Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
ska
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
58 ms
·
871.
▲
by
ska
4y ago
A branding problem doesn't mean all of the content is bad, no?
872.
▲
by
ska
4y ago
I've seen similar clauses in a bunch of contracts, it can't be that rare. You are unlikely to be able to add your own language (as per your case of non-exec) but you may not have to.
873.
▲
by
ska
4y ago
Risk of injury comes from two things: Environment and stresses on the body. Environmental risks can be mitigated, if the job location inherently has risks, you will be better off with a regulated industry where the regulations have teeth,
874.
▲
by
ska
4y ago
The world isn't printf debugging or debugger only. Some things are really amenable to the debugger, especially simple bugs. Really especially ones that should have been caught by static analysis anyway but that's a separate issue
875.
▲
by
ska
4y ago
It's definitely both. Ideally the debugger is never your first resort. An extremely common failure mode for less experienced developers is messing about in a debugger all afternoon for a problem that should be solvable in 5 minutes.
876.
▲
by
ska
4y ago
Lots of places have fixed rate, but not for long periods like the US. Does France offer 30 year low rates?
877.
▲
by
ska
4y ago
I can't think of another country that has anything like it, and it's largely because of the government market interference (via Fannie/Freddy especially) It's going to continue to have some weird knock on effects.
878.
▲
by
ska
4y ago
> non-competes should be banned, yes. There isn't really any problem with them if they are properly compensated. Which isn't to say they are a great idea, but not worth banning if it is a contractual agreement on both sides.
879.
▲
by
ska
4y ago
Acquisition deals often include clauses to address exactly these sort of things, they include both carrots (e.g. earn out details) and sticks (specific prohibitions).
880.
▲
by
ska
4y ago
Ah ok,then my next thought is grinder. For what it's worth I found a huge difference in consistency after getting a decent quality burr grinder.
881.
▲
by
ska
4y ago
another possible factor here is that roasted beans aren't so stable over time, so "the same" materials, aren't.
882.
▲
by
ska
4y ago
> The only "value" they provide is.. As far as I can see this is mostly an incentives problem. Bureaucratic control of academic hiring has ended up emphasizing the short term measurable (e.g. #of pages published, so-called imp
883.
▲
by
ska
4y ago
This is a wildly inaccurate take. There probably exist examples of both the extremes you describe, but in such small numbers as to probably be irrelevant. On the whole the industry is conservative but changes fairly quickly.
884.
▲
by
ska
4y ago
>I think it's a pretty basic law of supply and demand: increase supply and value goes down. Aren't you assuming photos are fungible? Smartphones have made "snapshot" photography more accessible than ever, and killed
885.
▲
by
ska
4y ago
> meaning many people will be underwater. Overall interesting overview of how amazon specific details combine right now, but I don't think "underwater" is the right term here - in normal usage that refers to an asset that
886.
▲
by
ska
4y ago
I assume because equity is more than 50% of total comp. Or was.
887.
▲
by
ska
4y ago
oh i haven't relented, I just don't think it will do much good :)
888.
▲
by
ska
4y ago
Remains to be seen, honestly. Clearly a noticeable leap has been made, but it's far short, so far at least, of many claims being made. For what it's worth, "except it's really happening this time" has also been sai
889.
▲
by
ska
4y ago
variations of these algorithmic techniques have been around for decades, yes.
890.
▲
by
ska
4y ago
Unfortunately it's an abuse of terminology that "stuck" a long time ago, and I think the ship has sailed for fixing that (at least in the image processing community). These sort of techniques were named in reference to more
891.
▲
by
ska
4y ago
Huh, it was but I can't take credit for being clever about it, just simple order of magnitude estimation with a factor: e.g. I thought of a handful I was pretty sure were 500k+, rounded up to 10, then doubled as I know I don't kn
892.
▲
by
ska
4y ago
so is "grow wary", which is the potential problem.
893.
▲
by
ska
4y ago
At a guess there's at least 20 companies with head counts that high, and most of them aren't tech and retail.
894.
▲
by
ska
4y ago
>> "I would've expected it to not get such a basic query so wrong." Isn't this exactly what you would expect, with even a uperficial understanding of what "AI" actually is? Or were you pointing out that t
895.
▲
by
ska
4y ago
I suspect you mean "owned" by ran. Trades are a decent entry into owning a small business with higher barrier to entry than say a restaurant but still pretty straightforward and can be made revenue stable with a small team (somet
896.
▲
by
ska
4y ago
I think you are partially right, in that it should be normalized that some IC's may be delivering more value than their managers, and compensation should reflect that. The broad version, that managers just don't contribute as mu
897.
▲
by
ska
4y ago
> and cost more. for the same spec, sure. I think virtuals make sense at both ends - either dynamic scalability for large N is important, or you only actually need a small fraction of a physical box. Paying 45/mo for something
898.
▲
by
ska
4y ago
> he recruiter said she averaged about 4 weeks a year (which is below the legal minimum) how does that work?
899.
▲
by
ska
4y ago
One hint is that you only see this in jurisdictions without strong labor laws around PTO. Some places, you have to give at least X weeks accrued a year for all full time employees (might be laddered by tenure) and you have to allow them
900.
▲
by
ska
4y ago
At a certain level, this tends to be balanced with severance. "Garden leave" is different from "I can't get a job".
More ›