I edit articles on
... and related subjects.
I have not had too many bad experiences here with admins on power trips and that's probably why I'm still here.
This is partly because the articles I edit are not terribly popular. Similarly, Reddit general communities (e.g. based on geography) are the most toxic while the people in niche subreddits are generally supportive.
If the topic or subject is very interesting to me, I will spend quite some time doing research.
Examples from cryptography: Mary D'Imperio, Edward Scheidt, and Two-square cipher.
The more I edit, the less I trust statements without any citation. Over time, these statements are cited outside Wikipedia, even in "reliable sources" like newspapers and published books, leading to entrenched mistakes (circular sourcing).
Example 1: The French article for the film Trafic stated it was funded by the partner of the lead actress Maria Kimberly (Alec Wildenstein) because the director Jacques Tati was bankrupt. This sounds entirely plausible but I could never find a reliable citation for it. Over several years, I started to see the claim outside Wikipedia in various places. The claim shouldn't be in a Wikipedia article.
Example 2: There was an image going around with a picture of the Shah and the heading "Guess who's building nuclear power plants". It even made it onto Nuclear program of Iran at one point. But there was no source given; it just looked interesting to some people and probably supported their prejudices. Discussion at Wikipedia:Files_for_discussion/2022_February_15.
I love going back over old games and seeing how terrible annotations really were!
Examples:
[edit]An opening: 1. e4 e5 2. f4 exf4 3. Nf3 g5 4. h4 g4 5. Ne5 Nf6 6. Bc4 d5 7. exd5 Bd6 8. 0-0 ... the best move is 8... Nbd7 which absolutely refutes the whole idea, but because it has not been published in any WP:RS sources it is not in the article.
[edit]I have been looking at this game since it started in 1999, with whatever computer chess program is rated highest, with the fastest hardware I could get.
I think my ultimate conclusion is that the only moves worth commenting on are the ones which definitely changed a win into a draw; or a draw into a loss. Michael Nielsen wrote a book Reinventing Discovery where he talked about the supposed important difference between 26... Bc5 and 26... f4; my current conclusion is that they are both about the same, and I don't think this conclusion will change.
So:
So, the page really needs updating but as there's no WP:RS source yet, these aren't in the article.
[edit]I have to be honest. 5. Nxf7 deserves a question mark because Black gets at least a draw, contrary to the Maarten de Zeeuw articles from New In Chess Volumes 63 and 65-68 in 2002-2003. I used to love this opening as a junior, but at higher levels very few people play it.
After 5. Nxf7 Bxf2+ there is no difference between 6. Kf1 and 6. Kxf2, contrary to the article, White can only draw with either of these.
Also, it seems the best move after 5. Bxf7+ Ke7 is actually 6. Bc4, not in the article.
NoWP:RS source for those yet.
Example: UFO (band) and UFO Club - nobody has found a published link between the two.
In "High Stakes & Dangerous Men: The UFO Story" (Neil Daniels, 2013) a nominally WP:RS source, he writes "Noel Moore of Beacon Records, at the UFO club in London in October 1969, caused the band to rename their outfit UFO." This cannot be true because the UFO club closed in August 1967. It looks like a quote from English Wikipedia, where the idea was introduced in 2004. I deleted this idea in August 2023.
Another problem with Wikipedia is that news sources are not ranked in terms of quality; they are only "black and white".
Example: the Daily Mail and The Sun (United Kingdom) are not WP:RS sources. But in the discussion about Denis Michael Rohan who was living in Australia, ABC News Australia was treated as equally as reliable as ynetnews from Israel; which cannot be correct.
Sometimes individual articles at The Conversation don't have a neutral tone; although The Conversation is listed as a reliable source here.
Three that come to mind are:
I have been on the Kryptos discussion group for a few years and unfortunately it has gradually become a wasteland. Contributors who could offer the most (e.g. former Kryptos section solvers Ed Hannon, Denny McDaniels and Jim Gillogly, professional cryptanalyst Jim Reeds, Mark Armitage and Robert Matson, who helped break the codes of "Can you crack the Enigma code?" by Richard Belfield, Steve Roberts former GCHQ/DSD, Bill Briere former cryppie; various American Cryptogram Association people) have left or post little; the moderator blocks links from some people but not others, and code; while posters who suffer from Dunning Kruger dominate the discussion with gibberish, theories about Nostradamus and alphabet soup. There are some gems but you have to grope through gigabytes of garbage. Truly, a classic source-ranking problem, which motivated me to write my own Kryptos paper for HistoCrypt.
Tirath Hassaram Khemlani.
There is no separate article about him here yet, just Loans affair.
He was a key player in the Loans affair, which brought down the Whitlam government.
He had connections with shady institutions like the Nugan Hand Bank and the American Mafia but there is no one comprehensive page where all of this is explained. There are a number of reasons:
Trove and Kissinger cables are good sources.
Australian Dictionary of Biography Obituaries Australia
To illustrate his obscurity:
Mr Game said that Mr Khemlani's daughter, Shanti, who is married to an Australian and lives in South Australia, only learned of her father's death on Wednesday when he had told her. (died 19 May 1991, article 13 September 1991)
Most National Archives of Australia material (search "Khemlani") ends in 1975.
There are many questions which don't have clear answers and WP:RS sources are paywalled, not online, or very obscure.
In particular, I distrust quotes that cannot be traced back to the original source. Particularly untrustworthy quotes are ones that confirm some sort of prejudice on the part of the reader or sound uncharacteristically "pithy".
"The world is a great book, of which they that never stir from home read only a page. " - not St Augustine.
"The three golden rules to ensure computer security are: do not own a computer; do not power it on; and do not use it." - very pithy! But not Robert Morris (cryptographer)orRobert T. Morris.
"The way a government treats refugees is very instructive because it shows you how they would treat the rest of us if they thought they could get away with it. " - in endless memes, but not Tony Benn, actually Neal Ascherson paraphrased
"Markets can remain irrational longer than you can remain solvent." - not John Maynard Keynes, but A. Gary Shilling
This is very important for Wikipedia research. There is a page by Gwern Branwen which is a list of sources and techniques. He explains how to use Google search with modifiers. I have always found that PDF files are likely to have higher quality content and over time this defeats people who use search engine optimization and listicles to try to drive your searches to their low-quality ad-infested site. "filetype:pdf" restricts searches to PDF files.
I have learned much from reading Hacker News - as Google has tried to "personalise" the search over time, results have become less relevant for the user and more useful from the advertisers' point of view. People often suggest "site:reddit.com" "site:stackoverflow.com" "site:stackexchange.com" as modifiers to hopefully lead to higher quality content.
Google Groups (formerly Dejanews) has a terrible interface; as you try to page through results sorted by date it may suddenly decide you've seen enough results and stop, just before you reach what you need!
The "books3" archive from "The Pile" of 196,640 books is a great reference for building large language models and searches -- obviously copyrighted.
Other great sites:
