byroot

28 ms·

301.
by byroot 5y ago
That was solved years ago by moving the refcounts into different pages: https://instagram-engineering.com/copy-on-write-friendly-pyt...
302.
by byroot 5y ago
> People follow it closely, because these speedups might be a fluke, and as the compiler becomes more compliant, they may disappear (that happened in past with some JITs). They're not. YJIT is really 100% compatible with regular int
303.
by byroot 5y ago
> Who cares if they do more than 150 FPS I don't, but you said "almost C like".
304.
by byroot 5y ago
That's TruffleRuby, not JRuby. And even then, I haven't benchmarked myself, but I wouldn't be surprised most NES emulators written in C or other native languages out there would do way more than 150FPS. Edit: and it doesn&#x2
305.
by byroot 5y ago
Without knowing against which benchmark these numbers were measured make them pretty much meaningless. YJIT achieve a 700%+ speedup on some micro-benchmarks [0] too (see `30k_ifelse` and `30k_methods`). But that's nowhere near the gain
306.
by byroot 5y ago
The last French nuclear test was in January 1996, so rather unlikely.
307.
by byroot 5y ago
> It takes 11 hours to take the train. How's that even possible? Looking at GMaps, there's roughly 650km of tracks between the two, so that's an average of 60km/h. Even regular speed trains in Europe easily average 10
308.
by byroot 5y ago
There are plenty of cases where exceeding the requirement is cheaper / easier / simpler than to try to match it more closely because you have some common and standard, hence cheaper, parts available. So IMHO over and under enginee
309.
by byroot 5y ago
> since we obviously can't get anything just perfectly engineered it's gonna be over engineered. A bridge will take more load than it's rated load, an elevator will take more load than it's rated load etc. etc. That&#
310.
by byroot 5y ago
> was alliance for practical purposes as seen with joint "victory parades" Alliance would mean joint fighting, not joint parading. Yes they secretly divided Europe on a map, and agreed not to fight over it, that not an alliance
311.
by byroot 5y ago
> They also allied with them As your own link states, this wasn't an alliance. Also if you read the "background" section of the article, you'll see that they very much saw Nazi Germany has a major threat and wanted to
312.
by byroot 5y ago
The source is me: https://gist.github.com/casperisfine/1c46f05cccfa945cd156f44... Note that this overhead apply as soon as a refinement is defined for a method regardless of wether it's ever active. That 40% figur
313.
by byroot 5y ago
> Seems worth reporting if not already done. I assumed it is expected overhead.
314.
by byroot 5y ago
> A shame that refinements don't really get much love Their performance is quite awful on MRI. Any method that was refined is tagged as such, which incur a 40% performance hit (for empty methods)[0]. That alone tend to disqualify th
315.
by byroot 5y ago
It's not "front" and "back" it's "left side of the boat" and "right side of the boat".
316.
by byroot 5y ago
That's why I said to some extent. Sorbet usage varies between teams and apps, it's not enforced across the org.
317.
by byroot 5y ago
Yep, totally agree with you, I was merely correcting a factoïd, and sorry for the very terse response, I was on the go. And even the "better tooling" part is somewhat subjective. The added strictness and explicitness is valued by
318.
by byroot 5y ago
Shopify did adopt sorbet to some extent.
319.
by byroot 5y ago
> France gets something like 75% of its energy from nuclear. That's not quite correct. We do get 75% of our electricity from nuclear, but electricity is only ~25% of the energy consumed in France. So in practice about 18% of our ene
320.
by byroot 5y ago
Even a one second cache would cap your homepage to 1 req/s which can be cheaply and trivially handled by any stack.
321.
by byroot 5y ago
That link is not responding for me, but here's a mirror: https://github.com/thorvalds/linux/commit/1da177e4c3f41524e8...
322.
by byroot 6y ago
It's not always a rational choice. Among other things, there's some social status associated to flying compared to taking a train. So the professionals tend to take the plane even when it's neither cheaper, nor faster, nor mo
323.
by byroot 6y ago
It's always good to hedge your bets. The two represent the common dichotomy between a reimplementation that potentially reach much higher performance, but will continuously have to battle for compatibility, and an in-tree development t
324.
by byroot 6y ago
I don't claim it's a long term solution, and that you should run of your cache for years. But having that cache means you have the time to get a clean solution rather than having to rush something because nobody can work anymore c
325.
by byroot 6y ago
Then the 0.3.6 version was released, and we could update because the license isn't much of a problem in this case. Alternatively we could have continuee to deploy for a few more days until Rails core shipped a new version of Active Sto
326.
by byroot 6y ago
Or you know... just cache them. If your CI or deploys broke because of this, it basically means you're constantly re-installing all your dependencies from scratch, which is totally silly. Configure your CI & other tools to cache th
327.
by byroot 6y ago
According to "Le Canard Enchainé" [0] this is because: - They agreed not to make profits on the vaccines for the first year or so (that was imposed by Oxford)[1] - There's no penalty for not reaching the delivery target
328.
by byroot 6y ago
If called on a Function it returns the source of the function, which allows for all sorts of meta programming.
329.
by byroot 6y ago
The old.reddit.com UI shows the last edit date: https://old.reddit.com/r/wallstreetbets/comments/ip6jnv/the_... This post was last edited 4 months ago.
330.
by byroot 6y ago
None at this stage. Chrome did add some instrumentation a few versions back to collect data and decide wether it's worth implementing. Problem is: who's going to bother sending early hints if it doesn't improve performance?

More ›