 (UTC)
Your example is inaccurate. There are FAR more mainspace articles than charm logs on this wiki, so FAR more chance of an IP incorrectly editing/vandalising a mainspace page, rather than a charm log. Chicken7 >talk 13:02, May 5, 2010 (UTC)

But this is just on charm logs, where it seems IPs are causing a lot of problems. It's not like the entire wiki is being closed off.

MagicSa1axLogFishing

21:22, April 27, 2010 (UTC)

IPs are likely the greatest contributors to charm logs as well.  Tien  21:29, April 27, 2010 (UTC)
IPs are the greatest contributers to charm logs. Without them, most of the logs would have very small sample sizes. --LiquidTalk 21:30, April 27, 2010 (UTC)
True, I suppose. 00:42, April 28, 2010 (UTC)

Oppose - More data can only make the logs better. Ancient talisman.png Oil4 Talk 20:00, April 29, 2010 (UTC)

Oppose and Comment - Per all whose stated that this defeats Wiki Policy. Also a cap wouldn't work "The more the merrier!". This problem will be EXTREMELY difficult to solve, needing a balance between controlling vandalism and IP freedom, personally I don't think this will be resolved with everyone happy. 222 talk 09:17, May 5, 2010 (UTC)

Comment - I think the original intent was to have correct charm log data, not vandalized logs that don't meant anything...which means sifting out the bad edits constantly. I mean, that's basically what we're doing for article edits: undoing vandalism...but there are many tools to do that. Sifting out charm logs is like editing regular articles, but a lot more difficult if it is vandalized and much harder to control and see what was vandalized. Lil cloud 9 15:39, May 5, 2010 (UTC)

Request for closure - No consensus, and this has been going on for quite a while. Ajraddatz Talk 19:35, May 27, 2010 (UTC)

Comment - I really think part of the description of this forum was misunderstood. I think the "cap" of 10000 was to be implemented for a single edit (as it very unlikely that a person could sit down and log 10000 kills at once...even killing 10000 chickens would take forever). There would not be a cap to the total number of kills. Farming cape (t).png Lil cloud 9 Talk 05:04, June 10, 2010 (UTC)

If that's the case, that is ironically exactly the limit for a single edit in the ever-about-to-be-tested TLULbot (I know I said I'd be done that by now, homework FTL). I'm sure it would need tweaking, but that seems like a good number to me if it's interpreted that way. I'm a regular user and I approve this message.  TLUL Talk - Contribs 06:47, June 10, 2010 (UTC) 

Closed - None of the above proposals will be implemented. C.ChiamTalk 09:55, June 16, 2010 (UTC)