#wiki
1 messages · Page 9 of 1
So poor grammar on the Item acquisition for Bottled Faith.
DON'T left the map before he finish the speech.
should be dont leave the map
Understand that some people that edit the wiki on their free time are not english natives speakers so they can have a "poor grammar"
And if you see any mistake, I encourage you to edit it
I have no idea on programming level. There is a user (L1mer) try to update the specification so that to make the pypoe some sort of working in some area (plus the repoe writer who make a separate fork), but probably I need to learn the code work to try to fix myself for the skill gem export.
And GGG literally wrote it in TOS reverse engineering the game content and data is forbidden, just they don't want to change the enforcement level yet. (The current enforcement level is no banning for reverse engineering but only banning those used it in cheating)
You can edit and refine the text your self. Just a reminder text as too many people click respawn in the mid of the speech and lost the loot forever
anyone have access to the database?
I've spotted some errors and was under the impression the bots updated the data, so it shouldn't be inaccurate...?
or is data still being updated?
@gritty shadow
data hasn't been updated in a few leagues 😦
hrmm
we did the minimap 12 hours ago
reminds me I have some work to do on the wiki.. but some other stuff seems out maybe I'll do it myself
I don't know what you're talking about I think you pinged the wrong person
Oh do you not know omega2k lol
no idea who that is
he runs the path wiki ...
excuse me this is a pathofexile discord please keep it on topic
thats what you get for pinging me because some dude has a similar name
He is not online in wiki since December, and you got his msg when viewing his github
https://github.com/OmegaK2/PyPoE
(and i guess L1mer stop to maintain his own fork too, he does not response the pull request https://github.com/L1mer)
(ok, i will try to run the fork from https://github.com/brather1ng/PyPoE or pull some file overthere to see is it work to try to run bot in some namespace (sorry no modifier namespace))
@livid moth It seem the pic from the Template QuoteLore is shifted due to the new skin css. How to fix it?
e.g. error pages
Reading some of the comments here on updating the wiki, and I cannot believe that GGG doesn't at least give a spec. Without PoeDB nd similar tools, there basically is no POE.
It is incomprehensible that GGG would leave the people making the tools that make their gama playable completely in the dark.
dude.....read it
https://www.pathofexile.com/developer/docs
Executable apps that interact with the game or game files.
This behaviour is strictly against our Terms of Use (section 7b).
Creation or use of these type of applications will result in immediate account termination.
the site i posted have all the official resource that the dev want the tool developer to know, but not the spec inside the game file
people should just stop making or updating tools for the next rlease or two.
and make people know why you stopped
seems to highly depend on the definition of "interact"
No, it is against TOS but they don't enforce it yet
lol yet. they don't hav e a choice
i doubt reading a file placed on your machine can violate TOS.
"We don't want you to" isnt really an arguemnt
Can you read?
not sure what you mean. This is in the context of executable applications and they shouldn't contain your OAuth client credentials. They can locally store tokens if they want since those are short lived
zao seems to have been under the misconception that we endorse extracting game data which we do not and never have. We just haven't acted on the accounts engaged in it
on the other hand us publishing these documents doesn't mean we're looking to change our existing level of enforcement
@pale bear why we should stop updating tools? the fact that they exists already means there is no issues with getting data from client (as long as ggg is neutral to it)
i just saw whoever it was say ggg doesn't endorse reading the pack files- not that they can really do much about it
im sure their company will say they can, but doesnt make it so
it would be a corporate death sentence to enforce it.
@jade mica ggg's horrible stance against the 3rd party devs that make this game what itis
peple bust their asses for no pay to help the game along and ggg basically tells them to fuck off
I am not suggesting OmegaK2 stop to develop is due to new ToS, but i am saying dev (Novynn) has said before those stuff, at your own risk on tool development so that GGG can change the enforcement level to ban people as they are literally saying it is against ToS but they does not enforce it yet (for cheating of course they will ban, for poedb or pypoe fork such as repoe, it seems not yet)
if just the wiki, pdb, and pob went dead next week, ggg would feel it.,
not saying ppl should (and ive oly make 1 edit to the wiki for bad info), but i would deinitely be closer to it
wiki has been semi-dead due to unable to fetch modifier (actually as of 3.13 still able to extract mod by pypoe, my bad not step in to request bot account right to carry actual edit)
ive thought about helping, but every time i get closer i back away
i work in computational finance and trading and have a lot of spare time right now even
i waned to do somethign with the passive tree, but i cant get the info
But now with life cost change, it definitely need someone to fix pypoe for wiki.
For poedb, i think website ad and Patreon revenue are not much, but at least there is a monetary reason to do so (unless you are saying the server cost is greater than revenue)
is pypoe real? last i saw was a mention of it on the ggunpak site
It is real
my python is better than my nonexistant other lang that it was in
the semi-working fork 1
https://github.com/L1mer/PyPoE
the semi-working fork 2
OmegaK2 stopped to answered pull request, and just like POB fork, it is a community based project that other people work on its own fork and then merge code
impt question: does it work on linux? im useless as a dev on windows
While the actual function of pypoe is read the game file, extract the values and parse as either wiki code, or in lua for wiki module. But GGG change the format of its own table time to time. which life cost is a major rework
i have to get some sleep, pm'ing yo my email
I think py can run in any console? (I use Windows PowerShell to run it)
i have 4 machines in front of me right now, 1 is windows and i need oher peope to help me when it goes down
and i guess notepad or Visual Studio Code or whatever can do for the editing the py code
How to update the gem level progression without hurting my hand changing the value manually ? Is there a method with a spreadsheet ?
I think last league (3.14) pypoe fork still works for support gem. For active skill gem, you really need to manually update it or learn py and try to figure out how to incorporate the 3.14 change on cost to pypoe
note that Vinifera7 changed the parameter so that pypoe need to incorporate the change in it too
Yeah. The new cost parameters are documented here: https://pathofexile.fandom.com/wiki/Template:Skill#Cost_parameters
However, I made sure to include backwards compatibility. So in case you do use the old way of specifying mana cost, it won't break anything.
Because I did this for Bladestorm and oh my god, that's was loooooong to update it manually
Or i guess i need to figure it out myself and read other people fork on the specification on "teaching" pypoe how to read the game file properly
Kind of three parts to it I reckon. Ensure that enough of PyPoE understands bundles, ensure that the exporter code does it right, and ensure that the spec matches up enough.
I never managed to get much farther than trying to specify which tables to export, as the syntax to pass multiple of them into pypoe_exporter is inscrutable.
Anyways, the wiki need a rework, like when you want to edit a skill for exemple and you know nothing about the wiki, you just want to edit the cost or the gem acquisition... It's not normal that you have to scrolls about 10K lines to got what you need
Like the wiki is very counter-intuitive if you don't have the patience or the fundamentals, so I maybe (not sure) try to rework some of the pages to put them more beginner friendly to edit or to update it
That's kind of unavoidable though. All of that data has to be defined somewhere.
Yes but like, it's not organized
The gem level progression is not in the gem level progression section, is in the gem presentation section
So if you want to just edit the gem level progression, you will need to pass all of the gem description, cost etc... To have what you want
Yeah, I know. Currently, the skill data for skill gems (and support gems) is stored by the Item template.
But I now that it will not be an easy task to rework this etc ... Because of time, private life, working etc... etc...
We could split it up so that the skill data is stored on a separate page in the Skill: namespace, but that doesn't solve the problem of it being convoluted.
Yea, I will think about how can we change that
I don't think there's any good way to use the Skill progression template to store the skill data though. I don't see how that can work.
Because like, I want to add the new gem, update the gem, add the new items etc.. Like what I did in 3.12, but currently it's very painful to edit it :/
I don't have the idea currently, will think about it
the wiki exporter understands bundles just fine, it's just a matter of adjusting the new files/specification and the relevant stuff in the exporter code. A lot of things are just mapped 1:1 from dat to lua export (maybe with minor adjustments for wiki readability)
I'd be wary about splitting things that belong together too much. Cargo is prone to issues, and the more you split things up the more likely you end up into missing/duplicate data for some stupid reason. Gamepedia applied a bunch of fixes, maybe it's better now. I was hoping there would also be proper multi-row stores at some point but that never happened afaik.
Also when you do split things up that means multiplying the pages a user would have to go through and edit to update a skill. Cargo strikes here again, that until all pages are updated and only when the main page containing the item is edited, it will actually update the skill.
So, when you say do seperate pages for each skill level, unless you force a cache update the skill would contain whatever old values it had before, which doesn't really make it any easier for editiors. It's the reason a lot of things are sort-of contained in a single page where it made sense.
The skill template is already simplified from the data variant in that it defines "static" stats. The exporter actually looks at the gem levels and finds stats that never change and puts them into the static value in an efford to reduce clutter on the wiki. One thing I had planned, but never got around to was removing the levelXX = True line since that can be implied from having other values, though that only makes it marginally less cluttered.
Omega_K2 is alive! 🙂
The line height applied in the template is messing with that. I'd suggest adjusting that.
so we have one month (edit: until August 23rd, 2021) to debug the bugs due to css change....
why ?
read the link above
Ah that's a bad news
I supposed fandom is the best free option, no other wikis that you might be able to migrate to?
WTF fandom has yet another color change on its new fandomdesktop skin.....
UX friendly colors, definitely. 😵💫 It feels like fandom colors competing with ads for attention 🥲
Someone already fork the POE wiki except that they still figuring out how cargo table work
actually it just matter time that someone made a fork that work, just like Destiny 2 wiki of fandom is way inferior than destinygamewiki.com or destinypedia.com
It should somewhat profitable to at least cover the server cost. Just it is not sure GGG' pov and the umbrella clause in the TOS. Bungie is quite friendly with tool developer (just they are not that happy to be datamined all the not release quest of the season) but GGG only friendly on the trade API but not the rest of the tools. It just take a blind eye on trade overlay (pricing one) and often stressing it subject to change on enforcement level. And wiki to a certain extend need to scrap all the graphic of skill gem and status icon to the wiki, which such DMCA case (if any), currently should handle by fandom
Yea but with all of the PoE content that we got and will get, it will be a hard and long work to have a good forked wiki, and if GGG it's okay with that
actually most map article are not fixed since 3.1 (or 3.5?)....content are left in the child article, modifier are outdated every league, skill gem and passive outdated every few leagues, actually wiki has not much content to fork. The real thing works are the traffic (which the rmt ad exists and i am not success to report it) and the active wiki editors and the templates, all are the backbone of the wiki
and the wiki fork literally copy all the img as far as i know (or at least img with cat that can be discovered by bot are copied)
I'm agree with you, I think to have a good wiki, we need to have the entire power on it, like how does it work (coding), database etc.. and the ad (to counter rmt ad, like you said), traffic etc... Since the wiki belongs to fandom, we have just the "content control". So I hope that the one who is currently copying it ? Will succeed
I don’t see why GGG would have any problems with it. There’s no commercial relationship between GGG/Curse/Fandom
It certainly will if every active wiki editors jump the ship and let fandom version be ruin with GGG nerf hammer every league. Some monster and monster article are actually created during 1.0 and never touched by other editors again (except bot edit)
I'm waiting the ship :>
But I think that we also need to currently update the wiki
tbh, I suspect part of the solution there is going to be merge/redirecting a bunch of articles into parent articles with lists
and BTW here is the all time rank
Top Wiki Editors - Official Path of Exile Wiki
Hey all, the Path of Exile Wiki will have the expanded width for anons and users by default now. It will take up to 24 hours for the cache to take effect.
Scratch the cache part, the cache was dumped so it should be live for everyone.
If it looks like a forked wiki can be sustainable, then I'll happily get on board.
I never have a look on its current state BTW, only via discord
Things start to moves : https://www.reddit.com/r/pathofexile/comments/ox4bvj/would_you_contribute_to_moving_the_wiki_out_of/?utm_medium=android_app&utm_source=share
I've seen a lot of "Why GGG doesn't help?", "Why GGG doesn't gives the data ?etc..." And I think the answer it's simple and can be found couple leagues ago when Chris said "It's been a long time where players was searching for a new recipe, mechanic etc.. And that's why we've added... ...."
GGG want us to struggle, to find recipe, to find data, to find everything because they think that's exciting us like the old times, when wikis was edited manually by fans with datas written on a book... And I think they have this nostalgic vision
GGG continues to delete .dat files, and the api has not been improved for many years. It is very likely that more .dat files will disappear in the next league.
Does that mean certain game data no longer exists on the client side? What's an example, if you don't mind me asking?
recently removed or all removed?
[
"AfflictionEndgameAreas.dat",
"AfflictionRandomSpawns.dat",
"AlternateQualityCurrencyDecayFactors.dat",
"AtlasQuadrant.dat",
"AtlasSector.dat",
"BestiaryRecipeItemCreation.dat",
"CharacterStartItems.dat",
"ChestItemTemplates.dat",
"ComponentArmour.dat",
"CustomLeagueMods.dat",
"DelveFeatureRewards.dat",
"DelveRewards.dat",
"EventSeason.dat",
"EventSeasonRewards.dat",
"HarbingerMaps.dat",
"HarvestDurability.dat",
"HarvestInfrastructureCategories.dat",
"HarvestObjects.dat",
"HarvestPlantBoosterFamilies.dat",
"HarvestPlantBoosters.dat",
"HarvestSeedTypes.dat",
"HarvestSpecialCraftCosts.dat",
"HarvestSpecialCraftOptions.dat",
"HarvestStorageLayout.dat",
"IncursionUniqueUpgrades.dat",
"InventoryId.dat",
"ItemSets.dat",
"ItemSynthesisCorruptedMods.dat",
"ItemSynthesisMods.dat",
"ItemTemplates.dat",
"ItemTradeData.dat",
"LabyrinthRewards.dat",
"LeagueFlags.dat",
"LeagueTrophy.dat",
"LegionRewards.dat",
"MapStatsFromMapStats.dat",
"MicrotransactionHudVariations.dat",
"ModSets.dat",
"MonsterAdditionalMonsterDrops.dat",
"MonsterChanceToDropItemTemplate.dat",
"MonsterStatsFromMapStats.dat",
"NPCAdditionalVendorItems.dat",
"NPCShopAdditionalItems.dat",
"PassiveSkillBuffs.dat",
"PathOfEndurance.dat",
"ProjectileVariations.dat",
"ProphecySets.dat",
"QuestRewards.dat",
"QuestVendorRewards.dat",
"RaceAreas.dat",
"RandomUniqueMonsters.dat",
"RitualDaemons.dat",
"ShopCategory.dat",
"ShopCountry.dat",
"ShopCurrency.dat",
"ShopForumBadge.dat",
"ShopPackagePlatform.dat",
"ShopPaymentPackage.dat",
"ShopPaymentPackageItems.dat",
"ShopPaymentPackagePrice.dat",
"ShopPaymentPackageProxy.dat",
"ShopRegion.dat",
"ShopToken.dat",
"StatSets.dat",
"SupporterPackSets.dat",
"UniqueFragments.dat",
"UniqueSets.dat"
]
GGG has never been keen on the community data mining PoE. However, it was also a much smaller game way back when, which made it possible – though through no small effort – to keep things fairly up to date on the wiki without relying on data mining.
I personally would like if game factors, such as gems and their associated values could be automated. While things like certain mechanic or drop locations are manual.
Pretty much how it is currently
gamespot used to have people written full notepad guide
and Chris lives in that era. Without Youtube, download trailer in fileplanet is meta, and then gametrailers.com is founded (and then folded)
Yet those guides written in notepad were for games that never changed once they were released, let alone sweeping meta changes every 3 months...
The big value-add the wiki has over say poedb is the coexistence of hand-written information with machine-updated tables of bulk information.
poedb pretty much by design only holds mined and scraped information, organized in many ways.
Yes but since it's data mined, it's easier to put it on a website than write every lines manually with some explanation
Like, I tried to create the page for Rage Vortex, and there was so many errors and some value I couldn't add because didn't find them
I guess you can ignore the per level parameter for now and just fill the stat text and mana multiplier , those essential info
Looking at page history of some gem article back to 2013, there must be some way to do it without datamining, just may be miss a few info
I've tried to ignore the per level info, but it doesn't work, it must have it and I dont know why otherwise the page doesn't load and we have a cargo error
Can you show me the error?
I will show you this tomorrow, can't now (UTC)
Some of these were removed because they no longer exist in the game. A bunch of the Harvest data files were for the Sacred Grove layout minigame, for example
And at least one of the other files you listed looks like a feature in development which was accidentally included. 🙂
QuestRewards and the one that mod sell prices are the real shitty ones. component armour was moved elsewhere I think?
though also shitty if entirely gone
#tooldev-general message re. ComponentArmour
ModSellPrices is… eh. It's annoying to not have but I suspect that most players don't really care that much. QuestRewards hurts more; I think part of the answer there is going to have to be a close reading of patch notes and manual spot-checking
I have created this page https://pathofexile.fandom.com/wiki/Expedition_Logbook, please help me complete the sub-pages and create templates.
It looks like the internal class name for the log books is actually "ExpeditionLogbook". I will fix this, and update Module:Game
Done.
Does it usually take weeks for new content to get a page on the wiki?
If there is an article that you think should be on the wiki, you're more than welcome to create it. How long it takes really depends on you.
I'm not familiar with creating wiki pages, and thus it would take me vastly more time than it would take someone who's very familiar with it. I'll leave it to other people, for the same reason you (presumably) leave solving mathematical theorems and studying traffic patterns to people who don't have to spend tons of time learning the basics just to get off their feet. Sorry for the nerve I struck, I really was just asking.
It's a meaningless question. No one here is on a schedule.
I'm not sure I understand. "Does it usually take weeks for new content to get a page on the wiki" is obviously referring to past patches where new content was added. I haven't played in a long time, so I'm not familiar with how long it usually takes the new content to get up and running on the wiki.
There’s two types of content on the wiki - there’s the hand-written sections on pages and for subjects that are pretty much all human text, as well as things that only humans can find and index like vendor recipes, gem quest rewards, things like that.
Then there’s the data-driven tables and bulk content like skill gem info, mods, level scaling. All that is sourced from game data files and require a fair bit of work from independent developers to keep up to date with the changes GGG make to the files. This aspect of it has been lapsing for leagues.
As many parts of the wiki are designed around semi-automatic updates of data, hand-updating them can often be tricky, especially for a newcomer to the structure.
Ah, interesting, thanks!
From a content PoV, what the biggest hurdle to importing the current wiki to another host? IIRC getting images was a bit of a pain but everything else was fairly simple assuming the correct Wiki extensions were installed
All of the new divinations cards has been added
I have added the patch note 3.15.1 but without any links because cannot finish it at time
Downloading 15k image files from the wiki may take a while 🤔
wow, iamacyborg is still alive too? 😉
just about 😅
good to see!
you too bud!
All page names https://docs.google.com/spreadsheets/d/1Odllve1gRFfBuoKwNIeEUr9Hj5kjPNw9CoGyk7D1nas/edit?usp=sharing
Sheet1
Page
O' Eternal
The Kiss Goodnight
The Nameless Play
10x Northern Lights
10x Passive Skill Refund
10x Skin Transfer
15x Dark Fireworks
1 Hour Lethal Solo
1 Hour Multiprojectile Solo
20x Classic Fireworks
3.11
3.11.1
3.11.2b
3.2.0
50x Passive Skill Refund
50x Skin Transfer
5x Skin Transfer...
Looks like Fandom's Export util break when trying to generate files > 50MB
Yeah so if you want to export revisions you can get ~3k pages at a time via the exporter
well, let's see if this works
I've switched over to the new Fandom theme, just so see what it's like to actually edit the wiki with it enabled. It's so goddamn slow to edit anything! I clicked the "edit section" pencil on one page and had to wait about 10 seconds for the visual editor to load, then I had to wait some more switching to the source editor. It used to be fairly slow to edit stuff on Fandom, but this is ridiculous! This is an absolute deal breaker. No joke.
Oh, and I forgot to mention. Section editing doesn't even work. Clicking the "edit section" pencil doesn't do anything different than clicking the main "edit" pencil at the top of the page.
The entire editing experience has been gutted. Where's the list of transclusions on page? Where's the debug log and the performance data?
There's a bunch of utility links that have gone missing too, like WhatLinksHere. Not sure if that's a recent loss or an older one
WhatLinksHere in the right side panel when logon
What links here is down here for some reason.
It's also here, but I normally have the side bar collapsed.
Okay, so it's somewhere that you can't get to without logging in. Not much better
Probably.
Yeah, I just checked. It's not shown anywhere if you're not logged in.
Oho! There is actually a way to switch back to the full editing experience. Go into your preferences and switch to the Editing tab. Select "Source editor" as your preferred editor.
There seems to be no way to get the full editing experience without doing this, since the Fandom theme defaults to the visual editor and clicking "Edit source" goes to the visual editor in source mode instead, which is useless.
I have set to source code edit as default btw (not visual edit source code mode). Old habit of source code by default carry over from wikipedia (and carry over my not preview edit bad behaviour)
I did too, but Fandom apparently decided to change my preferences for me.
testing import scripts now I've got the whole thing dl'd
you probably need template and module
and modifier are stored in modifier, area in area, monster data in monster namespace and so on
despite if pypoe is fixed modifier and area and monster and some module that only store game data in lua can be skipped and just export as new
Templates and Modules are all in the import too
Basically grabbed everything from every namespace
~3GB XML file
Slow as hell though
I think that page is trivial and some third party site is doing the same thing
https://pathofexile.fandom.com/wiki/Daily_Deals/2021?action=history is primarily updated by Arrowneus and it has stopped since June 2021
There used to be a third party website that kept a historical record of the daily deals at https://mtxhistory.com/. It no longer exists.
Either way, I don't really care about keeping a record of it on the wiki, but there was no harm in letting Arrowneus do his thing.
Oh yeah, it's fine it just took ages for that one page to run through as it's huge and there were so many updates on it.
The import I'm working on is currently at ~6k pages
What extensions are critical to the wiki, other than Cargo?
What comes pre-installed? Anything?
Ones that I can think of that might need to be installed are EmbedVideo, Math, Scribunto, SyntaxHighlight, and Widgets. There might be others.
I ticked all the boxes and got these
easy enough to remove stuff or add new extensions though
I think Tabber might be used on a few pages, but I'm personally not a fan of it.
Cool I'll take a note of those and get them added. No point doing much until I can confirm the import ran and I've run the image import
Looks like it'll take at least another day at this rate though
I'm generating another XML dump from Frandom which is just all the newest revisions of pages to try speed up the process, I can add the historical revisions after that
if you update pypoe you can probably ingore several tens of thousands of pages in the export, since you can reexport them with pypoe
specifically mod/skill/monster namespaces
I suppose the version history is lost, but I don't think it matters much in those cases
I have no idea how to run pypoe though 😬
well you'd need to update it first since it has been on the poe version of december or so
hah yeah I think I'll stick with my current plan for now just to get the basic site up and running
A little more real estate than Fandom
my point is it is low priority to export the page history and it is no harm for people to keep on their own project (except never answer the talk page and sometimes i have to double check the meta data id is fabricated or genuine )
Tabber is used in a limited scope . Mainly by wonderful on Maven's Invitations
@crimson gulch Is the site live or not? So that i may need bot account right to test pypoe fork on thing that actually exported before in 3.14 (area, monster, lua)
live but not open right now
I can open it up though
Might be worth waiting until the import is done before you try run pypoe on it though
And meanwhile reddit has someone asking for funding for yet another fork and claiming they have talked to POB and poedb on exporting data.........just sound too fishy to not done anything and start to ask by funding...
I've spoken to the guy, he seems like he has decent intentions but doesn't have any experience with running mediawiki
or my be i can try on a limited scope in the fandom to see it need specification update or not (modifier and gem definitive need update)
and the previous fork https://pathofexile.miraheze.org/ now it is rotting....
The problem is folks are putting up wikis without doing anything about the content
also hosting on another wiki farm seems counterintuitive
He sent me a message asking for help with the data mining part
@analog abyss I can create an account for you so you can dig about
DM me what you want here
I'll open up all this stuff as soon as the content's in
~208k revisions added 😅
The site is here until I bother moving to the proper URL, no rush on that front though https://phpstack-640597-2086210.cloudwaysapps.com/w/index.php?title=Main_Page
I'm curious, how would fandom react if current wiki admins "sabotaged" the wiki to remove all its information and instead point to a non-fandom site?
Obviously the site is community managed so it's not like a fandom admin has any standing to object (and also something like the poe wiki is probably going to be too small to ever concern themselves over).
But I have to imagine they'd object to literally salting the earth to help with SEO.
Yeah, that's not an option
they'll revert edits and ban users
also just a bit too petty
That's really unprofessional, so I wouldn't do that – not that it hasn't crossed my mind.
Are we professional ?
But yeah, no. It wouldn't be impossible to revert, and I don't want to entirely burn all of my bridges with Fandom. A lot of their staff came from Gamepedia, after all.
I'm against sabotage and if we can make it without do it, it will be fine
Yep, although more and more of those folks are leaving
Looking forward to having a fourth site in the mix of poedb, fandom wiki, and that bullshit RMT site in my search results in arbitrary order 😄
I have a question, because I continue to update the wiki, should I stop ? Or continue and you will merge the data after ?
I don't think professionalism is part of the picture when fandom are forcing extensive negative changes to the experience.
But I agree it is objectionable.
Just that if they're not going to listen to their own communities then abandoning Fandom for better options might force them to reconsider some changes.
Ranking a new domain is going to be a shitshow tbh
Wait, what bullshit MTX site?
Sorry, RMT ^_^
Google strongly take registration age of domain into consideration when it comes to ranking
It's not that hard to put the website at the first position
It is very hard
But it's costly
The current site has a decade of ranking and history behind it
You can continue to update. We'll worry about merging later.
I ran the export yesterday but getting a few days worth of revisions is no big deal
You can very fast be the first page, like some have tested it if fake website etc... But it cost around 400 to 2000$
I'm not convinced that's true
Yes and no, like to add all the prophecies cards, it took me 2h30
Check on YouTube or Reddit posts talking about
But from a script perspective it won't take that long
And I'm going to re-enter in my full-stacks dev, so I will ask if you want
I've worked with a bunch of SEO professionals over the years and if it only cost a couple k to rank a domain they'd be all over it
seeing a page from 0.10.0 is a blast from the past http://phpstack-640597-2086210.cloudwaysapps.com/w/index.php?title=Unique_item
A unique item is an item with a specific name, unique artwork, and a predetermined list of modifiers. Unique items have the same implicit modifier as a normal item of the same type.
Unique items are meant to be less powerful overall than high-level rare items, but offer unique gameplay and allow special character builds to be based around them. ...
I'm agree with that but I don't think that Fandom is investing in SEO specialy for PoE
And if we need to transfert :
- Reddit post announcement
- Fandom announcement
- Discord etc...
Players will moves by their own
yeah, it'll be a long process either way, and that's assuming folks like vini want to go edit a new wiki in the first place
Yeah, I'm hyped.
where's my party parrot gif when I need it
I don't have Nitro, RIP
Import is currently running through V's on the standard namespace, all the other namespace up next 😬
jesus christ these maintenance scripts are slow
I'm sure most of you have seen this already: https://www.reddit.com/r/pathofexile/comments/ozqy5w/project_poe_wiki_an_effort_to_move_the_wiki_out/
865 votes and 114 comments so far on Reddit
Is JourneyToJah on here?
I was DMing him yesterday but doesn't look like he's in this server
I'm just wondering why they didn't try getting in touch with anyone on the current wiki before posting this.
I think he reached out to @stone crater but yeah, that'd have been a logical step. I guess community folks want to move the wiki but didn't think to speak to editors about it, they just got stuck in
It's not really a problem. I just think we should coordinate our efforts, rather than setting up competing new wikis.
It's one of those lovely problems where a lot of souls can try their hand at the first step, yoinking data, while the long-term data ingestion and curation is something else.
I would like the bikeshed to be blue, btw 😉
Yeah, either way once this import is done it'll be much simpler to do a DB dump than have to run the import scripts again if someone wants the wiki moved to another host
For sure. Maybe I am biased but if iamacyborb is already migrating data than I would say he's ahead of journey's efforts, who sounds like he's waiting to see if he can rely on poedb for datamining and still needs to work out porting that info into his wiki.
For me, whatever assistance I can render in getting tooling and DAT stuff working I would rather channel it into this existing community, wherever we end up going.
Is there anything folks need setup from a bot perspective on the wiki?
I don't think so. Not at the moment. We didn't have any true bots running on the wiki, just individual users' bot accounts.
page history of legacy mod still useful.
I can't believe the Daily Deals page has ~3.6k revisions
I think as long as the "bot" account edit was marked as "bot" so that some people like to filter it out, it is fine. I guess if using pypoe with wiki write access, rather on how to set up password instead
fandom has 2-step password BTW.
Ah interesting, I'll look at how I could replicate that if it's something people want/need
All content pages done, looks like user talk is being imported now, still be a while before Modules/Templates and all the rest get done though
Is there an edit history for this that I could port to Vector.css? https://pathofexile.fandom.com/wiki/MediaWiki:Hydra.css
Does the import script not also do deleted pages?
I don't think so 😬
Well, I can restore it.
I'm just running the latest revisions right now, I'll do the full dump of past revisions later once I can confirm modules and templates are working
Isn't this the stylesheet you want though? https://pathofexile.fandom.com/wiki/MediaWiki:Hydradark.css
It is, but Hydra is a proprietary Gamepedia skin
so I don't have access to the PHP files for it
looks like it's just a forked version of Vector though
Of course. Why didn't I think of that?
weird, running into a caching issue that's logging me out when I try access Mediawiki namespaced pages
Oh yeah that css is going to take a bit of work http://phpstack-640597-2086210.cloudwaysapps.com/w/index.php?title=List_of_unique_rings
This is a list of unique rings.
Ah, Reddit
Reddit is full of upvote mob that no idea what they are vpvote. There are lots of reader want to change but not many have experience on editing.
It is not the same case as POB that before the fork, people already forking and submit pull request to openarl, which he is so busy and people finally find someone as central depositary
The previous wiki fork at least teaching anyone that want to fork, they need to deal with CSS and lua table and everything that is datamine . It needs to have the wiki admin and editor hands dirty on actually editing instead of just talk about ask poedb or what is the domain or don't even start their own wiki articles forking
I find it odd they've setup a new Discord rather than reaching out to folks here
I inquired about that, they don't quite like the drama of this server.
End up what I feel it just ask for money and the project bust or not is not the real concern
As for what that might imply, heaven knows. I gave them some tech info on how horrible it is to work with the game data.
That's not even talking about getting that data into the wiki then using Cargo to actually do stuff with it
or the Module files written in LUA
What I saw Poe cdn stop to public some game asset as PNG that need zao's exe to export DDS instead. Even poedb has dead link pic due to the cdn.
Without hand dirty to know how to actually work . Just good luck on their own project.
With 3.16, legacy CDN URLs will be fully gone.
I quite feel with GGG on that, as it's both a big bandwidth cost of hotlinking and a risk of revealing unfinished content by URL guessing.
Yeah, that's fair, but they could find other ways to support community efforts to get these files
What I done it quite often on editing the scalar value of the cdn kmg
There was some nice value from the CDN when it came to composing map icons for example, they're all loose glyphs in the data files which you have to tint and composite on the base map icon.
Flasks are a big one too.
I think map article the Dev still public the parameters of the new map series. But ggg write it black and white content creator should host the fork IMG themselves. Poedb currently not and I guess it save a lot of space from that
weird error
I think audio file can be replaced. The steam folder has all the latest file in .ogg
Now that is an old video
worst case scenario I can just disable embedvideo and rerun the import
is it really a huge costs with cdns and caching?
Ah joy, the sight of Lua errors all over a page http://phpstack-640597-2086210.cloudwaysapps.com/w/index.php?title=Aegis_Aurora
Aegis Aurora is a unique .
This required a blank edit to fix on Modifier pages, right?
I assume they jam some good coin into CloudFlare.
No idea about the scale of things of course, don’t remember if novynn talked about it anything.
@rich widget Any idea where the item table is declared?
Hang on. I'm looking
Feel like I'm missing something obvious with having to declare the tables
Yeah so I've created those
Maybe it's being populated?
How many tables do you see on Fandom?
Oooh actually, that helps me find the links I need
Cargo tables - Official Path of Exile Wiki
just running the rebuild script to see if that fixes anything
SMW was definitely less of a faff
This template defines the table "items". This table has not been created yet.
It says the table hasn't been created yet
Where do you see that?
🤦♂️
What's weird is I was able to create some table but the button isn't showing up to create that particular one
And now it's showing

generally I'd say, make sure all tables exist
then make sure mods are in the db, then base items, then unique items
some data is commited via lua modules, might need to null edit those (can find them via the search)
Yeah I've found how to create all the tables now so that's working
mods are all in the db but might require null edits
Some stuff is working http://phpstack-640597-2086210.cloudwaysapps.com/w/index.php?title=Shrapnel_Ballista
Shrapnel BallistaAttack, Projectile, Totem, Physical, BowLevel: (1-20)Cost: (5-11) ManaAttack Speed: 50% of baseAttack Damage: (38-44)% of baseEffectiveness of Added Damage: (38-44)%Summons a ballista totem that fires multiple arrows with extreme force, breaking them apart into shrapnel. Enemies can be hit by multiple arrows from the same attack...
Slowly burning through Lua errors
Seems like stuff is working, need to run rebuild scripts again
I hate Sunday and insomnia
And this is brather1ng's fork. Trying to exporting league currency
PyPoE.poe.file.specification.errors.SpecificationError: <ERRORS.RUNTIME_ROWSIZE_MISMATCH: 3002>: "ActiveSkills.dat": Specification row size 124 vs real size 125
Well i guess i need to dig out other people's fork to check what has changed once again....
And world (area) export too
PyPoE.poe.file.specification.errors.SpecificationError: <ERRORS.RUNTIME_ROWSIZE_MISMATCH: 3002>: "WorldAreas.dat": Specification row size 480 vs real size 504
While the other team which has reddit admin blessing, well not sure they need get their hand dirty on how datamining work or not......
And monster as well
PyPoE.poe.file.specification.errors.SpecificationError: <ERRORS.RUNTIME_ROWSIZE_MISMATCH: 3002>: "MonsterVarieties.dat": Specification row size 445 vs real size 437
I’m not quite sure they’ve got Reddit admin blessing quite yet
they claim to have
I am waiting L1mer online and reply me, and if the pypoe is partially fix. It is good to go at least exporting this league basetype and currency as well as some of the monster data. Modifier part i guess it take longer as i have no knowledge on programming itself (well, matlab and R is for engineer and mathematician)
I would assume that it's derived from that one response they got to the first post.
I think blvcksvn is not entirely means all the admin or talking about switching the link in reddit into their. blvcksvn has edit in POE wiki (fandom) i guess he is talking about he will not edit fandom and will edit the new forked wiki.
for WorldAreas.dat, add the following fields to the end of the definition:
('HeistChance', Field(
name='HeistChance',
type='int',
)),
('RitualChance', Field(
name='RitualChance',
type='int',
)),
('Unknown110', Field(
name='Unknown110',
type='ref|list|int',
)),
('UltimatumChance', Field(
name='UltimatumChance',
type='int',
)),
('ExpeditionChance', Field(
name='ExpeditionChance',
type='int',
)),
It just a matter of who get the fork actually work. Sadly the previous attempt was stuck in cargo table
thx
should be immediately after the definition for DeliriumChance
It is way too messy that i have a fork to tune the black list of item export. and then brather1ng's has their own fork. Merging the change by me and then by brather1ng is a nightmare as they were not meant to be the central repository
or may be i can submit pull request i guess
still have row size 500 vs real size 504
may be due to once again i have 3.15.1 ggpk and they change something midleague?
did you already have DeliriumChance above HeistChance? if not, add that too
omegk2 version already have that. omegk2 stop at HeistChance. Unless there is something add before instead
oh, right
go back up to Unknown32 (near the top of WorldAreas) and change its type from int to ulong
I forgot that changed too
I have to view brather1ng and L1mer's fork to try to deduce what has changed . managing the version now really really messy
And i guess instead of "Unknown32", which have 5 in the same file, the naming of unknown may need cleanup to be more specific
the numbering of unknown fields is always sequential within that single data file, sometimes there's gaps when the purpose of a field is discovered
well, it works. I guess i need to take a break to deduce the change of the rest (monster , etc ) to make some of the function of pypoe back to functional
I mean may be i would name it as "AreaUnknown32" to be better searched and distinguish with the unknown from other game files
Largest I have is 496, but not sure how late in the league that is.
My data ingest is still partial.
WorldAreas.dat? 3.15.0 was 504 bytes/row, and still is
well that part is only a portion of specification that need to fix and brather1ng has done the part that they are using for repoe. And i still need to fix active skill gem, monster and a lot more for pypoe as the wiki exporting script
and i guess may be better based on brather1ng 's fork instead of have multiple version of fork which all have some changes that i need to catch up from Jan 2021
Start to have even more drama on not asking what pypoe even is or try to use it on 314 ggpk or whatever and decide that start from guessing the file specification I guess (or setup a scrapper to copy value from poedb?)
People always want to complain against things because they want to be a part of something. Like for the wiki transition, I've seen on the Reddit post that people has the same complain as him. "why now ? They could do it since 3.9...".. Whatever just don't look at them, and continue what you are doing
Where was that even posted?
in their discord I think
Just comparing the page load stats to the Fandom wiki...
It'll be slow right now, I turned off the Varnish cache while I work on stuff
Also probably missing a lot of cache optimisation
It's still way faster though lol
the magic of not running a million tracking scripts 😬
Even with an ad blocker, the main page on fandom loaded 6.03MB of data and took over 20 seconds to do so. Plus it keeps trying to load shit in the background.
Yeah, I was wondering about that.
It's probably the stylesheet using images from the fandom domain.
Yeah, that's mental...
Dude who built that tool is rad he's an ad fraud researcher
I did a podcast with him https://www.mql.fm/002-60-million-60-billion-ad-fraud-question-augustine-fou
Oh, you mean the tool for graphing the resources?
yeah
cool
We may need the Variables extension. https://www.mediawiki.org/wiki/Extension:Variables
The Variables extension allows you to define a variable on a page, use it later in that same page or included templates, change its value, possibly to a value given by an expression in terms of the old value, etc.
It is much like a template, only very lightweight and scoped to only a single page, so you can use many variables on a page without s...
I really don't worry myself as editing has teach me how to fix the cargo query code (or somewhat). Or at least have some time on using pypoe . I never worry myself on assuming i know everything. Problem need to be solved by people, but the shortcut always observed other people work and deduce the solution to my own problem. Or just humble to ask and networking
Going through and creating all the remaining Cargo tables now
Is anyone able to help with null edits of pages?
Script error: The function "store_data" does not exist.
🤔
I think all tables have now been created
so need null edits to populate everything, it doesn't look like the MW rebuild script helps on that front
Okay so just missing _pageData and area_spawn_weight_overrides neither of which appear to be defined anywhere which maybe means they were defined in an old revision
strange error 🤔
Wiped the dupe table in the DB but that's obviously not a good fix 😂 https://www.poewiki.net/wiki/Stiletto
StilettoDaggerPhysical Damage: 7-27Critical Strike Chance: 6.10%Attacks per Second: 1.50Weapon Range: 10Requires Level 15, 30 Dex, 30 Int30% increased Global Critical Strike ChanceAcquisitionArea level: 15Vendor PricesVendor Offer1x Scroll FragmentWeapon DPSPhysical: 25.5Phys+Chaos: 25.5Total: 25.5MetadataItem class: DaggersMetadata ID: Metadata...
All Modifiers created and deduped for this, now getting this error
Genuinely surprising
LOL, this is getting way better lighthouse scores than the WordPress site we have at work
Keep going, you are doing a good job, I can't wait to edit the wiki !
It's available to edit now!
I was just setting up caching and stuff but it's all good to go
Nice, does the leaderboard is still here ?
Because with the system of points, badges, leaderboard, it help people to have some result with what they are doing so it will encourage them indirectly
Yeah that makes sense
Hey guys, just to clarify - i'll contribute personally to the new wiki, but we would need to see some progress before we can endorse it in an official capacity on the sub
What sort of progress are you looking for?
I dont think i can give you a benchmark at the moment but we'll be keeping an eye out
Fair enough!
Outta curiosity does the wiki still use the old wikis edit styles and stuff?
What do you mean by edit styles?
Like the editing ui and the way the metadata tags are set up and stuff
I think like time, if the wiki is edited continually it will help for sure, because I care about the "hype" of the transition. A lot of contribution will be needed and in long term
Yeah it should all be the same as the old one
I have the Visual Editor installed right now but not 100% sure that's a good idea tbh
Gotcha. It would be great to have some user manuals since that stuff can get annoying
I think source editor is way better
Yes I was thinking about it
But like even the drop locations for div cards and stuff
Its almost impossible to edit them
Almost because don't have the knowledge for this or like it's really impossible ?
Ah, stuff like that is genuinely more complex because it's handled through transclusion of other pages
There's very few userguides on how to edit those linked data sets which is quite limiting
Gem levels too, not for now because it's not a priority in time, but the template of page will need to be reworked
A lot of this stuff is hidden in Template: documentation but it makes sense to make that stuff more easily accessible
Not that I think that people should have to read tons of docs before they can do anything though, that's counterproductive
We've talked about it with Liu because when you want to edit a page, you have the whole gem level at the start
I'm agree with that
Yeah just something to keep in mind
Most people get turned off from contributing due to having to learn the systems
Maybe with commentaries ? In the code source of the page ?
There are definitely ways to help there
but also from a data PoV you need to balance ease of editing with ease of accurate data through imports
which is a fine-line to walk
Plus.. "putrid clositer" as the div card location is kinda sus
Haha
Thw joys of automation
with the transclusion system and Cargo the benefit is a single page gets edited and that propegates through the entire site
Most people get turned off from contributing due to having to learn the systems
I don't think there's any way around that. If it's just information (like skill/unique interactions) or even formulas, a low end user can handle that easily, but I think only experts would be able to produce or tweak templates.
Or anything that relies on bots/datamined information
Even just a guide on how to edit the template would be great for more dedicated users
And "it's not fun" to edit the wiki, like it took me 2h30 hours to add all the div cards from 3.15
Yeah, I think bots will remain necessary unless GGG can help provide data but that seems incredibly unlikely
GGG will never do that
Because they are nostalgic, I encourage you to look upper, when we've talked about it
Anyway i gotta go for now but yeah let's discuss again later how to improve accessibility
Otherwise i fear the wiki might stagnate again
The fandom ad machine was only part of the issue
Yep, I'm afraid of the hype too
For things like {{il|X}} we just need to wait that you fix cargo data problems ?
Got a link to a page where you're seeing that?
Could be Cargo or a missing extension
Astral Projector is a unique .
Or here, it didn't find any of the attack skill : https://www.poewiki.net/wiki/Attack
Attack is one of four damage sources, which deals damage to targets based on the damage values from weapons or unarmed. It can also be based on the defensive values from shields like with or . Skills that deal attack damage have the Attack gem tag.
Yeah pages need to be null edited so the Cargo DB can be updated
so like just reload the page with a blank edit ?
Until the Cargo DB is populated these errors will occur but I'm not sure how to resolve that short of null editing the entire site
Yep
I need to learn how to setup a basic wiki bot, been a while since I had to do that
No, need to do a blank edit
it's dumb
The maintenance script to purge all pages also doesn't do it
Yeah need to do it to the pages it's trying to find 🙃
aaaaaaah
Seeing if this will help
Oooh I've understood, like if for exemple a page (page1) is using the data for "Spell Cascade Support", you need to blank edit the page of "Spell Cascade Support" and then blank edit page 1
But there is only 1 page
Yeah that's another issue
Try now!
Data is getting saved twice in Cargo tables for some reason
purged the cache, works now
and can we purge the cache ? the button seems to not working
You can but maybe the caching I've put in is a bit agressive
it works for me but takes a few seconds
okay
same, aaah if it's doing it each time, i will always need you xD
Yap it's doing it for all the different rings
It's normal that we don't have the access to import page like : https://www.poewiki.net/wiki/File:Inpirational_(Champion)_passive_skill_icon.png ?
be careful, page like that, will disconnect you, i dont knowwhy
Yes, every time I reload a page by blank editing, Cargo seems to readding the data, :
Disconnect you?
Yea, if try to logging with this page, it will disconnect you
Weird, will take a look when I get back home
okay!
I've gone down the templates and cargo route on a corp wiki before, imo you want to make 2 sites, one with all the datamined / extracted / databased info in the best possible way. And a wiki for human notes as readable as possible, with some pulls in the wiki on the other to display some convenient info. Maybe even ditch wiki completely and just do a static site generator with a "edit this page" ala github backed docs using a WYSIWYG.
but in the end people will complain the wiki is wrong, but not fix it even if they know the answer, it is incredibly hard to promote the shared ownership shared caretaking feeling, not even talking about how such a community invariably starts setting up standards that raise the barrier to entry / mods undoing edits discouraging newcomers (no matter if it was warranted or not) etc
that's my 2c anyways
That Reddit thread is going as well as I was expecting 😬
poe reddit is a shithole
I honestly wouldn't expect anyone from there to seriously help with anything
having admined a mediawiki for a decade and a half I would never use it again if I had to start over. would use something much simpler and straightforward
the power just isn't worth it imo
as far wiki bots are concerned, I'd recommend mwclient library in python. It's not strictly a bot, but very very easy to program a bot to do what you want
to give a concrete example of null-editingthings
import mwclient
site = mwclient.Site('www.poewiki.net')
site.login('YourUserName', 'YourPassWord')
for page in site.categories['Mods']:
page.save(page.text())```
the duplicated entry has happened before in fandom verskon
it can happen afaik if the wiki itself goes through pages and updates them, causing to add duplicate data if they are already in the db
https://pathofexile.fandom.com/wiki/User:OmegaK2/cargo
currently gone in fandom version. but map article often has this problem.
but passive skill need pypoe export anyway, last time i only done to rowid 1500 ish and some ascendency was skipped . i will have a look for the notable and basic passive ,it need spec update or not
Thanks! Will have a dig into that
dupes deleted
Any idea what's causing this expand/collapse functionality on certain templates?
Makes the page look crappy by default
Looks like stuff is starting to propagate through
That's fucking bizarre
Seems to be something related to the Varnish setup
Looks like this is finally running through
dunno, I've sent mods a DM but no response yet
looks like it's back, weird. Probably got reported into oblivion
yeah I'm guessing autohidden
s'all good though
I always forget how much of a pain in the arse setting up email is for new websites
I'm impressed, already got the first spam bot
I'm petty but that delete reason seemed appropriate
x)
Looks like the Cargo table rebuild I’m running is slooooowwwwlllllyyy fixing the lua errors across the site, it’ll take a while though
You can follow progress here https://www.poewiki.net/wiki/Special:CargoTables
Skills are starting to populate now https://www.poewiki.net/wiki/Ancestral_Warchief
Very glad that means null edits across the board won't be required
That's a javascript, my dude. Is it causing issues?
No issues, it’s just ugly 😅
Vini, giving you admin user rights in the new site if that’s okay?
Yes. Thanks
I’m going to have to turn on email validation for new users to try limit the spam bots
Yes please
I saw all of the bots there
Email validation isn't too much of a hurdle. For ordinary users to deal with, I mean.
You will probably need something stronger than email validation, but at least it's a start.
Yeah for sure
I’m out getting tattooed today but will make sure that’s done by the end of the day
I’ll keep an eye on recent changes and block/delete when I see spam though
You'll probably want to set up one of the third-party CAPTCHA plugins (like reCaptcha or hCaptcha) as well. You'll still get some spambots even with that; one reasonable defensive measure may be to lock down external links and require all domains to be approved
There's a pretty short list of sites I can imagine it making sense to link to -- official site, reddit, poedb, github, a couple dozen well-known tools
Yeah I'm looking at hcaptcha now
wikipedia, fandom (lel), ...
Trying to avoid the big G for service 😅
Fandom are getting a single disclaimer link in an about page and that's it
Added
If you previously created an account you might need to confirm your email
Single question but where and how passwords / emails are stocked ?
Emails and pw's are stored in the MW database
Let me find out how pw's are stored
but use a password manager
yes but they are encrypted ?
yeah
Old topic but relevant https://webmasters.stackexchange.com/questions/74014/how-does-mediawiki-save-the-passwords
Okay, I willl look at it
The user table is where MediaWiki stores information about users.
If using Postgres, this table is named mwuser.
Check the user_password section there
Normally passwords are hashed. I expect MW is no different.
I believe they are hashed, not encrypted. Storing encrypted passwords in a database really isn't that secure.
So in other words, the DB doesn't know what anyone's passwords are. The only thing it knows are the hashed passwords. When you log in, the website applies the hashing algorithm to the password you enter in the login form and if it matches the hashed password stored in the DB, then it says, "Okay, you must have entered the correct password."
I'm not 100% sure but I think we'll need cargo:attach templates for all the cargo tables for the Cargo maintenance script to work
Yes. They don't all have attach templates.
I'm not sure what the Cargo maintenance script is though, exactly
It’s a script to rebuild Cargo tables, basically avoids the null edit step if there’s an attach template
Okay, yeah. I thought that was what you were talking about. It requires the table to have an attach template. My understanding of the way it works is that the attach template is transcluded on each page that stores data to the table. Then the rebuild script uses "What links here" of the attach template to find the right pages.
Ah I see
Which table(s) need to be rebuilt?
Basically all of them that don’t have an associated attach
I'm looking at the list of tables now. The areas table, for example, doesn't have an attach template. It also currently has 0 rows.
Could we run a script to null edit each of the area pages? Wouldn't that effectively build the table?
We’d need a bot for that, but yeah
It can be done with AutoWikiBrowser
Will have a dig when I’m home
Currently getting my thigh tattooed 🙃
Just need to be careful with the null edit route as it was generating dupes in Cargo
Well, time to create an account and testing pypoe after repoe creater say they fixed some part of spec
👀👀👀
any offline dds to png tool recommendation?
First "article" done
Necrolord BootsBoots SkinChanges the appearance of any equipped boots to the Necrolord Boots.156x156pxMetadataItem class: MicrotransactionsMetadata ID: Metadata/Items/MicrotransactionItemEffects/MicrotransactionNecrolordBoots
Necrolord Boots is a boots skin microtransaction. It can be purchased from the shop for 70 points.
The item is part of Ne...
Worst case, ImageMagick's convert probably can do it.
Well the wiki keep on asking me to login when i try to upload the file for the Necrolord Boots
I guess 4G wireless intenert problem?
If you want to do it more interactively in a graphic manner, I use xnviewmp to look at thumbnails and do bulk conversions.
Well, i uses xnviev free version before. I will have a look but for now i just export img one by one
@rich widget Just exported the Iron Flask using pypoe fork, i guess it miss the tag in module
I think it’s a cache issue with pages ending in image file extensions
Will look when I’m home
Currently enjoying this 🙃
@rich widget and need to add Expedition to the list of leagues allowed for the drop_leagues parameter
visual representation of the wiki structure
Done
is the implict_text transposed to the unique went wrong?
I have 4 lines of Restores Ward on use, but rest seems fine
https://www.poewiki.net/wiki/Elixir_of_the_Unbroken_Circle
https://www.poewiki.net/wiki/Vorana's_Preparation
I wonder if it duped somehow
No. God no
BTW is @honest topaz interested in moving too ? But it would be another project for exporting version history. If not, i wonder the zh: (which basically nonone but me used )de: fr: ru: interwiki link need to be discuss their fate
I don't have any interest to export zh wiki as there is not much real stuff to be export and i won't have time to write it one by one (and i guess Mainland China has its own POE wiki already)
I didn't forget about the RU wiki, but there's so many issues to discuss. I'm sort of avoiding heaping everything on all at once. Maybe Ruba can continue to hang out on Fandom until we get things figured out.
The split with Fandom does unfortunately mean that the interwikis no longer work though.
I could import them if they’re wanted
@crimson gulch How do I get AutoWikiBrowser to work?
I get this when trying to connect.
Okay, we're getting somewhere. What does this mean? I get this when trying to log in.
That’s weird
Just googling to see what I can find
Can you try ticking the https option if not done already?
Already set
Weird
I wonder if I’ve set something in the settings file to block it
Or maybe a user permission thing?
Are you using something called "autoproxyblock"?
@crimson gulch How do you check user groups?
Fucking hell. I figured it out. It needed www.
🙃
The thing is that it let me connect without it, but showed that error when I tried to log in.

I’m finally heading home, will be able to fix shit in 90 mins or so
an it is late night for me. May be tomorrow i would check the pypoe again on how much function left from the fork
Sounds good! Anything you need me to look at?
Other than fixing the cache issue on File pages
@crimson gulch I believe you need to create the namespaces that are missing.
https://www.poewiki.net/wiki/Area:1~1~town This should be in the Area namespace, but it appears to actually be in the main namespace
Create these namespaces, please. Try to keep the ids the same, if you can.
That's pretty bad actually, you'd need to kill all pages before creating the name spaces
I believe it goes nuts if you do the name space after the fact and gets confused
Shit, I wonder if the import got fucked up because the namespaces didn't already exist
Ok, looks like I can assign the name space to a given ID
Need to test though
Where did you see the id numbers? Trying to check the ones I've created are correct
Looks like there's a script for it https://www.mediawiki.org/wiki/Manual:NamespaceDupes.php
namespaceDupes.php is a maintenance script located in the maintenance directory that can be used to clean up inaccessible titles after adding a namespace (or interwiki prefix).
<option value='10000'>Buff</option><option value='10001'>Buff talk</option>
That worked
Yeah. It's correct
Just creating the rest then running the script
Created
<option value='2300'>Gadget</option><option value='2301'>Gadget talk</option><option value='2302'>Gadget definition</option><option value='2303'>Gadget definition talk</option><option value='10000'>Buff</option><option value='10001'>Buff talk</option><option value='10002'>Passive Skill</option><option value='10003'>Passive Skill talk</option><option value='10004'>Modifier</option><option value='10005'>Modifier talk</option><option value='10006'>Monster</option><option value='10007'>Monster talk</option><option value='10008'>Stat</option><option value='10009'>Stat talk</option><option value='10010'>Skill</option><option value='10011'>Skill talk</option><option value='10012'>Area</option><option value='10013'>Area talk</option><option value='10014'>Guide</option><option value='10015'>Guide talk</option>
Starting to come through
Done
That number looks wrong with Modifiers though 🤔
This looks correct now though https://www.poewiki.net/wiki/Area:1~1~town
Okay no, looks good https://www.poewiki.net/wiki/Special:AllPages?from=&to=&namespace=10004
We now have 6 lines 🙃
yay
Just deleted the dupe in the DB, now only shows Restores Ward on use x 3
How do you delete?
I'm literally having to delete manually in the DB right now
which is not the right way but it's best I can do until I have a fix for CARGO DUPING STUFF
minus caps
The table rebuild script is slow, huh?
Yeah, mega slow
I'm going to increase the size of the Droplet I'm using to one with more CPU this evening and re-run everything again tomorrow
I've taken the whole week off as I'm getting tattooed Wednesday and Thursday too so have a lot of time this week to make shit work
This might be a good workaround while I trying find the real solution https://www.mediawiki.org/wiki/Extension_talk:Cargo#Cargo_duplicate_row_workaround
It's going, but oh so slowly.
yeah 🙃
It's having to delete everything and recreate them all with the correct namespaces
Just creating the _Pagedata table for Cargo now
not sure if that's used anywhere but saw it's on Fandom and it has to be created using a separate script
I don't think it's used anywhere, but I guess it won't hurt.
yeah
It does let you query by page category, so there's that
Yeah, might be useful for a bot
Looks like it gets used in a few places https://www.poewiki.net/w/index.php?title=Special:Search&profile=all&search=_pageData&fulltext=1
Hmm. That said, I have no idea why this query is joining with _pageData. https://www.poewiki.net/wiki/edit/List_of_fossils
List of fossil items.
yeah I'm trying to work out what that's achieving
@grand magnet Any idea why you set it up that way?
It's not in the query though?
I just removed it
well it does some unnecessary shrug
Oh I renember now I think
I think it used to query categories on the page
Probably just forgot to remove when I copied the query
So it was a holdover from an older query. Gotcha
I do feel like it might make sense to "pre-bake" more data, instead of making Cargo figure it out at runtime. In particular, it'd probably be good to move away from queries which depend upon substring matching, e.g. to identify items with modifiers related to a particular topic
I just null edited a page and didn't get a dupe 🎉
I spoke too soon, page was rebuilding so they're all duped in the new table 😮💨
Downloading Cargo 2.2 which is the version on Fandom to see if it's an extension error
Fandom version is heavily edited if I recall correctly.
🙃
I've managed to break the whole site doing that anyway
and we're back
but 2.2 is not playing nice with things
Yes, duplication of entries in Cargo is an issue - when pages are saved
while a recreate process is ongoing, or two recreate processes for the same
table run at the same time, it can lead to duplicate rows.
🤔
https://sourceforge.net/p/semediawiki/mailman/message/36016946/
How tho
Yeah and that’s with Varnish turned off for most of the day
And mostly file transfers to the box itself
@rich widget hey, I was wondering if you had a moment to chat real quick. sent a friend request so we can speak in dm
@crimson gulch if you have a moment that would be fine too
when pages are saved while a recreate process is ongoing
well if the cargo table duplication is due to pages are saved while a recreate process is ongoing, i guess i have messed it up by edited some modifier
Time to have my character climb to Act 10 for my Templar project then https://www.poewiki.net/wiki/Templar#Text_and_Speech
(sure i can copy from https://oriath.net/Audio/Templar/ BTW)
oriath.net should be all up to date fwiw, feel free to copy from there 🙂
and btw, the reason the oni-goroshi drop line shows up is because the game treats that as a character audio event for each character. it's not part of the sword's lines because you aren't wielding it yet
in wiki it is done by timeline basis, so that i need to have a playthrough. But to safe my time on typing all over again i guess i just copy and paste is the best way to do it
I never done all ear since 3.0 so that other editors have done amazing job on documented the 3.0 era dialogues. Just may be missing a lot of new league dialogues for whatever reason on editors not active on it anymore
I ran All Ears and No Stone Unturned based on the Google Sheet, the instructions still work.
There has been no additions since whatever it was, 3.9
i mean in wiki it miss a lot of new dialogues . I think post 3.13 (Meven) some dialogues has changed too
the all ear wiki page may be up to date. but i am unsure in the npc article it has documented the correct up-to-date dialogues or not
Right, I have no experience with those 🙂
anyway it is the fundamental problem of the wiki. You can have a website dedicated to static content of the game by datamining. But for the wiki, which meant to be edited (not much page except template and module are protected from edit), seems either rely on the player community or setup bot to pour content in it
Is there too much information in the wiki, if there's such a thing?
yeah, may be the solution is leaving an ext link and let people to click it to read the latest info. poedb also has the more complete quest log of each quest. Some quest article is not written in a walkthough manner (as quest log sometimes unclear and misleading) but a partial collection of quest log and quest dialogues
but for the area, monster, skill and modifier, since it is used by wiki editors to describe the interactions, it needed to be datamined into the wiki but currently pypoe is not working in 100% state.
I admit poedb open to "wiki"-like edit would make that site have even more advantage that any traditional Mediawiki-based POE wiki and the forks
for example https://pathofexile.fandom.com/wiki/The_Trialmaster, it has the monster datacard datamined and displayed, but the monster skill is not yet datamined (it "can" be mined but way too messy to do it for all monster skill), while poedb did datamined the monster skill. Both site does not have descriptive info to digest the datamined data or from the experience of player that actually killed the boss, which poedb would have an advantage if player referring to the internal stat ids of the monster skill in their walkthough on the boss fight
In fact, the wiki is maintained by very few people. I have 4 people who have applied for editing permission, but now 0 wiki have written it.
This is another reason. In the past there are more editors
and for the dispute on twinned POE wiki fork (actually triple for the first fork attempt). I think the site ownership can entrusted to Vinifera7 or subreddit mod or GGG themselves and date to date management without dispute Vinifera7 and other existing fandom wiki admins can do that job in new forks. I just never able to understand Vinifera7 was not invited in JourneyToJah's team and let Vinifera7 (and other wiki admins and vet) absent from the discussion and decide to go ahead and decide to insist JourneyToJah himself needed to be the leader in the project
I don't think any of us were contacted, or even mentioned
I have left my comment in their discord after reading the comment (which i re-posted part of it here) of JourneyToJah , and then in their team there is people try to settle the dispute between JourneyToJah and iamacyborg. Don't bother to go there to try to read the followup (which i don't want to boil my emotion and my usually not so assume good faith bad habit).
btw more often than not it's just "armour", not "armours".
You mean grammar in the wiki article or just wrong channel? I admit i have bad grammar and a lots of stuff related to lanuage
oh sorry, i was referring to the templar blurb above 🙂
"The armors associated with the Templar's strength"
English has a few tricky bits, like most languages I imagine 🙂
i think the context is refer to the item (Body Armour, Shield, Helmet) but in-game Armour means Armour (rating) and then the actual armor piece
so that i think the word armor should be substituted but i don't have one in mind
perhaps "The armour types associated with the Templar"?
or even "The types of armour associated with the Templar" but this is less pretty imho
changed
ugly spelling/grammar hurts my eyes, i can't help it lol
regarding the future of the wiki(s), i **really **think GGG needs to help by providing data like they do for the passive tree and item filters.
for example, would it really kill them to release a JSON or whatever containing skill gem info on the day of the league start?
In that case the email should be send by subreddit mod such as blvcksvn (has another alias in discord), not random guy (the thread has upvote count as crowning as the leader of the project?). Or i am not sure OmegaK2 as wiki admin has asked before or not
i'm sure he has before
In their team discord they are wishing GGG bless them and decide to pour data to their project. If it is that easy what not GGG host a wiki like a few years ago?
wiki are meant to be edit by editor (i believe player have way more time than GGG employee to this voluntary job) unless GGG decide to host one. And they decide to not host one for a long time why suddenly they will change ?
I still hope they release json so that i have day 0 POB btw
🙂
can you imagine trying to keep a wiki up-to-date every league without datamining anything?
7.i. Reverse engineer, de-compile or disassemble the Website, Materials or Services or seek to establish the technical processes, operations and communication protocols of the Website, Materials or Services through any means, including without limitation by reference to the input or output of the Website, Materials or Services or the internal structure and workings of the Website, Materials or Services.
well, the wiki already have outdated modifier . it is not updated in 3.13 as desptie pypoe still work at that time. no one run a wiki bot . While ruba has done it in ru-wiki.
exactly @stone crater , GGG are actually the opposite of helpful at the moment 😉
when they remove QuestVendorGems.dat from client, I ask for help. and they response they will remove anything that client don't need to prevent datamining
this is just the umbrella clause that reserved to use but not using right now . At least i have used pypoe and submit some pull request to the pypoe fork and i am not banned yet. I even talk about pypoe in public in official forum. For sure it is more than happy if they release all mod in json, but i don't think they will.
ZiggyD, well-known friend of GGG, has asked mod group problem in public and we still rely on datamine to know mod group or learn it in hard way by bricked item from woke orb
the current situation is certainly frustrating from our point of view
maybe they could make full data for an item available once it's "dropped" once?
That's correct. I posted some concerns on the initial reddit thread posted by JourneyToJah, and was basically told, "It won't be an issue." JourneyToJah has not reached out to me before or since.
have they completely removed QuestVendorGems.dat from the client? they don't add it back in on launch day?
no. i think. So that some quest reward in wiki was stuck in 3.11 as it need people like me to read the patch note and manually change it (Luckily they did mention the change in quest reward, but at the same time for sure they can display in their website that the quest as of that league, ALL the quest reward instead of changed rewards only)
well that seems like a disproportionately aggressive response... how many hours until someone finds all the new gems on a vendor anyway?
they are in patchnote, but sometimes wrong
the company is terrible in doc in general. Someone asked question in official forum and he has read all the item filter announcement since 2.0 and spotted unannounced change and then the official page https://www.pathofexile.com/item-filter/about is somewhat outdated
patchnotes are manual intervention too, which is not ideal
their official web site still link to https://pathofexile.gamepedia.com/ as official wiki
do they actually gain anything from GGG as the "offical" wiki though, other than the title?
at least they renamed the text in the pull down menu from official wiki to community wiki
any plan to change it to the new one?
oh sorry chuanhsing i missed your point entirely 😉
Don't know, other team has sent email to wish to have GGG blessing . But until the poe wiki fork is stable i don't think it is the time to cut the traffic to our (maintained by existing active editors of fandom wiki) wiki fork
so that some people spotted the error and then report to GGG as missing quest reward, since so many people play league to have a fresh start and check the gem they want from the quest (or most of the time false positive due to rely on outdated wiki that since we all are volunteer, so that can't make it up-to-date all the time)
i think chuanhsing was pointing out that the link should be changed to https://pathofexile.fandom.com/
for now at least
this is minor detail that gamepedia is a redirect to fandom
What other team though?
@crimson gulch Do you want to install this extension? https://www.mediawiki.org/wiki/Extension:Interwiki
Sorry man was late my time, free today
Will do
JourneytoJah and co have created a new Discord and are working on another fork there
They’re struggling with a lot if basic MW stuff though and there’s a lot of talk about stuff they want to add or have on the wiki that shows they don’t really know or understand what’s already there
I’ve mentioned multiple times over the last couple days they should come speak to editors here but…
They’re also trying to fork PyPoe without talking to Omega
PyPoe is public project, should be fine just fork
Yeah just seems weird to not speak to the creator to understand what needs changing
He should think that the wiki has not been updated for a long time, even if he finds the original team, there will be no improvement. Therefore, he intends to form a brand-new team by himself to solve the current dilemma.
I can understand a group breaking off and doing their own thing if there is an irreconcilable difference in vision between them and the mainstay group, but again, that discussion never took place.
On the bright side the creator of Cargo responded to a query I had about dupes so hopefully we can see some movement there https://www.mediawiki.org/wiki/Extension_talk:Cargo#Also_suffering_from_duplicate_rows
The problem is they never really start anything and in the stage of guessing and reverse engineering the wiki template and already want their project is one and only one for the community
It just can't play a blame game in us for allegedly suddenly had an idea of fork in the past weekend. It is success or not still depend their own team .Even we don't fork , they still need to be assume people have a choice to read their and fandom.
Running this again https://www.mediawiki.org/wiki/Manual:Rebuildall.php
Should be faster now there are more CPU cores available
Just noticed the Math extension is missing, will add
The creation of data structures, and storage of data, is done in Cargo exclusively via templates.
Any template that makes use of Cargo needs to contain calls to the parser functions #cargo_declare and #cargo_store; or, more rarely, calls to #cargo_attach and #cargo_store.
#cargo_declare defines the fields for a table of data, #cargo_store stores...
@crimson gulch You were wondering whether we set fields that should be unique as unique
Yeah so one thing I'm thinking is maybe we've not set an actual unique key on the tables, hence why they're storing duplicates
Well, every table has a unique key built in with cargo. It's the _pageID field
At least, I think so. Let me actually check on that
The creation of data structures, and storage of data, is done in Cargo exclusively via templates.
Any template that makes use of Cargo needs to contain calls to the parser functions #cargo_declare and #cargo_store; or, more rarely, calls to #cargo_attach and #cargo_store.
#cargo_declare defines the fields for a table of data, #cargo_store stores...
My bad, it's the _ID field. The _pageID field is not unique
Yeah, that might be the issue
I'm not familiar with Lua but this looks like it has potential?
`-- mw.logObject(m_cargo.map_results_to_id{results=mw.ext.cargo.query('mods,spawn_weights', 'mods._pageID, spawn_weights.tag', {where='mods.id="Strength1"', join='mods._pageID=spawn_weights._pageID'}), field='mods._pageID'})
function m_cargo.map_results_to_id(args)
-- Maps the results passed to a table containing the specified field as key and a table of rows for the particular page as values.
--
-- args
-- results : Table of results returned from mw.ext.cargo.query to map to the specified id field.
-- field : Name of the id field to map results to
-- the field has to be in the fields list of the original query or it will cause errors.
-- keep_id_field : If set then don't delete _pageID.
-- append_id_field : If set then append the id to the table sequentially as well which allows preserving
-- the id order they were found in.
--
-- return
-- table
-- key : The specified id field
-- value : Array containing the found rows (in the order that they were found)
local out = {}
for _, row in ipairs(args.results) do
local key = row[args.field]
if out[key] then
out[key][#out[key]+1] = row
else
out[key] = {row}
-- Append the ids sequentially, this allows preserving the order
-- the ids were found:
if args.append_id_field ~= nil then
out[#out+1] = key
end
end
--Discard the pageID, don't need this any longer in most cases:
if args.keep_id_field == nil then
row[args.field] = nil
end
end
return out
end`
What am I looking at?
This is part of this page https://www.poewiki.net/wiki/Module:Cargo
Oh, okay
So I'm looking at how the mods table is declared. It looks like the id field is not actually defined as unique.
that is, the id field, not the _ID field
I wonder if that may be the issue?
It might be. Thing is, it doesn't explain why cargo dupes rows
No, it doesn't, and what's weird is doing a null edit saves two rows but further edits don't create additional dupes
yeah so if you check here you can see what's stored in the table https://www.poewiki.net/wiki/Special:CargoTables/mods
I've been finding a mod that's not in there, and then null editing it
it creates two rows
doing another null edit doesn't add anything
In the backend I've also deleted one of the rows to see what'd happen, doing another null edit in that case recreates the dupe row
Click on Modify query to see the query
It looks like it's just a straight up dupe. The only difference is the value of the _ID field.
Yeah, so maybe on that table we need to set the pageName or pageID as a unique key
Is that even possible to do?
Yeah so I think you can add the unique parameter to a field definition to define it as unique
The creation of data structures, and storage of data, is done in Cargo exclusively via templates.
Any template that makes use of Cargo needs to contain calls to the parser functions #cargo_declare and #cargo_store; or, more rarely, calls to #cargo_attach and #cargo_store.
#cargo_declare defines the fields for a table of data, #cargo_store stores...
Okay, but the _pageID field is declared automatically. You don't declare it.
maybe on this field definition on line 98 of Module:Mod?
id = { name = 'id', field = 'id', type = 'String', wikitext = i18n.tooltips.id, },
So what happens when Cargo decides to create a duplicate row then? It throws an error? That's not very helpful.
I think it just won't save the new value
or at least, another value
Hopefully it'll still allow the value to be overriden when it needs to be
Let's put that idea on hold. What I want to try is setting up the attachment template for the mods table and rebuild the table that way.
That is, if you're done running scripts and stuff.
54,000 pages in, another ~12k to go
Okay. I'm not going to edit anything until that's done.
But if you try create the attachment table that'd be fine
and then we press to button to rebuild when the other script is done
Cargo is just sensitive. You remember that we created the Mod namespace after the mod pages were already in, right?
yeah
Well, I'm pretty sure that's what's screwing it up.
It was happening before then I think
We just need to let Cargo rebuild the table.
let's do that
Shouldn't I wait?
How long do you think it'll take to create the attach template?
Five minutes...
ah yeah then let's wait a bit
This take a while https://www.mediawiki.org/wiki/Manual:Rebuildall.php
I don't know what that does exactly, but I just thought I should wait while it's running.
It should be safe to abort via ctrl+C if you get bored.
I am definitely bored
It should actually be fine if you go ahead, this script won't touch cargo tables
Oh, okay
@grand magnet If you have a minute, I could use your help. Where does Module:Mod actually store rows to the mods table? I'm just not seeing it. https://www.poewiki.net/wiki/Module:Mod
m_util.args.from_cargo_map
there's a store at the end of it
m_cargo.store(frame, cargo_values)
Seems a little circular though...
That util function should probably move to m_cargo if it isn't there already
Yeah I see. That util function seems like it's doing things over and beyond what it's purpose is. It should just validate args...
m_cargo.parse_field_arguments might be the "modern" way
monsters uses that one
Might be possibly to replace m_util.args.from_cargo_map with m_cargo.parse_field_arguments.
Good to know
1 row populated every 5 minutes. https://www.poewiki.net/wiki/Special:CargoTables/mods?_replacement
I'll just check back next week then.
It runs based on traffic to the site, I'll run the jobqueue script once the other is done
actually fuck it, this one is more important
This should run through 10k items in the queue
Running through the cosmetics table right now
Looks like we've somehow got some triplicates in the Areas table https://www.poewiki.net/wiki/User:Iamacyborg/test
I wonder if we can just modify the queries on pages so it ignores dupes
That way stuff works, even if the DB is messy
Reassuring from the extension dev 😬
Okay, yes, I see the problem. That's very strange, and I don't know why it's happening.
😅
Ok looks like the mods table is populating now
They seem to all be going at the same time.
yeah, it's weird, at least the number is going up in the right direction 😅
no dupes in those tables so far
you can query the replacement tables like this
{{#cargo_query: table=cosmetic_items__NEXT |fields=_pageName, COUNT(_pageName) |having=COUNT(_pageName) > 1 |group by=_pageName |order by= COUNT(_pageName) DESC }}
193k items in the job queue
The cosmetic_items table appears to be done, so I'm going to replace
sounds good
No dupes.... yet
running again, things might slow down as I've set it to ignore memory limits
queue is down to 170k
~149k looks like this should clear relatively soon
~100k
seems to be running through a bunch of cache update jobs at the mo but slowly clearing down
~41k
🤞
Data's coming in without dupes https://www.poewiki.net/wiki/Special:CargoTables/mods?_replacement
Looks like it's stalling at 2661 mods, I wonder if that's to do with the linktable or something
On the bright side, null editing a modifier that's in the new table didn't create a dupe
Hi, I heard the poe wiki is forking from fandom. Im curious if that's true and if theres some sort of current status on the new wiki. I edit quite a bit on the oldschool runescape wiki, which forked from fandom a few years back and the POE wiki has been talked about a bit among the editors there. Might be useful to talk to some of the folks there if new forks of the poe wiki are spinning up and you want help from a wiki that has successfully split from fandom
no worries if not, just thought Id throw it out there
Yep. The situation is that we're currently working on building the data tables on the new wiki poewiki.net. It's not quite ready to go, so we haven't been promoting it.
I should also tell you that there is a competing wiki project that was started by someone else. They began work on their project without contacting any of the existing wiki editors, so it seems that they want to go it alone or form a completely separate group of editors.
Oh interesting. Maybe thats the reddit post I saw then 👀 Cool to hear that youre working on a fork though, hopefully it goes well. Feel free to reach out if you want help or advice, I can forward you to the right people on the osrs wiki
Just check this space from time to time.
Yeah, Ill keep an eye on this for sure
Definitely would recommend trying to consolidate the various forks
You're going to have a hard enough time winning against a dead fandom wiki
The technical migration is probably the easiest of the three migrations you need to successfully fork away from fandom
What do you identify the other two as?
Migrating the editor community, and migrating the reader community
Yeah, that's understandable.
I don't have much of a sense of where the editor community is/whether any people are interested in staying on fandom, but even if you're very successful, you'll still have some low-information editors not know about the fork inadvertently working on the old one
And it can be challenging to communicate with them
Yeah you can’t really advertise the fork on fandom itself
And then "migrating the reader community" is basically code for SEO
Right lol
I run the runescape/osrs wikis btw - we've been through those exact hurdles before
The seo thing I see as a multi-year process tbh, even optimistically a dead fandom wiki is going to outrank a new one even if it gets a lot of links from other high ranking sources
Doesn’t help that the comms guy for fandom is a mod on a bunch of gaming subs
It's a multi-year process for sure, but you can probably get to a majority hold in a matter of months if you know how Google works
There are a lot of discoveries we made along the way
and FWIW I think most folks' dislike of fandom far outweighs loyalty to a particular reddit mod
as you can see from the fact that all of these "fandom sucks, let's make a new wiki" posts get like 3000 upvotes, lol
We had an opportunity to go to Gamepedia in about 2014, and boy am I glad we dodged that bullet
Yeah, previously it was "Gamepedia sucks", even though Gamepedia was awesome in comparison.
anyway, if you guys get stuck anywhere, I'd love to help out
I’m honestly pissed that a single company has a stranglehold on gaming wikis
if you were a bit earlier in the process I'd offer to host POE on the same infrastructure we do for osrs/runescape
Uh, yeah...
I’ve got it on a DO box but would love to pick your brain on that front
touhou wiki had that exact problem when they forked -- it looks like touhouwiki.net is finally outranking the mostly-dead fandom wiki but it took years for that to happen
The main difficulty of SEO-forking is beating down Google's duplicate content penalty
most people don't know about this
but quite often, your forked wiki will be #2 for some query behind fandom
but due to the way google tries to not show the "same page" twice, your 2nd-place result won't show up at all
so you need to do some clever things to make it show up
and then eventually, the goodwill from your community will mean enough people click the #2 result to get it up to #1
but you won't be able to do that last step unless you do the clever things to get it to show up in the first place
You’ve given me a lot to think about
that's definitely a phase-3 type of issue though
Still need to iron out some issues with Cargo 🙃
getting the technical stuff and getting the editor community are much more important earlier on
I will also say it's pretty important to get the editor community as on-board as you can, before fandom starts locking things down
Or at least, have some way to communicate with most of the power-users
How do you mean "locking things down"?
I guess we have the benefit of most folks being here
they will probably desysop and block all of you eventually
I guess so. I don't expect that will happen any time soon.
We have admins on there who haven't been active in a long time as it is.
I guess I don't have too much experience with the current crop of fandom community managers, but previously things got very nasty very quickly
If it’s anything like Gamepedia they’ll pay someone a couple hundred bucks a month to take over and run things
it's all fun and games until you're actually threatening their revenue stream
And expect them to work 20+ hours a week for that