ec109685

18 ms·

241.
by ec109685 2y ago
Did WWDC end up filling in any additional blanks from your article? Thanks!
242.
by ec109685 2y ago
> If you do apply, there is a whole process to exclude you from the position. You'll be too qualified or under-qualified or your salary expectations won't match the advertisement. Or they'll find some other reason to strik
243.
by ec109685 2y ago
If provisioned concurrency is used, then fetching the database at startup would be a valid strategy given you could still achieve flat response times.
244.
by ec109685 2y ago
Surprised they don’t have a slower rollout across multiple regions / kubernetes clusters, given the K8s API’s are a spof as shown here where a change brought the control plane down. Also, stale-if-error is a far safer pattern for servi
245.
by ec109685 2y ago
The author of SQLite commented here about accepting contributions: “They have a really high bar”, but are accepted, occasionally: https://news.ycombinator.com/item?id=34480732
246.
by ec109685 2y ago
Every simultaneous request is in its own microVM, making concurrency a non-issue / thing to optimize in most cases at the function level.
247.
by ec109685 2y ago
They aren’t closed for contribution. From the author: “They have a really high bar”, but are accepted, occasionally: https://news.ycombinator.com/item?id=34480732
248.
by ec109685 2y ago
They could license the test suite from SQLite (and a lot of tests are open sourced): https://www.sqlite.org/prosupport.html#th3
249.
by ec109685 2y ago
Are lambdas guaranteed to be warm on the first request to serve in single digit milliseconds?
250.
by ec109685 2y ago
GPT trademark was rejected: https://www.saul.com/insights/alert/what-gpt-and-who-owns-it
251.
by ec109685 2y ago
This is why I always introduce myself as David Mayer and share my love of counting r’s in strawberries.
252.
by ec109685 2y ago
But it confidently implied they are the mayor, which would make you look stupid if you sent the letter. That is the Achilles heal of these models :(
253.
by ec109685 2y ago
Apple trained their LLM on Google TPU’s: https://www.theregister.com/2024/07/30/apple_google_tpu_ai/
254.
by ec109685 2y ago
Why would running the model with Firefox crash his Mac? At worst, it should start swapping: > The first time I tried it consumed every remaining bit of available memory and hard-crashed my Mac Also, hallucinations so easily dismissed the
255.
by ec109685 2y ago
There was so much noise about autonomous vehicles in SF from first responders and other officials. There was a segment that dismissed this as histrionics about self-driving cars. As soon as Cruise closed up shop, the complaints went way dow
256.
by ec109685 2y ago
Microsoft is their path into the enterprise. You can use their so-so enterprise support directly or have all the enterprise features you could want via Azure. They also are still leading in the enterprise space: https://www.linke
257.
by ec109685 2y ago
Agreed, feels like virtue signaling.
258.
by ec109685 2y ago
The Malazan response is below the deck response.
259.
by ec109685 2y ago
Given we have engines that can render complex 3d worlds, can maintain consistency far longer than a minute and simulate physics accurately, why put all that burden on a GenAI world generator like this? It seems like it’d be more useful to h
260.
by ec109685 2y ago
The way YC closed ranks was my favorite anecdote from Chaos Monkey’s. I always find it interesting though YC leaders are extremely against cancel culture, but when it comes to business, its okay to threaten that someone will never work in t
261.
by ec109685 2y ago
If the nvidia driver has a bug, can one workload access data of another running on the physical machine? E.g. it came up in this thread: https://news.ycombinator.com/item?id=41672168
262.
by ec109685 2y ago
Escape hatch feels like the lambda the author talks about. Ejecting is abandoning the generator and owning the code from that point on completely.
263.
by ec109685 2y ago
Most of their app is a mismatch of web views and native with spinners galore. It doesn’t seem like performance is a high priority for them anymore. E-commerce is fundamentally small data once the user lands on a product page, so every subse
264.
by ec109685 2y ago
Employers want to pay the minimum, clearly, but until a person’s salary exceeds the value they bring to a firm, there will be other firms willing to pay more and attract that talent. So provides some upward pressure on wages, which the auth
265.
by ec109685 2y ago
You might be interested in this nice write up: https://shopify.engineering/how-we-built-shopifys-bfcm-2023-...
266.
by ec109685 2y ago
Total curiosity, but what’s the limiting factor of scaling out to multiple regions day one.
267.
by ec109685 2y ago
It async replicates to s3, while providing a consistent interface to storage clients.
268.
by ec109685 2y ago
On the server, applications are generally super close to their backend dependencies, so multiple round trips to data stores, caches and whatnot are no problem. On the client, this would be deadly. So it’s not just easy to take code that run
269.
by ec109685 2y ago
Apache where I used to work was configured to restart each child after 1000 requests. So many outages prevented at very little cost.
270.
by ec109685 2y ago
Serverless means you don’t care about how many hosts you’re spread across.

More ›