It is shorter, but only by 19 removed words and 6 removed answers. Compare for yourself:

https://www.nytimes.com/games/wordle/main.bd4cb59c.js

https://web.archive.org/web/20220210034340id_/https://www.po...

Oh, hmm. Perhaps the list has been changed multiple times? It was originally 12972 words (see https://www.poirrier.ca/notes/wordle-optimal/) and was 10638 words as of yesterday. Anyway, some of the words that have been removed are slurs/profanity.

I believe you're mixing it up. There are two arrays of words in the source code. One that contains all the solutions in order, and one that contains all the possible words without repeating any words in the solutions list

So the 10638 words are the set of all possible guesses minus the set of all the solutions.

Oh… I think you’re right. Thanks.

The one you linked also describes the old word lists similarly: https://www.poirrier.ca/notes/wordle/#variant-cheat-mode

It allowed cunts as a guess today :p

There's a bit of confusion here because there's actually 2 word lists. There's a list of ~2,300 words that are ACTUAL words that will some day be the word of the day and there's a list of 12k POSSIBLE words you can enter in

I think OP is correct about them removing ~2k of the possible words

The “possible” words are added to the “answer” words to get the full list of guessable words, which is still around 12k just like it always was.

What's an example of a word that was removed? Any pattern there?

The day they moved to the NYT they removed words that could refer to the female anatomy (like "pussy"), as well as words referring to certain unpleasant concepts (e.g. slavery). They retained words referring to male genitalia (like "boner"). Make of that what you will.

They removed them from the allowed guesses list, or the possible answers list? I might be convinced about removing them from the possible answers, but the allowed guesses should be basically all possible English 5-letter words. Even if they make some argument about kids typing rude words in a classroom activity, well, kids are going to do that anyway...

Removed from the allowed guesses. If you guess "slave" and hit enter you'll receive an error.

I suppose they're both vulgar, but one difference is that "pussy" has a derogatory use that "boner" doesn't have.

Perhaps, but it also has an entirely innocent use that boner doesn’t. At least here in the UK it’s pretty common to use “pussy” to refer to a cat.

Pussy and boner both have innocent uses. A pussy is a cat, a boner is a silly mistake.

It seems inconsistent that dick (which can refer to a penis, a jerk, or a Richard) is still allowed, if they're worried about derogatory meanings.

Quick edit: DICKS is allowed, DICK is only 4 letters

My wife was returning something from the store Dicks. She needed the receipt and knew it had been emailed to me. A quick check and I didn’t see any such email.

So in all seriousness, while I’m at work on a open floor plain facing computer she asked me to search my spam folder for Dicks.

They notably did not remove CUNTS.

"Boner" can also mean a tremendous mistake -- negative connotation, but neither a slur nor a genitalia reference.

Which makes it distinct from a pussycat how?

The game has two lists -- one list that includes 'valid guesses' (things you can enter), and one that is a list of things that could actually end up as the solution.

From the possible solutions list, they removed words that have negative associations -- stuff like "slave."

From the guesses list, they mostly removed things like slurs.

What's the game's reason for having two lists instead of just one?

A lot of valid 5-letter words are so obscure as to make the game not fun when you inevitably lose.

If the answer to a day's Wordle was QIBLA or AALII or RAHUI, I'd probably stop playing.

Here's the full list of valid guesses. There are a lot of words here that most people don't know.

https://gist.github.com/cfreshman/cdcdf777450c5b5301e439061d...

At least in the first version of the game, the 'solutions' list was actually just a sequential list of what the answer would be for each given day. So, calling them 'possible solutions' on my part was actually a little misleading (sorry), there's no probability going on or anything like that. Not sure if NYT changed it.

The value is that we all get the same answer, so which makes it more fun to compare our results (this does mean that you can just cheat and look up the answer, and it might have been possible to obscure the list somehow, but probably getting into an escalating war with cheaters was not a priority for such a simple little free game).

I assume the logic is that they shouldn't have "unusual" words as answers but shouldn't gatekeep too much about what words you enter as that would be frustrating.

https://boingboing.net/2022/02/11/pussy-pupal-and-agora-amon... "Lynch", "slave", a series of racial slurs, etc.

A previous post here has the list: https://news.ycombinator.com/item?id=30294688

I think you're technically correct originally. They only removed less than two dozen from the ACTUAL word list but they did in fact remove ~2k from the list of possible words you can enter in (which is the list most players interact with)

Do you know of any rehosts of the old Wordle code?

I made a copy the day The New York Times bought Wordle: https://buchh.fun/wordle

Thanks to Josh for making Wordle with a free license.

Hmm. The #241 word was changed between original and NYT, and I just played your copy and it was using the NYT word instead of Josh's. I don't want to spoil the 2 words here so here's a link to Mashable that covers the two different words to the same puzzle number. https://mashable.com/article/wordle-241-glitch-answer

I checked the other commenter's rehost and it also used NYT's word. So does this mean all supposed "rehosts" don't and can't work if the word list still defaults to NYT's?

Nevermind, the Mashable article misreported what happened. NYT didn't replace Agora with Aroma, they removed Agora and Aroma was simply the next one in line. My apologies, your rehost is fine.

Original: #241 agora, #242 aroma, #243 caulk, #244 shake

NYT list: #241 aroma, #242 caulk, #243 shake, #244 <not out yet>

Because I got Aroma instead of Agora on your rehost and didn't check the puzzle number, and because I took Mashable's word that Aroma was a replacement for Agora, I mistakenly thought your rehost was using NYT's word list.

It could still be improved. e.g., guessing right on the first word should evoke "lucky", not "genius".

Did the original Wordle have a free license? I don't recall seeing license information anywhere.

It's still there in the JavaScript file.

> Copyright (c) Microsoft Corporation.

It appears to be the Zero-Clause BSD, but for some reason, it says Microsoft holds the copyright. Could original author must have used something with the same license to write Wordle? If it isn't a free license and I am misinterpreting the code, I would be happy to take my version down.

Why?

We can only guess, but perhaps they've applied some sort of commonality filter and dropped words that are rarely used? Or perhaps they've simply moved more words from the guess list into the challenge list? (the challenge list was produced from what words one person immediately recognised, after all)

Or, most likely IMO, they were not 100% certain they could legally use the existing word list (now, or for what they plan for the future) or were certain they couldn't without paying for some sort of license, so they produced a new list from scratch from a source they have licensed for their other word games.

There are two lists: guesses you are allowed to make, and possible solutions from the puzzle. From the solutions list they removed some words that had different British and American spellings, as well as some words with negative associations, like "slave."

From the possible guesses list, the words removed appeared to mostly be quite offensive slurs.

Ah, good. The Americanisms did my nut in. 5 thumbs up.

* Americanizmz