2D data thread

"Scanning" is for the mobile age, and it also have use for archiving with Paperback
Tagging the streets with QR code is basically Zig Forums's passtime (see twitter.com/ahundredhands)

FIle embedding is about hiding data in weird places for times when a website only accepts images, videos, audio and such. See for YouTube

Challenge 1: bash script that hides data in EXIF/*IFF metadata blocks, or even MKV/MP4 metadata blocks
Challenge 2: hash visualization similar to hashblot and polyhash

UGH!

Translation: I'm too stupid to make something, someone else please make it for me!

Just for fun? Or do you want to solve a real problem? Making 2d barcodes isn't difficult. Making something more useful than existing solutions is difficult.

Just a FOSS version of a circular/annular/dart barcode, because it looks good for sharing PGP keys, IPFS/Torrent hashes or hyperlinks.

The Messenger code looks pretty obvious with 4 "eyes" 90 degrees apart for position/alignment, but we don't know where the timing patterns are.

The Kik code is even weirder as there is no position/alignment pattern or Timing pattern.

Snapchat's code (non-annular) looks pretty arbitrary with the logo, the format is static and scanning is easy, adapting it for Zig Forums or some other project would be sweet.

There's a reason why color isn't used for printed barcodes because blacks are cheap.

negligible cost and phone-to-phone data transfer makes it easy to use color barcodes.

First of all, cost of ink amounts to millions of dollars. Second, color printers have quadruple the complexity and failure rate of black and white printers. From which follows the third, color printers are vastly more expensive to operate.

Firstly, if you have the money to print colored posters, the side cost of colored barcodes are negligible.
Second, you have not answer the phone-to-phone or screen-to-camera data transfer question.

Printers are garbage and produce vastly different color results, cameras are garbage and produce vastly different color results, time is garbage and changes printed colors over time, the whole idea of using color as scannable information is a bad idea.

Ah and let's not forget how ambient color (daytime, surrounding materials and colors, etc) change the way things look.

When you only use primary and/or secondary colors it becomes negligible
Ditto
Then you are printing on the wrong type of material in the first place
Color adjustments, people, it is EASY (especially with phone camera lights)

If you do that then the cost of ink is definitely less worth it. It'll probably be a better idea to use shapes rather than colors. For example I imagine you could change squares into horizontal & vertical lines pretty easily.

And that is how you reduce data capacity of a barcode! Well done, faggot!
Seriously, standard line barcodes have been around for decades with shit capacity, then , and came and nothing is the same ever again

Squares = pixels in the QR code or whatever. Rather than just reading whether it's black or white, you also check whether the pixel is tall or wide.

increases unnecessary complexity, and the tradeoff is that you have to make larger prints to prevent print blurs.

Thirty Kikcodes as example!
my.mixtape.moe/ewwpgz.zip
Now, all we need is to find which patterns are used as position/alignment pattern

my.mixtape.moe/iajgen.zip
There, 100 examples sorted by resolution. Surely you can make something outta this

Bumping this for fun

2DPD

also uncalibrated screens lol

That's why you put a calibration region on the image.

I was thinking of making a small 2 hour project where you could encode strings into small images that would be made up of 1x5 "tiles", representing 5 bits/32 unique characters,
00000 would be a space obviously
00001-11011 would be the characters of the alphabet
11110-11111 would be 5 extra special characters, the ones i have in mind are ",", "-", (apostrophe), "?", and "."
I'd have the decoder two commas into a colon, a comma and a period to a semicolon, a comma and a hypen to a to an exclamation mark, and two hypens into parenthesis.

This is fascinating. I've always wanted a means of storing data on giant tapes of some ridiculously strong material that could stand the test of time and be simple enough to implement on SBCs.

No. That is not creative
en.wikipedia.org/wiki/PDF417
en.wikipedia.org/wiki/Code_39
en.wikipedia.org/wiki/Code_93
en.wikipedia.org/wiki/Code_128
en.wikipedia.org/wiki/Telepen

bump

CQR code-5, CQR code-9 and HiQ codes FTW
pdfs.semanticscholar.org/95c7/fe455a084b0f0ce9923e996652156c1f7998.pdf
arxiv.org/pdf/1704.06447.pdf
researchgate.net/publication/313674473_CQR_CODES_COLORED_QUICK-RESPONSE_CODES

home.ie.cuhk.edu.hk/~hlxu/papers/TIP2018.pdf

Baudot code or ITA2.

There are numerous protocols currently in development and they all have unique abilities. One is even challenging wireless protocols like NFC in terms of throughput. The main thing is being able to decode with a 99%+ success rate in ANY lighting. The next few months should be interesting, I mean, serving data off your screen like its some kind of antenna, that's freaking sweet. Good luck fellow developers, let us make this a good run. - JP

jabcode.org/