kbolino

18 ms·

301.
by kbolino 1y ago
But what does "heap my_heap_var" actually mean, without a garbage collector? Who owns "my_heap_var" and when does it get deallocated? What does explicitly writing out the heap-ness of a variable ultimately provide, that
302.
by kbolino 1y ago
There's a matrix provided by Amazon themselves listing the level of IPv6 support in their various services: https://docs.aws.amazon.com/vpc/latest/userguide/aws-ipv6-su...
303.
by kbolino 1y ago
Both interface endpoints and gateway endpoints are also called VPC endpoints. The former get distinct IP addresses in your VPC subnets while the latter get distinct entries in your VPC routing tables. They are even created with the same API
304.
by kbolino 1y ago
There were still a couple of services/features that choked on IPv6 last time I looked (1.5-2 years ago) but it works with most things and they do seem to be making progress on the others.
305.
by kbolino 1y ago
Fair point, and valid for S3 (the topic at hand) and DynamoDB. Other AWS services, though, don't support gateway endpoints.
306.
by kbolino 1y ago
I doubt the traffic ever actually leaves AWS. Assuming it does make it all the way out to their edge routers, the destination ASN will still be one of their own. Not that the pricing will reflect this, of course. The other problem with (i
307.
by kbolino 1y ago
The problem is that VPC endpoints aren't free. They should be, of course, at least when the destination is an AWS service in the same region. [edit: I'm speaking about interface endpoints, but S3 and DynamoDB can use gateway endpo
308.
by kbolino 1y ago
Wait, why large asset sizes? Aren't we talking about "reduced texture sizes" etc.? To me, this also implies reduced video sizes, compressed audio, and so on. So the scenario as I understand it is that, broadly, no individual
309.
by kbolino 1y ago
In this scenario, low latency would be a lot more useful than high throughput.
310.
by kbolino 1y ago
No, we're talking about games with invasive anti-cheating mechanisms. The proposed alternative was to just have "community servers where cheaters get banned" -- and I responded specifically to the "cheaters get banned&qu
311.
by kbolino 1y ago
No, the question is "why do we even need invasive anti-cheating mechanisms" with the proposed alternative being to simply ban people (of course, bans are still used even with anti-cheat). Yes, banning a license key (assuming you h
312.
by kbolino 1y ago
Because the unmoderated servers are full of other cheaters. Some companies have tried a strategy of quietly shunting cheaters off to cheater ghettoes but the cheaters figure it out pretty quickly. With some limited exceptions, the cheating
313.
by kbolino 1y ago
Yes indeed, but this is basically a particular kind of "private, invite-only server".
314.
by kbolino 1y ago
Well, some of them are smart enough not to immediately log back in and spam "I'm the guy you just banned! Ban me again!" in global chat. And the admins, even paid ones working for big corporations, have finite patience and ti
315.
by kbolino 1y ago
For many cheaters, $60 is nothing. Either because they have money to burn, or because they're not getting it legitimately in the first place.
316.
by kbolino 1y ago
> The solution existed starting in the late 90s through the early 2010s [...] the cheater gets banned. This attitude also reflects the naivete of the 1990s. You can't ban a person. You can ban an IP, you can ban "all known Tor
317.
by kbolino 1y ago
It's not a global on-off switch. With a proper IOMMU, the hypervisor/operating system can lock out specific devices from DMA access, or confine them to specific address ranges.
318.
by kbolino 1y ago
If you think of a commit as a Merkle tree, then a file's content is a leaf node, and thus has nothing to verify. It either exists or it doesn't. Non-existence creates usability problems, but not verifiability problems. Of course,
319.
by kbolino 1y ago
What matters in the long run for making a good wiki are the people and policies. Young and niche wikis are happy to take any contributions they can get. The quality and timeliness of any given bit of text soon ends up wildly different from
320.
by kbolino 1y ago
It actually starts with Gogs. Gogs is FOSS but basically BDFL. What does and does not make it into Gogs (it's still around) is ultimately decided by one person, and he's fairly conservative. Gogs is very fast but lacks a lot of fe
321.
by kbolino 1y ago
In the past, at least, RAM upgrades weren't just about filling in the slots you couldn't afford to fill on day one. RAM modules also got denser and faster over time too. This meant you could add more and better RAM to your system
322.
by kbolino 1y ago
Is that really the long pole in the tent, though? No doubt the most tightly integrated package can outperform a looser collection of components. But if we could shorten the distances, tighten the tolerances, and have the IC companies work o
323.
by kbolino 1y ago
As I said, consumers play a part here too. But I don't see the causal line from the physics to the stagnation, stovepiping, artificial market segmentation, and cartelization we see in the computer component industries. Soldering RAM ha
324.
by kbolino 1y ago
Is the problem truly down to physics or is it down to the stovepiped and conservative attitudes of PC part manufacturers and their trade groups like JEDEC? (Not that consumers don't play a role here too). The only essential part of soc
325.
by kbolino 1y ago
AGP came too late for the 486. It may have been VLB, a short-lived step between ISA and PCI (and then AGP).
326.
by kbolino 1y ago
In order to collect reliable statistics on compile times, crates.io would have to be doing the compilation themselves. There's no way to trust telemetry from random users, both due to situational variability (individual machine load, o
327.
by kbolino 1y ago
There are at least as many ways to pepper as there were to salt before salts became integral to the definition of a good KDF. To wit: KDF(password, salt XOR pepper, ...) KDF(password + pepper, salt, ...) KDF(password, AES128(salt, p
328.
by kbolino 1y ago
Not reusing passwords across sites greatly limits the blast radius but verbatim password exchange still carries its own risks. The widespread adoption of TLS addressed most of the issues, as I alluded to already, but there are still insider
329.
by kbolino 1y ago
To wit, each word drawn from a 10,000-word dictionary adds about 13 bits of entropy. At 4 words, you have (a little over) 52 bits of entropy, which is roughly equivalent to a 9-character alphanumeric (lower and upper) password. The going re
330.
by kbolino 1y ago
Right, but peppering was not part of my comment. You can't always pepper, and there are different ways to do it. It's (mostly) orthogonal to the matter. You do not have to do any transformations on the input when using Argon2, w

More ›