The author discovers the wild world of financial engineering and "float".
Instead of settling usage changes for it's image service with a single charge, Cloudflare settles it with two charges in such a way that the customer "floats" Cloudflare money for some portion of their billing cycle.
This is the bullshit that you start to do when you become publicly traded. It adds zero value for the customer.
If Figma continues along this path I worry that they are doomed.
Generally, there are three paths to profitability: 1) increase pricing, 2) decrease expenses, or 3) offer new products and services with superior margin.
It seems like Figma has chosen to implement 1) but sell it to customers as 3). However, in contractionary periods like we are in now, 1) does not work -- no matter how much lock-in you have and how you sell it.
Weirdly, they don't preference contracts for shell corps owned by, say, poor white trash from Middletown, Ohio -- despite these people meeting your definition of "someone who historically hasn't been paid well".
But I digress. In the biz, if the contract has to be trafficked through a shell corp where the owner takes a 20% cut, everyone simply bids a 20% higher rate.
It's the taxpayer's prerogative if they would rather their government pay an inflated contracting rate -- as opposed to using that money for education, infrastructure, social services, etc.
From subpoenaed communications we even know that certain figures who publicly denounced the lab-leak theory, privately believed the theory was plausible or even likely.
This article is part of what I like to call the narrative media. You decide what the facts are and don't give a damn about reality -- because that's what keeps your readers coming back for more.
What a wonderful way to cut headcount/expense and lock-in profitable margins on healthy annual revenue.
Can only work when you have the advantage of being the dominant product in the marketplace -- but I gotta hand it to the board, I couldn't have done it better myself.
So here's a little brain teaser about what you have to do when dealing with potential nation-state actors. This scenario is for the folks who are calling "hyperbole" when the actor is clearly, potentially a nation-state. This scenario is based upon an event that actually occurred.
1. You have a $200 million piece of defense-critical equipment.
2. You know that there was a 5-minute period where a potential member of a foreign intelligence service was alone and unattended in the same room as this piece of equipment.
What do you do with the equipment? You can:
a) Put the equipment into service
b) Disassemble the equipment on both a hardware and software level and try to detect if anything was altered
c) Destroy the equipment
If you choose anything other than c) you have probably never been, nor should you ever be, in charge of securing critical assets that can be targeted by a nation-state. This incident seems to indicate that the leadership at Microsoft would choose a).
Also, bear in mind that these are the people that you just sent all your ChatGPT data to.
1. While JCH will usually be the most performant hashing method, naively, removing a node will affect all nodes of higher order. This makes the logic of node deletions somewhat more complex than (say) Discord's hash ring. This is why JCH is more common for long-term, distributed, redundant storage -- where the topology changes far less frequently.
2. For sharding, what makes distribution hard is not so much the hashing but consensus on the cluster state -- this is the hidden problem. Bryan Hunter's talk on Waterpark (https://youtu.be/9qUfX3XFi_4) is a excellent example of what you can do when you can set things up so that the topology is fixed. In fact, this approach makes things so straight forward that it is shared by Riak, where the number of vnodes is fixed.
However, if you have a rapidly changing topology (like several Kubernetes clusters that are frequently scaling up and down), you can often need some sort of consensus mechanism to make sure every node has a consistent view of the cluster. In my experience, this usually ends up being the most complex part of distribution problem to solve.