Just tried asking Llama 3.2:3b to write a YAML file with Kubernetes Deployment definition. It spit the yaml out but along with a ton of explanations. But when I followed up with below it did what I want it to do.
>>> Remove the explanation parts and only leave yaml in place from above response.
apiVersion: apps/v1
kind: Deployment
metadata:
name: my-deployment
spec:
replicas: 3
...
Alternatively this worked as well
>>> Write a YAML file with kubernetes deployment object in it. Response should only contain the yaml file, no explanations.
... ions.
```yml
apiVersion: apps/v1
kind: Deployment
metadata:
name: example-deployment
spec:
replicas: 3
selector:
matchLabels:
app: example-app
template:
metadata:
labels:
app: example-app
spec:
containers:
- name: example-container
image: nginx:latest
ports:
- containerPort: 80
```
If anyone has an idea how in the modern day one could run a search engine, be sustainable and provide quality results without being influenced by advertisers, SEO hacks and also maintain security around it all - infected pages etc. - they are not on HN and are probably too busy implementing their idea. Kidding aside as a thought exercise what can people think of to kind of sort of make modern search engine work?
I am asking because I get the feeling that there's not much you can do and Google's way is the only thing anyone can do and still be in business. Now how much they rely on ad revenue and how much more they have to compromise is another story. Probably Google should be thinking of including better search to paying Google One customers?
Biggest problem is inertia - that would have to be solved - one way is that Google search gets useless and people have an option to move to something better. I doubt this will happen. Other way is someone builds not just a marginally better product but a 10x better product and people move there - I don't see that happening anytime soon either.
We could have an “organic” search engine which only positions itself for “normal” webpages, and defines normal. For example, the webpage should have a title and max 4 paragraphs, on the topic, and shouldn’t include storytelling. Technical problem-answer oriented pages should only contain various aspects of the problem, but no storytelling either. Fewer phrases to index, more density of keywords, easier to index. And maybe we should come back to the rules of 1. speed 2. content being in the original HTML 3. the react hydration shouldn’t dilute the HTML.
It doesn’t matter what its artificially-defined rules are good. But people would enjoy going there better than on Google, because you’d find the organic pages.
It doesn’t matter that Google would also index them. It’s like the Panamax, it defines rules but others can use those rules to.
Ads aren't the real problem though, SEO and dark patterns are.
Ads, even when mixed in with legitimate content, are labeled. A user can reasonably infer an ad website's goal when they click on it: to get you to pay for something.
The real problem is the system enables bad actors to do the same, without gaining the Ad label, by gaming the system to outrank legitimate (or free) sources of information.
Seems premature. You don't write successful browser engines using relatively new language that is optimized for one platform.
But it was predictable - Andreas comes from Apple background so.. The other thing is no one can build a browser alone - the ability to attract the right people is what differentiates a mainstream usable browser and a toy one. IOW focusing on adding another language that is not as popular amongst the right developer demographic seems like the wrong thing to focus on at this stage.
Ok. But the good old push code to staging / canary it before mainstream updates was a simpler way of solving the same problem.
Crowdstrike knows the computers they're running on, it is trivial to implement a system where only few designated computers download and install the update and report metrics before the update controller decides to push it to next set.
It would mitigate the problem, but not solve it. You can still imagine a condition that only occurs after the update has been rolled out everywhere. Furthermore, such a bug would still be extremely problematic for the concerned customers, even if not all of them were affected. In addition, it would be necessary to react very quickly in the case of zero-day vulnerabilities.
Yes, I am not arguing against having the ability to deal with it quickly - I am saying canary/ staging helps you do exactly that. Because as we see in the case of Intel CPUs and Crowdstrike some problems or scale of some problems is best prevented.
"Mitigation" is dealing with an outage/breakage after it occurs, to reduce the impact or get system healthy again.
You're talking about "prevention" which keeps it from happening at all.
Canarying is generic approach to prevention, and should not be skipped.
Avoiding the risk entirely (eBPF) would also help prevent outage, but I think we're deluding ourselves to say it "solves" the problem once and for all; systems will still go down due to bad deploys.
with the way they handled the debian crashing a little while ago, frankly they are happy to still go ahead with testing this way. still much better way to handle things than pushing to everybody at the same time.
Why trust somebody else not messing up? With that in place for windows and crowdstrike billions of dollars would be saved and many lives not negatively impacted
...
You said data doesn't show they drag their feet and then proceeded to present anecdote of your personal preferences and use cases while adding that thoughtful UI and battery life are the features and not web standards or the implementation quality of it nor the lack of 3rd party browsers on iOS - https://news.ycombinator.com/item?id=31902707
Sure they have recently implemented some features like IndexedDB but the data does indeed show that they dragged their feet!
They did years ago, as of late they are in fact moving faster than others. I think my point stands, they are no longer clearly behind in features, and in fact probably are near tied if you subtract the Chrome-only stuff, and take into account there's a variety of things Safari has that others don't now.
Site is copyright 2022 - 2024 and 8 jobs are currently listed (7 of which were sourced on the same day) - unless I am missing a "next page" box.
I would humbly suggest dropping the $99 job listing fee, reach out to employers and manually vetting postings to keep some quality and try to get some critical mass.
I recently got a UDM and don't recall having to sign in and I was able to set it up and everything functions including automatic updates. Where did you have to sign in and for what product?
At some point they changed to doing everything via the 'cloud'. Obviously their customers didn't like that so they gradually backed down.
I want to upgrade my unifi AP so I ask on HN or other forums once in a while. Last answer I got was that you can disable the need for cloud but you still have to sign on somewhere once to be able to disable it.
So I repeat: are you absolutely sure you didn't need to give Ubiquity any info or go through their servers this time?
With the APs you can set them up in standalone mode without the need for a controller at all, but you get more out of them with a controller so if you have the means to run a controller you absolutely should. With the controller you're heavily incentivised to sign in, but you don't have to.
With their "Cloud Gateways" (e.g. UDM Pro, UDR, UX, UCG-Ultra) it used to be difficult if not impossible to get them into a usable state without signing into a UI account, but that is no longer the case since a few years ago.
There are some exceptions, if you wanna use their cameras you lose out on some features if you don't sign into the cloud to activate those, specifically facial recognition stuff.
All I want is a standalone AP. I don't care about advanced administration, I just want it to connect my wifi using devices to the rest of my network.
I switched from consumer to Unifi just because of the better signal.
Unfortunately, even in just this thread there are very mixed answers. One yes you can (and a downvote that probably means the same thing), your 'maybe' and one 'i couldn't'.
OK so I just setup a blank controller (version 8.2.93, latest stable), no UI account (you just click "Advanced setup" and then skip when it tries to get you to supply an account), I then adopted a factory reset UniFi AP AC HD I had laying around, it accepted it without complaints, and automatically updated it to the latest firmware available (this is an option that's on by default in the controller software).
I then factory reset the AP AC HD again, and set it up as a standalone AP using the Android app on a freshly reset Pixel 6 running the latest version of Graphene OS.
Both methods worked fine to setup the AP without any UI accounts involved.
This is available on all APs able to run the current version of the AP firmware, which is literally all the APs released since and including the AP AC Lite (which is from 2014 or thereabouts, they support their devices for a very long period).
7 Pro Max is the current top-end model.
I highly recommend setting up a controller for managing the AP though, even if you don't keep it running 24/7 (you can simply start it on any PC whenever you want to make configuration changes), as when they're in standalone mode a lot of the cool features don't work, you only get one SSID (per radio) and no roaming support for example, whereas when managed by the controller you can have up to 8 SSIDs per radio on the latest APs, you can access telemetry, setup roaming, etc.
I honestly can't recall that ever being true for the controller, but it absolutely was true for the UDM/UDM-Pro on release and for quite a while after, so I guess it might have been true for the controller as well for a while around then too, I don't setup controllers that often.
Maybe it was just hidden, like it is now, but it pissed me off enough to not research it.
> I don't setup controllers that often
I've only done it twice; once when I installed the AP the first time, and once when I retired the machine with the controller so I needed to install it on a new desktop. Second time is the one where I failed by refusing to make an UI account.
I bought one of their entry level wifi routers and was completely unable to configure it without first setting up an account on their cloud service and using that to activate the router. Maybe it's possible, but if it is they hid it very well. I know their hardware tends to get good reviews, but the whole experience left a really bad taste in my mouth
I bet this was between 2018 and 2022? Yeah, that used to be the case, not anymore though, there's literally a big fat "Setup Console Offline"-button on the latest firmwares (since at least a year back I want to say), if you somehow find one with older firmware you can simply boot it into recovery mode and update firmware before setting it up for use.
reply