LINKS AND COOL HACKS

links-and-cool-hacks.txt

RTL relay testings
https://github.com/facebook/relay/blob/master/packages/relay-test-utils/__tests__/RelayMockEnvironmentWithComponents-test.js
https://github.com/entria/entria-fullstack/pull/109/files

Very easy native splash screen on Xcode
https://medium.com/@kelleyannerose/react-native-ios-splash-screen-in-xcode-bd53b84430ec

Init rn app on specific rn version
(rn cli 2.0.1+)
react-native-cli init --version="[email protected]" my project

// correct CRA usage
yarn create react-app my-app

// mongo regex example
db.Event.find({ title: {$regex: new RegExp("tom", "i")} })[0]

// rn network debug
https://github.com/jhen0409/react-native-debugger/issues/382
```
// add to index.js
global.XMLHttpRequest = global.originalXMLHttpRequest || global.XMLHttpRequest;
global.FormData = global.originalFormData || global.FormData;

if (window.FETCH_SUPPORT) {
  window.FETCH_SUPPORT.blob = false;
} else {
  global.Blob = global.originalBlob || global.Blob;
  global.FileReader = global.originalFileReader || global.FileReader;
}
```

Renato Bohler's tmux config
```
set-option -g default-shell /bin/zsh
set -g utf8
set-window-option -g utf8 on
set -g status off
set -g default-terminal "screen-256color"
set -g prefix C-a
unbind C-b
set -sg escape-time 1
set-option -g base-index 1
setw -g pane-base-index 1
bind r source-file ~/.tmux.conf \; display "Reloaded!"
bind | split-window -h
bind - split-window -v
```

// calculate ssh key fingerprint
- cat public key, remove leading strings, copy base64payload
- echo -n $base64payload | base64 -D | md5

// replaceAtIndex helper
export const replaceAtIndex = <Item = any>(array: Item[], index: number, item: Item): Item[] => {
  return [...array.slice(0, index), item, ...array.slice(index + 1)];
};

export const ONE_SECOND_IN_MILLISECONDS = 1000;
export const ONE_MINUTE_IN_MILLISECONDS = 60 * ONE_SECOND_IN_MILLISECONDS;
export const ONE_HOUR_IN_MILLISECONDS = 60 * ONE_MINUTE_IN_MILLISECONDS;
export const ONE_DAY_IN_MILLISECONDS = 24 * ONE_HOUR_IN_MILLISECONDS;
export const SEVEN_DAYS_IN_MILLISECONDS = 7 * ONE_DAY_IN_MILLISECONDS;

json to typescript type conversion
https://transform.tools/json-to-typescript

awesome phone regex
https://regex101.com/r/MNWXbW/3

convert salaries between year/hour/month
new Array(130).fill(0).map((x, i) => (i+4) * 5000).map(year => ({ year, hour: (year/1920.1755589082431).toFixed(0), month: (year/12).toFixed(0) }))

cool thoughts articles tweets ideas
https://stopa.io

github gif how to
![](name-of-giphy.gif)

kill branches that have no remotes
```bash
git branch -vv | grep ': gone]'|  grep -v "\*" | awk '{ print $1; }' | xargs -r git branch -D
```

twitter bot translation google translate npm library
https://github.com/vitalets/google-translate-api

how to grep git source code
```
git grep <regexp> $(git rev-list --all)
```

how to grep git commit msg
```
<your fav git log alias> --grep="foo"
```

git push a few commits to origin
```
git push origin 91143c3:fix-52-2
```
where 91143c3 is the last commit you want to push, and fix-52-2 is the branch on origin to push to.
this actually pushes all child commits of 91143c3 to origin because commits are chained together.
so, if you'd like to push a range, push the head commit with this technique.

when doing complex stuff, find the simples code that works and build from there

Mac hack: make macos always open an app in full screen.
System Preferences > General > "close windows when quitting an app" turn this off
Open app, make it full-screen.
Quit while in full screen to "remember" full screen for that app

yt-dlp youtube download

 yt-dlp --list-formats <video id>
 yt-dlp --format 251 <video id> // medium audio only webm 48khz
 ffmpeg -i in.webm -codec:a libmp3lame -q:a 0 out.mp3 // webm to VBR mp3 max quality

tmpfs firefox

follow this page on how to setup, script is good to use.
about the cache, here reallocate cache to tmpfs under the profile, so the sync script will also sync your cache

# about:settings add new string key
browser.cache.disk.parent_directory = /dev/shm/firefox-mpakm5ej.dev-edition-default-vacation/

when automating script, cron is not required as systemd has timers.
create the templated systemd user service (note the @), enable and start it once.
create a [email protected] in ~/.config/systemd/user/

   1   │ [Unit]
   2   │ Description=Firefox sync timer
   3   │
   4   │ [Timer]
   5   │ OnUnitActiveSec=5min
   6   │ AccuracySec=1us
   7   │ Persistent=false
   8   │
   9   │ [Install]
 10   │ WantedBy=timers.target

enable and start the timer [email protected]

firefox optimizations

https://bbs.archlinux.org/viewtopic.php?id=45518
https://bbs.archlinux.org/viewtopic.php?id=41624
https://en.wikipedia.org/wiki/Swiftweasel
https://wiki.archlinux.org/title/Firefox/Tweaks
-O3, -funroll-loops, and -march=native
accessibility.force_disabled
https://www.howtogeek.com/557929/how-to-see-and-disable-the-telemetry-data-firefox-collects-about-you/
https://www.makeuseof.com/tag/speed-up-firefox-immediately-with-these-6-simple-tweaks/
permissions.default.image | 3 // 2 disable all images
use image disabler extension
network prefetch settings
privacy.resistFingerprinting
https://firefox-source-docs.mozilla.org/contributing/contribution_quickref.html

pipewire audio games

pw-link -i #list inputs
pw-link -o #list outputs
pw-link -l #list links
pwlink input output # create a new sound mapping, usually from game channel to your speaker

waydroid android emulator (no arm64 apps)

it needs wayland, use a program to create a wayland session inside x11. like cage.
you'd normally run cage -- <normal waydroid command>
it needs binder kernel support, use binder_linux-dkms
install waydroid and enable the waydroid-container.service.
run waydroid init or waydroid init -s GAPPS
google apps are useless since it won't be able to use those services anyway
waydroid session stop to stop using waydroid
waydroid show-full-ui to show window

stop waydroid-container.service
root rm -fr /var/lib/waydroid ~/.local/share/applications/*aydroid*
waydroid init -f
start waydroid-container.service

to hard reset
if internet doesn't work, try using a firewall (confusing I know) like ufw and use recommended setting for waydroid
clipboard needs some setup, see arch wiki
use https://apkpure.com to find apks
resolution, resize window to see effect, restart session
waydroid prop set persist.waydroid.width 2712
waydroid prop set persist.waydroid.height 1930

Redroid

docker pull redroid/redroid:13.0.0_64only-latest
https://hub.docker.com/r/redroid/redroid
https://aur.archlinux.org/packages/anbox-modules-dkms

mongo shell tricks

db.collection.find() returns a cursor.

There are a bunch of methods for the cursor object which can be used to get the cursor information, iterate, get and work with the individual documents from the query result (in the Mongo Shell). For example,

let result = db.widget.find( { id : "4" }, { quality_level: 1, _id: 0 } );

The methods print, printjson and tojson are useful for printing the query results assigned to a cursor and iterated over it, in the Mongo Shell.

In this particular example, each of the following statements will print the expected output:

    result.forEach( doc =>  print( doc.quality_level ) );
    result.forEach( doc =>  print( tojson(doc)) );
    result.forEach( printjson );

Note the query must be run each time the cursor is to be iterated, as the cursor gets closed after iterating thru it. But, one can use the cursor.toArray() method to get a JavaScript array, and iterate the array and work with the query result documents without running the query again.

NOTES:

    The db.collection.findOne() method returns a document (not a cursor).

https://stackoverflow.com/questions/25527617/print-document-value-in-mongodb-shell

QEMU commands

#  download qcow2 disks from arch-boxes github mirror
qemu-system-x86_64 -m 2048 -accel tcg -usb -device usb-tablet -drive file=disk.qcow2,if=virtio
# snapshot a qcow disk, do NOT use base image after snapshot (hidden file)
qemu-img create -f qcow2 -b .disk.qcow2 -F qcow2 disk-$(date +%Y-%m-%d).qcow2 # add %H-%M to get HH:MM

New Arch commands

sudo pacman -Sy
sudo pacman -S gpm man nano vi
sudo systemctl enable gpm && sudo systemctl start gpm
sudo useradd --create-home --groups wheel --shell /bin/bash vacation
sudo passwd vacation

# add file to /etc/sudoers.d/foo with lines
youruser ALL=(ALL:ALL) ALL # can use sudo with own password
Defaults timestamp_timeout=60 # minutes between passw prompts

SSH

ssh-key with passphrase, with ssh-agent, passing passphrase to ssh-add from script

While it might seem like a straightforward idea to pass the passphrase to ssh-add from a script, e.g. echo "passphrase\n" | ssh-add, this is not as straighforward as it seems as ssh-add does not read the passphrase from stdin, but opens /dev/tty directly for reading.

This can be worked around with expect, a tool for automating interactive applications. Below is an example of a script which adds a ssh-key using a passphrase stored in the script:

#!/usr/bin/expect -f
spawn ssh-add /home/user/.ssh/id_rsa
expect "Enter passphrase for /home/user/.ssh/id_rsa:"
send "passphrase\n";
expect "Identity added: /home/user/.ssh/id_rsa (/home/user/.ssh/id_rsa)"
interact

Note that as the passphrase is stored in plaintext in the script, from a security perspective, this is hardly better than having a passwordless ssh-key. If this approach is to be used, it is important to make sure that the expect script containing the passphrase has proper permissions set to it, making it readable, writable and runnable only by the key owner.

https://unix.stackexchange.com/questions/90853/how-can-i-run-ssh-add-automatically-without-a-password-prompt

socks proxy using ssh and redirecting traffic through it

open a socks proxy on localhost using something like

# google cloud
# C compression, N no command on remote ssh server, q quiet
gcloud alpha cloud-shell ssh --project=your-project --authorize-session -- -D your-port-example-31337 -CNq

from stackoverflow, how to start the proxy and how to connect to it using proxychains

To do what you are wanting, I recommend sshuttle.

You use it like this:

./sshuttle -r username@sshserver 0.0.0.0/0 -vv

It will tunnel all your TCP traffic automatically for you. You can add the --dns argument to have it tunnel your DNS traffic as well. The remote server only needs to have Python installed.

If you only want to tunnel specific programs I would recommend proxychains.

Once it is installed, start your ssh socks proxy like this:

ssh -fNTD 127.0.0.1:local port username@sshserver

This will start a "SOCKS" proxy listening on local port.

Then edit /etc/proxychains.conf to point to the same port as local port:

socks5 127.0.0.1 localport

Finally start your program that you want proxy-ed like so:

proxychains program name

It should just work. However, a few programs will have trouble working with Proxy Chains. Also keep in mind, that with Firefox, you have to change additional items under about:config to force it to do DNS lookups through the proxy instead of bypassing it.

As an additional note, on web browsers. If they support socks proxies, you don't need to do anything additional to get them to use the above mentioned, ssh tunnel, just enter 127.0.0.1 for the SOCKS proxy server and the for the proxy port.

EDIT 3/29/16

Since this post is still seeing some upvotes, I thought I'd update it. Proxychains is still in most Linux repos and still works on Linux. However, the project is effectively abandoned and does not work on OSX. For either Linux or OSX, I highly recommend upgrading to a still-maintained fork: proxychains-ng: https://github.com/rofl0r/proxychains-ng

Besides working in both Linux and OSX, it is easy to compile, and also has much better support for DNS tunneling.

I should also mention another option, which is redsocks. It works similarly to proxychains(-ng) and is also likely in your dist repo: https://github.com/darkk/redsocks

EDIT 11/27/19 If you go the proxychains route, please use proxychains-ng. There are some serious bug fixes over the legacy version, like: rofl0r/proxychains-ng#292

https://superuser.com/questions/62303/how-can-i-tunnel-all-of-my-network-traffic-through-ssh

How to allow a user to ssh in

edit /etc/ssh/sshd_config and add a line

AllowUsers          foo bar

use a tab before user list and spaces between users, then

systemctl restart sshd

How to discover local computers on the network

make sure both machines are on the same network, wifi or cable

ip neighbor

To SSH into a neighbor, make sure your username is added to that machine a a allowedUser to ssh, then do
assuming your user is the same

ssh 192.168.0.213

ip is from previous command

ssh copy via rsync

rsync -re "ssh" $targets user@machineorip:/path/to/destination

WIFI troubleshooting

List list SSIDS with information

nmcli dev wifi

Connected to network
type password on prompt

nmcli dev wifi con <SSID>

Fix connection to old routers

• sudo systemctl stop dhcpcd
• sudo systemctl stop NetworkManager
• sudo nano /etc/dhcpcd.conf
• add lines noipv6 and noipv6rs to disable ipv6
• sudo systemctl start dhcpcd
• sudo systemctl start NetworkManager

Tricks.txt

nmcli --ask con up ip foo
wpa_supplicant -B -i wlp4s0 -c <(wpa_passphrase china-in-b0x-adm keep....)
groups #to list groups youre in
wpa_supplicant -B -i $IN -c /etc/wpa_supplicant/config
to use wpa_supplicant stop NetworkManager
iw dev $IN link
sys start network-online.target

linux input hardware key remmaping

How to remap keys from somethings such as a keybord or gaming mouse

scancodes to keycodes

This step maps scancodes (hex that hardware sends) to keycodes (system defined key names)

use sudo evemu-describe from evemu to get a dump of your current keyboard.
you might need to guess what your keyboard is called here, it's input number event_ (event15, event4)
also try ls /dev/input/by-path the files here can be used as arguments to sudo udevadm info <file> this will dump useful
info about the input device.

The /dev/input/event_ number is the path of your hardware in the system.
see this
run sudo evtest on a new tab, choose the input device from the list. Now when you click a key, it will dump scancodes (called value in the output) and keycodes as strings. Scancodes are hexadecimal values.

• make file /etc/udev/hwdb.d/70-custom-keyboard.hwdb. In this example, 700e3 scancode is left windows key, it should send a now.

evdev:input*
 KEYBOARD_KEY_700e3=a
 KEYBOARD_KEY_<scancode>=<keycode>

• evdev:input* line is a hardware selection, note the *. See link for more info. Try lsusb to find bus ID. udevadm also helps here.
• KEYBOARD_KEY_XXXX is a scancode that comes from evtest
• a is a keycode. It may be found at keymap.xkb file created by xkbcomp $DISPLAY keymap.xkb. Or header file /usr/include/linux/input-event-codes.h, like this KEY_A -> a
• once you have the file in place and edited with your changes, do sudo systemd-hwdb update && sudo udevadm trigger to load changes to kernel. Wait around 5s.
• do sudo udevadm info /dev/input/by-path/<your keyboard> expect to see KEYBOARD_KEY_700e3=a (your changes) in the output. Using the keyboard should also reflect the keys remaped at this point. If you don't see changes in output try a reboot, sometimes it works with no reboot, sometimes it doesn't.

keycodes to symbols

You might not need this step, if it works already, good.
This step maps keycodes to symbols, locale and user oriented input characters.

• make file keymap.xkb by xkbcomp $DISPLAY keymap.xkb if you have not already. Open it in a text editor.
• search for a keycode, expect a line like this

<LWIN> = 121;

• search for <LWIN> now. Expect something like this

key <LWIN> {         [         Meta_L ] };

• this means keycode 121 maps to LWIN (left windows) which maps to left meta key.
• search around the file and make changes.
• to test it do xkbcomp output.xkb $DISPLAY

testing

• use xev and put mouse on top of the little window to see what keycodes and symbols are being sent.

Ports and processes

netstat -tulpn # dump all processes using ports
lsof -i :3100 what on port 3100 on localhost?

sudoers

sudo -i
cd /etc
echo "Defaults timestamp_timeout=45" > sudoers.d/timestamp
echo "Defaults editor=/usr/bin/nano" > sudoers.d/editor

fstab tips & tricks

• user sets nodev, nosuid, noexec, unless you specify dev or suid or exec after user
• defaults already includes auto, rw and nouser
• auto and user in this order makes little sense, auto causes root to mount, blocking users from unmounting.

list open files

files preventing a partition from unmounting
lsof /media/diskX

linux gpg keys handling

see https://stackoverflow.com/questions/5587513/how-to-export-private-secret-asc-key-to-decrypt-gpg-files

when installing packages, if running into gpg errors to import keys, check this: https://sks-keyservers.net/status look for a server that is green on HKPS column. then import key using that server gpg --keyserver sks.pod02.fleetstreetops.com --recv-keys <pub key>

pubkey encrypt
gpg --sign --encrypt --armor file
will output file.asc and ask for recipient email and keypassword
provide a gpg key you own

pubkey decrypt
gpg -d file.asc > out
provided you have the priv key of the reciepient used to sign the data!

set key trust level
gpg --edit-key $EMAIL
select key in prompt by typing a number (usually 1)
type trust to set trust level
quit

tar

tar -cvf out.tar in.file
tar -xvf out.tar

linux dns

use resolvconf -l and resolvconf -a $INERFACE < file
where file is like /etc/resolv.conf:

# resolv.conf from wlp3s0.ra
# Generated by dhcpcd from wlp3s0.ra
nameserver 8.8.8.8
nameserver 8.8.4.4
nameserver 2804:14d:1:0:181:213:132:2
nameserver 2804:14d:1:0:181:213:132:3
option timeout:1

if ip link shows interfaces are up, but still ping fails, try dhcpcd <interface>

resolv.conf confliict

NetworkManager wants to start dhcpcd as a child process as part of the connection process. If dhcpcd is running as a systemd service it will conflict, the child loses and the connection fails.

Disabling NetworkManager dns config solves the conflict, no more child and dhcpcd service handles DNS, but causes the tray icon to bug out and Spotify doesn't work

Disabling dhcpcd service allows connection, but NetworkManager writes a bugged, garbage DNS server to resov.conf that causes several issues. Additionally, dhcpcd is required for CLI environment connection.

Set NetworkManager to use resolvconf: rc-manager=resolvconf, and set dhcp=dhclient to fix dhcpcd conflict. Enable dhcpcd to allow cli environments to connect normally to networks.

CLi dns and brazil registro.br

dig a.dns.br // get dns server ip
dig @200.219.148.10 golang.dev.br // use server IP to query a domain in it 

calls to dig will return more information each time, keep adjusting the invocation with output

linux networking

if using a GUI, network manager should take care of everything. Any issues can probably be resolved with its config file.
on console

• stop NetworkManager.service, to avoid conflicts over interface. it relies on kwallet which relies on x11.
• use wpa_supplicant to connect
• use iw dev $INTERFACE link to check link status
• ip link ip link set $INTERFACE up to manage interfaces
• check that dhcpcd.service is running on all interfaces, or run dhcpcd $INTERFACE

Find your internal ip address

hostname -i

linux broadcom wifi

kernel module is wl
check dmesg | grep wl to find if it's loaded or not
also check lspci -k look for Network controller
good:

 Kernel driver in use: wl
 Kernel modules: bcma, wl

bad (no drive in use):

 Kernel modules: bcma, wl

fix 1: unload wl and load it again

sudo rmmod b43
sudo rmmod ssb
sudo rmmod bcma
sudo rmmod wl
sudo modprobe wl

fix 2: edit mkinitcpio blacklist to avoid problems causing wl to not load
see arch wiki broadcom wifi

linux importing pfx windows key files

IN=path/to/pfx
openssl pkcs12 -in "$IN" -legacy -nocerts -nodes -out out-filename.key
openssl pkcs12 -in "$IN" -legacy -clcerts -nokeys -out out-filename.cer

files will be in $PWD
store in /etc/ssl
cert and pem might be the same thing

systemd

/etc/systemd/system has symlinks to enabled services, unlink to disable service

journal

When checking journalctl, use journalctl -r -u <unit> as root to see all output for <unit>
-r means reverse (new on top)
running as regular user hides output

rsync

https://lzone.de/cheat-sheet/rsync

community

German Guaigua linux specialist on react br

Amdgpu

low level gpu overclock/underclock

echo "manual" > /sys/class/drm/card0/device/power_dpm_force_performance_level

echo "s 0 300 750" > /sys/class/drm/card0/device/pp_od_clk_voltage
echo "s 1 588 750" > /sys/class/drm/card0/device/pp_od_clk_voltage
echo "s 2 952 800" > /sys/class/drm/card0/device/pp_od_clk_voltage
echo "s 3 1041 800" > /sys/class/drm/card0/device/pp_od_clk_voltage
echo "s 4 1106 800" > /sys/class/drm/card0/device/pp_od_clk_voltage
echo "s 5 1168 800" > /sys/class/drm/card0/device/pp_od_clk_voltage
echo "s 6 1209 800" > /sys/class/drm/card0/device/pp_od_clk_voltage
echo "s 7 1270 800" > /sys/class/drm/card0/device/pp_od_clk_voltage
echo "c" > /sys/class/drm/card0/device/pp_od_clk_voltage

echo "m 0 300 750" > /sys/class/drm/card0/device/pp_od_clk_voltage
echo "m 1 1000 800" > /sys/class/drm/card0/device/pp_od_clk_voltage
echo "m 2 1000 800" > /sys/class/drm/card0/device/pp_od_clk_voltage
echo "c" > /sys/class/drm/card0/device/pp_od_clk_voltage

echo "r" > /sys/class/drm/card0/device/pp_od_clk_voltage

OD_SCLK:
0: 300MHz 750mV
1: 588MHz 765mV
2: 952MHz 912mV
3: 1041MHz 981mV
4: 1106MHz 1037mV
5: 1168MHz 1106mV
6: 1209MHz 1150mV
7: 1270MHz 1150mV
OD_MCLK:
0: 300MHz 750mV
1: 1000MHz 800mV
2: 1750MHz 900mV
OD_RANGE:
SCLK: 300MHz 2000MHz
MCLK: 300MHz 2250MHz
VDDC: 750mV 1150mV

pacman

404 on package fix

• google package find dl link with higher version, copy link
• pacman -U

gpg key error

pacman -S archlinux-keyring will refresh your keys

groups

add a user to a groups

• usermod -a -G adm vacation

hibernation

https://wiki.archlinux.org/title/Power_management#Sleep_hooks
https://bbs.archlinux.org/viewtopic.php?id=247036
https://www.kernel.org/doc/Documentation/power/swsusp-and-swap-files.txt
https://btrfs.readthedocs.io/en/latest/Swapfile.html#hibernation

# no btrfs
sudo -i
truncate -s 0 /swapfile
chattr +C /swapfile
fallocate -l 2G /swapfile
chmod 0600 /swapfile
mkswap /swapfile
swapon /swapfile

# btrfs
btrfs filesystem mkswapfile --size 2G /swapfile

hibernation needs offset for file as cmdline boot arg or in /sys/power/resume_offset

sudo -i
btrfs inspect-internal map-swapfile -r /swapfile > /sys/power/resume_offset

btrfs resize

use sudo btrfs filesystem show /dev/nvme0n1p5 to get the devid of the fs to be resized
to resize a partition, delete and recreate it.
btrfs filesystem resize 1:max /mnt where /mnt contains a btrfs and devid is 1
for some specific size, try btrfs filesystem resize 1:+300G
if expanding, expand partition first, then expand fs. if shrinking, shrink fs then partition

btrfs subvolumes

btrfs subvolume set-default /path/to/subvol # will be mounted by kernel
btrfs subvolume snapshot -r /source/subvol /snap/name # read only
btrfs subvolume get-default / # currently @root is default
btrfs subvolume list /

to restore a backup create a rw snapshot from a ro snapshot

useful commands

grep @ /proc/mounts | bat --language fstab
sudo mount -o "remount,exec" LABEL=Archlinux /home/vacation/.local/share/Steam
/usr/bin/btrfs check --check-data-csum /dev/disk/by-label/Archlinux # fsck equivalent
root btrfs send --compressed-data /toplevel/@root | btrfs receive /media/data/toplevel # send ro snapshot, pass folder to receive

Infra

AWS

samples
https://github.com/aws-samples
cdk
https://github.com/aws-samples/aws-cdk-examples/tree/master/typescript

K8S

exec command in pod

kubectl exec --stdin --tty $POD -c $CONTAINER -- /bin/sh

psql

PGPASSWORD=$(gcloud auth print-access-token) psql -U [email protected] -d hub -h /home/vacation/.eleanor_sql_sockets/ele-qa-436057:us-east1:eleanor-postgres

get secrets

kubectl get secrets -o jsonpath='{.data.FOO_DATA}' $CONTAINER

kubectl edit secrets
kubectl get pods
kubectl describe pod
kubectl logs --tail=100 $POD $CONTAINER

k8s scheduling/cron scheduling syntax helper comment

# ┌───────────── minute (0 - 59)
# │ ┌───────────── hour (0 - 23)
# │ │ ┌───────────── day of the month (1 - 31)
# │ │ │ ┌───────────── month (1 - 12)
# │ │ │ │ ┌───────────── day of the week (0 - 6) (Sunday to Saturday;
# │ │ │ │ │                                   7 is also Sunday on some systems)
# │ │ │ │ │
# │ │ │ │ │
# * * * * *

k8s jobs and crons

# job from cron
kubectl create job --from=cronjob/hub-server-sync-athena-ticklers my-job

gcloud install

link

plugin update

gcloud components install gke-gcloud-auth-plugin
export USE_GKE_GCLOUD_AUTH_PLUGIN=True
gcloud components update
gcloud --project=ele-qa-436057 container clusters --region=us-east1 get-credentials cluster

pod details

kubectl get pod hub-server-insurance-cleanup-4wb86-9nw44 --output=yaml

Typescript

extract type from relay type

type FragmentRefs<T extends string> = T

type FeedList_query = {
    readonly posts: {
        readonly endCursorOffset: number;
        readonly startCursorOffset: number;
        readonly count: number | null;
        readonly pageInfo: {
            readonly hasNextPage: boolean;
            readonly hasPreviousPage: boolean;
            readonly startCursor: string | null;
            readonly endCursor: string | null;
        };
        readonly edges: ReadonlyArray<{
            readonly node: {
                readonly id: string;
                readonly " $fragmentRefs": FragmentRefs<"Post_post">;
            } | null;
        } | null>;
    };
    readonly me: {
        readonly " $fragmentRefs": FragmentRefs<"Post_me">;
    } | null;
    readonly " $refType": "FeedList_query";
 };


type ExtractNode<T extends {edges: any}> = NonNullable<NonNullable<T['edges'][number]>['node']>
type Post = ExtractNode<FeedList_query['posts']>

tsconfig.json typeroots needs to manually add node_modules/@types, like a default that is lost when you specify manually another path

"typeRoots": ["src/types", "node_modules/@types"]

https://stackoverflow.com/questions/39040108/import-class-in-definition-file-d-ts
ambinet vs local type declarations and different syntax to import in them

declare module 'repl.history' {
  // import some types from node
  type AddHistory = (repl: import('repl').REPLServer, file: import('fs').PathLike) => void
  declare const def: AddHistory
  export default def
}

type Primitive =
  | string
  | boolean
  | number
  | bigint
  | symbol
  | null
  | undefined;

type object = any non primitive.

Sql and psql with docker

connect to posgress using psql inside docker
sudo docker-compose exec postgres psql -U postgres -d hub

describe table
\d table

connect to db
\c

quit
\q

list all tables
\dt

list all dbs
\l

delete row
DELETE FROM table_name WHERE condition;

clone db
create database new_database_name (Name of new database which we have cloning from another database) WITH TEMPLATE old_database_name

kill off containers and volumes using docker-compose
docker-compose down --rmi all -v

if fail, you may need to docker ps -a and remove any images that errored with docker rm <image id>

dump docker dp to local file
sudo docker-compose exec postgres pg_dump -h localhost -U "postgres" "hub" --format plain > "hub.dump.sql"
to restore
psql -d hub -f hub.dump.sql

docker cp from container to local fs

use sudo docker cp 5810d91ee2e5:/out.txt $PWD/out.txt
where
postgres is the container for docker-compose, and 5810d91ee2e5 is the container for docker
5810d91ee2e5 is from sudo docker ps
/out.txt is from sudo docker-compose exec postgres pwd which replies with / and out.txt is the file I want. exec postgres ls also works btw!

docker nuke

removecontainers() {
    docker stop $(docker ps -aq)
    docker rm $(docker ps -aq)
}

armageddon() {
    removecontainers
    docker network prune -f
    docker rmi -f $(docker images --filter dangling=true -qa)
    docker volume rm $(docker volume ls --filter dangling=true -q)
    docker rmi -f $(docker images -qa)
}

postgres

-- seq error fix
SELECT MAX(the_primary_key) FROM the_table;   
SELECT nextval('the_primary_key_sequence');
-- nextval needs to be hight than tha table
SELECT setval('the_primary_key_sequence', (SELECT MAX(the_primary_key) FROM the_table)+1);
-- just getting the value fixes, or set it with setval

-- drop all tables quickly
DROP SCHEMA public CASCADE;
CREATE SCHEMA public;
     
GRANT ALL ON SCHEMA public TO postgres;
GRANT ALL ON SCHEMA public TO public;

psql

psql -U postgres -d db-name-here
\dt to list definitions

psql -U hub-server -W -d hub -h /home/vacation/eleanor_sql_sockets/ele-qa-436057:us-east1:eleanor-postgres
-U user
-W prompt pass
-d dbname
-h host/socket

column information
SELECT column_name, is_nullable, data_type, column_default FROM information_schema.columns WHERE table_name = 'foo';

postgres upgrade

systemctl stop postgresql
sudo -i
mv /var/lib/postgres/data /var/lib/postgres/olddata
mkdir /var/lib/postgres/data /var/lib/postgres/tmp
chown postgres:postgres /var/lib/postgres/data /var/lib/postgres/tmp
exit
sudo -iu postgres
cd /var/lib/postgres/tmp
initdb -D /var/lib/postgres/data --encoding=UTF8
# it is the old version upgrading from, ls /opt might reveal it
pg_upgrade -b /opt/pgsql-14/bin -B /usr/bin -d /var/lib/postgres/olddata -D /var/lib/postgres/data
exit
systemctl start postgresql
# 6 is number of cores
sudo -u postgres vacuumdb --all --analyze-in-stages --jobs=6
echo echoing cleanup commands
echo sudo -i
echo rm -fr /var/lib/postgres/olddata
echo rm -fr /var/lib/postgres/tmp
echo exit

x-work

get list from website

// do not use really sucks. need to concat with paragraphs and titles
re = []
sel = ".JobOpeningListItem_JobDescription__1DPoi''
document.querySelectorAll(sel).forEach(x => re.push(x.innerText))
('vaga: ' + re.join('\n\n vaga: ')).replace(/Learn more >>/g, '')

// real dolar
var makeBrlToUsd = rate => brl => Math.round(brl*rate)

High level

speed

In posts talking about how someone made something fast, I see the same idea repeat time after time: it has linear time complexity, has great cache locality, and saturates all available processors.

Interview questions

technical, high-level

How do you handle the diversity of patterns within the JS ecosystem? maybe follow-up: What are some good solutions or patterns you have used that worked for you?
What's your thoughts on JS itself and it's current direction?

execution and team work

Have you worked with designers? what were some highlights there?
We all know it's very hard to give estimates, how to do you approach that? what's your attitude?
What is a big mistake in product process you've seen or committed yourself?
Your favorite meeting and why?
Tell me a good idea for a project that your company should take on, but will not. Explain from both sides, even if they are wrong.
Walk me through a project you enjoyed start to finish. Explain design decisions (why x and not y?)

code review

https://www.michaelagreiler.com/respectful-constructive-code-review-feedback/
https://phauer.com/2018/code-review-guidelines/
https://mtlynch.io/code-review-love/

Datadog

APM traces

query samples

Service:hub-server @url:*drivers-license*
trace_id:2934837843

tricks

search for logs using kubectl, find trace id, then do https://app.datadoghq.com/apm/trace/{id}

if code looks like this

infra.Logger.Info().Str("event", "message.new")

query to find field would be @event:message.new

RUM custom actions

find with @action.target.name:foobar

Cookie cliker

// node
prestigeLvls = Math.cbrt((allTimeCookies + popingWrinklers + sellingBuildings + difference) * 10 ** 18 /* 18 for quintilion cookies*/)/10000

// load cookie monster into cookie clicker
Game.LoadMod('https://cookiemonsterteam.github.io/CookieMonster/dist/CookieMonster.js');

Go

Installing private repos as packages

update ~/.gitconfig with

[url "ssh://[email protected]/"]
      insteadOf = https://github.com/

run go env -w GOPRIVATE=github.com/eleanorhealth/* and add to shell init file export GOPRIVATE=github.com/eleanorhealth/*
run ssh-keyscan github.com > ~/.ssh/known_hosts
try to run go mod download "github.com/eleanorhealth/member-server@<latest commit on main>" and see if you get no errors, if so, run go get "github.com/eleanorhealth/member-server@<latest commit on main>" to add the package

notes

omitting the commit hash doesn't work
ssh-key add may be needed on mac, it will prompt you for password
code must be merged to main

coverage unit

go get golang.org/x/tools/cmd/cover
go test -race -coverprofile .coverage ./svc/server/application/application/...
go tool cover -html=.coverage

/**
 * Maybe captures the result of some operation that may fail.
 * If there is an non null error, attempting to retrieve result will throw.
 */
export class Maybe<T> {
    #result: T | null
    #error: Error | null

    // constructor(data: Data | null, error: Error | null) {
    constructor() {
        this.#result = null
        this.#error = null
    }

    /**
     * throws unless error() returns null.
     */
    result(): T {
        if (this.#error) {
            throw this.#error
        }

        if (this.#result === null) {
            throw new Error("null data")
        }

        return this.#result
    }

    /**
     * if null, result() is safe to call
     */
    error() {
        return this.#error
    }

    /**
     * error must be null for result to be read
     */
    setResult(data: T) {
        this.#result = data
    }

    /**
     * blocks result from being read
     */
    setError(message: string) {
        this.#error = new Error(message)
    }
}

AI

52.7k words on member-server codebase
244 non mock non test files
find . -type f -name *.go | grep -Ev test.go | grep -Ev mock
avg 314 words per file (handlers)
avg 215 word per file (codebase)
say 5k words for a massive conversation context with ai
have 59k words of context to use with ai