here’s some of my progress, please view my final project for the final result:

Android App development proved harder than I thought, and I didn’t want to phone it in, so this is very unfinished. Still, since I really like my idea and plan to continue working on it, I’ll share what I have so far.

My project is a messaging app called Rebus Chat that automatically converts typed messages to Rebus puzzles like the one below, and sends them that way.

Even the usernames and the buttons like “back” will be images rather than text. All the images will come from The Noun Project. A mock-up of the design I was thinking is below (can you guess all the messages?)

I really like this idea, because it reminds me of the puzzlehunts I love to participate in, and requires a little tricky but fun thinking to decipher each message. To convert text to Rebus, here is the pipeline I had in mind (mostly unimplemented):

• Strip out punctuation/capitalization/etc.
• Try to correct spelling errors (I am looking for an existing tool to help me do this; if I can’t find one then any “words” that don’t parse as words jump to the last bullet)
• Split everything into syllables (this will be done using Rita). We only care about pronunciation here, so “Where are you” could just as well be “Wear arr ewe”.
• For each syllable, if one interpretation of it is as a (non-abstract) noun, look it up in The Noun Project and use that picture (if multiple noun interpretations, use the most common one–there are tools for ranking words based on popularity)
• If a syllable still doesn’t have a picture, try combining it with neighboring syllables to see if it makes a two-syllable noun (if so, use that image).
• If that doesn’t work, try near rhymes.
• If there still isn’t anything, then I’m not sure what to do. Some words/phrases just aren’t 100% Rebus-able (eg. “Relevant Job Experience”–what should that even look like?). I have thought of a few options:
  • • Option 1: Use “word math,” like R + (elephant) for “relevant” or (smile) – S for “mile.” This seems pretty hard to do programmatically, at least robustly, and there will still be words that don’t work. Like “the.”
    • Option 2: Just put those parts in as text, like “(eye) (half) the relevant job experience”. It will be as Rebus-ified as possible, and still a bit of a puzzle to decipher, but not purely images, which is too bad since I like the all-images look.
    • Option 3: Just remove those parts, keeping only what can be Rebus-ified. This might turn “I have the relevant job experience” to “(eye) (half)” and then… nothing. That’s no good, because it loses important content. However, maybe in the case of just small words (a/the/and) it’s okay. This could perhaps be fused with Option 2, then?
    • Option 4: Prompt the user before the message gets sent, marking that word as un-rebus-able and encouraging them to try something else. This is a little clunky and less smooth from a UI perspective, but might result in the best Rebuses.

I am leaning towards Option 2, but would be interested in hearing your opinions on this. I really do want to make this a reality, because I think it could be super fun and it really is time I learn Android App development.

My telematic project is a website where you can publish a text work and watch it get vandalized by visitors in real time. Once a work is published, each visitor is given a random word in the work that they can replace by submitting a text form. Every time a word is replaced, this is communicated to each client using sockets. The visitor also gets a new word to replace. This project is implemented in glitch.com using JQuery, Sequelize, and Socket.io.

This project’s concept has evolved far from its starting point. Originally, I was looking into text RPGs and sought to create a cairn where each visitor can add a new “page” to the text RPG. This text format naturally led me to look more into the implications of language rather than taking on a game-like structure. At that point, I fully intended to write some form of poem remixer in which visitors could do things like change the rhyme structure of a written poem or transform a word to a synonym, antonym, or other related word. This still felt lacking to me, because what was the point of this transformation? I realized that I could create a more interesting work (at least to me) by focusing on that transformative aspect. I have a long history with transformative works and the futile notion of authorship and content ownership on the internet. That, and it would be kind of funny to watch a troll vandalize your work before your very eyes.

So I opened up the format to freely change any single word in the published piece. A few limitations apply. Instead of choosing which words to change, the programming assigns you a random word. You can only replace words with single words–no spaces. In order to go to a different word, you must submit your change (or lack of change) for the current word. Additionally, the text supports HTML, which provides another avenue of mayhem in exposing the ugly syntax by getting rid of an end-bracket. Where I found the most interesting interactions to arise was in the networking. Once the work is published, the author is simply another visitor to the site. The author can see their work be vandalized in real time. Just as the vandalizers can bastardize words in the original text, the author can defend by restoring their original words or even communicate with the vandalizers by sacrificing a word with an angry, caps-locked “STOP”.

1. I made an Alexa skill with Abby Lannan, a Euphonium performance major Graduate student at CMU, where it will help you to practice music in various ways. It allows you to play music with other people’s music recordings or upload your own music. It helps you with music ear training, intonation and rhythm accuracy. This project is pitched to Amazon’s Alexa Day as a future Alexa project.

It tracks users’ practice progress and helps them achieve their practice goals. It synthesizes sounds on a cloud server based on voice commands.

Basic Functions:

Skill Development: It has a metronome and drone that offer a variety of sounds. It also allows users to upload recordings to play music with each other.

Tutoring: It has ear training game that teaches music theory.

Record Keeping: It saves user info on cloud database and allows them to track longterm practice progress.

A detailed slides is attached here:

2. Also, I attempted to make an Alexa skill that sends you an email of a pdf of pictures that ‘will make you happy.’ First of all, you tells Alexa what makes you feel happy/grateful. Then, she sends you a pdf to remind you the happy things in the world. The results look pretty like trash.

[hands, anime, butt, beach]

[spagetti, good weather, chris johanson, music]