Every time I see one of these posts and the ensuing comments I always get a little bit of inverse imposter syndrome. All of these people saying "Unless you're at 10k users+ scale you don't need k8s". If you're running a personal project with a single-digit user count, then sure, but only purely out of a cost-to-performance metric would I say k8s is unreasonable. Any scale larger, however, and I struggle to reconcile this position with the reality that anything with a consistent user base should have zero-downtime deployments, load balancing, etc. Maybe I'm just incredibly OOTL, but when did these simple features to implement and essentially free from a cost standpoint become optional? Perhaps I'm just misunderstanding the argument, and the argument is that you should use a Fly or Vercel-esque platform that provides some of these benefits without needing to configure k8s. Still, the problem with this mindset is that vendor lock-in is a lot harder to correct once a platform is in production and being used consistently without prolonged downtime.
Personally, I would do early builds with Fly and once I saw a consistent userbase I'd switch to k8s for scale, but this is purely due to the cost of a minimal k8s instance (especially on GKE or EKS). This, in essence, allows scaling from ~0 to ~1M+ with the only bottleneck being DB scaling (if you're using a single DB like CloudSQL).
Still, I wish I could reconcile my personal disconnect with the majority of people here who regard k8s as overly complicated and unnecessary. Are there really that many shops out there who consider the advantages of k8s above them or are they just achieving the same result in a different manner?
One could certainly learn enough k8s in a weekend to deploy a simple cluster. Now I'm not recommending this for someone's company's production instance, due to the foot guns if improperly configured, but the argument of k8s being too complicated to learn seems unfounded.
I've been in your shoes for quite a long time. By now I've accepted that a lot of folks on HN and other similar forums simply don't know / care about the issue that Kubernetes resolves, or that someone else in their company takes care of those for them
k8s makes it easier to build over engineered architectures for applications that don’t need that level of complexity
So while you are correct that it is not actually that difficult to learn and implement K8S it’s also almost always completely unnecessary even at the largest scale
given that you can do the largest scale stuff without it and you should do most small scale stuff without it, the number of people for whom all of the risks and costs balancr out is much smaller than the amount that it has been promoted and pushed
And given the fact that orchestration layers are a critical part of infrastructure, handing over or changing the data environment relationship in a multilayer computing environment to such an extent is a non-trivial one-way door
You really need to learn about the burden of proof, since you'd (hopefully) be able to figure out that in this case, the burden of proof is on you to prove what you're saying is accurate. HN isn't a place for hand waving arguments where your only supporting proof is "logic".
The truth of the premises do not require empirical evidence as you seem to require, they require the premises be true.
- Oncology is a for-profit business
- Curing customers reduces profits
- Treating customers increases profits
- Curing cancer is bad for business
Do you understand how logic works or is your brain fried by "where's the study bro..." Give me a break, man. Do you have any interest in sincerity or is your job at stake otherwise?
While I mostly agree with your point of Apple being rather aggressive with forced upgrading, I don't think the device requirements for these features were based solely on the desire to push out people with older devices, but rather due to the hardware requirements for a lot of the ML/AI features being based on the Apple Silicon, at least for the Mac side of things. As to why they drew the line at the iPhone 15, perhaps it's a similar reason regarding performance requirements. While obviously, I'm not intimately knowledgeable of their basis for the device requirements, I'd wait a few more years to see how the device requirements for these new features cascade. If they continue requiring newer and newer devices, only supporting the trailing generation or so, then I'd agree wholeheartedly with your sentiment.
I'm with you here. As a proud owner of an iPhone 13 Mini, I refuse to switch to anything bigger than that, but I do concede that any moderately useful AI pipeline will require more power than my aging phone is capable to provide.
I'll always slightly regret not getting a Mini, but 2020 was a really bad year for it to launch (when I didn't feel like the extra work needed to see one in person) and 2021 I actually needed a better camera.
In retrospect though, it may be best that I don't know what I missed.
Yes, it's not arbitrary at all — they're only offering it on devices with at least 8GB of memory.
The iPhone 15 Pros were the first iPhones with 8GB. All M1+ Macs/iPads have at least 8GB of ram.
LLMs are very memory hungry, so frankly I'm a little surprised they support such low memory requirements (especially knowing that the system is running other tasks, not just ML). Microsoft's Copilot+ program has a 16GB minimum.
If a would-be founder doesn't want to start a company because of their salary, then it's probably for the better. An early-stage startup isn't a vehicle to funnel VC money into your account at the rate that multi-trillion dollar companies give it out.
Moreover, one thing NIH syndrome handwavers miss is that developing a custom situation-specific solution allows you to expand your knowledge in that discipline, create a more tailored solution, and avoid supply chain vulnerabilities that can come with using third-party packages. In my experience a handbuilt solution to a problem is going to be much more efficient in most cases than OSS out there, except in spot instances where the scale of the task is not able to be achieved in a small amount of code (e.g. a versatile graphing library), although these are /very/ few and far between.
My horror story of Ticketmaster; I recently bought standing-room-only tickets on short notice (<1wk) for an event near me, declining the additional fee to be able to refund my tickets. After more discussion with the others I was going with, I bought seated tickets instead through SeatGeek. Understanding I declined the ability to refund, I attempted to sell my tickets, but their system kept encountering an internal error preventing me from selling the tickets.
I reached out to support for assistance, and after several days of wasted time and run-around, they finally sent my issue to their engineering team saying they'd get back to me in 5 business days. Keep in mind I said I bought these tickets a week before the event, and they'd already wasted a few days giving me the run-around, functionally meaning I wouldn't be able to sell my tickets.
I attempted to charge back the purchase since they did not provide what I paid for (tickets I could sell), and they fought me and won somehow.
So thanks Ticketmaster, for sucking me out of hundreds of dollars for nothing more than bytes in your database that I couldn't do anything with. I hope they go bankrupt.
For anyone who is in my shoes and hasn't used Ticketmaster yet and might be tempted to give them a chance thinking all of these horror stories are just unlucky people- don't. I was naive to think that all of those companies with bad reputations are just the loud minority but Ticketmaster is the only one I've had the misfortune of finding out is seriously awful. Use SeatGeek or countless other platforms instead. Gun to my head to use Ticketmaster again I'd probably take the lead instead.
> I attempted to charge back the purchase since they did not provide what I paid for (tickets I could sell), and they fought me and won somehow.
This has happened to me twice now (though not with TicketMaster) and I was 100% in the right, and I lost. When I mentioned it on HN I was met with a lot of doubters. I think something has really changed regarding chargebacks.
Subjectively, I've seen a lot more conversation on the internet in the last few years about people using chargebacks, often in contexts where it's obvious to me as an outside observer that they're abusing the system by doing chargebacks compulsively without even trying to resolve things with the merchant.
It wouldn't surprise me if we're seeing a tragedy of the commons: chargebacks were easy as long as people were inculturated to use them as a last resort. Now that enough people reach for them first, banks have to look at each one more closely and they're going to err in the direction that takes less work.
100% chargebacks have changed in the last 2-3 years. I had a vendor send me the wrong part and refuse a refund. Even showing that they sent the wrong part despite ordering the correct part, my chargeback was denied.
I think a lot of banks have gotten weary of chargeback scams and taking the brunt of Amazon's binning practices.
Frankly, I'd think it better if they just cut off those bad retailers from the system, which is where the failing is. Alas, monopolies in -that- sector as far as I know prevent a single bank from doing a whole lot, especially when it's a vendor that does so much volume that all the legitimate chargebacks won't risk their standing with the payment processors.
Amazon has a lot of flaws, but I've never once had an issue returning an item for a full refund. I'm sure chargebacks are up in recent years, but I'm not sure it's Amazon that's to blame.
i think this is a reference to amazon's co-mingling of inventory regardless of where it came from, which results in businesses selling legitimate items having customers that get sent counterfeit goods.
Chargebacks are only as aggressive as the bank's customers are willing to enforce by leaving / suing against the vendor's level of customer expectation of service. Outside of a vocal minority, no one is going to want a card that doesn't work on amazon.
The playing field has been rapidly shrinking, and the customer base is much more stressed and unwilling to fight.
Not to mention that's also roughly around the timeframe that binding arbitration really got pervasive.
I'm not sure why. In many cases the merchant is charged a chargeback fee regardless of whether they are in the right or not. The bank gets paid either way.
I think a lot of people have been abusing chargebacks (e.g. "I didn't like the item, so I chargedback rather than returning it") and they clamped down on it, and it affects us normal people moreso.
They most definitely have some BS in the fine print about how they're not responsible for their awful system, as CYA for things like this. Truly scum of the earth.
If I had to guess, it is probably in their fine print, and the ability to pay for a refund would be a further refutation in a chargeback case.
That the ticket could not be sold via their system for whatever reason, is not a 'simple' act, although TBH maybe they should write to the DOJ or whatever... given some of the other stuff they've been caught doing, it would not at all surprise me to see some `if (!ticket.HadRefundOption) throw` hidden in their sales system.
TBH OP (Not a lawyer, not legal advice) you could always try small claims, they might not even show up and then you can collect a default judgement
Isn't this where the hilarious "sheriff showed up at the office, graciously giving them 30 minutes to cut a check before he started to confiscate the chairs" stories come from?
> how they're not responsible for their awful system,
Yea, this is something that has to be protected by consumer rights laws. Otherwise companies will be like "It's unfortunate we have a monopoly, but fuck off and give us your money. Thank You. Your case has been closed".
So you didn’t purchase ticket insurance, you got the tickets you paid for (which they can allow you to sell, at their discretion) and you filed a chargeback… why is that Ticketmaster’s problem? Like it sucks, to be sure, and Ticketmaster is awful, but I’m not sure why that chargeback would be considered legit.
If the tickets were sold as marketable (it sounds like they were) but were not in fact marketable, that's a problem.
If they were sold as marketable pending function(situation) with the implication that function(situation) was not simply "return false" but it turns out that function(situation) was actually "return false," that too is a problem.
Time and potentially some filing/service costs. You may be able to claim some of those as well (When I almost had to sue for a security deposit, in that Jurisdiction I could get some filing fees but not service costs for whatever reason...)
That said, if they don't show up, you'll get a default judgement. And if TM doesn't pay, they can have fun with it if there is an office nearby. A while back someone got a judgement against a bank, they didn't pay out. He came by with the sheriff and they started loading up chairs/etc when they hesitated to cut a check. :)
Or, whatever other 'collection' action you may have to motion for after the fact if they don't pay.
At least where I live (Texas, United States) from what I saw it's allowed to have legal counsel, although it may be uncommon. I'll have to look into the process more and see if there's anything I can pursue.
I vaguely recall in california you can opt-out of arbitration within 30 days of a contract. don't know if there are details or if that is still the case.
> Legal counsel isn't allowed in small claims courts.
That's only true in a handful of states. Most allow you to bring a lawyer.
Small claims courts will generally have simpler and friendlier procedures so that even if a lawyer is allowed you will be fine without one in most cases.
Even after reading the article I'm a bit confused as to how $250mm is even enough to play ball in this industry, a single EUV lithography machine can cost more than their entire funding already. I'd be curious to know their plan to work around the significant investment required to even enter this market, as I doubt $250mm is really enough to develop anything rivaling the results of Nvidia or others.
> there aren't enough boxing fans in the US to fill a single NFL stadium, much less 16 a week.
Do you have any proof for this at all? The most anticipated boxing matches usually get over a million pay-per-view buys, not including the tickets sold for the actual arenas.
And the Superbowl gets over 100M viewers -- a two-orders-of-magnitude difference.
[edit]
Yes, there are literally more than 80,250 (MetLife stadium's capacity) fans of boxing in the US. However e.g. the Kingdom Arena (location of Fury v. Ngannaou) seats about 30,000. If there were a 17 week boxing match season in the US, it clearly would be a much smaller draw than gridiron football, and that doesn't even include the college game, which has 8 stadiums that seat over 100k (and 6 more that are larger than MetLife).
PPV Sales can't be compared to Super Bowl viewership numbers, considering there's a pretty significant difference between watching something included with your existing TV subscription versus paying another $50+ to watch a one-night event. I'm not trying to argue that boxing matches are necessarily on the same level of viewership as the Super Bowl, but you can't compare them directly like that. I'm also not arguing that boxing could sell out NFL stadiums over multi-week spans, I'm directly arguing the point that "there aren't enough boxing fans in the US to fill an entire NFL stadium", which is completely false.
> There is nothing stopping us from doing the same thing.
However achieving property rights for the land needed for the project is a much bigger issue in the US than in China, where I doubt they receive much dissidence from people affected due to their infrastructure improvements.
I mean there are more maladies that affect the colon and rectum than just colon cancer, so I expect colo-rectal surgeons will survive, if with a decline in positions.
Plus not everyone responds to treatment. Allergies, side effects. They may slowly age out but we will probably always see a few in the nearest large city.
Every time I see one of these posts and the ensuing comments I always get a little bit of inverse imposter syndrome. All of these people saying "Unless you're at 10k users+ scale you don't need k8s". If you're running a personal project with a single-digit user count, then sure, but only purely out of a cost-to-performance metric would I say k8s is unreasonable. Any scale larger, however, and I struggle to reconcile this position with the reality that anything with a consistent user base should have zero-downtime deployments, load balancing, etc. Maybe I'm just incredibly OOTL, but when did these simple features to implement and essentially free from a cost standpoint become optional? Perhaps I'm just misunderstanding the argument, and the argument is that you should use a Fly or Vercel-esque platform that provides some of these benefits without needing to configure k8s. Still, the problem with this mindset is that vendor lock-in is a lot harder to correct once a platform is in production and being used consistently without prolonged downtime.
Personally, I would do early builds with Fly and once I saw a consistent userbase I'd switch to k8s for scale, but this is purely due to the cost of a minimal k8s instance (especially on GKE or EKS). This, in essence, allows scaling from ~0 to ~1M+ with the only bottleneck being DB scaling (if you're using a single DB like CloudSQL).
Still, I wish I could reconcile my personal disconnect with the majority of people here who regard k8s as overly complicated and unnecessary. Are there really that many shops out there who consider the advantages of k8s above them or are they just achieving the same result in a different manner?
One could certainly learn enough k8s in a weekend to deploy a simple cluster. Now I'm not recommending this for someone's company's production instance, due to the foot guns if improperly configured, but the argument of k8s being too complicated to learn seems unfounded.
/rant