srvmshr

I came to know when Jeff Dean tweeted about it in September [1]. Truly a great loss.

[1]. https://twitter.com/JeffDean/status/1705042899677884630

I genuinely don't want to sound pedantic or critical - but anyone who dabbled in maths - from olympiads in school or undergrad or did math for research - would have several such notes for e.g. of graph theoretic models or combinatorics doodles. I used to color mine, or had different dashes/bold schemes for visualization of my problem space.

I think blog author came across this & found it beautiful (good for him!). But IRL this type of doodling is commonplace among math folks.

From their viewpoint, most mathematician's I've met, admire someone's blackboard skills - handwriting, problem layout, & Hagomoro chalks. And the second one would be the witty/intelligent communication in their papers, book or seminal work - which sometimes gain cult status in their small circles.

I still use two fully specced out 15" MacBook Pro Mid-2015. Battery lasts me close to 3 hours still on a regular work day.

My previous 2010 MacBook pro similarly lasted 6 years (until about 2016 when I got two of these machines).

Apple used to have a different league of build quality back then. It all started going down in late 2016.

PolyCam is apparently not free but considered one of the best. I used 3D Scanner from Laan labs (& it's a free app. No ads thankfully). I used it to build a few dataset for NeRFs based on MIPNeRF. The data quality seemed pretty fine. I use a iPad Pro with M1 btw.

https://apps.apple.com/us/app/3d-scanner-app/id1419913995

Ever observed how legal documents, product documentation, books etc., are typeset - with exact placement of comment boxes, US code reference/legal disclaimers/ barcode, precise footnotes or margin notes. Many of those are meant to be also machine readable when printed out. Exactness is a very pressing need.

I would recommend taking a good look again. It might answer you why it is preferred in some situation to be typeset in PDF over a format where text could reflow.

About the immutability in Word, it seems optional & not something by design. You can edit any *docx & 'Save it as' back. This feature doesn't absolve immutability as a principal feature.

They have become a reliable way in office & legal processes around the world in terms of fixed layouts & content immutability (in a sort of layman view. Knowing Acrobat Pro exists & PDF editing too - but I'd argue that in majority of cases its not as trivial as modifying text file or markup sources, with intention to change or forge).

Correct me if wrong but Word/LibreOffice layouts could change depending on the machine & version number - but with PDF you get what you intended to show. I think that has always been the winning proposition for PDF

> Very hard for little Timmy to ask a quick question via email these days.

There was a time in HS when I had several to-fro with DMR about Unix philosophy without fully realizing what a giant he was in our CS galaxy. I truly miss him.

I think the view is a bit myopic. Certainly, government projects have inertia & cost overrun. But over the last six decades, here are some of the initiatives on my mind, which couldn't have been done without government investment in technology:

* Internet, as many people pointed out already.

* Nuclear energy & commitments to nuclear projects e.g. nuclear medicine, nuclear chemistry, medical research, particle physics & major instrumentation grants

* Space technology. Ability to understand the cosmos, origins and the unanswered questions about the eventual fate of our world & mankind. Including some flattering moments such as Apollo program.

* Meteorology & mapping: This is usually the underdog. We owe a lot to meteorological studies and USGS. Humanity (not just the US) has managed to stave off famines, several natural disasters or be prepared to face the next simply because how well we have mapped earth and it's resources by remote sensing technologies, many of which are government developed or sponsored in large.

I think what is rather true is government projects are endurance races. They aren't as flashy as SpaceX or ChatGPT, but without them the next glitzy tech will be arduous to get. As with all races, one can't win everything & the slow nature of it makes us question its worth sometimes.

Although Soviet-era books were notoriously terse & difficult to digest, they had some very aesthetic typesetting. I have owned a few (Problems in Physics by Irodov, & another by Krotov) and they all share similar design aesthetics.

For the author, he misses an important point that except sometimes (and rare case) for a few core libs and functionality, majority zero-days and zero-clicks are based on new features and their integration into current OS iteration. Software revisions begets new security events. Patching software bugs/loophoples is not like treating cancer - something that has existed always and keeps popping up. It is new software added to old stack and made to gell with it. That 'gluing' process churns up new security issues.

If the phone is physically okay and you depend on few core functionalities, then it is perfectly okay to keep using it for majority of (non-critical) tasks. Most bugs in the system & features of old iOS are limited to that old OS anyway - and most likely addressed. If some advanced utility are going to be involved e.g. work communications, some security protocols, I'll perhaps work with a device which still gets critical updates at the least.

I have a iPhone 8 which I use just to Facetime my MIL and receive her iMessages (I am on Pixels since 2019). I can't foresee a old patched- often iOS with a older no-frills Facetime version to have a major risks. For these tasks, I don't see it necessary to get a new iphone.

Mostly a ton of JavaScript. As for the concepts explained visually, ICYMI there is a small footnote at the end of the article:

> To generate the 50D word embeddings we used the GloVe 6B 50D pre-trained model and converted to Word2Vec format. To generate the 2D representation of word embeddings we used the BERT large language model and reduced dimensionality using UMAP. The self-attention values and the probability scores in the beam search section are conceptual.

I am slightly concerned people would still find more insidious ways to discriminate, under the guise of other reason(s). And conversely, people looking for protection would on occasion, weaponise this law.

(I have some first hand experience being burnt by the caste factor, when it was genuinely not intended or thought from my side. But perf review comments were reported as caste discrimination. No idea why would I remotely be interested to do that)

Visited Dubai couple of years ago. Nice place to visit for a few days if you remain oblivious to how its society functions; but taking up residence will feel like living in a shopping mall always. Also, the general idea of labour abuse conveyed by this linked article[1] is mostly true. Some excerpts

> This is all supposed to be illegal. Employers are meant to pay on time, never take your passport, give you breaks in the heat – but I met nobody who said it happens. Not one. These men are conned into coming and trapped into staying, with the complicity of the Dubai authorities

> There's a huge number of suicides in the camps and on the construction sites, but they're not reported. They're described as 'accidents'. Even then, their families aren't free: they simply inherit the debts.

I was made aware of such labor traps meted out to Philippine and Thai women as well, who were forced to work in beauty & massage parlors, or as house maids. I was accompanying my parents a decade ago, who were rapporteurs on behalf of UNHRC. In many places, we were physically stopped from proceeding. The situation I am told has improved a tiny bit, but still remains grim. When you look at Dubai's manpower resources - you aren't looking at a industry, but more of a cartel working at a nation-scale.

[1] https://www.independent.co.uk/voices/commentators/johann-har...

Sorry about late reply. I wish there was some way to get notifications of comments. Here's a survey paper which connects TDA & Graphics in several ways:

https://arxiv.org/abs/2212.09703

I tried my level best and got 2/20. I repeated with random guesses and got 5/20. I have used AWS on-and-off since 2018.

AWS icons are next to Egyptian hieroglyphs in difficulty. Google iconography isn't so bad (but bad enough still.)

From an earlier post, copying my answer which seems relevant to your query

> Being a Japanese resident (PR) now, I did look up this option for a summer home in the woods (Akita, Aichi-ken etc). This is what I learnt:

* Akiya houses come with an agreement to reside. Not just be a summer home. Their goal is to repopulate the countryside.

* You have to cultivate any farmland that comes with the deed. You cannot sell it without special permission. There is a whole lot of paperwork to deal with such situations.

* Most old Akiya will have strong regulation not to change the frontage significantly. It isn't permitted. Architecturally these changes have to be consistent with the older house and rest of the community. Building permits are very stringent. (Basically, you can't build a sprawling European villa in a Japanese countryside for e.g., even if land wasn't a concern.)

* Land tax is levied the day from purchase - not from the day of moving in. (This could pinch given that time to renovate can be 1-2 years in remote places, and taxes aren't cheap necessarily)

* Residents have to contribute to local development funds which take care of Matsuris etc. You like it or not, local government will knock on your door with a bill.

* Connectivity is poor. Cell reception outside NTT can be spotty. Internet is even harder.

* Language proficiency is a must. No one speaks English. Not even the local government officials.

* House will need a significant amount of renovation. With the stringent restrictions - anywhere between 100 & 200 grand (depending on the disrepair)

For those interested, there is a really nice course with lecture notes taught by Vidit Nanda (Oxford).

https://people.maths.ox.ac.uk/nanda/cat/

Unlike what people commented elsewhere, TDA has growing application in computer graphics, 3D reconstructions and computer vision.

PS: I am told the new version of this course is narrated/illustrated by Robert Ghrist. He's famous for his foundational calculus course

Just a benign observation: The list starts with brutal and graphic, going progressively benign to harmless ones. Shock effect expected.

I have a ML PhD - maybe slightly qualified to answer this question.

Yes, absolutely doable. Immerse yourself into learning things well. Learn the basics. Don't do course hopping & book hopping - pick a rock solid book or lecture series, devour it & get building. The last part is the most important part.

People with ML MS/PhD only have an additional (1) degree & (2) networking. If you invest time, you can overcome (2) by asking good questions in Twitter/Reddit & making connections. I still do it after finished my degree. Twitter is the Linkedin for ML.

As for (1), YSK that most advisees are getting advised by professors who made it big before deep learning took off. So everyone is still on the learning curve of sorts - advisors, advisees and your peers. Sometimes student's intuitions could be better than the professor's. Don't sweat over it. Focus on building.

> Americans are the world’s leading refund seekers; consumers in Japan seldom return anything.

Lived in America. Permanent resident of Japan. Imagine my culture shock.

(Full disclosure/explanation: I was more shocked by the return culture, than the no-return culture here. More used to carefully choosing & keeping the item until it falls apart entirely. We are also very particular about non-burnables & e-waste. I am S Asian but with more ties to Japan & hence settled here. My culture shocks are a whole different story & worthy of a book chapter. American return culture, tipping & fanaticism for football/baseball team is unique.)

I deliberated for a long time to choose between Remarkable 2 and the conventional iPad - and in the end just stuck to the iPad (after trying a borrowed Re-2 for a few days)

* The reading quality & feel to the hands is far superior in Remarkable 2 and it looks much more affordable to entry level iPad, but at the end of the day it is constrained to be just a reading/annotating device.

* For people like me who prefer Android as day to day driver, but occasionally require use of FaceTime or iMessage to connect with others (friends abroad or in group), iPad is very convenient.

* Apple Pencil objectively feels more polished and versatile as a stylus for variety of tasks - note taking, sketching, minor photo touch-up, evaluating paper & reviews etc. The nibs are replaceable as well.

* The software ecosystem around Remarkable 2 still has enough rough edges. It is not that smooth to sync, transfer or update the device. iCloud+GoodNotes integration wins by large margin.

* iPad (and the Pro more so) has a camera which comes handy at scanning and annotating documents quickly (loaded directly to Goodnotes) versus going through scanning and transferring that to Remarkable 2

That is what struck me too after reading the article.

With the digital (and maybe Bigdata/ML) age, copyright protections need some revisions & broadened scope of "fair use". But digitizing and making media items available over internet face the arguments that piracy sites get accused of. IA is a respectable agency, and these media is from 50s & 60s. But the optics of it from legality remains the same unfortunately. I wouldn't also believe these having no commercial value. Among the collection, I found popular artists such as Louis Armstrong - who I still listen to over Spotify/Apple.

IA would probably cut some losses if they disable all streaming or only provide 30sec clips (as preview), while keeping the music archive more restricted. Not that I like this proposition, but if we are serious about preserving the legacy of 78rpm, some more consideration could be useful. Copyright laws in their current form is (unfortunately) unforgiving to IAs efforts in this context.

Elaborating a tiny bit on #2: It is a dangerous situation when non-technical leadership comes up with very specific technical opinions - based on hearsay & evangelists on their Twitter/Instagram.

"I saw X on his Reel telling changing everything in our frontend to purely Coffescript will be more durable" is a kind of advice which does not bode well generally

> Bautista Corona said city workers have collected 6,000 tons of trash from storm drains over the last three years through a program that specifically targets big items like couches and refrigerators.

Good griefs. That is a new one for me. I would not expect to see couches & refrigerators dumped into manholes anywhere.

Same sentiment heard elsewhere. Pulling stick up in a stall midflight is a rookie mistake. Pulling up & increasing thrust is a takeoff peculiarity. Even if AF447's altimeter was faulty, they both missed something else in their panic - they fell 20,000ft in few minutes, going faster & faster. The g-forces are palpable enough to experienced pilot (who reky on it sometimes when they fly in complete darkness).

This incident was rookie flying behavior coupled with a complete disregard of situational awareness. The flight UX of Airbus isn't that bad: 330 & 340 have a much better panel layout & less clutter than their Boeing counterparts from late 90s.

Bonin in his panic hit his priority button to take back control but didn't say he was taking back control. He should have announced he has the controls - and it is obvious knowledge to pilots that flight behavior us from the averaged stick inputs. But only when they know both are handling flight surfaces.

If a pilot stubbornly wants to keep control maybe out of desperation or panic - and performs actions which are counterfactual to the knowledge of aerodynamics every pilot is taught, there is no instrument system which can salvage this catastrophe.

["Bonin acted like an idiot" - not my words, but two of my CFI friends, who discuss aviation topics on weekends often over beers.]

I won't like to criticize someone with a different trade skill (much less when they're already dead), but to paraphrase two of my CFI friends, their summary view was "Bonin acted stupidly". He didn't communicate he had controls, kept pulling the stick to nose up. The pilot psyche element had lot to play in this accident. He was sticking to a naive understanding of aerodynamics in panic - which was counterfactual - rather than counting on prior training experience (like what we developers would say an antipattern). Relaying a similar point, his training still wasn't enough to gather situational awareness & apply common-sense instead of a panicky behavior.

> Researchers have found that humans quickly stop remarking on hot temperatures; after between two to eight years, what was once a record high starts to feel like the new normal

Al Gore sensitized his layman audience of this effect in his invited talks (c. early 2000s), as slow boiling the toad. It is also a part of the "An Inconvenient Truth" documentary (2006), which is quite a relevant discourse even today. If you haven't watched it, I recommend you should - it is quite accessible, & well explained considering it was made almost 2 decades ago.

For me Eudora was the first client I installed. They had a unicorn logo & even came as a Lite edition. Fun times those were, in early 2000s (along with ICQ, ACDSee32, & Quake3)

https://computerhistory.org/blog/the-eudora-email-client-sou...

Kitaboshi Pencil Factory is within cycling distance from my home in Katsushika ward. This news item makes me so happy! You don't see things from your neighborhood make to the trending HN section everyday :)