It's an example and what I had. It's intermittent and not 100% reproducible. So what value are you adding by complaining, which also absolutely won't help that you are accusing me of?
Looks promising! I have a project in the summer that is collecting various data (lidar, drone images for photogrammetry etc) in a remote area and I will definetly try this out. The idea is to sync files up as soon as possible via 4/5G and Starlink using BondiX. Have you considered supporting multiple uplinks/bonding in Keryx?
Great use case! Remote data collection with mixed uplinks is exactly the kind of scenario where intelligent routing matters.
Currently, Keryx doesn't natively support multi-uplink bonding at the application layer, it relies on the OS routing table and picks the best available route.
However, if BondiX is presenting the bonded connection as a single virtual interface to the OS, Keryx should work transparently over it. We've tested over various network types (fibre, cable, 5G, Starlink, etc individually) and the ML strategy selection adapts well to changing conditions.
Multi-uplink support at the Keryx level is on the roadmap - essentially treating multiple paths as independent channels and doing our own aggregation/failover. The transport layer we're using makes this technically feasible.
For your summer project, my suggestion would be:
Test with BondiX doing the bonding (should work today)
Let me know how it performs - real-world feedback from field deployments is gold
If you need native multi-path support, happy to prioritize that feature based on your timeline
Happy to stay in touch as your project develops. Feel free to email david@netviper.gr if you want to discuss the specifics of your setup, or if you are interested in testing out the GUI application.
Feels like everyone completely forgot about this. The White House made a big stink about AI deepfakes in their first couple of months and then proceeds to publish them at blinding speed. No hint of irony or self-awareness.
With both RF and optical you could see FEC or ARQ being used for something that isn't 100% signal loss. Downlink is optical, uplink is RF. Downlink transmits with FEC, user terminal fixes as many errors as possible, still missing packets so requests ARQ and either gets retransmission on optical or RF.
+1. I bet it's because of this confusing verbiage, the AI also got the gist of the article wrong, and lead me to believe that this article shows "post-hoc exploit" , when in fact there's no mention of the word 'exploit' in the article. See the screenshot linked below [1].
On a tangent, in the process I learnt that Firefox (at least on desktop) now has an "AI preview" feature where if you long-press on a URL, it brings up the pop-up. The first time, it notifies that the "AI" processing is local-only to preserve privacy.
To me, “hijacking” a passkey sounds like credential disclosure, which is quite worrying for a core team member to talk about. I know what you mean, but it’s probably the wrong term to be using if we want to emphasize that passkeys cannot be stolen.
Weighted votes throw secrecy out the window and they also don't allow hedging. I've been playing with the idea of 2x ballots for min_voting_age_years. So voters aged 18-36 will get two ballots in each election and can either do them identical or hedge their bets.
The author here, I ran another test here https://jsben.ch/QCXCY . https://jsbenchmark.com/ seems to have issues with longer running code.
1500 repeats, for me for-of is now on par with classic for. As I said, for-of loops are not optimized as easily and reliably because v8 has to prove certain things to be able to do so. For-of's default is iterator protocol, and its more sensitive to deoptimization.
100% / 92% / ~32% (forget exactly on this one) for for-of, classic, forEach for me, on a Ryzen 5700 AMD desktop. Other things running, not a fantastic test, but I did turn the cooling system up ahead of running. This defied my expectations & I had to try & see!
Side note, I am fantastically glad to see at least some kind of online javascript benchmarking site is alive again. https://jsperf.com was so good to us for so many years. Obviously there's flaws but the way that it not just was a benchmark site but also would record folks results & aggregate them was incredibly incredibly incredibly fantastically sweet.
Got similar results but then tested on an older win10 i5 system (8+ years old?) chrome 145 and get variable results but classic is always about twice foreach or forof and foreach often performs 2x better than forof. I fear this might be more connected to CPUs than anything.
Weird. If V8 produces different optimized codes for different loops, this might be true on some CPUs. But when I get 100% convergence when increasing the repeats, this becomes a bit suspicious. Also, it might be some timing difference related to initial warmup or some adjustments in V8 for specific CPUs (including some power saving). I also ran on Pixel 3a , both classic and for-of were 100%, foreach was ~9% (on Windows 11 and AMD Ryzen 5000U was the same, except foreach was 62.96%).
I created another benchmark https://jsben.ch/sdaEM . The difference is that doSomethingWithValue now assigns the passed value to some variable. I think the optimizer could still notice that the function was dummy and do weird things, but not sure 100%. So I store in a variable to prevent this. Now the tests run slower, but still both classic and for-of tend to be 100%. forEach is sometimes slower, sometimes on par with the other 2 (on both AMD Ryzen 5000U Laptop and Pixel 3a).
Could you please run this test on your mentioned machine? Also could you gradually increase the number of repeats in each test case and see if there is some convergence? Also make sure the tab and the browser window are active during the benchmark. Otherwise the browser / OS might give less priority to the task. And BTW, can you also test on Chrome 143 and even Chrome 144 (it will be released tomorrow)?
reply