Hacker News Reader

Comment by femto | Hacker News Reader

> ...update the JS each day with the new word

Even more hands off than that. The script contained the entire list of words and each day it just moved to the next entry in the list. It makes sense, in that originating as a fun personal project the author wouldn't want to be burdened with a daily update task.

Edit:

Out of curiosity I had a look at the Wordle on the NYT site, and it's still the same script.

View on HN

That’s a perfect example of solving for the problem you have and not over engineering.

But if that is the case, then that would not apply to the wordle archive, right?

If the list of words for every day is encoded into a static code, why does the wordle archive need to crawl the site each day, or set the system time to different days, just to get the words for different days... Seems to be a bit over engineered to me...

It's also a perfect example of actually understanding the problem you're trying to solve (which is a prerequisite for not overengineering).

Wordle is a tiny word puzzle game with no stakes. So, while this architecture makes it trivially easy to cheat, there's also zero reason to do so. You're just robbing yourself of the fun in order to... do what, exactly? Brag about it?

Unfortunately it also allows to rob other people of the fun, as there has been multiple Twitter bots spreading the future words. I think one even replied to people's Wordle score tweets.

That's a good example of how just a few malicious knobs ruin the fun for everyone else. I suppose blocking those bots isn't difficult but the fact that someone went out of their way to code one is so sad.

Well... Aren't the twitter bots then the bad actor?

I wouldn't complain that wikipedia exists when I am answering a quiz...

I don't think a more complex wordle implementation would solve this. If the whole world is guessing the same word, there will always be someone that tries to spoil it for others, just because they can.

> Well... Aren't the twitter bots then the bad actor?

Of course. But what does that matter? All you've done is assign blame. The person still had their game experience ruined. With a different design, that might not have been possible. This is like spam protection -- it's not Gmail's fault that you're receiving spam, but they're the ones in a position to stop it.

At a certain point, it becomes more efficient to consider it societies problem.

If bad actors in person went door-to-door attempting to break in by picking locks, would you say that it is purely the responsibility of the homeowner to ensure that their locks are strong enough?

At a "certain point" sure, but I don't think Wordle spoilers rise to the level of "society-level" problems. If we're not at that point with spam email, I can't imagine we're at that point with Wordle spoilers.

If it weren't stored in the JS, the twitter bots could have just played the game quickly and found the answer. Making is slightly more annoying wouldn't deter them.

That'd only apply to today's answer. Now they're able to tweet the answers for the next week.

Don’t you think that a bot could not brute force everyday’s wordle every morning for a tiny amount of processing power? There are solving strategies out there that will solve it with a minimum amount of attempts. So all it would do is add a tiny delay to todays bot being a nuisance.

That’s still slightly different though. If the bots were just for today’s word, and I were the type of person who likes posting my streaks, I could realistically avoid checking twitter until I’ve solved today’s puzzle. If instead the bots are posting future solutions, it’s much harder to avoid having something spoiled.

Agreed but I do think a lot of people project into the future about “what could be a problem” and solve for that. I would guess it’s very common. I’ve seen it done and done it myself. You say to yourself “well if this becomes popular, I won’t want people to know the next word”, which sounds reasonable but it stops things from getting done.

Hence the "actually understanding" part, that's precisely the point. What is, vs what could be.

A hard lesson I've learned for naming things, is that it's almost always better to name something for what it _is_, _right now_, than for what you think it should eventually become.

The NYT removed a few words from the original list that were seen as more obscure, but otherwise it seems like they just copy-pasted the client-side code.

> Out of curiosity I had a look at the Wordle on the NYT site, and it's still the same script.

Really? I had a feeling the game/wordlist changed somewhat after the move to nyt - I guessed they'd put some crossword cryptist on it... But I suppose I was wrong.

Ed: AFAIK the js uses the client side date to pick a word - and today nyt and the ipfs mirror posted here does not share the same solution:

https://www.nytimes.com/games/wordle/index.html

https://bafybeic4blel5vf4il73n3nzt6vw7npsov6or3cp3myjms4npii...

The wordlist did change. https://arstechnica.com/gaming/2022/02/heres-how-the-new-yor...

Thank you. But that still indicates I was wrong - the possible solutions appear to be essentially the same - even if particularly different. I was surprised to see "slave" as an illegal guess.

But it would appear that the difference in previous solution: stairs vs royal(I think ?) might just be slightly different word lists for the solutions... I thought there was much more of an edit.

They changed some obscure and offensive words initially, but there have been recent news articles around days where people had 2 different solutions to the same puzzle.

Initially I suspected that this was NYT watching for how unusually difficult words play out and then changing the answer on the fly to be a simpler word. But I think in reality these were the same changes they made shortly after the acquisition and somehow people have the old JS file cached locally.

https://techcrunch.com/2022/03/31/why-wordle-game-284-had-tw...

Reply on news.ycombinator.com