Hacker News new | past | comments | ask | show | jobs | submit | pathless's comments login

That was an amazing read as well. I really wish the internet had more game-programming-for-beginners or whatever you might call them style microblogs.

Shamus will be missed! RIP.


I didn't know he had passed away. While not what's being discussed now, DM of the Rings remains one of the funniest comics I've ever read.


Telegram is genuinely the best general communication platform I have ever used, by far. I really hope he has a good lawyer and this doesn't end up getting essentially murdered for creating it. When you create something that is objectively great, everyone will use it - including bad actors.


I use telegram for some group chats, but I'm not sure why tech-savvy people would like it so much – messages are not end-to-end encrypted which makes it an inferior choice compared to even whatsapp


For one, it's the only major messenger that has an actually lightweight, well-written and full-featured desktop client rather than yet another boxed-up web browser. I might be more enthusiastic about using the alternatives if I could use the Telegram client.


It's very bizarre to see all these comments downplaying this, or implying the lack of E2EE by default somehow makes it less attractive to the average user than something like Signal.

Most people care about usability and interconnectivity first and foremost because the majority of their messaging activities are not so sensitive that they feel the need to sacrifice those things for mandatory E2EE. Call that shortsighted if you like, but it's far more common than this "encryption or bust" mindset around here.

If signal or some messaging platform could find a way to be E2EE capable all the time, with all the same usability and design as telegram, without unnecessary restrictions on users, and without it being a completely walled off garden from which your data can never be self-extracted, it would win this argument.

Same goes for things like Tutanota and a lot of these other data prisons that are cropping up which create privacy through taking away user agency.

Until then users will pick what they want for their own needs. Telegram met those needs for many.


> the same usability and design as telegram

From the recent Tucker Carlson interview of Pavel Durov, Telegram has:

  - 1 PM (Durov)
  - 1 owner (Durov)
  - 30 developers
  - 0 HR, they hire contest.com winners
https://news.ycombinator.com/item?id=41343845


a very impressive guy. too bad he didn't name the oss libraries.


One could list OSS libraries (if any) that Telegram uses today, then diff against OSS libraries used by other E2EE messengers.


isn't only the client side oss? server side logs/libs is more likely. isn't it amazing 30 guys handle a billion users and who knows what sort of ddos is unleashed against them.


Line has a fully featured desktop client that's even more lightweight than Telegram. They also have a white paper that explains their encryption which is decent enough https://d.line-scdn.net/stf/linecorp/en/csr/line-encryption-...

You do need to turn off "display stickers suggestion" before the app becomes nice.


Intriguing (and surprising to me that they offer E2EE at all), but there is seemingly no Linux build. I can't seem to find source code either (Telegram Desktop's is released under the GPL).


Does Line offer server based/cross-platform messaging with a similar upload allowance? I’m curious to find telegram alternatives.


Some of these are native apps; not Electron, Java, or what have you:

https://matrix.org/ecosystem/clients/

(Of course, I don't know what qualifies as a major messenger to you.)


They used to have a page wittily named "feature matrix", which made it apparent that only Element was really kept up to date, with other clients missing features ranging from channel search to embedding images. I don't know if this situation has improved and whether the original page still exists somewhere.


Several of them have been reporting improvements over the past couple years in the weekly Matrix development blog, and I know at least one of them has both search and embedded images. You might want to have another look some time.


> which makes it an inferior choice compared to even whatsapp

I'd rather have a good privacy policy with a good enough server-side encryption than some closed-source implementation of E2EE, that we can never audit.

WhatsApp actually disallows you from reverse-engineering the app and looking into the algorithm. That begs the question, what percentage of E2EE is it really? 20%? 50%? 100%? Because there's still no way to confirm their claims of E2EE. All we have is a company with a really good track record in lying publicly, telling you that it's safe.

Looking at WhatsApp's privacy policy, I really wonder why people even support it compared to Telegram: https://privacyspy.org/product/whatsapp/

https://privacyspy.org/product/telegram/


This is no longer true, whatsapp have taken steps [0] to make their e2ee auditable and honestly I disagree with the idea that no e2ee is better than closed source e2ee. I'm not sure why you would trust a privacy policy more than you would trust encryption, with a court order Telegram would provide your chats to law enforcement, while Whatsapp would not be able to.

[0]: https://engineering.fb.com/2023/04/13/security/whatsapp-key-...


> to make their e2ee auditable

This is not the algorithm being audited, it's the key. Telegram's complete algorithm is auditable, including the open source client apps. Server code is always unverifiable, so let's not bring that up.

Secondly, WhatsApp channels and large groups (copied from Telegram) are not encrypted in any way (cmiw), as opposed to Telegram's MTProto 2.0 Cloud encryption. The app is completely closed-source even with all their claims of privacy and its TnC even discourages you from reverse-engineering it.


WhatsApp Communities are indeed E2E encrypted. About channels, why would you want a channel to be encrypted when you are just a follower and cannot communicate back? In fact WhatsApp's guidelines explicitly state the following:

> Channel updates should be used to share information with followers and viewers, not as a way for admins to communicate back and forth.

https://faq.whatsapp.com/671443411431514/


I prefer Signal and it provides encryption.


Signal has the better protocol and the better organisation behind it, but inferior apps and UX. It's unfortunate, really.


Apps? I just use it for messaging.

I haven’t noticed any major UX issues but I use it for one thing only


Signal for mobile and Signal for desktop are different apps with different code bases. Neither is as good as Telegram's, in my opinion.

Signal is fine for messaging. Not bad, not amazing. I'd have a much easier time convincing people to switch to Signal if it would've had a client as good as Telegram's, especially for the desktop application.

That said, Telegram has been adding more and more annoying premium features that distract and annoy.


Thanks I only use mobile so now I understand


Yes, the messaging apps. They suck big-time. By far the worst communication app I’ve ever used. So many bugs…

Telegram is indeed the best like others are saying, and WhatsApp is a distant second.


Can you give an example?

Never had any problem with Signal.


On iOS, if you turn off your Internet connection and receive a message, you won’t get a notification when you restore your connection. This problem doesn’t exist with WhatsApp, Messenger, Instagram. Quite strange.


On my Signal app screen, I have Chats, Calls and Stories. I could probably do without Stories. Periodically it prompts me for my PIN.

Other than that it just works


You can actually turn off Stories in the settings, then the option also disappears from the bottom bar.


Why are you assuming tech savy people care primarily about end-to-end encryption?


Maybe because tech-savvy people understand the need and importance of encryption? There's of course always the exceptions that say they don't care about privacy, but that is fortunately usually a small group, at least in the tech-savvy world.


Both tech savvies and laypeople expect private/encrypted messaging app to provide the basic property that only the sender and the intended recipients can read it. This is achieved with end-to-end encryption. Techies know the term, and can understand it's not present. Non-tech people don't understand, and just rely on word-of-mouth that it's super secure, when it's not.


Same reson people choose macbooks for work over running Linux, it's just a nicer product


Where's windows than?


Not used in serious contexts


At least on the beginning, when I looked into it, it had a very simple and well documented API. I guess it was the only messenger you could send a message with one line of code (of course not e2e encrypted). So it's very simple to send you a message from your home project.


I think it is because Telegram is a communications platform, not just another chat app. For example, it has good APIs to build apps on top of it.


WhatsApp doesn't save my history. And secret services of governments of certain counties are not a realistic adversary that I'm trying to defend myself against. The usual scammers which are going to steal my identity are not the people Durov will sell admin access to his server to.


because telegram just works really well tbh.

it is lightweight, pretty much never goes down, has reasonable features.


I prefer it because of the bot API.


Hot take: E2EE is overrated!

Almost all conversations that most people have are benign. I used telegram to follow journalists (essentially as a twitter replacement), how would E2EE benefit my use case?


Yeah no shit. But sharing cute cat pics deserve human right to privacy. Also, when everything is E2EE, that reduces the metadata about when you say something private. You don't want your opt-in encryption to reveal metadata about how close you are to someone.


i gotta agree, i basically live out of telegram.

even with the recent trend towards adding incremental bloat to the client, it’s managed to stay a simple, straightforward tool for communicating with minimal advertising and enough of the features that i need front and center.


The fact bad actors are also using it is not the problem. His unwillingness to moderate content and cooperate with authorities is. Great UX doesn’t suddenly put you above the law.


Video and Audio calls are hit and miss. The history and search are not reliable. The interface is not really suited for big group chats... I could go on and on.


How big is a "big group chat" in your definition?

It's perfectly fine for a few thousand people


i find history and search very reliable in my experience but agree about the calls compared to e.g. messenger


And tactics exist outside of control of communications, to capture these bad actors, to infiltrate their ranks; why are these alternatives to fighting the production of child exploitation and abuse content not brought up in conversation ever?


The masses do not care much if ones do not do agains bad actors what are in their power just their pretty platform shall keep running, they will keep this one alive too and argue for it to the death, don't worry, the masses could argue for any malicious thing that they find pretty or nice or like for some reason. Can organize some protest or even riot too in a - unencrypted by nature - group channel, there will be scores to participate, as recent example show in other precious matters, maybe can loot some good scores too on the side of the big party about a dear matter for the heart! Paris deserve the revenge! : /


Good lawyers won’t make much difference for him as the French government is tired of not being able to look at all our conversations. They want to start scaring people into compliance and verifying all their actions with the government or at least scaring the companies providing a (semi) private experience. This is mostly like just phase 1 of getting the keys that open up telegram servers to 5 eyes by getting Durov under their thumbscrews.


I actually despise it. I'm not sure if this has changed, but after being forced to make an account under my phone number, it proceeded to send a message that I had joined to everyone who had my number in their contacts and was foolish enough to share them with Telegram. This included a rather vile woman whose number I apparently inherited from a deceased relative some years before. She didn't understand this and accused me of stealing his identity. While it was simple enough for me to brush it off, I couldn't believe they would allow and even encourage such a thing, so I almost immediately deleted my account and instead tried out one that wasn't so eager to lap up my personal details.


Signal is much better, even if the UX is not quite as good as telegram's.


It's also the least private one (compared to whatsapp and signal.)


They were great until they put fast downloads behind a paywall. It's like they're going out of their way to throttle download speeds nowadays


You like it better than Signal? The only thing I know Telegram for is several of my girlfriend’s relatives being exposed to crazy scams and right-wing conspiracy theories and misinformation on it.


Signal has better governance, plus e2ee mandatory, while on Telegram is optional and rarely used. Telegram also has a “social media” aspect with huge groups and channels, which attracts many people, but is a depart from the whole secure chat messaging it’s still known for.

IMHO, Signal is way better.


frankly signal is a lot buggier than telegram and people like having their chat history.


This is exactly my experience as well. I have never actually used telegram as I was early a signal user and never needed it but my ex used it. All she ever used it for was conspiracy garbage she would follow. Anti covid vaccine doctors and groups mainly. The amount of misinformation she tried to show me and every time I would show her how it was fake she still would not believe me. Then she was even scammed out of $10k from telegram when she fell for a romance crypto scam. The conspiracy stuff is a main reason we broke up it was every single one from flat earth to fake moon landing to all the covid world economic forum world take over and on and on. Most of these came from telegram.


The reason you consider all that misinformation is because it's politically sensitive information and every single other social media company and the vast majority of western-aligned media censor it, so the only place you come across such information is in the uncensored Telegram platform, and assume it must be false because all the other media you consume tells you so.


No I do not assume the moon landing was fake because other media outlets won't cover it. I do not think the earth is flat because news outlets won't cover it. The list goes on. I also do not think the covid beliefs I have come from the fact that mainstream news outlets won't cover it. I strongly believe a narrative was painted and information censored and controlled when it did not fit the narrative they wanted. I was first to criticize the Canadian government on how they handled the vaccine rollout and mandates. I did however spend a lot of time doing my own research and listening to any information my ex showed me and also researching those sides to things and formed a well informed opinion. I sought out other opinions and studies. I did my best to follow the science. Not at all what you are implying. In the end my ex was only interested in the opposite which was any narrative that said the vaccine was poisonous, we are being forced to take it because this is a world culling and in the next few years population is about to crash as it has now effected our sperm and dna so we will not be able to have children and the crazy thought kept on coming almost all fueled by telegram chat groups pushing that stuff.


Did she also believe the "fine people both sides" thing was a hoax?


I don't understand why anyone uses Telegram if it there is no proof that it is secure. Their code isn't open source?


There is no random hassle that get your account/group/channel deleted(unless you are doing real CP, which is a giant red line of telegram) or random limit for size of the group. And easy to use. That's it.

Also the apis are almost completely free, so you can do lots of creative projects for fun.


Most people dont give a shit about security. Telegram is easier to use than signal and has more features because auto encryption makes stuff like public chats difficult.


Exploit their smartphone then email tidbits of their more outstanding Facebook chats, SMS messages, maybe some nudes, and recordings of their phone calls. You will see the "I don't care about security attitude" do a hard 180 real quick.


Sure, but for now they dont care.


Unity has become an unfocused, poorly optimized, half-baked mess. This WAS bad for the independent and small developer market that actually uses their engine for final builds, but they've begun a mass exodus to Godot, which happened to go from "ok" to "great" JUST as Unity ruined their platform with their short lived "20 cents per install" policy.

Someday soon, we will see Godot eclipse Unity in the same fashion that so many other proprietary juggernauts were slowly cannibalized by laser-focused open source projects over the years:

In 2022, the split among GMTK participants was 16% Godot to 61% Unity. In 2023, it was 22% Godot to 49% Unity. This year, it was a whopping 37% Godot to 43% Unity: https://pbs.twimg.com/media/GVfo5-0WQAAIMAQ?format=jpg

This is major, because Godot has just had another round of home run improvements that brought in even more developers. I think 2025 is the year that Godot effectively replaces Unity for new developers.


> In 2022, the split among GMTK participants was [...] This year [...] 37% Godot to 43% Unity:

Jam and commercial games have different requirements.

> I think 2025 is the year that Godot effectively replaces Unity for new developers.

For game jams, possibly. Unity and Unreal have still a very large market share for commercial games.


> Jam and commercial games have different requirements.

Indeed, but it's a clear signal that something is shifting. In jam game environment, the shift is obviously faster, as the timeframes are different. But you extrapolate that things will change for mainstream games as well, it'll just take way longer time.

It would be interesting to see the number of jam games being done at Unity when it first appears, although I think the whole "game jam" thingy wasn't as big then as it is now. But maybe that could give some indication on how long it'll take before we see a difference in mainstream game engine marketshare.


The 'thing', be it a game engine or whatever else, of choice for small projects often end up taking over for commercial products too, devs bring their favourite tools into their work and slowly corporate adoption of it grows. It happened with Slack, it's very quickly happening with Blender, hell it even happened with Unity itself I'm not sure I'd be as optimistic as OP in terms of how quickly Godot will take over Unity, but I do believe it will happen, or at least something similar


This is the sort of thing people said about Blender for a long time, and my understanding is that it’s now often used in commercial contexts.

Not a given by any means, but it’s happened before and it will happen again.


Commercial developers and jam participants want the same thing. And even if they weren't much alike, this is still a huge indicator of trajectory.


Yeah but people use the tools they know. Commercial may follow


I think Unity just publicly figured out what has always been the case - there is no market for a game engine targeted to developers that want to make PC games. They all have no money, don't want to pay license fees, and are rarely successful enough to pay their success forward back to Unity.

However for realtime applications other than "standalone games", Unity has absolute (and growing) dominance outside of where Unreal is carving out a niche (Film/TV). Automotive, simulation, robotics, etc. are all leading Unity adoption.

If you're (royal "You") the type of dev that wants to make games for Steam, Godot is definitely leaning more in that direction these days than Unity, but I think Unity isn't seeing anything really "bad" coming out of that besides lack of good PR in places where game developers post screenshots/etc.


> lived "20 cents per install" policy

Funnily enough nobody who paid attention was going to pay the $0.2 since it was only there to funnel developers to the pro tier which would have given them a significant discount on the per install fee (IIRC there weren't even any current customers who would have paid that much because the personal tier had a $100k limit which was removed and the fee only applied after it).

Unity just had such a horrible PR release and did an inconceivably bad job at explaining the changes that they mad an already horrible pricing model seem 5x worse than it actually war. It was so incoherent that nobody read past the $0.2 per install...


It wasn't the quantity of money that was the issue for most people. It was the "I am changing the deal. Pray I do not change it any further"

They were asserting their ability to unilaterally bring in new terms, and specifically with the per-install issue bring in terms that require data that was poorly defined and could potentially be arbitrarily decided.

It wasn't a money issue. It was a trust issue.


The quantity of money legitimately is an issue too though. Especially for mobile games -- which rely heavily on advertising and scaling up thin margins -- 2.5% of revenue (plus $2,000/yr/seat) could be quite a significant percentage of profit.


To be fair that % of revenue Unity was historically getting from most of its customer that were actually making meaningful amounts of money (i.e. enough to have a permanent paid workforce) relative to the value it provides (especially compared to Apple's 30%) was very low. Of course not surprising given how relatively little pricing power it always had.

But I don't really buy the pricing argument. Most of that type of games are all made with Unity meaning that it wouldn't be too hard for all of them to raise prices by a few % with minimal impact. That mostly only applies to IAP not to ads but then again Unity was offering a discount/waiving the fee if you were using their platform (which is of course scummy and predatory but probably wouldn't have had a massive financial impact).

Overall yeah... they completely messed up with aligning their pricing model with their long term goals as a company. It was always highly suboptimal (the per seat license is almost insignificant for some companies while a significant barrier for smaller ones, contractors etc). They would probably still have been fine if they hadn't started almost literally burning money like complete madmen after the IPO for no reason. Instead they now have >$2 billion in debt (albeit low/zero interest) and nothing to show for it whatsoever (well besides IronSource to some extent I guess..) in addition to a still extremely bloated but also unproductive and heavily demoralized workforce.


Yeah certainly, changing the terms retroactively was just something else.

However I still think that the $0.20 headlines did a lot of reputational damage (even if most medium-major customers did the math) and that could easily have been avoided.

The way they introduced the already the hardly fathomable changes was IMHO as amateurish and half backed as it gets. Literally a corporate suicide attempt, can't imagine what were they thinking...


Any such policy can not work and would have been gamed heavily to damage competitors.


> Unity has become an unfocused, poorly optimized, half-baked mess.

Unity has always been a relatively unfocused, poorly optimized, half-baked mess.

It's just that there used to be no concurrence in this segment and their overall attitude towards makers made the mess somewhat endearing. Now that they act like a typical listed company, well, people feel less inclined to give them a pass.


I tried Godot, unity is light-years ahead of Godot for small team game dev


If you start using Unity, you're not an independent developer any more : has there been a point where Unity did not require a subscription yet ? And when did it became a platform ?


> If you start using Unity, you're not an independent developer any more

How so? The usual definition is that if you're self-publishing, you're independent, but you seem to go by some other heuristic.

> has there been a point where Unity did not require a subscription yet ?

Yes, up until 2016 Unity sold "normal" paid licenses (as well as offering a free version), meaning pay once and keep using, rather than subscription-based like it is now.


Remember AMD's Skybridge project? They wanted to put ARM and x86 together on one chip, but they cancelled it. I don't know why, because it seemed like a cool way to bridge the gap. Maybe they were worried it'd become a reverse trojan horse like the Apple Macintosh clones that came about shortly before Steve returned in the 90s.

Seeing as Raspberry Pi has no dog in the game, and simply wants to offer the best product possible, I wouldn't be surprised if this is their intentional way of giving people RISC-V with the intent to allow them time to port their code to it so they can eventually pull the ARM stack. I say this because Jeff Geerling has a video about the Pico 2 where he explicitly says you can only use RISC-V or ARM cores. Not both sets at the same time.

I think it's probably because a full migration to RISC-V would allow them to innovate a bit more and cut prices, although I don't know what the restrictions really are with ARM's IP control and pricing. At $5, every penny counts though.


AMD also had a project to make ARM CPUs that would work with its AM4 x86 socket for Amazon (Project Seattle, replace by Amazon's in-house Graviton). I see the Pico2 in a similar light. If customers want ARM or RISC-V Raspberry Pi has them covered. if the market moves away from ARM, they'll be well positioned.


You can use both at the same time, but you can only use two cores. eg 2xARM, 2xRV, or one of each.


Intel has had very good marketing for decades, and they've also surrendered to marginless sales just for the sake of appearing dominant. AMD's strategy of late seems to be for higher margins, but lower share.


The redesign is terrible. It looks nicer, but ultimately it behaves terribly. It loses connection with reddit all the time, makes my PC fan speed up to a volume that not even games can achieve, and is downright just behaviorally bizarre. Videos stop and start randomly, or refuse to buffer. Images are too big and stick out of frames sometimes, things take 10+ seconds to load even if the rest of the page loaded just fine.

I switch the URL to old.reddit and poof! Performance problems are magically gone.


The performance is awful. If they just did a site wide CSS and HTML makeover, that would've been fine. But they somehow crammed enough JS in there to make my computer beg for mercy.


It is atrocious. Reinforces every possible stereotype about web apps. One tab of the reddit redesign after scrolling will eat more memory and cpu then most AAA games on my rig.


This makes sense.

A person harboring the most replication of a virus is more likely to produce a prolific strain.

An immunocompromised individual is also much more likely to harbor other viruses that can share genetic material with each other when replicating.

If any human body in the world can be predicted to produce a radically different new strain, it's one whose body doesn't suppress viral replication.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: