04:54 babyfaceold: ran out of time a bit not sure i narrowed it down to right thing, folding of models might be the thing in spacetime and minimisation
04:55 babyfaceold: http://lifeware.inria.fr/wiki/Net-WMS-2/Software
10:57 pepp: hakzsam: the only result change is a Crash but the test logs say the result is Pass :/
14:47 babyfaceold: what i was thinking was that you might be messing with the wrong guy overall, if you could even stop it , it might be too late.
15:08 babyfaceold: it was conspiracy where i suffered under, those dudes are gone soon
15:08 babyfaceold: they get knocked down one by one
15:08 babyfaceold: one after another, they deserve that too
16:43 babyfaceold: other than that , yes we cash out the victory in computing soon, i looked at the cryptography end of things, and some kits live up to to the expectation, and are not the hardest nightmare or task to pull in, you need to to just discuss what are the third party licensing requirements, then you can proceed to patch those last holes you have.
16:45 babyfaceold: I switched my focus to defense cause i got the performance theories enough mature under the hood of my research. It's that in this world of conflicts its back against the wall to land some security things to be successful at all too.
17:07 babyfaceold: I have actually bunch of patents that do pay attention to the security side of things of network cards to avoid danger mentioned, the danger of hardware race at the memory buses. So i am not saying this all to be inventor, personally me i know thay have done security ontop of pci too but those are spin offs outside the mainline trees.
17:19 Caterpillar: is Mario Limonciello here?
17:41 babyfaceold: Where as with miden-vm and gini the performance extensions can be already exposed, pci and kernels chunks is different type of effort, its clear that i had my own painful experience , in other words i got exploited like a little outsider, there is no question about pride or honor that got hit, but i do not want some details to be transparent to all the mad society. But rescue prime optimization permutes are seeming very powerful in miden-vm.
21:47 babyfaceold: The whole idea is to make LBA/CHS indexing in compressed format, and have the software computer chiplets deal with computing things. That's actually not hard to be achieved, my coach always said bring the nuts out of the fire, that with added security works would bring the linux nuts out from the fire indeed. But work is enjoyable and interesting too. The highest underlying physical accelerators to design chiplets to, are GPUs and FPGAs.
21:56 babyfaceold: Cryptography is actually fairly simple too, if the opener move is enough correct, taking the key generation to certain base modules to culminate into ciphering the fixed banks, fastest is diffie-hellmann elliptic curve , this keygen has a dense layout of keys as it covers the whole bit length, it is in other words very hard to break. just two directional cryptography
22:14 babyfaceold: and it's not like miden-vm is hosting bugs of security kind either, just the arp/ndp out of spec packets that cause a race on the bus where they fix results through surveillance intrusions, its what i try to secure in kernels code. I have to look at pci sig stuff from kernel or efi tree.
22:14 babyfaceold: and reorder the things a bit
22:16 babyfaceold: so that is hard requirement for people who want to post their GPU crypto mining results, such as archival nodes
22:16 babyfaceold: requirement is not to get fooled and lose all your results, in military also that kind of vulnerabilities is though even worse, that is a matter of life and death
23:20 Venemo: Caterpillar: I don't think so but he is active on the gitlab so you can ping him there
23:20 Caterpillar: Venemo: we spoke on gitlab yes, I was just wanting to send him an informal comment about a thing