#Pictopy
1 messages · Page 6 of 1
Hey @gaunt heath ! "Not planned" PRs didn't align with our direction. To stay on track:
Check good first issue or help wanted labels
Confirm with @marsh idol sir before starting any work
I've just checked, it's a good feature idea: let users share albums via secure token links. Generate tokens with secrets module,
store in shared_links table with optional password + expiration. Add route to validate
tokens with rate limiting to prevent abuse.
Now, Rahul sir will confirm whether it's aligned or not
while exploring the app, I noticed a few UI issues that affect the overall look and feel
https://github.com/AOSSIE-Org/PictoPy/issues/748
I thought of creating a feature request in github but again as rahul sir will be confirming so ...
Hey@marsh idol,
Sir Quick clarification about issue #725.
I had mentioned my intent to work on it when I opened the issue, but noticed a few PRs were opened before any assignment. Just wanted to confirm the usual workflow here — should contributors wait for assignment, or is it first-PR-wins?
https://github.com/AOSSIE-Org/PictoPy/issues/725#issuecomment-3649095342
@young torrent You are now level 4! :confetti_ball:
Hey , I am Vanshika Agrawal , I'm new to open source , and I've been exploring your projects.I'd love to contribute.
Hi everyone, I'm Swastik, a second year Computer Science student from Delhi, India. My skills are Javascript, Typescript and Python. I've setup Pictopy locally. If there is any unassigned issue with which I can start contributing, please let me know. I'd be happy to contribute.
start by reading README file
Hi, I am shivendra Prasad , a passinate CSE student from India, i am new to open source and love to contribute to you project Pictopy , i will try to add another aspect to this project. Thank you
probably you don't have linker for rust so need to install Build Tools for Visual Studio 2022 bcz this provides you a linker
Is there any project mainly focused on a website?
you can check out #1345044736515379210 or even #1311271974630330388, both are built on web technology so you can checkout them
Hi everyone 👋 I’m Gowri Shankar, new contributor here.
I’d like to work on BUG #705 related to metadata being returned as a JSON string.
https://github.com/AOSSIE-Org/PictoPy/issues/705
I plan to deserialize the metadata properly and add unit tests as suggested.
Please let me know if I can proceed or follow any preferred approach. Thanks!
Other Projects are not that active
Pictopy is great but it's really hard to start ....
@tranquil gulch You are now level 5! :confetti_ball:
Yes, once you set it up correctly, you’re good to go. If anyone faces issues during setup, they can ask the community, and the members will guide them.
I am talking about contributing stuff hard to standout ...
@warm glen You are now level 2! :confetti_ball:
yes sir i have succesfully installed whole project and now i can run this project it's amazing i solved the problem
and now i will go to contribute like solving issues
i want to open landing page can anyone help me how to start it?
hey gowri, good to see your interest in this issue, but already 3 pr raised on this issue
see is there any existing issue raised that you can solve, or if you can solve any pain point of the maintainer that would be really helpful
Hi everyone 👋
I’m Vatsal, a 3rd-year bachelor’s student , Joining Swiggy as an AI/ML Intern and an active open-source contributor.
I’ve been going through projects under the AOSSIE group, and InPactAI, PictoPy, and BabyNest really stood out to me in terms of impact and alignment with my interests in AI/ML and applied systems. I’m planning to prepare early for GSoC 2026 and would love to start contributing meaningfully ahead of time.
If there are any maintainers or contributors here, I’d really appreciate guidance on:
- beginner/intermediate level issues or areas to start with
- recommended issues and features i can contribute to
Hey Vatsal! congrats on the Swiggy internship!
For getting started with InPactAI, PictoPy, and BabyNest, here's what I'd recommend:
Check out the repos' issue trackers for labels like good-first-issue or help-wanted
Go through the CONTRIBUTING.md/ Readme.md files to understand the workflow
Set up the projects locally and familiarize yourself with the codebase
Review recently merged PRs to see what kind of contributions are being accepted
For specific guidance: The maintainers usually hang out on the AOSSIE Discord/Slack—drop a message
@marsh idol Sir, as requested, I’ve shared a short video demonstrating the bug, followed by a video showing the fix with the proposed solution, Whenever you have time, please take a look.
(https://github.com/AOSSIE-Org/PictoPy/issues/706#issuecomment-3651740450)
Thanks for the clarification 😊
Since this issue is already addressed, I’d be happy to work on another simple issue.
Could you please suggest the one that would be suitable for a new contributor?
Thank you!
Hello everyone!
I am Aaryan a FE CE(Computer Engineering) I am preparing for Gsoc and found PictoPy project very intruiging I would like to start contributing I have read contribution.md and setup all the necessary dependencies required for this project. I cannot personally ping the maintainers so I request maintainers to assign me issue
Thank You.
Hey @frank quest since Rahul sir might be busy, it might be best to wait for their input first and then decide whether to open an issue.
hey @old yarrow Usually, issues aren’t directly assigned on request — it’s best to comment on an open issue you’d like to work on (especially a good-first-issue) and mention your approach. Maintainers can then take it forward when they review.
https://github.com/AOSSIE-Org/PictoPy/pull/781
@marsh idol please review my PR for the issue - memory implementation page .
I have also attached a screenshot after implementation .
That’s fair. My thought was mainly to gather community feedback first so I am clear and more aligned with the idea before Rahul sir or any other maintainer weighs in.
Thanks for the reply!
@marsh idol please review my PR for the issue - added profile page and logout button.
I have also attached a video after implementation .
Hi @radiant mauve !
myself Aaryan (SnippyCodes on GitHub), I've expressed interest in working on Issue #640 (CORS restriction) on GitHub. I've shared my proposed fix there. Could you please check it and assign the task to me? Thanks!".
@old yarrow You are now level 2! :confetti_ball:
I have also tested the necessary commands...
Hey @marsh idol — PR #792 fixes the logging RecursionError crash. All checks pass & ready for review
@marsh idol Sir,
I have raised this issue: https://github.com/AOSSIE-Org/PictoPy/issues/795
The issue highlights that PictoPy currently does not have:
A direct image-level deletion workflow, and
any undo mechanism to restore images after accidental or destructive actions.
I wanted to confirm whether this problem aligns with the project’s priorities and whether it would be appropriate for me to start working on it.
I would really appreciate your feedback on the scope and relevance of this issue before proceeding further.
Describe the feature Overview PictoPy currently lacks a direct image-level deletion workflow and provides no undo mechanism for destructive actions. This feature proposes a safe image deletion syst...
Hello everyone,
My name is Sonu Chowdhury, a third-year engineering student from Kolkata, India. I am exploring open-source projects for GSoC 2026 and came across the Pictopy project, which aligns well with my interests and technical skills. I am excited about the opportunity to contribute and look forward to collaborating with the community.
Hello everyone 👋
I’m Anvesha Srivastava, a BCA 2nd-year student, and I’m aiming to prepare for GSoC 2026.
I’m particularly interested in contributing to PictoPy and understanding the project in depth.
I work primarily with Python and I’m actively improving my skills in DSA and open-source workflows.
I’d love to start with beginner-friendly issues and gradually take up more responsibility.
Looking forward to learning and contributing—thank you!
I have solved two issues in this project.
Could someone please guide me on what I should do next?
Hello maintainers,👋🏻
I’ve raised an issue related to UI layout and responsiveness on the PictoPy documentation homepage:
https://github.com/AOSSIE-Org/PictoPy/issues/796
The issue highlights some remaining concerns around content alignment, readability, and overall responsiveness that could improve the user and contributor experience.
I wanted to confirm whether this aligns with the project’s current priorities and whether it would be appropriate for me to start working on it.
I would really appreciate your feedback on the scope and relevance of this issue before proceeding further.
Thank you!
What's wrong with the existing documentation Description The documentation homepage currently has several UI/UX issues that affect readability and overall user experience, especially on differe...
@marsh idol Sir,
I have raised a PR for Issue #795: https://github.com/AOSSIE-Org/PictoPy/pull/801
This PR implements image-level deletion with a safe and user-friendly workflow:
-Images can be deleted individually or in bulk
-Deleted images are moved to Recently Deleted
-An Undo popup is shown immediately after deletion
-Images in Recently Deleted can be restored or permanently deleted
I also attached the demo video in the PR.
Kindly review the PR and share your feedback. Thank you.
Hey @distant thorn how are you? Its been quite some time
How's it going guys
Hello everyone,
I noticed that on the landing page where users can download the project, there is no footer.
Because of this, users don’t have a direct way to contact the company or learn more about it.
So, I have raised this issue (issue #803).
Hey everyone
I’ve fixed the performance issue in the /images/toggle-favourite endpoint where it was fetching all images and iterating (O(n)).
Updated it to directly fetch the required image (O(1)) and raised a PR for review.
Would appreciate it if someone could take a look https://github.com/AOSSIE-Org/PictoPy/issues/805
can anyone just tell me is this issue i raised is correct or not?
The server’s been pretty lit these past few days, apparently.
i was also inactive these days
good rn exams are going on
Hello everyone!
I’ve implemented a fix for this issue by expanding the hover and click hit area around the image navigation arrows.
The solution introduces a larger invisible interaction zone while keeping the arrow visually compact. When the cursor approaches the arrow, it highlights and slightly scales, and clicking anywhere inside the area triggers navigation — similar to native photo viewers.
PR: https://github.com/AOSSIE-Org/PictoPy/pull/808
Feedback welcome 🙂
Heyy ,
@marsh idol
i am now open to work
if there is some issue that i can solve please let me know
Heyyy @marsh idol , please let me know if I can contribute on any issue, would love to contribute.
@marsh idol Sir,
In PictoPy, image paths are selected from the local file system. If a user deletes an image directly from the file explorer, the database still retains the old image path and PictoPy continues to show it. This leads to data inconsistency between the filesystem and the database.
like this broke the sync so, I can start working on fixing this issue?
@eternal venture You are now level 2! :confetti_ball:
Hello everyone 👋
I’ve joined the PictoPy community and have successfully completed the local project setup.
Looking forward to contributing and learning from you all!
Hi @marsh idol Sir & everyone!
I’m Vansh 👋. I’m a computer science undergrad who enjoys working with Python and getting involved in open source. Pictopy really caught my attention because it’s a clean, Python-based project with practical use, and it feels like a great place to learn while making meaningful contributions. I’m excited to get involved and would love to know where I can start contributing.
@marsh idol and others as well,
I have opened this issue https://github.com/AOSSIE-Org/PictoPy/issues/793
while solving what i did was i tried chaning few parameters and wrote few new functions
eps = 0.2 for tighter clustering
min_samples = 1 to include as many faces as possible
conf_threshold = 0.5 for better detections
similarity_threshold = 0.8 for conservative assignment
FaceNet_util_is_face_quality_good() : Rejects blurry, partial, or too-small faces
I tested these changes using around 200 images of me and my friends, and the results improved significantly. Previously, the system was detecting two different people as the same person, but now its able to separate them correctly.
However, when I tested with images from a Kaggle dataset (about 50 images each of two people), the system unexpectedly combined the two individuals as one, even though they were being detected as separate previously with original configurations.
Then, I placed all the images (both from my friends and the Kaggle dataset) into a single folder and ran the clustering again. This time, the results were as expected, the system correctly identified them as separate individuals.
I am trying to understand why the performance differs when the images are in separate folders vs when they’re combined in one folder. Any insights would be greatly appreciated!
Hi Sid, how're you doing.
Lets try and resolve this issue...
Your observation matches quite a lot with this paper: https://www.sciencedirect.com/science/article/abs/pii/S0020025521008367
Global vs. Local Density problem
The paper discusses how "bridge points" (noisy images/ outliers or low quality faces that sit in the space between two clusters) cause distinct clusters to merge
Also, according to the paper, having min_samples = 1 makes the system extremely vulnerable to "chaining" ||(the merge problem)||
Could you try increasing min_samples from 1 to 3?
Does that make a difference?
The Scikit-Learn documentation https://scikit-learn.org/stable/modules/clustering.html#dbscan itself warns that min_samples = 1 is too low for real-world face data
Those are some great inputs @real gale , Thanks a lot, I'll give it a go and keep you updated!
@real gale @distant thorn guys can you tell me in which issues are you **contributing ** ?
Hey Guys ! I’ve set up the repo locally.
I noticed most of the good-first-issues are already taken,
Is there any area you’d recommend I start with?
Yes @flint token you can go through this good first issue.
https://github.com/AOSSIE-Org/PictoPy/issues/828
rn, exams are going on
but i was contributing in other projects
@real gale That helped a lot, those were great insights, thanks 🙌
That along with few more changes solved that problem
Hi everyone! 👋
I’ve set up the project locally and gone through the docs to understand the workflow.
I’m excited to start contributing and was wondering if you could suggest any good first issues or beginner-friendly tasks where I can add value and learn the codebase better.
Thanks in advance!
So Happy , can you please assign me this issue , i've started working on this
Niceee
Hey i just solved this issue and created my first PR , can you guys help me with getting this merged
Hi everyone! My college is now completed and I am also done with most of the hackathon's work. Glad to be back and see many new contributors. Please be patient as I review all the remaining PRs and review the new issues. I am the sole maintainer for this project right now 😅
I primarily do this work in my free time, so please expect a slow review and response time. Let us take this project to new heights 💪
I am planning to schedule a Google Meet in the upcoming weekend to interact with the community. Will 10PM IST be fine for everyone?
Congratulations!!!🥳
We’re so glad to have you back
Congratulations sir 👏🏻🎉
That would be such a great initiative
We would love to interact with you !!!
are u talking about devr.ai of aossie or something else
Congratulations!
smth else
🥶 🚀
Good afternoon @marsh idol Sir 👋
I hope you’re doing well. I’ve opened my first PR in PictoPy to improve the documentation homepage layout and sidebar spacing, focusing on better readability, spacing, and overall visual hierarchy.
🔗 PR: https://github.com/AOSSIE-Org/PictoPy/pull/834
The changes are limited to documentation UI/styles only. I would really appreciate your feedback or any suggestions for improvement. I’m happy to iterate and learn.
Would really appreciate your feedback. Thank you! 😊 🙏
@severe haven You are now level 2! :confetti_ball:
Thank you for highlight my issue i am uploading one more good first issue
Ccongratulations!!🔥 🥳
@warm glen You are now level 3! :confetti_ball:
Hey @marsh idol Sir, I’ve just opened a PR to migrate the framer-motion dependency to the recommended **motion **package.
PR link : https://github.com/AOSSIE-Org/PictoPy/pull/838
Hello everyone! I’m new to this server and interested in contributing to this project. I’ve worked with YOLO for object and face detection in my projects, and I believe that experience could be useful here. Could you please guide me on how to get started and where I can contribute effectively?
Hi everyone, I’m Prathiba.
I’m setting up Pictopy locally and plan to start contributing with small bug fixes and documentation improvements.
Happy to collaborate and learn from the community.
Hey everyone!
I'm Rudra, a student passionate about AI/ML and Python development. I recently discovered AOSSIE through GSoC and I'm really excited about the projects here!
I'm really interested in PictoPy and I've already cloned the repo and gone through the codebase. Looking forward to contributing and learning from this amazing community!
Would love to start with a beginner-friendly issue if anyone can point me in the right direction. 🙂
Thank you so much!
Merged the 1000th commit to this project just now 🔥
Congrats to everyone 🎉👏🏻, huge milestone for the project
@marsh idol Sir,
In PictoPy, image paths are stored from the local file system. If a user deletes an image directly from the file explorer, the database still retains the old path, and PictoPy continues to reference it. This breaks synchronization between the filesystem and the database and leads to data inconsistency.
I wanted to confirm whether this is a valid issue to work on and get your feedback before proceeding with a fix. I already raised issue - https://github.com/AOSSIE-Org/PictoPy/issues/812
Kindly please have a look when you get a chance.
This is already addressed by the sync microservice.
Okie Sir thank you to clarify !!
Hello @marsh idol sir!
I’ve been exploring the PictoPy codebase and I’m keen to start contributing. I’m currently going through the project structure and existing issues to understand the workflow.
I wanted to check what the preferred approach is for contributors, should I start by picking up existing issues, or is it encouraged to raise new issues for bugs or feature ideas identified during exploration? I’d appreciate any guidance on high-impact areas to focus on early.
Thank you!!
Hey @marsh idol, code is ready for
https://github.com/AOSSIE-Org/PictoPy/issues/811
https://github.com/AOSSIE-Org/PictoPy/issues/837
—can I submit PRs?
Describe the feature I’ve observed that when scanning large folders (e.g., 100+ photos), the frontend remains stuck on “No Images to Display” until the entire process completes — even though the ba...
Is there an existing issue for this? I have searched the existing issues What happened? Description Currently, running the app requires multiple manual steps: cd backend && python main.py c...
Hello @marsh idol sir! The album feature has been added! please take a look at my PR. I have successfully implemented all the necessary features. Feedback is most welcome https://github.com/AOSSIE-Org/PictoPy/pull/851
Yeah this is from the hackathon right?
Crazy
Unfortunately nope
How was the crowd, even I am intruiged 🤔
How were the people
i also wasn't there
Good morning @marsh idol sir pls address this one so I can process further 🙏🏻
https://github.com/AOSSIE-Org/PictoPy/pull/814
Hey @real gale can you please review the changes
It was great! We got 50 new project submissions and 650+ new pull requests
around 220+ attendees
Hi all!
Myself Varsha, and I’m a CS student preparing for GSoC 2026.
I’ve selected PictoPy as the project I’d like to contribute to, and I’ve started exploring the docs and codebase.
I’m excited to learn, contribute, and be part of the community.
Happy to be here and looking forward to collaborating!
Hi everyone !!!
I’m Phanindra, a CSE (Cyber Security) student preparing for GSoC 2026.
I’m here to learn open-source workflows, understand large codebases, and start making meaningful contributions. I’m currently focusing on building strong fundamentals, reading project docs, and preparing to pick up beginner issues.
Looking forward to learning from the community and collaborating where I can add real value.
Hey folks! 🙌
I’m Shreya, learning the MERN stack and starting my open-source journey 😄
I’m a beginner, but I love digging into problems and trying to solve complex issues step by step 🧩🔥
Super excited to collaborate, learn from everyone here, and contribute as much as I can 🚀
Glad to be here!
wow 😳
Hi everyone ,
I’m Prakhar Agarwal, a student learning Web Development and preparing for GSoC 2026.
I’m exploring PictoPy and would love to start contributing through beginner-friendly issues.
Looking forward to learning and collaborating with you all .
Hey @unreal quartz I've tried reviewing the PR, kindly do the needful, otherwise good job 👍
@real gale You are now level 20! :confetti_ball:
Hi @marsh idol, the Advanced Zoom on Scroll PR is also ready for review now, and this is a very high value PR and an end user consumer facing improvement. I'd really appreciate and request you go through it, try the feature out...
I've tried implementing the feature in as much depth and with as much precision as possible. Refined across as many edge cases I could think of/ stumbled upon while working on this. centering, anchoring, bounds, panning safety, cursor feedback, etc
You may find the PR here. It's end user ready, properly reviewed ||by AI 💀 (coderabbit, copilot...)|| many times.
Hi all , 👋
Iam , Shashank Tandan, a student developer from India .
I have started setting up the pictopy repository locally and going through the code base . Please let me know if there are any issues or areas where I can start contributing . Happy to collaborate and learn from all of you !
https://calendar.app.google/Qotu5ZTvuNPh6bqHA
Hi everyone, I have planned weekly meets for maintaining this project. Today will be our first meet.
Hello Rahul Sir,
I’m new to open source. I’ve raised a few issues and also submitted some PRs.
I just wanted to ask what I should focus on next and whether I’m going in the right direction.
Any guidance or feedback would really help me.
Gentle Reminder for the meet.
facing this issue while reinstalling
@marsh idol https://github.com/AOSSIE-Org/PictoPy/issues/871
@marsh idol would you like me to share the MoM (Minutes of Meeting) here in the chat?
yea pls, i joined very late
Sure
https://docs.google.com/document/d/1ZkDr6gQjPyImEwD-UJ86zXMd3S6F5lFG6tUsSBcdNzk/edit?usp=sharing
here you go chat :)
PictoPy Weekly Meet 10:00 PM 27-12-2025 Key Takeaways Release Deadline: Pictopi version 1.1.0 must launch by Monday or Wednesday, targeting cross-platform usability and core features. Critical Bugs: Current version unusable for Windows and Linux; 1.1.0 aims to resolve major issues and enhance f...
This is a great summary, how did you make it? 😄
damm bro
Firefly 💀
I think everything is neatly covered here.
Damn.
True
Great tool
From now on, you are sending these notes for every meet.
Everyone who attended the meet, please read this meeting summary as well. It was great meeting you all and I hope to meet you all again soon 😄
@marsh idol please review this PR https://github.com/AOSSIE-Org/PictoPy/pull/801
Hello @marsh idol sir, while stepping up the app Picotpy, came across these few issues
During the initial set-up, the back button doesn't seem to work
So if i want to go to step 2 and want to change my avatar or name, I can't
And the second issue is that.. there is step 4/3 instead of vice versa 😅
Hey @marsh idol, I couldn’t join yesterday’s meeting due to an exam today, but I’ve gone through the notes and am fully aligned with the action items.
Could you please review my PRs when possible?
#706 is still under triage
#713 and #707 are awaiting review
#811 is nearly complete and will be ready for review soon
i would also like to contribute
My issue was not mentioned because I was writing my issue on the chat during the meeting.
No issues, please share it here in the chat, i’ll add it in the document.
Sure!!
I wanted to highlight a few issues I’m currently facing in PictoPy:
The Settings section is very slow, especially while checking folder status. It takes a lot of time to load.
Folders are not getting deleted properly, and currently any type of folder (images, code files, etc.) can be uploaded without restriction.
Due to the slowness, uploading another folder often fails or gets stuck.
Additionally, I have already raised a PR for image-level deletion:
After deleting images individually or via multiple selection, they move to the Recently Deleted tab.
From there, users can permanently delete the images as well.
I also explained this PR during yesterday’s PictoPy meeting.
Also add what sir gives reply for the same .
@real gale
Good morning @marsh idol sir yesterday I was busy with my odoo hackathon that's why, I was unable to attend the meet .
Next time I will make sure to attend it.
Great. I’ve added it in the doc
Issue no. 638 and 764 already address this.
I will currently only review those issues and PRs which are on the high priority for the 1.1.0 release.
Can you open an issue for the slow settings page, with detailed step by step guide on how to reproduce it? As suggested in the Google Meet also.
Okie I'll !!
@eternal venture You are now level 3! :confetti_ball:
Okie , same doc ??
As my exams are ongoing, so i'll start working on issue #871 after 1st jan.
https://github.com/AOSSIE-Org/PictoPy/issues/748
is anyone else able to reproduce this issue? On my system it seems fine.
yes @marsh idol sir i have opened it and in my pc i faced this #748 issue so i have submit a pr
and here is my all issues which i have raised in past times :-
1.https://github.com/AOSSIE-Org/PictoPy/pull/857#issue-3762926393
2,https://github.com/AOSSIE-Org/PictoPy/issues/828#issue-3757112701
3.https://github.com/AOSSIE-Org/PictoPy/issues/803#issue-3747567198
📝 Description On the landing page where users can download the project, there is currently no footer section. Because of this, users do not have an easy way to contact the company, access social li...
Is there an existing issue for this? I have searched the existing issues What happened? Description I noticed a UI/UX issue in the button section of the website that affects both visual clarity and...
Fixing - #856
Description
The issue where the Home button in the navbar was not functioning on the landing page has been fixed.
Previously, clicking the Home button did not scroll or redirect the u...
Share a screen recording please.
I know about these issues, you reminded me in the meet also. Please be patient, I am reviewing them.
Guys please make this one thing very clear, DO NOT open a PR for an issue which is not yet reviewed by me. What do you think, all the issues you report are always valid? NO.
You are wasting yours as well as my time. I have to spend hours cleaning up the mess made by the new incoming contributors. I might be sounding frustrated, it is because I am. I can anytime ban a user for opening too many irrelevant PRs for their irrelevant issues. Open as many issues as you like, but first let me review the issue and THEN YOU START THE WORK ON IT AND OPEN A PR.
Make this thing very clear.
ig you should pin this msg
Yes, I think I am not having that access to pin the message.
@marsh idol
The “Windows protected your PC” warning is controlled by Microsoft Defender SmartScreen at the OS level and cannot be removed through backend or application code. Unsigned executables will trigger this warning, and standard signed executables may show it initially. To avoid the warning on first run, the application must be digitally signed using a trusted Certificate Authority, preferably with an EV Code Signing certificate. MSIX packaging alone does not eliminate this requirement.
@marsh idol If we want to avoid any popup notification, the application must be digitally verified using a trusted code-signing certificate; there is no alternative technical solution.
Hello Rahul
I have finally been able to implement the proper shutdown task we talked about yesterday...
Below is a screen recording of everything in action - displayed alongside task manager for proof.
There's one thing about graceful shutdown though... The frontend itself seems to be talking some time (~2 seconds) to close after the close button was pressed....
This happens because frontend will need to wait for “shutdown accepted" response from the backend to make sure backend is shutting down and that the frontend doesn't ever just mistakenly be like "okay I've sent the shutdown message, now I can close, now its backend's headache it closes or not". In this case, in case backend doesn't get the "Shutdown" message it may keep running in the background...
Is this wait okay? Or do you want me to force ki-
There is one thing we can do as a workaround , if we publish the app on the microsoft store it automatically signs it and probably this will remove that warning
@frank quest You are now level 5! :confetti_ball:
Not completely sure but for open source projects I think microsoft does provide free tiers for uploading applications on the store
https://github.com/AOSSIE-Org/InPactAI/pull/253 the PR is ready for review
over there also they ask for verified certificate.
@drowsy parcel You are now level 6! :confetti_ball:
not free they charge small amount , but after paying that we are good to go
Update:
So I asked ChatGPT to compare this behavior with Spotify ||Spotify because I saw it running in the background as well😭||
GPT said "This is good behavior". I asked it why and this was the response 👇
TLDR: Should I
- Force kill the processes
- Keep graceful shutdown but add a feedback like a spinner for user, saying "Closing..." ??
or you can first send close request to backend and without waiting for response, you can close frontend
Fair enough, i thought about this as well but what happens incase the request gets interrupted or dropped? We’ll be risking orphanage again, and the issue will persist
With the response check, we can force kill if response doesnt come back (ive kept the timeout at 3 seconds)
In my opinion, a modal spinner saying “Closing…” is the best option. But I’ll wait for Rahul‘s response in case he wants something else or anyone else has a better suggestion.
The spinner thing will create poor UX Experience , doesn't it ? The graceful shutdown is better I think , without spinner .
But no feedback can lead to the user thinking the app froze or is slow and heavy... bad impression
my bad , I thought frontend closes immediately and backend shuts down gracefully . Maybe we can try this approach but modifying tauri build config
No issues man
I looked up a similar method that popular closed source applications use:
- User clicks “X”
- Frontend immediately closes the window
- Frontend sends /shutdown to backend
- Frontend waits up to 3 seconds
- If backend doesn’t shut down, frontend force kills
But here there is a requirement that a background process will still need to be running, since a closed front end cannot both - be gone and enforce force kill…
Apps that do this usually have a tray process running in the background
I will see if it's possible @real gale
@marsh idol https://github.com/AOSSIE-Org/PictoPy/issues/876
I raised slow setting issue please review it once.
Has anyone encountered this issue ?
When I try to add photos to the photo library, I can’t see any local images in that folder. It shows ‘No items match your search,’ even though the folder originally contains images in supported formats (PNG, JPG, JPEG).
This isn’t a bug but rather a windows feature… the function in itself is asking for a “folder” and not image files. So as soon as you open a folder, and if the folder doesn’t contain any more folders but only pictures, the folder is shown empty and windows asks you to “Select folder”
For it to showcase any image files, the function will have to ask for images, not folders. And that is not the case since we are looking for folders.
Oh, got it — that makes sense. However, wouldn’t this interrupt the user experience? Naturally, a user would keep opening folders until they reach the one containing the images. In that case, the folder appearing empty could be confusing.
@wild patrol You are now level 2! :confetti_ball:
Understood your concern. You're always welcome to open an issue and propose your solution approach 😄
Sure !!
also the App has been giving me troubles actually, it wouldn't' t open and says "Failed to connect"..
I believe, few others were facing the same difficulty and were working on it if i am not wrong
is there any update?
What OS are you using?
@marsh idol Hello , I was digging into the face clustering code to understand the pipeline better and noticed that several important decisions happen implicitly like embeddings can be skipped, DBSCAN can mark faces as noise, and post-merge logic can collapse clusters.
While the steps are logged individually, it’s hard to tell at a glance how many faces were excluded, how much noise DBSCAN produced, or how much merging affected the final result without stepping through the code
I opened a small issue proposing high-level summary logging (no behavior changes) in case that kind of observability is useful: #878
https://github.com/AOSSIE-Org/PictoPy/issues/878
This Error seems to be happening for me as now... I did see it today in the morning but it only occurred once and got fixed by itself when I ended all pictopy related tasks and reopened the app. But now it seems to be happening consistently with me as well
windows
To Check for other bugs, I tired downloading the app in my friend's laptops as well..
The issue appeared in theirs as well 😅
I initially thought this was caused by my recent changes and spent hours debugging, only to realize now that it wasn’t related to them at all. 😕
@neon meadow You are now level 7! :confetti_ball:
Can anyone tell on which issue we are currently working on, so that I can go further
did you find out, to which it was related then ?
Its an SQLite failure error, backend is unable to open database files because DB file is often being created and found at Temp folder.
And because it's being created at Temp, sometimes the path happens to exist which leads to a successful startup, else this error
So if the DB file is created in a stable, app-specific data directory instead of Temp, this should resolve the issue.
Yes, that's the solution
is anyone working on this?
@wild patrol You are now level 3! :confetti_ball:
I am again reminding everyone to follow this guide strictly. I have just now closed 2 such issues, where the user opened a PR without any maintainer reviewing the issue. Even if your issue is valid, do not start the work. Please have patience.
@real gale @wild patrol Let us meet over a google meet.
is 9 PM IST fine?
but how the request will be interrupted? Its a very rare case
Most definitely
Video call link: https://meet.google.com/ueb-cmvo-wff
Use this link. @real gale @wild patrol
@marsh idol please review this issue and tell me that I start working on that ???
It triggered suddenly for all , so maybe there is an another trigger of it
Hi chat, please find below the key notes and takeaways from today's meeting 👇
https://docs.google.com/document/d/1ZkDr6gQjPyImEwD-UJ86zXMd3S6F5lFG6tUsSBcdNzk/edit?tab=t.hvq4nfwf0wa5
PictoPy Weekly Meet 10:00 PM 27-12-2025 Key Takeaways Release Deadline: Pictopy version 1.1.0 must launch by Monday or Wednesday, targeting cross-platform usability and core features. Critical Bugs: Current version unusable for Windows and Linux; 1.1.0 aims to resolve major issues and enhance f...
Any bugs on high priority before release ? Happy to help
Hi @marsh idol — I’ve reported a bug related to duplicate React keys in the AI-Tagging section. I’ve included details, reproduction steps, and a suggested fix in the issue here:
https://github.com/AOSSIE-Org/PictoPy/issues/888
I’d be happy to work on a PR once it’s approved.
Hi @marsh idol , I’m still facing this issue. Could you please help me with it?
@marsh idol sir can you please review this pr
https://github.com/AOSSIE-Org/PictoPy/pull/579
is this issue during setup ?
After installation when I am trying to use the application
ok
There’s a clash in port number on localhost:8000
try ‘netstat -ano | findstr :8000’ and terminate all processes running on port 8000 using ‘taskkill /F /PID <PID>’
Already tried it
Can u share screenshot after running the netstat command?
@wicked silo You are now level 2! :confetti_ball:
@real gale
Hello @marsh idol can you please check this PR https://github.com/AOSSIE-Org/PictoPy/pull/886 and let me if any changes there are to be made!
Restricted CORS middleware to only allow specific localhost and Tauri origins. This fixes the security vulnerability where any origin could access the local API. Closes #640
Summary by CodeRabbit
...
yeah
port 8000 is listening
so the app isn't able to start its backend (because app also needs port 8000) hence the clash
now try taskkill /F /PID 13032
You'll similarly have to make sure port 8001 is also free
I already did that
It's still showing same error
Hey everyone
Sorry for being inactive for so long. I had my exams going on, and they’re actually still continuing till 1st January, so things got pretty hectic for me.
A lot has changed during this time, but I’m still here and staying connected with you all. I’ll slowly start being more active again and joining in properly (||but still contribute ||).
Thanks for understanding ❤️
Congratulations! 🎉 And sorry for the late wishes —|| just saw the update now ||.
No worries 😄
Hi @marsh idol, I'm unable to correctly build and test graceful shutdown for linux, could you share me the exact build commands we're using for Linux .deb builds?
Are you talking about implementing a graceful shutdown on the frontend side???
Try running the build workflow on your fork
i.e. publish a release on your fork. But for this, your fork's main branch should have the commits of the "automatic shutdown" PR.
Hello everyone, i am is Vishal.
I am a Computer Science student with a strong interest in open-source development and building web applications as well as end-to-end systems. I enjoy transforming ideas into practical implementations and believe in learning through hands-on experience. I look forward to contributing meaningfully, collaborating with peers, and growing both technically and professionally as part of this team.
I’ve set up PICTOPY properly. Could you please guide me on how to start
still getting error ??
Yes
Could you please tell me exactly where you’re getting the error?
start by reading README file
After starting the app on the loading screen
Which command did you use to start the app?
Just clicked on it
Oh, so you’re talking about after installing the app and then running it?
Hey @marsh idol ! I’ve implemented the changes you suggested for issue https://github.com/AOSSIE-Org/PictoPy/issues/608
— sanitizing image_ids and securing the IN clause. Could you take a look and let me know if it’s good to merge or needs anything else?"
Hi @marsh idol
When you were creating deb package for pictopy, does it ask for private keys to sign the package? If so then can you give me private key to sign appimage
nvm i found it in CONTRIBUTING.md
@distant thorn You are now level 17! :confetti_ball:
Hi everyone ! @marsh idol
My name is Sanskriti , a Full Stack & Android developer exploring this organization’s projects.
I’m planning to apply for Google Summer of Code 2026, and I wanted to ask if it is confirmed that this organization will be participating as a mentoring org in GSoC 2026 this year.
If the participation is planned or confirmed, I’d love to start contributing early to active project repositories and prepare a proposal accordingly.
It would be great if you could share any updates or the expected timeline for the official confirmation/announcement.
Looking forward to contributing and being part of the community!
GitHub: https://github.com/Sanskriti10247
Thanks a lot for your guidance
Can anybody tell me in any bundler (like deb or .exe), where is the code (in src-tauri) to run backend and sync microservice executable on startup?
or we have to manually start it
yeah
ok can you tell me
oh
@marsh idol can i try to integrate backend and sync micro with my appImage so that we manually don't need to start the server
Hi everyone,
I’m currently exploring the Pictopy project under AOSSIE and going through its repository and documentation.
I’m interested in starting my first contribution here.
Could you please let me know if there are any beginner-friendly / good first issues or small features or documentation improvements that I can start with?
@marsh idol What are the plans for the video page, i have experience with processing and analysing videos with local YOLO models.
@tall bone You are now level 3! :confetti_ball:
They cannot give you their private key. If you fork the repository and release your own version, you must use your own private key to sign the package.
@wet lion You are now level 13! :confetti_ball:
The video page is still being planned, but it may include advanced video features later. Your experience working with video analysis could be very helpful for shaping it.
huh, but its publicly available in contribution.md file
@marsh idol should i create a gpg key or use that one is in that file
If you want to build and release packages yourself, you need to generate your own signing key.
they are not giving you their (||or may be||) private key to (||not sure||), you can also use that or generate your private key on your fork and then build using that key
sure
Hey @wicked silo ,
You’re getting this error because of this part of the code:
if (!(await isServerRunning())) {
const serverPath: string = await invoke('get_server_path');
const command = Command.create(
isWindows() ? 'StartServerWindows' : 'StartServerUnix',
'',
{ cwd: serverPath },
);
const child = await command.spawn();
command.stderr.on('data', (line) => console.error('Error:', line));
console.log('Server started with PID:', child.pid);
}
The issue is that the backend is already running, but the frontend checks it by sending a request to http://localhost:8000. Since the backend doesn’t respond properly, the check fails and the code assumes the server isn’t running, so it tries to start it again — which causes the port error.
you can check that manually by type this URL in your browser and send req and see the res
Hey @marsh idol@wet lion, I'm back after completing my sem Exams, pls let me know any help you want, now I'm available.
Hey@Rahul@Rahul vyas,I'm done with the javascript and some mini-project continuing my study Nodenjs and Express,EJS i want to contribute if you need some help ,pls let me konw , because i need some experience.
Hey everyone!
I’m Parth, excited to join this project team and start contributing.
I’m currently building my skills in open-source development, and I’m also planning to participate in GSoC 2026, so I’m hoping to learn, collaborate, and grow with you all.
Looking forward to working together!
hey @young torrent and @radiant tusk both start by setting up the project locally and if you face any issue then please let us know
It's an example private key. Actual private key is not shared with anyone.
hmm..
private credentials cannot be shared ...
to generate a build for yourself to test @distant thorn, you need to generate a private and public key using "npm run tauri signer generate"
and update this public key in the tauri.conf.json
Can @distant thorn also use this referenced key ?
yeah, maybe this is totally doable
sure i'll update the pub key
I have added the appimage in bundler but in workflow file, we need some changes for this appimage beacause it depends on linuxdeploy and some minor system packages
Hi!@wet lion , I’m setting up the project locally.
Could you please let me know the correct steps to run it?
You can check out this guide to run it.
to run it. There is also a video guide by Rahul Sir, which is very helpful for beginners and well explained.
@wet lion I'm facing issue here please can you help me ,issue with npm run setup
The first thing is that if npm is not recognized, it means you need to check whether Node.js is installed on your system or not.
First, run node -v and then npm -v.
If these commands don’t work or show no version, install Node.js first, and then start again.
And for adding the upstream remote:
First, check if an upstream already exists by running
git remote -v
If an upstream is already present, remove it using git remote remove upstream .
Then, add the upstream again with the correct repository URL.
Hey @marsh idol , I’ve implemented the image search feature based on tag name or image name. If you agree with the approach, I can go ahead and raise a PR.
see
@marsh idol SIR please review my doc so that i can work on that feature
when will be next meet ?
@marsh idol We can implement celebrity search, i.e we will hardcode the face encodings of celeberities (i have them all encodings ready) and compare all of them with the face encoding in the image. This way we will be able to tag celeberities by their name
- Different feature : We can also add OCR feature to detect text in images
Hey @tall bone , this feature looks good, but it can’t be implemented because it may give false matches for look-alike faces, requires frequent updates when appearances change, and doesn’t scale well as the number of celebrities increases bcz requires lot of celeb faces to be encoded. But don't know what Rahul sir thinks about it
No, I’m not a mentor
I was just sharing my thoughts to be helpful.
Discord is just a community where everyone helps each other. I was only trying to help—no offense intended. 😊
Most the modern face recognition services like AWS Rekognition have this feature
People always have old and authentic celeberities photos, it is not like if Ronaldo ages to 60 everyone will have his photos of only age 60
@tall bone You are now level 4! :confetti_ball:
And offline models are less smart but smart enough to handle a gap of good 10 years
I have tried this with YOLO offline model + dlib, it was sufficient. So, this is feature is elastic
hey as you are talking abt AWS Rekognition as this have Pre-trained celebrity database and also uses vector embedding ..
Ya they store faces as vectors
hmm
@zenith hound You are now level 2! :confetti_ball:
soo we also need to store faces as in vector db's
Only of the celeberities
also that we need to store all the mostly known celeb's faces in our app
Then we can detect face and calculate vector encoding in uploaded image on the fly and tag it with celeberity name
No images, only vector encodings of their faces like this https://github.com/Raais/Celebrity-Face-Encodings/tree/main/encodings
but for this we need to introduce Vector DB for this
No, it can be simply stored as json
Let me give you an example,
User uploads an image -> We detect face in that image -> We calculate face encoding of that -> we compare that with celebrity face encodings -> if matches we tag it with the celebrity name
hello all i am Uday and i'm interested to contribute to this project.
yeah this is quite possible but as you are talking abt YOLO offline model this can not detect the image if lightning changes or face of image/celebs changes (like one image can have beard and other can not have ) in this etc type of conditions
@marsh idol my PR is ready but i need to make changes and add a stage for appImage compilation
A small model can sometimes confuse it, but user can choose to increase compute and get bettwe results
@tall bone You are now level 5! :confetti_ball:
I got this on small
yeah same thing happened with me
It will be a good feature, as you definitely donot want to search person for an image of Ronaldo
And it will not consume that much compute, its simple and easy
small model makes so much errors
same
@marsh idol Do the application has any way of deleting images from within the window? If not i will be happy to implement this
I am proposing these new features
- Celebrity detection and tagging
- Delete option
- OCR for textual images
- Duplicate and best shot
- Direct uplaod option
It is a good idea i also want to help you in this 😇
Include me too together we can implement and test these features with more accuracy and efficiency, @severe haven as well.
Ideas for the Albums page
-
AI will classify images into screenshots, person, group, architecture, docs, vehicles, streets, festival, animal, trees.
-
User able to cretae new albums and drag and drop or ctrl+x and ctrl+v images there
-
User be able to deleted albums
A secret vault idea
-
A separate page for a secret vault, there will be a secret hidden button to open it and it will have a password
-
A key will be generated from the password and it will be used to encrypt docs, images, videos in that vault.
-
They key will never be stored, it will be generated, used and deleted.
I can implement this if you could approve @real silo
whats direct upload option
is face collection feature working properly?
Sometime works something don't
I uploaded multiple photos of the same person with one duplicate photos.
Only the duplicate photo got clustered.
@wicked silo You are now level 3! :confetti_ball:
It can cluster duplicate photos and also same person photo if they have similar lightning condition similar bg bcz this offline model only can detect same env like photos
I’m experimenting with improving face clustering so that the same person gets grouped even across different angles and lighting.
At present user have to move images through their file explorer, with this they can directly upload from anywhere
@marsh idol@wet lion I opened an issue regarding face detection accuracy and submitted a PR that resolves it.
Issue: #915
PR: #914
please review it
Hey everyone! I just pulled the latest changes from main (commit d07d817)
and I'm getting an infinite recursion error when trying to run the backend.
The error is in setup_logging.py at line 250 where emit() calls logger.log().
Has anyone else encountered this? @marsh idol @wet lion Should I open an issue or is this already
being addressed?
Error: RecursionError: maximum recursion depth exceeded
Location: backend/app/logging/setup_logging.py:250
emit() should not call logger.log() because it re-enters the logging pipeline and causes infinite recursion
and likely fix would be to write directly to the handler output inside emit()
thanks for confirmation, looking for minatainer approval before raising issue and pr, I'm ready with code
@young torrent You are now level 6! :confetti_ball:
I tried setting up PictoPy locally on Windows (Tauri dev flow) and ran into the same recursion issue in backend/app/logging/setup_logging.py that was mentioned earlier.
Specifically:
InterceptHandler.emit()ends up callinglogger.log(...)- which routes back through the same handler
- leading to
RecursionError: maximum recursion depth exceeded
I’ve paused further setup on my side for now since this blocks the backend startup.
Just wanted to confirm this is a known issue and avoid duplicating work.
Happy to help once there’s guidance on the intended logging flow 👍
Yep, same recursion issue — I’ve already fixed it locally and will open a PR once the maintainer assigns me. Feel free to explore other parts in the meantime!
sounds good
In which command you are getting this error? @north lion @young torrent .
During backend startup (Tauri dev flow).
The error occurs while initializing logging in setup_logging.py, before the backend server starts.
Hi guys, I think many people are having doubts and feature suggestions regarding this project. We have a weekly call scheduled and fortunately, it is on Saturdays. So please join the call at 10 PM IST.
Thanks for the update! I’ll try to join the weekly call.
For context, the recursion issue I mentioned occurs during backend startup when running the Tauri dev flow.
Thank you for organizing this meeting.
@wet lion @marsh idol , will you also be attending???
Possible, yes
Yes.
@young torrent @north lion have you tried using
.env\Scripts\activate.ps1
fastapi dev ```
I tried running the backend standalone using fastapi dev.
The server starts successfully and the OpenAPI docs load at :8000/docs.
The recursion issue seems to occur only in the Tauri dev flow, not during standalone backend startup.
Thanks for pointing me in the right direction, this helped clarify the scope of the issue.
check dm
I think issue was brought up earlier as well by someone , but I think there is no progress on it so I am bringing it up again
The issue is that the folder check API is continously hit while on the settings page and stops if changed
I’m planning to open a docs issue proposing a Windows troubleshooting section + clarifying user (EXE) vs dev workflows.
A lot of confusion seems to come from the difference between packaged builds and Tauri dev on Windows.
Let me know if that sounds useful @marsh idol
@north lion You are now level 3! :confetti_ball:
Yeah, this sounds like a real issue.
Since it’s being observed repeatedly and seems settings-page specific, it might be best to open a GitHub issue so it doesn’t get lost here and we can track it properly.
Yes but first I think I will make sure Rahul sir allow it by putting this forward in today's meeting
Hi everyone, I am working on an important change to the project so, I am rescheduling the meet for tomorrow 10 PM IST.
Hi everyone! 👋 I’ve just finished setting up PictoPy locally and have both the backend and sync microservices running smoothly. I’m a GSoC 2026 aspirant and am very excited about the project’s goal of local-first image management. I’m ready to dive into the codebase—are there any 'good first issues' or specific areas in the backend/sync service you’d recommend I look into first? Looking forward to contributing
Hey@short wraith, go through the readme.md and contributing.md. For the sync microservice, examine how the system handles file system events or metadata extraction. Do not open any PR without the maintainers' review; help the maintainer maintain the codebase.
Thank you for the guidance @young torrent ! Understood. I will spend the next few days deep-diving into the README.md and CONTRIBUTING.md as suggested. I'm particularly interested in tracing the flow of file system events and how metadata extraction is structured within the sync microservice. I'll make sure to discuss any ideas here with the team before opening any PRs. Excited to learn the codebase and help maintain the quality!
@marsh idol I start work on this issue please tell me ?https://github.com/AOSSIE-Org/PictoPy/issues/876
Hi @marsh idol, I am setting up pictopy project on windows os but with node version as 22 and python version 3.13. During this process encountered some issues so made changes in setup script. Can I raise pr which supports both the versions? Want your feedback for it.
Try python 3.12 setup is supported in that version perfectly
Hey @marsh idol
I’ve pushed the requested updates to PR #629 https://github.com/AOSSIE-Org/PictoPy/pull/629 and addressed all review comments.
CI is now fully green and there are no pending issues.
Whenever you get some time, could you please take a final look?
Happy to make any further changes if needed. Thanks!
@marsh idol @real gale I have a better solution for that when we close the application, backend should also be close.
Right now as i can see that we manually need to open Pictopy frontend and backend.
But In tauri there is plugin of sidecar which helps us to a seperate executable alongside with main application. On closing the frontend, the backend would automatically closes. It can be added in our src-tauri/src/main.rs file
So whats your opinion?
here is some code snippet for this approach
source - https://stackoverflow.com/questions/75913627/how-to-use-tauri-app-and-python-script-as-a-back-end
So this would solve this problem also
this is our current src-tauri/src/main.rs code. Here, after getting the resource path for backend, we ain't doing much
Hi everyone, I have made 2 major changes to the project.
-
Read the updated setup guides because we now have to start the sync-microservice manually.
Guide1: https://github.com/AOSSIE-Org/PictoPy/blob/main/docs/Script_Setup_Guide.md
Guide2: https://github.com/AOSSIE-Org/PictoPy/blob/main/docs/Manual_Setup_Guide.md -
I have changed the port numbers of the backend services from 8000 and 8001 to 52123 and 52124 to avoid clashes with other processes.
Everyone should first pull the latest changes and then again setup the project.
just sharing what worked for me on Windows after pulling latest main
Local dev order that worked:
- Start backend (52123)
- Start sync microservice (52124)
- Then launch Tauri
if Tauri is started before the sync service, I was getting“Waiting for servers”.
Might help others on Windows
looks like a nice idea and with lesser no. of line changes. Do join today's meet to discuss this.
@marsh idol You are now level 32! :confetti_ball:
Hi everyone,
I’m Panav , a sophomore in the Mathematics & Computing student at IIT Kanpur, and an aspiring GSoC 2026 contributor.
I’m currently focusing on web development using MERN stack and learning how to contribute effectively to open-source projects.
I recently started exploring this organization and would love some guidance on:
how to get started.
areas where my contributions could be most helpful
I’m excited to learn, contribute consistently.
Looking forward to working with you all!
@marsh idol just flagging a small UI observation (not blocking):
On Windows/Tauri I’m seeing double vertical scrollbars on the main content view (window scroll + inner container scroll).
Looks like a nested overflow issue (likely overflow-y on a child container). Not critical, just sharing in case it’s already on the radar.
Happy to look into it later if maintainers feel it’s worth addressing.
sure
https://calendar.app.google/rwvtXZGRXksgZKKn6
Yesterday, we could not meet. So we will be meeting today at 10 PM IST.
hey can anyone help me how to start contributing
i had setup localy and started it and i was unable to add any pics or folders
Can anyone help me guys ?
@short wraith You are now level 2! :confetti_ball:
this usually happens when only the frontend is running.
PictoPy needs backend + sync microservice to be running locally.
On Windows, try this:
- Start backend
cd backend
uvicorn main:app --reload --host 127.0.0.1 --port 52123
- Start sync microservice (important)
cd sync-microservice
python main.py
(or whatever command is mentioned in the setup guide)
3. Then launch Tauri / frontend
cd frontend
npm install
npm run tauri dev
Once both services are running, adding folders & images should work. If you’re just testing the app (not developing), the Windows .exe is the easiest way.
Let me know what OS you’re on.
also, on first run windows may block network access, make sure firewall permission is allowed.
I’ll try this let you know
👍
while the microservice command?
yes
do you have any resolution for this ?
the app crashes with a RecursionError: maximum recursion depth exceeded
its the same logging recursion issue in sync-microservice/app/logging/setup_logging.py
@north lion You are now level 4! :confetti_ball:
it looks like there’s an infinite loop in the logging setup the custom logging handler’s emit() method is calling get_logger() / logging.getLogger(), which ends up triggering the same handler again and again.
it’s not a setup problem on user end, the logger is recursively re-initializing and hitting Python’s recursion limit (especially on Windows / Python 3.12)
ig this would be discussed today in the meeting
yess
Yes, It was recursively calling, I’ve just seen the backend and it’s seems to be the issue
I have opened issue #918 regarding this topic and am waiting for @rahul sir’s review
https://github.com/AOSSIE-Org/PictoPy/pull/923/files
it was closed as there were not active discussion on the issue and nobody was assigned with the problem
I noticed, it was closed yesterday
Hi @marsh idol, I have made some changes regarding project’s compatibility with python 3.12 and 3.13. And worked on one script which first runs the frontend later runs backend so that app will be available to use. As I have to manually run both servers on different terminals. Can I proceed with creating issue and submitting my changes? Correct me if going in wrong direction.
rn, you need to manually start backend and sync microservice
bruhhh 💀 🥶
so we need to increase the depth also
Hi @marsh idol @unreal quartz
I opened issue #924 about improving onboarding and documentation clarity from a first-time contributor’s perspective.
I’ve now set up the docs locally and have a small, scoped reorganization in mind (no new content, just structure/clarity).
Before opening a PR, I wanted to confirm if this direction sounds okay and if there’s anything specific you’d prefer I focus on.
Thanks!
I have tried it too, but still showing the same
@short wraith run all three commands in different terminal
Yes, I did that only
@short wraith You are now level 3! :confetti_ball:
are you on latest commit?
Yes I’ve synced it
it was showing unsupported windows v11
i think that's the problem ?
https://docs.google.com/document/d/1yK9agOY1BAX0MRzvJNLapJ8AWZsGRfht7xyUFpD1DSw/edit?usp=sharing
Hi guys, the weekly meet is at 10 PM. Till then, update your agendas in this doc. I will address each person's concerns/suggestions/doubts in the order listed in this doc only.
https://github.com/AOSSIE-Org/PictoPy/issues/937
Hi everyone i found out a bug and raised an issue on it.
I would appreciate if someone could confirm the issue and provide more details on the bug.
https://meet.google.com/ueb-cmvo-wff
Meeting link. Starts at 10PM IST.
windows 11 should be fine, I’m on win11 as well.
Just to confirm the setup:
- are you running backend, sync-microservice, and tauri in separate terminals?
- and does the recursion error appear when starting the sync-microservice?
If yes, this is likely the known sync-microservice logging recursion issue, not a setup problem.
Fill this doc as soon as possible. I will only address queries mentioned in the document during the meet.
Now, it’s not recursively and I’m running in separate terminals as well,
is it not working yet?
No it’s still showing me that
For videos and albums : feature in development
assuming that you can now open image directory?
and yeah for videos and albums, its under development
is the name section for contributor/user name?
Yes
But someone in issues added a video that we can add pics and all
Yes, adding images works
Currently, images are added by placing them into the watched directory externally (via file explorer), not via drag-and-drop in the UI yet.
Once added, PictoPy will pick them up and they’ll appear under Home and AI Tagging views.
In-app folder/image adding is still under development.
@marsh idol i'll join after 10:15 due to some work
Sorry I won't be able to talk right now (priyansh)
@marsh idol
@north lion You are now level 5! :confetti_ball:
https://docs.google.com/document/d/1ZkDr6gQjPyImEwD-UJ86zXMd3S6F5lFG6tUsSBcdNzk/edit?usp=sharing
This is the summary document for all the meetings.
If you were not able to attend today's meet, please read the summary.
@marsh idol Sir, as you assign sawan on recursive logging issue on issue number #918, should I share actual issue in Discord?
Apologies for not being able to join the meeting. I was not in touch and away at that time. I’ll make sure to stay connected going forward.
Thanks for sharing the summary! @marsh idol sir, did you make this?
Thank you for the doc i missed out the last 5-6 minutes of the meeting
Hey, @marsh idol as you mentioned on Meet, I have raised the issue with the functionality. I will add
https://github.com/AOSSIE-Org/PictoPy/issues/942
I have a follow-up drag-and-drop feature that can be implemented pls have a look
https://github.com/AOSSIE-Org/PictoPy/issues/944
and tauri port detection error I will send within couple of hours
Describe the feature The Issue: Right now, the sidebar is always visible, which takes up screen space that could be used to show more photos — especially on smaller screens or in fullscreen mode. S...
After updating Pictopy and reinstalling it on my system, I encountered issues while starting the backend service. As a result of these backend issues, the frontend application is also unable to start successfully.
@marsh idol I have created the issue for AUR support for arch linux https://github.com/AOSSIE-Org/PictoPy/issues/946
Hi everybody, hi @marsh idol, just noticed there's a huge change to the architecture
Oh
I see
@marsh idol i have some suggestions regarding our final release package
- Currently, we are copying all our backend folders into the release package which is not a great approach. This should be the desired folder structure
usr/
├── bin/
│ ├── pictopy_py #Tauri application
│ └── python_backend
| └── python_sync_backend
├── share/
│ ├── applications/
│ │ └──.desktop
│ └── icons/
(demo opensource project - https://github.com/Senhaji-Rhazi-Hamza/tauri-vue-python-demo-todo)
-
We are creating a database file in that release package itself. which is also again not a good practice . There is python function
user_data_dirin the packageplatformdirswhich would automatically chose the best plateform specific path available for our database and logger to get copied.
(https://platformdirs.readthedocs.io/en/latest/api.html#platformdirs.user_data_dir) -
After solving these two issues, we can use
@tauri/Command/Spawnto directly spawn our backend and microservice alongside with our tauri application
After that pls suggest any other dependencies that needs to be sorted our so that we can run python backend binary in standalong environment.
No, it is fine.
There is a tool called fireflies.ai
The recording you shared for the collapsible side bar is from another app. Can you make a prototype for our app?
What is the issue exactly? screenshot seems to be fine.
- How do we ensure that backends are in a separate directory after installation?
- Seems like a good idea because whenever the app updates, the database will get overwritten which is not a good idea. This will allow us to keep the database intact even after an update.
Hey @marsh idol, I pushed changes to my Graceful Shutdown PR, updated it with the merge conflicts. And coderabbit said “macOS support removed.”
I traced this… but I’m pretty sure my changes didn’t change anything regarding this as all…
- you mean the executable? or its files?
- yea the Resouces directory would be overwritten not the database file
both executables and files
but why we need files? in final release application
can you tell me which files are being used by backend executable
@marsh idol I’ve opened a PR fixing the duplicate scrollbar issue on Windows/Tauri (UI-only, non-blocking).
I’ve tested it locally and addressed review comments.
Would appreciate a review when convenient.
ig pyinstaller will bundled them in 1 single executable file only
It can but there is a flag in pyinstaller command which keeps those things separate. I have tested both ways, and the single executable is really a bad idea. Because it unpacks everything before startup and takes a long time.
Valid
Even i tried onefile builds while testing graceful shutdowns in prod builds and the startup times were horrendous
My frontend is not running i mean Pictopy is not running and just telling "Waiting for Servers"
Is anyone here running PictoPy on Windows 11 where face clustering is working correctly (i.e., faces are grouped into collections, not just tagged as person)?
I’m testing on Windows 11 + Python 3.12 and the backend runs without errors, but face-clusters always returns empty results. Trying to confirm whether this is a known Windows limitation or environment-specific.
Thanks for sharing this with us
@marsh idol as discussed in our last meeting, I’ve raised the issue here:
https://github.com/AOSSIE-Org/PictoPy/issues/876
Please review it.
It looks like the backend service is not running yet, so the frontend is currently waiting for it to become available.
@ornate jungle refer to this.
It seems like issue related to python version on windows 11+ version. You can check it with log and try if it is actually clustering faces or not ..
@marsh idol please review this
I checked the backend logs, face detection is running, images are tagged as person, but no face embeddings / clusters are being generated. /face-clusters consistently returns empty.
I’m currently on Windows 11 with Python 3.12, and I also see ONNX Runtime warnings about Windows support.
Hi Isha, sorry for the delay. But due to other high priority tasks, It will take few more days before I review that issue.
I'll try reproducing this with Python 3.10 / 3.11 and report back, this might be a windows + py version compatibility issue
Yeah, that's can be the issue (window+python version) bcz on windows 11 this is running correctly
Hi @marsh idol , could you please review the PR for issue #918?
Hi maintainers,
I’ve been exploring the project and setting it up locally to understand the development workflow.
Since the project uses Tauri + React (TypeScript) + FastAPI + Rust, the initial setup can be a bit challenging for contributors, especially those new to Rust or Tauri.
I’ve been working on office and personal projects involving Docker, FastAPI, and React, so I wanted to ask if the project would be open to adding a Docker-based development setup (e.g. docker-compose) to simplify onboarding.
Potential benefits:
• One-command setup for contributors
• Improved environment consistency
• Helpful for GSoC students and first-time contributors
If this sounds useful, I’d be happy to work on:
• Dockerfile(s)
• docker-compose.yml
• Basic setup documentation
Looking forward to your thoughts
Hi, it is a good idea. I believe an issue was already raised around a Docker-based development setup for improving onboarding.
Dockerizing the frontend is tough. There are only hacky ways to do it. We need to use a remote desktop software to run desktop app on docker.
For backend, the setup process is very easy. So no need for that too
We had a docker setup guide, I removed it only because of this reason.
Plus docker is way too heavy for windows users and requires minimum 16gb ram to work nicely
Docker-based development setup
Docker takes a lot of memory and space
Totally agreed using docker consumes too much ram and storage.
I think for now instead of docker it would to be good to work on improving current setup experiences itself.
hmm
@marsh idol its working ig
i was overthinking but its just a simple one line command
evrything is intact. in main.rs, i am changing my current working directory to my pictopy server and then spawing that process
just need to fix this backend side error (logging related)
Yes I had raised a similar issue.
@marsh idol I have raised a pr resolving issue: #939
PR: #955
Please review it
Hi everyone
I’m Ayush, a 2nd year undergraduate student with an interest in open-source development and collaborative projects. I’ve been exploring the Pictopy codebase and I’m excited to start contributing and learning from the community.
If there are any good-first issues I’d be happy to pick one up and start working on it. I’m comfortable with documentation, bug fixes, and beginner-level feature improvements, and I’m eager to grow as I contribute.
Looking forward to contribute
@marsh idol As many people are facing issues due to wrong python version or they are just running it without creating proper virtual environments.
I am proposing two things
-
Using miniconda to create environments, as with venv they can't mention python version for that venv
-
I can make a tutorial video for doing setup from zero, covering debug for most commonly faced issues
Pyenv can be used for that , it is eassy to implement as welll
Is this really a big problem? Why aren't people able to setup the backend? setting up virtual environment is very easy
Valid
Yes its easy but most of the messages recently are people getting stuck in basic setup issues and thinking that this is a legit bug
Sir please guide us for this with the help of virtual meet.
Don't know why but people just need to follow the README file and they are there
I will take a meet tomorrow
and cancel the upcoming weekly meet, because I have some other commitment for this week only.
Hey @tall bone really appreciate your efforts but even if you make a full guide then also people will get the error bcz of their system env
No need of tutorial ig. But we might think on managing using miniconda instead of venv and pyenv. Only reason is its easier
If someone is using venv, he literally have to deleted other python versions and download the 3.12 one (novice thinking). Also, the guy might be working on some other projects where there is a different need
Then pyenv is the better option I think
because, with conda, we need to change the installation commands if I am not wrong.
and I want to have pip as the package installer for all systems
miniconda is light weight version of conda
I have personally used pyenv and miniconda. The problem with pyenv is that sometimes terminal and vs code confuse between versions and cause horrible unexpected errors
Like terminal is using pyenv venv version but vs code still uses global version. Even on changing interpreter this issue might occur
suppose i am using a miniconda environment, then do we have to use pip or conda to install packages from requirements.txt?
pip command
The commands will be like
conda create -p venv python==3.12
conda activate ./venv/
pip install -r requirements.txt
conda deactivate
Also miniconda has better dependency resolution than pyenv because pyenv uses your global python but miniconda has its own setup for these things
Even for pyenv we have to update commands
@tall bone You are now level 7! :confetti_ball:
I will update the steps in the setup guide and the code in setup scripts. First I need to test conda for myself.
Then you can make a tutorial
Ok, Thankyou very much 😀
Then, does it close the server automatically? and have you tried running the sync-service too. Make sure that both the processes are spawned.
okie , Can I take up any task which is in high prioriy ?
exactly , it will create problem for the user which have only 8 gb ram
Right now, I do not have any issues in mind which are high priority.
Okay, I’ll look into another bug then.
Please also review this issue when possible:
https://github.com/AOSSIE-Org/PictoPy/issues/795
@eternal venture You are now level 4! :confetti_ball:
Hey @marsh idol, I've implemented the collapsible sidebar toggle feature as discussed in yesterday's meeting. I've raised issue 942 for this. Pls review it, and suggest any changes, and should I raise pr for this.
sure i'll implement and tell you by tomorrow
what's the issue?
Got the issue while starting the app?
@frank quest You are now level 6! :confetti_ball:
yea
the backend is running properly
Exactly
😭
but the sync microservice executable is not giving me correct health
For me it looks all good but still it says could not start
these build issues never end 😢
sad
can you share screenshot?
what are the logs displayed when you startbackend?
oh i see
what is the command you are using to build @distant thorn ?
Is there a scheduled meeting for tomorrow?
Yes
The upcoming weekly meet will not be held
instead, we meet tomorrow
Okay, i’ve got a question related to the new architecture. I will discuss this tomorrow only in that case.
sure
see first i packaged my application through pyinstaller, then copied app directory into that
@distant thorn You are now level 18! :confetti_ball:
i have followed that is in action workflow
wait
I know why is this happening
where have you stored the PictoPy.db file for the built PictoPy_Server?
the sync service needs that file
backend/dist/PictoPy_Server/app/database/PictoPy.db
sync microservice doesn't have that ig
and where is the sync executable located at?
Hello everyone,
I’m Muneer Ali, an MCA student (2025–2027) and a Full-Stack & AI developer based in India. I’m excited to start contributing to AOSSIE and collaborate with the community.
Background
Strong experience with JavaScript, TypeScript, Python, Java
Frontend: React, Next.js, Tailwind CSS
Backend: Node.js, Express, REST APIs
Databases: PostgreSQL, MongoDB, Firebase
Familiar with Git, GitHub workflows, CI/CD
Good foundation in Data Structures & Algorithms
Open-source & project experience
50+ merged pull requests across multiple repositories
Built AI-driven platforms including:
DocMagic – an AI document generation platform
SignifyAI – an Indian Sign Language translation system
Regularly contribute, review PRs, and follow clean commit practices
What I want to contribute
I’m particularly interested in:
Web platforms (React / backend-heavy projects)
Developer tooling
AI/ML-assisted educational or productivity tools
Improving documentation, tests, and scalability
I’m currently exploring AOSSIE repositories and issues and would appreciate guidance on:
Beginner-friendly or “good first issue” tasks
Ongoing projects that need contributors
Best practices followed in AOSSIE contributions
Looking forward to learning, contributing consistently, and collaborating with mentors and fellow contributors.
Thank you!
— Muneer Ali
GitHub: https://github.com/Muneerali199
LinkedIn: https://linkedin.com/in/muneerali
let me show you the built directories
keep the PictoPy_Sync folder in the same directory as the PictPy_Server
Bro 😭
see this is directory struct
@marsh idol Sir I had raised an issue regarding Developer Experience. Please review and assign it.
bro has very vast knowledge 😂
Can anyone guide me I am new here tell me some issues I can work too on it
Currently I setup project locally
Welcome to the community Munner, but when there is an important conversation going on, please do not send a message which is not related to it.
wait
Ok sir
micrservice can't reach the db
because its in different folder
resources/backend/PictoPy_Server/app/database/PictoPy.db
resources/sync-microservice/PictoPy_Sync/app/database/ -> doesn't have any db
k
Hi, I'm Sohail, a final-year B.Tech student(CSE) and a MERN stack developer with three years of experience. I'm here to make meaningful contributions.
Also, would it be okay if I joined the VC?
Hi! I’m Piyusha, a second-year ECE student. I’ve set up PictoPy and explored the backend and frontend. I’m excited about the project and would love to start contributing wherever I can!
I am, Alivia Hossain, a State-level scholar (JBNSTS 2022) and a Semester 1 AI Engineering student with strong interest in open-source, AI, and web development. I have contributed to multiple open-source programs including GSSoC, SSoC, and served as Project Admin/Mentor in Hacktoberfest and SWoC. I have built an AI-based disease prediction system, an AI deepfake project, and developed my university department’s official website from scratch. Looking for to be a contributor in GSoC too !
My LinkedIn: https://www.linkedin.com/in/alivia-hossain-513a3a365?utm_source=share&utm_campaign=share_via&utm_content=profile&utm_medium=android_app
@marsh idol i have implemented to kill backend and sync server when application closees
Hey @marsh idol !
I noticed that the top Nav Bar shifts/jitters whenever the gallery reloads or hits the end of a page, so I’ve raised a GitHub issue to stabilize the layout. I also want to add a Pull-to-Refresh feature to the gallery to make it easier to update the image feed manually. I’m keen to jump in and fix this could someone please assign the issue to me?
@short wraith You are now level 4! :confetti_ball:
@distant thorn could you please share this with the Task Manager?
yeah, if you can share this with Task Manager this will be easy to understand
@zenith hound You are now level 3! :confetti_ball:
@marsh idol
I put together a short screen recording showing how I’m running PictoPy locally on Windows using a small personal PowerShell script that simply orchestrates the backend, sync-microservice, and Tauri together after the official setup.
I’m not proposing this as an official script, it’s just my local workflow, but I wanted to share the video for context and ask if you’d prefer this to stay purely personal, or if it’s worth opening a small documentation / ergonomics issue later.
Important: This script is meant for local development after setup is complete. It does not install dependencies or create environments, and assumes the project has already been set up using the official setup scripts.
ok
@marsh idol sir will there be any class today??
see this
You are using which version of node in this?
@severe haven You are now level 3! :confetti_ball:
v22.16.0
Try with node 20
why you are using node 20
the latest lts version is 24
With the sidecar plugin?
yes
can you send the ss of the issue you facing
just to understand the intended workflow on Windows, with the sidecar plugin managing backend/sync lifecycle, is that expected to be the primary dev path there as well?
I ask mainly because I’ve seen Tauri sidecar process handling behave a bit differently on Windows compared to Linux/macOS, so I wanted to align my local dev setup with what you consider the supported path.
Let me check
can you test that feature on windows?
you can try 22+ maybe 22.20 that I'm using but i don't think so latest version gives error
@bold estuary You are now level 2! :confetti_ball:
by forking my repo
sure, i can test it on Windows
@north lion You are now level 6! :confetti_ball:
dm
@marsh idol we need to change the DB path also
These issues were coming while trying to start backend server and microservice
@marsh idol
I’ve implemented a small UI polish where the “Go to Settings to add folders” text in the empty gallery state is now clickable and directly navigates to Settings.
It’s a minor UX improvement and doesn’t touch any backend or routing logic. Let me know if you’re okay with including this
@marsh idol Hi, just checking in to see if there’s any meeting planned for tonight?
It works perfectly for me lookout for the command to run backend and sync-microservice in doc.md
Yes, open a PR.
@marsh idol Do we have a meet today?
Yes
What did you tought of the miniconda thingy
I was testing it yesterday
We are going ahead with miniconda setup
I will commit those changes before today's meet.
I am glad you liked it
I've raised the PR 👍
I was thinking of adding a feature where we should be able to manually reassign images from one cluster to another, or even manually assign an image to a particular cluster, instead of performing global reclustering every time a new pic is added. it would be a good way to compliment the AI tagging by adding a human-in-the-loop component
@marsh idol @radiant mauve is this something we can try implementing ?
https://github.com/AOSSIE-Org/PictoPy/issues/891
Something similar is already open
alright thanks for letting me know !
Hey @marsh idol !
I noticed that the top Nav Bar shifts / jitters whenever the gallery reloads or reaches the end of a page. I’ve raised a GitHub issue to track and stabilize the layout.
I’m also interested in adding a Pull-to-Refresh feature for the gallery to make manual refreshes smoother for users.
Issue link:
https://github.com/AOSSIE-Org/PictoPy/issues/969
I’d love to jump in and work on this, could someone please assign the issue to me if it’s available?
Describe the feature The Issue: Right now, Folder Management in Settings is always expanded. When a user has many folders, they have to scroll all the way down just to reach other settings, which h...
Meet at 10PM IST, update your agendas here. I will only address the topics mentioned in this doc. Order will be decided based on the priority of the agenda
Sr. No. Contributor Name Agenda/Doubt/Suggestion 1 Vasu Sahu Bulk AI tagging process, folder management pr #787 Logging recursive error using CLI at root folder so that both frontend and backend run using single command Fixing tauri auto port detection when a port is engaged somewhere, ...
Hey everyone ,
I’m Kalpana. a final year BCA student. My main tech stack is MERN, Next.js, TypeScript, Tailwind CSS, and I’m also exploring AI/ML with Python.
I’ve recently started exploring PictoPy and set it up locally. While going through the project, I raised an issue around improving real-time folder sync with the filesystem.
🔗 Issue link: https://github.com/AOSSIE-Org/PictoPy/issues/968#issue-3785422101
I’d love to know if this issue is worth working on from a maintainer/mentor perspective, or if there are other areas in PictoPy where contributions would be more impactful. @marsh wyvern @marsh idol i kindly request you to review the issue and if it is aprpoved then assign me this issue.
Also, if anyone has suggestions for good first issues or high-impact features I could work on, I’d really appreciate the guidance.
Excited to learn and contribute more here.
Thanks for the review, sir 🙂
@tall bone
https://github.com/AOSSIE-Org/PictoPy/blob/main/docs/Manual_Setup_Guide.md
Can you check these updated instructions?
Is the meet link the same as last time?
Added a smooth exit animation to the Media Info panel for consistent open/close behavior
Will today’s agenda be discussed today?
Yes
I will be late sorry
@marsh idol
I have updated the meeting summary on the doc. If you are not able to attend the meet, it is highly recommended that you read the summary.
The message with the links is always pinned, so it is easier to find it in the pinned messages.
@marsh idol I've created the pr for that minor change of the unused open folder button in the image viewer.
https://github.com/AOSSIE-Org/PictoPy/pull/976
Hi @marsh idol, just for the sake of doc-ing, my approach of fastapi/python based graceful shutdown is ready, it works on both windows and linux as of my checks.
https://github.com/rohan-pandeyy/PictoPy/releases/tag/v0.0.9-test-shutdown
Tested on this release...
Its completely alright if you want me to work with @distant thorn on the Rust based approach, but I'd like if you could give a moment and test this out on MacOS as well. All the "how it works", we had discussed in one of the previous meets, are the same, I had just polished after that meet's suggestion
@marsh idol sir i wasn't able to attend the whole meet, just wanted to ask about a few issues that I have raised, whether I can try implementing them :
- Add a button to the AI Tagging Window to navigate to the Settings tab easily where the tagging feature can be enabled. (issue#925)
- Add a confirmation dialog for the Reclustering option in the settings. (issue#966)
- Allow user to choose the cover photo for a particular cluster, as right now, the first photo becomes cover
Describe the feature 🌟 Feature Description I propose adding a "Configure Tagging" (or "Manage Folders") button to the header of the AI Tagging page. Currently, the AI Tagging pa...
@marsh idol
-
Tried reusing
tauriUtilsas suggested by you, but it caused false positives in the web-only landing page, so I kept a localisTauricheck to fix the scrollbar issue.
PR - #941More details in the PR
Issue - #938 -
PR updated: removed hard-coded navigation and switched to using the existing route constants.
PR - #964
Issue - #965 -
Opened a PR for the Image Details exit animation fix (UI-only polish), waiting for review.
PR - #978
Issue - #972
Hi everyone,
I set up PictoPy on another laptop following the documented setup steps. The backend starts successfully and /health returns 200, but the Tauri app exits automatically with the error:
“Failed to connect to one or more local services. Exiting…”
After debugging, it looks like the app performs a startup health check and exits if any expected local service is unreachable. In my case, the backend is running, but it seems to be listening on a different port than what the frontend/Tauri expects, so the health check fails and the app exits by design.
Could you please confirm:
Which port(s) the backend is expected to run on during local development?
Whether there is a config/env variable that needs to be updated if the backend runs on a different port?
Happy to share logs or open an issue once confirmed. Thanks!
I explored running PictoPy using VS Code tasks to meet two main requirements.
Requirement 1: Display live logs for the backend, sync microservice, and Tauri frontend in separate terminals.
This is fully supported by VS Code tasks, with each service running in its own integrated terminal, which improves visibility and debugging.
Requirement 2: Automatically terminate all running services when the Tauri app is closed.
VS Code tasks do not support coordinated shutdown across dependent tasks, so this behavior cannot be reliably implemented within the integrated terminal.
The PowerShell runner achieves both requirements by managing process trees directly, which is not possible in VS Code’s task model. For this reason, both approaches are kept to balance log visibility and full lifecycle control.
@marsh idol
Which ports are your backend and sync service running on locally?
@north lion You are now level 7! :confetti_ball:
I think you might be not running sync-microservice or else used a different comand to run it can you share your terminal
Thanks a lot everyone 🙏
The issue is resolved now.
It turned out the sync-microservice didn’t have its own virtual environment.
After creating a dedicated venv and installing the dependencies, everything is working as expected.
Hello 👋
I’m Puneeth, a Computer Science Engineering student from BMS College of Engineering, Bengaluru.
I recently set up PictoPy locally and got the frontend + backend running successfully.
My primary background is in React and JavaScript, and I’m starting to explore the FastAPI-based backend gradually. I’m excited to contribute to PictoPy and would love to begin with frontend or beginner-friendly issues.
Please let me know if there are any good first issues or areas where I can start contributing.
Thanks!
Hello sir @marsh idol if this approach is good please assign me this issue
https://github.com/AOSSIE-Org/PictoPy/issues/983
Is there an existing issue for this? I have searched the existing issues What happened? 📋 Issue Title Feat: Implement lazy loading for images in gallery to improve performance 📖 Description Current...
Hello @marsh idol, I’ve opened PR #982 addressing this by centralizing SQLite connection handling and enabling WAL with a bounded timeout to reduce lock contention—happy to adjust based on feedback.
https://github.com/AOSSIE-Org/PictoPy/pull/982
@marsh idol
I made a few small docs navigation fixes:
- the Overview section now properly lists Features and Architecture
- the Setup section groups quick/manual/script guides under one expandable section
- the Frontend links now match the actual existing docs.
I’ve opened an issue for this.,
Issue - #985
Hey everyone, I opened a small frontend refactor PR today while going through the Home page code.
Sharing it here for visibility.happy to make changes if needed.
https://github.com/AOSSIE-Org/PictoPy/pull/987
While reading through the Home page implementation, I noticed that the component was accessing data?.data directly even though usePictoQuery already exposes successData for this purpose.
This chang...
forgot to upload the after image
Hi @marsh idol sir, I raised a feature request for adding a “Back to Top” button. If this is okay, please let me know and assign the issue to me. If there are any requirements , I’m happy to discuss and will wait for your confirmation before raising a PR.https://github.com/AOSSIE-Org/PictoPy/issues/979#issue-3786296691
Describe the feature Currently, PictoPy has a side scrollbar that allows users to navigate between sections, which works well. However, to improve user convenience, especially on pages with many se...
Hey @marsh idol I have implemented the Duplicate and best shot feature. I have attached the video in the PR. Here is the PR
Also, I have implemented the "Celebrity detection and tagging" for 1000 celebrities. I have attached the video in PR. Here is the PR
https://github.com/AOSSIE-Org/PictoPy/pull/990
Hi everyone 👋
I am Sami, a 12th grader and i have some experience with full stack development. I went through the project and it seems very interesting, so from today, I'll be contributing alongside y'all.
Lets have a great time working together 🙂
Hi everyone,
I’m Sarabdeep Singh Bilkhu (GitHub: @SarabdeepSBilkhu), a student contributor.
I’ve set up PictoPy locally and gone through the codebase, tests, and documentation.
Looking forward to learning from the project and contributing to approved issues.
guys i wanted to test onboarding for a bug but there wasnt a simple way to reset the application state to go back to the onboarding page
i have added a delete user functionality
which is essentially resetting all the preferences, removing the folders and taking us back to the onboarding
i am opening an issue, the pr is ready and i fixed a validation error(if you dont select a folder during onboarding you will be met with validation error in the console) along the way.
@marsh idol please take a look
(i received a mention spam warning?!)
Describe the feature It would be a good and handy feature to have a "Delete User" button in the settings. The idea is to let users completely wipe the slate clean, remove the profile, pre...
@toxic temple You are now level 2! :confetti_ball:
Hi
Hi, I’m Himani, a first year BTech in CSE student exploring open-source and currently learning by working on real projects. I’ve been going through PictoPy’s codebase, documentation and I've set up PictoPy locally, and I’m interested in contributing to small fixes and improvements while I understand the system better. Looking forward to learning from the project and contributing where I can.
Hi Everyone !! I'm Keerthana , a third year Student at Bangalore Institute of Technology. I would like to contribute on this project , I have already set up the Project locally. Aiming to give and get more from this Project . Thank you
I have opened an issue https://github.com/AOSSIE-Org/PictoPy/issues/995 . PictoPy has no gallery viewing option I think it would be better to add this feature for users convenience. And I would like to contribute on this issue . So @marsh idol Sir please confirm me before i proceed.
@junior moth Sir please confirm me before i proceed
Hi @marsh idol , I’ve opened a PR that adds WebP image support by extending backend validation and updating the UI text.
The fix is scoped to validation only (no ML or clustering changes).
Would really appreciate your review whenever you get time. Thanks!
Hi everyone! 👋 I'm new here and really excited to contribute to PictoPy.
I've just submitted my first PR (#997) which adds a collapsible "Folder Management" section to the Settings page. This helps declutter the UI, especially for users with many folders.
I'd really appreciate any feedback or reviews. Looking forward to being part of the community! Thanks!
?
@marsh idol I put together a contributor-focused project structure doc to make PictoPy easier to navigate for new contributors.
Shared it as a gist here for a quick review: Gist would love your feedback.
Hi @marsh idol , I’m Harsh Shah (2024 graduate), currently working as a web developer at Media.net. I recently set up Pictopy locally and am new to the project.
I opened Issue #977 regarding the performance problem in the /toggle-favourite endpoint. Currently, the endpoint calls db_get_all_images() and scans the entire images table just to determine the favourite status of a single image, which makes the operation O(n) and causes unnecessary memory usage. As the number of images grows, this will not scale well and can waste a lot of memory.
My proposed approach is to change this flow to query only the specific image by its ID and then toggle/update that value directly in the database, instead of loading all images into memory. This significantly reduce memory usage.
I wanted to check if I can proceed with this fix .
Thanks!
is there any reason why this MR hasnt been merged yet? https://github.com/AOSSIE-Org/PictoPy/pull/689
guys, i saw multiple bugs in onboarding and created an issue for it.
https://github.com/AOSSIE-Org/PictoPy/issues/999
(ik you are busy sorry for tagging you again @marsh idol )
(i needed the delete functionality #993 for fixing these)
@marsh idol @radiant mauve , is there a demo video of PictoPy that I could look at?
This is a little old video, there have been multiple developments since then
@marsh idol Here is the zip folder of the sample photos I used.
regarding Issue: #915
PR: #914
@marsh idol @radiant mauve , here is how I manage my photos:
1 - I first organize the photos in folders based on events (e.g. "My Son's Birthday", "Christmas", ...).
2 - I also have some folders for photos that do not belong to any particular event (e.g. "Photos taken at Home", "Photos taken in the neighbourhood", ...)
3 - When all photos are in folders, at the end of the year, I go through each folder and select a few photos that I then copy to a "Screensaver" folder. (I have an old mac connected to my TV, and this Mac shows this screensaver photos in the TV.). My priority is to select photos where my family members appear. The more family members appear, the better.
If PictoPy allowed me to search photos that contain at least one person from a subset of people and allowed me to sort the photos by the number of people who appear in each photo, this would be very useful to me. (As far as I know, most photo management photos allow searching for photos that contain a single person, but not subsets of people, as I am suggesting.). This would be very helpful in step 3 of my workflow.
@marsh idol @radiant mauve , in PictoPy, are "albums" a concept that exist only in PictoPy's database? What is the relationship between folders and albums?
Yes it exists only in the locally stored database
Each image id can be a part of multiple albums
The ID will be used as a reference to view the image from the actual folder location
Thank you for the feature suggestion. I will get started on planning this feature in the upcoming releases
Could we also store the "albuns" of a photo in the EXIF metadata of the photo somehow?
Yes, we can store it. But I have a question. How would we utilise this stored information. Would it be for some other device?
A long time ago, I used some photo management tools that stored all information in their own databases (I don't remember their names anymore). This made me trapped to using their software forever, even if the database was stored in my own computer. When the software didn't suit me anymore, all the tagging and classification information that I had constructed with it over time became useless. This is what led me to prefer to organize my fotos with folders (i.e. steps 1 and 2 of my workflow described above) instead of software-specific databases. (There is nothing more general and software-agnostic organization approach than using folders.)
EXIF metadata could be a good middle ground between using only folders and becoming locked in a software-specific database.
Yes this is a valid issue. I would gather more ideas on how to export the tagging information along with storing it in the photos itself.
There may me a common standard setup by some other organisations that we can follow to store this data. I would need to research more about this. If we use some common standard,then the data can be portable.
Or we can start some common standard under AOSSIE'S name that others could follow.
Like a new file format which just stores rhe metadata
Or one more thing that we can do is that we can give user an option to create a folder for the photos with same category/tag in the file system itself. So even if you don't use the software anymore, the folders would still remain.
I think the EXIF metadata approach would be better than storing metadata in a file that is separate from the photo itself.
The EXIF metadata lives within the photo itself. So, it does not get lost if the photo itself is moved.
Yes, correct.
Yes, I think something like this could help me with steps 1 and 2 of my workflow as well.
Generally, I think that, ideally, the source of truth should be EXIF metadata, and the database should be only a fast cache (that can be reconstructed by scanning all the photos and their EXIF metadata again, if necessary).
Another thing that would help me with steps 1 and 2 of my workflow would be using AI to sort photos by similarity (and not just to find faces).
For example, sometimes I have lots of "Christmas", "New Year" and "Home" photos taken at approximately the same time. In other words, I have lots of "Home" photos interleaved with photos of various events. It is a pain to sort them into folders, because of this chronological interleaving.
But my "Home" fotos all look very similar. So, it should be easy for an AI to group them by similarity and allow me to move them all in batch to a separate folder.
The AI could use not only a visual analysis of the photos, but could also cluster the photos by the dates they were taken.
This clustering by the date is different from simply separating fotos by day.
Sometimes multiple events happen in a single day. Sometimes a single event (e.g. vacation) happens over several days.
I am somewhat getting your issue. Can you please explain with the help of an example like image1.jpg and so on?
It would make the requirements more clear.
I have a fantastic idea for this, We could add a friends and family section where user can creates groups like family, school friends, college friends, etc. and upload photos of their friends and family with their names. Then we already have celebrity tagging, similar to that we can tag photos with names of the people and group they belong to appearing in that photograph. Further on clicking that person it will show all images in which that person appears. Same for group.
I am proposing two features (inspired by bruno's message)
- Advanced People Search & Sorting
- User will upload single/multiple photos -> application will search all images for those people -> Group them under a name (given by user) -> sort the photos by number of matched people appearing
- Use Case: Finding the best group photos for a screensaver where the goal is to see as many family members as possible.
- Friends and family feature
- User will create groups like family, children, school friends, college friends, office, etc.
- User will add members to a group (say family), he will upload images of members + their names
- Application will automatically search and group all those images and tag them with the name of the person
- User can see all photos of a member by cliking on his icon in the group
- The group will also contain all pictures of all the members, might sort this on basis of date then number of people appearing
- Give user option to create a folder for a group or member
WE CAN ALSO combine feature 1 into feature 2
@marsh idol I know current application has so many bugs but i am going to debug it real hard upcoming week using separate OSs (windows 11, mac, linux) and some tools to find performance bugs. I am just proposing these features for now, we can implement these in next to next week.
Regarding EXIF metadata
- We need to be carful not to corrupt the file
- Use XMP (Extensible Metadata Platform): The industry standard (used by Adobe Lightroom, Apple Photos). It allows for rich tagging and hierarchical keywords.
- Use PyExifTool library (as piexif doesnot support XMP) for write and read metadata
- Will store all the tags in metadata
- Use file first approach, update enteries of database if new file with metadata arrives
- Event Detection & Clustering
PictoPy will automatically group photos into event-like clusters using:
- Scene-level visual embeddings (not face-based)
- Temporal proximity with flexible time windows
- Density-based clustering to support variable event durations
- Event Review & Folder Suggestions:
event_id, event_label, start_time, end_time
- Detected event clusters are presented for review
- Users can rename, merge, or split events
- Entire event clusters can be moved into filesystem folders in a single action
Then we can also have , Event-Aware Search & Filtering, and not just People search, like we can add both the versions.
Once events are identified:
- Users can search photos by event name
- Filter events by time range or duration
- Quickly retrieve all photos belonging to a specific event without navigating folders
we could also store stable event labels (e.g. event name and date range) in XMP, while keeping event clustering logic app-specific and rebuildable.
This is a complex problem, we need to implement multi-stage solution for this.
We can implement this in PictoPy using a 3-stage process: Temporal → Visual → Semantic.
LOGIC
Step1: Adaptive time clustering (Calculate the time difference (delta t) between consecutive photos.)
- If delta t > 5 hours, make an event boundary.
Step2: Visual Similarity clustering
- Inside the "Dec 23-26" bucket, we convert every photo into a Visual Embedding
- We run a clustering algorithm (like DBSCAN) on these vectors.
- OR use OpenAI CLIP
Step3: Semantic Labelling
- This solves your specific "Festival" requirement.
- We check the date. If it's near Dec 25, we activate specific "Zero-Shot Classifiers" for Christmas.
- We ask an AI model (CLIP): "Does this cluster look more like 'a Christmas Party' or 'a normal living room'?"
TECH STACK
- OpenAI CLIP (completely free and open source) for step2 and step3
- Might use Clustering algorithm like DBSCAN if OpenAI CLIP underperforms step2
@marsh idol I have created relevant issues for these feature requests.
i am going to focus on debugging first then we can come to these features if you approve
Hi @marsh idol , I was going through the album route tests and noticed that many of them patch internal database helper functions directly. I’ve created an issue to discuss refactoring the tests to focus on API behavior rather than implementation details, which should make the tests more maintainable and resilient to internal refactors. Here’s the link to the issue: https://github.com/AOSSIE-Org/PictoPy/issues/1007
Would love your thoughts on this.
don't copy paste the response from LLM directly please, bcz this creates confusion.
By combining the features, do you mean implementing a 'Smart Groups' feature where PictoPy automatically organizes photos from the entire library into these groups?
Thankyou very much but that's not the case my friend
Doing backend since 2021
here is proof https://github.com/DeveloperAmrit/HeroicBloging-1 check its creation date (sorry for going irrelevant guys)
@tall bone You are now level 9! :confetti_ball:
By combining i mean giving button for feature 1 inside the page of feature 2
As it creates groups and "friends and family/ social circle" page is for groups
Check this out guys, this can be a game changer for us https://openai.com/index/clip/
it support Zero-shot classification for christmas thingy and natural language search as well
But this modal is 10x heavy than the models we are currently using
I found a lighter version of CLIP called CLIP_ViT_B32_Quantized.onnx (90MB) trained by OpenAI and is open source and free
@marsh wyvern @marsh idol i have implemented the smart event based sorting (christmas one).
I don't have nitro so i have uploaded the video here
https://www.veed.io/view/fff559f6-9d2d-43ae-8eec-b1dba3b29a99?source=editor&panel=share
Currently it is only based on OpenAI CLIP local model (size 154MB). Only Visual similarity clustering.
It has some deviations (like in the diwali section), i am working on that
So, it's like a "Groups" dashboard with button: "Find Group by Photos" right?
hi i am new to this project i have try to learn how the pictopy is working through documentation and all
can i work on this issue #979 Suggestion: Add a floating “Back to Top” button for easier navigation?
can anyone help me what should i do? Can i start working on the issue?
@marsh idol I think we can solve with steps 1 and 2 using a combination of image timestamps + AI-based visual similarity.
@marsh idol i have created a PR (https://github.com/AOSSIE-Org/PictoPy/pull/1009) please review it
I have created this, smart event based sorting.
For this "smart event sorting", we might have to give user two options
- Sort based on event and time - for camera folder where photos arrive in real time
- Sort based on only event - for downloads or whatsapp folder, where date of photo is irrelevant
Hi @marsh idol , I'd like to work on issue #852. Would you kindly assign it to me?
I've reviewed CONTRIBUTING.md and understand I should wait for the green light before opening a PR, so I wanted to check in first.
Also, just a quick follow-up: I previously submitted work for issues #969 and #992, but the PRs haven't been merged yet. If there's any feedback or changes needed on those, please let me know and I'll address them right away!
@marsh idol hey, will there be any meeting tonight
@north lion You are now level 8! :confetti_ball:
Hey guys I am new to the community. My name is Deep Shekhar Singh and I find this project interesting and aligned to my knowledge. I hope to learn a lot and add some value to the project !!
Hey @marsh idol I am doing robust testing and debugging of pictopy, till now i have found about 13 frontend issues and 1 backend issue. I will continue debugging and will update you
I tested it with pyright and have found around 63 type mismatches
@radiant mauve please review this issue
https://github.com/AOSSIE-Org/PictoPy/issues/1029
Hi @tall bone and @marsh idol 👋
I’ve reproduced this issue #1017 locally and have been working on it.
Both actions described in the issue now behave as expected in my local setup.
I haven’t opened a PR yet, in line with the project guidelines.
Please let me know if you’d like me to proceed with a PR or if there’s any preferred direction before that.
Thanks!
I appreciate your work @final vessel but I am already working on the issues i found. You may try to find some other bugs yourself and then solve them
Try to test backend with edge cases there might be some bugs
@tall bone i have already solved the issue locally , but i didn't noticed that you want to work on this issue on your own until i rechecked.
Ok no worries keep care now, let Rahul decide who should make a PR
@marsh idol
@knotty shore @bitter wind @unique grove Hey everyone! Just got PictoPy running locally. I was digging through face_clusters.py and noticed the DBSCAN logic re-processes the entire dataset every time. That seems like a pretty big OOM risk for large libraries (saw Issue #810 mentions this too).
Since there aren't any active PRs for it yet, I was thinking of mocking up a POC for 'Incremental Clustering'—basically just checking new faces against existing centroids to save memory. Is anyone else looking into this right now, or should I give it a shot?
Hi @bitter wind @knotty shore ike to work on issue #852. Would you kindly assign it to me?
I've reviewed CONTRIBUTING.md and understand I should wait for the green light before opening a PR, so I wanted to check in first.
Also, just a quick follow-up: I previously submitted work for issues #969 and #992, but the PRs haven't been merged yet. If there's any feedback or changes needed on those, please let me know and I'll address them right away!
@marsh idol ☝️
i would like to work on this issue #1033. Can i get assigned to it?
@marsh idol please review this issue
https://github.com/AOSSIE-Org/PictoPy/issues/1029
https://github.com/AOSSIE-Org/PictoPy/issues/1004 i would like to take this issue can this be assigned to mee i have worked such things before i would like to try? @marsh idol @bitter wind
Describe the feature As discussed in discord with bruno Problem - "Users often have photos from different events and everyday contexts interleaved in time, making manual sorting by date ineffe...
Hi team! 👋
I just submitted a fix for the Watcher service crashing permanently on exceptions (#1032).
I refactored the logic to use a Producer-Consumer pattern with a supervisor loop, so the service now handles threading safely and self-heals after network errors or timeouts.
Would love some feedback when you have time!
PR: https://github.com/AOSSIE-Org/PictoPy/pull/1037
Hey , i have a question about this project , is this project going to be running project for GSOC 2026 or is it going to be closed??
you can checkout #1458761086558801940 for gsoc 2026 list
Hey everyone, I’ve opened a PR that solves issue #1020 (white flash and app reload when clicking the logo or profile in the Navbar).
The fix switches the navigation to router-based links and is linked to the issue.
PR: #1039
Would be great if someone could review it when free. Thanks!
PR:https://github.com/AOSSIE-Org/PictoPy/pull/1039
I think there already exists a fix for it in pr
Oh okay — I couldn’t find an existing PR for this when I checked.
If there is one already, could you please share the link? Happy to coordinate.
Yeah i’ll just cross check and we’ll let you know
@everyone I have opened about 20 issues regarding bugs in pictopy library. I am strictly informing all contributors to not submit a PR to issues created by others unless assigned to you. Also don't ask me to assign those to you, i am not a mentor.
Thanks for sharing. I checked PR #637 and also found PR #917.
Both address the same issue, but they are larger in scope and still have some failing checks. My PR focuses only on the navbar navigation reload with a minimal change. Happy to go with whatever the maintainers decide.
@marsh idol are you able to have some free time, it would be great if we could have those weekly meets continued
Yeah the issues and pr’s are piling up…
@bold estuary You are now level 3! :confetti_ball:
Hi Amrit, the meets are still happening and were never discontinued... the next one would take place on 17th Jan at 10PM IST according to the schedule.
@marsh idol https://github.com/AOSSIE-Org/PictoPy/issues/1030 and https://github.com/AOSSIE-Org/PictoPy/issues/852 both aren't the same issue ?
Describe the feature Description Currently, PictoPy reliably supports common image formats like .jpg, .jpeg, and .png, but several widely used image formats are not fully supported or handled consi...
Finally the mentor is here
I am busy with shifting to a new city for my internship and I will be able to reply to you all within few days. I hope you understand 🥲
It has been stressful few days due to shifting.
ig it would be a headace now
good guess tbh
It's been 3 weeks since issue #852 was opened with no update from @frosty sundial , so I solved it. If the mentors assign it to me, I'll raise a PR
@marsh idol @knotty shore
@livid shuttle You are now level 2! :confetti_ball:
there's already a PR #996, it adds webp format, if you want you can suggest under that PR
How many same issues are open? 🤔
tbh the issues are assigned mostly during weekly meetings
When are these weekly meetings?
@sinful pumice ☝️
In this discord channel?
Yes
Rahul you've been a great help, thanks
Wlc
Community exists for helping each other
When is the weekly meeting held?
Have a look at this
@sinful pumice You are now level 2! :confetti_ball:
Weekly meetings are held on google meet. Mentor rahulharpal usually inform us some days prior
Thanks! he informs on this channel only right?
Yes
Hello evryone! ive setup pictoPy on my device and able to run it now can anyone guide how do i go about if i want to contribute
like ive a feature in mind what do i do to add it??
i am pretty new to open source sorry!!
You can open a new issue explaining the feature and ask if you can work on it
Thanks for pointing that out! I checked PR #996, allowing WebP is great, but my implementation is a complete overhaul that solves #1022 (SVG, AVIF, JXL, BMP, PSD) and #852 (Modern formats) in one go.
Crucially, I updated the backend to handle these files. Since the frontend web view can't natively display formats like HEIC or PSD.
I implemented everything except .exr( I figuring out for .exr)
I'd like to open a new PR for this. I'm ready to push it now if that works for you.
you can discuss those on meetings 👍
https://github.com/AOSSIE-Org/PictoPy/issues/1042#issue-3816282575 I have opened this issue now what do i do
sorry if i sound naive
@main jay You are now level 2! :confetti_ball:
Hi, I’m Pranjali Bhardwaj, a final-year CSE student. I’ve made a few contributions to InPactAI and would be excited to explore more AOSSIE projects. Looking forward to learning and collaborating with you all.
Hi everyone, just a small follow up, I had submitted 2 issues and 1 feature request:
here are the links 🙂
https://github.com/AOSSIE-Org/PictoPy/issues/994
https://github.com/AOSSIE-Org/PictoPy/issues/999
https://github.com/AOSSIE-Org/PictoPy/issues/993
looking forward to the meeting on 17th
Is there an existing issue for this? I have searched the existing issues What happened? If I am not selecting a folder in the onboarding folder selection step, the console is throwing a validation ...
Is there an existing issue for this? I have searched the existing issues What happened? Onboarding section has multiple inconsistencies: validation error when folder is not selected (#994 ) layout ...
I have many same similar and duplicate issues while scrolling
issues like 1030, 1022, 973, and 852.
White/Flash Screen Problems 1020 and 636
Image Deletion - 1044 and 795
Open Folder/File Buttons - 1017, 959, and 688
Silent Failures - 710 and 709
@livid shuttle You are now level 3! :confetti_ball:
Database & Threading - 943, 936, and 824
Sync Service - 1034, 1033, 1032, and 1031
Settings Page Lag - 876 and 612
Face Clustering - 961, 915, 810, and 350
Text Search - 902 and 661
Favourite Toggle Bug - 977 and 805
Navigation Bar - 988, 856, and 962
Scrollbar Problems - 1019 and 938
@marsh idol When will be our next meet??
Hello everyone! I’m Ata Ul Hai, a first-year CSE student. I’ve been exploring the AOSSIE ecosystem and am very interested in contributing to PictoPy.
I have already studied the official contributing guidelines, but I wanted to check if there are any other community-specific norms or additional documentation I should be aware of before I get started
Looking forward to working with the community and making some meaningful contributions
Fork and clone the repo locally, run it, and start with solving simple issues to get more familiar with the codebase. If you find any bug, don't raise issue and then the maintainer will assign it if it aligns with Pictopy.
Thank you for the guidance! That makes perfect sense.
If I encounter any bugs, I'll be sure to discuss them here in the channel first with my findings before taking any further steps. Looking forward to making my first contribution
One quick follow-up: once I have a fix ready for a minor UI/UX issue or a bug, would you prefer I submit a Draft PR for initial feedback, or should I go straight to a Full PR once I've tested it locally?
test it locally, if it works good locally and you think it's an important bug so raise issue and wait for maintaier to review. it might take time bcz Rahul sir is busy.
@young torrent You are now level 7! :confetti_ball:
Fork the repo, Add upstream and then setup the project. After that make your changes and push it to your forked repo. This will help you to create a PR (Pull request). Follow the PR format of AOSSIE and open a new PR for the issue you raised.
Although i would suggest you to get yourself familiar with the project first. Because opening issues for small additions will lead to an overflow of open issues.
I would suggest you tinker with the project find pain points and open a good issue targeting crucial improvements.
@chilly wren You are now level 2! :confetti_ball:
Sure, Thanks for the guidance
Hello folks!
I noticed a lot of new contributors were facing problem with the project setup and with contributions, so i built a platform tailored for PictoPy which can help you get started.
Also i have included a frontend preview for the app so users can tinker with the project firsthand before setting it up.
Here's the web app- https://pictopyhub.vercel.app/
@marsh idol Kindly review my PR #916 and let me know if you find any issues.
Is there a meet today?
@real gale any updates regarding today's meet ?
No meet today
Hi, I am Nikhil Gupta, pre final year CSE student. Looking forward to collaborating with you all
welcome to the community man
thankyou sir
I was trying to setup the project on my local environment but kept getting this error, even though the backend is working, I followed the steps here https://pictopyhub.vercel.app/
could you please share the log
yes
should i send image or the copy of the logs?
terminal 1
terminal 2
terminal 3
sync microservice is not listening
your sync services got shutdown
could you share SS with your dev tool
sorry but is this what you are asking for?
@hollow quiver You are now level 2! :confetti_ball:
"No folders found in database"
Don’t say sorry. I am asking about DevTools inspection.
like this
Just a heads-up: the pictopyhub.vercel.app site isn’t the canonical source for local setup and isn’t maintained by the core team, so some steps there can be misleading.
For the correct and up-to-date local/dev setup, please rely on the repository itself, especially
.github/workflows/build-and-release.yml, which reflects how the project is actually built and wired together.
wait, you pressed ctrl + C right?
logs shows the sync microservice was manually interrupted
yeah, that is the interruption
good observed
🙂 that is manually done, rerun the microservice and the frontend will work normally imo?
yeah, just bcz of manual interruption 👍👍
no i did not press ctrl+c, i've tried it a bunch of times, the same thing keeps on happening
can you record it on obs?
or like any screen recorder
yes
yeah, that will be helpful
that would help us understand what's actually happening
more when you can share ss of dev tool
this is expected
your backend is running correctly
as you don't have any specific folder for pictopy to works on with sync-service
yep i removed the folder
@north lion You are now level 9! :confetti_ball:
I was checking what npm run setup actually does and traced it to scripts/setup.js.
The script explicitly handles only:
Windows (win32) via setup.ps1
Debian-based Linux via setup.sh
macOS (darwin) isn’t handled and the script exits with “Unsupported operating system”.
are you using manual setup or script-based setup? @hollow quiver
he used this
yeah
I'd suggest you to get through manual setup from the start, I can help you with it
okay so, redo everything from start, cuz if the setup is broken then some internal features might as well break
ah ill give you the code
you mean restarting from the cloning step
@hollow quiver You are now level 3! :confetti_ball:
i mean you can
okay
clone it from your fork
ok
cd PictoPy```
ohh i am using github desktop 😁
cd backend
python3.12 -m venv .env
source .env/bin/activate
python -m pip install --upgrade pip
pip install -r requirements.txt
uvicorn main:app --host 127.0.0.1 --port 52123
do these for backend
after running backend, go to another terminal
cd sync-microservice
python3.12 -m venv .sync-env
source .sync-env/bin/activate
python -m pip install --upgrade pip
pip install -r requirements.txt
uvicorn main:app --host 127.0.0.1 --port 52124
run the microservice, open new terminal for frontend
okk
cd frontend
npm install
npm run tauri dev
then frontend
if you want more details,
.github\workflows\build-and-release.yml
can check this
the cloning hasn't ended
its normal
okay, thankyou very much😁
lmk if you get stuck
yeah👍
are you done with it?? @hollow quiver
downloading the requirements
@hollow quiver if you still face any issue, please refer to https://github.com/AOSSIE-Org/PictoPy/blob/main/docs/Manual_Setup_Guide.md
yeah, i am still facing the same issue
try going through the link rahul gave
same error, I don't know why, I followed each and every step from the start
@hollow quiver You are now level 4! :confetti_ball:
I will retry in the morning
all the requirements are installed, i don't know why it shows error for that
i get this error when i try to reinstall the requirements
What it does:
- Starts the backend and sync microservice in the background using their respective virtual environments
- Launches npm run tauri dev as the main process
- Streams logs from all services into a single terminal
- Automatically shuts down backend and sync when the Tauri process exits
- Fails fast if either backend or sync crashes
- Reduces log noise by collapsing repeated log lines (e.g. status polling)
Why I made it:
- On Windows, running PictoPy currently requires multiple terminals
- Backend and sync keep running even after closing Tauri
- This provides a cleaner dev workflow and avoids orphaned processes/ports
@marsh idol would love to know the feedback on this
Its good but i have already implemented the spawing and closing of all services via tauri application itself. So no more scripts to run, just open pictopy and all services will run
https://github.com/tushar1977/PictoPy/releases/tag/release
you can test this by downloading a release file from my fork
What we’ve added is for Production releases. @north lion mentioned “npm run tauri dev” I think he’s proposing for development mode
oh mb
but this is also applicable in development also
it is necessary to get issues assigned by mentors before raising any pr. i couldn't find any instructions related to this in the contributing.md