Hacker Newsnew | past | comments | ask | show | jobs | submit | more cpuguy83's commentslogin

I guess this time the pro really is for pros. The 120hz on the base model is great.


It’s a phone, who or what are those pros that use this in a pro way? And I’m saying this as a pro owner who only got it for the camera, which I use to take casual photos


The same Prosumers who need AirPods Pro.

I'm the same, thousands and thousands of casual photos of people, animals, flowers, landscapes with a 15 Pro Max. My DSLR sits unused in a case. It has to be the 17 Pro but I'm debating regular or Max.


The best camera is the one you have with you.


videographers are the only pro market left.

obvious by the extra storage and additional camera features


This is extremely narrow minded. As another commenter pointed out, you are driving on easy mode in terms of environment and where a majority of the training was done.

This is not a general solution, it is an SF one... at best.

Most humans also don't get in accidents or have problems with phantom breaking within the timeframe that you mentioned.


> Most humans also don't get in accidents

Have you met any humans? Or seen people driving?


Way to cut my sentence I half. That's not what I said and you know it.


Oh please - people excuse and dismiss major accomplishments, you can send a skyscraper to mars and people on HN will still be calling you a fraud.

The Bay Area has massive traffic, complex interchanges, SF has tight difficult roads with heavy fog. Sometimes there’s heavy rain on 280. 17 is also non trivial.

What Tesla has done is not trivial and roads outside the bay are often easier.

People can ignore this to serve their own petty cognitive bias, but others reading their comments should go look at it for themselves.


I have ridden in many Tesla-based Ubers with human drivers using autopilot.

Here outside of Los Angeles, about an hour east, they do not do well at all on their 'auto-pilot.'

Your area has the benefit of being one of the primary training areas, and thus the dataset for your area is good.

Try that here. I'll be more than happy to watch you piss yourself as the Tesla tries to take you into the HOV lane THROUGH THE BARRIERS.


Auto-Pilot is not FSD. It's akin to a regular carmaker's Automatic-Braking-System and Lane-Keep-Assist. If you're seeing it used dangerously that's user error.


> you can send a skyscraper to mars and people on HN will still be calling you a fraud

To date, SpaceX has sent nothing to Mars. Not to understate the company's accomplishment, but "people on HN" are fed up exactly with statements like yours.


People here just whine and complain - yes they’ve “only” just sent a skyscraper to space for now and caught the booster on reentry, it’s a work in progress (along with their reusable rockets, earth scale telecom side project etc.)

My point is people will still be calling him a fraud when they do get it to mars, no evidence is sufficient for the HN cynic that thinks their “above the fray” ethos makes them smart.

Tesla has had massive success despite the haters, the model y becoming the literally best selling car on earth and you wouldn’t know it from HN. FSD has gotten really good, good enough to use more than not as they continue to improve it.

The best thing about capitalism is the losers here don’t matter - the winners get rich and keep going.


I said nothing about SpaceX here nor did I condemn Tesla ... or even mention Tesla.


You downplayed what Tesla FSD can do and said I was being narrow minded and the Bay Area driving is "easy mode" and said vision isn't a general solution. I think none of this is true.


BTW, undrstandable to not have an example on the spot.

In general we do actually try to provide full context for errors from dockerd. Some things can be cryptic because, frankly, they are cryptic and require digging into what really happened (typical of errors from runc), but we do tend to wrap things so at least you know where the call site was.

There's also tracing data you can hook into, which could definitely be improved (some legacy issues around context propagation that need to be solved).

I've definitely seen, in the past, my fair share of errors that simply say "invalid argument" (typically this is a kernel message) without any context but have worked to inject context everywhere or do better at handling errors that we can.

So definitely interested in anything you've seen that could be improved because no one likes to get an error message that you can't understand.


Do you have an example?


Concerned about cve's but doesn't pay attention to the massive list of cve's for rootless setups which have a much broader scope/impact.


Anyone running this on Linux? I find it works fairly poorly there. To be fair, vscode is also not great for me (especially vim mode) on Linux.


Running on Linux here. Working great for me. If you are referring to font rendering, unfortunately the Rust ecosystem for it is still young, so there are improvements to be made.


I don't notice any major issue yet, except for that freezing on wayland which seems to be fixed already.


Zed was working great for me on linux. It recently broke support for using a remote env over ssh, but locally still works fine.


That is a ridiculous statement. In the case of userns exploits there have been many and it means that every unprivileged user can obtain root on the machine.

Whereas rootful docker is a well known thing, run on millions of machines, and none of the vulnerabilities discovered in its entire existence is as bad as any single priv escalation issue caused by allowing unprivileged users to create a user namespace.


Lol that's just like saying having no door is better because some people can pick locks and occasionally open doors.


No, it's saying the door is a lie. At least within the context of your statement about vulnerabilities.


Escalating from an unprivileged user to root by creating userns and exploiting various things in the kernel along the way.


https://microsoft.github.io/wassette/ does just this, using wasi components.


I mean obviously it's about chip production, but shitting on the chips act is a political stunt and nothing more.

Building fabs takes lots of money and time. Intel also doesn't have customers except themselves and have fallen far behind in the fab business and has a decade+ or mistakes to make up for.

What we have here is picking a winner and potentially insider trading/market manipulation with Trump shitting on Intel leading up the "deal".


That was what I was arguing against. It seems like people cannot stop themselves from making this about Trump. I guess he does that himself though. The problem with saying he was shitting on the CHIPS Act is that Biden himself didn't even pay out the money. It's because Intel made no progress on the fab. It's pretty clear behind both Biden and Trump is a desire to have Intel's foundry working. With Intel making no progress on it, I am assuming the call to take a stake in Intel was done to mitigate all the benchmarks Intel said they would make but failed to do. Perhaps the government figured they could pass the money to Intel without giving up the reigns this way. Either way, I am very convinced there is a national security motive under all of this and neither Trump nor Biden went down this path to grandstand on it.


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

Search: