DEV Community: kurab

Why "Not Deciding" Breaks Agile Teams - Decision-less Agility

kurab — Sun, 21 Dec 2025 07:27:19 +0000

I've been seeing a familiar pattern lately:

A team adopts Agile.
They start using AI across the workflow.
Motivation is high.

And yet leadership says: "Nothing got easier."

When I join those teams, I rarely find laziness or bad intent. People work hard. They care.
But the same structural symptoms show up quietly, repeatedly:

Important decisions stay "pending" for months
There's no measurement that tells whether we're actually improving
Retrospectives happen, but learning doesn't reliably change the next iteration

From the inside, this can feel reasonable. Often it is the most reasonable option available.
And that's exactly why it persists.

I call this recurring breakdown Decision-less Agility.

This post isn't a tutorial. And it's not a best practices list.
It's an attempt to name a failure pattern so teams can recognize it early before it hardens.

Uncertainty isn't the problem. Losing decision boundaries is.

Agile assumes uncertainty. That's fine.

The failure begins when teams lose the distinction between:

Not deciding the future (appropriate under uncertainty)
Deciding the minimum necessary now (required to stay governable)

When that line blurs, "we'll decide later" becomes a default posture - not a temporary choice.

In multi-stakeholder environments (consulting, internal platforms, regulated domains, or simple politically complex organizations), "later" often means "never", because:

decisions create accountability
accountability creates risk
risk gets redistributed... until it evaporates

Work continues. Releases continue.
But decision-making capacity quietly collapses.

Decisions don't stay unmade. They get embedded.

Unresolved decisions don't remain empty. They get filled usually by the path of least resistance.

They become:

scattered conditionals
default values that become doctrine
"temporary" flags that never leave
implicit contracts that drift without anyone explicitly acknowledging the change

Over time:

the system's "correctness" fragments into code shards
temporary behavior becomes permanent
nobody can explain what is safe to change

Eventually, "we can't touch that" becomes a rational survival strategy.

And the organization learns the worst possible lesson:

Not deciding is safer than deciding.

That is the core of Decision-less Agility.

Why this becomes "rational" inside the team

From the outside, this looks like avoidance.
From the inside, it often looks like risk management.

Common forces include:

uncertainty that won't disappear
a bias toward shipping visible progress
accountability that becomes politically expensive
documentation and alignment costs that don't show up in short-term metrics
missing measurement, so trade-offs can't be compared

None of these forces are "wrong".
But together, they resolve into a structure where the safest move is to keep moving without deciding.

AI changes the speed, not the nature of the failure

In the AI era, this pattern hardens faster.

AI can produce plausible output even when assumptions and constraints are missing.
That's powerful - but it can also make unresolved context less visible.

So AI isn't the villain here.
It's an amplifier:

implementation progresses even with unclear assumptions
short-term success masks contradictions
local assumptions harden faster

If the frame is missing, AI still outputs.
Humans do the same.
The difference is speed and scale.

Naming the pattern changes the conversation

Naming a situation doesn't fix it.
But it often restores something essential: the ability to talk about it without blame.

Once a team recognizes Decision-less Agility, they can ask more useful questions:

What is the minimum decision set required to proceed safely?
Where is our current Source of Truth?
What assumptions are we letting code silently decide?
What single measurement would tell us if we're actually improving?

The goal isn't certainty.
It's restoring decision-making under uncertainty.

If this feels familiar

I'm collecting patterns like this in Failure Patterns: a pattern language for recurrent breakdowns in software systems.

This post is one slice of that work.

If you want to explore more patterns (and the forces that sustain them), start here:

https://takt.dev/failure-patterns

How to upload blobs to Azure Storage from Next.js

kurab — Wed, 02 Mar 2022 11:11:39 +0000

Summary

Create Next.js + TypeScript + Prisma App
Upload blobs such as mp4 to Azure Storage
Give a unique name *1
Save it to Database as well

The URI to reference a container or a blob must be unique. Because every account name is unique, two accounts can have containers with the same name. However, within a given storage account, every container must have a unique name. Every blob within a given container must also have a unique name within that container.

...

The Blob service is based on a flat storage scheme, not a hierarchical scheme.

via. Naming and Referencing Containers, Blobs, and Metadata

Tech

Next.js
Prisma
PostgreSQL
TypeScript

package.json is as following.

{
...
  "dependencies": {
    "@azure/storage-blob": "^12.8.0",
    "@prisma/client": "^3.9.2",
    "axios": "^0.25.0",
    "next": "12.0.10",
    "react": "17.0.2",
    "react-dom": "17.0.2",
    "react-hook-form": "^7.27.0",
    "uuid": "^8.3.2"
  },
  "devDependencies": {
    "@types/axios": "^0.14.0",
    "@types/node": "17.0.17",
    "@types/react": "17.0.39",
    "@types/uuid": "^8.3.4",
    "eslint": "8.9.0",
    "eslint-config-next": "12.0.10",
    "prisma": "^3.9.2",
    "typescript": "4.5.5"
  }
}

Create Next.js app

Place upload file form on top page
Using react-hook-form to handle form
Prisma ORM + PostgreSQL
Create get/register API and call them by axios

Install Next.js and modules

$ yarn create next-app azure-storage --typescript
$ cd azure-storage
$ yarn add @prisma/client react-hook-form axios
$ yarn add @types/axios prisma --dev

PostgreSQL by docker

docker-compose.yml

version: "3.8"

services:
  db:
    image: "postgres:12"
    ports:
      - "5432:5432"
    volumes:
      - ./pgdata:/var/lib/postgresql/data
    environment:
      - POSTGRES_USER=${DB_USER}
      - POSTGRES_PASSWORD=${DB_PASS}
      - POSTGRES_DB=${DB_NAME}

.env

DB_NAME=sample
DB_USER=johndoe
DB_PASS=randompassword

$ docker-compose up -d

Prisma

$ npx prisma init

.env

+ DATABASE_URL="postgresql://johndoe:randompassword@localhost:5432/mydb?schema=public"

prisma/schema.prisma

generator client {
  provider = "prisma-client-js"
}

datasource db {
  provider = "postgresql"
  url      = env("DATABASE_URL")
}

model Item {
  id        Int       @id @default(autoincrement())
  name      String
  createdAt DateTime  @default(now())

  @@map(name: "items")
}

$ npx prisma generate
$ npx prisma migrate dev

Code

├── pages
│   ├── api
│   │   ├── items.ts     // get item API
│   │   └── register.ts  // register item API
│   └── index.tsx        // form and display blob list
├── hooks
│   └── useItem.ts       // hook to call APIs
└── types
    └── ItemType.ts      // type of Item

detail code is: kurab/next-azure-storage:baseForm

$ yarn dev
or
$ yarn build
$ yarn start

OK.

Azure implementation

Read a quick start document and apply it to Next.js + TypeScript.

Quickstart: Manage blobs with JavaScript v12 SDK in Node.js

install modules

$ yarn add @azure/storage-blob uuid
$ yarn add @types/uuid --dev

only @azure/storage-blob module is needed to upload blob to Azure Storage.
uuid is for giving a unique name such as "fd629c13-dcc5-4503-a0be-934e96b99b27".

Access with sasToken

This time, I use sasToken (Shared Access Signature) to access Azure Storage. Create sasToken on Azure Portal and note it (sv=blah blah...).

.env

+ NEXT_PUBLIC_STORAGESASTOKEN='sv=.....'
+ NEXT_PUBLIC_STORAGERESOURCENAME='xxxxx'

NEXT_PUBLIC_STORAGERESOURCENAME is Storage Account.

Initialize BlobServiceClient by these information like:

import { BlobServiceClient } from '@azure/storage-blob';

const sasToken = process.env.NEXT_PUBLIC_STORAGESASTOKEN;
const storageAccountName = process.env.NEXT_PUBLIC_STORAGERESOURCENAME;

const blobService = new BlobServiceClient(
  `https://${storageAccountName}.blob.core.windows.net/?${sasToken}`
);

Easy!

rename and upload

To keep code simple, easy to understand, I ignore transaction for this time. That is assuming both upload and save database never fail.

Add onSubmitSave method on index.tsx that rename file, upload and save. (actual upload and save are in hook)

...
import { v4 as uuidv4 } from 'uuid';
...
  const onClickSave = (formData: any) => {
    if (formData.files[0]) {
      const newFileName =
        uuidv4() + '.' + formData.files[0].name.split('.').pop();
      uploadFileToBlob(formData.files[0], newFileName);
      registerItem(newFileName);
    }
  };
...

You can write smarter. For this time, I wanted to separate steps to explain.

Now, uploadFileToBlob method in useItem.ts hook.

...
import { BlobServiceClient, ContainerClient } from '@azure/storage-blob';

const containerName = 'sample-container';
const sasToken = process.env.NEXT_PUBLIC_STORAGESASTOKEN;
const storageAccountName = process.env.NEXT_PUBLIC_STORAGERESOURCENAME;
...
  const uploadFileToBlob = useCallback(
    async (file: File | null, newFileName: string) => {
      setLoading(true);
      if (!file) {
        setMessage('No FILE');
      } else {
        const blobService = new BlobServiceClient(
          `https://${storageAccountName}.blob.core.windows.net/?${sasToken}`
        );

        const containerClient: ContainerClient =
          blobService.getContainerClient(containerName);
        await containerClient.createIfNotExists({
          access: 'container',
        });

        const blobClient = containerClient.getBlockBlobClient(newFileName);
        const options = { blobHTTPHeaders: { blobContentType: file.type } };

        await blobClient.uploadData(file, options);
        setMessage('uploaded');
      }
      setLoading(false);
    },
    []
  );
...

This code is just following the document.
If you are sure container has already existed, you can remove:

await containerClient.createIfNotExists({
  access: 'container',
});

This line will warn you everytime when container exists.

PUT https://.... 409 (The specified container already exists.)

after uploaded blobs, it's up to you. In sample code, I set a new blobs list to state.

const blobs = await getBlobsInContainer(containerClient);
setBlobs(blobs);

like this.

Done

kurab/next-azure-storage

I use AWS most of the time, so when I hear "Azure," I think "Damn..." But Azure is so richly documented that I don't have much trouble finding what I'm looking for.

It also works well with github and VSCode, of course. Good.

Check list before re-install Xcode Command Line Tool

kurab — Wed, 04 Nov 2020 15:38:25 +0000

Environment & Phenomena

Mac Pro (Late 2013)
macOS Catalina ver. 10.15.7
Xcode ver. 12.1 (12A7403)

failed to xcodebuild.

$ xcodebuild -version
xcode-select: error: tool 'xcodebuild' requires Xcode, but active developer directory '/Library/Developer/CommandLineTools' is a command line tools instance

Really installed?

check the path

/Library/Developer/CommandLineTools

Check Location

Xcode > Preference > Locations

For me, this setting was empty. and now it works.

$ xcodebuild -version
Xcode 12.1
Build version 12A7403

Still not work, then Re-Install

$ xcode-select --install

$ rm -rf /Library/Developer/CommandLineTools
$ xcode-select --install

or download installer from Apple Developers

Serverless chat bot by Google Apps Script - step/step

kurab — Wed, 04 Nov 2020 15:27:09 +0000

TL;DR

By using "GAS(Google Apps Script)", notify today's events on "Google Calendar" to "Google Chat", "Chatwork" and "LINE Messenger" every morning.

All you need is

Web browser
Accounts (Google, Chatwork, LINE, LINE Developer)

and it's free.

source code: kurab/notifyEvents

What's Google Apps Script?

It's a JavaScript-based programming language provided by Google that allows you to do a lot of things, and it's easy to integrate with Google's services.

Nowadays, even non-engineers can let Python do all the routine work for them. But with Python, it's a bit more difficult to get started in terms of having to install this and that. With Google Apps Script (GAS), on the other hand, all you need is a browser and a Google account, so even non-engineers can easily get started. Furthermore, there are many companies that use GSuite, so routine work may be relatively related to Google's services.

GAS is also basically JavaScript, so you can solve most of the problems by just Google it.

Hello, world

First of all, access to Google Drive. I'd like to work in "My Drive > bot" directory for this time.

Select "Google Apps Script" from "+New" menu on top left or "Right click" menu > more.

Then, you will move to the script editor like this.

Let name helloWorld as project name and update code like this:

function myFunction() {
  console.log('Hello, World');
}

After saving your code(Ctrl+s or save button), click Run button(▶) on menu. It seems nothing happens but you can find log from menu "View > Logs".

Congrats!! Now you have done your first Google Apps Script. This is the basic working flow of GAS.

Get Calendar Events

Here is the reference of Google Calendar API.

Calendar Service

This time, I'd like to get today's events information.

Let's create GAS file and name "notifyCalendar" or anything you want.

To start, access your calendar from script and get the calendar name.
Find Calendar ID in "Settings and sharing" of calendar > Integrate calendar > Calendar ID and paste it to {Calendar ID} in following code.

function notifyEvents() {
  const calendarId = '{Calendar ID}';
  const calendar = CalendarApp.getCalendarById(calenderId);
  const calendarName = calendar.getName();

  console.log(calendarName);
}

When you run this, there will be a review of your calendar access and you will be prompted to grant it. Then wait for the execution to complete, open the log and you will see the calendar name. If there is no log yet, wait for a while.

Now, let's continue to get today's events in the calendar and put them in the log.

function notifyEvents() {
  const calendarId = '{Calendar ID}';
  const calendar = CalendarApp.getCalendarById(calenderId);
  const calendarName = calendar.getName();

  const today = new Date();
  const calendarEvents = calendar.getEventsForDay(today);

  console.log(calendarEvents);
}

If you have events today, you will get data like [{}, {}, {}]. If you don't have event, you will get empty array like [].

List up events if you have, and display "no event" if you don't have. Well, if you want to specify a period of time, not just today, use the getEvents method.

function notifyEvents() {
  const calendarId = '{Calendar ID}';
  const calendar = CalendarApp.getCalendarById(calendarId);
  const calendarName = calendar.getName();

  const today = new Date();
  const calendarEvents = calendar.getEventsForDay(today);

  if (calendarEvents.length) {
    for (var event of calendarEvents) {
      var eventTitle = event.getTitle() ? event.getTitle() : '(no title)';
      console.log(eventTitle);
    }
  } else {
    console.log('no event');
  }
}

getTitle method doesn't return "(no title)" for no titled events. You need to set it by yourself.

Some events on your calendar are timed and some are not, so if you have a timed event, show it as well. At the same time, you don't want to get sent just the name of the event, so add a thoughtful message as well.

function notifyEvents() {
  const calendarId = '{Calendar ID}';
  const calendar = CalendarApp.getCalendarById(calendarId);
  const calendarName = calendar.getName();

  const today = new Date();
  const calendarEvents = calendar.getEventsForDay(today);

  var message = 'Good morning!\n' +
                'Notification from ' + calendarName + 'at' +
                 Utilities.formatDate(today, 'GMT+0900', 'MM/dd') +
                '\n-----------------------------------------';

  if (calendarEvents.length) {
    for (var event of calendarEvents) {
      var eventTitle = event.getTitle() ? event.getTitle() : '(no title)';
      message = message + '\n- ' + eventTitle;
      if (!event.isAllDayEvent()) {
        message = message + ' at '+ 
                  Utilities.formatDate(event.getStartTime(), 'GMT+0900', 'HH:mm') + '~' + 
                  Utilities.formatDate(event.getEndTime(), 'GMT+0900', 'HH:mm') + '(JST)';
      }
    }
  } else {
    message = message + '\nno event';
  }

  console.log(message);
}

We have successfully get calendar events now.
You can get more information by the API, check documents like getXX method or isXX.

Class CalendarEvent

Post to chat

Google Chat

Issue webhook url on chat room that you want to post calendar events.

and paste it to {Webhook URL}.

function notifyEvents() {
    ...
  } else {
    message = message + '\nno evnet';
  }

  postToChat(message);
}

function postToChat(message) {
  const chatWebhook = '{Webhook URL}';
  const messageText = { 'text': message };
  const options = {
    'method': 'POST',
    'headers': {
      'Content-Type': 'application/json; charset=UTF-8'
    },
    'payload': JSON.stringify(messageText)
  };
  var result = UrlFetchApp.fetch(chatWebhook, options);
  Logger.log(result);
}

You will ask to review access, so grant it.

Chatwork

Issue Chatwork API Token on Chatwork settings and memo chat room ID that you want to post. xxxxxx(number) of #!ridxxxxxx in URL.

Now on GAS editor, from menu > Resources > Libraries, add:

M6TcEyniCs1xb3sdXFF_FhI-MNonZQ_sT

I set the Latest version, 18.

Paste your Chatwork API Token and Room ID.

function notifyEvents() {
  ...
  //postToChat(message);
  postToChatwork(message);
}

function postToChat(message) {...}

function postToChatwork(message) {
  const chatworkToken = '{Chatwork API Token}';
  const chatworkRoomId = '{Chatwork Room ID}';
  const chatworkClient = ChatWorkClient.factory({token: chatworkToken});
  chatworkClient.sendMessage({room_id: chatworkRoomId, body: message});
}

LINE Messenger

This time, you can send a message only to yourself, so you can send the message to a fixed ID.

Flow is:

on LINE Developers Create "Provider" and "Messaging API Channel"
In Messaging API settings, issue "Channel access token (long-lived)"
In Messaging API settings, disable "Auto-reply messages" and "Greeting messages" (Cuz we don't need them)
In Basic settings, copy "Your user ID".
On GAS, publish as Web App and copy URL (menu > Publish > Deploy as web app)
On LINE Developers, in Messaging API settings, set URL that you copied at step5 as Webhook
On your LINE Smartphone app, scan QR code to become a friend
Run GAS

For a detailed explanation of LINE Developers, I'll leave you to look at other articles and so on, but the GAS code will be following. Put in the Token and user ID that you have copied in steps 2 and 4 above.

...
function postToLine(message) {
  const lineToken = '{LINE Token}';
  const lineMessagePushUrl = 'https://api.line.me/v2/bot/message/push';
  const lineHeaders = {
    "Content-Type": "application/json; charset=UTF-8",
    "Authorization": "Bearer " + lineToken
  };
  const linePayload = {
    "to" : "{User ID}",
    "messages" : [{
      "type": "text",
      "text": message
    }]
  };
  const options = {
    "method": "POST",
    "headers": lineHeaders,
    "payload": JSON.stringify(linePayload)
  };
  var result = UrlFetchApp.fetch(lineMessagePushUrl, options);
  Logger.log(result);
}

Set Trigger and execute it every morning

Finally, let's set up automatic execution. When you click "Trigger" button next to "Run" button in the menu, you will be redirected to "Trigger Settings" page. Create a new Trigger.

In this case, I want this applet to notify only once every morning. About the mysterious range from 9am to 10am, I tried to run it for a few days and it's 9:46am and every day is different. I don't feel uncomfortable, but in my case, I thought it's good as long as the notification comes before the start of work.

That's all, but... I'll get notifications on Saturday and Sunday too, which will ruin my weekend feeling. So I'll add a process to GAS to not send messages on Saturdays and Sundays. Holidays...I can do that, but this time, it's okay.

function notifyEvents() {
  const calendarId = 'kura@indigames.net';
  const calendar = CalendarApp.getCalendarById(calendarId);
  const calendarName = calendar.getName();

  const today = new Date();
  const dayOfWeek = today.getDay();
  if (dayOfWeek === 0 || dayOfWeek === 6) return;

  const calendarEvents = calendar.getEventsForDay(today);
  ...
}
...

PS

When I googled "Google Apps Script", there are a lot of articles start from Spreadsheet..., and I had the impression that it is equivalent to the Macro of Excel, and it reminds me like the secret Excel Macro that no one can maintain.

Well, anyways, I think it's good to have the option of using Google Apps Script for small things.

Sample script for this article is here.

kurab/notifyEvents

A one-liner that turns DUMPed table definitions in MySQL into a Confluence Wiki or Markdown notation

kurab — Mon, 02 Nov 2020 11:39:07 +0000

I want to easily paste a DUMPed table definition in MySQL into Confluence. I'm sure you've had that happen to you.
I want to paste DUMPed table definitions in MySQL to Dev.to easily. I'm sure you have.

It's a bit tedious to write table definitions for a DB in a document to be delivered. It's not productive at all, and I want to get it done in one shot.

Target

From

+--------------------+------------------+-----------------+------+-----+---------+-----------------------------+---------------------------------+--------------------------------------------------------+
| Field              | Type             | Collation       | Null | Key | Default | Extra                       | Privileges                      | Comment                                                |
+--------------------+------------------+-----------------+------+-----+---------+-----------------------------+---------------------------------+--------------------------------------------------------+
| id                 | int(10) unsigned | NULL            | NO   | PRI | NULL    | auto_increment              | select,insert,update,references | ID                                                     |
| name               | varchar(256)     | utf8_general_ci | YES  |     | NULL    |                             | select,insert,update,references | 名称                                                   |
+--------------------+------------------+-----------------+------+-----+---------+-----------------------------+---------------------------------+--------------------------------------------------------+

One-liner

Confluence Wiki

$ mysql -h (host) -u (user) -p (schema) -e 'show full columns from (table)' | sed 's/^/|/g;s/$/|/g;s/\t/ | /g;1 s/|/||/g'

Markdown

$ mysql -h (host) -u (user) -p (schema) -e 'show full columns from (table)' | sed -E "s/^/\| /g;s/$/ \|/g;s/\t/ \| /g;1{h; p; s/\w+/:-:/g}"

By using xsel or pbcopy, save it in clipboard in one shot.

Shell script to monitor network on MacOS

kurab — Mon, 02 Nov 2020 11:17:56 +0000

TL;DR

In the context of a global pandemic, more and more people are working from home. In software development, it can be important to know which network to access from, or network speed for a comfortable video conference. It's useful to have a quick check on such things.

I'd like to be able to look up

Outgoing IP Address
Internal IP Address
The Wi-Fi SSID
Network Speed DL/UL

with bash script.
And I'd like to display it on the desktop with Geektool, which may not be used by many people these days.
If you just want to see the results, skip all of them and go to github.

kurab/networkMonitorMac

Environments

Mac Pro (late 2013) / macOS Catalina version 10.15.7
MacBook Air 2020 / macOS Catalina version 10.15.7

Outgoing IP Address

When connecting to a protected network based on a reliable, fixed IP address, it's important to know your external IP address.

You can get it with following command.

$ curl --silent api.ipify.org --max-time 5

So, the script will be like this:

#!/bin/bash
ext_info=`curl --silent api.ipify.org --max-time 5`

wait

if [ -n "$ext_info" ];
then
    echo "External: $ext_info"
else
    echo "External: OFFLINE"
fi

unset ext_info

"wait" may not be necessary, but we'll put it in from the beginning later, as there's heavy processing involved.

Internal IP Address

First, find out the MAC address of your Wi-Fi adapter.

Go to Preferences > Network and select Wi-Fi, click Advanced to display Hardware and remember the MAC Address. Just remember the last two letters or so.

$ ifconfig

When you hit this command, you'll see a list of things, but try looking for sections like en0: en1: en2: ... In the "ether" section, you can see something like the MAC Address, so look for the MAC Address of the Wi-Fi.

In my case, en2 for Mac Pro and en0 for MacBook Air. The explanation continues with en0.

The IP address of the assigned network is written in en0 of ifconfig, so I extract it.

#!/bin/bash
en0_info=`ifconfig en0 | grep inet | grep netmask | grep -v 127.0.0.1 | awk '{print $2}'`

wait

if [ -n "$en0_info" ];
then
    echo "Wi-Fi: $en0_info"
else
    echo "Wi-Fi: INACTIVE"
fi

unset en0_info

SSID

On MacOS, you can learn a lot of things about Wi-Fi by typing the following command, which was the same on pre-Catalina OSes.

$ /System/Library/PrivateFrameworks/Apple80211.framework/Versions/A/Resources/airport -I

The SSID is the information you want to know. You can get various other information as well, so please scatter what you need as needed.

#!/bin/bash
en0_info=`ifconfig en0 | grep inet | grep netmask | grep -v 127.0.0.1 | awk '{print $2}'`

wait

if [ -n "$en0_info" ];
then
    wifi_network=`/System/Library/PrivateFrameworks/Apple80211.framework/Versions/A/Resources/airport -I | awk -F: '/ SSID: / {print $2}' | sed -e 's/.*SSID: //'`


    echo "Wi-Fi: $en0_info"
    echo "- SSID: $wifi_network"

    unset wifi_network
else
    echo "Wi-Fi: INACTIVE"
fi

unset en0_info

Network Speed DL/UL

Use speedtest-cli to measure the network speed.

It's a bit different in function, but either one is fine. This time, I'll use the latter one, pypi's. I'd like to get it with curl and put it in /usr/local/bin.

$ curl -Lo speedtest-cli https://raw.githubusercontent.com/sivel/speedtest-cli/master/speedtest.py
$ chmod +x speedtest-cli
$ mv speedtest-cli /usr/local/bin/.

versions:

$ speedtest-cli --version
speedtest-cli 2.1.2
Python 3.8.3 (default, Jul  2 2020, 11:26:31) [Clang 10.0.0 ]

If you switch to a Python version, such as 3.9.0, the following will still work.

$ speedtest-cli --simple

You'll see that it takes quite a bit of time to measure. The code we've been using so far will give you instantaneous results of execution, so if it's annoying, you can separate the scripts.

$ speedtest-cli --simple | grep Download | awk '{print $2}'

You can get it like this, but this is not practical because you have to measure twice with DL/UL, which takes a crazy amount of time. Let's use the measurement once and use it well.

#!/bin/bash
speed_info=`/usr/local/bin/speedtest-cli --simple 2>/dev/null | grep load | sed -e 's/^.*: //g;s/ M.*$//g'`

wait

if [ -n "$speed_info" ];
then
    dl_speed=`echo $speed_info | awk '{print $1}'`
    ul_speed=`echo $speed_info | awk '{print $2}'`

    echo "Download: $dl_speed Mbps"
    echo "Upload: $ul_speed Mbps"

    unset speed_info
    unset dl_speed
    unset ul_speed
else
    echo "Download: -"
    echo "Upload: -"
fi

speedtest-cli spits out a standard error if you're not connected to the network. I discarded it. "Mbps" is fixed. We'll think about it again when the network is much faster.

LAN cable?!

There may be a case that you are not using Wi-Fi but LAN cable to connect to the Internet. The way to check the network interface is the same. This time, en0 in Mac Pro and MacBook Air do not have the interface of LAN cable, but it is possible to use it via the adapter of USB-C. In that case, it was en2.

In the explanation.

Wi-Fi: en0
LAN: en2

In the case of a LAN, set only the IP address in the network.

#!/bin/bash
en2_info=`ifconfig en2 |grep inet |grep netmask |grep -v 127.0.0.1 |awk '{print $2}'`

wait

if [ -n "$en2_info" ];
then
    echo "Ethernet: $en2_info"
else
    echo "Ethernet: INACTIVE"
fi

unset en2_info

Final Source Code

Put all of the above together and adjust the spaces and so on, the source code will look like this

#!/bin/bash
speed_info=`/usr/local/bin/speedtest-cli --simple | grep load | sed -e 's/^.*: //g;s/ M.*$//g'`

ext_info=`curl --silent api.ipify.org --max-time 5`

en0_info=`ifconfig en0 | grep inet | grep netmask | grep -v 127.0.0.1 | awk '{print $2}'`
en2_info=`ifconfig en2 | grep inet | grep netmask | grep -v 127.0.0.1 | awk '{print $2}'`

wait

if [ -n "$speed_info" ];
then
    dl_speed=`echo $speed_info | awk '{print $1}'`
    ul_speed=`echo $speed_info | awk '{print $2}'`

    echo "Download   : $dl_speed Mbps"
    echo "Upload     : $ul_speed Mbps"

    unset speed_info
    unset dl_speed
    unset ul_speed
else
    echo "Download   : -"
    echo "Upload     : -"
fi

echo ""

if [ -n "$ext_info" ];
then
    echo "External   : $ext_info"
else
    echo "External   : OFFLINE"
fi

unset ext_info

if [ -n "$en0_info" ];
then
    wifi_network=`/System/Library/PrivateFrameworks/Apple80211.framework/Versions/A/Resources/airport -I | awk -F: '/ SSID: / {print $2}' | sed -e 's/.*SSID: //'`


    echo "Wi-Fi      : $en0_info"
    echo "- SSID     :$wifi_network"

    unset wifi_network
else
    echo "Wi-Fi      : INACTIVE"
fi

unset en0_info

if [ -n "$en2_info" ];
then
    echo "Ethernet   : $en2_info"
else
    echo "Ethernet   : INACTIVE"
fi

unset en2_info

executed result is like this

$ ./networkmonitor.sh
Download   : 37.69 Mbps
Upload     : 54.78 Mbps

External   : XXX.XXX.XXX.XXX
Wi-Fi      : 192.168.1.19
- SSID     : XXXXXXXXXXXXXXX
Ethernet   : INACTIVE

Configure Geektool

Refresh for 120 seconds, time out for 60 seconds, set Font to Ricty and run this bash script, oh my goodness! On my desktop, at all times!

The network speed isn't very nice tho...