HNHacker News
TopNewBestAskShowJobs

kitsuac

81 karma · joined November 16, 2016

submissionscomments
kitsuac··on I've been writing ring buffers wrong all these years
It's odd that you were using ring buffers in 1992 for low level code but don't understand the value of avoiding a modulus instruction. Masking is far more efficient and often a ring buffer will be used in code where performance is absolutely critical.
kitsuac··on Bill Gates et. al. launch massive clean-energy fund to fight climate change
The only thing which matters when considering "The People" is their aggregate "generalization" -- contrasted with these uber wealthy where a handful of them can make an actual positive impact on climate change where the people failed. Sure, plenty of people are gung ho about making a difference, but if their sum doesn't add up to actual change then that's the aggregate reality.
kitsuac··on Bill Gates et. al. launch massive clean-energy fund to fight climate change
There are big minds without power/wealth and also far sighted people without. You pretty much need both in order to do something significant which doesn't align with the overall will of the general public.
kitsuac··on Bill Gates et. al. launch massive clean-energy fund to fight climate change
But that's a reflection of "The People" and their will. People want their existing jobs and status quo. They care about next week's pay check. Not a theoretical tidal wave that people they don't know say will happen if they don't stop working on fossil fuels and switch to being a cashier at Wendy's. To separate that reality of will from democracy would make it no longer democratic.
kitsuac··on Bill Gates et. al. launch massive clean-energy fund to fight climate change
Everyone gets it wrong much of the time, seems like the question may be more about who gets it wrong a little less often.
kitsuac··on Bill Gates et. al. launch massive clean-energy fund to fight climate change
But it isn't the role of the representatives to come to terms with science, it's their role to come to terms with the will of "The People" (the tangled mess of what that means). I think it could be argued that the will of "The People" isn't to enact strong action to fight climate change. It's more to retain jobs in old energy industries and procrastinate until something absolutely dramatic and visceral impacts their day-to-day.

I'd suppose the Achille's heel of "The People" is their short-sight. You need big minds with big wallets to see and act on the long game.

kitsuac··on Bill Gates et. al. launch massive clean-energy fund to fight climate change
It would have some interesting implications, politically. The democratic sense of "The People" rejected the issue and failed to take care of it through democratic and mundane day-to-day decisions. But the uber-wealthy who are often considered absolute enemies of "The People" had to step up and save the entire planet.
kitsuac··on Fast Fourier transform in x86 assembly
I'm not sure to what extent exactly, but iirc they instrument APIs and pass them through IPC. It may not really be the best example, for some reason it popped into my head but maybe it was an incorrect assumption.
kitsuac··on Fast Fourier transform in x86 assembly
I work with video codecs now, but anything computationally intensive and/or close to the hardware will benefit from some assembly code. The three that you mentioned are good examples. The web browser you're using to view and respond to this thread is another example. Among other things, the JavaScript engine most likely does dynamic recompilation and it may use sandboxed webviews.
kitsuac··on Fast Fourier transform in x86 assembly
I program assembly as a large part of my daily job, and we generally target between 500% and 1500% speedup for our assembly routines. That number being compared to C code with all compiler optimisations enabled. 1.0x - 2.4x is indeed pretty awful, but it does sometimes happen. For example, many algorithms are fundamentally serial in nature (e.g. entropy decoding).
kitsuac··on A fast multi-producer, multi-consumer lock-free concurrent queue for C++11
Not true at all - the core atomic operations are very well understood by now and are supported on any modern CPU.
kitsuac··on A fast multi-producer, multi-consumer lock-free concurrent queue for C++11
Cooperative multi-tasking doesn't always mean a per-worker queue is feasible. Really the only time I could see that being feasible is if the work queue is constantly saturated. Often times you just don't know in advance which worker to queue up with which task.
kitsuac··on Learning to Read X86 Assembly Language
If you want to learn how to read X86, don't choose this AT&T style syntax as it's a monstrosity.
kitsuac··on I made an iPhone game with PhoneGap and won't do it again
Being a person who programs actual ARM/Neon assembly for both iOS and Android, your post is hilarious. ObjectiveC and Java are anything but assembly.
kitsuac··on Constexpr-8cc: Compile-time C Compiler
If it's too difficult for you to use, then use another language which holds your hand more. But for many of us, it's an incredibly powerful language and there isn't really anything comparable when you need to squeeze every instruction cycle out of the CPU while still using HLL features. I don't find it difficult at all to use. Sure it's complicated, but not everything worth doing is trivially simple.
← PreviousPage 2 of 2