Refreshing to see a dev channel that isn't selling courses, vague motivational speeches, and cinematic coffee pours. I'm still too new for this one but saving it for later
I'm writing a multi-threaded elevator system (Scala3/ZIO2) that receives requests via a socket connection (mimicking inside elevator and outside floor requests). Tasks: Experiment with different dispatch strategies and data structures (Priority Queues, Fibonacci Heaps). Optimize for wait time vs travel time. Find ways to optimize fairness. For example elevators could mutually monitor and "steal" requests from others if they are better suited for handling it. Have a global dispatcher re-balance load (floor requests) across elevators. Deploy elevators with different strategies (one optimized for wait time the other for travel distance). Optimize for commonly seen patterns during office hours (9 vs 5). Capture data and visualize statistics to evaluate each approach. Visualize elevator motion. That'll keep me busy for some time :)
This is exactly what I needed man. I've been so burnt out with the idea that I can't get any better outside of work. It sucks being super slow at learning man.
Loved this! I actually ended up making my own ETL pipeline for RUclips comments recently! I had some issues with airflow, but the full pipeline consists of extracting raw text data into a source table, applying transformations, and finally moving the cleaned data into a new table. The database server is PostgreSQL. I've been on the edge on whether I should now move into an analysis stage, but you inspired me to begin!
7:16 I'd also add a sudoku generator feature as extension suggestion - it requires a solver and finding a way to generate random ones more or less quickly can be interesting too.
Awesome idea! You'd need some mechanism for determining if a Sudoku board is "solvable" (i.e. has one and only one solution), which in itself could be a cool project.
Thank you. It’s overwhelmingly helpful to have you breakdown the projects; expectations of concepts; and a clear concise direction to work towards; and lastly, why and the impact of learning everything. I know other channels do this. But your method is clear and very digestible. Thank you.
Hi Bennet! Thank you so much for this video, this is what I was looking for. I'm a security engineer and I want to level-up my software development skills. The OAuth2 Authentication Service is like great project for me. Thank you again. P.S: I appreciate how natural this video is. This is what technical content should be in RUclips. I'm tired of excessive edition, stock videos, unnecessarily loud music.... I've been subscribed to your channel for a while, and I like our style❤
Great ideas, I have done several Sudoku solvers and had a laugh when you mentioned doing that. The last one was in Javascript which blew me away by its speed since I am old school and had the impression that JS was slow.
In terms of Sudoku solver a good bonus point would be to implement a Sudoku app feature of being able to process input as an image file. That would require some image perspective processing and OCR logic
I thought about doing that but it is extremely difficult to do I think without using an OCR library. Instead, I web-scraped the cell data from websites that publish Sudoku puzzles since they use code to generate a table for the grid.
I definitely need more advanced project experience. I have 3 years of professional web dev experience, looking for my next job in the industry right now at a Mid-level. Hard to find.
Please timestamp this, most people don't want to weed through a video like this for a project they're interested in: 1. Image Manipulation Program 1:21 2. Full Stack Application deployed with Docker 2:57 3. OAuth 2 Server (PKCE) 4:18 4. Sudoku Solver 6:24 5. Pulling Tweets 8:06
@@Coral_dude Yes, because the desire to sit through 10 unstructured minutes of rambling to find out all 5 "advanced" projects are basic ass ideas you can find on any blog post is truly conducive to being or becoming a developer.
Thank you for this video. This inspired me and gave me more ideas on how to improve skills, and maybe for more fun projects. Sorry if I made mistakes, because I'm just started learning English and left this comment also for practicing. Thanks again😊.
Wanted to do the last project involving Twitter and performing Sentiment Analysis but unfortunately Twitter removed the ability to scrape the site to get data :(
For the second project, what would be the best way to architecture a Django backend and React frontend application? Would they be in separate repositories?
Project 5 sounds very interested to me, I'm trying to access an internal CRM system to get customers comments and store it in database they run them through LLM to perform sentiment analysis. I already have an app that does it but data collection is a manual process so not scalable and efficient. Ideally I'd love to create a dashboard that would display customers sentiment and we can then take action accordingly.
Good video. One clarification - flood fill is filling connected areas - pixels, if you will. So if you have red and you are flooding (replacing) with blue you start on a particular red pixel and then any other red pixel that is connected to that red pixel changes to blue - but others do not. Happy coding!
Is the OAuth2 authentication system like an Auth0 clone ? Basically the app is the middle of another backend app and the provider (Google OAuth for e.g), if that's what you mean?
Awesome video, definitely going to try throwing more in a docker container. tentorflow keras, scipy, biopython with numpy and matplotlib is the beast of a project I am conquering rn. just another great set of skills.
Hey! I've got a lot of questions regarding project 2 so sorry in advanced if this comment is long. When you say deploy an entire stack using docker, do you mean just be able to spin up the entire stack locally? Or do you mean actually deploying it to a service like AWS with the docker-compose file. Also, would it be best practice to containerize your database and redis separate from your server, and if so, how would you go about doing that with services like planetscale and redis-labs.
I can't reccoment implementing your own OAuth server as a side project. Won't work and the scope of the features too large. Maybe clone the Duende Identity Server repo and poke around and get it up and running, request access tokens, hook up a SPA to it for OIDC authentication etc.
I didn't say you had to do it from scratch with vanilla code! Roll your own auth is always a bad idea (unless you really want to learn the underlying mechanics). Luckily there are a bunch of libraries out there to help you do this. I've done it when I worked at a startup, so I know it's possible & will teach a lot.
Hey man I loved this video and was thinking to implement a login service which will use PKCE as well as normal login. I was wondering to see if there is a way to let the user decide the configuration such as choice of db sql or nosql and cookie based or jwt based auth in terms of normal login. Or choosing various oauth flows, something like that. The user in this case would be the consumer of the project not the actual end user or client
Sounds like you're trying to build an identity provider like Auth0, Okta, Duo, etc. Definitely possible, though the scope of your proposed project is huge & would take a lot of effort! I'd say start with one - e.g. PKCE backed by a SQL database - and then expand from there if you want to add options.
Hi, can you give me an example of a project that would fit the no.2 idea? would building a facebook type platform be one?or could you list out a couple. thanks
You can dockerize almost any type of software, so the only limit is your imagination on this one. In the video, I suggest a web application & recreating any app on the internet is fair game.
idk... concurrency, or alghorithms are not it. in the real world repos are huge and simplicity and uniformity are the main objective. getting fancy with it is ussually trouble down the line. here and there u might use it but 99% is finding the right place to do something simple. u are not gonna design/implement authentication, u are just gonna use it. ur project ideas are interesting, the problem is that they are too much work without teaching u the actuall skill u need. which is contributing to large thing without damaging it. ofc if u are enjoying doing these projects, do them. im just saying that if u are getting stucked, loosing motivation and feeling overwhelmed, there is no need to torture urself cause these projects are not where its at in the end. ive never done it myself but trying to contribute to some open source might be the way to go.
I beg to differ. I got paid to implement OAuth2 as authorization for our API at a startup. Of course, I used a library to do so, but I didn't say you had to code it from scratch in the video! Also, at nearly every company I've worked, concurrency is important. You limit your scalability without it. If you've got better project ideas that are accessible & teach the "right" things, I'd love to hear them.
@@BennettGarner its not hard to start a bunch threads and let them do stuff. its hard to discern the right scenario and have repo that is useable afterwards. those arent tasks for juniors with one side project experience. again good project ideas, but i dont think that side projects can teach u what u need to actually learn. u are not gonna become senior dev on solo projects, but by seeing and understanding lot of projects and by contributing to them. and working solo on a project that has no deadlines, no demands, no rewards is way harder than to work with bunch of people on something that is gonna be actually used. im not trying to be mean but i think this needs to be said.
Does someone know a sample project that could be used for number 2? I'd like to do that one but don't want to have to go through the hassle of building the millionth crud app of my life first.
How about not a CRUD app then? What if you built a log aggregation API that allowed a user to POST a new log and GET a list of recent logs, that's it. No update, no delete. For that project, I can think of a couple things that would be useful to add to your docker-compose file, like a Celery worker & a Redis cache. It wouldn't even need to have a UI at first, just the API with maybe some docs published via nginx?
Refreshing to see a dev channel that isn't selling courses, vague motivational speeches, and cinematic coffee pours. I'm still too new for this one but saving it for later
Lollll if this isn’t the most accurate summary of coding youtubers idk what is! 😂
another good one is Dorian Develops. He's real as hell.
oh boy don't expect much if you're new. Its tough out here
You named three very specific channels LOL
Is selling courses a bad thing?
I'm writing a multi-threaded elevator system (Scala3/ZIO2) that receives requests via a socket connection (mimicking inside elevator and outside floor requests).
Tasks: Experiment with different dispatch strategies and data structures (Priority Queues, Fibonacci Heaps). Optimize for wait time vs travel time. Find ways to optimize fairness. For example elevators could mutually monitor and "steal" requests from others if they are better suited for handling it. Have a global dispatcher re-balance load (floor requests) across elevators. Deploy elevators with different strategies (one optimized for wait time the other for travel distance). Optimize for commonly seen patterns during office hours (9 vs 5). Capture data and visualize statistics to evaluate each approach. Visualize elevator motion. That'll keep me busy for some time :)
This is exactly what I needed man. I've been so burnt out with the idea that I can't get any better outside of work. It sucks being super slow at learning man.
Loved this! I actually ended up making my own ETL pipeline for RUclips comments recently! I had some issues with airflow, but the full pipeline consists of extracting raw text data into a source table, applying transformations, and finally moving the cleaned data into a new table. The database server is PostgreSQL. I've been on the edge on whether I should now move into an analysis stage, but you inspired me to begin!
Have you made it open source, mind sharing the code?
Very cool! Great idea.
Sounds like a good time! I love a good ETL pipeline. "I had some issues with airflow" , don't worry, its a feature not a bug LOL
7:16 I'd also add a sudoku generator feature as extension suggestion - it requires a solver and finding a way to generate random ones more or less quickly can be interesting too.
Awesome idea! You'd need some mechanism for determining if a Sudoku board is "solvable" (i.e. has one and only one solution), which in itself could be a cool project.
Thank you. It’s overwhelmingly helpful to have you breakdown the projects; expectations of concepts; and a clear concise direction to work towards; and lastly, why and the impact of learning everything.
I know other channels do this. But your method is clear and very digestible. Thank you.
Glad it was helpful!
Hi Bennet! Thank you so much for this video, this is what I was looking for. I'm a security engineer and I want to level-up my software development skills. The OAuth2 Authentication Service is like great project for me. Thank you again.
P.S: I appreciate how natural this video is. This is what technical content should be in RUclips. I'm tired of excessive edition, stock videos, unnecessarily loud music.... I've been subscribed to your channel for a while, and I like our style❤
I'm definitely not a senior engineer, maybe not even mid level, but I was looking for some advanced projects like those, thanks for the ideas!!
Glad I could help!
Great ideas! Will try one or two of these. Good way to improve my skills and have a project to show interviewers. thumbs up
The best project idea video I've seen! Will definitely try to build a project or two👍
Great ideas, I have done several Sudoku solvers and had a laugh when you mentioned doing that. The last one was in Javascript which blew me away by its speed since I am old school and had the impression that JS was slow.
In terms of Sudoku solver a good bonus point would be to implement a Sudoku app feature of being able to process input as an image file. That would require some image perspective processing and OCR logic
I thought about doing that but it is extremely difficult to do I think without using an OCR library. Instead, I web-scraped the cell data from websites that publish Sudoku puzzles since they use code to generate a table for the grid.
Now for the sudoku solver as a follow up attempt to prove NP=P or disprove it :)
I definitely need more advanced project experience. I have 3 years of professional web dev experience, looking for my next job in the industry right now at a Mid-level. Hard to find.
Please timestamp this, most people don't want to weed through a video like this for a project they're interested in:
1. Image Manipulation Program 1:21
2. Full Stack Application deployed with Docker 2:57
3. OAuth 2 Server (PKCE) 4:18
4. Sudoku Solver 6:24
5. Pulling Tweets 8:06
we just need to chill, it's only 10 min of our time
Spoken like someone who is not and will not be a developer
Ps thank you for the stamps you’re the real MVP
@@Coral_dude Yes, because the desire to sit through 10 unstructured minutes of rambling to find out all 5 "advanced" projects are basic ass ideas you can find on any blog post is truly conducive to being or becoming a developer.
@@facundodominguez1063 10 minutes I didn't want to spend listening to project ideas ive heard a thousand times.
Great advice thank you for the suggestions.
Love this, this seems to exactly match where I am!
Great ideas man will surely try them...
Add API using micro services architecture to the list
Those dishes/plates on the top right looked like golden and silver play buttons of RUclips 😁
Haha, nowhere near that yet!
Thank you for this video. This inspired me and gave me more ideas on how to improve skills, and maybe for more fun projects. Sorry if I made mistakes, because I'm just started learning English and left this comment also for practicing. Thanks again😊.
Awesome! Glad the video was helpful & thanks for the comment!
Awesome! Subscribed immediately.
Welcome aboard!
Wanted to do the last project involving Twitter and performing Sentiment Analysis but unfortunately Twitter removed the ability to scrape the site to get data :(
For the second project, what would be the best way to architecture a Django backend and React frontend application? Would they be in separate repositories?
Why not put them in one repo with 2 folders such as frontend and backend?
create a client folder for front end a server folder for the backend
but it really does not make a difference you can create two repos
Project 5 sounds very interested to me, I'm trying to access an internal CRM system to get customers comments and store it in database they run them through LLM to perform sentiment analysis. I already have an app that does it but data collection is a manual process so not scalable and efficient. Ideally I'd love to create a dashboard that would display customers sentiment and we can then take action accordingly.
Cool idea. Data extraction & cleaning will certainly be some of the hardest parts. Sounds similar to dovetail.com/
Good video. One clarification - flood fill is filling connected areas - pixels, if you will. So if you have red and you are flooding (replacing) with blue you start on a particular red pixel and then any other red pixel that is connected to that red pixel changes to blue - but others do not. Happy coding!
Not always no and most image editors worth their salt will let you toggle between "contiguous" fill and non-contig.
Great vid. Thanks for posting.
these will look great on my github
Is the OAuth2 authentication system like an Auth0 clone ? Basically the app is the middle of another backend app and the provider (Google OAuth for e.g), if that's what you mean?
Awesome video, definitely going to try throwing more in a docker container. tentorflow keras, scipy, biopython with numpy and matplotlib is the beast of a project I am conquering rn. just another great set of skills.
beautiful... was looking for this
Amazing. Happy to help!
a video that is true to it's title 👏👏
This made my day 😊
Bennet!! Met you through jct you look like such a baby in this video ! Aww.
Great video, thanks for the tips!
Let me know when you build one of the projects!
Hey! I've got a lot of questions regarding project 2 so sorry in advanced if this comment is long. When you say deploy an entire stack using docker, do you mean just be able to spin up the entire stack locally? Or do you mean actually deploying it to a service like AWS with the docker-compose file. Also, would it be best practice to containerize your database and redis separate from your server, and if so, how would you go about doing that with services like planetscale and redis-labs.
all of those questions should be google searches, starting by finding out whether AWS has a free tier
I can't reccoment implementing your own OAuth server as a side project. Won't work and the scope of the features too large. Maybe clone the Duende Identity Server repo and poke around and get it up and running, request access tokens, hook up a SPA to it for OIDC authentication etc.
I didn't say you had to do it from scratch with vanilla code! Roll your own auth is always a bad idea (unless you really want to learn the underlying mechanics). Luckily there are a bunch of libraries out there to help you do this.
I've done it when I worked at a startup, so I know it's possible & will teach a lot.
Can’t do the tweet thing anymore. API costs
3:46 - is it a good practice to deploy your application with docker-compose?
There are many ways to deploy an application, but I'd say docker-compose is among the industry standard best practices right now.
Hey man I loved this video and was thinking to implement a login service which will use PKCE as well as normal login. I was wondering to see if there is a way to let the user decide the configuration such as choice of db sql or nosql and cookie based or jwt based auth in terms of normal login. Or choosing various oauth flows, something like that.
The user in this case would be the consumer of the project not the actual end user or client
Sounds like you're trying to build an identity provider like Auth0, Okta, Duo, etc. Definitely possible, though the scope of your proposed project is huge & would take a lot of effort!
I'd say start with one - e.g. PKCE backed by a SQL database - and then expand from there if you want to add options.
loved this video man!!
Appreciate it!!
Great video.
Glad you enjoyed it!
Hi, can you give me an example of a project that would fit the no.2 idea? would building a facebook type platform be one?or could you list out a couple. thanks
You can dockerize almost any type of software, so the only limit is your imagination on this one. In the video, I suggest a web application & recreating any app on the internet is fair game.
You're cool dude
idk... concurrency, or alghorithms are not it. in the real world repos are huge and simplicity and uniformity are the main objective. getting fancy with it is ussually trouble down the line. here and there u might use it but 99% is finding the right place to do something simple.
u are not gonna design/implement authentication, u are just gonna use it.
ur project ideas are interesting, the problem is that they are too much work without teaching u the actuall skill u need. which is contributing to large thing without damaging it.
ofc if u are enjoying doing these projects, do them. im just saying that if u are getting stucked, loosing motivation and feeling overwhelmed, there is no need to torture urself cause these projects are not where its at in the end.
ive never done it myself but trying to contribute to some open source might be the way to go.
I beg to differ. I got paid to implement OAuth2 as authorization for our API at a startup. Of course, I used a library to do so, but I didn't say you had to code it from scratch in the video!
Also, at nearly every company I've worked, concurrency is important. You limit your scalability without it.
If you've got better project ideas that are accessible & teach the "right" things, I'd love to hear them.
@@BennettGarner its not hard to start a bunch threads and let them do stuff. its hard to discern the right scenario and have repo that is useable afterwards.
those arent tasks for juniors with one side project experience.
again good project ideas, but i dont think that side projects can teach u what u need to actually learn. u are not gonna become senior dev on solo projects, but by seeing and understanding lot of projects and by contributing to them.
and working solo on a project that has no deadlines, no demands, no rewards is way harder than to work with bunch of people on something that is gonna be actually used.
im not trying to be mean but i think this needs to be said.
Does someone know a sample project that could be used for number 2? I'd like to do that one but don't want to have to go through the hassle of building the millionth crud app of my life first.
How about not a CRUD app then? What if you built a log aggregation API that allowed a user to POST a new log and GET a list of recent logs, that's it. No update, no delete.
For that project, I can think of a couple things that would be useful to add to your docker-compose file, like a Celery worker & a Redis cache. It wouldn't even need to have a UI at first, just the API with maybe some docs published via nginx?
Wow! His eyes are beautiful!
Holy shit do you know Forrestknight, you look just like him 😂😂
twitter api is expensive now...
Yeah true. Can you tell this is an older video? You'll probably want to find a different API for that project
beautifull hair
should have mentioned compiler or vm
you've got very interesting eyes.
improve sound brah
r u a man or a woman
You are a beautiful woman
Someone pls give a list of what projects he is saying.
If Bistro Huddy was a programmer
a video that is true to it's title 👏👏