One, you don't need this. The vast majority of people working on the web are now so thoroughly overserved by their frameworks, especially the way that benchmarks like this measured only the minimal overhead the frameworks could impose, that measuring your framework on how many nanoseconds per request it consumes (I think time per request is a more sensible measure than request per time) is quintessential premature optimization. All consulting a table like this does for the vast majority of people is pessimize their framework choices by slanting them in the direction of taking speed over features when in fact they are better served by taking features over speed.
Two, you are performance bound, in which case, these benchmarks still don't help very much, because you really just have to stub out your performance and run benchmarks yourself, because you need to holistically analyze the performance of your framework, with your database, with any other APIs or libraries you use, to know what is going to be the globally best solution. Granted, not starting with a framework that struggles to attain 100 requests per second can help, but if you're in this position and you can't identify that sort of thing within minutes of scanning their documentation you're boned anyhow. They're not really that common anymore.
This sort of benchmark ranges from "just barely positive" value to a significant hazard of being substantially negative if you aren't very, very careful how you use the information.
Framework qua framework choice doesn't matter much anymore. It's dominated by so, so many other considerations, as long as you don't take the real stinkers.
Tl;dr: 89 fresh cadavers used since 2017 (cost ~860k$) for “hands-on training on non-perfused and perfused cadaver bodies” to simulate battlefield injuries.
Seems like “noble” (?) life-saving trainings to me. I personally wouldn’t oppose to my body being used for that borderline medical use-case.
I’m super interested in this topic of “how to keep a community alive”, (1) introducing new blood - with the risk mentioned above - at a regular pace, and (2) keeping existing members as alive/active/connected as possible.
Would anyone here have good reads linked to that topic, I’m interested to share - writing this, I just realize I should’ve asked my friend Copilot too.
One interesting framework linked to this I found a couple years ago and still refer too frequently is the Orbit model (https://orbit-model.vercel.app)
You want everyone to gain a little more than they give up, then your logistics bring people into interaction with your value proposition and the community grows.
French here, working/living in France, and not aware of any state-backed “shorter work week” test.
According to my research, only SME applied it so far (with 2 exceptions of 1000 and 11k-employees).
And no particular consensus on the results (many companies stayed on their 4d agenda after the initial phases, but again, we’re talking SME here).
> How exactly do you ensure that you only hire the best person for the job?
I understood Scale’s stance as: « out of the candidates we have in front of us for any opened job position, we’ll select the one providing the best immediate value on their concrete field - as opposed to the one offering less of a direct outcome but more « indirect » value, such as helping make our workplace a diverse, intercultural and safe place. »
I’d bet this article answers very concrete decisions they had to take internally recently and the article’s writer simply decided to turn his opinion into a company « value ».
The exact opposite stance would’ve been possible (as in « we’re an intercultural, diverse place to work and we try to make the world a better and safe place for everybody, one hire at a time »), and companies usually mix a bit of the 2.
Don’t assume evil intentions or a big « socio-political » plan. That just looks like a company leader trying to make an opinionated decision public, so anyone working there could subscribe. Better in my opinion than nothing.
The MEI (as opposed to DEI) acronym is however unfortunate as it can easily raise unneeded binary conflicts, as seen on this forum ;)
Assuming that framework is true - which I’m not challenging at all -, why would you need to _fire_ the builders, instead of pivoting them to new greenfields projects - as I assume Tesla has many.
Unless your CEO isn’t a maintainer/optimizer himself?
Based on my experience (electric motorbike), the regenerative breaks are super strong indeed.
You can tweak their « force », but usually, the strongest level is the most comfortable - that you eventually keep all day long.
It will cause you to full stop even on the steepest slopes.
Once you’re used to it, you dose your deceleration by focusing on how much you release the throttle.
And the only situations where you have to hit the breaks are the unexpected events - e.g., a car coming at the last moment and which you should give priority to.
I don’t think those company goals should be looked at face value.
What is interesting to watch is the dynamics (what was there last year and isn’t there anymore), read between the lines (what is not mentioned as a goal) and the ordering (what is left at the end, what is more important than what).
You can’t expect such an exercise to be treated with transparency (#5. Lay off 20% more of those guys) as there’s just too much to lose by doing so.
On the other hand, those goals are subtle but clear enough for the top managers of the company to inspire themselves and define their own - more concrete - yearly ambitions.
And the lower you go in the org, the more concrete the goals (assuming there are any at all).
No yearly goal is worse to me than unclear ones, as it gives red flags of management not giving a shit and cruising.
reply