Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
bcherny
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
20 ms
·
271.
▲
by
bcherny
5y ago
Good documentary about Steward Brand, the author of Whole Earth Catalog: https://www.weareasgods.film/ He also founded The Long Now. Come to a talk if you’re in SF — they’re great. https://longnow.org/ htt
272.
▲
by
bcherny
5y ago
Does this marketing piece for Lexus really belong on HN?
273.
▲
by
bcherny
5y ago
Sorry, I must have misunderstood the question! :)
274.
▲
by
bcherny
5y ago
The process is more complicated under the hood. I’m describing it from the POV of an engineer that wants to update a field — from that POV, it really is that simple.
275.
▲
by
bcherny
5y ago
> how does it handle ever-changing business-requirements that can retroactively apply to existing objects and associations If the change is backwards-compatible: 1. Update the schema definition 2. Run codegen 3. Done If it’s not: 1. Crea
276.
▲
by
bcherny
5y ago
I mean... https://www.wsj.com/articles/facebook-to-hire-10-000-workers...
277.
▲
by
bcherny
5y ago
I just finished Egan’s excellent Permutation City, which has a fun and CS-heavy take on this. https://www.simonandschuster.com/books/Permutation-City/Greg...
278.
▲
by
bcherny
5y ago
In what ways is this different than today’s world, where depending on what you want to do (connect with people with similar interests, talk to friends, look at cat videos) there’s a bunch of options for what service to use?
279.
▲
by
bcherny
5y ago
This is neat! Is it open source? Or, do you have a write up about the algorithm somewhere?
280.
▲
by
bcherny
5y ago
Ah, seems reasonable!
281.
▲
by
bcherny
5y ago
What would you be trying to learn from the sample? A small sample would not be representative of what people see; a large sample would give the same results as the report.
282.
▲
by
bcherny
5y ago
/s aside, I wonder if there is evidence showing that breaches really are effective at reducing our reliance on SSN-as-auth.
283.
▲
by
bcherny
5y ago
Why would the government do a better job at digital security than TMobile?
284.
▲
by
bcherny
5y ago
Clarification: OC doesn’t seem to be arguing that these govt functions shouldn’t exist; they’re arguing that it’s wrong to trust govt with digital security more than we trust private companies.
285.
▲
by
bcherny
5y ago
These breaches keep happening, and I’m super pissed how little I can do. What concrete proposals exist for phasing out SSN as proof of identity in the US? And how can I (as a person in tech) get involved?
286.
▲
by
bcherny
6y ago
This is one of those “Product Management 101” things that engineers are surprisingly bad at. I couldn’t agree more that this is an essential skill for engineers (especially sr engineers), whether they’re working on infra, growth, or product
287.
▲
by
bcherny
6y ago
> Don’t attribute to malice that which you can attribute to incompetence.
288.
▲
by
bcherny
6y ago
I agree that decentralized moderation is right, in theory. What does that look like in practice — Reddit-style downvoting? And do you really think the really bad stuff — child porn, terrorist recruitment, and so on, that FB/Twitter
289.
▲
by
bcherny
6y ago
Some really interesting design choices here: Haskell/ML style FP base, fibers as a concurrency primitive, both exceptions and ADTs. More than those, I LOVE the macro system. Such a clean syntax!
290.
▲
by
bcherny
6y ago
Yeah, that’s been the thinking at least since Lippmann’s time. If you only vote on local issues, you don’t need media to inform you since you can see it first hand. But as soon as the average citizen is expected to keep up with state, natio
291.
▲
by
bcherny
6y ago
This piece makes the case that the only way traditional media can hold powerful people accountable is by being the arbiters of their distribution, by controlling who gets published and how their story is told. The way things used to be befo
292.
▲
by
bcherny
6y ago
GDPR, CCPA, etc. ban this, right? And without a permanent user ID (cookies etc. are super volatile), I’m not sure how valuable this data is.
293.
▲
by
bcherny
6y ago
If what ways would an employee-owned company produce better results? Unclear how that’s related (especially since much of FB stock is already owned by employees).
294.
▲
by
bcherny
6y ago
How do you tag someone that doesn’t use Facebook?
295.
▲
by
bcherny
6y ago
Sorry if it was unclear, but my argument is that while many libraries solve pieces of the problem, no one library (or standard) solves all, or even most, of these.
296.
▲
by
bcherny
6y ago
Author here. Data fetching, caching, consistency, and UX are all closely related. If you treat them as separate problems, you're punting the problem onto product engineers, who won't solve it well. (See the last paragraph of the p
297.
▲
by
bcherny
6y ago
I'd agree Relay is relatively far along, compared to other frameworks. I might lump its support into a few groups: - Full (eg. request batching) - Full, but painful (eg. colocation) - Partial (eg. mutation queuing) - Non-existent (eg.
298.
▲
Data fetching on the web still sucks
(performancejs.com)
155 points
by
bcherny
6y ago
|
96 comments
299.
▲
by
bcherny
6y ago
You’re both right. In philosophy, you call this an “ultimate” cause, vs. a “proximate” cause. You push a marble with your finger, and it moves. Why? Because the momentum from your finger was transferred into the marble when you touched it (
300.
▲
Ask HN: technical book club recommendations?
2 points
by
bcherny
6y ago
|
2 comments
More ›