Hacker Newsnew | past | comments | ask | show | jobs | submit | lousken's commentslogin

Haven't touched gemini api since they did not support having a $ limit per api key. Is it possible now?


Finally!

With a 10min delay, via aistudio

Good enough for a side project, not good enough for transferring banking system from cobol

That is actually what companies like IBM and Unisys are already doing today, LLM assisted porting.

https://research.ibm.com/publications/enterprise-scale-cobol...


Why not? I think we are perfectly capable on generating a test and validation environment that we can use for correctness. Most likely llms could do this better than engineers with zero to none domain and language knowledge can do these days. From that point on, rewrites would become feasable (not easy, feasable).

I would hope publishers would take note and remove it, having hundreds of megabytes of junk in the executable is just wasteful to put it mildly

Denuvo is there to prevent piracy within the first 90 days of release. Something like 60 to 80% of a game’s revenue is during that period. They don’t care that it’s eventually cracked, and they absolutely do not care about performance.

> Denuvo is there to prevent piracy within the first 90 days of release [...] They don’t care that it’s eventually cracked

Ah, so Denuvo is always removed after ~90 days after release, as there is no point for them to keep it there?


Not strictly after 90 days, but Denuvo is usually removed after the peak sales period for a game. It's really at a publisher's discretion when to remove it, as the sales model for Denuvo is that you have to continue paying for it on a subscription basis to keep it active.

This is untrue. Yes Denovo got removed from some games relatively early, but mostly it was long after this "peak sales window" I would have to make a list of how long it took for games, and I am too lazy to even ask AI, but I think it took years in some cases and a lot of community outrage for the devs to remove it, and they did not just remove it after some peak sales window but when the games were actually cracked and the steam forums were flooded with pissed of people who realized pirates had a better experience then actual buyers. THEN they removed it.

So it's more like after they were cracked rather than some time window, sometimes these may have been overlap.

1 year after release is for sure not "peak sales window".


Yet I have a bunch of games on steam wishlist which I've been waiting for years to buy.

The stopper is of course denuvo, which they keep renewing the license of, for no good reason.


Maybe because a lot of users still have those games wishlisted?

Having the game wishlisted is a signal of players waiting for a sale, or future patches/correction, or simply not bothering to cleanup wishlist, not a signal of someone is eager to pirate the game.

Denuvo is sold as a subscription to developers, and it is often removed 6–12 months after release.

>often removed 6–12 months after release

that's not true. Only denuvo after ~2020 is subscription based and the contracts usually are between 2-4 years.


- Devil May Cry 5: released March 2019, Denuvo removed February 2020

- Forspoken: released January 2023, Denuvo removed July 2023

- Final Fantasy XVI: released September 2024, Denuvo removed March 2024

- Dead Rising Deluxe Remaster: released September 2024, Denuvo removed September 2025

These are just a few examples, there are many more. I can't say whether it was removed because the contract ran out or another reason, but, as I said, Denuvo demonstrably is often removed 6–12 months after PC release.


Pointing out exceptions doesn't invalidate the rule

I said Denuvo is "often" removed 6–12 months after release. Often means many times, frequently. I have given you examples that this has happened many times, so I'm satisfied with the wording I used.

You said it "usually" lasts 2–4 years. Usually means most of the time. What I said is not incompatible with what you said, but in any case, you've presented no data or evidence that Denuvo is kept for 2+ years most of the time.


There have been well over 150 games with denuvo the last 5 years. 4 games in that time is not "often" by any stretch of the word.

This isn't a good faith argument. You made a claim that has now been shown to be false, and now you're trying to reverse the burden of proof for your claim.

Whatever the pedantic meaning of "often" is in the context of this conversation, one thing is clear, your statement that Denuvo switched to a subscription service is entirely unsubstantiated. If you have evidence to back up your claim then the burden is squarely on you at this point to provide it.


My claim:

Usually removed between 2 and 4 years since 2020

Your claim:

No because here are 4 games from the last 5 years which where removed after ~6 months which means it is often removed before that time.

My rebuttal:

That's 4 games in 5 years from well over 150 denuvo games since 2020. Simple math should tell you that below 3% is not "often".

But somehow that means my claim is false...

It is well known that denuvo DRM is a SaaS subscription software for many years. I'm not gonna entertain your tantrum further for something you can trivially look up and should already know if you where knowledgeable enough to actually discuss the topic.


Denuvo should charge more for every month extra since the release

A number of publishers have retroactively added Denuvo to their older games, inexplicably.

Any list?

With the hypervisor method they get 0 to 1 day protection

Then DRM should automatically remove itself after that period. Copyright durations should also be adjusted to that same time frame.

The bigger problem with Denuvo is that it appears to significantly impact game performance as well

It can, but that seems to be more related to poor implementations by the game devs, and not inherent to it. There are plenty of examples of games with Denuvo that still run fine (give or take your opinion on whether the presence of DRM is inherently "impacted performance").

If many of your users misuse your tool, that's a design problem not user error

There's a valid point there, but that's stating it way too strongly.

I haven't seen their documentation, maybe the problem lies there. But there's no tool or documentation so perfect that nobody can use it wrong.


Yeah, goddamn hammers needs to be way softer, do you know how many thumbs worldwide have been hurt by them? Clearly the fault of the hammer.

If I have you hammer with a wax coated handle then it will regularly slip out of your hand.

One could blame the user for not "just" holding it right. Or alternatively reconsider if the handle should have a grippy coating instead.


More modern version: No you are holding your iPhone wrong, it is not a design fault that makes a ground loop in the antenna if you hold two metal surfaces with your hands.

Isn't Denuvo actually implemented in a game by the DRM developers, though? I remember reading that they have a process where the game dev sends Denuvo an unprotected executable, who adds the DRM to that executable and sends it back.

Thus, I believe the poor implementations are directly the fault of Denuvo.


I think that used to be the case years ago but isn't anymore.

The games run terribly on release because they have Denuvo, and then when the sales volume no longer justifies the licensing costs of Denuvo, the devs strip it out and sell it to the players in patch notes as "optimizing performance."

Someone else mentioned GTA getting more aggressive copy protection out of nowhere. It's not out of nowhere. With GTA6 ads out for a while, sales of GTA5 are up as people either play it for the first time or replay it. Sales going up means they can justify copy protection.

Denuvo has layers upon layers of obfuscation that inflates nearly every instruction and function call, extra code execution that does nothing to throw off someone trying to follow code execution paths, and constant moving around where the game stores stuff in memory, again, to throw someone off watching via debugger.

It's pathetic because one company has been almost entirely responsible for people needing to buy faster and faster CPUs and GPUs trying to eek out more and more performance. CPUs, GPUs, memory - all of it has gotten enormously faster, we have more cores, etc. Despite all that, every new game barely runs at 60fps.

Do you really believe that year after year game developers and game engines get worse and worse at performance? Of course not.


> With GTA6 ads out for a while, sales of GTA5 are up as people either play it for the first time or replay it. Sales going up means they can justify copy protection

How does that justify it? Adding stronger DRM when cracked copies of the same content are already out there is like trying to get insurance after your house has already burnt down.


> Do you really believe that year after year game developers and game engines get worse and worse at performance? Of course not.

If you strictly want to blame Denuvo then that assumes game developers cannot think of a way to spend their extra performance either. Which is obviously not the case.


I would assume the cache misses alone will destroy any performance.

The evidence for this supposed performance hit is basically zero.

False. There's lots of side-by-side recordings of Denuvo and non-Denuvo versions of games on YouTube clearly showing that Denuvo does impact performance.

No there aren't, because having identical builds of games with Denuvo actually removed and present is vanishingly rare.

If you compare a game that's had significant performance patches over a period of years and had Denuvo removed to the launch version (as so many of these videos are) then no shit you see performance differences, but it doesn't tell you anything.


https://www.youtube.com/watch?v=07NMuobVVwQ

This channel has many comparison videos like this one.

Loading times and 1% or 0.1% low FPS are usually hit the hardest and those stutters are the most immersion-breaking.


Should be noted that even scene cracks don't fully remove denuvo and all the performance intense checks so the impact should be even larger on actually denuvo free versions.

I would hope that users would just refuse to buy games that use Denuvo and similar malware. I do, but I know most users don't care.

Why would they care for a few hundred MBs when the games are in the 10s of GBs?

CPU cache space for code is much smaller than GPU memory for models (and the former is more important for performance since many CPU operations like pipeline parallelism are latency bound, not compute bound).

This. Why spend extra on x3d cpu when you can have a reasonable game size (not that it has large enough cache anyway)

>This. Why spend extra on x3d cpu when you can have a reasonable game size (not that it has large enough cache anyway)

Because game(SW) devs/publishers don't care about spending money to optimize for reasonable size, and the enthusiast gamers want to play the game either way and will gladly fork out the cash for the HW to play it, if anything for the bragging rights.

Remember "will it run Crysis?" vintage 2007? Yeah, enthusiasts will be enthusiasts.

I'm a fan of the free market here. Badly optimized games will hurt their sales and force the studios to change or go bust, if the market decides so.


That was not enough because of DRM junk though

Remove DRM and let buyers suffer less? Crazy talk.

> This update moves Windows toward a single monthly restart by consolidating OS, .NET, and driver updates

I just can't, gotta ask - what about c++ updates? What about integral os components that were migrated to the store and if you disable it, you won't get updates? What about defender updates (not definitions but app update) that won't get applied if you have another anti malware?

The thing I hate about windows updates is that microsoft can't even update all their own stuff with a single button.

edit: almost forgot - why is office not in windows update, and what the hell is wrong with teams and why it is seperate from office updates

Just updating windows is a complete and utter mess and every single Linux distro is 100x better


trixie security just out, perfect


AI was always slowing me down, only recently it has become somewhat useful.


Can't they use Mythos to figure out their uptime?


They weren't able to use it to prevent Claude Code source code from leaking, or from some random Discord server from gaining access to Mythos.


> prevent Claude Code source code from leaking

That's silly. It's a JavaScript app, they are more or less open source by design. There was no secret sauce in Claude Code.


Odd hos they still DMCA'd the rehosts of the leak. Clearly they dont consider it "open source".


Mythos prompt: Hey Mythos, make me 20,000 H100s.


Careful, you might just get turned into an H100 along with everything else in the observable universe


Availability is priority? Does not seem like it is https://mrshu.github.io/github-statuses/


One of the reasons I no longer work on windows

1. Log onto a server

2. Need a browser to run some stupid tool, everything besides edge is removed

3. Edge shows "setting up profile" with multiple steps (cant be skipped ofc) and after setup it goes to new tab showing shit ton of ads and stuff lagging the RDP session

explain that, macroslop


My favorite part of that dance is, "Do you want to import from Chrome? [no] Should we import your data from chrome? [no] Ok, time to import your data from Chrome. [skip]" It asks you three times right in a row. Really weird.


My favourite part is saying "Reject all" to the cookies and sh*t on my new tab before I even open any website.


> One of the reasons I no longer work on windows

Amen. I can't take the abuse anymore so I run Linux at home and if I have to interact with Windows I try through Wine before firing up a VM or breaking out the Surface that mostly collects dust. I'd rather run ReactOS in production than use Windows 11.


Can you still curl the chrome installer and run it as a user without elevated permissions? Seems a lot faster.


  Invoke-WebRequest 'https://download.mozilla.org/?product=firefox-latest' -OutFile firefox_setup.exe


Technically yes, but then you wait for the download + install + you have to select your ad preferences, also non-skipable


Of course, did you think fleeing from Microslop to Google would spare you the ads?


I like the word macroslop. But I think we give Microsoft too much credit with that word. To me it is just miniscule in thinking - so the name I use is microslop, the spam-AI company. Showing ads is what these microslop companies do - by abusing the end user.


They are entshittifying their products at scale


Fully agree, I can't wait for the day when developers finally stop using javascript for shit it was never designed for. .NET is decades ahead at this point.


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

Search: