#Pictopy
1 messages · Page 2 of 1
Ok thanks, trying for doing some quality work
Keep it up👍
you get stuck somewhere contact us"""
and listen mentor is out for a week he informed us so don't think he isn't active
keep up thee good work
I AM PRETTY MUC DONE
Just making final touch.
Yeah.
Having some issues rendering returned image.
will be solved in some time
??
@abstract bear You are now level 19! :confetti_ball:
Hello I am Divyanshu Shekhar. I am looking forward to contributing to this project.
I have good knowledge of MERN, Go, Python. I have done in the past, simple project in Rust and Tauri too.
Okay.
so you can go through codebase .
and try to understand workflow.
Specially Tauri and Rust part.
Yes sure
Let me know if you face any issue in set-up
I will try to understand it to my best
Great!
whenever i am trying to run " npm run tauri dev" my system is crashing again & again
that sound great clone the repository and listen there some issue when u run npm run tauri dev
some bug
typoo error
i made the changes and but and pr for the same
our mentor is out for week (he informed us) so we have to wait a little a while
but you send me screen shot i can helpp
tuh bas ss bhejjjj mai dekhahta hoon 🧐
ok
I mean i followed the steps from the github file
till the npm run step
till the last step you see on this SS
@opal cradle You are now level 2! :confetti_ball:
it just crashed again
bro listen trying to run scriptt
go the frontnend folder
and then npn run tauri dev
ok
run npm run taui dev
it is still crashing
did u installl all the dependeciess
npm install
yeah i did
yeah
?
venv?
bro just give me a minute show me your screen i might get to know better what is the problem
?
um......no
🙃
i told you i just followed till the npm run trauri dev command
why is this issue coming ??
No worries it is solved
@opal cradle You are now level 3! :confetti_ball:
greattt listenn can u add more themesss
brooo i can give you something
good for new bie
good first issue
I have a concern about optimizing the overall performance of the application. Why are we querying the database using SQLite in Python instead of Rust? Won't it increase the latency
Exactly. There is this same issue with image fetching and posting. Why not fetch with Rust itself?
you are thinking that while rust is faster than the python
right that is what are u saying?
beacuse Rust may complicate maintenance without significant performance gains unless queries are highly compute-intensive.
Oh i see ,so your intuition is that fetching of images is faster with the help of Python ,right?
Not exactly. Fetching images using Python can be efficient for smaller-scale applications,
But the high memory usage of the app ... Don't you think it's due to the overhead of Python only ??
Yes but when the application will be scaled further then for that I was having the concern
but for that we are using effecint library like numpy or pillloww
if u saw
Proper profiling and optimization in Python might resolve the issue without the need to migrate to Rust.
Well with libraries it would be better to stick with python but had the general doubt so yeah
Maybe certainly, just had the idea to do so. I guess I will too have to check some benchmarks like pillow and rust image libraries.
share them tooo
i wanna have a look at them
but generally spaeking even the senior engeenir takes around 1 year to be comofortable be rust
Exactly and it might be tough for us to implement certain advanced features now.
its better we stick with python
Pillow is pretty efficient while it comes to image processing.
@worldly pulsar You are now level 19! :confetti_ball:
Its scalable and can be easily integrated.
And also, we can implement any AI feature easily .
Yeah.
It will be good for future improvements.
Pillow uses Numpy for most of the processing and computation.
And Numpy is implemented in C++ . So that makes pillow efficient because of underlying C++ implementation
New AI models will not be compatible with rust.
wahheee
And Python itself has huge library support to implement anything.
yeah for implementation it's definately easier,i got curious looking at the database codebase and thought of asking . Well got to learn something new !!
There is so much more to learn.
Hello everyone, I've been trying to figure out to contibute, can anyone tell me what's under developement and what are the places i can contribrute to
All the projects in AOSSIE are under development and are being actively maintained.
What's your preferred tech stack?
I am still waiting 🙂
ooh listenn my recent pr itt need to be memrgedd
see i have added a new feature through which user can control the entire ui
so in that i have only three themes where you can add more themes
You can try to find bugs.
Raise the issue and solve it.
It will be great experience for you.
It just feels different to solve an issue from scratch.
Ok 👌
Just a suggestion.
You can also pick existing issue.
let my recent pr merge then you can work on that
then
well parag @worldly pulsar
what about
the
documentation
working on it.
will add today.
till when ?
mtlb u have to go somewhere
so bro just make sure pr is there before that
documentation
is necessary
i mean.
i will finish it till evening.
I will
ok when are u leaving ??
and would be back when ?
tomorrow.
on 19 .
okk best of luck bro
thamks!
i'll be working on the face detection things live and then sort imagess hope fully it will merge soon to
It will be nice.
and listenn paragg requirement .txt i pretty much added all the dependienies would make a pr for it by tonight ??
sound great and that model we need to soon
Sure.
??
the model u would add to right
Yeah.
Hey @abstract bear .
I have added the documentation .
Check it out .
GREATT """"
@mystic sigil
hy tell me this is the better place to discuss
would be happy to have your views
?
CAN WE clearify because u were talking about onnx compatibilty and then were saying onnx won't be supported for long
@abstract bear I have mentioned all the problems there https://github.com/AOSSIE-Org/PictoPy/pull/244
for your reference
i. Models size is 4 GB that's too large from my end
ii. As ONNX does not support it so , I have no idea it works for all divices or not
Description:
This PR introduces the following updates:
Added a new section for Model Documentation in the main README.md to guide users toward more detailed information about the AI models used in...
but you recently said there is certainity that onnx would be supported in the future
??
model size is a good point
i am working on it
shukriya for a great feedback
but can u clearify this
Yeah.
Whats your point?
the current focus is on delivering a functional feature, with flexibility for future adaptations if needed.
I have told "In the future, there is no certainty whether ONNX will continue to be supported." means I have no idea ONNX in future will support or not the model you provided
it remains a widely adopted format with strong community backing and frequent updates
So, Why are we focusing so much on ONNX compatibility?
but firstly u were saying why its not onnx compitable
so i am not getting you point
can u elaborate
can i provide u file u can try it out with all the filess
@worldly pulsar @abstract bear , let me clarify what I am saying. Your diffusion model is not supported by ONNX, which violates the backend Pictopy architecture.
Since ONNX does not support it, there is no guarantee that it will work across all devices
@mystic sigil You are now level 7! :confetti_ball:
And 2ndly it size is too high around 3.8 GB
On what basis you are saying this?
Have you tested it on multiple devices?
hile ONNX doesn’t support the diffusion model,
I have told there is no gurantee
size i am working on it
I agree that size is a concern but it can be optimized in the future.
ONNX doesn.t support stable diffusionn models
How can you be sure that it will work on all divice but if ONNX supports then Yes , I can say it will work on all divices
Its working perfectly fine on Windows and Linux.
I am sure because I have tested it.
The key point is that while ONNX provides a broader guarantee, the current model may still work across a significant range of devices,
@worldly pulsar If that only matters then Yolo model was not used here , for your reference you check the previous discussion of Pictopy
okay i wasn,t aware that we have go through the chat around a year go
can u add that in the documentation
Your point was multi-device compatibility.
Which you didnt even back with proof.
all my saying @mystic sigil can u just try it out
please so that we can get an feedback and work towards it
i am working on the model reducing its size
Try it on your system . And then we can have a discussion.
Let it be merged first then surely I will test it in my system
well the concern you showed us great full i have already started the workk for the size of the model and onnx
and if possibble can u help us out in testing
@mystic sigil asking as a fellow developer
We would love to hear your feedback!
@worldly pulsar You are now level 20! :confetti_ball:
Surely
@mystic sigil
just wanted to say you were your conern is real
at mid night going through all website onnx is a must
it would take a lot of effort but right now the model size has been shrink too around 2 gb
i am downloading automatic 1111 to optimize the midel and convert it itno onnx format
shurkiya bhai @mystic sigil
Looking forward to new Model.
do you guys know a good dataset for training images for gallery with which I can add more classes , I searched literally for hours and found no properly labeled dataset , few found were too large which were not possible to train locally
you can try imagenet : https://www.image-net.org/
I think its smaller versions exist
it has numeric labels do you have any idea how can I train yolov8n model wiht it?
also when loading it I dont see any labels folder, txt files required for training yolov8
ohh thanks
Great.
Now, we also have ONNX compatibility.
how to solve "failed to add folder" in ai tagging page but album is loading images from the folder but not ai tagging
Can you send console log messages.
is your backend up and running
??
INFO: Application startup complete.
INFO: Uvicorn running on http://0.0.0.0:8000 (Press CTRL+C to quit)
INFO: 127.0.0.1:45796 - "GET /images/all-image-objects HTTP/1.1" 200 OK
INFO: 127.0.0.1:45810 - "OPTIONS /images/add-folder HTTP/1.1" 200 OK
INFO: 127.0.0.1:45810 - "POST /images/add-folder HTTP/1.1" 400 Bad Request
yes its running
BHAI EK MINUTE WHEN DID U FORK THIS REPO
JUST ASKING
the repo has update a lot since then listen?
@viscid estuary
just go sync you your repo to the lastest pictopy repo
then it should work fine
You are running both frontend and backend on WSL right?
If not , please run both on WSL.
You might be facing CORS error.
And also try syncing the latest changes from main repo. Just as @abstract bear said.
Let me know if it works.
ok sure
..
https://medium.com/@mandalsouvik/safetensors-a-simple-and-safe-way-to-store-and-distribute-tensors-d9ba1931ba04 @worldly pulsar
SafeTensors : a simple and safe way to store and distribute tensors. Why and how it is faster than vanilla torch loading?
@abstract bear You are now level 24! :confetti_ball:
have a look at this
this might be the key for the speed
i read some articles about safetenors it we go from onnx to safetensor it would helpp
will help to keep the size low and i talk to some seniors about this
they said it won't effect accuracy if we use safetensor format and try quanitized from fp16 to int8
1.8 gb the size can easily be decreased to 800 mb or even less
@worldly pulsar
@abstract bear , You are right.
SafeTensors can significantly reduce model size while ensuring fast, efficient, and secure loading. It's perfect for keeping things lightweight without accuracy loss.
Combining SafeTensors with quantization (FP16 to INT8) can shrink a 1.8GB model to 800MB or less, maintaining performance and reducing memory overhead.
Any minor quality drop from quantization is negligible if done right.
SafeTensors boosts loading speed and security, ideal for real-time or edge applications. It’s the key to lightweight, efficient deployments.
Sure.
In that way, we can also expect more contributors to add more AI features.
Yeah.
It will be great.
we must workk on accuracyy
Yeah.
We will.
i am going through some docmentation
hey guys, what all issues are currently people working on?
Also, there have been a lot of merge conflicts lately
what is the issue
@radiant mauve we recently added a image genration feature
working on pr on requirement txt and and documentation
and alsoo some feature in ui and face detection for feature in backend
@radiant mauve, please check it out.
We have also added setup docs for the image-generation model.
I am working on some frontend bugs and also adding detailed documentation of existing AI models
I am working on writing workflows for generating executables for all platforms, I need to test on Mac, all the other stuff is ready, will raise a PR soon. Also, I am working along with @mystic sigil to remove the "images" folder from the backend of AI tagging and getting the images directly from the folders.
@rahul can u elaborate i was thinking to add a video in readme .md on how to setup pictopy locally so should i waitt for few days
You can make the video now also, for setting up the project locally for development.
great then
I have addded generate memories feature, secure folder , done some optimisations in face tagging backend for more accuracy
also I found out their a seperate thumbnails folder generating specially for thumbnails, I dont know who did that but that make things more slow and take even more of user's space
also I am now working on improving the existing functionalies, adding necessary features for a gallery app instead of good to haves as good of haves will make the app more bloated only
To solve uncessary merge conflicts and errors because of that we can add tests in our app and a github action workflow for it so after someone resolves the conflict it is easy to know if that contains any error or not I'll create an issue for it and start working on it first as it seems more important
I have create issues for it
No, no the thumbnails folder was added recently in one of the PRs. It doesn't take much space, only like 50kB per image. But with that we are able to drastically improve performance because we first render thumbnails while displaying rather than full resolution image in the grid.
@marsh idol You are now level 8! :confetti_ball:
You have got a point.
It improves performance drastically.
ok then its fine, I didnt look at how much space it takes
@wraith creek You are now level 13! :confetti_ball:
@wraith creek ,
yes?
You were saying to collaborate on buliding advance editing features
yeah we can if you want I'll dm you after sometime describing everything well discuss
yeah, a pipeline would be great
the project has becoming big now
yess Im almost done with tests for fast api then will start with rust test and lastly frontend and a pipeline on push
that's great
would be greatly helpful for me
it's becoming tougher to review each PR manually
End-to-end testing will decrease your workload exponentially.
i just commented on your PR
great work, but doesn't it diverge from our goals(as of now)?
image generation in a image gallery seems useless to me, until it can mix different images
rather than generation
what do you mean by end-to-end testing here?
Thank you for your feedback!
I understand your point about the feature diverging from the current goals. My intent with adding the image generation functionality was to explore AI-based capabilities that could enhance the user experience and provide a unique edge to the gallery.
yeah, the work was great
it's just, it's too early for the project and I am not sure, if it's right for the project
We can definitely consider adding the ability to mix or edit existing images in the future to make this feature more relevant and aligned with the project’s goals.
thank you for the feedback
yeah, that would be a pretty cool feature
We will add AI image generator soon.
its a great idea would workk on this right away
yeah
also, try making it modular
like suppose I don't want to use the image generator part, I can have a lighter variation of the app
and suppose I want to use it, I can download the model, and try it out
Sure.
we will give user an ability to generate image with low medieum and high quality
@radiant mauve we have resolve the merge conflicts
Yeah.
And we are also planning to give user choice to select between generating image with CPU or GPU .
hmm, what about doing this, based on load
like use the GPU, and if you have some CPU left, use that too(upto a limit)
user never knows which setting is best for it
Thats a great idea.
We will implement it for sure.
So, we will dynamically check User system resources and based on that use them.
yeah
we would make feature more robust and keep working on the model
great!
We will add it after Image-Generation PR gets merged.
no, didn't we just now concluded that image generation is not project's goal?
better would be image mixer
also, we decided to have modularity
got it
Got it👍
Can you elaborate a little bit, on Image-mixer.
So, that we would have a idea of what features to add.
why not you guys work on some other feature currently
image mixer is also too early, although it's a great Idea
if you wanna work, work on packaging the project
the project is still not executable
Got it, we are getting into it right away! I’ll focus on packaging the project and making sure it’s executable first. Thanks for the heads-up!
yeah, and it would be better, if you make a pipeline
so that we get an executable
after all the changes
Sure, I'll work on setting up a pipeline to ensure we can generate an executable seamlessly after implementing all the changes.
We will work on that.
Without pipeline, it will be manual work every time new changes appear.
@marsh idol has added that
he was working on it ig
sorry missed this
@worldly pulsar @abstract bear this is duplicate
seems like Rahul is already working on it
no problem then can i make a landing page for picto py from where the picto py is available for the user to download
@radiant mauve
a suttle sweet landing pagee
@radiant mauve , so now the executables are done. Should I start on something new, or do you have anything specific in mind?
Whatever you suggest, we can make it happen.
For example, we could develop a mobile app for PictoPy, as most people use image galleries on mobile.
if so, I would love to help you @worldly pulsar on developing it
are you guys also facing issues after the recent merge, there seems to be lots of error after @wraith creek 's recent merge.
Let see what @radiant mauve thinks about this.
I noticed duplicate dependencies and '}' missing in rust after resolving merge conflicts I've fixed that in my latest PR where I'm adding tests, apart from that I checked everything locally their were no errors, when I pulled the main origin their were few lines in mediaview.tsx whiich were causing ig some type errors those lines were not added by me
these merge conflicts problem will be resolved soon after adding tests and a pipeline which I'm adding
why nott we need people because adding some major things requires people
@marsh idol there is some merge conflict i noticed it iff possible just resolve it
if mentor allows for the landing pagee you can come in we can build it togetherr
Absolutely.
just @rose crane be readyy for some night fixing issue and bugsss🤪 and then going through documentation mid nighttt
@abstract bear You are now level 25! :confetti_ball:
@radiant mauve can u remove this enguage
I have fixed all meerge conflicts caused by other PRS too
should I push that
in my tests PR?
by the wayy the thing you are about ot make pr
i have alreadyyy workedd on it and pushed
but mentor was out after that
so it was'nt merge
@wraith creek no worries we gotta wait for my pr evreything would be good to go
I have fixed all the bugs and pushed
it's a;readyy there fixedd
have alook around a week beforee
pleasee
just chill broo tuhh kaare ya maii kya farak padhtha hai
ok I thought it was causing problems for many people so I just went ahead and fixed everything every file wherever the error was coming for easy setups
yeah
no worries these conflicts will be resolved soon after adding tests
and pipeline
not resolve fully but reduce significantly
yupp greatt work
what css update? I never updated any css
gnn
gn
I am getting a lot of errors in the MediaView.tsx file after a recent merge.
The app is not able to run and I am not able to resolve conflicts on my PR due to this.
okay wait let me checkkkk
let me get my handss dirtyy
give me some time @marsh idol paaji
paaji @marsh idol i just synced my repoo to the latest versionn
just give me an hour i'll fixx it
No no, you don't need to do that.
did it brother
I think it should be resolved by the person who coded that part
yup i get youu but its just some syntaxx errori 'll fixx it
@marsh idol working upon it give me timee
@marsh idol hy everything is up and running everything is up running
https://github.com/AOSSIE-Org/PictoPy/pull/270 i have madee pr which solve all the issue there were too many typoo too which isfixedd
and listenn everyone please don't use motion(framer-motion) in the frontend just stick to react type sccript tailwind css espicallyyy in src tauriiiiii
I saw that too after pulling main I've fixed that with all others in my latest PR yesterday but seems like @abstract bear has also fixed so their should be no issue after merging
bro please @wraith creek add your testt as soon as you cann
on it! can be a bit of delay because of exams but will be definitely done by the end of this week
hmm theek hai thodaa time lagea
if u need helpp in that
i can helpp
sure I'll tell you if I need any help in that!
i'll be happy to helpp
Someone solving this issue also add a 'bcrypt' module in requirements.txt , as 'bcrypt' used in https://github.com/AOSSIE-Org/PictoPy/blob/main/backend/app/database/albums.py but not in 'requirements.txt'
@abstract bear ,
Add bcrypt with your PR.
okkkkkkk i'll make a commit in that pr itself
let me solve this tooo
guys is anyone else also facing issues after the last merge
my app is not running after pulling
There were some errors.
@ KARNX7 has fixed them in his PR.
The app will be running after that PR gets merged.
why are the PRs getting merged so late?
@rose crane You are now level 3! :confetti_ball:
do anyone face same error?
Hey @viral cobalt , This error has been resolved but PR for that is not yet merged.
So, You will have to wait till that PR gets merged.
Or you can pull changes directly from the PR:
arham everything resolved just wait mentor will come and merge it
Thanks for your help @abstract bear @worldly pulsar
Hey guys, now there are lots of issues with merge conflicts
I advise you to, keep on pulling the main repo before making a merge
As now, to resolve the backlog of these
can you all, just create a google doc or something, with all MR's, and after I merge it in that order
you guys resolve conflicts, and fix yours(until your turn comes in the order)
it's needed otherwise, there's going to be a pile of MR's like this
can someone please take the initiative of creating a doc
and adding the PR's in it
hy @radiant mauve also do we need landing page for this app from where can download our app(pictopy
@marsh idol ,
I am getting this error while running frontend after recent merge..
Do, we have to run some commands for creating /dist folder?
If yes, please update Redme.md .
@everyone listen i am going to restructure the frontend what i mean by this is that becausee too many things are added so it making look cluttered (i don't think any user is going to use in such a condition ) so would make everything well structured and it should follow a simillar design i will create a notionn dockss for it so pleasee from now onn before you add anything before you add anything update in that notion dockk
I will check that after few hrs
I am out right now
No problem👍
I think just creating an empty folder named dist inside the backend folder should do it.
I should probably have updated the readme or created this folder
in my PR.
I will correct this.
@worldly pulsar I have opened the PR for the fix, thanks for pointing out.
Glad I could help!
https://docs.google.com/document/d/1UDnSGBjlTkDlibs9J85_cY4G8YCUFIHSsJS_NVg_K3A/edit?usp=sharing
here you go!
bro he deosnt meant thatt
ohh understood just the PRS with merge conflicts?
but that's variable too for eg if one PR gets merge their is a chance that it would cause conflicts with other
Though if still if doc with all MR are needed I'll update it by tonight
if you want to you can work no problem with me otherwise I'll update this tonight
naah if u can just do it
updated them now it contains all PRs with MRs
The order goes from top to down right?
The first one will get merged first
?
Yup
@wraith creek You are now level 14! :confetti_ball:
I guess so that's upto the maintianer the order in which he merges
guys the frontend looks ugly, is anyone working on it ? @abstract bear you previously mentioned working on it once are you doing ?
sorry can u elaborate i did nt get that @rose crane
the sidebar ui and everything is looking quite ugly, are working on revamping designs?
bro just have a look i already
made a pr customization and everything named such
i worked on teh side bar
it not mergedd
but it would be merged soo
soon
is there anything to be done on frontend where I can help?
i am right now working on something backend
and frontend pretty much done
there are some some routing issue in the frontend
people have not used react roiter well so its causing issue
ok i'll look into it
surre
Let me know if you face any issue in that.
I've fixed the AppRoutes.tsx file. There was issue with conditional rendering of Layout Routes
Do you need help with that?
No problem.
https://github.com/AOSSIE-Org/PictoPy/pull/287, fixed the routing issue here
https://github.com/AOSSIE-Org/PictoPy/pull/293
Can anyone please have a look in it ?
The PR is regarding the missing package that led to breaking of code in object detection in yolo model and also corrected the file path .
hey guys the images and design in the memories section look too big and don't go with other pages, planning to change them to have similar look throuhout the app
@rose crane You are now level 4! :confetti_ball:
bro hy i
i am already done with it
i was about to make a pr when i was about to get back group the collage
So, you are done with it?
yup i was working on it till like morning but couldnt push it
as i had like leave for collage
so i thought after collage i would come back and make a pr
ok then
great sorry i should have make a issue for it bea cause i thought i would fix it in my frontend restructring
things
sorry if u like started the work @rose crane
I hadn't stared yet
Hey guys, I want to contribute to pictopy. If you need assistance with your assigned isseue related to frontend please do let me know
Is the request being sent to the backend?
yup
see na i have loaded images from the frontend it showing uip here in the terminal \
but images are not there frontend
hy @everyone
i am make the web ui for the picto py where we will allow user to download it from
would need you all reveiws for it
great idea, do let me know if you need help
This is needed.
Sure.
Let us know when you are done building.
Hi @everyone
I'm looking forward to contribute to PictoPy
Would love to get connected with you all
Hey @radiant mauve ,
https://github.com/AOSSIE-Org/PictoPy/pull/308
I opened a new issue and created a pull request for the same
Can you review it once??
That's something I need to be good at..cp
Is it done?
guys when am trying to upload pic no pic is coming...can someone tell if u r also facing the same prob? images not processing
did you start the server?
It was working fine few days back.
please visit http://localhost:8000 and see if you get a message from the server.
not getting
which platform are you using? windows, mac..?
windows
ok, I can help you setup the project. Did you follow the instructions provided in the readme file for the setup? Or you are getting any errors following those
sure..am woking on eduaid for now..will let uk once i start working again on pictopy
..
👍
@marsh idol You are now level 9! :confetti_ball:
I would like to collaborate
When did you fork the repo?
Recently or its been many day?
Today
@frail rover You are now level 3! :confetti_ball:
If you are getting a message at localhost:8000,
there must be issue with server setup.
Are you able to see these lines when server is running?
Loading existing face clusters from database...
[2025-02-12 19:25:06 +0000] [6502] [INFO] Running on http://0.0.0.0:8000 (CTRL + C to quit)
Hey, can anyone help me with this, I following the steps in the readme file and trying to run the backend but getting this error
nvm got fixed
Great then
Great.
it will be better if you run backend in WSL.
.
i have worked on hero.tsx and just decided more thing
i will contact you when i finalized for some sectionss
Hey everyone! 👋
This is my first contribution to this repository! I'm still exploring and understanding the codebase, but as I ran the application for the very first time, this was the first thing that caught my attention. So, I decided to fix it!
It's a minor UI fix for the folder picker icon alignment on the Setup screen when the app opens in a smaller resolution.
You can check it out here: 🔗 https://github.com/AOSSIE-Org/PictoPy/issues/323
Looking forward to learning more and contributing further! 🚀
oh I am using endeavor(arch linux) not sure how wsl will help?
Oh.
Then it will not be an issue.
I thought you are using windows.
yea, I was actually trying to use uv but ig that dosent work for some reason, I switched back to the normal pip workflow and it started working
its great start to start from something small if u get stuck somewhere and you need help message me i'll helpp
I am not able to upload image folder. Not sure what the problem is? Is anyone else having the same problem?
Yeah I too had this..haven't tried after that..will try later on..
To reproduce the problem, I cloned the repository again and setup the project. I am not getting this issue. If you were able to see those "retry" buttons, then did you try to switch to another tab like settings and then going back to the AI tagging tab? This will re-fetch the images and the images will load without any error from then on.
If this still doesn't work, please let me know what do you get when you visit this url: http://localhost:8000
You should see a message like this:
If you are not able to see that message then, the server isn't running.
I am also not getting this issue.
Folders are being uploaded without any issue.
Everyone here is my implementation idea for https://github.com/AOSSIE-Org/PictoPy/issues/210
I would greatly appreciate any suggestions from your end.
@mystic sigil , right now, we are saving thumbnails in a folder.
So, I didn't get your point about fixing it.
Can you elaborate a bit?
Your idea of storing edited images seperately is good one.
My motive , what ever operation we do with our application should not reflect in actual folder , that's why decided to shifted all thumbnails to a single folder images/pitopy.thubnails
Oh.
Got it.
Go ahead.
It will be nice improvement.
Thanks @worldly pulsar
I've identified a potential memory leak issue in the ONNX Runtime sessions during batch image processing. The problem occurs in both YOLOv8 and FaceNet components where ONNX sessions aren't being properly released.
I've created an issue here: https://github.com/AOSSIE-Org/PictoPy/issues/335
Would love to get some feedback.
I think we should add tone curve in editing , what do you think?
I think as Bruno sir mentioned in the GSoC announcement, we need to suggest changes/features which align with the project's vision. This project is an image tagging/classification and face recognition-powered gallery. So adding more photo editing features will not align with our app. This is my take.
Agreed
I have seen such thing have just made the app more cluttered it would be better if u have something in my mind d related to face tagging
Face sort
That would be better
This is my takes
I agree.
Features related to Face tagging will align with the app.
I think when you open pictopy it looks now cluttered
Yeah.
A little bit.
There is too much scope in this project
In the backend part as people pretty much stop and moves forward
So we gotta rethink
And yes one more thing
Are u familiar with electron ??
Are you asking me?
Yup
I tried to build something with electron but way back
So I think after I spend some more time
I'll be able to get back to it
It must be somewhere back in the mind
If you have used it once.
Then you catch up on it pretty quickly.
But, there are some issues with the electron.
Like?
That why they choose tauri because of it
Yeah.
There must be some way I feel
Is there library support for this
There should be.
We need to find a way to solve that problem.
We gotta research more about tauri but
Electron
Yeah.
It has huge library support.
Just like Python.
Let see how we are going to take things forwardd
Yeah.
We will find a solution if we go deep enough.
Hi guys, I don't think we can come up with a fix for so many MR's
so what we are gonna do is, we will discard all the previous MR's, and now review all line by line
in your MR's, what you can do is, add a MR: TITLE(SR. NO.)
as the project has become so big, it has become difficult to maintain the style
as there are lot of contributors
I will be active from today to tomorrow
you guys, create new MR's and ping me up here
do mention the serial number
@marsh idol @wraith creek @worldly pulsar @abstract bear @mystic sigil
and any other contributors, that are currently active
https://github.com/AOSSIE-Org/PictoPy/pull/337
I closed my previous PR and reopened it with sr. no. 1
Sure
queue, your PR after this guys
i am waiting for the pipeline, and will merge this then
pipeline meaning workflows?
for the builds?
yep
https://github.com/AOSSIE-Org/PictoPy/pull/309 closed my previous PR and reopened it with sr. no. 3
I've also added pipeline to run tests , closing reopening and renaming my PRs
do it after @abstract bear
wait for my signal
let me review his PR
ohh ok
I think you would have to close the PR and create a new fresh PR, for it to show at the top ig. Or it is the same thing ig.
same thing, ig he took the fresh main pull
doing it right away
But the workflows won't run on the older PRs I think.
they will if we merge the main branch
oh right
then @wraith creek you do yours first then
So everyone has to create a new pull request by copying all the description from the older one.
yes doing it should I close and reopen my old PR or create a new one?
new
btw rahul are you exporting PROD env variable in your workflow because your tests are failing because of that
Let me check quickly
I'll add that in this workflow PR, just tell me what to export in PROD?
I thin due to this change in you PR, the workflow is failing
import.meta.env.PROD is a variable provided by vite
while process.env.PROD is a varable that we have to define in .env file
yeah but cant do that otheriwse frontend tests would fail that's why I intentionally added vite-enviroment-variable plugin
and that needs a prod env
@radiant mauve hy i wanted to ask i am making a landing page for the pictopy where we can give our app to the end user shall i continue with it?
yes, that's a great idea
go ahead, do add AOSSIE's logo in it
thanx would update you in few daysz and would add aoosie logo on the navbaar
great!
https://debate-ai-prototype.vercel.app/ i recently made a complete ui for ai for debates ( only frontend depolyed)
@wraith creek open your PR with sr. no. 2, because workflows PRs are to be merged before the other
ok wait
https://github.com/AOSSIE-Org/PictoPy/pull/338 @radiant mauve
Once this merges, I will open my another PR.
it is merged you can go ahead
merged
@radiant mauve You are now level 14! :confetti_ball:
Fixes #161
Demo Video: https://www.youtube.com/watch?v=4EM2_lW2S54
Major Changes:
Now user can select one or more folders to display on the home page.
Added Refresh button to the home page as well...
@marsh idol You are now level 10! :confetti_ball:
The backend tests failed for the generate thumbnails function because I have made modifications to it
We have to change those tests.
also anything wrong with the build pipeline for windows?
it's failing for every PR
Yes I found the issue it was related to environment variables
@wraith creek You are now level 15! :confetti_ball:
The error was in this PR right? https://github.com/AOSSIE-Org/PictoPy/pull/275 @wraith creek
Fixes #258
These tests check for any errors and wether these components render correctly or not. We can add more tests if we have a final UI/UX for this app, for eg a test for image should scale b...
Yes I fixed that in my actions workflow now no errors are coming
Yeah need to change the tests according to your latest changes to the function ig
yep
I'll change that
@radiant mauve Do we want to build the app after every PR merge or we can set the workflow for builds as a manual trigger, which will generate builds with updated version number and detailed changelog in the release section?
Ig we can create certain milestones and release versions when those gets completed and trigger a build manually?
We can add few Prs to those version milestones
yes, that sounds good
manual trigger
else we might use up all the github resources
or the org
The windows PR check passed since the problem was fixed in the last PR, now the app should build without errors for all platforms.
Yes
I will modify the workflow for builds
@radiant mauve ,
please merge this.
https://github.com/AOSSIE-Org/PictoPy/pull/315
@abstract bear build failed for your PR's
all build tests have passed.
yeah just wait, waiting for Karn's MR
else we will continue with yours
your MR has conflicts
fix it and create a new one
working on it.
i am fixing it right away
@worldly pulsar follow after Karn
@marsh idol ,
My backend tests are not passing, But I havent changed anything in backend.
Yes, it is due to my changes in the backend
Rishab will add the new tests
it will work
So, what should I do to pass the current tests?
I think its fine for now, mainly all the Tauri build tests should pass.
Okk.
beacause the build tests checks for TS errors
https://github.com/AOSSIE-Org/PictoPy/pull/341 @radiant mauve i everything is up and running and each test is passed
@radiant mauve ,
Here's PRs with SR No.- 5
https://github.com/AOSSIE-Org/PictoPy/pull/342
and SR No. - 6
https://github.com/AOSSIE-Org/PictoPy/pull/343
All tests are passed.
Description
This PR introduces several enhancements and fixes for video playback:
Features Added:
Current Video Time Tracking: The video player now tracks and displays the current playback time in...
@radiant mauve I have disabled the auto-triggering of build workflows in my latest PR (Sr. No. 7). Also, I have modified the backend setup instructions and removed the server scripts which were not up to the mark now we are using the standard approach to run the backend, using "fastapi dev" command for the dev. Environment.
reference taken from : https://fastapi.tiangolo.com/tutorial/first-steps/
my backend is running correctly
its just blank
@radiant mauve should I put my another pending PR in queue?
Let me check I think it is due to PRs merged yesterday
found the issue in frontend/src/Config/Backend.ts file remove process.env.backend_url I removed vite enviroment plugin yesterday because windows build was facing few issues of that but forgot to remove process.env in backend.ts
I'll fix this along with my next PR
For now In this file
frontend/src/Config/Backend.ts
replace
export const BACKED_URL = process.env.BACKEND_URL || 'http://localhost:8000';
to
export const BACKED_URL = 'http://localhost:8000';
queued my PR, fixed this issue in that
https://github.com/AOSSIE-Org/PictoPy/pull/347
Fixes #259.
Improves testing and CI
i don't think that is due to that because re forked it yesterday after the merge
it worked fines
Oh, must be some other PR then
@radiant mauve I have made a PR (https://github.com/AOSSIE-Org/PictoPy/pull/336). I have followed all the guidelines. Bassam Adnan has approved my PR. Can you review it?
@hallow saffron You are now level 3! :confetti_ball:
I was trying to work on this issue
rust 🏳️
I was trying to work on an issue and i was running into an error endlessly
The error indicated that DATABASE_PATH is missing from the settings file.
I noticed that there are multiple database paths being used (FACES_DATABASE_PATH, IMAGES_DATABASE_PATH, etc.) but DATABASE_PATH is not defined.
is anything like this previously reported?
.
I just opened a PR to fix the face recognition clustering bottleneck. It adds batch processing for embeddings, improved memory usage, and various database optimizations. Some highlights:
- 98.1% performance improvement in clustering operations
- Smart batch processing for face embeddings
- Advanced database optimizations
- Real-time performance monitoring
Before: Processing 100 images took ~45 seconds
After: Same batch now processes in < 1 second!
PR: https://github.com/AOSSIE-Org/PictoPy/pull/354
Would love any feedback on the approach!
Your PRs look really good and detailed 😄
@marsh idol You are now level 11! :confetti_ball:
@hallow saffron I think you are using the run-server script to start the backend, In my recent PR which is open, I have simplified the backend setup process,please prefer to use that : https://github.com/AOSSIE-Org/PictoPy/blob/9c918b26c0d643300b36a42787693b989867ab28/README.md
can you tell, which python file was giving the error?, refer the stack trace
I think that all the database tables should be in a single .db file, currently there are different DBs for different tasks, which makes foreign key references impossible.
There are many changes which could be made to the backend
i am working on the Landing page for the pictopy I JUST COMPLETED hero.tsx and
guz any suggestions
???
yes I noticed that too, need to change that
I went through it thoroughly and it looks good and very detailed
Looks great! In PictoPy, the last letter is small
Also when I was contributing to some python libraries I noticed they had comments on every function which made it easy to understand the codebase it would be great if we add hints and comments throughout the fastapi and rust codebase
looks good to me but I may not be the right person to comment on UI/UX
Yes, I think we can make this as a good first issue
Also the PictoPy documentation is outdated, after so many changes to API and frontend
We can open up small good first issues for new contributors
yes totally agree
We also need to make a CONTRIBUTING.md file to standardize contribution process and code preferences
yess, also need to change the issue template with few checklists like for changes did you update the docs or added unit tests etc
Yep.
I think we should go ahead and create these issues with ("Good first issue") in the title after @radiant mauve sir will add good first issue tags, we can remove that
yeah sure I'll create the issue for hints and issue template
I've created them, you can create for updating docs
done
Very nice looking! 👍
@mystic sigil You are now level 8! :confetti_ball:
Looks great.
@marsh idol @abstract bear ,
I was thinking to add a script to update requirements.txt automatically whenever new PR is made.
Right now, new depedencies need to be added manually.
doesn't running the command pip install locally, update the requirements.txt automatically?
the person needs to include the .txt into their commit
No.
It doesnt.
Oh
Yeah.
So, this will ensure , that all the requiremnts are up to date.
I thought it works like npm install
thats why npm is better package manager.
how will you check for all the changes and find out if a dependency is missing?
By using .git hooks.
And github Actions .
and also local script that can run on command to update the requiremnts.txt
Okay, so you need to scan each .py file?
Or something else?
Yeah.
for import statemnts.
Or I can just scan backend conda or venv environments.
You can look into another approach also, there is a pip command which lists all the installed packages
yeah.
pip freeze
.
So, this command scans env.
But with this you can't judge which package is redundant.
@worldly pulsar I think this is what you are talking about?
https://stackoverflow.com/questions/31684375/automatically-create-file-requirements-txt
pip freeze is great when you have seperate virtual env for project.
But if seperate env is not present,
we need to think a way for that situation.
@worldly pulsar check this out:
https://github.com/marketplace/actions/automatic-requirements-txt-for-python-projects
Project:
https://github.com/bndr/pipreqs
@marsh idol You are now level 12! :confetti_ball:
This is exactly what we need.
I will start working on this immediately.
Thank You :D
I was trying to do pytest in backend folder and was running into errors which basically indicated what i wrote
Thank you :D
if i put a photo to secure folder, how do i remove it from secure afterwards?
This feature not present there
We need to implement photo removal.
Yes
If you want may work on this ...
Also i saw that we use PBKDF2 for secure folder pw
@hallow saffron You are now level 4! :confetti_ball:
What if we change that to ARgon2
What's the benefit of it?
Argon2 makes it harder to crack the pw than pbkdf2
performance wise, theres almost no significant change for the user
more of an industry standard (from what ive read on reddit and articles)
im working with ref to this issue
https://github.com/AOSSIE-Org/PictoPy/issues/349
pbkdf2 uses SHA (hashing algo) while argon2 uses Blake2b
In terms of Speed and Security I think for modern development using argon2 is better
There is one drawback also I found , in terms of security argon2 is better but in terms of system compatibility pdkdfd2 is better ...
@worldly pulsar have you any suggestion on this ?
@mystic sigil , @hallow saffron ,
I think Argon2 will be better because, it’s more resistant to brute-force and GPU/ASIC attacks than PBKDF2 due to its memory-hardness and better key-stretching.
it is their
their is an unlock button their
you can if you want to but their's not much of a need i think for this
too late already sstarted
ooohh haan thanx reminding i think i was working on this last night so made that mistake
This was the issue #349
Added secure folder functionality to PictoPy with:
- Implemented Argon2id for stronger password hashing.
- Added password strength validation (min length, uppercase, numbers).
- Enhanced file deletion with multiple overwrite passes.
- Ensured AES-256-GCM encryption with proper salt and nonce handling.
This is my PR
Is there an existing issue for this? I have searched the existing issues What happened? Current security measures for file system operations are basic and need enhancement. Current issues: - Basic ...
Fixes #349
This PR addresses the security vulnerabilities identified in the secure folder feature, focusing on encryption, key management, and access control.
Key Changes:
Argon2id Implementation...
I'd love to get some feedbacks on these changes!
Thanx bro still working on the entire page would keep updating on this
Hello everyone, I have opened a PR which fixes issue#209: https://github.com/AOSSIE-Org/PictoPy/pull/364
In this, some important changes are:
- Accessing images directly from the folder given by the user instead of copying them into the "images" folder.
- Modified the database structure to contain every table in only one .db file called PictoPy.db.
- Modified backend tests to align with the latest changes.
Please look into the PR and give any suggestions.
@mystic sigil You can now start your work on the isolation part of the issue#209.
@wraith creek You can start with the progressive loaders that you talked about. Because I have added the folders table in our DB.
Okay 👍
To access my changes, you will have to create a branch which is the copy of my branch. Because my changes are not yet merged.
yeah
Not a problem , I'll handle
Hey @marsh idol ,
How can i test the Github workflows locally?
means on your github fork or locally on a PC?
Local PC before commiting.
Got it.
I have not tried it
Thanks!
I tested my workflows on my fork.
How?
But, workflow is in new brnch and not in main branch.
@worldly pulsar You are now level 30! :confetti_ball:
temporarily you can allow triggering for any branch, you can create a test pull request on your fork
there workflows will run
You want to test which workflows?
Oh.
Automatic requiremnts.txt generation.
I will try that.
For using this, it will utilize docker
Hi community, I have contributed to other AOSSIE projects like Edu-Aid. I'm comfortable working with JS and python. Could you assign me any first issues?
@molten kindle You are now level 2! :confetti_ball:
Hy listen
Try cloning the repo 9f picto py
In the frontend navbaar color many of us doesn't like it
You can work on that good for beginners
Ok thnx
Will work on it
Also @marsh idol and I've created some good first issues you can check them out too
Hey guys! I really need help setting up locally for frontend. I have tried and When i run the environment, i keep getting white screens. How do you setup locally? please i need a guide to start contributing to this wonderful project
hey @glass seal , there is some issues with the app.
I think someone has raised a PR to solve those.
but it is not yet merged.
So, wait until it gets merged or pull that pr in your code.
hey @worldly pulsar , thanks for the update. Can i still apply for the gsoc program without contributing? I don't know when the issue might be merged
The issue is about url in that is there in Backend.ts file , go to frontend/src/config/Backend.ts instead of two url just add one url 'http://localhost:8000/'
I hope your problem will be solved , thank you.
hy @fading pilot \
recently the backend setup was changed
make sure you have python 3.12 as as all the dependencies goes with it
or you could wait like a 2 days i am a pr on how to setup pictopy locally video
sure , thank you
thanks
go on
Hey @radiant mauve !
I went over the pictopy project and it looked really interesting to me, and the tech stack resonates with mine a lot as well!
I would like to start contributing to it asap, specially the rust backend and tauri, would you be able to point me to any feature requests or issues that I can attend to?
I would potentially like to apply to pictopy for this gsoc term.
I see that electron has been proposed for the native client gui framework, which would be ported from tauri...
Why is that? Isn't tauri supposed to be cross-platform?
Moreover if the backend is written in rust, using Tauri as the gui framework would be a better choice in my opinion. Correct me if I am wrong here.
hy penguizz is the only mentor
I have previously contributed to organizations such as Graphite, https://graphite.rs/.
Even they prefer using Tauri as their gui framework of choice for their upcoming desktop app.
mentor
I see
Thats great.
Hoping to hear from you soon ^^
Yes, Tauri comes with many advantages, including better efficiency in terms of app size and RAM usage
Tauri is the best option right now.
Yes absolutely, thats the main advantage of using Tauri over electron.
@marsh idol if you don't mind, would you be open to discuss what exactly are the problems that we are facing in pictopy? And what is being done right now to address them?
Sure
So the idea page mentions there being some problems with using Rust and tauri, what is it specifically?
@young jasper You are now level 3! :confetti_ball:
It is mostly the dependency issue with Tauri, every new contributor faces problems while setting up this project
@abstract bear is working on making setup videos
We have solved the shipping issue of this project with the newly added workflows
I think the ideas page is tentative, Tauri is the framework that we should stick to.
The main issue with our project is the accuracy of the YOLO models
They are not good enough
Yes I heard about this.
I'm gonna try to work on this issue https://github.com/AOSSIE-Org/PictoPy/issues/237
@marsh idol what is the preferred way of running the project locally while testing and developing? Through the docker file or just running it as is?
Without docker
I have the project running locally without docker
With docker the experience is laggy, because you need to connect to a ip address for remote access to the window.
have you guys ever considered using poetry as a package managing tool?
it can be used for both virtual environment management and package management, and sharing poetry projects is very easy as well
@worldly pulsar
I'm unable to play any video through the app, is this a bug in the upstream branch or is it just me facing this problem?
I think the issue they're referring to in the project idea is we need to install different types of packages and dependencies on different platforms like on windows we need to setup visual studio c++ tools and on linux we need to install other packages
But I agree tauri has lot more advantages, this is a very small issue, i think in repo their already exists few scripts to install platform related dependencies for windows and ubuntu/debian based systems
So I too think porting from tauri to electron is not such a great idea
Yes it is issue on linux, you need to install some packages for it. On windows it works fine
we can simply put up a PR to fix it and setup will be a breeze then, so yeah setting up is not a issue
hmm. I am using arch and I installed all the packages which were mentioned in the install scripts
had to do it manually because the packages are named differently in ubuntu and arch, but I believe I installed all of them
There are specific packages for playing videos, you can check console by right clicking and then inspecting. You may see some errors when you switch to videos page
When I run cargo tauri info I get this output, seems like everything is installed properly here
❯ cargo tauri info
[✔] Environment
- OS: Arch Linux Unknown x86_64 (X64) (KDE on wayland)
✔ webkit2gtk-4.1: 2.46.6
✔ rsvg2: 2.59.2
✔ rustc: 1.85.0 (4d91de4e4 2025-02-17)
✔ cargo: 1.85.0 (d73d2caf9 2024-12-31)
✔ rustup: 1.27.1 (54dd3d00f 2024-04-24)
✔ Rust toolchain: stable-x86_64-unknown-linux-gnu (environment override by RUSTUP_TOOLCHAIN)
- node: 22.14.0
- yarn: 1.22.22
- npm: 11.1.0
- bun: 1.2.4
[-] Packages
- tauri 🦀: 2.0.0-beta.23
- tauri-build 🦀: 2.0.0-beta.18
- wry 🦀: 0.41.0
- tao 🦀: 0.28.1
- tauri-cli 🦀: 2.3.1
- @tauri-apps/api : 2.2.0 (outdated, latest: 2.3.0)
- @tauri-apps/cli : 2.3.1
[-] Plugins
- tauri-plugin-fs 🦀: 2.0.0-beta.10
- @tauri-apps/plugin-fs : 2.2.0
- tauri-plugin-dialog 🦀: 2.0.0-beta.10
- @tauri-apps/plugin-dialog : 2.2.0
- tauri-plugin-store 🦀: git+https://github.com/tauri-apps/plugins-workspace?branch=v2#66e8dbfc1a5cbeff3846d9ccffb57c45eb0de51e (2.0.0-beta.9)
- @tauri-apps/plugin-store : not installed!
- tauri-plugin-shell 🦀: 2.0.0-beta.8
- @tauri-apps/plugin-shell : 2.2.0
[-] App
- build-type: bundle
- CSP: unset
- frontendDist: ../dist
- devUrl: http://localhost:1420/
- framework: React
- bundler: Vite
Inspecting the console gives this output
[Error] Unhandled Promise Rejection: NotSupportedError: The operation is not supported.
play
togglePlay (NetflixStylePlayer.tsx:69)
handleContainerClick (NetflixStylePlayer.tsx:147)
callCallback2 (chunk-RPCDYKBN.js:3674)
dispatchEvent
invokeGuardedCallbackDev (chunk-RPCDYKBN.js:3699)
invokeGuardedCallback (chunk-RPCDYKBN.js:3733)
invokeGuardedCallbackAndCatchFirstError (chunk-RPCDYKBN.js:3736)
executeDispatch (chunk-RPCDYKBN.js:7014)
processDispatchQueueItemsInOrder (chunk-RPCDYKBN.js:7034)
processDispatchQueue (chunk-RPCDYKBN.js:7043)
dispatchEventsForPlugins (chunk-RPCDYKBN.js:7051)
batchedUpdates$1 (chunk-RPCDYKBN.js:18913)
batchedUpdates (chunk-RPCDYKBN.js:3579)
dispatchEventForPluginEventSystem (chunk-RPCDYKBN.js:7173)
dispatchEventWithEnableCapturePhaseSelectiveHydrationWithoutDiscreteEventReplay (chunk-RPCDYKBN.js:5478)
dispatchEvent (chunk-RPCDYKBN.js:5472:93)
dispatchDiscreteEvent (chunk-RPCDYKBN.js:5449)
I'm not able to make out what package I'm missing 🤷♂️
No that package is not part of the rust dependency, you need to install it with your distros' package manager. I forgot that package's name on Ubuntu
@marsh idol You are now level 13! :confetti_ball:
You should check the console when you are running the app, maybe you can find some error being thrown, I was able to find that error via console of the app on Ubuntu
I understood that part, I was just making sure that tauri is running as it should
@young jasper You are now level 4! :confetti_ball:
this is the error message I get in the console, and its not helping much 🤷♂️
[Error] Unhandled Promise Rejection: NotSupportedError: The operation is not supported.
play
togglePlay (NetflixStylePlayer.tsx:69)
handleContainerClick (NetflixStylePlayer.tsx:147)
callCallback2 (chunk-RPCDYKBN.js:3674)
dispatchEvent
invokeGuardedCallbackDev (chunk-RPCDYKBN.js:3699)
invokeGuardedCallback (chunk-RPCDYKBN.js:3733)
invokeGuardedCallbackAndCatchFirstError (chunk-RPCDYKBN.js:3736)
executeDispatch (chunk-RPCDYKBN.js:7014)
processDispatchQueueItemsInOrder (chunk-RPCDYKBN.js:7034)
processDispatchQueue (chunk-RPCDYKBN.js:7043)
dispatchEventsForPlugins (chunk-RPCDYKBN.js:7051)
batchedUpdates$1 (chunk-RPCDYKBN.js:18913)
batchedUpdates (chunk-RPCDYKBN.js:3579)
dispatchEventForPluginEventSystem (chunk-RPCDYKBN.js:7173)
dispatchEventWithEnableCapturePhaseSelectiveHydrationWithoutDiscreteEventReplay (chunk-RPCDYKBN.js:5478)
dispatchEvent (chunk-RPCDYKBN.js:5472:93)
dispatchDiscreteEvent (chunk-RPCDYKBN.js:5449)
still im trying to read about it on some forums if I can find the cause of this issue, maybe it is some missing package that I am not able to locate
@young jasper , I was working on adding a script to generate requirements.txt.
But got busy with other work.
if you want, you can work on it.
my idea on it was, if we use poetry, we never have to worry about requirements.txt
Will you explain a bit, how it works?
whenever you add a python module using poetry its lock file is generated, anyone with that lock file only needs to run poetry install to get all the dependencies installed
So, we need to use poetry instead of pip, right?
so like npm's package.json, the .toml file will get automatically updated whenever someone installs a package?
yes
because pip doesn't update the .txt file
?
something like this
[tool.poetry]
name = "dbot"
version = "0.1.0"
description = ""
authors = ["Your Name <[email protected]>"]
readme = "README.md"
package-mode = false
[tool.poetry.dependencies]
python = "^3.10.12"
discord = "^2.3.2"
python-dotenv = "^1.0.1"
discord-py-slash-command = "^4.2.1"
tk = "^0.1.0"
[build-system]
requires = ["poetry-core"]
build-backend = "poetry.core.masonry.api"
That would be great.
then poetry is better option, we don't need to have workflows
yeah.
yes, for example-
poetry add <package>
poetry run python3 main.py
then what about virtual environments?
the dependencies are automatically added in this file when you use poetry add, and lock file can be generated with poetry lock
we need to activate that right?
can Poetry add and change package-lock.json when new package gets added?
poetry install creates a virtual env in ~/.local dir
we need to activate it?
then it can either be activated with poetry shell which activates the env or you can add poetry run before any command and its ran via the virtual env
Hey everyone, this is really interesting! I’ve heard a lot about Poetry but haven’t had much hands-on experience with it yet. One thing I’m curious about—how does Poetry handle version conflicts between dependencies? For example, if two libraries require different versions of the same dependency, does Poetry resolve it automatically, or do we have to intervene manually? Also, how does it compare to pip-tools in terms of managing lock files and resolving dependency trees? Would love to hear your thoughts
I've been using poetry for quite some time now and it can even be added to github workflows
thats great.
It can be set to trigger on every push.
2 versions of same package cannot exist, in the same environment that is
each project has its own envrionment generated by poetry
and as for comparison to pip, I would say I have yet to encounter a situation where pip was better than poetry
it automatically resolves dependencies and installs them with whichever package you require it to install
Got it @young jasper! but one more thing as far as i know pip can reuse packages from a shared system cache, while Poetry installs everything fresh for each project, doesn't this lead to redundant storage and slower setups?
poetry can retrieve packages from shared cache as well
@young jasper You are now level 5! :confetti_ball:
Although I have encountered a bit more storage being used with poetry, but it is not much
And when it saves you from the headache of falling into dependency hell, in my opinion it is worth it.
Ohh...then this sounds like a great initiative we could take!
Poetry increases reproducibility, in the end that is what matters more.
Should I help you with this? If you think I could contribute in any way, let me know about this. I've heard about Poetry before but haven't had the chance to work on it with someone—would love to explore it together! 🙂
sure
but I think moving to poetry would be the maintainer's call?
yep
@radiant mauve sir's call.
Yeah, obviously, only after the maintainers give the green flag for this.
should I make a issue on this untill then?
yess
go ahead.
Hi mentors,I hope you're doing well. I'm a first-time participant in GSoC and excited to contribute. I would love your guidance on understanding the organization’s projects, contribution process, and best practices.
Could you please share any tips or resources to help me get started? I’d really appreciate any advice on how to make meaningful contributions.Im specially interested to contribute to Pictopy @radiant mauve can u help me
You can check out good first issues #355 and #356
Also, you should setup the repository using these instructions:
https://github.com/AOSSIE-Org/PictoPy/blob/97286a91a395200e80e532ff2b11672a4107f1dd/README.md
don't refer the readme from the repository directly, the above link has updated readme file
Thank you so much for your guidance!Rahul I’ll start working on it and will keep you updated on my progress. Really appreciate your help! 😊
Before starting the work on any issue, write a comment on its thread so that somebody else also doesn't start to work on it.
Hey @marsh idol , I already commented on the issue and just saw your message. I really want to align with issue #356—I haven’t asked for #357, so if someone else wants to take it, they can. But I did notice someone had asked for that issue before too, so can I go ahead with #356? Also, I’m new to Pictopy, but I’m a Master Contributor for the Internet Archive. I’d love to work on these cool ideas, so it would be great if you could assign me that issue!
@round laurel You are now level 2! :confetti_ball:
I have asked ritik for any updates, if he doesn't answer, you can start the work.
Sounds good, @marsh idol ! Let me know if Ritik responds. If there's no update, I’d be happy to take it up
yes Rahul done I commented on the issue #357 and started working on it
That issue has already been taken care of i.e. a PR is open for fixing this issue.
okay @marsh idol i will find out another issue I can contribute to
Hey @marsh idol I saw issue #355 (Adding comments throughout the codebase). It looks like a simple task—can I take it up? Let me know if it's available!"
Yes
comment on its thread
how does the AI tagging here work?
I have loaded a couple of folders but nothing is coming up in the AI Tagging section
Btw, @marsh idol @worldly pulsar I raised an issue regarding the package management we were talking about. https://github.com/AOSSIE-Org/PictoPy/issues/374
hy you can recent sr .no write that in you pr
when you make a pr
I didn't get it.. what did you mean?
refer to the above message
from the mentor
this one
this one
So should I name my SR. NO. after the latest merged PR or the most recent PR raised
Yeah.
That would be better.
Hy @rahul
I want to add download files on the landing page
Provide me those link
@marsh idol
Paaji jalde thoda
Sun x64 windows hai arch is Linux and third one Mac?
.dmg wala mac, .deb for ubuntu and .exe windows
This weekend
Ping me up