Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
groestl
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
16 ms
·
241.
▲
by
groestl
3y ago
I think non-continuous succession of the same entity comes down to using or grandfathering the law, claiming the same assets, but also honoring liabilities of the predecessor. It's the same as with companies.
242.
▲
by
groestl
3y ago
Or stuck in traffic while you walk by
243.
▲
by
groestl
3y ago
Tresor: As a tourist, could enter because I knew the DJs on the lineup and was wearing a black hoodie. Guy behind me with a fancy shirt got rejected subito, no questions asked.
244.
▲
by
groestl
3y ago
This is definitely not the case with their core cloud products. > almost every project not using Spanner was moving to Spanner This even includes Datastore. Even Datastore moved to Spanner.
245.
▲
by
groestl
3y ago
Avoidant attachment at its best.
246.
▲
by
groestl
3y ago
Three rules of crisis communication for your spokesperson: - They say what they know - They don't say what you don't know - Let them know only what you want them to say
247.
▲
by
groestl
3y ago
> Make music > Theorize Not sure about those.
248.
▲
by
groestl
3y ago
I must say I really like the crowd behavior, where awesome related posts pop up the days after something from years ago hits the front page :) In this case, it must be the Wintergatan post https://news.ycombinator.com/item?i
249.
▲
by
groestl
3y ago
So you can trade off 300ms of additional roundtrip time (on anything non-CDNable) at a cost of 3% revenue and reduce your infrastructure complexity a lot
250.
▲
by
groestl
3y ago
Something I also learned not long ago: human eyes can see the polarization of visible light https://en.wikipedia.org/wiki/Haidinger%27s_brush
251.
▲
by
groestl
3y ago
Composition is better than inheritance. So I refactored this into a ListConfigDispatcherServletInitializer which can accept a list of ConfigStrategies for dispatching. Added AbstractDaddyLongLegsConfigStrategy, to help with implementing con
252.
▲
by
groestl
3y ago
I also found that when building a feature iteratively, with feature flags for rollout, a simple feature degradation path often appears natively.
253.
▲
by
groestl
3y ago
> and avoid surprises I think for most people and organizations, auto-deletion of a server _is_ the surprise ;)
254.
▲
by
groestl
3y ago
Now. Does lateral bias in cradling determine handedness in infants as well?
255.
▲
by
groestl
3y ago
Depending on the country / culture / job description, "making stuff up" is sometimes a viable option for "adjust accordingly", on all levels of expertise.
256.
▲
by
groestl
3y ago
I appreciate your elaborate answer to my short post very much!
257.
▲
by
groestl
3y ago
> What would the rate of our universe's expansion tell us about the black hole we are inside of? I'm only a layman, but AFAIK, depending on the size and uniformity of the black hole, we could notice a non-uniform redshift, mayb
258.
▲
by
groestl
3y ago
After all, we might be in a very large black hole, without noticing.
259.
▲
by
groestl
3y ago
I've seen this. Usually a combination of no economical constraints and technical curiosity on the engineers side.
260.
▲
by
groestl
3y ago
> I'd rather rewrite this than understand how it works Sounds like "how should I know what I think before I hear what I say" ;)
261.
▲
by
groestl
3y ago
No, a single HTML page and 20 lines of Javascript is clear cut. But there's a _lot_ of instances where it's not that way, and still rewrites are being proposed.
262.
▲
by
groestl
3y ago
I mean public final was there before as well.
263.
▲
by
groestl
3y ago
I've seen broken hash implementations many times, so I'm not sure it would be a bad idea to require tests for them ;) My other favourite of trivial code that's broken: returning the same Iterator instance in Iterable.iterator
264.
▲
by
groestl
3y ago
Usually, everyone knows. It's just that they need to hear it from somebody else (or will ignore it intentionally and did so up until this point).
265.
▲
by
groestl
3y ago
"Only when looking at things globally" is a very interesting thing to say. I'm with you, generally, only that I consider myself part of the "them", and us v them definitely makes sense, given the absurd distribution
266.
▲
by
groestl
3y ago
They produce largely for the rich, is what I'm hinting at.
267.
▲
by
groestl
3y ago
Not for all of them in the same manner though. The richest 1% have a carbon footprint equal to the "poorest" 66%, for example.
268.
▲
by
groestl
3y ago
> Your worst offenders are industry Which produces for?
269.
▲
by
groestl
3y ago
To each their own. Some people enjoy giving parties, and then it's just another reason to throw a big party.
270.
▲
by
groestl
3y ago
> it’s already priced into your profits Assuming you have any. You might not, because of AWS.
More ›