Hey. Im thinking about creating matrix client based on telegram. I think their UI/UX is one of the best among messengers. Plus their android app is open source.
I have to take couple of things in consideration, like: licensing, technical details and some more. Its very possible from the technical perspective and pretty unclear from the licensing/legal perspective (since they could change license any moment or even go private).
In this post i would mostly like to know if open source community even wants anything related to telegram. Not everyone likes it. The project would target mainly open source community and will be public as well.
So the questions: Do you think we would want it? And would you personally try it?
Thank you <3
EDIT: :V If you think its bad idea, please tell me. I see much of positive response (and i feel pretty happy about it), but i have to be real, so if you think im wrong or you see any problems, please speak up :)
Telegrams UX is excellent.
Bringing some of that to matrix would be neat.
Definitely
:3
it would probably be easier to take an existing matrix client implementation and make it look like telegram.
i don’t think the ui will be much of a problem, the backend stuff: communication with the server, enryption, device validation will probably be the most complicated to implement.
But if you do it, I’d probably try it. The existing matrix clients are all not very good.
Yeah, the thing is, even tho telegram is horrible to work with, it has all needed widgets and ui elements. It would be very convenient to have telegram as starting point. (Not the most horrible codebase i worked with btw hehe)
From what I understand, Matrix is a pretty hard service to write clients for because they’re constantly updating the server architecture and so clients end up very much behind very fast.
considering telegram is basically the only modern and stable non-electron chat that runs buttery smooth, such a matrix client would be a game changer.
i’m stuck with telegram for now because i have an m1 mbp and an ideapad tablet with linux, neither which can run more than 1-2 electron apps simultaneously without screeching to a crawl and putting itself on fire.
like, 99% of my life revolves around hating on electron; it feels like electron has become standard for modern much needed productivity software. and it makes using computers that aren’t super high end basically useless. developers need to start reconsidering their end-users, because electron will never be optimized for non-highend hardware (i’m not even convinced electron run snappy on a highend computer, but i would love to be proven wrong because currently i just assume coders have gotten too lazy to care about the end-user).
Yeah, we try to “taurify” the world 😆
i haven’t used gomuks, i’m not a big fan of terminal applications
nheko has so many issues. one version it was impossible to login. then when they fixed that, the encryption kept breaking and a lot of messages from my friends were lost. it also lacks a lot of features. last time i tried nheko was in january. so maybe they have fixed the stability issues and added some modern features; but otherwise it can’t compete because it can’t be used as a daily driver.
I’ve been using nheko as my Matrix client and main Telegram client via mautrix-telegram for the past 4 years. I fixed a couple of annoying bugs myself, and it has been working great for me since. Shame your experience isn’t the same, maybe you can report the bugs to the devs (who are amazingly responsive).
Ofc you are welcome to share any of your thoughts about it. I would be really glad if you would do so
Sorry, I’ve never used telegram and don’t typically use IRC type stuff because info is lost in the void, and I need time to collect my thoughts. I tried a matrix client around a year ago, but do not recall which one. I don’t use it often enough to have an app. I don’t like anything that keeps a persistent icon on my phone. I don’t even like the Andy WiFi and mobile data visible. IIRC the Matrix client I tried put an icon on my status bar, reduced my battery life substantially, and defaulted to notifications for nonsense.
I might use a matrix client if it were very light weight and didn’t intrude in any way. I would use it like SMS but for the internet. I’m weird though, like completely degoogled, use a whitelist firewall, and only accept messages and notifications of any kind if they are human messages by people in my contacts. I have a dozen FOSS apps total on mobile and do most things in browser… So probably not your best measure of user base, but no comments yet and I think what you’re doing is great.
Neither the Matrix clients nor servers are lightweight—largely by design. All of the clients take literal minutes to start up if you don’t use it regularly & chew thru data. Even if they managed to hide how slow syncing is, it is just being hidden & taking just as many resources in the background. The whole decentralized eventual consistency + chat history is permanent is a model that makes it this way. At least IRC understood the chat was ephemeral & the protocol—even v3—isn’t bloated.
Please! I think that all mobile matrix clients are either bad or just somewhat good with many caveats. For me, Element X develops at a glacial pace and it seems the philosophy is to just make a minimal client and nothing more. FluffyChat is too opinionated and rough everywhere for me. Beeper is by far the best client however you’re limited to a beeper account.
We really need more clients that aren’t just Element or trying to reinvent the wheel.
I understand you. I use element X and SchildiChat mostly, but i wanted to see something else, something we get used to.
Neat idea, I personally never used telegram but heard a lot of positive feedback about their ux. I only have a feeling that, at the protocol level, the two ecosystems might be too different to make it viable.
That’s just a feeling tho, so go for it.
I like telegram, I like matrix, I’ll give it a shot
Yep, that has actually been a dream of mine for a while, I went as far as downloading the source code and setting up the build environment, but then got a bit overwhelmed by the task. It would be amazing to have this.
I use element to try out matrix, I haven’t used telegram. But I’d try it a new client.
The problem with telegram is the recent revelations that there is little to no backend privacy. So I’m not totally sure who your userbase might be?
Open source afficionados who also don’t mind random cops trolling through their contacts and messages? Seems like a rather small overlap.
Or maybe I’ve misunderstood the recent media attention that telegram has gotten or your proposal.
I am not sure how feasible it is to do, but the coolest would be if both Telegram and Matrix could be used from the same app with the original Telegram UX/UI.
In the sense that after installing the app you can choose whether to log into a matrix server or a Telegram account. And then you could also add more accounts of either service later, as you already can add multiple Telegram accounts in the Telegram app now.
Probably quite a complex task to make an app like that but it would be cool.
Lets see if there is more ppl interested in that feature. Personally i dont like telegram as a product, but like their mobile UI/UX
Edit: It is feasible, but would require more work and would make codebase even more unintuitive
I think it could be another gateway for more people to start using matrix, because they don’t need to bother with two apps and wouldn’t need to learn a new UI for matrix.
But maybe also not since it would require another app install anyway.
I guess its not really worth it if maintainance becomes too complicated.
It could. Im not planning on syncing with upstream. So if there would be any important telegram fix of official client, i would need to implement it manually or copy-paste code.
Also, i want full rebrand and change styling of many UI elements. My final goal is to make it original product
@majestic
Maybe a Rust App with multi account with Telegramm UX is fine. Rust is a secure language and multi platform.Can’t be enough Open Source Clients.
