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

OCI registries.

Harbor + Notary + admission controllers - AKA private image repository with image signing.

Sigstore. Another method for signing & verifying artifacts.


It did, but only as a test. https://en.wikipedia.org/wiki/STS-49


This sounds kind of like Tilt and DevSpace, but for just general purpose containers. From a quick look through the website, I didn't see these features, but these would be great additions:

- File sync with the local filesystem and the container (2-way)

- Port-forwarding to localhost for debugging (I guess the DevPod way is to run the entire IDE in the container, but I love Tilt/DevSpace for allowing me to work in my local fat IDE)


Port-forwarding and using your local IDE is already working in DevPod today. We also added auto-port-forward where it watches what happens inside the container and then starts port-forwarding automagically.

File sync: That is a great idea. We got that in DevSpace already as you mentioned and we definitely think this could be super valuable in DevPod as well. Right now, a git push and then pull is required to get things from inside DevPod updated on local but with sync this would be even easier and faster.


We have a designer on the team that uses a Windows box. This could be a great way to get him spun up without having to build him a local setup and manage two sets of local dependencies for him on Windows and us on macOS.

How well does it work for web dev - Django specifically? Is it as simple as running the dev server and accessing the locally-forwarded port through a local browser? Are there any gotchas to be aware of? Thanks!


Exactly. add a devcontainer similar to the example repo in [1], install the devpod app and point it to the repository and your designer is good to go, including port forwarding to their localhost

[1] https://github.com/backendclub/example-django-devcontainers


Thanks! Is it trivial to “copy” container connection information between machines? For example, can I write a devcontainer.json, spin up a back-end on Digital Ocean, make sure it has all the necessary envs and system dependencies and is working correctly, then just copy-paste connection information to that pod and send it to my designer, where he just pastes it into his DevPod desktop app and is off to the races?


Once you set up the provider on the designers machine, you can embed a deep link to the desktop app in your repo, similar to the „open in devpod“ button in our repo [1]. [1] https://github.com/loft-sh/devpod/blob/main/README.md (See line 11). This makes sure they open the correct repo with the provider you configured and are ready to go

For your use case it’d be cool to share provider information between parties. We‘ll think about it, although that might also be a potential selling point for teams and a commercial product…


Perfect, thanks for your time. Provider setup + deep link will work fine for us.


It is made by the same company that makes DevSpace so that tracks


This doesn't surprise me much. From what I've seen consulting/contracting, SaaS-based observability tends to cost 30-50% of cloud spend--EC2, storage, S3, RDS, maybe k8s, and other cloud services, or whatever the equivalent is on GCP/Azure. I wouldn't be surprised to see Coinbase with a >$150M quarterly cloud spend, so $65M on observability would make sense.

That said, managing observability yourself should result in <5% of cloud spend. So I'm figuring someone at Coinbase said "WTF" to this bill and migrated to Grafana/Loki or Kibana/OpenSearch or Kibana/Elastic. Well, that, and Coinbase's business also dropped off a cliff. Combined, I could easily see a one-time influx of $65M from one customer, gone the next quarter.


The beauty of EVs with respect to the grid is they present an opportunity to make our grid much more resilient and reliable, all while enabling renewables much greater percentages of our overall power generation. Renewables (well, specifically solar and wind) are very peak-y, with peaks that can overwhelm a grid and blow out transformers, and troughs that require additional power generation from other sources. This limits the base load power that can be generated from wind/solar to around 40% of total demand. Unless you have a way to store and retrieve power on demand. You’d need enough storage to handle roughly 2 days of power usage needs in order to smooth power usage to cover peaks and troughs. And, lo and behold, the typical EV car has enough power to cover a typical home for about 3 days!

If we incorporate inverters into home building standards today, we can guarantee that our EV fleet can be used to provide storage and auxiliary power for our grid and allow renewables to approach 100% of power generation. But we will need to deploy wind and solar in a distributed fashion so that power generation and storage is local to where the power is consumed. And a grid like this will be much more resilient to power outages and weather.


Watching the mission, I noticed that engines kept shutting down. There are 33 of them. When they had turned off 7 of them, with 26 engines remaining lit, it started to lose control. The engineers all seemed ecstatic about the mission. I have to wonder if they weren't intentionally shutting off engines to see how many could fail while still retaining control. If that was indeed tested, the answer appears to be 6--they can operate nominally with 27 engines lit.


It didn't start to lose control because the engines. It tried to do a stage separation flip and the stage didn't separate. Only then was it doing something that might have been out of control, or maybe it was trying to do the seperation again.

Control can be maintained with very few engine. I depends if its center engines or outer ring. Only the central cluster can gamble, 2 or 3 of those should be able to handle to rocket depending how full it is.

With only outer engines the could use differential thrust but the rocket isn't really designed to do that.

The Soviet N1 used differential thrust only. SpaceX prefers gimbled engines.


Does it also emulate the wickedly slow I/O times of the CD/RW drive (er, Canon Magento Optical drive, apparently)? I seem to recall waiting really long for anything to read or write to that thing, despite the wonder of having a CD that I could erase, with 660MB of storage.


How about "The Return of the System76 AMD Laptop"?


I have one about ChatGPT. ChatGPT ends up not being useful as a replacement for junior software developers. But ChatGPT does end up taking over middle management. It is more reliable, with better results than humans, at getting software updates from developers and communicating them with coherence to upper managers. Someone will use it to develop an Agile ChatGPT, then another with a Scrum ChatGPT, and then a third, initially a joke, but later accepted as the preferred one, the Waterfall ChatGPT.


Very good. I made a similar prediction years ago, yours is better.

I predict middle managers will rush to use one of these offerings, initially fantasizing about reduced dependency on supervising junior staff. I expect this to be boringly controversial, as some workplaces embrace it and in others, managers rely on it in secret. Retaining the vestigial office of scrum master or agile poobah will be a corporate status symbol to signal safety to VC.

On the top floor, AI systems styled to emit the messaging and maneuvering of, say, your company’s very own Elon Musk cutout, will be a useful fad for predicting political fallout. I predict AI will not compete with strong cultural networking among the kind of people who report to boards.

I expect stories will emerge where certain roles simply stand out as more committed to tending internally-important AI systems and services. For example, resume-vetting cutely predicts the kinds of applicants skilled in tending itself.


Yep. I've passed on this generation of fully electric vehicles because none of them offer physical buttons to operate the climate control.

Automatic climate controls don't cut it--change directions such that the sun starts beating down on me and I'm going to need to turn up the fans beyond what automated systems would choose. Test driving the Tesla Model S, on two separate occasions while attempting to set the climate controls I almost got into a wreck--it not only is only on the touchscreen, but also buried under something like 5 menus. WTF? Volvo XC-40 and Mustang also have no climate control buttons. In my area, all the other electric cars have a waiting list longer than a year. I want to buy electric, but safety is paramount--driving a car is far and away the most dangerous thing I do on a daily basis and I'd like it to be as safe as possible. I ended up with a Subaru this time.

Hopefully someone will make an electric car with physical buttons for the climate controls.


> electric car with physical buttons for the climate controls.

I drive a VW e-up! for the same reasons. It is a simple city car with analog controls and it's perfect for the Berlin city.


F150 Lightning (base model) has non-touchscreen UI.


The new Lyric has very nice tactile controls.


Consider applying for YC's W25 batch! Applications are open till Nov 12.

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

Search: