DEV Community: Claire Froelich

Free automated backups of Obsidian notes to GDrive for Mac, Windows and Linux

Claire Froelich — Mon, 18 Dec 2023 14:34:55 +0000

No one wants to lose all their years of note-taking to a spilled cup of coffee on the keyboard.

Obsidian has a robust paid Obsidian Sync proprietary service that enables two-way sync between devices both on and offline, version history and end-to-end encryption. If these features are important to you, I recommend this service.

However, if you just want simple free, automated backups of your local notes to GDrive you have options.

These solutions are scalable in that they only sync file changes and do not copy your entire vault every time.

Free backups for Mac/Windows

Luckily this one is easy. Google has a free Google Drive for Desktop client available for these operating systems that enables automated real-time, two-way syncing of local directories on your computer.

Once installed, follow these instructions to sync your vault folder to GDrive.

Free backups for Linux

Unfortunately there is no Google Drive for Desktop client for Linux today. Here is something I hacked together using FreeFileSync and a shell script run on a 5 minute cron job.

Note FreeFileSync does not encrypt your files. If that is important to you, check out this thread for some solutions.

1. Install FreeFileSync

This tool can be used 100% free for the purposes of this article.

Consider downloading the Donation version with some extra thank-you features if you appreciate their service.

Don't be fooled by the malwaresque look of their website. The service is great, community is friendly, and I haven't been frauded yet since making a donation.

2. Create a new sync configuration

Open FreeFileSync and create a "New" sync configuration from the upper left.

3. Configure source and destination directories

On the left, select the directory of your Obsidian vault.

On the right, click the cloud icon to configure a connection to your GDrive

When you click the cloud icon, a connection menu will appear. Authorize a connection to your GDrive, select which drive you will save to, and select a destination directory.

By default, it will sync to the root of your drive - this is messy! I recommend specifying a destination directory for your vault files, (you can create one if it doesn't exist yet). I named mine the same as my local vault, "claire-thinkpad"

Once set up, your configure screen should look something like below

FreeFileSync lets you set the sync mode to "mirror" (one-way, left to right), or other sync modes such as "two-way", which in theory would enable you to pull changes from GDrive and share work across multiple devices. I didn't try two-way sync yet - comment if you do and find it useful!

4. Save configuration as a batch job

Once you have your source and destination directories selected, click "Save as" in the upper left

Accept the defaults and click Save as

Save the .ffs_batch file inside your vault. For me that was ~/Documents/claire-thinkpad

5. Create the backup script

First create a shell script file that will execute the sync.

touch ~/cron/backup.sh

You can call it something other than backup.sh.

Paste the below script in the file:

#!/bin/sh

if ping -c 1 google.com >/dev/null 2>&1; then
  # Internet is up, run batch job via FreeFileSync
  env DISPLAY=:1.0 /usr/local/bin/FreeFileSync /home/claire/Documents/claire-thinkpad/SyncSettings.ffs_batch
else
  # Internet is down, do not run cron job
  exit 1
fi

Replace /home/claire/Documents/claire-thinkpad/SyncSettings.ffs_batch with the path to your .ffs_batch file.

The script checks whether you are connected to the internet by pinging google.com. If true, it will execute a headless run of your batch job via FreeFileSync.

I was saved here by Joe Lotz's post about running background jobs with FreeFileSync. The DISPLAY=:1.0 is the trick Joe found to run this headlessly, without opening the GUI.

Test your script

Run ~/cron/backup.sh and then check your GDrive destination folder and see that your files are there. Make sure your local vault is not empty first.

6. Create a cron job to execute your script

In your terminal, run crontab -e to enter the cron editor.

Add this cron job at the bottom of the file :

*/5 * * * * /home/claire/cron/backup.sh

Replace home/claire/cron/backup.sh with the path to your shell script from step 5

This will run your backup script every 5 minutes. Feel free to change to a desired interval, with the help of a cron editor tool or ChatGPT.

There you have it

Your changed files should now sync every 5 minutes, so long as you have internet connection.

Technically, instead of FreeFileSync you could code up your own diff calculator and Google Drive API calls to post changes directly to Google Drive. The benefit of using FreeFileSync is that it will calculate diffs for you out of the box and handle all the Google Drive API logic. Google APIs are not fun.

Quick HTML/CSS hack for prettier Japanese titles

Claire Froelich — Tue, 23 Nov 2021 18:43:37 +0000

As you read this article in English, look at the end of each line of text. You can thank the <space> character for cleanly separating words into unbreakable chunks that aren't interrupted by newlines. English: words trump newlines.

Japanese does not use spaces, in general. Humans that read Japanese just have a 'spidey sense' of where words start and stop based on grammar and context.

For large blocks of text like the body of articles, it is common for multi-character "words" to be split at newlines. This is OK: characters are mono-spaced and paragraphs keep an aesthetically pleasing blockiness. Japanese: newlines trump words.

From Asahi Shimbun (text in bottom section is read top-to-bottom, right-to-left).

Note above how each newline breaks in the same spot, regardless of position in the word. No hyphens necessary in Japanese!

While “newlines trump words” looks nice in large blocks of text, it looks awkward on lonely one-line text like titles.

For example, this text "Learn body parts":

<h2>体の部分を学ぶ</h2>

Default "PEZ dispenser" behavior

Without a space character or hyphen telling the browser where to delimit words, the default behavior is to just pop characters off one-by-one like PEZ candies in a PEZ dispenser.

This is the English equivalent of what's happening on resize:

Learn body part
s
Learn body par
ts
Learn body pa
rts
...

Can we agree this is ugly and unreadable?

To prevent PEZ dispensing and force compound characters to stick together in this Japanese title, you'll want to wrap the desired chunks in display: inline-block style.

HTML

<h2>
  <span class="w">体の</span>
  <span class="w">部分を</span>
  <span class="w">学ぶ</span>
</h2>

<!-- 

CSS style:

.w {
  display: inline-block;
}

-->

Not PEZ dispensing

Much better!

Automate it!

This solution is fine quick fix for the occasional eye-catching title, but requires knowledge of Japanese grammar to know where to put the <span> tags.

Fortunately there is a cool tool called Budou that automates this formatting. You don't even have to know Japanese - it uses AI to semantically parse text into words and insert the spans for you (it works for Chinese and Korean too!).

This has been the first of what will probably be many articles on "things I overlooked when translating my website into Japanese".

See the pretty Japanese line-wrapping in action on my children's book website: The Power of Vegetables

Deer illustrations in GIFs by Varvara Fomina

This article was originally posted on Medium

Automate daily Tweets with Postman and the Twitter API v2 🌞

Claire Froelich — Sun, 14 Nov 2021 03:23:39 +0000

Has your Twitter feed been blowing up with "gm's" lately?

// Detect dark theme var iframe = document.getElementById('tweet-1459628182810857481-72'); if (document.body.className.includes('dark-theme')) { iframe.src = "https://platform.twitter.com/embed/Tweet.html?id=1459628182810857481&theme=dark" }

If you're not in the know, this is crypto speak for "good morning", and is a customary way web3 and crypto nerds clock in for the day.

Do you always forget to gm? Do you not care but want to give the illusion you're hip and GMI? Use Postman monitors with the Twitter API v2 to automate posting your daily salute to Twitter.

1. Get a Twitter Developer account

On top of a normal Twitter account, you will need to set up a free Twitter developer account here.

Make a hobbyist account and explain that you will use your account to post your own tweets. Part of your on-boarding will be to "create a project".

2. Create an App

Once you have your developer account and project set up in your Twitter developer dashboard, you need to add an App to your project in order to use the API.

Give your new app a name

3. Get your secrets

Next you will be given some secrets. Treat these credentials like passwords and store them somewhere secure. The keys are hidden in this image, but you'll see something like this:

4. Give your app read and write permissions

You need to enable your app to post tweets on your behalf. Go to your app's "Settings" tab and click the button to edit the App permissions.

Update the app permissions from "Read" to "Read and write". Click save to apply the changes.

5. Generate additional credentials

You will also need to securely jot down an Access Token and Access Token Secret, which can be generated in your app's "Keys and tokens" tab.

Copy the generated token and token secret with your other secrets. These are the final credentials we need to authorize requests to the Twitter API.

6. Log in to Postman

Postman is a free tool for building, consuming, testing, documenting, and monitoring APIs. You can create a free account here.

Once you have an account you can access the tool from the browser without downloading anything, simply by going to https://postman.com

7. Create a personal workspace

To make API requests in Postman, first you need a workspace. You can create a workspace from the "Workspaces" dropdown.

IMPORTANT: The workspace visibility should be set to Personal, since we will be storing our secret keys in a way that would make them visible if the workspace were public.

8. Create a collection

In Postman, API requests live inside Collections.

Create a collection either with the "Create a collection" button, or the plus "+" icon in the collections tab. Name your collection Twitter API v2

9. Add a request to post Tweets

Now we can build a request to the Twitter API that will post a tweet on your behalf. First, "Add a request", and name it post a tweet

Using the Twitter API v2

We will use this endpoint in the Twitter API v2 to tweet:

POST https://api.twitter.com/2/tweets

You can read more about it here in the docs.

Paste this endpoint into the request URL in Postman, and set the request method to POST.

Add a body to your request

The Twitter API v2 expects a JSON body, where a text property is the text content of your tweet.

In the "Body" tab of your request in Postman, select "raw" > "JSON" as the datatype and paste this JSON data in the body:

{
  "text": "gm"
}

Authorize your request

You can't just go tweeting as anybody. In order to tweet you will need to authorize your request with your credentials.

This endpoint accepts OAuth 1.0 authorization. We can use Postman's "Authorization" tab on the request to set this up. Select OAuth 1.0 as the Authorization type.

First, make sure the request is set up to add the authorization data to the "Request Headers":

On the right, we will enter our Twitter app credentials: API Key, API Key Secret, Access Token, and Access Token Secret for the fields Consumer Key, Consumer Secret, Access Token, Token Secret, respectively. Leave everything else as the default.

NOTE: Usually you do not want to hard code sensitive data like this directly in the Authorization tab.

Normally you should use variables to obscure your secrets. Variables defined only in the "Initial Value" column are local to your Postman app and won't be shared.

However, in order to automate our tweets in a later step we need the secrets to be accessible by Postman servers in your workspace. Be sure not to make your workspace public or people could steal your API credentials and tweet abuse your Twitter.

Save all your changes and send the request. In Postman you should get a status 201 response from the Twitter API with the id of your new Tweet. Go check out Twitter!

10. Automate daily tweets in Postman

Postman allows you to create monitors that run your collections at specified intervals to alert you when your servers have caught on fire.

This feature can double as a way to schedule cron jobs for running API calls, which is exactly what we'll do to run our post a tweet request every morning.

Set up a Postman monitor

In the same personal workspace you made, click the "Monitors" tab on the far left menu, then "Create a Monitor".

Give your monitor a name. Select the collection you made and configure it to run every morning. Set it to a bright and early time to make yourself appear industrious ☕.

Scroll down and click the big orange "Create Monitor" button to set up the scheduled run. Give your monitor a test run by clicking the "Run" button manually - check your Twitter and make sure it works.

Congratulations! Sit back and let the robots take over.

BONUS: Add random emojis for humanizing effect

If you want to further hide the sham that you have a bot tweeting your gms, add a personalized random emoji to each tweet using scripts in Postman.

Add a pre-request script to get a random emoji

Head back to the Collections tab in your workspace and click on the post a tweet request.

Open the "Pre-request Scripts" tab on the request. Any Node.js code we write in here will be executed right before our request is sent off. Paste in this code, but use any custom emojis that suit your personality:

const emojis = ["🌞","💪","🍠","🍵"]

const emojiDeJour = emojis[Math.floor(Math.random() * emojis.length)]

pm.collectionVariables.set("emojiDeJour", emojiDeJour)

To explain, first we are defining an array of emojis. Then we select a random emoji and assign it to the local variable emojiDeJour. Finally we use Postman's pm helper object to programmatically set our emojiDeJour as a collection variable of the same name. This will allow us to access the emojiDeJour in the body of our request.

Add the emoji to our tweet in the request body

Finally, interpolate the emojiDeJour from the collection variables using Postman's {{variable}} syntax with double curly braces. Update the request body to:

{
    "text": "gm {{emojiDeJour}}"
}

Save your changes! Now each time request is sent you get a random emoji at the end of your tweet. I got sweet potato.

Conclusion

Postman monitors are a simple and free way to run scheduled API calls, making it easy to automate tweets using the Twitter API. Twitter has a fabulous API for interacting with the app, just take care your use case is not prohibited before making automations. Also note that the Twitter API is rate limited and you should make no more than 200 tweets via the API in a 15 minute window. But if you tweet that much you will have lost all your followers anyway.

You can see a completed example of the Postman workspace here (just make sure YOU make a personal workspace)

gn 🌚

Disclaimer: I work for Postman, but the thoughts and promotion of gm culture in this article are my own.