Lessons learned from making web app Smart Brain

2020-03-29   6 min read

Smart Brain Demo

GitHub repo for the view

GitHub repo for the server and database

(1) What is .desktop file?

After the installation of Postman with Snap, I somehow couldn't open it up although I did see it installed in my /home/Desktop/ (I normally don't like to put anything in /Desktop but the page I referred said so I followed.) I wouldn't have minded starting it up from command line so much but also it would've been nice to have an icon on app launcher.

That's how I got to learn that there is such a thing called .desktop file, which is basically what lets your computer show you the icon of an application. I didn't see the Postman icon because I didn't have .desktop file pointed at Postman. Yep, makes sense.

Postman official documentation says

To start the app from a launcher icon, create a desktop file, naming it Postman.desktop and saving it in the following location: ~/.local/share/applications/Postman.desktop

So I added this cat Postman.desktop Exec is where the program needs to look for Postman.

.desktop files

and it worked! Postman icon did show up, and when I clicked it, the app did turn up. I was pleased for 2 seconds until another question popped up.

app-launcher

If .desktop file is necessary for my machine to show me icons, why am I not seeing files like atom.Desktop, Bitwarden.desktop, htop.desktop in /.local/share/applications ? Where are they...?

I found my precious .desktop babies in /usr/share/applications after some searching (I avoid saying "googling" because I don't want the equation search = google to be norm. In fact, my default browser & search engine combo is firefox & DuckDuckGo), usr/share/applications

This means that .desktop files can be put in /usr/share/applications as well! What's the difference? That's another valuable lesson I learned. Globally installed (hence globally accessible) applications go to /usr/share/applications and locally installed (hence only locally accessible) ones go to /.local/share/applications.

(2) What is $PATH & .bash files

Basically $PATH is a list of directories where your OS look for the program to run when you type commands. I got to learn this because when I installed snap it told me that it wasn't set in the $PATH and I indeed couldn't run snap command. Instead of setting $PATH variables manually, you can set it in .bashrc file, which gets run every time you open a new terminal, so that it automatically sets it for you. Some Ask!Ubuntu pages say I could set it in .bash_profile as well but it only works if I set it in .bashrc. The part I have added looks like this bashrc

(3) PostgreSQL

Relational database and Non-relational database

PostgreSQL is relational database; That is, schema (relationship between each tables) needs to be defined first before you stsrt adding data to them. This also means that each columns need to be set first, whereas non-relational database like MongoDB can accept a new column as you add data.

User Interface - psql and pgAdmin4

PostgreSQL comes with CLI based controller called psql with which you can interact with databases, but also there is GUI called pgAdmin available.(latest stable version is pgAdmin4) What they can do for you is the same except for the richer interface with pgAdmin, obviously.

Separation of concern

It is a better practice to have multiple tables that contain less columns than to have less table that contain a massive mount of columns. You can JOIN to relate each tables using a primary key and a foreign key.

SQL - Structured Query Language

Once you logged into the database, you MUST type ';' at the end of command otherwise it won't work since this is SQL, not shell script nor JavaScript (we are not good at that, are we?)

SQL commands I learned

Create Database

createdb DATABASE_NAME

Log into the database

psql 'DATABASE_NAME'

Create Table

CREATE TABLE table_name (column1 datatype, column2 datatype, column3 datatype);

Constraints: primary key, NOT NULL, UNIQUE

Each table can have only one primary key, which is a combination of NOT NULL constraint and UNIQUE constraint.

CREATE TABLE users (ID serial NOT NULL PRIMARY KEY,
name text UNIQUE NOT NULL,
age smallint,
birthday date NOT NULL);

Since Postgre is relational database, you need to define data type for each columns. In addition to Pre-defined datatypes available PostgreSQL, you can also set your own datatype using CREATE TYPE command.

List all tables

\dt

Insert a row

INSERT INTO table_name (column1, column2, column3th) VALUES (value1, value2, value3);
INSERT INTO users (name, age, birthday) VALUES ('Shimba', 99, '1921-01-01');

Show data in a table

SELECT datatype, datatype, datatype FROM table_name;

For example

SELECT name, age, birthday FROM users;

if you want all columns to show anyway, you also can use * (= all)

SELECT * FROM table_name;

Add a column to a pre-existing table

ALTER TABLE table_name ADD target_column datatype;
ALTER TABLE users ADD location text;

Update data

UPDATE table_name SET target_column=value WHERE target_column=value;
UPDATE users SET location='Toronto' WHERE name='Shimba';

to update data for multiple users

UPDATE users SET location='Toronto' WHERE name='Shimba' OR name='Elliot';

Delete row

DELETE table_name WHERE target_column=value_to_be_deleted;

DROP table

DROP TABLE table_name;

Conditional Selection

SELECT * FROM table_name WHERE target_column LIKE condition;
SELECT * FROM users WHERE name LIKE 'E%';

Sort a table

Descending order

SELECT * FROM table_name ORDER BY target_column DESC;

Ascending order

SELECT * FROM users ORDER BY age ASC;

Basic functions

SELECT AVG(target_column) FROM table_name;
SELECT SUM(target_column) FROM table_name;
SELECT COUNT(target_column) FROM table_name;

How to JOIN 2 tables

SELECT * FROM table1 JOIN table2 ON table1.target_column = table2.target_column

Typos/comments/suggestions are welcomed on GitHub issues

ELLIOT SHIMBA  ©  2020

Made with Bulma