Communicators

Jan 06, 2011 23:40

Physical:
The device is about five by seven inches (12.8 by 17.8 centimeters) and half an inch thick. The screen does not completely fill the surface of the device - on one of the short edges are three buttons, two centimeters (0.8 inches) wide, with respective lengths of 3, 4, and 5 centimeters (1.2, 1.6, and 2.0 inches). Their labels say proudly "text", "audio", and "video". It looks like they're printed on, but if you get a magnifying glass it will become clear that this is actually a very fine LED display (and the words will change to the native language of whoever is reading). Due to the way the labels are positioned, it would be agreed that the side with buttons is the left.
On the left edge are three dials, labelled "volume", "microphone", and "brightness". (The labels are affixed in the same manner as the buttons above.) The right edge bears a slider with three unlabeled settings; they are in actuality "on", "vibrate", and "off". In "vibrate" mode, the backlight is turned off, and the machine will vibrate briefly instead of playing an audio notification or the audio of the message.
These devices aren't as delicate as their current real-life versions (they will probably survive a drop on the floor), but they are breakable. A new one appears in your drawer the next day, with no posting access for 24 hours.

Functionality:

When the "text" button is pressed, a beautiful physical keyboard unfolds gracefully from the device, justifying its unusual thickness, and you will be taken to a post-submission window. Its keys are similar to the buttons in the nature of their labelling -- they will rearrange to the alphabet of the user. With the exception of this keyboard, the device is controlled entirely by touch (reference the Accessibility section for possible alternate mechanisms). The microphone and mic are both internal to the machine, so you can look at the center of the screen and be looking straight at the camera. Waveforms will display while recording audio, video will display while recording video. Afterwards is a preview-submit-cancel dialogue, during which filters can be added. Text posts can be retroactively edited, but no posts can be deleted entirely (although nothing's stopping you from putting "THIS POST DELETED" in a text post).

Everything on the network is translated into Common, like speech is in the asylum.

Browsing:
All digital windows in Carriero cannot be closed - both the main window and any popup windows. The main differences from the Carriero page visible on Livejournal are as follows:
  1. The usernames and userpics used by players do not appear. Instead, each post is accompanied by the same picture of the patient (one taken during the interview), with the patient's ID underneath. 
  2. When clicking the ID of a patient, instead of being taken to a personal journal, you will see all posts that user has ever made.  
  3. The tags function as "c: ____" tags instead of the "p: ____" tags they now resemble, due to the aforementioned feature. Tagging is automatic.  
  4. All posts that are tagged "carriero itself" do not appear. 
     
  5. Posts involving audio or video, obviously, feature embedded audio or video instead of descriptive text. 
     
  6. Links to information that is public IC, such as the map and the menu, appear in the sidebar.
Differences between posts as visible on Livejournal:
  1. Only one comment shows up for each thread; tapping this will result in the entire conversation being played back in the case of audio or video, or going to a separate thread page in the case of text.  
  2. There is a prominent "play all" button in the place of the "read all comments" button, which will play all audio and video conversations in chronological order.  
  3. In a long text comment thread between only two people, there will be no thread-indenting or necessity to expand again.  
  4. No comments that are actually action will show up, obviously.

Filters:
Hope you've memorized all your friends' IDs, because you'll have to recite or type each ID you want to see the message. You'll have to submit your posts separately IC if there's portions you don't want others to hear; OOC it's still acceptable to have them all in only one post.

Accessibility:
Linda has taken elaborate precautions to ensure that she can accept any patient to her facility. The buttons are also labelled in rather small braille, which is apparently altered by the device. (The cells will visibly rise up and down to shift from standard braille to tenji, for instance, if the device is passed from an English-language blind user to a Japanese-language one.) The keyboards lack braille by default and will only gain it in the hands of a blind user.
Audio transcripts and described video are available, but they will only appear on the device of a deaf or blind user, respectively. All button-based dialogue can be controlled by voice commands; there are even elaborate voice command trees that will only appear for the blind-mute, which take input by tapping the screen.
Those who are deafblind, with severity in both impairments to the extent that none of these resources are of sufficient help, and are familiar with handsigning will receive one of these. Each one is always this high relative to their master. They respond to voice commands, but will immediately sign to their masters the following information: A tap on the stomach will initiate transcription of whatever information follows - the default fashion of doing this is transcription of handsigns for a text post. Tugging on the device's ears will force it to record actual audio, and likewise pokings its eyes will record actual video. There are the additional options of ending your message with the signs "A-U-D-I-O P-L-E-A-S-E" or "V-I-D-E-O P-L-E-A-S-E", in which case the inmates will be treated to a lovely machine reading of your message. (The inflection is very good, but the pronunciation is quite wooden.)
Whether you receive an "Electronic Interpreter for the Deaf and Blind" or "Robotic Assistant for the Blind and Deaf" depends mostly on which would sound more like your voice. The Electronic Interpreter is voiced by a woman (not Linda) of about 25, with a low, calm voice, and has more roundness in its design, while the Robotic Assistant is voiced by an eager man of about 20 with more of a square aesthetic.  

Bonus:

Somehow you earned these. Click on the "tag" page and scroll all the way down; they have small icons like on this site but with no titles.The options available so far:

  1. A platformer with two levels. In the first you try to avoid drowning as the water level increases; in the second, you now have gills and must avoid the sharks in the facility. All characters who were in the asylum at the time are playable, although TSH-330-001 is not unlocked unless you beat the game with all other characters. The real benefit of the game is being able to explore the asylum, albeit only the parts that were open at the time.
  2. A turn-based strategy game with suspiciously similar gameplay to Devil Survivor. You're fighting against the orcs.
  3. A brief visual novel. You play as TSH-866-001 or TSH-270-001 and try to convince the others to go looking for the ring. Every ending is different somehow. In one ending you actually convince them but then all of you are killed by orcs. In two you die by angering other subjects. In three you die by the hands of other subjects by accident. In most of the others they come to the conclusion that they can go rescue TSH-800-004 and TSH-802-001 "tomorrow", and get the ring if they find it along the way.
  4. Tetris. No score is kept, but level-ups (not visibly tracked) occur every five minutes.
  5. A short little piece that would be labelled "life and style" on many websites. You go around with the goal of kissing the person who has your number, but they're never on the screen at the same time as you. If you go for 90 seconds without kissing someone, you die. It's pretty boring.


Links to sketches for those interested:

Homepage
Entry
Game icons

Navigation

#technology

Previous post Next post
Up