HNHacker News
TopNewBestAskShowJobs

tgb

5,680 karma · joined September 9, 2012

submissionscomments
tgb··on Google shifting to “upstream first” Linux kernel approach for Android features
Even Google misspells it: https://blog.xkcd.com/2010/05/03/color-survey-results/

I checked the other day and it's still true.

tgb··on A mathematician's guided tour through higher dimensions
It's often repeated but describing spheres as "spikey" isn't right at all. The argument made is that they seem spikey in rectangular coordinates. But that's actually a statement about cubes, which are spikey. That's easy to see even just considering 2 and 3 dimensions : the corner of a cube is better for stabbing than a side. But the sphere is rotationally symmetric in all dimensions and is not spikey in the least bit.
tgb··on Unmanned submarine earmarked for Irish Sea freight crossings
There are "underwater gliders" that propel themselves by ascending/descending via changing their buoyancy and using fins to direct this forwards. They have low speed but are much more efficient than most subs. Would be interesting to see such a glider take on freight, though it would probably need thrusters for navigation in harbor.
tgb··on Unmanned submarine earmarked for Irish Sea freight crossings
Doppler velocity log is likely what they'd want.
tgb··on Scikit-Learn Version 1.0
Matplotlib is my go-to despite being mediocre. I recently found proplot library built on it which seems to solve a lot of the warts (particularly around figure layout with subplots and legends). I haven't had a chance to use it yet - does anyone know if it's worth it?

I like to stick to basic, widely used tools when possible so I'm biased against it versus just wrangling it out with matplotlib. But proplot does look compelling, like it was written for exactly my complaints.

tgb··on Scientists create matter from pure light, proving the Breit-Wheeler effect
I find your comment confusing. What exactly is the "good example" you're referring to? The person you are replying to appears to have done exactly what you are saying to do, by my reading, so I'm wondering what you're referring to.

'The map is not the territory' is the phrasing I've heard: https://en.m.wikipedia.org/wiki/Map%E2%80%93territory_relati...

tgb··on Eyeway's Retina-Tracking Foveated Display
I was going to say impossible, reasoning from how I understand your eye's lens to focus. But that doesn't rule out the possibility of it working if out-of-focus. If you look at a small point light source and let your eyes go out of focus, the light smears to a larger range. The shape and texturing of that range depend upon the light source: for example search for bokeh shots during a partial solar eclipse. This should be emulatable with a small projector.

However the size up of the image should be no bigger than what you see when you let your eyes go out of focus, which isn't really that much. And you'd have to find a way to make your eyes want to focus like that, which could be tricky.

tgb··on Alternative Milks
Where is the oat milk getting it's extra calories? Using the usual 9kcals per gram fat and 4kcals per gram protein and gram carbohydrates, I get the expected 120ish amount for cow milk but only 71kcal for oat milk (listed at 100kcals). I guess the table is missing non-sugar carbohydrates?
tgb··on The grandmaster diet: How to lose weight while barely moving (2019)
Why do you say that? I mean, I just pointed out it burns me 15000 calories a month and frankly I haven't built any muscle from running.
tgb··on The grandmaster diet: How to lose weight while barely moving (2019)
> Exercising is almost irrelevant

I get where this is coming from, but I really feel like everyone has over-corrected on this and ended up being wrong again in the opposite direction. Exercise burns a significant number of calories. An hour run can easily burn 700 calories for someone my size (~70kg). Running an hour a day 6 days a week burns more than one pound of fat a week. I run less than that amount and my phone estimates I burned 15,000 calories last month from running. That's not a get-slim-quick magic pill, but that's really significant!

Exercising for 20 minutes once or twice a week is pretty negligible and people over-correct by eating more. But a consistent, easily-doable exercise habit is a huge source of additional "calories out".

tgb··on Duolingo reaches $6.5B valuation on day of IPO
Audio is the most important and hardest component of learning a language so apps have a huge advantage over textbooks any day. I really don't think recommending textbooks is at all relevant these days. Graded readers are better for simulating immersion. Apps are better for getting your feet wet. Textbooks can teach grammar but so can apps and really you have to be pretty far along for grammar to be the main stumbling block (over listening, pronunciation, and vocabulary).
tgb··on Amazon has acquired Facebook's satellite internet team
If your original claim were true, then you'd also be saying that there is twice as much area at 2m above sea level as there is 1m above sea level, and infinitely more than at sea level. If you truly believe this, then I will be willing to make you a trade where I give you one tenth floor condo in exchange for 10 first floor condos.
tgb··on Amazon has acquired Facebook's satellite internet team
They're at 100x higher altitude relative to the sea level but nearly the same altitude relative to the Earth's center which is what matters here.
tgb··on The worst volume control UI in the world (2017)
On Android I installed an app that lets me scale the global volume for this. It's called "Precise Volume" and seems to work well. I wish I could tweak the "gamma" of the volume slider though: I want to make quiet side quieter without making the loud end quieter so that it works both with headphones and with other devices.
tgb··on EdgeDB: EdgeQL
I'm a SQL newbie so maybe I lack context for this decision, but why did you decide to keep the "SELECT" keyword? Going through your tutorial, SELECT seems redundant and could just be dropped without any loss of information. In that way look a lot more like most non-SQL languages. Why do I need to SELECT "string" instead of just "string"?
tgb··on The Private Life of a Cat (2020)
Like this one, where they put cameras and GPS on cats? https://www.bbc.com/news/science-environment-22821639

Or this one:

https://www.sciencedirect.com/science/article/abs/pii/S00063...

> Twenty-three percent of cat prey items were returned to households; 49% of items were left at the site of capture, and 28% were consumed.

tgb··on Juul bought an entire issue of a scholarly journal
Your citations don't back up your claim that "the main reason smoking causes cancer" is polonium 210. They just note that polonium 210 is one possible reason for smoking causing cancer. From your [1]:

> Do these doses lead to lung cancer? It’s hard to say, especially since the effects of polonium are only part of a wider range of damaging consequences caused by inhaling cigarette smoke. But animal studies certainly give us cause for concern.

Moreover wikipedia cites two articles saying:

> In contrast, a 1999 review of tobacco smoke carcinogens published in the Journal of the National Cancer Institute states that "levels of polonium-210 in tobacco smoke are not believed to be great enough to significantly impact lung cancer in smokers."[224] In 2011 Hecht has also stated that the "levels of 210Po in cigarette smoke are probably too low to be involved in lung cancer induction".[225]

https://en.wikipedia.org/wiki/Health_effects_of_tobacco#Radi...

tgb··on Software Engineering principles to make teams better
How do you make something idempotent if one of the effects is sending an email?
tgb··on What mRNA is good for, and what it maybe isn’t
You mean the degeneracy in the last base of codons that I mention? I wouldn't even call that an error correcting: would you count a file type that just skipped every third byte 'error correction'? I wouldn't. This isn't a robust system and most changes to any bases will change the protein sequence. Contrast that to DNA which does have error correction, since there are two strands instead of one and also two chromosomes with nearly identical information. Both are used to repair errors, at least some of the time. I'm not aware of such a system for RNA and it's not clear where it would get the information from - though maybe the most likely system would be a 'self-destruct' where errors are detected but since repair is impossible, the RNA is just destroyed.
tgb··on What mRNA is good for, and what it maybe isn’t
What error correction does RNA have? Are you just referring to the third base in the codons? DNA has some but I don't know of any for RNA, and it hardly seems worth it for most organisms to have RNA error-correction: it's better to just degrade the RNA and retranscribe it.
tgb··on Challenges in the diagnosis of magnesium status
I was saying that "90% water" stat is super misleading because it makes it sound like <10% is anything of value, but that's off by a factor of 3. Baking absolutely makes a bigger than 10% water loss by mass in broccoli, and that would of course double the non-water fraction even if it were only 10%.

But the point is, it doesn't matter how much water there is. No one is wondering whether they'll feel more full on 100g of chicken or 100g of broccoli. Anyone trying to control weight would ask about 100 calories of broccoli versus 100g of chicken. I don't know what the answer there is, but they're probably comparable. A head of broccoli is just 205 calories but still contains 17g protein and 16g of fiber. Equivalent calorie chicken breast would be like 40g protein and no fiber. Both are healthy food choices unless you deep fry them or something. You're making lots of fast-and-loose comparisons here that are simply misleading. (Why is the comparison broccoli + fast food versus just chicken? Why do you suddenly switch to talking about iceberg lettuce? Why pretend that broccoli isn't a healthy food that is substantially more satiating than fast food - as is chicken?)

tgb··on Challenges in the diagnosis of magnesium status
> most green vegetables like broccoli are 90% water.

That's super misleading. The nutrition facts for broccoli here [1] gives 89% water but actually over 1/3rd of broccoli by mass is carbs, fats, proteins or fibers. And comparing raw broccoli to cooked chicken is deceptive - cooking broccoli obviously gets rid of a lot of it's water depending on how you cook. Moreover, comparing gram-to-gram, I'm sure broccoli is less filling, but a lot of people care more about calorie-to-calorie.

[1] https://www.healthline.com/nutrition/foods/broccoli#nutrient...

tgb··on Introduction to Locality-Sensitive Hashing
Well written article. Does this get used at all bioinformatics, for maybe de novo sequence construction from short reads or something? Seems like it would have many applications in that field.

Edit: I see now that user inciampati already addressed this.

tgb··on Scientist finds early virus sequences that had been mysteriously deleted
If SRA is like the closely related GEO database, then you can upload data without it being public yet. You frequently do so prior to publication (but give the reviewers special access) and then make it public later. So it may never have been publicly accessible.
tgb··on Your CPU may have slowed down on Wednesday
Sure, but you don't need to mitigate timing attacks on untrusted code if that code already has root access.
tgb··on Wasmer 2.0
> NaNs (Not a Number, usually referring to positive/negative Infinite values in floats and integers)

Huh? This sounds totally incorrect to me.

tgb··on The most copied StackOverflow snippet of all time is flawed (2019)
I iterate over rows in pandas fairly often for plotting purposes. Anytime I want to draw something more complicated than a single point for each row, I find it's simple and straight-forward to just iterrows() and call the appropriate matplotlib functions for each. It does mean some plots that are conceptually pretty simple end up taking ~5 seconds to draw, but I don't mind. Is there really a better alternative that isn't super complicated? Keep in mind that I frequently change my mind about what I'm plotting, so simple code is really good (it's usually easier to modify) even if it's a little slower.
tgb··on Time to sunburn
The FDA's article gives a meaninglessly useless. The multiplicative one is clear. What is the plausible case where someone gets confused by this in a way that's harmful? Would anyone ever over-estimate the time they can spend in the sun? I can only imagine under-estimating that time (which is fine) since our base reference for sunburns is always to picture the hottest day of summer at the beach at middday - the worst case scenario. The sun setting makes this better not worse!

Either way the multiplicative one is clear: 30 minutes of midday exposure with 30 SPF is approximately equal to 1 minute of midday exposure and so is minimal. That's helpful to know! And 50 is only moderately better than 30, but 30 is a lot better than 10 - the FDA page says nothing about this.

tgb··on Time to sunburn
This model doesn't take into account time of day, but many do. But those mostly seem to do so incorrectly, as far as I can tell. They tend to use the measure of "UVA/B incident on a horizontal square meter". However humans are not flat horizontal surfaces - and in fact a lot of my exposed skin is more vertical than horizontal. Therefore I think the correct measure should be "UVA/B incident on a meter-diameter sphere" or similar (where absorption from the atmosphere is the only contributor, not merely foreshortening). I was unable to identify numbers for this at all when I looked into it - does anyone know of a model that does this properly? Or just number of how much UVA/B is impacted by this?
tgb··on Time to sunburn
I don't see anything in that article that actually contradicts the definition of SPF. It's basically saying:

> It's a common misconception that speed determines how fast you get to your destination, but actually distance also matters. So speed does not reflect time to destination. Instead, speed lets consumers compare the level of relative quickness and that 60mph is quicker than 6mph.

Obviously we need to hold other factors constant. It's not a useful point in when talking about speed and I don't see the point in talking about SPF. And it certainly doesn't mean that time to destination is inversely proportional to speed.

← PreviousPage 4 of 34Next →