solatic

19 ms·

241.
by solatic 11mo ago
You cannot divorce a product from the people who built it. The product reflects their priorities and internal group well-being. A different group of people would have built a different product.
242.
by solatic 11mo ago
Sure, but if you did a good job, the gradual deployment can go relatively quickly and smoothly, which is how $FAANG roll out new features and products to very large audiences. The actual rollout is usually a bit of an implementation detail
243.
by solatic 11mo ago
India's UPI (digital payments) is almost as big a scale as it gets, and it's pretty universally considered a success: https://en.wikipedia.org/wiki/Unified_Payments_Interface
244.
by solatic 11mo ago
That's what works for products , not software systems. Gradual growth inevitably results in loads of technical debt that is not paid off as Product adds more feature requests to deliver larger and larger sales contracts. Eventually yo
245.
by solatic 11mo ago
Shell is the best language for data science. Pick the best tools for each of getting data, cleaning data, transforming data, and visualizing data, then stitch them together by sheer virtue of the fact that text is the universal interoperabl
246.
by solatic 11mo ago
This. I may not work with AI training workflows, but I struggle to understand why they supposedly require launching a thousand pods per second to use GPUs that need to fundamentally be installed across different baremetal machines. Once t
247.
by solatic 11mo ago
> The C-level folks seem to think they are buying some kind of indemnity with these "enterprise" grade solutions, but there is no such thing. Something you need to understand about executives, is that they're not really in
248.
by solatic 11mo ago
Auth providers are among the hardest systems to secure. It's not just a question of the underlying code having vulnerabilities - for companies with Internet logins, auth systems (a) are exposed to the internet, (b) are not cache-friend
249.
by solatic 11mo ago
AMP is an internal proprietary fork of Cortex, they're not up-streaming their changes, also in large part due to the scalability limits of Cortex's design. It has the same scalability limitation I described earlier with the lack o
250.
by solatic 11mo ago
> combine traces from frontend and backend I'm scratching my head a little bit on what your expectation is here. Traces and real-user-monitoring are not the same thing here. Distributed tracing is specifically a microservices thing.
251.
by solatic 11mo ago
There were other problems with VictoriaMetrics - a failed migration attempt by previous engineers made it politically difficult to raise as a possibility, lack of a promise of full PromQL compatibility (too many PromQL dashboards built by t
252.
by solatic 11mo ago
I'm personally not convinced OpenTelemetry is the future. I get the desire to not be vendor-locked to a single provider, but Prometheus and Jaeger are very solid, battle-hardened, popular, well-maintained, easily self-hosted open-sourc
253.
by solatic 11mo ago
The project where I looked at Mimir was a 500+ million timeseries project, with the desire to support scaling to the ten-figure level of timeseries (working for a BigCo supporting hundreds of product development teams). All of these systems
254.
by solatic 11mo ago
Mimir is just architected for a totally different order of magnitude of metrics. At that scale, yeah, Kafka is actually necessary. There are no other open-source solutions offering the same scalability, period. That's besides the point
255.
by solatic 11mo ago
> Go's performance advantages are basically irrelevant in an AI agent, as execution time is dominated by inference time Inference time is only the bottleneck if you are running a single agent loop, for a single consumer, with a sing
256.
by solatic 11mo ago
It's exactly why taking a trip through the ops/infra side is so important for people - you learn why LTS-style engineering is so important. You learn to pick technologies that are stable, reliable, well-supported by a large-enough
257.
by solatic 11mo ago
Ninety-degree angles do not exist in nature. So you're going to get two schools of thought: UI should look more natural (and therefore round off any hard edges), or UI should intentionally embrace hard edges, as a declaration of defian
258.
by solatic 11mo ago
Most people in this thread, it seems, just want a simple way to manage Kubernetes manifests, something that keeps track of different settings for different environments and what's in common for each environment in order to generate the
259.
by solatic 11mo ago
100%. Three decision-making styles are dictators (make decisions without feedback), arbitrators (listen to feedback, then the arbitrator makes the call), and mediators (mediators guide discussion until consensus is reached). Arbitrators are
260.
by solatic 11mo ago
> searchable archive Have fun with that. At the end of the day you run into the same problem as most wikis, where people just "threw stuff over the wall" into a big wiki database, and it gets to be really difficult to find rel
261.
by solatic 11mo ago
> top DSLR / digital medium format But these cost several thousand dollars, due to them still being relevant for contemporary professional applications. Medium format bodies can be found for hundreds of dollars on eBay. Even at $1&#
262.
by solatic 11mo ago
I don't feel like you really countered my argument. If you're concerned about cost/shot, digital beats 35mm, no question. > I shoot film because it makes me slow down and think a bit. 100%, because of the higher cost per s
263.
by solatic 11mo ago
Film doesn't really make sense anymore outside the realm of luxury-budget art. And if so, why 35mm? Why not medium format / Hasselblads, or large format? Why not actually go for the format sizes where there's a reasonable arg
264.
by solatic 11mo ago
Precisely. People don't need to carry wallets anymore. No cash, no physical plastic credit card. It's little surprise that the purse will trend smaller as people need to carry less and less inside it. Apple is hitting a revenue gr
265.
by solatic 11mo ago
A model, no, but an agent with a calculator tool? Then there's the question of why not just build the calculator tool into the model?
266.
by solatic 11mo ago
VAT is horrible. It's a regressive tax on consumption, so it's doubly economically inefficient and is the worst offender for reducing quality of life for the most people. Reductions in VAT rates are always populist, but government
267.
by solatic 11mo ago
> Society changes It's rather disingenuous to claim, in context, that society has changed to enjoy advertisements. Indeed, I would make the argument that society will never , ever , get to the place where we enjoy advertising. A
268.
by solatic 11mo ago
Fully agree with you, and people who don't understand this should follow through to read the actual link. Marsf explained that sumobot deleted the volunteers' translations, zero coordination, and Kiki's response is "I&#x
269.
by solatic 11mo ago
I also got a bad impression from that, but upon further reflection, it's a more interesting choice. Keeping the plug port in the trunk means that the charging cable (kept in the trunk) can stay semi-permanently plugged into the car, wh
270.
by solatic 11mo ago
Reliable systems require hard limits imposed by designers. When systems hit the hard limits, it's a sign somebody's assumptions are wrong: either the designer built too small, or there's some bug pushing up against the hard l

More ›