HNHacker News
TopNewBestAskShowJobs

pca006132

634 karma · joined February 7, 2021

submissionscomments
pca006132··on I'd rather read the prompt
There are many issues here. The lack of incentives is probably the most important one. For new professors (in research universities), good teaching is usually just a good thing to have, but it is not a deciding factor for their tenure. When they get their tenure, they probably have enough students, and they need to work hard to apply for funding and keep the students paid. Administrators care most about ranking, and teaching isn't really evaluated in the ranking. They just push the professors to do more research and apply for more funding.

It is also hard to evaluate university teaching because there are no benchmarks for that (compared with high school, for example), and it is hard to judge if teaching is good from student feedback. You can only know if someone fucked up or did really well, which are outliers.

There are other issues as well. Professor IMO is a ridiculous job, you are supposed to be an expert in the field, be a researcher, be a manager, be a teacher, be a salesman, all at the same time. There are people who can excel in all these, but these are probably just outliers. It doesn't help when PhD training doesn't train you to be a proper manager and teacher. While there are some teaching training, I think we are not really held to a high enough standard. E.g. One can pass the teaching course if they just show up and spend some time, even though their teaching is horrible.

pca006132··on What went wrong with wireless USB
There are websites using WebRTC for p2p transfer.
pca006132··on Only Teslas exempt from new auto tariffs thanks to 85% domestic content rule
It probably also means that companies will not do it: It seems impossible to keep the tariff this high for years, and Trump will only stay in office for a few years...
pca006132··on Vacheron Constantin breaks the world record for most complicated wristwatch
Sadly we don't consider things that are economically efficient as craftmanship :)
pca006132··on Google Is Winning on Every AI Front
What if the models are somehow trained/tuned with Ads? Like businesses sponsor the training of some foundational models... Not the typical ads business model, but may be possible.
pca006132··on Has the decline of knowledge work begun?
I feel like top management and share holders don't really care about things that may happen ~30 years later. And even if they care, there are a lot more other issues that may ruin a company, e.g. the entire industry is replaced due to technological advancement.
pca006132··on I want a good parallel computer
While financial incentives is important to some, a lot of people write books to share their knowledge and give the book out for free. I think more people are doing this now, and there are also open collaborative textbook projects.

And I personally think that it is weird to write books during your working hour, and also get monet from selling that book.

pca006132··on France rejects backdoor mandate
The issue is that something like this will not affect the criminals, they can just use something with encryption.
pca006132··on FOSS infrastructure is under attack by AI companies
There are off-chain solutions to handle most of the payment, and only put a summary on-chain. I think there are already micropayments in Brave or something.
pca006132··on Sorting algorithms with CUDA
1. Parallel radix sort (CPU or GPU) can also be pretty fast. For input distributions without too much pattern and keys that are not too large, radix sort is probably the fastest if you can use it.

2. For typical applications, memory transfer speed matters more than sorting performance on the GPU. If most of your work is done on the CPU, transfering the memory to the GPU may take more time than sorting the array. Not sure if unified memory (apple M series chips and AMD new APU) can remedy this though.

pca006132··on The DOJ still wants Google to sell off Chrome
I would argue that subscription is the only sensible model for browser (if payment is needed). For browsers you do want to keep up to date with security patches as well as new features, that is pretty critical for both usability and security.

For normal software you can buy and use some fixed version without too much issues, for browsers you probably need the updates.

Security-wise maybe something like buying a ESR version may be possible, but it will not be cheap and you may miss the new features when some webpage you visit breaks.

pca006132··on uBlock Origin is no longer available on the Chrome Store
And I don't understand what is the benefit of lying as well - everyone on the internet knows what this is about, at least if they used ad blockers. A lot of people don't, but they will not be affected anyway.
pca006132··on An Attempt to Catch Up with JIT Compilers
When things are performance-sensitive, you want things to be tunable and predictable. Good luck playing with the JIT if you rely that for performance...
pca006132··on An update on Mozilla's terms of use for Firefox
I'm just curious if it is possible that some former/current employees in Mozilla can just form an org, say they will maintain a fork of Firefox, and accept donation from the users that were pissed and maybe apply for some funding from EU NGI?
pca006132··on Hyperspace
I thought reflink is provided by the underlying FS, and Hyperspace is a dedup tool that finds the duplicates.
pca006132··on Hyperspace
Is this the dedup function provided by other FS?
pca006132··on Hyperspace
CoW is not aliasing. It will perform the actual copying when you modify the file content.
pca006132··on How Core Git Developers Configure Git
Wow I never knew this, I always just type `--no-ext-diff` when piping the output. Thanks!
pca006132··on Thailand to Cut Power to Myanmar Scam Hubs
Tbh I feel like the people didn't care much about it because the people who got kidnapped there were mostly attracted by job offers that were known to be shady (quick money), so there wasn't much discussion around this. The actor getting kidnapped is different: People suddenly found that you can be kidnapped even when the offer seems legit and only staying in Thailand. This forced the government (both China and Thailand) to take action to address the concerns.

But I think this will come back again. This kidnapping thing is not new, and it caught media attention a few years ago as well, they just go silent for a period of time and continue their operation later.

pca006132··on More than 40% of postdocs leave academia, study reveals
Or maybe try not to judge researchers based on the amount of grants that they can get...
pca006132··on More than 40% of postdocs leave academia, study reveals
I feel terrible for the idea of jugding academics based on the amount of grant money they can get... It feels like encouraging a lot of smart people to find ways to waste money, even when they know that they don't really need that much for their project.
pca006132··on More than 40% of postdocs leave academia, study reveals
Meh, not actively messing up the course is already better than some of the other professors...
pca006132··on More than 40% of postdocs leave academia, study reveals
I sometimes just wonder, a lot of professors are bad at teaching because they don't have to be good at it. Is it the same for universities? It feels like reputation for universities are quite detached from the courses they have or teaching quality. Rankings focus a lot on quantitative measures, but teaching quality is hard to measure quantitatively. The output of universities, i.e. the quality of their students, depend on both the teaching quality and "IQ" of their students before admission, which is mostly a feedback loop because universities with good reputation get the best students... Optimizing for teaching quality also means that professors spend more time on teaching and less on research, which may reduce their research output and reduce the ranking, which has a more immediate effect on the reputation than teaching quality.
pca006132··on More than 40% of postdocs leave academia, study reveals
It depends on incentives. From what I can see in CS, a lot of young researchers are focused on short-term projects, disconnected from actual problems, and spend lots of effort to package the result to increase the chance of getting a paper in top conferences/journals, because they need this for their career. They will be forced to leave academia if they don't have enough results in time, from what I know. And even for established researchers, they have to do something similar, so their students can have enough results. And they need to try really hard to get funding, because institutions want researchers that can get lots of funding, so institutions can get money from that.

This is probably not the complete story, and probably a bit too pessimistic, but I think this is true...

pca006132··on Dear sir, you have built a compiler (2022)
This depends on languages. For c/c++ without a package manager, people hesitate to add external dependencies (and especially when high performance is needed, e.g. game engines).
pca006132··on Do AI detectors work? Students face false cheating accusations
Yeah, LLM is kind of just making expensive cheats cheaper. You can do it without LLM, and indeed students did similar things prior to the release of ChatGPT, just less common.
pca006132··on PEP 760: No more bare excepts
No. A serious language designer will try to make things they like at the beginning, not trying to patch it later. Patching via breaking public API (language design) is never a good thing.
pca006132··on tolower() with AVX-512
RISC-V has SIMD extension as well. Even when there is no SIMD, prefetching or instruction selection/scheduling will have a big impact on the performance, so it is unlikely one can easily write a few lines of assembly and get to a similar level of performance.
pca006132··on CrowdStrike Update: Windows Bluescreen and Boot Loops
Another way to look at it is that you can have more hospitals using systems with a lower cost, thus saving more lifes comparing to only a few hospitals using an expensive system.
pca006132··on Disruptor-rs: better latency and throughput than crossbeam
Probably doesn't make sense. Busy wait is fast when you can dedicate a core to the task, but this means that you cannot have many tasks in parallel with a small set of physical cores. When you oversubscribe, performance will quickly degrade.

Tokio and other libraries such as pthread allows thread to wait for something and wake up that particular thread when the event occurs. This is what allows scheduler to schedule many tasks to a very small set of cores without running useless instructions checking for status.

For foundational library, I think you want things that are composable, and low latency stuff are not that composable IMO.

Not saying that they are bad, but low latency is something that requires global effort in your system, and using such library without being aware of these limitations will likely cause more harm than good.

← PreviousPage 2 of 12Next →