urbanriskmap / cognicity-grasp Goto Github PK

Requirements for prototype:

• Get user location via web geolocation (ref: https://developer.mozilla.org/en-US/docs/Web/API/Geolocation/Using_geolocation)
• User input: water level
• User input: free text
• Submit button

Suggested fields:

• card_id
• created_at (ISO 8601 timestamp with timezone from user device)
• text (UTF8 @ 140 chars)
• water_depth
• image (optional)
• location (GeoJson (point))

urbanriskmap/riskmap#132 (comment)

• Trigger request for card one-time-link via Twitter
• CogniCity server to accept card input data
• Create database table for card data?
• Push out via report API

Hi @talltom , after our discussion with BMKG (Meteorological, Climatological and Geophysical Agency), they are agree to feed our map with the realtime rainfall data in mm/h. Can we show it on the map like the waterlevel or sensor data in a realtime graph per hour?

I have made the wiki page and added this on our list. Please check on this link .

BMKG also will provides us the location data of the rainfall station for Jabodetabek, Bandung, Surabaya. Can we integrate their .csv/.kml data (with a column for coordinates)?

cc @anexact

Setting global values in the 'before' function, may return false positives where these are not updated in the test due to some other errors.

Need a more robust method of testing changes to logs.

Do we really need image_id or can we make card_id part of primary key in grasp_report_images?

Right now we are constrained to either testing on a phone or on Desktop. The touch and click apis are different. We should add code to make this more streamlined.

As we are collecting data directly from users we need to add validation processes for report cards:

1. Client side, how to inform user if input is incorrect (e.g. report length too long?)
2. Server side, how to catch malformed input data from report cards (e.g. null geolocation etc)

Putting this here so I remember to fix it later.

I think we're not doing error checking correctly in ReportCard.js.

/Users/abrahamq/Documents/urop/cognicity-grasp/ReportCard.js:157
else if (result[0].received === false){
^

TypeError: Cannot read property 'received' of undefined
at /Users/abrahamq/Documents/urop/cognicity-grasp/ReportCard.js:157:29
at null. (/Users/abrahamq/Documents/urop/cognicity-grasp/ReportCard.js:85:6)
at Query.handleReadyForQuery (/Users/abrahamq/Documents/urop/cognicity-grasp/node_modules/pg/lib/query.js:114:10)
at null. (/Users/abrahamq/Documents/urop/cognicity-grasp/node_modules/pg/lib/client.js:172:19)
at emitOne (events.js:82:20)
at emit (events.js:169:7)
at Socket. (/Users/abrahamq/Documents/urop/cognicity-grasp/node_modules/pg/lib/connection.js:121:12)
at emitOne (events.js:77:13)
at Socket.emit (events.js:169:7)
at readableAddChunk (_stream_readable.js:146:16)

Right now, the client asks the server for a signed url, the server (which has credentials)
asks s3 for a signed url where the client can PUT an image. There's a corner case where the
server asks for a signed url, but the client doesn't actually PUT the image (connection drops, user backs out, so on.)

Here's a fix: use the lambda code to touch our database after it does the compression so that it tells us that an image url is present.

BNPB released regulation (no 2 year 2012) to standardise Indonesian hazards parameter. Please look into this link to see the original document.

*start from page 15 - table 1 on the document

Finding:

Basically, the alert levels for all hazards described as:

1. low (green)
2. medium (yellow)
3. tinggi (red)

The thresholds for each level in each hazards are referring to the national standards that are released by related agency (i.e extreme weather by BMKG, etc).
We will follow up to BNPB next week if there is any update on this regulation while we have a discussion with PDC as well.

cc @talltom @anexact @prittandrani @nashinmahtani @clgeros

Right now photos are uploaded immediately when you click the upload button, but we should instead hold off until the user signs the terms and conditions.

We should write tests around the image upload functionality.

Hi! We were discussing about the bot language, which is may be the best for the users if we provide Indonesian & English.

We were talking deeper about how the bot will determine the language to be used for the rest of the report, and come up to this idea that the bot will send the first message to request the user to choose the language, like in the screenshot here:

Any comments? :) @talltom @fadmastuti @anexact @nashinmahtani @clgeros

cognicity-grasp should follow the template of other cognicity-reports submodules:

• independent configuration via local config.js
• shared database parameters
• shared log file
• index.js exposing module API