The Seint's Christmas CTF — Part 6

The theme of flitting around the world continues in the final Part of The Seint’s puzzle:

The last message I got was: "This trip is over. I'm glad you accompanied me virtually. Can you figure out just one more thing? Last year I found a nice café. It was just beside one of the places I visited last year. I will give you a clue of where I was - look at the attached files. The last one is from there. I remember I had to head north for about 10 minutes, under some green bridges and by a sign telling that the maximum height beneath that bridge is 4.2 m. The name of the café was something something café and one more word after that. Will you be able to find it?".

Unfortunately one of the files got corrupted. There must be another way to find out the last, fourth location.

The answer is that unknown "something something" before the word "café" in its name, written in lowercase and turned into a hash.

PS: If you want to go back to the beginning of this walkthrough, you can find it here. Spoilers ahead!

We’re then presented with four files, like this:

Each one, apart from the fourth, corrupted one, looks a bit like this:

This isn’t a format I’m familiar with, so I Googled $GPGGA to get some ideas. This suggested the data was likely GPS data, which meant that 3959.307,N,11623.467,E were some variation of northings and eastings. I wondered whether there was a tool to automate the interpretation of this data and Googled GPGGA decoder, which led me to the RL.SE GPRMC & GPGGA decoder. Plugging in the first line of the first file got me this:

Zooming in eventually got me to a stadium, but the labels stayed in Chinese. Thankfully, the decoder had provided a more familiar format for the co-ordinates, so I could plug them into Google Maps. I then realised that I was outside Beijing National Stadium.

Working my way through the other two files got me to the London Olympic Stadium, and the Maracanã in Rio. Suddenly, I was getting an idea where the fourth location would be…

These are all Olympic stadiums, and a quick check confirmed they are all in order:

This means that the fourth location must be Japan National Stadium in Tokyo, so we should try following the directions from there…

Looking north, there are a number of bridges, and it doesn’t take long to find some green ones, including one with the required height marker:

Continuing north, we find the something something café something The Seint’s friend is talking about, and the first two words give us our hash!

And it would be onto our next step, but opening the Zip for Part 7 gives us this:

We’re done! There is an Easter egg in the unlocked Zip, but I’ll leave that for you to discover.

Takeaway

I’m not sure about the takeaway from today’s puzzle. It was great to need a piece of trivia knowledge to go along with the Google mapping, and I enjoyed learning about the GPS data. It’s a tribute to internet culture that whatever bit of encoding you find, someone will have built a tool to decrypt it online, often simply for the joy of solving a puzzle.

On the wider puzzle, I think I would soon replace a friend like The Seint’s. If I got a text from one of my friends saying Guess where I am? It's somewhere pink, and rhymes with the middle name of the third Duke of Northumberland..., I’d reply Just tell me you idiot, I'm busy. But I’m glad The Seint has friends like this, even made up ones, because it gave me a bit of enjoyment on some dark winter nights.

Thanks Seint, and hopefully you’ll have another set of puzzles for us at Christmas!

Postscript

If you’re not already convinced that AI-powered chatbots are going to change OSINT after seeing my coding example in Part 3, check this out:

🤯

What amazes me about this dialogue with ChatGPT is:

1. I am able to ask my follow-up question in natural language, and ChatGPT just deals with it.
2. ChatGPT has an awareness of context, albeit limited to the current conversation. So when I ask “Can you convert that latitude and longitude…”, it is able to infer that I’m referring to information it provided me in the previous answer.

This is a much more intuitive way of exploring a problem space and coming up with useful answers. There is the temporary bonus that I don’t have to look at a dozen ads every time I ask a question, but GPT is an expensive tool to run, and the end user will need to pick up the tab at some point.

There is an aphorism doing the rounds that goes along these lines:

No, AI isn’t going to replace [insert job here],
but [insert job here] using AI will.

It’s fairly clear [insert job here] is going to apply to a lot of roles, and the investigator will definitely be one of them. Time to start learning how to prompt effectively!