dragandj

> What's your source for the claim that a typical Roman citizen is white? That's a pretty bold statement. Much bolder than the relatively non-controversial statement that Rome had a diverse population.

We're circling back to the definition of "white". How's that a bold statement? Today's Great Britain has very diverse population, too, including lots of people from Africa, and yet I wouldn't think that a statement that a typical Brit is white is controversial. Typical Brit is not black, although there are many black British people, of course. And even if we constrain discussion to the most narrow definition of "white" (say, 18th century Englishmen) let's not forget that English people are the result of centuries of mixing of Britons, Germans, Celts... So, white Britons are themselves of diverse ancestry, but all these diverse ancestral people are "white". Let me repeat that race is an artificial construct, but if I'm pressed to say if someone is "white" or "not quite white" I can only go by outside appearance. Someone in Europe can have some genes from one Zimbabwean ancestor, but if he "looks white" to me, I can only say he's "white".

> The problem is simply with depicting all of Rome as (essentially) British.

Yes. That is a TV series. The point of TV series (in my opinion) is not to be 100% accurate, it's to entertain and (somewhat) educate. And the people in TV series are not real ancient Romans, but current actors. The series is for British audience, so it's normal that it's filled with British actors. If the series was produced in Bulgaria, the senators would be played by Bulgarian actors. Both are not exactly of the same color spectrum as ancient Romans, but are close enough to pass as Romans to their viewers. If the series was produced in China, and Romans were played by Chinese actors, that would probably pass as normal to Chinese audience. Maybe Zimbabwean actors could pass as Romans, at least in sub-Saharan Africa; why not? I'm not sure why people insist so much that British actors are bad representation of Romans. They seem quite OK to me (I'm not British).

I don't know whether Lybians consider themselves white or not. The Lybians that I've seen in person (maybe a dozen) didn't appear much darker than my European compatriots. Yes, I can spot the difference, but it was more in facial features than in darkness of their skin color, which was fairly "whitish".

WRT the Lybian Roman emperor: no one claims that ALL Roman citizen were white, or Italic, or European. Any such large Empire was heterogenous. But it is safe to say that a typical Roman citizen, espicially a Senator, would have been white for any reasonable definition of "whiteness", especially in a TV series. There was one black US president (his mother was white and father black if I remember correctly), but a typical US president so far was white.

I don't have anything against hiring a bunch of Chinese or Zimbabwean actors to play Senators, but that would certainly look odd. In the same way as hiring a bunch of Ukrainians to play Zulu warriors would look odd.

He was from Lybia, though. How dark is that? A funny anecdote from a book of my friend who backpacked Africa 15 years ago: A (dark skinned) Sudanese man was referring to some other Sudanese people from his village as "they are black". He himself also appeared black to my friend, but the man didn't consider himself black. It appears that (many?) people from North Africa are quite precise in differences of their spectrum of skin tone :)

Although Persians are on the edge of the Roman Empire, so I don't think this is related to this discussion too much, it depends on what you consider "white". When I google "iranians" and go to images, I see people that, if I must classify them into a "race", I can most closely classify them as white(ish).

Here's an interesting (USA) article that illustrates this issue: https://www.latimes.com/projects/la-me-census-middle-east-no...

Are people from Teheran, for example, "white, black, Asian, American Indian, or Native Hawaiian" (copy-pasted the choices from the article)? In my opinion, American obsession with race is silly.

OK, what is "white"? Define it. I don't know about USA, but from european perspective, Armenians are as white as any other European nation. What is USA term for that? Caucasian? Ironically, Armenians are as close to THE Caucasus region as anyone can be :) Additionally, the few Armenians, that I know in person, would consider and insult to be called Persian (for political reasons I think).

Yes, they were ethnically diverse. That does not mean that Roman senators were any other than a bunch of white male people, and that the majority of its citizens weren't what anyone from Europe today would consider "white" or "caucasian". Of course, lots of people with even totally dark skin lived in Rome and other regions, but not to the extent of "Roman senators have to be depicted as black".

Ancient Rome and Persia were rival empires. How many Persians (other than usual small number of traders, mercenaries, slaves, etc.) lived in Rome? Even Persians are (largely) of Indo-European ancestry, migrating from the same steppes that filled Europe. Modern people of Iran consider themselves white, although they do typically have slightly darker skin than someone from Italy. It all boil down to what you consider white. If only Swedish people are white, then OK. If Hungarians are white (they typically are) then people of Iran are pretty much white, too. Anyway, race in humans is a pretty much artificial thing, but it is real in our culture nevertheless.

At the time of ancient Rome, Turkic tribes lived somewhere around Mongolia. Gauls, Celts, Goths, and even Greeks were as white as any typical European is today, if not more white.

> His most important research, undertaken in the last years of his life, involved travelling to remote areas of Yugoslavia to make recordings of local singers, whose improvised songs offered clues about how the Homeric epics might have been performed millennia earlier. These recordings revolutionized the understanding of oral literature, but when Parry died no one had yet listened to them; they were just a pile of thirty-five hundred aluminum disks sitting in a Harvard storage room.

This is exactly what I think every time I read about whether Homer was one person or a bunch of anonymous bards. In Serbia and surrounding countries, there is a long tradition of folk epics going back 500-1000 years[0], that is the cornerstone of national identity, which was finally collected and written in early-mid 19th century (mainly) by Vuk Karadzic[1]. This is the area that is being mentioned here (ex Yugoslavia). One of the most well known bards from that time is Filip Visnjic[2], who was blind too. It is difficult to say when these songs were actually composed, but since the milestone of events is 1389, and some songs describe earlier events, it is safe to say that they evolved through centuries until they were finally written down. BTW these bards were singing older songs, but they were also adding heavily their own material...

[0] https://en.wikipedia.org/wiki/Serbian_epic_poetry [1] https://en.wikipedia.org/wiki/Vuk_Karad%C5%BEi%C4%87 [2] https://en.wikipedia.org/wiki/Filip_Vi%C5%A1nji%C4%87

This is relevant to HN mainly because Nubank strongly relies on Clojure and bought Cognitect (Clojure founder's company) last year.

Clojure & Lisp FTW!

General ideas (such as most of repl.it stuff) isn't covered by IP. I doubt that the (former) employee copied exact code in this case. It seems to me that he re-implemented some vaguely similar functionality. Moreover, most of that stuff existed before repl.it...

Why? Assuming that there isn't a non-compete signed by the former employee, the (former) employee has the right to use all their skills and knowledge to do whatever they want on the free market.

I wonder if they'd be abandoned if companies using them considered these 3rd party libraries valuable enough to contribute/fund the development.

It was the Phoenicians who did this, while Egyptians ordered and paid for it.

Let me chip in with some self-promotion.

This book explains and executes every single line of code interactively, from low level operations to high-level networks that do everything automatically. The code is built on the state of the art performance operations of oneDNN (Intel, CPU) and cuDNN (CUDA, GPU). Very concise readable and understandable by humans.

https://aiprobook.com/deep-learning-for-programmers/

Here's the open source library built throughout the book:

https://github.com/uncomplicate/deep-diamond

Some chapters from the beginning of the book are available on my blog, as a tutorial series:

https://dragan.rocks

How does CUDA support any of these (.Net, Java, etc?). It's the first time I hear this claim. There are 3rd party wrappers in Java, .Net, etc. that call CUDA's C++ API, and that's all. Equivalet APIs exist for OpenCL too...

I don’t know why everyone here is talking about dollars, while there is zero dollars involved all the time. The author did not put in any dollars, but his ETH, which he has plenty, since he created the Ethereum thingy. That may be a large part of the answer why no one “exploited” that hole: most people would have to put real dollars in, expose themselves to all kinds of crypto gymnastics, and end up with ETH, which may or may not be convertible to real dollars where they live. And, yes, they would also have the always pleasant exposure to tax authority now (which may or may not be solvable, depending on the place of Earth one happens to live at).

But, the sentence says “what is now Venice”. It’s Venice now, ant it was Venice then. It’s been Venice for the past 1000 years. I guess that it wasn’t “Venice, Italy”, but “Venice, The Venetian Republic” then, but I ascribe this oddity to there being “Paris, Texas” for the USA audience while the rest of the world knows that when you say Paris, you mean the capital of France and not “Paris, the nail salon down the road”. Likewise, 500 years ago, it wasn’t “Venice, the Venetian Republic” since anyone who would understand what Venice or The Venetian Republic is, would not need any qualifying of Venezia, or whatever local name they used for that city.

Anyway, the OP referred to the weirdest place in the sentence for that qualification because nothing else was qualified (was Alaska even Alaska back then? ;)

> The Company can issue more stock, diluting all shareholders, and sell to the short-sellers.

It seems to me that most of WSB crowd forget this kind of things.

IDK but another sane option would be that the regulator just pause trading on such stocks "until further notice" (I don't even know whether this is possible, but I guess that politicians can do whatever they agree to and pass a law or regulation)

Even if they don't, even if the funds go bust, only the early sell-offs from WSB will profit, while diamond-hands might be left with tons of $15 stocks that they paid $350 for...

If I may drop in with a bit of shameless self-promotion.

My "Deep Learning for Programmers: A Tutorial with CUDA, OpenCL, DNNL, Java, and Clojure" book explains and executes every single line of code interactively, from low level operations to high-level networks that do everything automatically. The code is built on the state of the art performance operations of oneDNN (Intel, CPU) and cuDNN (CUDA, GPU). Very concise readable and understandable by humans.

https://aiprobook.com/deep-learning-for-programmers/

Here's the open source library built throughout the book:

https://github.com/uncomplicate/deep-diamond

Some chapters from the beginning of the book are available on my blog, as a tutorial series:

https://dragan.rocks

Yes! They are just like the Bible. Concise, applicable, on the point, and you can learn so many useful (and true!) things from them.

Bayadera is much, much, faster, but it's not a probabilistic language. It's for data analysis with hierarchical Bayesian models.

According to google, the average salary in Kazakhstan is 212894 KZT/month, which is $500. How much does an American worker making $500/month have to pay to the IRS for SS and Medicare?

OTOH, isn't there a poverty-line amount in USA that makes some income level practically tax-free (through deductions, etc.)? I imagine that in most countries $15.000 / year is a rich person's income, while in USA it puts you under "starving". What I want to say, if you're living in, say, Kazakhstan, earning a good fat paycheck of $1000 a month and living like a king, shouldn't you file "I'm starving poor at $12000" to the IRS and get zero taxes, or, perhaps, even some food stamps or something like that?

(Just guessing, I don't have experience with this)

Hmm. One has to wonder how do one test the test of your tests, then. And those tests of the tests of my tests too...

Thanks!

convo is smart enough to cover 1D, 2D, 3D, and any other convolution layer that the backend can support. You only need to specify the data that it can't figure out, but you can specify more if you want.

Please read a few of tutorials from my blog. Most programmers in your situation told me that they had no problems following it; it only gradually introduces advanced Clojure concepts, and code snippets are usually extremely short + completely executable interactively as-is.

How is "conv" arbitrary? There is a function object that represents a convolutional layer in the network. It is bound to two symbols (because why not). You can either use "convolution" if you prefer full names, or "conv" if you prefer shorter. It doesn't represent the operation, but the layer. There are functions (with longer names) representing the convolution operation, which follow cuDNN and DNNL naming schemes.

Regarding the magic, I believe you haven't read my writings related to this. Exactly the opposite - there is no magic other than usual Clojure-fu, which I explain in a layered way.

But it's difficult to exactly reply to your critique, because you haven't given any example of an approach that would be good Clojure. Ok, give me an example of how you would do it in a comprehensible way (if what I provide is incomprehensible). You don't have to actually implement it. Show a non-working alternative. How would it look like?

Clojure does what I want.

Perhaps Cling can do something, or not, but I guess that's up to people who prefer Cling to find out and utilize.

If only C++ supported interactive REPL and the rest of Clojure/Lisp goodies, that might be possible. However, the code is CLOSELY related to the actual CUDA/C++ api. It's a lot simpler, concise, and everything, but I explain everything so that you can use the relevant parts with cuDNN and DNNL APIs in any language that you're most proficient in.