Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
haimez
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
18 ms
·
121.
▲
by
haimez
5y ago
Is it too soon to ask whether their mongo instance was publicly available without password protection?
122.
▲
by
haimez
5y ago
https://www.usenix.org/system/files/1401_08-12_mickens.pdf If you have a mossad problem, big integers were never going to save you from that before or indeed, after QC.
123.
▲
by
haimez
5y ago
TIL factoid of the month, easily.
124.
▲
by
haimez
5y ago
> most fragrant violations I love the smell of marginally improved security practices in the morning.
125.
▲
by
haimez
6y ago
FWIW: that 40MB single stream limit doesn’t seem to exist from my empirical measurements in the last two years. When did you work on S3?
126.
▲
by
haimez
6y ago
Really seems like you don’t know what sargable means (and didn’t bother to find out if this is your reply).
127.
▲
by
haimez
6y ago
Alternatively, where it _would_ be a good idea to produce spores if you want to hitch a ride on some passing foot traffic.
128.
▲
by
haimez
6y ago
This will be the fastest method of getting bank software rewritten in rails.
129.
▲
by
haimez
6y ago
You seem like you have a more detailed picture of the ground truth here, but that part you’ve basically said yourself (and I agree with, under the disclaimer that I found an “About” page and that’s the end of my knowledge): AIBS appears to
130.
▲
by
haimez
6y ago
Good to know, but that’s probably moving the goal posts a bit. This looks like a private industry group set up to proactively influence adopted standards (ie: a lobby) and does so internationally. At least in the US it’s fairly common for l
131.
▲
by
haimez
6y ago
I hadn’t heard of the ICC before, but their “About” page claims chapters in 38 countries and their “Find a Chapter” page ( https://www.iccsafe.org/membership/chapters/icc-chapters-and... ) has links for Canada, Aust
132.
▲
by
haimez
6y ago
> Why is this relevant? Rules and laws are there for a reason. I agree. > If society lets taxi drivers steal from people because they are taxi drivers, where do you draw the line for expected fair treatment of everyone? Do only rich p
133.
▲
by
haimez
6y ago
Hey, I don’t blame you for being angry- but, and this is only an empathy exercise: where do you think that taxi driver is today? Right at this moment, even?
134.
▲
by
haimez
6y ago
So, PoW then.
135.
▲
by
haimez
6y ago
Well, we should all definitely take his opinion on the matter, because he’s Gregg. Do YOU have an opinion?
136.
▲
by
haimez
6y ago
With business requirements that you never reduce speed.
137.
▲
by
haimez
6y ago
There were multiple attack vectors and some of them were significantly better at leaking information with hyperthreading enabled but the basic cache side channel leak worked just fine in single core systems before kernel patches mitigated t
138.
▲
by
haimez
6y ago
Read the article, but also: sortability.
139.
▲
by
haimez
6y ago
> Instead, you need to make separate parallel requests for those datapoints and let the dataloader be in charge of merging them into larger batches of requests for the db to fullfill. > This can result in some additional latency on a
140.
▲
by
haimez
6y ago
For SQL, always BEGIN first. If you’re unsure, run it as an EXPLAIN first. Then fire. Then commit it or roll back.
141.
▲
by
haimez
6y ago
I enjoy the irony of getting to be the one to point out and the poetry of “Andre Carnegie” immensely. I’m sorry this came at the expense of your typo.
142.
▲
by
haimez
6y ago
GPT-3
143.
▲
by
haimez
6y ago
The performance gains are very impressive https://tech.wayfair.com/2013/04/devnull-vs-mongodb-benchmar...
144.
▲
by
haimez
6y ago
Rust is statically typed, so it’s not quite the same effect
145.
▲
by
haimez
6y ago
Why would you think that? It seems like you’re assuming some limitation that isn’t inherent to the problem.
146.
▲
by
haimez
7y ago
Singletons (globals) considered stupid, a Yeggie classic: https://sites.google.com/site/steveyegge2/singleton-consider...
147.
▲
by
haimez
7y ago
Indeed, but it also doesn’t maintain locality of access nearly as well for young objects (the most commonly manipulated ones) and even older ones that survive.
148.
▲
by
haimez
7y ago
I’m not sure I follow. GC implementations that don’t copy (relocate) are inherently subject to the performance cost of “fragmentation” (in the sense of scattering memory accesses over non-adjacent regions). This is a very high price to pay
149.
▲
by
haimez
7y ago
Strong disagree with your thesis here. Most navigators of code are looking for something, and that something is not the most elegant refactor to make to cut some particular Gordian knot. Navigation tools are essential especially because the
150.
▲
by
haimez
7y ago
Are the emergent properties of a near random system preferable to live alongside/underneath? I suspect your answer mostly depends on your current socioeconomic status and overall faith in “the system”.
More ›