The resurrection of atari-wiki.com

[Smonson]

Nice work! Let me know if I can help in any way with the server stuff. Otherwise I'm ready to edit some articles.

[Cyprian]

Great initiative @exxos :

Btw, there is a mirror of Atari Wiki:
https://www.temlib.org/AtariForumWiki/
https://atari-forum.com/viewtopic.php?p=444358#p444358

and there:
https://beyondbrown.mooo.com/post/hardw ... sting-8.2/

[exxos]

Btw, there is a mirror of Atari Wiki:
https://www.temlib.org/AtariForumWiki/

I believe the guy hosting that page has been in touch with icky about it. I don't know what changes were made to that wiki over the original database. I don't think there will be many pages different but I'm sure we can copy them over at least if needs be.

We are still fighting the spam currently. Icky thinks something got hacked on the wiki years ago allowing the creation of spam accounts. So finding that vulnerability is also taking some time. But progress is being made in fixing and solving all the issues.

We want to clean out all the spam and make sure it's secure before it officially goes live again.

[Cyprian]

We want to clean out all the spam and make sure it's secure before it officially goes live again.

great, thanks

[sporniket]

Thank you for your effort

[Icky]

We are posting progress of the wiki reboot on the holding page.

https://www.atari-wiki.com

Screenshot 2023-06-10 at 11.20.13.png (85.27 KiB) Viewed 2367 times

[JezC]

Is this to be renamed the Icky Atari Wiki?

[stephen_usher]

Is this to be renamed the Icky Atari Wiki?

[exxos]

I have been looking at methods for deleting spam in the database. problem is there is literally over 2.2 million posts

I've wrote a script (Well ChadGPT did) which dumps all the pages, sorts them, then does a percentage match .. so it looks like this...

Code: Select all

5_Things_To_Hold_In_Idea_When_Choosing_The_Best_Automobile_Mending_Concentrate_2021..._Advice_Num_26_From_673) 68 20
5_Things_To_Hold_In_Idea_When_Choosing_The_Best_Car_Amend_Halfway_2021..._Information_Number_32_Of_118) 67 48
5_Things_To_Hold_In_Idea_When_Choosing_The_Best_Car_Animate_Shopping_Mall_2021..._Tip_Num_24_From_819) 75 53
5_Things_To_Hold_In_Idea_When_Choosing_The_Outdo_Automobile_Remedy_Nub_2021..._Tip_No._42_From_375) 72 43
5_Things_To_Hold_In_Idea_When_Choosing_The_Outflank_Motorcar_Resort_Halfway_2021..._Info_No._27_From_105) 74 46
5_Things_To_Hold_In_Idea_When_Choosing_The_Scoop_Motorcar_Quicken_Shopping_Mall_2021..._Info_Number_36_From_958) 74 43
5_Things_To_Hold_In_Idea_When_Choosing_The_Trump_Machine_Indemnify_Meat_2021..._Tip_Number_25_From_651) 73 43
5_Things_To_Hold_In_Intellect_When_Choosing_The_Best_Auto_Recreate_Mall_2021..._Info_Num_21_From_374) 67 21
5_Things_To_Hold_In_Intellect_When_Choosing_The_Best_Machine_Reparation_Focus_2021..._Information_Num_20_From_959) 80 53
5_Things_To_Hold_In_Intellect_When_Choosing_The_C._H._Best_Motorcar_Stamping_Ground_Gist_2021..._Info_Number_27_From_791) 74 48
5_Things_To_Hold_In_Intellect_When_Choosing_The_Charles_Herbert_Best_Automobile_Fixture_Centerfield_2021..._Info_Number_24_From_567) 73 49

Where the first number on the end is the % match to the previous line, and the number after is matching characters from the start of the line.

I've only output matches over 10% and below 90%.. as there are things like "my atari book part 1" then "my atari book part 2" kinda thing. Which gave over 95% match.

Then there are ones like this. Which makes things a bit tricky.

Code: Select all

SQSBillie485780) 55 2
SQSPriscilla) 48 3
SQSRoger23191) 37 3
SQTAnke411381) 42 2
SQTCandice) 48 3
SQTClaude8587) 64 4
SQTMitchell) 38 3
SQTRae8576630) 38 3
SQUCallum31747) 34 2
SQUDon423980) 35 3
SQUGia112726407) 41 3
SQVBrigida) 37 2
SQVJacques) 45 3
SQVMercedes) 60 3
SQVRaquel224657) 35 3
SQWAntoinette) 26 2

I've had to put some filters in there for "atari" and "jpg" etc, but with so much spam, there is no way to know if its binning genuine topics. I doubt it's caught all the spam, but it's no doubt got most of them.

I also had to put a exclusion filter rent for the words "Atari" and "AL_" and "JPG". The "AL_" seem to be Atari legend pages listing games I think.

So below is the output of entries which should be all be spammers people can have a look through it as much as possible to see if there looks like there is any Atari content in it anywhere...

results5.zip

(22.19 MiB) Downloaded 87 times

It is pretty amazing that software as well known as wikipedia has basically zero tools for combating spam which has already polluted the database, and yet, it has many tools for the prevention

There ares still over 200,000 posts to clean up. But as long as the bulk is gone, then it is at least a start.. I am still thinking of spam detection methods. But I think we could potentially end up losing the page here and there because human verification of so many pages isn't really feasible.

[chronicthehedgehog]

Had a few scans up n down - didn't see anything that needed to be kept