The documentation reference a “decode” function, and it’s imported to the example code, but it’s never called. I’m not sure what the API is after reading the examples.
It’s the same assumption is required for any hole punching handshake (including STUN).
> This is a property called “equal delta mapping”
FWIW I’ve worked in computer networking for 20 years and have never heard it called this. This blog is the only source that comes up when I search for that exact term. I wonder where the author got it from.
> It’s the same assumption is required for any hole punching handshake (including STUN).
This is incorrect.
Hole punching requires being able to predict external port. That's it. If the port remains the same, it certainly simplifies things, but ports going up (or down) by 1 (or 2, or 5) with each new mapping is quite common, trivial to detect and to punch through.
I wonder how many new technical terms are going to be created by LLMs - not to say that this post was N
necessarily written by an LLM (but, who knows!)
It comes from academic papers on categorizing NAT behaviors which (trust me) is hardly the page turning research most people are used to. In these papers they talk about patterns NATs use between successive external port allocations -- which they call the "delta."
The name "equal delta" just means a type of NAT with a delta that tries to preserve the source port. Not to be confused with "preserving" type deltas (that preserve "the same numerical distance" between successive mappings -- e.g. a "preserving delta" type with a value of +1 means each successive NAT allocation is one more than the previous.)
The foundation models themselves will be cheap to deploy, but we’ll still need general purpose inferencing hardware to work along side them, converting latent intermediate layers to useful, application-specific concerns. This may level off the demand for “gpu/tpu” hardware, though, by letting the biggest and most expensive layers move to silicon.
The output of the transformation layers are a collection of embeddings in the latent concept space. Those can be fed into an addition model to extract semantic segments, bounding boxes etc. IIUC this is how dinov3 works.
The issue is how do you interact with other industries/trades who protect their profit making potential.
Ok great - all software and networks are "free." How do you pay for Doctors and Plumbers and Electricians whose earnings are legally protected by the state but whose skill bases are also freely available to be used within the margin of error of a professional or a layman?
Issues like this are great to have conversations about, but if people don't start broadening the scope very quickly, it just turns into the IT/CS worker's worth going to 0 in a world where others worth are protected. And history states, if only 1 group sees the threat, the remaining trades/industries will let it die.
It's not clear to me what your argument has to do with the license laundering service that Malus (Malice?) is offering. Their stealing from the digital commons does nothing to address paying Doctors and Plumbers and Electricians.
It's directed at the person I replied to. It's not directed at the top level OP or Malus which is hilarious, monetized satire.
Focusing overly on corporate structures or specific skills tends to miss the point of how value is assigned in a capitalistic structure when knowledge is cheap. Knowledge has been the capital used by the labor force for hundreds of years. The reason some jobs are resistant is 100% the result of legislation at that point, not anything unique about the job.
"The Trades" seems to be the sales pitch used on the public. In the end they're just labor at that point since I can pump a 20 year old with a master electricians knowledge, keep one master on staff and fire every other person who hits that level when their earnings demand it in the same way we're firing many mid/upper level people in their 30's and 40's now instead of 50's and 60's which is the scenario in Tech today.
Software/IT is just the quickest to be absorbed. Many other industries are just in the slow boil, not seeing it yet.
The value from FOSS is the collaboration between all parties.
There is a mutual agreement between all collaborating parties that "hey we ALL need these core fundamental building blocks of software. why dont we all collaborate in this open space?" And everyone wins.
There is tremendous value in the Linux kernel, and these large open source programs. And this is basically an attack by corporations to attempt to privatize it all.
It's nothing new. This is simply the latest example of capitalist "growth at any cost". We sailed past any immorality hazards a LONG time ago.
What's this 'fun' you mention? As far as the incentives in our systems are concerned, anything that's not done in pursuit of monetary gain is certifiably insane. What really matters in life is using all the tricks, manipulation, abuse and loopholes to attain the biggest number in your asset counter. Anyone who doesn't follow the only thing that matters in life is alien, inhuman even. How do they not see it?
> Right? A “contract” that only one party needs to abide by is not a contract… it’s an abusive relationship.
I think you're absolutely right morally, but I think you've made a pretty important technical error: they're not abusive because "only one party needs to abide...by the contract", they're abusive because only one party can unilaterally change the deal. The companies that make these "contracts" can actually follow them, but since they can change them at a whim, it only really binds the other party.
There are plenty of other abusive aspects besides the fact that they can be changed unilaterally.
What I really don't understand is how it's supposed to be a fundamental part of contract law that there's a "meeting of the minds" where both parties agree to the same thing, and there are these click-through agreements that nobody reads, and everybody knows that nobody reads them, but they're still enforceable. I get why there needs to be a general presumption that you've actually read a contract that you've signed, otherwise you'd be flooded with people saying "actually I didn't read that" to get out of contracts they don't like anymore. But that presumption doesn't make any sense when one party doesn't read the contract, the other party knows nobody reads it, and everybody knows nobody reads it, but we all just sort of pretend.
I particularly love the pretend play of software forcing you to scroll the dozens of pages of contract text all the way to the bottom before the Accept button is enabled. Because obviously the reason I didn't read through the entirety of these eulas before is because I wasn't sure of how scrolling works.
The only way they should be enforceable is if they use that scrolling trick, then quiz you on all the terms (with at least multiple choice), every time the TOS is updated.
Despite sounding absurd I think that would actually work really well. It would make it functionally impossible to include arcane BS without driving off customers while also filtering out people too stupid to be trusted with any sort of online account.
I'm reminded of Mitch Hedberg's bit about getting a receipt when buying a donut. "I don't need a receipt for the donut. I'll just give you the money, and you give me the doughnut. End of transaction!"
Why do we need massive TOS for stuff? I'll just give you the money, and you give me the service. End of transaction!
Presumably because an ongoing service isn't a clean exchange of physical goods. It's more analogous to a gym membership which definitely does come with a contract.
By eating this donut you agree that we are not responsible for any health problems that might result, either directly or indirectly.
Wonder how a court would treat it if users just reply to the email updating the terms of service on our behalf and claiming that they have accepted the terms by not doing anything. (Eg add stringent PII protection, no tracking requirements…)
My guess is that you would probably get kicked off the service if anyone reads your TOS, so make sure to add onerous cancellation charges due to the user in your updated TOS.
I could imagine an AI sidekick that does all this work for you, and always has the last word because it'll never give up.
A place like Meta or Microsoft would tell you to pound sand, but an aligned army of collective-bargaining agents might succeed in removing a specific term from a smaller service.
But generally the ToS has few, if any, requirements for the company. Usually the ToS is just a list of demands they make of the user in exchange for the service. But the company usually reserves the right to terminate service for any reason, as well as change the serice in any way they want, and change the terms of the "contract" at any time.
Ok that's no way to build a functional society, though. Humans are certainly not the entities in this conflict with the time or resources to go to court.
>If the company violates their ToS, you can take them to court (or arbitration).
This is my favorite...how exactly can I monitor compliance? No evidence of non-compliance - get tossed out of court. No court order for discovery - no ability to monitor/gather evidence compliance.
The idea that this is even a potential for mutuallity on a TOS is just farcical.
Often I see a popup to accept TOS after the update, which was run without me agreeing to anything.
At which time the company has unilaterally denied my access to something I already paid for without seeking my affirmative consent.
In theory I could stop whatever I'm doing, go email the company a brief to the point letter indicating they've broken their ToS and are unacceptably impairing my ability to use my property under the contract that I did agree to, and giving them an opportunity to amend their problem and give me a rollback path.
Realistically the outcome of this is a brushoff and needing to file a consumer protection complaint or get a lawyer.
If the feature is something like "my car" I can't afford that opportunity cost and am coerced into accepting their contract by the way they presented the amended terms.
I figure ToS for physical devices should be blanket outlawed. They're fraught enough for purely online services. Physical devices keep all of that baggage then add additional questions about whether or not I own physical objects that I purchased.
> Quality of code has never had anything to do with which products are successful. I bet both youtube and facebook's codebase is a tangled mess.
The code’s value is measured in its usefulness to control and extend the Facebook system. Without the system, the code is worthless. On the flip side, the system’s value is also tied to its ability to change… which is easier to do if the code is well organized, verified, and testable.
