Hacker News new | past | comments | ask | show | jobs | submit | jlogsdon's comments login

Important to note that AWS doesn’t control spot prices: it’s a market where you bid on what you’re willing to pay. Customers control the cost through their bids.

edit; and prices tend to surge at end of quarter, regardless of provider. I’ve seen that for years at AWS and it’s only increased year over year.


asdf is a pluggable version manager for many languages and tools, not just one language.


Resource patches get richer the further you go out always in vanilla, not just rail world. Richness is a function of distance from your spawn point.


There are plenty of video interviews where Musk himself shows using auto pilot with his hands off the wheel.


It's even on their official sites demo video! As shared in another comment https://www.tesla.com/autopilot


It clearly states that the functionality in the video is coming in a future update. Its a video showing what the model 3 hardware is capable of once the software and regulation catches up.


No it doesn't.

On the Model S page one feature is called "Autopilot: Future of Driving".

The Autopilot page then leads with a still shot of a driver not touching the wheel. If you press play it says "He is not doing anything. The car is driving itself".

Sure, about 2/3 page lengths further down, hidden in the middle of a paragraph, it caveats it. But unless you read the fine-print, it's advertising hands free driving.


> once the software and regulation catches up.

What regulation needs to catch up?


There is none.


Every single automaker demos their lane-keeping technology with hands off the wheel.


Show me.



What's the benefit of this over using cert-manager?


I see some benefits:

1/ I got many issues for upgrading cert-manager and for managing logs. With Certs I do not have those issues, it is a light and simple Chart.

2/ Cert-manager is asking too much role permissions. Certs uses only api permissions for getting Ingress annotations and getting/updating secrets.

3/ Cert-manager is keeping a container running. Certs uses CronJob to check if issuing a certificate is necessary or not.

4/ Cert-manager does not support Godaddy dns validation for example. Certs supports all dns providers supported by acme.sh (more than 80), you can find the list here: https://github.com/Neilpang/acme.sh/wiki/dnsapi


I came to this thread having the same questions (why use this over cert-manager) and this answer gives some pretty solid reasons.

Can you elaborate more on /3 ? Why is this kind of service better suited for a CronJob over a running container?


Let’s Encrypt CA issues short-lived certificates (90 days) and official documentation recommends to check twice a day if certificates need to be renewed. It is a good choice to use Kubernetes Cronjob for this periodical task, resources and monitoring are saved.


I am using a free tier instance in GCP and using tinc to accomplish exactly this. It's not "click and play" but setting it up is simple. If I ever outgrow the f1-micro its easy enough to add another node with a public IP.

It has the added benefit of being a full-on VPN, though I don't generally use it for regular internet browsing.


It's not 100x as stated by the OP, but I wouldn't use it (async_hooks) in situations where performance is actually a concern. I wouldn't use node there either, but that's besides the point.

async_hooks are 100% usable in production right now, we use it at my current gig for tracing and its had minimal impact on the services using it.


No, they are not usable in production. We tried and was forced to rewrite whole app after this ASAP and eliminate usage of async hooks. We used them exactly same way: logs and some query-level context data.

Bluebird is bad for measuring async hooks because they have custom scheduler and most of the execution is done in single tick.

Overhead is for each tick, not each promise. If you have ~10 promises per request then you have N overhead if you will have 1000 promises you will get 100N overhead.

Our backend barely survived this after async hooks was just enabled and just disabling them make everything x100 faster.


Are you saying that async hooks perform fine with Bluebird but not native promises? Why not use Bluebird everywhere then?


Because this is not a solution, you still get penalty and bluebird is slower in latest nodejs versions.

Not really a long term solution.

We adapted approach with context from golang and now it works really well.


I see this "I'd never use node" sentiment quite a bit in the Java community. I'm always curious for the reason. Is it the JavaScript? or is Node unable to perform? After reading so many success stories from Netflix and Walmart, I'm wondering what the missing piece is....


Specifically in the area of high CPU load tasks. NodeJS is perfectly fine for heavily async tasks, especially calling out to external services. When you have CPU intensive tasks the lack of control over the event loop (task list) can be painful.


It’s not just Java, it’s any programmer with proficiency in any other language, including those that are quite proficient at JavaScript. The sentiment after our team wrapped up a recent Node project for a client was basically “the internet keeps lying about the maturity of the node ecosystem”. That was with the latest TypeScript and an architect who was very good.

The event loop model that Node uses is easily replicated in other programming languages. It’s not a sensible default though, blocking code is better for most domains.


From their Privacy Policy (https://www.gurn.io/privacy-policy/):

  4. We may collect the following Data, which includes personal Data, from you:
    ....
    l. A list of URLs starting with a referring site, your activity on the Website,
       and the site you exit to (automatically collected);


Hi jlogsdon,

This clause exists as we run Google Analytics on the main gurn.io webpage. It was a standard clause suggested by our lawyers here in the UK to ensure our users are fully informed about the possible data collection being conducted by the GA analytics.js tracking code.


This library doesn't run SQL, its only a template engine that gives you back a parameterized query and parameter array back. It's up to you to actually use them with an adapter, which means you can use prepared statements.


IndexedDB is standardized for IndexedDB, which is not at all designed for use on a backend.


What about IndexedDB's design is incompatible with running on a backend?


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: