#Rein
1 messages · Page 2 of 1
sure , should I Add the issues which are yet to be evaluated in the proposal ?
i didn't get it
can you clarify?
I mean you said right you would be reviewing the PRs related to 208 during the evaluation period
so should I add those in my contribution list while making the proposal ?
yeah you'll have to
alright
@slim pollen You are now level 2! :confetti_ball:
I had a small query - is it necessary to have the contributions only on the repo related to the project ? or contributions from other repos are also considered good to go in a proposal ?
for example if the project is Rein - so is it necessary that contrubutions made into the repo rein should only be listed ? or contributions made in different repos can be considered as well ?
can you please ans it
Hi @oblique bay
I can test on Windows and Linux (using VMs for X11 and Wayland), but I don’t have access to macOS hardware.
Would validating macOS builds via GitHub Actions and documenting any platform-specific limitations be sufficient?
contributions to project you are targeting is good to have but not mandatory if you can demonstrate thorough understanding
but PoCs for 130 and 208 is mandatory
I think its possible to install macOS on QEMU
but you'll have to check for VirtualBox
alright thank you , I've already attached screenshot for the webrtc issue in the PR
Ok sir
Is this good?
@oblique bay please check PR #249
Hello @oblique bay I am Abhik Mudi and I am interested in contributing to this project. I raised issue #248 about The sensitivity and invert scrolling option should be the same in mobile and desktop. I want to work on this.
Does anyone use ubuntu wayland? There only the cursor is moving from mobile. Other operations like scroll, click does not work as there is no ydotool scripts for them.
Make sure the demon is running in bg ydotools should work fine in wayland
Yes it is running. Only "move" is working other operations like "scroll" is not working
Other operations are only using nut js
Try debugging you can use tools like libinput
@elfin abyss @bitter urchin
directly use https://github.com/AOSSIE-Org/Rein/blob/main/shell.nix with nix or install those deps
also test https://github.com/AOSSIE-Org/Rein/pull/236 as mentioned in the issue 220
Update:
I was able to get almost everything working except for mouse movement. There’s a strange issue where the command works correctly when I interact with ydotool directly, but it fails when triggered from the app, even though both are supposed to execute the same command.
During debugging, I noticed it reports “event not found.” I’ll investigate this further. Since the command works when run manually, there should be a way to make it work from the app as well. I’ll also look into what you have mentioned.
@pure fossil create the issue for screen mirror in executables
Ok
[Bug]: Screen Mirror in Executatbles · Issue #254 · AOSSIE-Org/Rein https://share.google/oeF1TbbgTaEI4kBKs
I have created issue, I will solve it and get back soon
Hi sir @oblique bay ,
I recently opened this issue after running a SAST security scan and manually reviewing the results:
https://github.com/AOSSIE-Org/Rein/issues/256
From the findings, the most critical concern appears to be the usage of "eval()" / "new Function()" in parts of the codebase that process websocket communication and trackpad input events. Since Rein handles remote device input streams, executing dynamic code in these paths could introduce serious security risks if the input is ever manipulated.
My proposed plan is to address the issues in phases:
- Highest priority: Remove all "eval()" / "Function" usages and replace them with safe alternatives such as structured parsing or command mapping.
- Reliability improvement: Add proper "try/catch" handling for the unhandled promises in "InputHandler.ts" and related modules.
- Maintainability: Refactor complex components like "ConnectionProvider.tsx" to reduce cognitive complexity.
Before opening PRs, I wanted to confirm
- Whether the secret flagged in ".github/workflows/label-merge-conflicts.yml" is indeed a false positive.
- If the proposed remediation plan looks good.
If everything looks fine, I can start working on the first PR for the "eval()" security fix.
Thanks!
In my case, I could not get 'scroll' working with ydotool. Even when calling the binary via Node child processes, the overhead makes it too slow for real time tracking, and it completely lacks other multi touch supports also.
However building a custom uinput bridge is working perfectly. I was able to feed it absolute coordinates and successfully trigger native multi touch gestures (like smooth scrolling) directly through libinput.
@bitter urchin You are now level 2! :confetti_ball:
@maiden thicket I managed to get the application working inside the VM after a bit of tinkering. If needed, I can create a PR with documentation for setting it up in a VM.
Also regarding issue #220. the PR works both on my machine and inside the VM. For the local IP detection, I don’t think it’s an issue in a VM because the ports can be easily exposed by adjusting a network setting.
Also i should mention running the app in vm is really laggy and at times causes 1-5 secs of delays and latency. But that might be also due to my machine .
your choice
yeah raise the PR
you can check the instructions once added by @elfin abyss if the issues are still not fixed after that you can investigate further
also we are already palnning to move away from nutjs check issue #130
those were mostly hallucination
Yes that's why I was trying libinput and it's working smoothly in my ubuntu wayland
try options other than libinput as well and also consider other platforms
Yes I am researching that
the hold buffer is not visible someone can try to address that
yes Rituraj i am looking into it
Yes @oblique bay i have fixed it, if you say i will raise the PR.
@cinder zephyr You are now level 7! :confetti_ball:
@oblique bay I saw #242 about reducing electron executable size and also pr #255 ig it already enabled asar and moved some dep but the issue was reopened so is there addirtinal optimization needed ??
Sir @oblique bay ,is it good now see
Tell me if it's ready for push
Please review it sir when you are free
https://github.com/AOSSIE-Org/Rein/pull/260
@muted ruin You are now level 17! :confetti_ball:
I've checked it. The implementation completely deviates from the description in PR.
It completely relies on NutJS.
Focus on #130 for now.
Thanks for the feedback. I’ve now worked on #130 as suggested and opened a PR for it. Whenever you’re free, could you please review it and share your thoughts?
https://github.com/AOSSIE-Org/Rein/pull/262
Hey @muted ruin regarding #130 we have to make a working prototype and not directly raising a PR. The PoC will be then evaluated.
I've already made a PoC using Koffi too you can check. https://gist.github.com/Rozerxshashank/fbaab306ecaac822a49242fcac7b969f
We can discuss the replacement of NutJS by koffi.
Thanks for sharing the PoC. I’ll review the gist and understand how the koffi FFI approach is implemented. After that, we can discuss the replacement strategy for NutJS.
Hi @robust coral , I checked your koffi PoC — it was really helpful and confirmed that using koffi as an FFI layer is a promising direction for Issue #130.
To better understand the approach, I created a small minimal PoC focused only on the trackpad scope mentioned in the issue (move, click, and scroll). The goal was simply to validate that native input APIs can be triggered through koffi without relying on NutJS.
Gist: https://gist.github.com/Muneerali199/12227af56abb5ae225c7649d762f4676
Quick run:
"npm install koffi && node virtual-input-poc.cjs"
Current approach:
• Windows → "user32.dll" via "SendInput"
• Linux → "/dev/uinput" (should also work on Wayland)
• macOS → "CoreGraphics" ("CGEventPost")
I’ve verified the Windows part locally so far. This PoC is intentionally minimal just to validate the feasibility before thinking about deeper integration.
Please let me know if this aligns with the direction you had in mind, or if you’d like me to adjust or expand the PoC further.
@oblique bay can you please review this pr once
https://github.com/AOSSIE-Org/Rein/pull/263
@pure fossil
Once issue #107 is resolved, you all should focus on the PoCs mentioned in the 2026 idealist, as noted in the related issues.
As those are mandatory for the proposal, they are not supposed to be finished products, so don't aim for perfection as the architecture will be finalized later. Just address the constraints mentioned (e.g., Wayland, cross-platform, and trackpad).
rest of the minor issues can wait
Ok 👍
Sir @oblique bay
raise it, no need to ask
@cinder zephyr review #265 instead
Yes @oblique bay Reviewed, its working good and expectedly 👍
open issues marked as high priority are mandatory for proposal
rest are up to you
@cinder zephyr
@pure fossil u'll have to check both PRs, 267 and 263
if needed you can ask for separate pr for screen cap and flatpak incase of 267
Ok
@pure fossil You are now level 12! :confetti_ball:
I am reviewing few mins please
check thoroughly whenever you've time, it doesn't have to be now
Ok, sorry for inconvenience currently drives are coming to clg and as well exams so I have not given much time, but I will try my best
I recommend focusing on placements then, as it much more important
I'll check them myself when I get time
@oblique bay should I make changes for the 2 issues u closed ?
@muted ruin close the existing pr, and create separate ones for flatpak and screen sharing
and add a screen grab to show they are working
No I will check, sorry bro if I hurt you, my intention was to not hurt you, but clearing the things was my intention
nope
@oblique bay You are now level 31! :confetti_ball:
okay fine...I can add those 2 issues in my Proposal right ?
Will do in the morning
only the PR that has been closed as PoC
or if you can do better add that instead
alr thank you
@slim pollen You are now level 3! :confetti_ball:
We can share our proposal draft in the group ??
#gsoc-how-to-apply
No i mean for getting some suggestion as for improvement
I actually meant what I said, reviewing is my responsibility after all and placement is more important
you can still create the PoC if you get time and submit the proposal. I'll consider it regardless
we dont have a template this year
Ok I am also trying screen mirror for electron
Ok ok thank you sir .
@oblique bay I think that PR #263 is correct from my side I have checked it, I think it's ready to get merged
the executables are able to connect to websocket?
if yes the ping (ms) will be shown in the navbar
@topaz wind close the current pr, and sync the repo
otherwise you wont be able to generate executables locally and test them
alright
Sure sir
Testing again
what about PR #249 ?
you can include whatever you feel relevant
make sure its easy to follow and issues labelled as "priority: high" gets more focus
but they aren't merged yet right ?
answer remains the same
fine
@topaz wind I think you are addressing the wrong area in PR 270
build the executable and test it, once you are sure the websocket is being connected i.e. ping is being shown in the navbar and lan IP is correct raise the PR
the generated executable should be tested not the dev server
okay I’ll build the executable and test the websocket connection from the packaged app to confirm that the ping appears in the navbar and the correct LAN IP is being used I’ll update once I verify it
@topaz wind You are now level 5! :confetti_ball:
@oblique bay We tested it, not fetching websocket LAN IP, we will try tomorrow
Hi @oblique bay i am working on issue #130 and #208 to make the POCs.
My question is that how to show the POCs to you or how to use POCs in the proposal ?
- does we need to implement complete functionality in the project and then record the video of working?
- Or we need to raise the PRs for POC reviews ?
- or we need just code of changed files and then to show just code of files not even implemented in project ?
- or we need to raise the PR and then use links in proposal?
is there any other way then please clear it.
for PoCs, I need access to the entire working codebase, so it's up to you
you can use a public branch, a private repo (giving my GitHub ID access), or any other suitable method
regarding the video, add a link to it in the proposal
@oblique bay I wanted to confirm my understanding of issue #208.
My approach is to keep the existing screen capture method on the desktop,
but replace the WebSocket streaming with WebRTC for sending the screen/audio stream to the mobile client.
WebSocket will still be used for signaling and and establishing the connection for WebRTC.
Does this align with the expected approach for the PoC?
screen capture will be via Navigator, as it's universal.
she rest of the sharing will change, utilizing WebRTC for low-latency streaming. Check the issue description for constraints.
screen sharing should be tested on Wayland, especially with other platforms as well
regarding WebSocket, it's up to you. If isolating screen sharing away from WebSocket makes it simpler, then go for it. I dont have experience with webRTC myself
@muted ruin share a video of running the flatpak with flatpak run
also check how we can push the flatpak to flathub. check the repos of other apps on flathub for reference
Sorry sir last night I am in very high fever so can't able to respond . I'll send you video in few hours and also I know how to publish flatpak to flathub . It's like app store
Regarding publish to flathub : the process is we have to fork the flathub repo and add our files to the apps / folder
Then we create a pull request
After that flathub mentor will review it . Once merged , users can install
Flatpak install flathub com.rein.app
Like this
?
@robust coral You are now level 12! :confetti_ball:
What ??
wrong channel
Hey @oblique bay I've sent you an invitation for my PoC on github. I've tested it from my side. all the main things like input injection and screen mirror(low latency) is working fine without nutjs. I just want you to take a look and test it from your side and give me feedback.
I just want to confirm if it is good for wayland
Then I'll improve accordingly.
Hey @oblique bay is it working on your system?
I'll be away from my machine for a few days, so you'll need to test it yourselves.
That's why I've been prioritizing issue #107. So everyone can test across platforms independently without any bottlenecks.
that video, didn't show the app itself
and i've added some more resources to the flatpak issues itself
@oblique bay I have created issue for npm run electron not working because if it we need to package(npm run dist) frequently for small changes which takes more time to package and requires more time to test the electron app
@oblique bay Please review my pr when you are free: https://github.com/AOSSIE-Org/Rein/pull/285
@oblique bay I am trying to access websocket for LAN IP & Screen Mirror but still not completed
That is some issue while recording it not show both thing but it open the app at the same screen
@oblique bay have you tested my POC of issue #208?
I have invited you on github for that repo.
just confirm me that if it is making sense then i can also improve it more and i will start on POC of issue #130 as well
No I've not tested anything yet,
I recommend testing PoCs on VMs from your side and try to resolve #107
I can merge PRs related to .md, as I wont have to test them
Hi @oblique bay
I have opened a documentation PR for Rein that adds instructions for testing the app inside Virtual Machines (VirtualBox / VMware).
PR: https://github.com/AOSSIE-Org/Rein/pull/287
I would appreciate any feedback or suggestions. Thanks!
Can you share a video where your host machine connects as a client to the server in the VM?
Also, is the IP shown in the VM still wrong? Recent changes should've fixed it.
And we should avoid closing the issue with the PR unless functionality has been tested on all three OSes.
@oblique bay check the IP here
Thanks for the clarification!
I'm currently setting up a VM to record a demo showing the host connecting to the Rein server running inside the VM. I'll share the video shortly.
I'll also double-check the IP inside the VM to see if the recent changes fixed it. Noted about not closing the issue with the PR until it's tested across all three OSes.
@azure isle You are now level 2! :confetti_ball:
You raised a PR on an issue on which it was clearly mentioned to check the feasibility of testing the app of virtual box without actually testing it on VM??
Hi @oblique bay ! Thanks for the clarification earlier.
I have updated my contribution based on the feedback. I removed the VM testing instructions from the README and added a “Testing Rein on a Virtual Machine” section in CONTRIBUTING.md with only the network configuration (Bridged Adapter) as suggested.
PR: https://github.com/AOSSIE-Org/Rein/pull/288
Please let me know if any further changes are needed. Thanks!
Addressed Issues:
Related to #107
Description
This PR adds a VM testing section in CONTRIBUTING.md describing how to configure Bridged Networking when running Rein inside a Virtual Machine (Virtual...
Thanks for pointing that out.
You're right — the issue mentioned checking the feasibility of running the app on VirtualBox. My PR focused on adding the network configuration documentation in CONTRIBUTING.md, but I did not yet complete the actual feasibility testing inside a VM.
I will test Rein inside a VirtualBox VM and verify whether the host machine can connect to the server running in the VM using bridged networking. If needed, I can update the documentation or open a follow-up PR with the verified steps.
Please let me know if you would prefer that I first complete the VM testing before documenting it.
Sir @oblique bay ,can you review this pr when you are free
https://github.com/AOSSIE-Org/rein/pull/289
add how you plan to design the logging system and how the logs would look like
close the PR and link it in the proposal if its final
given IP detection is working on the Linux VM, I'll mark linux as complete on #107
i've also updated the Linux-specific note in the README
once someone confirms for macOS VM, I'll close the issue
PoCs can now be tested on wayland with VMs
once done, can optionally help with the Electron-related issues
or even check screen sharing on non-Chromium browsers (not much important)
“Got it, thanks! I’ll add the logging system design and example log format, then close the PR and link it in my proposal.”
Hey @oblique bay , this was the test of my PoC. It is working on wayland both input injection and screen share. Kindly confirm this from your side too
if you unblock port 3000 of the vm you should be able to access it outside
Yes I tested it outside too. This screenhshot was just an example
We're getting close to the proposal submission period, so everyone should start focusing on that with PoCs and issues marked as high priority for proposals.
Include link to source code of PoC (and working videos if there) in the proposal itself, prominently above the rest. Otherwise I wont be able to find them
Project size is large
if PoC repos are private, grant access to @imxade (and maybe bruno) on github. Also include a video testing the PoCs across different platforms via VMs (MacOS, Windows, Linux ...)
You all need to share the expected end-to-end architecture, both as text and diagram, including frontend and backend. This will show what you finally want to deliver not what it is
I've added a rough expected design as a placeholder generated with Claude based on the description, but I didn't check it thoroughly. Still, you can use it as a reference, improve accuracy and details in your proposals.
Koffi seems good for now as we wont have to manage our own bindings, go with it for PoCs if you can't find better alternatives.
PRs can wait. I've already closed 175 and thats enough for now. Even if you haven't contributed, you can still apply.
Proposals are to be generated via the Google Form mentioned in #gsoc-how-to-apply and directly uploaded to the gsoc portal for review
First words of project title should be "Rein:"
Avoid trusting LLMs too much, as you'll have to pass multiple official evaluations during the GSoC period with actual deliverables.
As far as I remember, you all can update your proposals as many times as you want in the GSoC portal until the deadline.
I won't be available after next week, so will try to review proposals before that no need to ping me specifically for feedback. Feedback will be publicly provided here, if I feel the need.
I can only see 2 proposals for now. Also note that one can apply for upto 3 projects but only consider those which you are absolutely sure about.
I'll share on Monday for review
My proposal and Prototype are ready you can review it. I also did the replacement of websocket in my prototype today only.
My proposal took a bit of time but is very close to being done I'll share by Wednesday
You all can follow my socials for sure, but since I can lose those accounts anytime, I don't carry convos or make connections there unless necessary.
Regarding help with proposals, check the pinned comment there's nothing more to add. For reference, you all can check previously accepted proposals. Carrying related convos outside of GSoC is against org policies.
@sage blaze
End to end expected architecture is a must
Also include a video testing the PoCs across different platforms via VMs in your proposal/repo
My proposal has the end to end architecture of how the screen mirror will work and second is how the input injection works. Do I need to add the full end to end architecture of working of rein?
yes
Like I have to create a new space in proposal with heading something like my understandings of rein??
thats up to you, I'll just check what how and why about the architecture and the PoCs.
Note that it should be the expected architecture not the existing one.
Like we are replacing websocket, webapi and nutjs so the expected architecture should have their replacement ?
Like we prepared the poc
ofc, it should show what you are planning to do and what the project will finally look like
Yes thanks for the clarification. I thought like there are things like the qr connection their architecture will be same so how will they be different from the existing one.
whatever remains the same will be shown as the same
thats obvious
Yeah now I get it. Thanks
in short the architecture will simply show what the project will finally be according to the candidate, thats it
I hope its clear now
Yes thanks 🙂
Hey @oblique bay , sorry for the delay. I completed the Windows VM testing in VirtualBox and recorded the process.
I was able to get Rein running inside the VM and open the /settings page successfully. The app generated the QR code and LAN URL correctly inside the VM.
One thing I observed during testing: on Windows PowerShell, npm run dev returned after predev instead of starting the dev server, while npx.cmd vite dev --host worked and started the app successfully.
ok, focus on the proposal for now
@oblique bay You are now level 32! :confetti_ball:
If you're confident about the project, don't let someone else's engagement or activities affect your hope. Those aren't the only things that matter.
- A haiku by me
Hi @oblique bay , I also observed an issue during the Windows VM testing, so I’ve raised an issue #315 for that separately for tracking. As you suggested, I’ll keep my focus on the proposal for now and will submit the proposal soon.
@oblique bay where do I check the size and details of the project for making proposal ?
All projects are large, check the pinned comment
Hello sir @oblique bay ,can I share my proposal now ??
Alright sir
Sir @oblique bay ,I have prepare my proposal please let me know if something wrong I have done
Include link to PoC codebase (and working videos if there) in the proposal itself, prominently above the rest
I'm not going to search my mails
also if you want to you all can explore bun+electrobun for packaging
First of all, it wasn't needed, but thanks for sharing it publicly so others can refer to it and make sense of the feedback. Note that this feedback is relevant for everyone. Also, go through the pinned post again.
I need no gist, but working PoCs to directly test, preferably based on Rein itself. I've got some invites to private repos, but none of the proposals contains the link to it at the moment. It should be there during final submission so I can test. No need to create separate codebases to show different PoCs; you can include #130 and #208 (anything more) in one.
No need to include code snippets in proposals; clearly describe the logic instead (maybe with a flow diagram?). If proposing UI changes, add an image showing the geometry. For example, if there's a controller UI, show how you'd position them including all buttons. Only if time allows will we go with controller support; it'll be a separate page from the trackpad page, allowing us to utilize the whole screen like cloud gaming apps.
Now coming to the particular proposal: be very careful while testing what you proposed.
The target architecture diagram is too simple and has no significance.
PR #317 doesn't do what it says. If that's otherwise, add a video pasting from phone's clipboard on host.
The problem of Electron executable unable to access WebSocket seems unrelated to 0.0.0.0 and might already be fixed when we move away from WebSocket.
The repo doesn't have anything related to Flatpak. Flatpak support will have a separate repo under Flathub org.
No need to have a separate section for problems; briefly introduce them with the solution itself. Everyone knows the problems, so focus on solutions and why they should be.
Better go with the project architecture description as a whole instead of going problem by problem, so one section doesn't contradict others and we get the overall idea. For example, we don't plan to use WebSocket, but some parts of your proposal say otherwise while others align.
https://drive.google.com/file/d/1rz-nQ9Rt_Frg9UIDwJCk38tABidIedE2/view?usp=sharing Please have a look at my proposal and share your thoughts.
@robust coral You are now level 13! :confetti_ball:
@oblique bay I've given you the access
I dont have the access, if you want to share it privately reupload it via the gform (if you cant tag bruno in #general ) otherwise share it as a pdf here directly
I'll be giving generic feedback if I feel the need
@robust coral @cinder zephyr @elfin abyss
you all can apply for triageur, only if interested
#announcements message
ok, but still reupload it via gform if possible
so that I can keep track of them together
This one?
#gsoc-how-to-apply message @robust coral
I've already filled it
you cant do it again?
Yes I can
you got it from github? cuz I've set it as private
yes
In the apache license
What is triageur?
#announcements message
Yes sure Rituraj
Why not
i am interested and am going to apply for it
Thanks for considering me for it
like its not necessary or anything (not even related this project) only do it if you want to, I've cleared the path already
just in case
Yes I am interested by myself
Actually i wanna go deep in open source so any this type of opportunity i would not skip
And also i am interested in Rein as well and understood it totally so i will do it
Same
Done
you all should upload the proposal directly on the gsoc portal from here on
it can be reuploaded as many times as you want until the deadline
I've got access to them and its much easier to navigate there
title/name of the project should start with "Rein "
Done uploaded it on the portal, the title starts with "Rein". Please have a look and share your thoughts
Hi @oblique bay I have uploaded my proposal on GSoC portal
please review it and am waiting for your feedback
Same
So here's generic feedback based on the proposals I've seen:
Current deliverable architecture diagrams are not detailed enough. Instead of high-level architecture, I'm hoping for mid-level architecture for more clarity on what you exactly plan to create. You can take the whole page either vertically or horizontally (or even break it down into levels, that's up to you). Just make it more specific than generic, as it'll be part of the actual documentation. Refer to the architecture in README and sequence diagrams from merged PRs; add details and improve upon them accurately as per your plan.
We don't plan to have mDNS or local URL, as it doesn't add much.
Gestures won't be handled by us but by the OS. We'll just send them as-is from the frontend, supporting as many fingers/touches.
Nut.js and WebSocket won't be used at all. Make sure your points don't conflict with one another.
No need to include code snippets in proposals; clearly describe the logic instead (maybe with a flow diagram?). If proposing UI changes, add an image showing the geometry. For example, if there's a controller UI, show how you'd position them including all buttons. Only if time allows will we go with controller support; it'll be a separate page from the trackpad page, allowing us to utilize the whole screen like cloud gaming apps.
Detailed description format:
-
Title: 'rein: ...', project size: preferably large (22 weeks) but you should decide yourself and be careful, Discord @, PoC codebase link, abstract
-
Resume
-
Maybe index/table of contents
-
Expected architecture diagram (showing what you plan to deliver finally as a whole, not just the current state. This should be detailed mid-level, not generic. Refer to the diagram on README and sequence diagrams from CodeRabbit under merged PRs)
-
If the mid-level architecture doesn't fit well for visibility, you can rotate it or break it down into several diagrams as per your choice.
-
Detailed description of entire architecture
-
Challenges (not limited to GSoC-only issues) and which part of the proposed architecture solves them. This is not mandatory tho
-
Packaging, test both electron-forge and electroBun
-
Future expansion
-
Rest, whatever you want to add
-
GForm response (this might likely be updated soon by Madhav so trying submitting the GForm again, so the entire formatted proposal will be provided by the form itself; everything till point 9 should be part of the detailed description, you may exclude resume)
Try to follow the order mentioned above and submit the proposal directly on GSoC portal, it can be updated until the deadline.
PoC:
Just incase, a very simplified expectation from PoC:
It should take the existing codebase and replace WebSocket and Nut.js with WebRTC and Koffi (or whatever you think will be good) for screen mirroring and trackpad. Test them in Linux, Windows, and macOS VMs and maybe add a video showing that to the PoC repo README and proposal.
repo: https://github.com/AOSSIE-Org/Rein
The PoC can be improved until we start ranking.
Also once again, you can apply for more than one project within the same org.
Cross-Platform LAN-Based Remote Input Controller. Contribute to AOSSIE-Org/Rein development by creating an account on GitHub.
I'll be barely active for the rest of the month.
So help newcomers navigate, and ask them to go through the pinned comments
OK
Thanks for the feedback
Thanks
@cinder zephyr You are now level 8! :confetti_ball:
Sorry for the delay I'm a bit late but I’ve submitted my proposal as well. I would really appreciate any feedback
We have to take rein codebase in new repo right? And then we have to implement PoC right? Means if I implement PoC in the same rein forked repo then how to mention it in the proposal?
Don't need to fork. Copy the codebase into another folder then work on it
I have uploaded the google form before , but I don't have any section about the packaging section. So should I submit another response?
Check the pinned comment for feedback, format, and expectations.
Don't stress about specific feedback.
Ask newcomers to do the same in my absence.
Will do
upload on gsoc protal as well
Ok I have to merge the pdf generated from that form and the detailed description right?
I think there is an option to edit your form response so you can edit it anytime
But the uploaded file can't be edited in gForm
There was also a field in form to upload the link of google doc on which you have written your response and you can update any time your that doc file
But if you submitted the google docs link
Hello @oblique bay After submitting Google Form I got PDF of my Response and in email it is written that "Add your detailed proposal at the end of this document" so I have to paste my proposal in that file or I have uploaded to to Google Drive, that link I have to add in my original proposal?
Hey you don't need to create a new proposal. You have to merge the pdf you uploaded in which you gave your detailed description with the mail pdf.
Okay means I have to paste my complete proposal made by me in that pdf right
Not paste actually first you got the copy of the response in mail. Create a pdf of that then merge your pdf of detailed description that you uploaded in section 6
Create pdf of your google form response then merge it with the pdf you uploaded on this.
Okay
@pure fossil You are now level 13! :confetti_ball:
So that means I have to merge the pdf of this form to my proposal pdf right? Like proposal pdf + form pdf ?
@robust coral And should my proposal contain repeated things that i already mentioned in the form?
No not to repeated thing you have to merge your detailed description only containing your project idea and all which should not contain things which are already in the google form
@austere flower you may have uploaded a pdf in section 6. You have to merge this pdf only with the pdf you made of your google form
Okay now understood
Thanks
attach detailed proposal at the end of google form pdf
@robust coral
what does detailed proposal exactly mean. Is it that pdf asked in section 6?
Yes
@robust coral You are now level 14! :confetti_ball:
Okay 👍
Right, but detailed proposal means not actual proposal no am I right?
Yes, at the end you just have to merge you google form pdf with the detailed description pdf which we uploaded on section 6 and then submit on gsoc portal .
Because the main details which usually we add in proposal are already filled in google form. The thing that remains is the detailed description of our idea
Ok starting pdf is generated pdf by gooe form then at end we have to merge our original prepared proposal or detailed proposal which we have upload in section 6 because detailed proposal I have uploaded different from my original proposal or in detailed proposal in section 6 we have to upload original proposal?
Yes but the detailed proposal you attaching at the end should not contain the things which are already filled in google form
Check this they mentioned it
Ok them we have to not upload our original proposal there in section 6 because I have uploaded consider description pdf in that section
Yes simple!
Ok
Hello @oblique bay I have uploaded my proposal on GSOC portal please review it when you are free and if needed give suggestions, changes or improvemens, I will be happy to make changes.
Thank You
Once again everyone, please carefully go through the parent comment providing the format for the detailed description PDF. Check it very thoroughly word by word and follow the order. Also, check the pinned feedback as well.
@oblique bay In Gform there is not coming edit response, so how to update my detailed description file in Section 6 ?
Or I have to only update original proposal which is different from detailed description in section 6?
Hey go and check the mail you may see the edit response button there
Ok thanks
Yha but we can't edit detailed proposal in section 6, we have to resubmit the Gform to change detailed proposal
Yes but there is an option to submit a link of your detailed description below that upload option
If you want to edit that then edit it locally upload it in google drive, and paste the link there
Ohh
Hey upload option is gone ??
Now its showing that we have to submit md
Yes now we got these two options.
Now I don't think so we have to merge anything. That thing is removed now
Is there any difference in the final format of the doc. Do we need to remake the old proposal
We have to fill the form again as there are some changes in the form
We have to remake it
Hey @oblique bay
I'm Upendra (@upendra512 on GitHub). I've submitted 6 PRs fixing open issues in Rein. Would love to get reviewed!
Memory Leak Fixes: #319, #322, #323
System Stability: #324, #325
Developer Experience: #326
All PRs are focused, single-file changes with proper error handling and backward compatibility. Happy to address any feedback!
I won’t be reviewing PRs for now, so please don’t spend your time there.
Focus completely on your proposal, as per the pinned messages.
also make sure diagrams are legible
Also, if there are formatting issues with proposals generated through the form, please inform Bruno and M4dhav.
Hi @everyone, I wanted to confirm one thing regarding the proposal submission.
Earlier it was mentioned that the detailed proposal should not repeat information already filled in the form. Since the form has been updated and proposals are now generated automatically, is it still necessary to avoid repetition, or is it fine if some information overlaps?
Just wanted to confirm to avoid formatting issues. Thanks!
I wanted to ask if I have already filled the timeline section in Google form then I should remove that section from my proposal doc ?
I would suggest not to repeat the things you already have filled in the form. Because at the end you will get a generated one with all the things you have written.
Thanks 👍
@azure isle You are now level 5! :confetti_ball:
It’s up to you, do check the generated PDF yourself.
Or, if you prefer, you can even create a proposal without the Gform.
yeah you can remove it
has anyone tried macos on vm?
here's a guide I found
https://iboysoft.com/howto/install-mac-os-on-virtualbox.html
I tried to do the setup but due to storage issue the vm got stuck. Had to remove it. But it is working.
I also had a video of it testing. I have attached it in my proposal.
doc needs to be updated?
Yes tried and testing is added in proposal
@oblique bay In gsoc portal we have to upload pdf or docx file?
Generated by filling gform
Ok Thank You
@oblique bay I have uploaded updated proposal on GSoC portal please review when you are free and suggest changes and improvements if needed, I will be happy to make changes accordingly.
Thank you
This is required in macOS
Hello @oblique bay I have resubmitted the gForm and also uploaded the proposal on the GSOC portal according to the changes. Please review it whenever you are free.
Due to storage issues I couldn't run mac on vm but in my proposal I have shown how my poc will work on Mac through my friend's pc.
@bitter urchin You are now level 3! :confetti_ball:
Hi @oblique bay , I have submitted my GSoC 2026 proposal for Rein on the GSoC portal.
Please review it when you’re free and let me know if any changes or improvements are needed. I’ll be happy to update it accordingly.
Thank you!
Hi! I made a few small improvements to my PoC after submitting my GSoC proposal — mainly README updates, clarification comments, and documentation fixes to better align with the proposal.
Just wanted to confirm — is it okay to continue making such non-functional improvements after submission?
@oblique bay I had a small query regarding the proposal format.
Will the original detailed description (Google Docs file) also be reviewed, or is the evaluation based only on the generated PDF from the form?
I’m asking because the formatting and structure in my original document is clearer (especially diagrams and layout), while the generated PDF appears slightly less structured.
Just wanted to understand this so I can ensure everything is presented in the best possible way.
Thank you!
We’ll review whatever is uploaded on the GSoC portal
-
Make sure your proposal is well formatted and readable. Go through the PDF yourself and ensure everything is legible. Do not write entire sections in italics with the same font weight.
-
docx provided by gform can be modified according to your liking as well or you can also create the proposal without gform as well. just include the info asked
-
Use tables, charts, flows, and visual aids to create clear visual distinction.
-
If you are using GForm, follow the format mentioned in the pinned comment for the detailed description.
-
If you are not using GForm, follow the pinned format for the entire proposal.
-
POCs should be easy to find. Include them on the first page of the detailed description, before the abstract, as per the format.
-
Do not put contributions in between proposals. If not using GForm, add contributions after the sections mentioned in the format.
-
If your POCs are private, make sure to share access.
-
Include a sequence diagram along with the architecture diagram.
-
If diagrams are too dense or unclear, with them provide an external link (image or gist) aswell, instead of tilting or compressing them which I myself recommended for earlier. So, you can include as mch details as you want. You may also host the entire detailed description as a gist if preferred.
-
Do not include code snippets in the proposal. Your POCs should demonstrate the codebase changes. You can go beyond just what I've asked for, in your POCs.
-
Preferably we'll avoid fallbacks such as ydotool or NutJS.
-
If you are not using Koffi, clearly explain why your approach is better. Avoiding manual maintenance of Node.js bindings is a strong advantage.
Also, don’t forget to upload the proposal first. You can keep improving the PoC for a bit longer after the deadline.
I have followed the format that you mentioned in earlier message of "Detailed format description" my whole proposal follows the orders of heading as is
We can upload only proposal pdf? If not uploaded gform generated pdf then ok?
@pure fossil You are now level 14! :confetti_ball:
anything
Ok Thank you for clarifying
Hey @oblique bay I used google form. It was automatically changing the font to italics.
thats fine and legible, i was talking about the detailed description
And also about formatting, in google form the generated proposal was having uneccesary empty spaces
Oh ok
gform provides pdf or docx?
Both
then modify the docx as per your self. however you like its up to you
I tried to do that but when I try to edit, the font went bad
Changing something in docx was like completely destroying the whole format
ok you can let it be or re format the entire thing if you want
@oblique bay If I use Gform then I don't have to include past contributions in detailed description file right?
yes, im simpler words dont put anything else between the format I mentioned
If you are not using gfrom, still you need to put it after the rest I mentioned in the format
(this is not a must, but makes it easier to navigate)
Ok
@oblique bay I have uploaded updated proposal on GSoC portal, please review my proposal when you are free and suggest changes and improvements if needed, I will be happy to make changes accordingly.
Thank You.
Hey @oblique bay till what date I can improve my PoC. Because I think I can do better in it
take the 1st week
@oblique bay Hi, I have a short query
In my submitted proposal, I included a compare link to my 2 separate PoC branches for the WebRTC and Virtual Input implementations. After submission, I realised that it was more convenient to review if I raised PR. So now I have to raise PR or it is ok?
@oblique bay how many people on an avg will be selected for the project ? Can you pls say about it since as far as I know for a large project 2 - 3 people are selected.
I am working on improving my PoCs and would like to take the first week to refine and stabilize them before moving ahead. I’ll keep you updated on my progress.
+1
@oblique bay I have updated my both PoCs please review whenever you are free
@oblique bay I have also updated my both PoCs so please look into them too
Same
@robust coral You are now level 15! :confetti_ball:
@oblique bay I have also updated my PoC please review
Shashank: Screen-sharing permission is often not requested, and even after granting it, the mirror remains blank.
Upendra, Anshupriya: Screen mirror stays blank.
Abhik, Pinjinx: Mirroring works but remove ws as a dependency. signaling and rest will be handled by the HTTP server.
Abhik: Native module build is failing because node-gyp build is failing. I'll try this again from my side.
So far, none of the input solutions seem any better than ydotool, since the visible cursor still does not move.
I tested the POCs on an Ubuntu VM with Chromium on Gnome, like in some of the demo videos.
If you can get them working in the similar environment, please share a video here and make any necessary adjustments by tomorrow.
also a single repo is easier to test at once then multiple
@robust coral @elfin abyss @cinder zephyr @bitter urchin @azure isle @radiant cave
@robust coral I managed to get both mirroring and input working after a few attempts, but the screen share permission is inconsistent. The cursor moves visibly in the mirror, but not on the actual screen.
@cinder zephyr The cursor moves visibly in the mirror, but not on the actual screen.
@oblique bay Could you comment on my input POC it was a bit out of the box. I wanted to know if it fits the project. because it worked well with linux distros I tested and stayed under 200 lines of code in implementing.
@oblique bay Please review my PoCs when you are free and provide suggestions and improvements. I will be happy to make changes as soon as possible.
I have added testing videos on
Windows
Ubuntu dual boot OS
MacOS Monterey VirtualBox
In my poc the ws is not doing anything I have changed all the functions that uses ws. It is just there in the package.json file should I remove it?
the cursor moves on mirror but not inside vm
I think both the koffi and your implementations are fine. the problem seems to be from my end
ok no need
@oblique bay till when can we keep updating our solutions ?
@slim pollen You are now level 4! :confetti_ball:
The time was already up, I’ll just finalize everything tomorrow
I have my solution ready...should I create a new repo for posting it ? like for #130 and #208 2 seperate repos
nah i'll just recheck whatever is there in the proposal
alr
the input injection is not working
@oblique bay how many people on an avg are getting selected ?
no clue, depends on the slots they get
ohky
Hi @oblique bay
I am @cinder zephyr and this is my friend's account,
I just want to infrom you that my discord account was hacked since yesterday and i have recovered it finally,
but it has been restricted to send any message for 24 or 48 hours, so i can not send message i have seen your messages,
i have sent you linkedIn request to send you message about POCs, please accept that also, i will inform you about my
POCs update according to your messages on linkedIn or on discord after fully recovered account.
By the way all things about POCs is working on my machine(windows) and no issue happened that you mentioned
I also will inform you about running linux and MacOS by tommorrow,
Thanks for your consideration
dw about that
the cursor movement issue seems to be from my side
So there’s no need to worry or rush about anything. I’ll finalize everything early tomorrow, and you won’t have much time. So just relax with what you’ve submitted.
I rechecked it on my friend's pc(ubuntu). The native module build was fine and the poc is also working smoothly.
Ok I try to fix it today itself sorry for inconvenience
Hi @oblique bay ,
I'm currently debugging the screen mirroring PoC on Ubuntu VM + Chromium (GNOME) as requested.
Screen capture permission is working, but I'm facing a network connectivity / ICE timeout issue between the host VM and viewer device. I'm actively fixing it and will share the repo update shortly.
I'll push the final adjustments once everything works reliably in the VM environment.
Hey @oblique bay , I think it is looking inconsistent because I made it like the screen sharing permission will occur only after the user has connected the mobile device that's why.
The Screen mirror may have remained blank in the first time because of the webRTC connection problems. For that only I made the system in a way like the screen share permission will occur only after the user has connected the mobile device
So should I remove this feautre?
@oblique bay I added that feature so that user get time in the starting to scan the QR. Making the experience better.
Yes, in webrtc poc, input injection is not working, really sorry for that and PoC 130 is working fine from my side.
I am trying to solve webrtc input injection
Hi @oblique bay,
The blank screen issue has been fixed. I’ve updated the repository accordingly, and the PoC is now working properly. Please review it once.
hey @oblique bay , can i pls work on issue #339?
hey @oblique bay , Bun can work as a complete node replacement for both mac ox and Linux I have also checked out the official document released by Bun and it has been confirmed from that. However it only works with node on windows but not without node. there is a reported bug on their official repo, this bug makes Nitro unusable on windows
can you check if theres a reason to keep nitro? given we dont plan to host the app anywhere
also add your findings to issue itself
thanks
Nitro is there for TanStack Start
If we wanted to replace Nitro ,would have to replace TanStack Start too which means rewriting the entire app. Once the bun's official repo fixes the nitro issue stated on their github , it will work on Windows too without any changes in this project. what do you think we should do?
nitro is not a hard dependency for tanstack start
working on this
hey @oblique bay , can you check this out and let me know what improvemets are needed whever you get time https://github.com/AOSSIE-Org/Rein/pull/341 . Kind regards
@elfin abyss Congrats you are in
We had fewer slots than ideas, which might be disappointing for some, and I’m sorry for that. I’m not sure whether I’m supposed to say this, but some highly ranked candidates were affected because of their associated institutes and other factors I’m not aware of, and that was not under our control.
Remember, you all are everything even without GSoC. And I can still vouch for all you've done.
Congratulations @elfin abyss 🎊
Congrats @elfin abyss
Thanks everyone ❤️ I’m truly grateful for this opportunity and excited to be part of the program. I’ll give it my absolute best. (Still cant believe im in)
Congratulations @elfin abyss 🎊🎉
Congratulations @elfin abyss 👏🎆
Thanks @void schooner @pure fossil 🙃
Congratulations @elfin abyss 🎉🎉
Congratulations @elfin abyss
@elfin abyss
coming to the actual work
I think Koffi is the right way to move forward without introducing a new language into the codebase until we hit limitations.
I’m also thinking of a new cursor behavior where the cursor follows the pointer position on the screen mirror. This would make it feel more like a touchscreen and also improve support for non-touch controls, including desktop-as-client use cases.
(Optional, if not common in desktop via accessibility settings) I’d also like an auto-zoom mode that follows the cursor for better accuracy on small screens.
Also, I believe we should aim to complete the planned deliverables (ASAP) by July at the latest. Ofc you can take help from AI, but ensure the quality responsibly. My availability may vary from time to time, so please don’t wait for merges and keep working on the other components. After that we can work on other obvious features or whatever you planned for future, so nothing is intentionally left for next year.
You are free to involve other contributors with you as well and supervise.
Hoping for the best...
Sounds good I am really excited to get started 👍
I agree with Koffi, we can rethink later if needed.
July deadline works. I’ll keep things moving without waiting on merges and maintain quality.
Will coordinate if needed and share my updates here.
@oblique bay I have to mention I do have sem exams till June 8th After which I have holidays I was thinking about starting a bit early and putting in extra hours during June to make it up
@elfin abyss Congratulations 🎉
now tell us your real name
@elfin abyss 😅
Well it's Rohith K Pradeep🙃. If you got a linkdin happy to connect
Hey @oblique bay, Can you check this PR and let me know what are all the improvements required https://github.com/AOSSIE-Org/Rein/pull/344 .
Summary
Implements cursor-follow interaction for Screen Mirror with accurate absolute pointer mapping and improved input handling.
Changes
Added absolute pointer mapping for cursor-follow behavior...
@mystic token You are now level 2! :confetti_ball:
I’ve pinned some issues in order of priority. I think we can tackle this once those are done.
@elfin abyss
#1467370739152846899 message
I’d think you should review coming PRs and opened issues before I do.
While reviewing, decide if it’s the best way to implement, whether it depends on something else (and should wait for it), and once done, test all functionality. Provide feedback accordingly.
We should maintain an organized list of PRs ready to merge in the proper order. I’ll review them in batches when I have time. Also, keep a list of issues that can be closed with reasons.
Link: https://docs.google.com/document/d/1upEKv4f6p2lGUcgaka0GaGFVmTHrcm4vCKVBPxlJrZE/edit
Sure I’ll start reviewing all active PRs and open issues, so it won't become a mess later on and update the sheet alongside
I have started working on Vite 8 and other dependency updates - Issue. Soon i will raise a PR.
Hi @oblique bay, Can you please have a look on this PR and let me know the improvements needed.
https://github.com/AOSSIE-Org/Rein/pull/347
I can see some out of scope changes. And I've updated the issue with more clarity
Also, you should directly consult @elfin abyss ragrding your work, as he'll have to ensure everything aligns with his plans
Unless there's some disagreement, I plan to review PRs marked as ready to merge by him in the doc
I have to add to that I have a lot academic work (so might take a few hours to reply back) for this week still I will be going through all of open the PRs whenever I get free time so you all can keep working on the project just ping me before starting work on an issue so that it doesn't interfere with anything or if you have already made a PR do ping. I'll make sure to go through them.
@elfin abyss You are now level 12! :confetti_ball:
@elfin abyss I have raised this PR. Please review it when you get a free time and let me know the improvements which are needed.
@elfin abyss, Can you please provide some context on this?
That’s likely due to outdated dependencies. You should still be able to run the app without issues it shouldn’t interfere.
Also, I went through your PR for updating dependencies, but it included some unrelated changes. It would be better to close it and open a new PR that only includes the necessary updates for deprecated libraries. I’ll review that and, if everything looks good, proceed with merging it.
Yeh I'll close the PR and open the new one. I have updated all the dependencies locally and the APP is running without any issues. I'm facing one more issue in the terminal during the run time of the APP.
@mystic token You are now level 3! :confetti_ball:
You’ve hit a hydration error which happens when the HTML rendered doesn’t match the client. What I usually do is do a git diff and go through your recent changes. anyway its upto you.
@elfin abyss This is the progress. Now I'm facing this issue.
Can I remove the dead code?
Just make sure while making PRs you have tested out everything and it doesn't break anything anywhere
Sure. What about dead code errors?
@elfin abyss, Can you please review my PR and let me know the improvement required.
https://github.com/AOSSIE-Org/Rein/pull/349
Addressed Issues:
Fixes #347
Description
This PR focuses strictly on upgrading and stabilizing the Vite 8 migration while ensuring existing functionality remains unaffected.
Changes made:
Updated...
Sure looks good overall I'll review it soon thanks for the contribution 🙂
@mystic token Do check your PR I have mentioned few changes
To add up
The React compiler is not enabled.
and description doesn't match and mentions dated babel-plugin-react-compiler
once compiler is enabled, all useMemo and useCallback hooks should be removed and functionality should be verified
check https://vite.dev/blog/announcing-vite8#vitejs-plugin-react-v6
and rest of the blog as well
@elfin abyss
Yeah I'll go through and review it soon
Now you please go through it.
Yeah I have enabled React compiler as well you please go through the PR and let me know for further improvements.
@mystic token You are now level 4! :confetti_ball:
I have went through the PR look good for me. Overall all dependency have been updated to latest major versions and the compiler has been enabled and callbacks rewritten
@elfin abyss You are now level 13! :confetti_ball:
As some of you maybe aware tan stack being compromised Link please be careful and aware while pushing PRs or while updating dependencies and make sure to verify package versions.
@oblique bay I have finished working on webRTC data channels and screen sharing I have some cleaning up to do after which ill push a PR probably by today or tomorrow. Also regarding this I have a few doubts/opinion I wanted to share with you if its possible could we have a meeting?
ok will do it
@elfin abyss when are you free?
Could we do it tonight whenever your free
I'm free right now, if you're not, we can start whenever you get free.
I'm nt audible right?
no
yes yo uare
its fine ill present then
lemme rejoin
k
@elfin abyss do update the docs with decisions you take with reason
Also anyone who is interested in working on something directly mention it here
My github notification is filled with a lot of things, making it difficult to attend anyone there.
@delicate berry
I'm Interested.
I'm also interested
I just mean if you all want to work on or are working on something, discuss it among yourselves and @elfin abyss here instead of GitHub.
Also, Rohith, if you need some specific help from others let them know here. For eg. the hotspot thing.
I dont have any specific task for anyone
@pure fossil @mystic token @delicate berry The workflow file appears to be outdated. It is still using an older Node version. Electron and Vite now require Node.js v22 for the current setup, so the workflows likely need to be updated. If anyone is interested and free, they can take it up and create a PR and Issue.
I will take a look at the issue and try to raise the PR by the EOD or within tomorrow.
Note that we are not moving to WebRTC because we need P2P, but for better latency over WebSocket, for more of a real‑time/UDP‑like response.
Most cloud gaming services use their own custom UDP protocols, so if there is any open or simpler option available for us, feel free to propose them. In case they would make things simpler/better.
Note that LAN takes priority over out-of-LAN access
@elfin abyss @pure fossil @mystic token @delicate berry and the rest as well
Also note that to take full advantage of WebRTC over HTTP/WebSocket, we'll need it on the backend as well for input streaming.
Another simpler option might be WebTransport's unreliable datagrams solely for input injection.
@oblique bay I have opened an Issue to fix the Node version in workflow can you please assign it to me?
https://github.com/AOSSIE-Org/Rein/issues/352
Description The GitHub Actions workflow configuration appears to be using an older Node.js version that may no longer align with the project's current requirements. Since the project now depend...
@mystic token You are now level 5! :confetti_ball:
no need for assignment
@oblique bay @elfin abyss Can you please have a look on this PR and lemme know if there are any implementations required.
https://github.com/AOSSIE-Org/Rein/pull/353
check why the vite build is failing
@oblique bay You are now level 35! :confetti_ball:
Nope vite isnt failing.
it's failing on CI i mean
"npm ci can only install packages when your package.json and package-lock.json or npm-shrinkwrap.json are in sync. Please update your lock file with npm install before continuing." - Its because package.json and package-lock.json are out of sync
raise a pr for it if possible
I have raised a PR. https://github.com/AOSSIE-Org/Rein/pull/355
I had merged the PR you made, but it doesn't appear to have fully resolved the issues. If possible, could you do a clean installation of all dependencies and raise a new PR?. You can check what is failing PR by reviewing the workflow jobs here:Link
It looks like the branch may be out of date with the latest changes on main. Since the lockfile synchronization fix was merged recently, could you please update your branch with the latest commits from main and rerun the CI checks?
@oblique bay I started with HTTP, but due to exams I didn’t get enough time to make significant progress. I also realized it needs more research. Since I want to learn Gstreamer as well, I’ve thought to push it to next week as my exams end this week.
The Koffi PR was almost complete and only the macOS part is left. I’m planning to borrow a friend’s Mac, test it, and push the PR this weekend.
You should also have the x86 version of MacOS installed as a virtual machine just in case the machine is not available
And yes i think we should go for gstreamer after koffi
in the Koffi PR, avoid other/unrelated changes
switching from WebSocket to GStreamer will be done at once for both A/V and input injection
I think we should treat each connection as a separate input device, so that we dont have to handle multi-input separately.
Please @oblique bay @elfin abyss review my PR when you are free.
Addressed Issues:
Fixes #284
Description
Screenshots/Recordings:
Functional Verification
Use "npm run electron-dev" for development
Use "npm run build" and &am...
@elfin abyss share both the social post in weekly update channel or somewhere, and tag bruno/admin if others did so that they can repost
@oblique bay I'm happy to report koffi works everywhere I tested it on all the platforms there is a bit more cleaning up to do. Mainly removing unused old code. I'll get a PR out by Tuesday night.
Nice