DEV Community: Dhravya

Supermemory - ChatGPT for your bookmarks

Dhravya — Mon, 15 Apr 2024 01:04:16 +0000

This is a submission for the Cloudflare AI Challenge.

What I Built

I built a way to build your own second brain, using Cloudflare Vectorize and Workers AI.

Here's the idea
Build your own second brain with supermemory. It's a ChatGPT for your bookmarks. Import tweets or save websites and content using the chrome extension (the extension on webstore is not updated, please use the one in the repo)

Well, here's the thing - me and @yxshv save a lot of content on the internet.

Twitter bookmarks, websites, snippets, etc.

But we never look back to it - to us, it's like throwing information in the void.

Supermemory fixes this.

Demo

Here's the deployed demo - https://supermemory.dhr.wtf

And here's the demo video https://www.loom.com/share/23b56648da3d4d8080b542a60cb9c8b6

My Code

supermemoryai / supermemory

Build your own second brain with supermemory. It's a ChatGPT for your bookmarks. Import tweets or save websites and content using the chrome extension.

SuperMemory

Interested in helping build the best second brain for everyone? Join the discord https://discord.gg/2X2XsKz5AU. Contributions welcome.

👀 What is this?

Build your own second brain with supermemory. It's a ChatGPT for your bookmarks. Import tweets or save websites and content using the Chrome extension

Well, here's the thing - me and @yxshv save a lot of content on the internet.

Twitter bookmarks, websites, snippets, etc.

But we never look back to it - to us, it's like throwing information in the void.

Supermemory fixes this.

Key Features

💡 Ideation: Capture and save ideas effortlessly.
🔖 Bookmarks: Import, organize, and resurface bookmarks when needed.
📇 Contacts: Store and manage information about people you know.
🐦 Twitter Bookmarks: Import and utilize your saved tweets.
🔍 Powerful Search: Quickly find any saved information.
💬 Chat with Collections: Interact with specific knowledge bases.
🖼️ Knowledge Canvas…

View on GitHub

Journey

It all started when I saw this company in YC batch 2024 (https://zenfetch.com) - I was determined that I can make something like that, even better, and open source it.

So, over the weekend, I built the version 0 of supermemory (and i called it anycontext) - it wasn't great. Very buggy and far from useful. But the bottom line was - I built a competitor to zenfetch in just a weekend.

Then, one day, I opened dev.to and found out about the cloudflare dev challenge! Hell yes! I've been a big fan of cloudflare and winning this would be amazing. So, I reached out to @yxsh for his help. Yash is an amazing design engineer and was able to completely redesign and revamp the user interface in just about 15 days, even though he's fully busy with classes and studies!!!

We also revamped the entire database, vector store, and even the AI generation to make it as good as possible. Added support for notes and spaces. Added import twitter bookmarks option.

As we did, the architecture became more and more complex, and seeing everything work so perfectly was just such an amazing feeling.

Now, Supermemory is the true second brain for both of us. Saving and forgetting content is a thing of the past.

So, are you ready to supercharge your memory? Go to https://supermemory.dhr.wtf and try it out for yourself!

Teammates:
@dhravya, @yxsh

Need some advice!

Dhravya — Thu, 02 Feb 2023 06:52:48 +0000

Twitter just announced that there will no longer be free access to the twitter API. (https://twitter.com/TwitterDev/status/1621026986784337922)

Hi! I'm Dhravya - a 17 year old maker. I have a small twitter tool - https://twitter.com/poet_this, which includes the website https://tweets.beauty and the API https://api.tweets.beauty

I got some donations at the start which helped me pay for the hosting costs, but that's been it.

And the main problems start here.
I'm a student and indie developer, and I have crucial examinations coming up in just 20 days. I wanted to monetise the bot after exams to create a revenue stream for myself, but the timing is very, very bad.

So, what do I do?

Loss for a couple of months and monetise later
Sell the project + twitter acc and start with a new side project right after the exams
Ask for donations and keep the project free for as long as I can
Start selling the entire codebase to multiple people (on license basis? not sure.)

If I choose to sell the project, where do I sell a side project? How do I get buyers?

Here's why I kept my project free:

Proof of work.
I wanted the maximum number of users so that I could have a boost for later projects
It wasn't expensive to host or difficult to make, and I don't think I could provide enough value.

What do you think? Any advice would be greatly appreciated!

Deploy your functions and use them everywhere⚡⚡

Dhravya — Fri, 20 Jan 2023 14:54:37 +0000

A beginners tutorial written in a funny way - You'll learn how to get started with cloudflare workers, and IOS shortcuts (and hopefully also get a laugh!)

I had a small API that me and my friends wrote in Python FastAPI - I hosted that API on my server. But the entire API was kinda useless so I shut it down.

I did the unthinkable. I forgot that one of my favorite functions, owofy.py was part of that project. After almost one year, I began on a quest to revive it.

It was a silly function that basically creates a funny "owofied" piece of text.
Oh... Hai thewe (。O⁄ ⁄ω⁄ ⁄ O。)It was a siwly function that basically creates a funny "owofied" piece of text. UwU
uhhhh... yeah, you get it.

I NEEDED to revive it. At all costs. I needed that in my phone, laptop everywhere. I had to harness the power of uwufying. It was my path to enlightenment. Without the owofy API, I would remain plebeian for eternity.

So, let's do this! Let's create a typescript function that uwufies text, set up cloudflare workers and create an IOS shortcut so that we can uwufy text everywhere 😱😱

If you don't know what a cloudflare worker is, or what IOS shortcuts are, this blog isn't...
jk. Don't worry. I'll explain it all.

Owofying text

Let's start with creating the owofy function itself. We need to take text, and use start with a dumb "Haiii OwO"-like phrase. add suffixes to words, replace "why" with "wai" and stuff. OK I won't bore you with the function itself, it's pretty simple.

I totally didn't take the help of ChatGPT and github copilot because I was having an issue with typescript and was trying to be extra cool solving it with generics.
totally not.

Anyways! Here's the TS code

WTF is a cloudflare worker?

To understand cloudflare workers, we need to understand what "serverless" is.

In short, serverless is a way to deploy code without having to manage a server.

OH- I hear that. You need an analogy.
You own a kitchen because you really like pizzas. You purchased the land (server), you hired a chef - or learnt how to make food yourself (system admin), and you are also the one who mops the floors and maintain the kitchen, keep it neat and clean. Your kitchen could burn down btw.

or, you can just order food from a food delivery app. Give it your exact recipe, pay for the food and delivery. And you get your pizza without any hassle. The difference is, ordering is much cheaper because the same network of kitchens provides pizza to a lot of other people. They have the best chefs, the fastest ovens, the best delivery network. Even if one of their kitchens burn down, all the others will stay active.

Got it? Cloudflare is, infact, the biggest "pizza provider" in the world.
Or in other words, Cloudflare has it's servers EVERYWHERE. You can't really move your own pizza kitchen, or make thousands of kitchens all around the world.

That's the power of serverless. Amazing performance, reliability, speed and scale, while also being convenient to use.
Does that mean serverless has no servers? Nope! Obviously, the servers do exist somewhere, they are just abstracted away and we don't need to deal with all that.

OK, NOW GETTING TO THE PROGRAMMING AGAIN.

Hey, this is very important - extremely important for our path to enlightenment. Yes, you have to join me. Together, we can uwufy the entire internet.
CODE ALONG!! Don't just read, you won't get anything.

Let get stuff done first:
First, Create a Cloudflare account

Ok, now, we have to do some installations (you could just do everything on the cloudflare dashboard but that would be cringe)

First, install nodejs if you don't have it yet (We would need NPM to download wrangler)

What is wrangler? It's basically a tool that cloudflare has made to make our lives easier. It helps us make cloudflare workers.

Let's install wrangler (run this in your terminal) -

npm install -g wrangler

(the -g flag is the global flag, it means that wrangler will be available globally, and you can run it with the wrangler command)

Now, log into wrangler with your cloudflare account.

wrangler login

Creating the worker

Congratulations! We have taken the first step to enlightenment. The uwufication of the universe. Now let's start a project.

Use the command wrangler init owofy-the-world
It will ask you some questions. It's your choice if you want to use typescript!!! In this project, I have used typescript because js isn't my cup of tea ☕

BTW - If you want to code in python or some other language, check this out - Languages supported by workers
I'll be completely honest, the python one seems completely broken. But other than that, all languages work!

This will create a folder named owofy-the-world. Open your editor in that folder

You should have a file called src/index.js (or .ts), and a file called wrangler.toml.

Let's understand!

// Path: src/index.ts
export default {
  async fetch(request: Request, env: any, ctx: ExecutionContext) {
    return new Response("Hello World!");
  },
};

Over here, the function is exporting a fetch() function - Which has

request - so we can access stuff about the request itself things like URL, HTTP headers, HTTP method, and stuff.
env - Any variables we might need (for things like KV database)
ctx - The context of the execution (don't worry about this for now)

First of all, to owofy, we need the params.
Let's say, I pass the text like this:
website.com/?text=Here's the text to be owofied
How do we get the stuff after text= 🤔

Wait - it's part of the URL! Let's just parse the URL and get the text out of it!

COOL! You can take any number of params, any param in general. You can even have authorization with headers, run your own functions and return whatever you want to!

But we are on a mission, soldier. We NEED to owofy the text! So, we'll just put our function there...

function owofy(text:string) {...}

export default {
  async fetch(request: Request, env: any, ctx: ExecutionContext) {
    const text = new URL(request.url).searchParams.get('text');
        return new Response(owofy(text || "You betrayed me. no text!"));
  },
};

That is IT! YES! It's that easy to create a cloudflare function. Now, imagine all the things you could do with it. One of my friends, Mohit, wrote a really good blog post on one such use case -

How to serve any website from your own domain (for free!)

Mohit Yadav ・ Aug 13 '22

#javascript #webdev #beginners #tutorial

Cloudflare also has a lot of examples that you should check out!!!

Deploying the worker

Just use the command wrangler publish. Yes. that's literally it. That's right - You don't need to anything.
Now you can sit peacefully, and owofy all the text you can in the world.

This will be published on a yourname.workers.dev URL - ughh, that's ugly! To change the world, we need to put it on owo.customdomain.com!
You can add a custom domain, custom routes and a lot more in the cloudflare dashboard workers page.

WTF is an IOS shortcut?

I know not everyone uses IOS so I'll keep this short!
IOS shortcuts are automations that you can run on your mac or iphone. Creating shortcuts is probably the most easiest things to do now - and it's really fun!

"Hey siri, make this cringe"

Use text from clipboard
Get contents of URL yourname.workers.dev/?text={ClipboardText}
Copy content to clipboard.

That's it. Here's the shortcut I made - https://www.icloud.com/shortcuts/48917df04e4c4838864d5a29b698c2cf
Feel free to use it. Copy any text and say "hey siri, make this cringe"

That's it!!!

If you liked this post, you'll like other tutorials of mine - follow me @dhravya

I post stuff on twitter @DhravyaShah too, if you're into that.

Every time I learn about algorithms, I either forget it the other day or don't understand anything at all.
Luckily, Algolab has a very good, video-based, interactive course that will get you from 0 to interview in no time!

Use my referral link (it supports me too) - https://algolab.so/p/algorithms-and-data-structure-video-course?affcode=1413380_pkegf5oy

Creating a Twitter bot in Python using Twitter APIv2 🚀🚀

Dhravya — Thu, 06 Oct 2022 19:57:52 +0000

When I was migrating @poet_this twitter bot to use the Twitter API v2, I wasn't able to find any good tutorials that do it, atleast none that worked properly. After the bot's success, many people reached out to me asking how they can also create a twitter by themselves! I'm here to help :)

🤔 What we're building

Let's create a twitter bot that replies whenever someone mentions the bot. To keep this simple, we'll just reply with the same text but capitalized.

What we wanna do:
(original_uploader): This is a cool tweet
(commenter): @ bot (*mentions the bot)
(bot): THIS IS A COOL TWEET

At the end of this tutorial, I'll also include ways to add more cool features to this, and some ideas for your own cool twitter bot!

⚡ Getting started

First of all, you'll need a Twitter developer account. Sign up for one on Twitter Developer dashboard

Here's a step-by-step guide by Twitter

Alright, once we're done with that...

Let's create a new python project!

Setting up the project

I recommend working in a virtual environment. Create a new folder, and open the terminal and use this command

python -m venv env

env/Scripts/activate // on windows
source env/bin/activate // on linux

Finally we just need to install the dependencies,

pip install python-dotenv python-twitter-v2

That is the only two dependencies we really need right now. python-dotenv to use environment variables, and python-twitter-v2 to interact with the twitter API!

🔐 Authenticating with Twitter

To authenticate with the twitter API with the bot's twitter account, we'll need a few things...

Consumer key (From developer dashboard)
Consumer secret (From developer dashboard)
Access token
Access token secret

Technically you can also use a bearer token, but let's keep it simple for now.

To get access token and secret (for testing), you can create them in the dashboard here:

Make sure that you have the Read and Write permissions, we need that for the bot.

(If you want to control another account, for example, a bot account, follow these steps to get the access token/secret for another account - https://developer.twitter.com/en/docs/authentication/oauth-1-0a/obtaining-user-access-tokens)

Once we have all these tokens, create a file, .env and fill it in, in this fashion -

CONSUMER_KEY=...
CONSUMER_SECRET=...
ACCESS_TOKEN=...
ACCESS_TOKEN_SECRET=...

👨🏻‍💻 Let's start with the code now!

Let's break the problem into little pieces...

We want to,

Authenticate with Twitter API
Listen for all mentions
Reply!

Cool, first, let's authenticate

import pytwitter
from os import environ as env
from dotenv import load_dotenv

load_dotenv() # Loads the .env file we created earlier

api = pytwitter.Api(
    consumer_key=env["CONSUMER_KEY"],
    consumer_secret=env["CONSUMER_SECRET"],
    access_token=env["OAUTH_TOKEN"],
    access_secret=env["OAUTH_TOKEN_SECRET"],
)

You can check if the bot is properly authenticated by using some endpoint, for example,

print(api.get_user(username="Twitter"))
# Should print Response(data=User(id='783214', name='Twitter', username='Twitter'))

Now let's listen for mentions.
There's one thing to keep in mind though - we don't want to reply to ALL mentions. This is because how Twitter works.

Let's say, I make a tweet and someone comments on it
@DhravyaShah: Hello world!
@SomeOtherRandomUser: Hi Dhravya!

But, instead of just being "Hi Dhravya", it looks like this-
@DhravyaShah: Hello world!
@SomeOtherRandomUser: @DhravyaShah Hi Dhravya!

Notice the difference? This can be a problem in a twitter bot, where it will reply even to the comments under the bot.
We'll fix this by checking if the text is ONLY mentioning the bot, and nothing else.

How? Just remove all the mentions and check for an empty string!


last_tweet_id = 0 # Just a variable to keep track of the last tweet we replied to.

if __name__ == "__main__":

    while True:
        try:
            mentions = api.get_mentions(
                user_id=str(BOT_ID), # get bot id from tweeterid.com
                return_json=True,
                since_id=str(last_tweet_id),
                tweet_fields=[
                    "conversation_id",
                    "entities"
                ]
            )


            if not isinstance(mentions, dict):
                continue

Cool, now we get a constant stream of mentions, well not really. We get all the mentions after since_id.

These mentions actually contain a LOT of data. I wanna keep it short so I won't get into the details, but feel free to print() this once and look through all the fields provided!

Now let's check if there's any data and it has the includes too (includes has all the extra entities)


    if not "data" in mentions.keys():
        continue # There's no more tweets

    for tweet in mentions["data"]:
        text = tweet["text"]
        reply_to = tweet["id"]

        # If it's not a reply to another tweet
        if not (tweet["conversation_id"] == tweet["id"]):
            if str(tweet["in_reply_to_user_id"]) == str(BOT_ID):
                continue

            # Get the parent tweet
            tweet_ = api.get_tweet(return_json=True,tweet_id=tweet["referenced_tweets"][0]["id"])
            text = tweet_["text"]


        # If tweet is a reply, it's in the format "@user @user @bot"
        users = [
                  mention["username"] for mention in tweet["entities"]["mentions"]
        ]
        new_txt = tweet["text"]
        # Remove all mentions in the start
        for user in users:
            new_txt = new_txt.replace(f"@{user}", "")

        new_txt = new_txt.strip()

If new_txt is "" (empty), that means it's a "direct mention" - just a "@bot" comment. If it's a direct mention, the bot has been explicitly called, which means we can respond.


if (new_txt == ""):
    api.create_tweet(text=text.upper(), reply_in_reply_to_tweet_id = reply_to

And finally, let's close the loop and also update the last_tweet_id to the new, last_tweet_id


            if "meta" in mentions.keys():
                if "newest_id" in mentions["meta"].keys():
                    last_tweet_id = mentions["meta"]["newest_id"]

    time.sleep(10)

That's basically it! Here's the full code in a Github gist

You can add more features to the bot my adding "Keywords", just like how @poet_this works on twitter - The comment "@poet_this blue 2" (blue 2 being the keyword) generates that particular style of image.

Finally...

This was just a quick rundown, but now...
Try making a bot yourself! It's a very fun and rewarding process, and you'll learn a lot through it.

Thanks for reading!
You can support me by sponsoring me on Github - https://github.com/sponsors/Dhravya
You can follow me on Twitter - https://twitter.com/dhravyashah

Disco.pics - Free and open source image hosting service

Dhravya — Sun, 28 Aug 2022 22:50:00 +0000

Overview of My Submission

I'm a huge fan of discord, and use it to communicate with my friends, family and teammates.

One day, a friend of mine told me how he has made a channel just to store and organize his images at one place.
All attachments uploaded to discord have a link automatically generated. These links are shareable, but they are really long.
Discord has a very high-speed content delivery network (CDN) powered by Cloudflare.

All of this made us think, "Why not create an image hosting that uses discord's CDN?"

And that's exactly what we did!
But this needed to be BLAZINGLY fast. We wanted to make a good image hosting, with custom embed support, an API, etc.

This would mean, we have to fetch from the database every time someone loads an image. Redis was the perfect choice for us.

Submission Category:

Wacky Wildcards!

Overview video

Language Used

Typescript

Link to Code

discopics / disco.pics

An image hosting service with all the customisation you need

This project has been discontinued. Use at your own risk

Disco.pics

Fast image hosting made easy.

Overview video

Here's a short video that explains the project and how it uses Redis:

[Insert your own video here, and remove the one below]

How it works

How the data is stored

There are two different schemas, the User and Image schema.

For each user, this is the information that is stored:

id: string
email: string
created_at: date
token_number: number
embed_title: string
embed_site_name: string
embed_site_url: string
embed_colour: string
embed_author_name: string
embed_desc: string

here, the ID is indexed, so we can find users by their ID.

The Image schema is the information that is stored for each image:

uploaded_by: string
slug: string
img_url: string
uploaded_at: date

Here, slug and uploaded_by are indexed.

Whenever someone visits the /[slug] route, the Image schema is queried for the image with the slug.

Then, using the uploaded_by…

View on GitHub

Additional Resources / Info

Collaborators

@yxsh

Check out Redis OM, client libraries for working with Redis as a multi-model database.
Use RedisInsight to visualize your data in Redis.
Sign up for a free Redis database.

How to ask for help

Dhravya — Sun, 17 Jul 2022 09:09:25 +0000

This blog was originally published on blog.dhravya.dev

Every couple of days, a beginner comes to our programming server, The Coding horizon, and asks for help - That's perfectly OK! Asking for help is natural and is almost necessary for learning. However, you need to HELP OTHERS to help you.

So here's some points to keep in mind, I'll try to keep it short and simple. You might like my blog about Learning to code, for beginners if you enjoyed this one.

What not to do

Let's start off with some things that you absolutely should not do.

Asking people to join a voice call

Many people don't know how to frame their questions very well, so they ask others to join a voice call. Please keep in mind that people that are willing to help you have a life, and they might not be in the situation to join a voice call, or even if they are, they might not be comfortable to - which is mostly the case.

Framing your questions correctly is all you need to do

"My code doesn't work"

Never ask for help with "just doesn't work". It doesn't help anyone in this situation - elaborate on what exactly doesn't work and why it's not working.

Pinging random people, giving bribes, and other things

Leaving the server because you didn't get help immidiately

This is kinda insane - People join a server, ask for help - don't frame their questions properly and then (since no one understood the question in the first place), they leave the server.

Joining a server just to ask for help isn't a problem here - even I did it as a beginner, and that's how I found amazing people and amazing discord servers. The problem is, leaving the server just because no one knew the answer to your question.

Ok, let's get to bribes - just yesterday, this happened, this checks out all points of what not to do so far-

As you can see they also pinged a random person without any context.

Shortly after asking this, they left the server, and I didn't get the chance to help them.

What to do instead:

Don't bribe lol
Frame your question correctly

Sending screenshots of your code

Or worse - sending photos of your laptop screen using your phone...

This happens an awful lot - and it really makes it difficult to get help. Some people just send screenshots, without any context as to what is going on, what's the issue, and what they have tried (More on that later).

What to do instead:

JUST COPY PASTE CTRL + C and CTRL + V

Make sure the code is properly embed, inside triple back ticks (code block)

What you SHOULD do

Google before asking

There's a good chance that the problem you're having has already been solved by someone on the internet, you should try to google it first.

Don't ask to ask, just ask

Ditch the "Can someone please help me" - Get to the point, ask your question right away.

Read: Don't ask to ask.

Frame your questions correctly

Ok this is very important - Framing your questions correctly will make it easier for others to understand your question, and you'll be much more likely to get an answer.

Here's how to do it.

So every question should have these points:

Language/Framework/Library you're using
What you're trying to achieve
A brief description of the problem
A relevant code snippet
In case of a visual problem, a screenshot of the problem (For example, web development-related questions)
Stack trace, if any
Solutions you have already tried

So here's sort of a template you can use for your questions:

Hey there, I need some help with a <language/framework/library/tool> problem.

I'm trying to <what you're trying to achieve>. But, <A brief description of the problem>.

Here's my code:

<Proper, embedded code :embed using triple back ticks,

your code

upload to pastebin if it's too long>

I've added the screenshots <attach the screenshot>

Here's the error I'm getting:


I've already tried the following solutions:
<Solutions you have already tried>

I can almost guarantee that you will get help from someone if you ask in such a format.

Be respectful even if you don't get help.

Sometimes, even after all this, your question might be ignored - or maybe no one could help you. That's fine! Please be respectful, even if you don't get help. Don't leave the server, or community, don't be rude.

TL;DR

Here's a TL;DR of how to ask for help.

"Code doesn't work" - Elaborate
"Can someone please help me" - Get to the point
Don't ask people to join a voice call, don't ping random people without context, don't give bribes
Don't send screenshots of your code - Just copy paste CTRL + C and CTRL + V
Google before asking
Frame your questions correctly
Be respectful (even if you don't get help)

That's it for today! I hope I provided some value with this blog. I've not been writing since like a month - I'll try to be more consistent again.

A whole new developer ecosystem: CakeCutter 🍰

Dhravya — Fri, 17 Jun 2022 04:36:43 +0000

A couple months ago, I wrote a blog Introducing Cakecutter, which was, essentially, a modular version for create-python-project. Both of these projects were very easy, because they were just simple projects for me to practice the Rust programming language. I made them and forgot about it. But little did I know, small projects go a long way.

Introducing Cakecutter🍰- Start projects quickly from Cakes (templates)

Dhravya ・ Mar 9 '22

#showdev #rust #productivity #devops

But then, a couple days ago, Yash (who is a fellow admin of the Coding Horizon community and an amazing developer), texted me, saying that he sees a lot of potential in my little pet project.

“Cakecutter is cool, let’s make it better”

And that’s exactly what we did.

We discussed new things we can implement, and decided to rewrite the entire thing in Golang

Ok, but what exactly is Cakecutter?

In a nutshell, Cakes are templates, which contain everything you need to set up a project. This includes:

The file structure of a project (for eg. some boilerplate files, README, src/ folder) [DOCS]
File content (for eg. Pre-generated licenses, or boilerplate code) [DOCS]
Commands to set up a project (for eg. Downloading dependencies, creating virtual environment) [DOCS]

Additionally, questions can be asked when setting up a project. Questions can be a selection between multiple things, or an input. Then, the answers to the questions can be used to fill in a template. [DOCS]

An example use case for this would be, asking the name of a user to auto-generate license, or you can even generate an entire README file with it!!

You can create files based on answer to a question, for example, creating a file only if the project is a typescript project.

And, you can even run commands based on answers. Example use case would be asking host OS to run setup commands!

This is a hell lot of power and functionality in just ONE CakeFile! (Just like real cakes 😆)

Have a look at the Basic example cake and you can also try out the Python cake

Sold yet? Install cakecutter.

Head over to https://docs.cakes.run/1-install/, or use with NPX:

npx cc

These are not even the biggest updates.

Like create-react-app , most people don’t like to make cakes. They like to eat them. We knew this, and a way to share cakes had to be made.

So, we created a publish command, so you can publish your cakes for others to use!

Try out this command to instantly generate a very simple python project template. Now you don’t need to create cakes!

npx cc cut python

Also, we created the website Cakes.run where you can find the documentation, installation guidelines, and also browse the cakes created by others.

To learn more about how to create cakes, Read the docs, and go through the examples repository

Your support is appreciated. Here’s how you can help

The entire stack - from the API, website, CLI, Docs - Everything is open source. You can find them on the Cake-cutter github organisation.

⭐ Star all repositories to spread the word. Especially this one, the main CLI 👇🏻

cake-cutter / cakecutter

Open source, cross-platform developer tool that helps you quick-start your project. Creates files, runs commands, templating based on question. Also works as a setup tool

Cakecutter

Create projects from pre-built cakes (templates) in seconds!

Read the full Documentation

👀 What is Cakecutter?

Sometimes, the most difficult thing is to just get started with a project. Cakecutter is a tool that helps you to cut the cake and start your amazing project instantly.

What Cakecutter does:

Users can publish, create or use a cake from Cakes.run. Cakes are basically TOML files which contain all the information needed to create a project.
According to the information in the Cakefile, Cakecutter will create all the files and (you can also fill them with content) in the correct location.
Setup commands (installing dependencies, etc.) can be defined in the Cakefile. These commands are run after the files are generated.
Cakecutter can ask questions to the user and take input. The input can then be used as variables for the project template. Read the docs here
…

View on GitHub

Follow Yash, the major contributor of this project, on DEV community and Github - https://github.com/kekda-py

Follow me, Dhravya Shah on Github - https://github.com/dhravya

How I (organically) grew a Twitter bot account from 0 to 750 followers in 20 days.

Dhravya — Fri, 20 May 2022 10:34:24 +0000

On 28th April, I published this blog, challenging myself to grow a twitter bot account to 1000 followers in 100 days. Well, I HUGELY underestimated the possibilities. I managed to get to 750 followers in just 21 days!

Can a twitter bot get 1,000 followers in 100 days? A challenge to myself and this community

Dhravya ・ Apr 28 ・ 1 min read

#discuss #watercooler

Beautify This! is a simple twitter bot that takes beautiful screenshots of tweets. To use it, all I had to do is mention @poet_this screenshot under any tweet and it will reply.

Earlier, the bot was powered by https://poet.so. And it was a selenium script that was really, really inefficient and super costly to host. So, I created my own custom image generator to do the same.

Since I did it using Python Image Library, it was quite challenging and I had to come up with interesting algorithms to, for example, vertically extend the tweet template based on the number of lines of the tweet itself. Now, I had reasonable hosting costs and a good enough script to be made into a twitter bot.

But why?

I've been growing my own twitter account since almost a year now (created it 3 years ago). In all these 3 years, I got 150 amazing followers, friends and stories to tell.
But well, you know, that's kinda a small number, isn't it? It is extremely difficult to grow a twitter account ... or is it? I thought.

I don't know. I just wanted a challenge where I manage to do something I thought was impossible.

Jumping in, I had no idea how to do this. I mean, I knew that more the people use it, the more publicity it will get. But how to spread the word?

This was also the time when I was super, super burnt out. I didn't have any motivation to code and I really had nothing to do (I don't watch movies, TV shows, I don't go out with friends).

How?

Here's a timeline of how it went:
Day 1 to 5 : 30 followers. It was just me using the bot everywhere on twitter. People noticed, people followed. But I was behind my goal.

The growth started to drastically slow down, I almost gave up.

But on day 5, I used the bot on an Elon Musk's Tweet. Luckily, it was literally 10 seconds after he posted the tweet (i genuinely wanted a screenshot to share to friends lol)
And HOLY SHIT! 5 followers, 10 followers, by the end of the day, I already had 70 followers! I had a trick now. Just turn on the notifications for Elon musk and (lmao) use the bot like anyone else would. This was amazing to spread visibility.

The bot got many, many loyal users now. Users who would use the bot on every one of their posts. by the 10th day, I already had about 150 followers.

I was now confident that I can probably hit the mark in 30 days. So I changed the about me of the bot to 30 (growth started slowing down again and I though 50 is a more realistic number) at this point.

by the 15th day, I had about 300 followers. Now, I don't know what the hell happened, maybe someone else used the bot on a very popular tweet, but people started supporting the challenge a LOT. I got emails, DMs, and what not - soon enough, I woke up to 400 followers (on about the 17th day). Since then, every single day, the bot has been getting (almost) 100 followers.

Today is the 21th day. It just hit 750 followers, and almost gonna hit 800! Help me get to 1000 followers in ... 30 DAYS!

twitter.com

Challenges

I also faced a ton of challenges. Like, the high usage meant that hosting costs also went up. Since the bot is image generation, a lot of image generation, I have to shut it down for some time to save costs. Whenever I'm on the laptop watching videos or relaxing, I turned on the notifications for twitter and listened to pings. And then used the IMGEN script on my own computer and manually sent the result image. Yeah...

Whenever someone used the bot in a wrong way, I manually send them messages on the correct usage, or sometimes just sent the screenshot manually.

Twitter flagged and muted many of the bot tweets, even though it's a "good" bot and has the automated tag. But it's understandable.

How you can help.

First of all, give a follow! https://twitter.com/poet_this. this will help it reach 1000 followers in less than 30 days.

You can also financially support me, so I don't have to ask my parents to pay for the hosting https://ko-fi.com/dhravya

Also, the bot is open source (The API, however, isn't open source yet. Still working on the API where the actual image generation happens and it has a lot of bugs)

Leave a ⭐

Dhravya / beautify-this-bot

A twitter bot that simply replies with a beautiful screenshot of the tweet, powered by beautify.dhravya.dev

Poet this!

Replies with a beautiful screenshot of the tweet, powered by beautify.dhravya.dev

Installation

git clone https://github.com/dhravya/beautify-this-bot.git

cd beautify-this-bot
pip install -r requirements.txt

Usage

python src/main.py

License

This project is licensed under the MIT license

Show your support

Leave a ⭐ if you like this project

Readme made with 💖 using README Generator by Dhravya Shah

View on GitHub

I was tired of making banners for my blogs, so I automated it

Dhravya — Sun, 15 May 2022 05:19:21 +0000

I'm a very lazy person, and always procrastinate on the smallest of things - opening the image editor, and making a good blog banner would always be the "extra" thing I had to do every time I wrote something. So, I made a script that generates them for me in milliseconds!

As Bill Gates says,

always choose a lazy person to do a difficult job because he will find an easy way to do it

The output is fully customizable, as it is 100% generated from code. You can change some values here and there, and boom! Generated. I ran it through all the posts of my blog and it quickly replaced many of the blog posts for me.

Check out the code Here (And ⭐ the post if you liked it!)

Dhravya / blog-banner-generator

Customisable, fast, easy to use blog banner generator

Blog banner generator

I was tired of making banners for my blog posts, so I created a script that does it for me.

Installation

git clone https://github.com/Dhravya/blog-banner-generator.git
cd blog-banner-generator
pip install -r requirements.txt

Config

The default config works well, but if you want your banner to be a different colour, or change the positioning of elements, feel free to edit the config file

Here are the config options (note that this may break the script, because I haven't tested it out much)

BG_COLOR : rgb(r, g, b)

ART_POSITION : Coordinates(x, y, size=(430, 430))

TITLE_POSITION : Coordinates(x, y)

IMG_POSITION : Coordinates(x, y)

FOOTER_POSITION : Coordinates(x, y)

FOOTER : "<your small footer>"

Running the script

You can run the script by using the generate() method of the ImageFactory class

if __name__ == "__main__":
    generator = ImageFactory()
    generator.generate(
        ... # Options here
    )

Generate options:

title, description…

View on GitHub

Introducing Audium - A "Micro-Podcast" social web app

Dhravya — Wed, 11 May 2022 15:02:30 +0000

Overview of My Submission

auDium

A simple, barebones "Micro-Podcast" web app

This was made as a submission for the Appwrite x Dev Hackathon

👀 Inspiration

Short content is blowing up nowadays, with the rise of social media that provide short-form content and popular websites making a shift from long to short, it's time to make short podcasts.

This was inspired by Twitter - which is considered a "micro-blogging" platform. It's a simple way to share your thoughts, ideas, and stories.

auDium is a simple web app that does exactly that, but with audio. It's like spotify + twitter.

Submission Category: Web2 Wizards

Link to Code

Dhravya / audium-app

A simple, barebones "Micro-Podcast" social web app

auDium

A simple, barebones "Micro-Podcast" web app

This was made as a submission for the Appwrite x Dev Hackathon

👀 Inspiration

Short content is blowing up nowadays, with the rise of social media that provide short-form content and popular websites making a shift from long to short, it's time to make short podcasts.

This was inspired by Twitter - which is considered a "micro-blogging" platform. It's a simple way to share your thoughts, ideas, and stories.

auDium is a simple web app that does exactly that, but with audio. It's like spotify + twitter.

📱 Screenshots

⚙️ Built with

🔘 Where Appwrite helps:

Database : All the posts are stored in an appwrite database instance. Appwrite makes databases very easy with it's python SDK
Storage : All the audio files are stored in the appwrite storage instance.

Authentication, however, is provided by Auth0, which provides really powerful and easy…

View on GitHub

Additional Resources / Info

For the purpose of this hackathon, the app is very barebones. It's not meant to be used as a full-fledged social media yet, but I'll be constantly working on new features and improvements.

So, for now, you can consider it as a proof of concept that works

Also, the app is not hosted because I don't have the funds to host Appwrite on my own, and this projects needs quite a bit of storage too, which I don't have. So, for now, you can only use it in the development environment, on your local machine.

📱 Screenshots

Create the fastest search for your website in minutes, without any dependencies ⚡🔎

Dhravya — Tue, 10 May 2022 06:03:10 +0000

The project I'm working on is written in Gatsby JS, but the solution itself is vanilla react and will work everywhere.

Jump to main content

The project I'm working on is written in Gatsby JS, but the solution itself is vanilla react and will work everywhere.

Today, I spent most of my time updating my blog, and thought to add more features like search, tags, MDX support, and a few design changes, including the sidebar.

I was deciding how I would implement the search function, because the only time I have done it, was using a
Self hosted version of Typesense

But well, that was costly to host server-side, especially for
something as simple as a blog. and their hosted solutions aren't that great price-wise either.

So one thing was sure, there is no need to use any API for this. After a quick google search, I came across this documentation on Gatsby's website which is about adding search to Gatsby

From that guide, under the Client Side section, here's what they recommend:

It is possible to do all the work in your Gatsby site without needing a third-party solution. This involves writing a bit of code, but using less services. With large amounts of content to index, it can also increase the bundle size significantly.

One way of doing this is to use the js-search library:

Adding Search with JS Search

There are two Gatsby plugins that support this as well:

gatsby-plugin-elasticlunr-search
gatsby-plugin-local-search

Now these search methods index everything which means higher bundle size. And they are also a hassle to set up.

The solution I went with

Now for my use case, it was probably a good idea to just make something simple by myself, and I can build on it as I keep updating this blog.

The idea is really simple, I just need to make a search box, and on every keystroke, loop through the contents and filter them like that.

const BlogIndex = ({ data, location }) => {
  // These posts can be anything,
  // I've just used the posts from a gatsby query
    const posts = data.allMdx.edges;

  // We need to filter the posts by the search query.
  // by default, we have all posts
    const [filteredPosts, setFilteredPosts] = useState(posts);

  // This will be the search query
  const [search, setSearch] = useState('');

  return (
    <div>
      {/* Our search bar */}
      <input
        type="text"
        placeholder="Search"
        onChange={(e) => {
          e.preventDefault();
          setSearch(e.target.value)}
      }/>

      {/* Simply mapping through everything and rendering blogs */}
      {filteredPosts.map(({ node }) => {
        <BlogPost post={node} key={node.id} />
      }}
    </div>
  )
}

Now, whenever something is typed in the box, the search state will be updated. Now, let's write a useEffect hook to update the filteredPosts state whenever the search state changes.

const BlogIndex = ({ data, location }) => {

    const posts = data.allMdx.edges;
    const [filteredPosts, setFilteredPosts] = useState(posts);
  const [search, setSearch] = useState('');

  //highlight-start
  useEffect(() => {
    if (search) {
      // post filtering here
    }
  }
  // only update the filteredPosts state when the search state changes or the posts state changes
  , [search, posts]);
  ///highlight-end

  return (
    ... // rest of the code
  )

And now let's write some very simple code to filter the posts.

...

  if (search) {
    const filteredPosts = posts.filter(post => {
        //highlight-start
        const title = post.title.toLowerCase();
        const description = post.description.toLowerCase();
        return title.match(search.toLowerCase()) || description.match(search.toLowerCase());
        //highlight-end
    }
  }
  setFilteredPosts(filteredPosts);
...

Since my blog has tags and stuff like that, I added functionality to search and filter by tags, too

    if (search.startsWith("#")) {
      return tags.includes(search.replace("#", ""));
    }
...

And that's it! But wait, there's more. This works, but you can't really share a search query to someone else, I can share google links - google.com/search?q=github
I think that's kinda important, like, for times when I have to share all my Rust blogs, it's just easier and convenient.

so well, let's update the URL to include the search query, in real time! I hadn't ever done this before, so it was great learning it. I got the inspiration from the IFTTT search engine

I found out about the window.history.pushState() method, which basically allows you to push a new URL without adding it to the browser history, or reloading the page. Read the documentation for the same over here -
History API | MDN

useEffect(() => {
        if (search) {
      //highlight-start
            if (window.history.pushState) {
        window.history.pushState(null, null, `/?q=${search}`);
      }
      //highlight-end
      const filteredPosts = posts.filter(post => {
        const title = post.title.toLowerCase();
        const description = post.description.toLowerCase();
        return title.match(search.toLowerCase()) || description.match(search.toLowerCase());
    }
  }
  setFilteredPosts(filteredPosts);
  }, [search]);

And now, we also need to parse the original request, using the window location object, and make it default for the useState hook we made for search

                      // 👇🏻 converts the URL from HTML encoded to a string (%20 to space)
    const initialState = decodeURI(location.href? // Use window location
                      .split('/')               // Split the URL into an array
                      .pop()                    // Get the last element only
                      .split("=")               // at this point, it's q=search, so we only need the "Search" parth
                      .pop() );                 

                                        // 👇🏻 We're using the initialState to set the search query.
  const [search, setSearch] = useState(initialState); // Now, only the blogs that match the query will be displayed on first load

That's it!

The full implementation can be found in the source code of this blog on Github

You can try out the search yourself

Feel free to visit the repository for this blog here

This website doesn't have an API, so I made my own, with extra features.

Dhravya — Fri, 06 May 2022 13:14:54 +0000

The poet_this twitter bot I made, used to use selenium to extract tweets from https://poet.so (literally going to the website, pasting link, taking a screenshot 💀)

This had several, several issues:

Very costly to host
Inefficient by design
Slow
Highly prone to errors and crashes

It was time to change this.

So I went ahead and created :

a python script that uses the Python Image Library to generate the image. This was quite challenging, because I had to (very painfully) plot stuff pixel-by-pixel, and also had to ensure that the tweet fits in the box (for which, I made my own trick where I vertically expanded the image, I'll write a bigger blog explaining the code)
An API that anyone and everyone can use
A frontend for the API
An overall framework, that works for twitter stuff for now, but I'll also be able to implement the same for reddit posts (and comments), etc. you get the point

It was a lot of hassle to make things like mentions, hashtags, and link appear in blue color, media support, and the replies, but I did it!

You can visit the website here - https://beautify.dhravya.dev

To use it instantly on any tweet, mention @poet_this
So yeah, it's time to change the name from poet_this to beautify_this lol

"But isn't this the same as poet.so?"
Well, it is for the most part, but I've also added (a very few, but important) extra features like replies showing up with a preview of the original tweet, with the website doesn't have.

I also think that my updated design is better

also, THANK YOU SO MUCH for showing so much love on the last post. Because of your support, I got a 50$ forem shop voucher. Even though I'll be unable to afford the shipping costs, it means a lot to me to just receive the voucher.

DEV Community: Dhravya

Supermemory - ChatGPT for your bookmarks

What I Built

Demo

My Code

supermemoryai / supermemory

Build your own second brain with supermemory. It's a ChatGPT for your bookmarks. Import tweets or save websites and content using the chrome extension.

SuperMemory

👀 What is this?

Key Features

Journey

Need some advice!

Deploy your functions and use them everywhere⚡⚡

Owofying text

WTF is a cloudflare worker?

OK, NOW GETTING TO THE PROGRAMMING AGAIN.

Creating the worker

How to serve *any* website from your own domain (for free!)

Mohit Yadav ・ Aug 13 '22

Deploying the worker

WTF is an IOS shortcut?

That's it!!!

AD

Creating a Twitter bot in Python using Twitter APIv2 🚀🚀

🤔 What we're building

⚡ Getting started

Setting up the project

🔐 Authenticating with Twitter

👨🏻‍💻 Let's start with the code now!

Finally...

Disco.pics - Free and open source image hosting service

Overview of My Submission

Submission Category:

Overview video

Language Used

Link to Code

discopics / disco.pics

An image hosting service with all the customisation you need

This project has been discontinued. Use at your own risk

Disco.pics

Overview video

How it works

How the data is stored

Additional Resources / Info

Collaborators

How to ask for help

What not to do

Asking people to join a voice call

"My code doesn't work"

Pinging random people, giving bribes, and other things

Sending screenshots of your code

What you SHOULD do

Google before asking

Don't ask to ask, just ask

Frame your questions correctly

Be respectful even if you don't get help.

TL;DR

A whole new developer ecosystem: CakeCutter 🍰

Introducing Cakecutter🍰- Start projects quickly from Cakes (templates)

Dhravya ・ Mar 9 '22

Ok, but what exactly is Cakecutter?

Sold yet? Install cakecutter.

These are not even the biggest updates.

Your support is appreciated. Here’s how you can help

cake-cutter / cakecutter

Open source, cross-platform developer tool that helps you quick-start your project. Creates files, runs commands, templating based on question. Also works as a setup tool

Cakecutter

👀 What is Cakecutter?

How I (organically) grew a Twitter bot account from 0 to 750 followers in 20 days.

Can a twitter bot get 1,000 followers in 100 days? A challenge to myself and this community

Dhravya ・ Apr 28 ・ 1 min read

But why?

How?

Challenges

How you can help.

Dhravya / beautify-this-bot

A twitter bot that simply replies with a beautiful screenshot of the tweet, powered by beautify.dhravya.dev

Poet this!

Installation

Usage

How to serve any website from your own domain (for free!)