What is CME gap? Why they fill & how to spot gap on ...

Investigation of (Dis-)Favor 1\3 questioning freedom of will (in Macro-society)

This item began as a simple idea to investigate "social construct" theory, a trendy theme in academia and identity politics. It turned into a staggeringly complex constellation of ideas, with surprises galore.
The idea occurs to me: validate (or not) 'social construction' of beauty. If not, then sense of beauty is innate (source is not one's society, but genetic or other episocial influences).
Natural Tendency towards Beauty in Humans: Evidence from Binocular Rivalry 2016 | plos (technical study)
Reading in Contemporary Aesthetics "Why Beauty Still Cannot Be Measured", by Ossi Naukkarinen, because beauty is a personal determination, and a metaphor of favor, but how is it determined? Example: which of these women looks beautiful to you? note: only descriptor for the AI search is "beautiful woman", AI learns about beauty by sifting mega-data; how effective is it? (achieves given goal?) AI results are socially constructed in the most explicit way possible! Beauty may not be measurable, but it can be selected (parsed) from non-beauty. Measurement is a comparison of some phenomenon to an abstract dimension. Selection is a go, no-go choice.
What is Nudge theory? This item straddles the fence between Macro and Micro societies, paradigmatic Nudges come from Macro sources, but include an option to choose without pressure to conform. Micro sources are always more direct; which side of the fence are you on, friend?. (Greener side, of course.)
What about searching for "good", images? note how often the WORD good is pictured. Good is a language construct that must be interpreted from the individual's perspective.
Ok, now search for "favor" images note that AI mostly interprets favor as a small gift, not as a preference (which is difficult to represent by image)
Well then, search for "preference", images note that the word preference happens to be used by a line of hair care products from L'Oreal, which dominates the returns... commerce rules!
social construct (def, search result)
validate (or not) 'social construction' of beauty
Is Socialism a social construct? (LoL) Socialism Defined (EVERY Country is Socialist!) 2.2k views Sep 29, 2019 Rokn'MrE
To (social) Construct, or Not to (social) Construct, is there a choice? (note most results are about gender)
Parsing gender
Discussion of gender is not my direction of choice in this part 1 investigation. I want to seek how an actor (esp. me) makes a choice, in a quest for freedom of will. Perhaps come back to gender in a future item.
person makes a choice, in a quest for freedom of will (selections available)
Exemplar Hyp (Harry) Frankfurt’s compatibilist theory of free will 2009 5pg.pdf
I notice my choice of article was partly determined, partly free, but parsing out those factors would be too much divergence from the goal here. But Truth (a two side coin) is my story, and I'm stickin' to it.
(previous link, compatibilism):
3 It explains our intuition that human beings, but not lower animals, have free will. Lower animals lack free will because they lack the second-order volitions which are constitutive of free will. (This item is unnecessary and probably not true; how do we know animals have no "second-order volitions"? Having no other language than "body", we can only surmise (guess) what their volitions are. Volitions come before actions, we cannot see them or interpret them in any way. Brain conditions might be interpreted with MRI scanning, but to put a subject in a scanning device is to prevent any other actions. Such measuring ruins the connection between mental state and volition being measured, except we can safely assume that every measurement of animals must default to the volition to escape the measuring device.)
That's the first-order, highlighted deviation from compatibility theory. Clarification of "second-order volition": a path from choice to action has an intermediate "middle-way" tunneling mode, contracting (taking on) a desire to make a choice, prior to making the choice. In order to prove freedom, one must establish the mental preference for an imagined outcome in order to prove that preference did come from within the person and was not forced by other external deciding factors (genetic factors are pre-determined).
incompatibilism Note: the approach is wrong by the universal assumption, IOW that the intersection of determined and free is zero. It's a supremacy position, or superposition principle (LoL), the error is in over-simplification. The Logic Argument (p.5) is not representative of reality, which is more nuanced. Therefore, Frankfurt's thesis is good (denial of incompatibilism), but not due to the case presented (superposition).
Take Frankfurt's case (p.4) of Black vs Jones4 to be analogy for State vs Individual. Silent Weapons for Quiet Wars (other sources exist, search for yourself)
The (myusername) determinism/free-will duality hypothesis (denial of incompatibilism due to non-zero intersection):
Most choices, including the choice of desires, are determined by contingencies of which one is the natural desire of the actor to optimize his/her outcomes ("best wishes"). Is a person always compelled to have best wishes? What is best depends on a person's mental state, which is usually determined by external factors, but those can vary in cogency (impact on behavior). Consider the choice to commit suicide, certainly not a trivial choice. (The Chosen means of execution (puns intended) is somewhat more trivial, but again, partly determined by external conditions.)
Some choices, nearly all trivial, are free because no interfering contingencies are apparent during the choosing interlude. It may happen in hindsight, that a past choice is observed to be a mistake, usually because some contingency was overlooked or unknown during the choosing. This observation should be remembered so as to avoid repeating a future choice like that mistake. Choices always have risks, including the choice to do nothing.
Different day, slightly different approach... parsing choice. 1 important choices that have many deep effects later, for instances a marriage partner, a new job, a new residence; 2 trivial choices which have minor effects, risks or physical involvement, for instances a choice of toothpaste at the market, to like or not a web-link or museum exhibit.
According to (myusername)'s determined/free paradigm, type 1 choices are nearly all determined by pre-existing conditions (not free). Type 2 choice is the arena of freedom. I suppose a person's low risk-aversion parameter could expand the envelope of freedom, but that's a characteristic that develops during maturation, one's history of choices and ensuing responses. Successful responses lead to more freedom, failures to less. So even when freedom exists, it accumulates a history (habits) which become a determinant.
Contracting the Social Construct Disorder (it's contagious) Take 1:
How does an actor (person in question who comes to an internal state, or inner-construct) interact with a community or society? Must it be IRL, or can virtual interaction suffice to construct internal states? And more to my point, must the interaction be two-way (containing feedback), or simply via broadcast medium? (broadcast includes published books, articles, records, radio, TV or Internet A/V shows, etc.)
Interaction with broadcast media can be summarized by: a choice, a degree of attention and focus (time spent on and attention given to item), a like/dislike or more complex reaction to item, having future behavior influenced by item, to continue a stream of behaviors (especially sequential item choices) as consequence of influence of item, to develop a complex of attitudes built upon stream of items (eg. just mentioned 'risk aversion parameter and habit).
Before going on, I notice that broadcast media is like Sunshine, Rain, and Grace. It is made available by participation in a community, and falls without curse or blessing, it's all there for the choosing (or ignoring), depending on the contingencies.
Mind control theory? (because mind is the inner source of volition... behavior, control the mind (easy), hence control the behaviors (difficult otherwise))
Mind control courtesy Tavistock Inst.
Construction of Favor (or any knowledge) upon Familiarity
What is Social Construction? (cntrlZ)
"For instance, trees are only differentiated from other plants by virtue of the fact that we have all learned to see them as "trees."
But we don't all know about trees to an equal degree. I know rather much about trees from my interaction with them: living among them, planting them, sawing them, moving them, burning them, etc., not from reading or talking about them. No doubt, there are many persons all over the world who have very little experience of trees, and cannot 'construct' treeness as well as me. Direct experience is more realistic and developed than social constructs.
Favor and Familiarity are interwoven by choice
I chose to live alone with trees and not alone with sea, or desert (for examples), because it was easier to go with trees. Was the choice free? I could have chosen city or suburb with even more ease than forest, so ease of choice was not the deciding factor, it was my preference of lonely forest over crowded urb that decided me. So maybe it wasn't really about trees, it was about independence or something else like that. When we choose, we may not understand the contingencies, but our decision (choice) may be due to habits or patterns that have developed in the maturity process. Habits are strong determinants, and they develop, according to Ian Plowman, 4 ways.
The cntrlZ article makes the case for 'Strong Social Construction' based on that 'knowing' which is all about language, certainly a social construct.
Within the social construction of language is the game. Outside the social construction is reality, the real world. (a list of social constructs follows)
That makes it clear. Experiences (direct ones) without resort to language are NOT social constructs. That observation makes another distinction clear: gender may be a social construct, as it's a language issue, but sex is not a social construct, it is a direct experience issue that develops in the maturation process: birth, infant, child, puberty, sexy adolescent, sexy adult, old (unsexy) adult, death. Prior to puberty, sex is incipient in its development, but comes to life, (like a flower blooms) after a decade or so. Knowing about sex as a child is by observation from outside (thru the looking glass), after puberty, it's direct experience, and much later, it's a fading memory.
Regarding Looking-glass self theory the notion of socially constructed identity (defining the self by differences/ affinities to others),
... the outcome of "taking the role of the other", the premise for which the self is actualized. Through interaction with others, we begin to develop an identity of our own as well as developing a capacity to empathize with others... Therefore, the concept of self-identity may be considered an example of a social construction.
... makes a spurious expansion of identity formation to include everyone (a unity), or nearly so. According to Reisman's Lonely Crowd, there is a triality of social nature, expounded by parsing people into tradition, inner, and other directed personalities. This theme was a scholar's response to the US trend toward consumerism and conformity to "norms", (local traditions, eg. "keeping up with the Joneses") mid-20th century. The social construct crowd would be Reisman's Other directed personality, which may truly be the majority, in USA certainly. However, the tradition-following and inner-directed personalities are a significant minority. Let's not ignore them (I'm in there.)
What is “Mob Mentality?”
Herd mentality | wkpd
Are All Personality Descriptions Social Constructions? Sep.2019 | psytdy
... that objective reality does not directly reveal itself to us, is true beyond a doubt.
The preceding statement author, JA Johnson, is way off (and his article is full of falseness). Objective reality IS direct experience, no more revealing modality exists. Denial of this obvious fact (just lied about above) is a redefinition of the term (a social construct). Experience is beyond language, thus beyond 'description'. However the following is a true reveal about (((Yews))) (the like of whom Dr. Johnson seems):
It is true that when we describe someone with socially undesirable traits... we are constructing for them a social reputation that might decrease their chance of success in life. This is precisely one of the concerns of (((social constructivists, like Dr. Johnson))), that certain categorizations (eg. a separate race) reduce power and status.
Proof that Truth is not a social construct (relative to culture, like morality absolutely is)... What do you believe in? Cultural Relativism
Conformity is a social construct (should be obvious, it's a social source of choices). What causes conformity? Social interactions, which traditionally occurred (Macro-version) in newspapers, magazines, cinema and radio programs. As culture changed the popular media to radio, TV and then to Internet, and church attendance fell out of vogue, the advertising industry became more powerful in defining social constructs. That's why Internet censorship is so important.
Who are the 'influencers' in society? (They used to be parents, teachers, peers... now it seems to be YouTubers, like PewDiePie. But an intentionally underplayed contingent of influencers is the predominantly Left-Leaning academia, who collectively promote Marxist preferences and political activism toward Socialist positions. Academia is pushing social construction because it provides an intellectual framework that denies the old (social injustice), and says ok to their preferred ideology, Cultural Marxism (new social "just us"; socially constructed ideas can be anything you want, their cogency depends on efficacy of publication).
Micro-Social Constructs are most cogent (due to conformity being human nature), discussed in part 2.
Bottom Line (part 1)
If you like freedom, and are serious about it, you must distance yourself from society, because it tries to reconstruct you according to the norm... conform!
Before you go, think about what is a hermit?, which should not be confused with Hermetic, name derived from Greek god Hermes. 7 Great Hermetic Principles – The Teachings of Thoth (illustrated)... same topic 2016
Investigation of (Dis-)Favor 2\3, Micro-Societies
Social Circles; Mates, Kin, Friends
note on Mates: school-, (prison) in-, marriage-, ship-, etc. note on my link choices, page rank has a strong influence
Social group (aka circle)
Is Conformity Human Nature? Don't blow this list off, if you want to understand social constructs. At least look at first item.
What is Social Proof?
Are Micro-societies any less 'constructing' than Macros? Or do some constructs exist for all realms, macro and micro? I think they are more constructing, because micros carry feedback, whereas macro is all absorption, individuals have negligible effects on society at large. They act in a statistical sense, with a few exceptions.
Concept vs Percept (concepts are stable mental recordings, and physical manifestations of them; percepts are changing sensations and reflexes which depend strongly on the situation, memories of which are variable too)
Favor, Good, and Beauty are words that belong in the same 'conceptual basket' (ward), they are alike, all refer to action 'like', as an affective (and affirmative) perception. Conversely for the word's opposites.
Perceptions are non-language reactions to stimuli, therefore not social constructs. They may be evoked into a social arena via language (or other virtual records), but these are only shadows of the perception, so what is evoked is drawn up from the receiver's own memories of perceptions.
Division of Labor (and role models) are Social Constructs
Sex is the most basic divider of labor, for all societies, especially the most primitive. As societies develop towards more technical, sex falls away from the divider, as natural talent and innate interest gain influence, until the basic operations of reproduction remain, the core division. What about rankings in the division?
natural tendency for dominance?
Are males naturally dominant in nature? | qra
(arguments opposing) Male Dominance (theory) with (bogus) "Explanations", by 2 feminist authors using Marxist ideology 2017 | verso While this blog seems to have obvious (to me) flaws, it does raise interesting ideas and references.
what attributes help males gain social status? Basic: status is competitive. It takes talent and effort to win.
To Raise Male Status (18 Rules) | @rctvmn (not because age 18 is best)
Dominance vs Prestige 2010 | psytdy Note: blatant bias toward Prestige via argument parsing Pride. (author is Jewish, maligns DJ Trump (nationalist), lauds John Lennon (globalist))
modes of thought: socially-controlled vs spontaneous
Major Component of Social-Construction: Public Education 3 Modes of Thought Jan.2019
Kaufman again: How Renaissance People Think 2011 | psytdy Note: We discussed concept vs percept, here Kaufman refers to fellow-Jew Seymour Epstein's dual modish rational vs experiential theory, same idea set.
polymath (short for Renaissance Man)
Favor-Goodness-Beauty paradigm
Favor is not favored in prior art, Truth takes Favor's place in the Transcendental Spectrum: Transcendentals 5pg.pdf
We have already seen the idea in part 1 that Truth is a disputed transcendental in the social-constructionism academic universe. Academics use the "universal fallacy" that their favored item is part of an incompatible pair, which by logic excludes everything not in their favor. They want to ignore the nuances in order to push an ideology toward a supremacy of thinking, just like in a totalitarian state.
Whereas the (myusername) principle of Truth, it has a dual nature, 1 relative to a society (democratic consensus); and 2 absolute to reality (math/science/technology). So 'Favor' is a better term because objective proof (no contest) is not required (except the meaning of objective that says 'objection!', meaning 'contest'). 'Favor' implies bias which is the subjective reaction that matches Goodness and Beauty better than 'Truth'.
Apply Truth-Goodness-Beauty paradigm to social construction
it is unconcerned with ontological issues...
because the aim of constructionists is to justify a collective "truth" of their own construction. A social construct is not absolute, it's anything a society wants it to be ("social proof"). That's a good description of tyranny... The Empowered Female Parasite 2014 (that's a surprising result, here is one not-surprising.)
Social Proof: established by culture media (mind control, a monopoly 2012 (scroll down long graphic), of the Juice 2015), go back to part 1, macrosocial constructs.
Does Appreciation of Beauty have any innate sources? (otherwise it's all a social construct) Neuroscience of Beauty; How does the brain appreciate art? 2011 | sciam (in brain)
Onward (Dis)-Favor Readers...
Investigation of (Dis-)Favor 3\3, House of Not-Friends
Contracting the Social Construct Disorder Take 2
Living outside the 'Normitory" (away from Dreamland (everybody's asleep), to where Nessun Dorma (nobody sleeps))
It so happens that an ethnic group which originated in eastern Mediterranean Middle-East evolved to specialize in intelligence, commerce, morally corrupt enterprises, and crime. Essential to their success was eugenic traditions that applied artificial selection to just those same specialties, which makes this ethnic group a formidable enemy. They have developed a very strong sense of in-groupness, and a vested interest in social construct studies. A unified collective is a more effective competitor than an inchoate population of diverse individuals.
This group has as ethnic traits: global dispersion (aka Diaspora), preference for urban environments (aka Cosmopolitan, or Globalist), covert inter-group rivalry (aka InfoWar), and deception (aka MOSSAD). This cosmopolitan group must operate covertly and deceptively, because those are effective tactics, and they are a small minority (2% of USA), therefore weak in the democratic sense.
Immoral Social Constructs enforced by 5th column subversives
wethefifth (political audio series)
serendipity: freethink
Another construct search, without gender reference
Is morality a social construct? If so, how can concepts such as 'good' exist? (note especially the links in top comment, to reddit posts)
"Good" can be understood as a variation of "Favor" as a direct experience (perception) of "like", rather than some idealized notion of an obvious social construct such as "greater good" (a theoretical derivation by interventionist actors-with-agendas trying to impose their own preferences upon others, IOW ideology hegemony pushers, for instance viz da wiz)
Cultural hegemony is the Chosen's mitzvah, that we all must go to Emerald City, land of Oz, where YHWH (impostor) rules.
Cultural hegemony
Concepts of Ideology, Hegemony, and Organic Intellectuals in Gramsci’s Marxism 1982
There is no universal morality. Morality is much like Beauty, in the mind of beholder (actor who holds to a specific moral code). Morality is a social construct, and varies between societies. (I think a fair definition of morality is a code of ethics which is community-specific.) For a society to sustain, it needs to be isolate from conflicting societies. If different societies, with different moralities must coexist, the natural tendency for actors in the same niche toward dominance will destroy or remake the subordinate societies, which reduces the conflicts.
Status Hierarchies: Do We Need Them? blog 2012 | psytd
a need for 'virtue signaling'? It's natural, and likely unavoidable, evidence pride displays.
Status Assignments: by birth (heredity) or merit (talent)?
Let's assume your morality values social effectiveness. The best path to that is to have talented persons dominant (meritocracy). Next we happen to know that talent is hereditary, but not perfectly so. Therefore birth (kinship, aka kingship) is only an indicator of talent, which is infrequent among low status groups, much higher among high status kinship groups. Thus we must conclude that awarding status by pedigree and family privilege is not the best way to effectiveness, but it often does work. What works best then, must be? a competitive system of merit-proving, with special attention to high-status families. (Helps if the natural tendency for snobbish repression is circumvented, for examples Han-style Civil Service Exams, and the Roman military promotion avenue, which occasionally led to top gun.)
Sustainable Competitive Advantages (aka moats): Network Effects 2019 | sEknα
Our Brain's Negative Bias 2003 | psytdA
Fear: it's the greatest (motivator) 2009 Owen Benjamin made a video about Fear over TIME 16 min.
Dominance Hierarchy employs FEAR to dominate
Dominance hierarchy | wkpd Social dominance theory | wkpd
scaring children is not ok, Sydney Watson blog 11 min
tools for social mobility and dominance (list)
9 Important Factors That Influence Social Mobility Social dominance orientation | wkpd SDO should theoretically be highly important to Jews, as their ethos tends strongly to emulate it among themselves and denigrate it towards outgroups (Goyim). Thus we should expect to see this field of study monopolized by Jewish scholars. Studying the Gentile: Fanciful Pseudoscience in the Service of Pathologizing the Covington Boys | OO
Contracting the Social Construct Disorder Take 3
Different day from Take 2. Re-consider interactions with a community or society: traditionally occurred locally, on Sunday meetings at church, parties, having a beer after work, town hall or children's group meetings, (eg. PTA, scouts) etc.
Re-consider "contracting". Original idea was meant to acquire, like a disease, not by design (choice), but determined by contingency (unlucky chance). Today, "contracting" means getting smaller, shrinking, like a cooling branding iron, or melting ice. Iron has several crystalline phases, the cooler, the more compact (more atomic order). Ice is contrary to most materials, as its crystalline structure is larger than its liquid phase, so as it melts (entropy always increases, going to less ordered) its atoms become more fluid. In both cases, the natural mode of change is toward ambient temperature. This trend (recursion to the mean) is maybe the most unbroken law of all physics.
Re-considering "Disorder"; original idea was meant as a mental disease, like ADHD (Attention Deficit Hyperactive Disorder), IOW anomalous condition, out-of-order, (order being assumed normal) in the human behavior dimension. Today it means individuals out-of-line, like discontinuities in a crystal. (Discontinuities are what make metal harder.)
When all the atoms of a metal are aligned (continuous), the state is called "annealed". This is the softest condition. When the metal has been "work hardened" by hammering, or forging, it acquires discontinuities (crystalline order becomes mucked up). This is a harder state. Hardness is measured by forcing a small ball into a test material and measuring the resulting depression (dent). Discontinuities resist dents and every other kind of deforming force (decreased plasticity (weakness) means increased elasticity and maximum yield (resilience, see Young's Modulus, Indentation hardness, Impact Toughness and Moh's Hardness)).
Now make analogy of metal with society. Non-conformist individuals (like followers of Marginotions) make society (if he-he-heeded) more resistant to outside forces (like George Soros, or seekers of Tikkun Olam) trying to make a dent in the established order (tradition, Protestant Ethic).
Contracting the Social Construct Disorder Take 4
Different day Re-consider "contracting" again. Today, it means make-a-deal, as in commercial contract. This kind of contract is in flux nowadays, as the advent of bitcoin has introduced a mathematical means of authorizing legal agreements (aka contracts) in a distributed ledger that makes such agreements social in a very direct sense. The social part of "social construct" is present in a world wide network of participating computer operators, while the construct part is present in a software package (app) that is now called "smart", meaning has built-in security and ongoing timely operations, like confirmation checking. In this contract-paradigm, the "disorder" part is due to it being outside of previous power-holding elites who are chagrined by the prospect of losing some of their powers to the Internetwork, which is out of their control. IOW disorder for elites, and made-to-order for independents. (note on that quote)
Social Contract per britannica (briefly) per wkpd
explicit vs implicit contracts Differences Between Implicit & Explicit Agreements (law) 2017 more specific, social contracts Social Contract Theory UT (includes videos, glossary)
to be continued: fairness is a social construct (contrast with deterministic fate)
study notes (all 3 parts, this series)
Gentrification, Displacement and the Role of Public Investment: A Literature Review 2015 pdf
https://duckduckgo.com/?q=Owen+Benjamin+made+video+about+Fear&atb=v81-4__&ia=videos
why is 'social construct' a popular theme?
https://www.success.com/8-daily-habits-to-build-your-mental-strength/
Pareto principle implications for marital harmony, a very brief summary of research by J Cacioppo)
https://en.wikipedia.org/wiki/Jewish_culture
https://en.wikipedia.org/wiki/Race_and_society
https://thejewishwars.blogspot.com/2019/03/aipac-traitor-jews-having-successfully.html
https://theevilofzionismexposedbyjews.weebly.com/14-what-zionist-and-anti-zionist-jews-have-said-about-education.html
submitted by acloudrift to AlternativeHypothesis [link] [comments]

[Table] I am a member of Facebook's HHVM team, a C++ and D pundit, and a Machine Learning guy. Ask me anything!

Verified? (This bot cannot verify AMAs just yet)
Date: 2013-10-02
Link to submission (Has self-text)
Questions Answers
What are your thoughts on the constraints vs static if debate on C++? D uses static if, and I remember seeing you championing it as the #1 feature you'd like seen in C++, but Stroustrup describes it as "a total abomination". Do you think that constraints do the job well? Possibly better than static if? I knew this was gonna come :o). Virtually everyone in the D community has an appreciation for static if - I have yet to find even a naysayer who's mentioning it as an unsavory aspect of the language.
That isn't a proof, but it is evidence. There's also plenty of evidence that C++ is worse off without it by comparison.
Now, the more debatable aspect is the use of Boolean expressions as constraints. The simple story behind that is that Walter Bright and I were looking at a simple means to constrain instantiation of templates. We had static if, we had compile-time function evaluation, so template constraints were a wonderfully simple and integrated solution within that context - a great "aha!" moment. People love it.
Now, in C++, there's no static if and C++14 is acquiring compile-time evaluation kicking and screaming. Within that language, I can totally understand how the context is less conducive to an appreciation of template constraints the way they're done in D.
Do you plan on writing more C++ books? or did you pretty much give up on the language and your focus is now on D? I have to admit my relationship with C++ has lost a fair amount of its romance. We know a lot about each other and we roll our eyes when witnessing once again each other's shenanigans.
I'd be hard pressed to write more books on C++. As a general rule the best driving force behind writing a book is "I feel I have something to say, that I believe is interesting and worth sharing." I do feel this in spades about D, but unfortunately not about C++.
I must add one note though. The C++ community has been very gracious and forgiving with my apostasy. I've continued to be bestowed respect from C++ programmers all over, and I am very appreciative and thankful for that.
I read Modern C++ Design a few years ago and really enjoyed it. It's been over a decade since it was published, however. How outdated do you think it is now, especially in light of the changes from C++11? It has aged surprisingly well, and to a C++11 user some of its implementation arcana are obviously easier to realize. I think if it aimed at describing the same exact designs in C++11, Modern C++ Design could do it in 250 pages instead of 350.
How much of FB's code is still plain PHP? How much is C++? Is HHVM really that much faster than ZEND? What are your other hobbies outside of programming / computers in general? Keith Adams has been gracious enough to run a "wc -l" on our main codebases just now. Fresh data! We're roughly in the 70/30 area for PHP/C++. Both codebases have grown a lot in absolute numbers since I joined, and most interestingly the ratio was somewhere like 90/10 four years ago. We've scaled up a lot since then, hence the increased emphasis on infrastructure.
As a follow up to that: PHP gets a bad rep these days for some poor language design. Is there any regret in using PHP or a desire to switch the code base to another language? Edit: Thanks for doing the AMA! I enjoy watching your presentations. "Regret" wouldn't quite be describing it, seeing as there is good evidence that our engineers are very productive with PHP. That said, few people if any would disagree with you. We have always been and still are working on a number of solutions to improve on PHP "the language" and also PHP "the platform".
What's the first thing you'd do if you were put in charge of creating PHP 7.0? Better arrays.
Would you mind expanding on this please? PHP arrays are very quirky and okay for a lot of things but best at none. They are used as straight contiguous vectors, tuples, singleton sets, maps, ..., you name it. To accommodate all these semantics PHP arrays have very complicated and non-obvious semantics. I think everybody would be happier if PHP had better-defined and more specialized types.
Thank you very much for replying. I guess this is where I realise I'll never be a proper computer scientist. For me PHP arrays do everything I need them to do, in a way that I find clear, easy to understand, and with a performance level that is more than adequate for the problems I have to solve. The beauty of it all is we may both be right!
Are there plans to use D in Facebook? Not for the time being. However, since recently it's become part of my job description to explore integration of D at Facebook. This is the first time I'm mentioning it publicly. assert(cat !in bag); I have a well-defined plan that is difficult but executable. If things go well, we'll make an announcement in a few months. Otherwise, well, we won't :o).
What do you think about Rust? (In the context of Rust vs D) Rust has a great approach to safe regions. That's a hard problem, and Rust has had to expend a considerable amount of firepower on it (four kinds of pointers etc).
D does not offer safe regions; I believe the language design precludes that without at least an amount of discipline.
So Rust is better than D at safe regions. However, like in chess, good language design is to not sacrifice the whole for the beauty of the part. I think D is better than Rust at a lot of other things, because it has firepower it can afford to expend at problems that are also hard, and just as important.
Other than C++ and D, what other languages are interesting to you in both a professional and personal capacity? I think Scala and C# are two fine languages. Haskell is a factory of good programming languages research. I've seen a talk on F# and it's been quite impressive - it's amazing what can be done with them reified types.
Generally I have a bias in favor of static types.
When will we be able to use D to make shared libraries that can be called from C? Good news: Martin Nowak, Walter Bright and others have done some wonderful, creative work on that. The next release of D (2.064) will include good support for dynamic and dynamically-loadable libraries from D itself, C, C++, and other languages.
The release is in the planning stages right now, so we're looking no longer than one month from now.
Software dev here. Facebook used to have a bit of a reputation for putting bugs into live (for example when all line returns on photo captions were replaced with "\n"), but it's my perception that things have improved vastly in the last couple of years. What sort of QA or other processes have you guys put in place? I'm glad you are noticing what has been a years-long effort to improve our process.
First of all, if one at Facebook talks about "process" in the RUP/Agile/etc sense, they'll get a smack on the head. We don't care much about such formalisms. What we do care about is making talented people productive, and for that we have a vast array of automated and semi-automated aides that I'll get into a bit below.
Facebook famously does not have a QA department. Engineers are responsible for testing their own code. The way that works is holistic, which checks and balances at each major bottleneck: (a) a sophisticated lint - all code must past linting; (b) then there's unittesting - all code must pass existing unittests; (c) code review - not one line of code makes it into our code repos without having been reviewed by at least one other engineer, who in particular looks for unittests added for new code; (d) various sandbox and production testing means.
Clearly it's good to move fast, and one thing that virtually all new engineers remark is how astonishingly fast things are happening at Facebook Engineering. A complete n00b could implement a feature visible to Facebook's entire user base literally within a week of starting. The converse risk is that of breakage, and the conventional-wisdom response to that is to increase bureaucracy and slow things down. Facebook has consistently striving to improve tooling and automation that enables people to still move fast, without breaking things.
When things do break we don't reprimand people who make mistakes. Instead, we look at eliminating the causes of breakages at their root.
The real question is: did you try llvm? Good one! Yes, I and others are working on having clang compile our codebase.
What's the biggest lesson learned from D that you wish other languages would apply? The scope statement. It's rare that I enjoy bragging about something, but I do like to brag that I invented a new control flow statement (which is awesome because they're so few!).
I think many languages implement some form of deferred execution, but most miss the point - Lisp's with-open-file, Java's try/finally, Go's defer, C#'s using are all sorely wanting.
Lisp's with-open-file with-open-file is just a wrapper macro on unwind-protect (for, as the name indicates, performing scoped opening of files). It's not a deferred execution primitive in and of itself. Sorry, yes, I meant unwind-protect but I had a lapse so I mentioned its cousin that I remembered. Both illustrate the same point.
What has been your happiest moment with programming? (thinking about this as I'm answering other questions, will edit later)
OK, I think most people left the stadium but this has been bugging me so I feel compelled to reply.
"Happy" is frequent enough to make "happiest" very difficult. Probably some of the happiest moments have happened when I had just started. Coding is like heroin - we spend most of our time trying to relive that first high.
I remember moments when I'd run a little program again and again with slightly different inputs just to revel in the joy of having done the proper incantations that make the machine do this and that and the other, like a clumsy but loyal genie. I mentioned I wrote this floppy disk formatter - it gave me a lot of joy to be able to tell it the complicated sequence of things I wanted to get done, to see how it ended up carving magnetized trenches into the physical world.
With support for more things inside constexpr just around the bend, do you think we can now do metaprogramming in plain old C++? Wouldn't it be more natural and easier to debug? I am glad that C++14 has increased the power of constexpr significantly. We have had an incredible time with compile-time evaluation in D for years, and I'm glad C++ "stole" that idea. I predict that constexpr-based programming will become a major idiom in C++14.
Also, a second question if you don't mind: How do I go about making proposals for consideration by the ISO C++ committee? I would like to propose a make_shared where I can provide the allocator rather than it using new. For submitting proposals to to the ISO C++ committee, mosey to the relatively new and unknown Link to isocpp.org
How did you acquire such a mastery of the English language (written + spoken minus the accent ) even though it does not appear to be your mother tongue? I must plead no-merit on that one. All of us get some good and bad inclinations upon birth, and it just so happens I inherited my Dad's inclination for foreign languages.
If I were to speculate it might have something to do with how thoughts are formed in the mind. Many people I ask tell me "I think in my mother tongue". For whatever reason, I don't think in a specific language. It's all that, abstract thoughts, and then the matter of rendering them in a language is simpler than translating them from one language into another.
talk about impostor's syndrome. Damn it’s reassuring to hear this from a smart, well-established guy like you. And I am serious. I would rate the impostor syndrome near the top of things that bother me.
I'm a big fan of your writing style (especially the phrase "hecatombs of code"). What are some of your favorite books? Thanks, that's very kind of you! Alas, I feel I missed out on reading lately, so I must dig earlier into my youth, from where "Remembrance of Times Past" comes to mind. Recently I've really enjoyed books by David Foster Wallace, Cormac McCarthy, and Neal Stephenson.
In the technical realm, I really enjoy the style of Scott Meyers, who's been a major influencer and mentor. I also like the writing of Herb Sutter, Bjarne Stroustrup, and the late John Vlissides.
Have you read any Thomas Pynchon or John Barth? If you're a Foster Wallace fan I think you'd get a lot out of those two. John Barth's Lost in the Fun House is an amazing piece of metafiction, somewhat fitting reading for the master of metaprogramming. Thanks for the recommendations!
Getting back to D's GC, I want to point out that you made it extra unattractive by choosing an old clunky one, you just dropped in the ready made, generic Boehm GC. So D not only has GC, but a GC that is far from the state of the art. As I mentioned elsewhere on this page, languages succeed if good people decide to work on them. We have a couple of GC experts on "team" already, and things are getting better. The better the GC, the less trash in the streets.
1) Is D used internally at Facebook? I'll understand if you can't answer, but 'no comment' means yes. ;) (answered above) 2) Yes, things are improving; however the improvement will be visible to the public later. I'm very busy right now with my efforts of exploring D within Facebook, which is not directly related but definitely related strategically.
I think binding rvalues to const references has been the small mistake that caused the rvalue references Hindenburg. Can you explain further? I'm not familiar with this mistake or the consequences. It would be a long discussion. Binding rvalues to const& made sense when first introduced (no templates, few subtleties) but in the long term made it virtually impossible to distinguish rvalues from lvalues on the callee side. That in turn forced an overly complex solution (rvalue references) as an expensive fix.
A big improvement for the phobos thing would actually be to make more use of output ranges instead of necessarily returning strings, etc. Then the caller can preallocate memory or whatever they want. Indeed! That's a sore point right there.
Are we going to see a D book on range based programming? I think ranges are really cool, and well deserving of a book treatment. Someone should write it. If you're asking whether I should be the one, I think I feel I have at least one book in me, but I hope to collect a bit more material.
At going native this year, you mentioned that C++ is closely tied in with the von neumann architecture architecture. Is this true of D as well? Does it mean that if the von neumann architecture were to be replaced by something shinier, C++ (and D) might vanish into oblivion? Yes, both C++ and D build on the same fundamental computing model: data has addresses, you know where it is and you know how it lies. But I'm not worried - if von Neumann is fundamentally replaced a lot of languages will need to adapt, and probably the best breed will be new ones that use the new model organically.
Also, do you think functional languages are better in that sense that their abstractions are not tightly dependent on the von neumann architecture? I'm not too good at predictions, but I did predict the next two paragraphs were to follow after seeing the first :o). There's clearly a lot of good in functional programming, and I think it's undeniable that features once considered hardcore FP are making it into the mainstream.
Notwithstanding, I'd like to hear what you have to say about functional languages and what you believe is their role in the coming years. Betting the farm on FP as the universal computational model would, however, take matters too far in my opinion. FP is doing okay today if the likes of 2-3x in performance is no matter to you. (I know, I know, there is this benchmark and the other etc.) It's possible that FP will be doing okay but just as mediocre on a new computing substrate. Also, it shouldn't be forgotten that FP offers, simply put, just crappy solutions to a variety of simple problems. It is nice in the sense that it's mathematically motivated, but there's plenty today that is also mathematically motivated. Formalisms have made good progress in the past two decades.
What's your opinion as a language designer about this feature-richness/simpleness trade-off? As there are languages all over the spectrum, clearly this is a matter in which reasonable people may disagree. Probably a good judge is "power offered per cubic inch of complexity" or something like that. Languages like D, Scala, or Haskell do well at that metric. Go is simple but in my opinion not the right kind of simple; in many ways it's a sort of a collective bummer that a language like it has emerged in this day and age. However, as I said, languages succeed because good people work on them, and there are good people working on Go.
Are there any features in D 2.0 that, in hindsight, you would want to drop? I'd drop the postblit this(this). It doesn't work well with qualifiers.
Any ones you wished you had incorporated but didn't? If I could have added a means for transparent reference counting without adding complexity to the language, I would have.
If there was anything you would today remove from your Modern C++ book what would it be and why? I'll tell you what I'd change in Modern C++ Design: the chapters on Allocators. Policy-based design can do wonders at creating layered allocators, see HeapLayers. I just missed that obvious idea, and I regret it.
What do you think about the Go language ? Which kind of feature would you borrow from it ? - What do you think about having more threading aware semantics in the languages instead of threading libraries ? (I'm thinking for example to the "synchronized" keyword in Java, or the Go, etc). (I assume there's understanding that I'm horribly biased.) I have a dim view of Go; I find it an unremarkable "me too" language that would be nowhere fast if Google weren't behind it. That said, it is executed beautifully and Google's engineers are true to form in delivering world-class server-oriented libraries for it. Go will probably never supplant C/C++/D for systems needs because it requires GC for core operations (fact understood by its proponents who rewrote the marketing message shortly after the initial launch). From Go I'd borrow the engineers who wrote its networking libraries.
There is no way around having the language understand concurrency at its core (unless it designs around it entirely, like PHP). The exact distribution of capabilities across language and library becomes a matter of taste. I'm a minimalistic guy who prefers "few axioms, many theorems" which applied to this would put few primitives in the language that allow for rich libraries.
D is a great language with a poor tooling. What are the plan to improve on the tooling part (and conquer the world) ? I noticed that many languages are successful because good people have made it a point to work on them. At this point my focus is to foster good quality in the language proper that would make it attractive for others to join. If I play my cards right and get other proverbial penguins to jump off the equally proverbial iceberg in the water, tooling will definitely improve.
Would you say D is different? Is there a strong incentive for C++ programmers (and lovers) to have a good look at D? There is an increasing amount of programmers who have decided they need the amount of modeling power and efficiency that C++ offers. This is because of a variety of phenomena (no more frequency scaling, heat is becoming a limiting factor, dynamic analysis has its limits etc). For those, I venture to say D is a very attractive choice because it offers all C++ does and a lot more, at a lower cognitive cost and with build times faster by one order of magnitude.
I must admit I'm reluctant to switch to a garbage-collected language (which is, pretty much any language that's not C++.) Knowing that everything (objects, memory, files, etc.) will be released/destroyed automatically at a predictable moment is comfortable. The simple answer to your question that surprises many non-language-pundits is that garbage collection is mainly a tool for memory safety. One may opt for a complicated typed regions system to avoid tracing (Cyclone, Rust), or simply rely on a tracing collector.
What in C++ resource model is considered so bad that every single other language is garbage-collected? (Let's forget those where you have to do everything by hand, like C, ASM, and butterflies.) I made more remarks on garbage collection in a different post.
garbage collection is mainly a tool for memory safety. Sure, it's merely one possible tool. Everything is a value. "a=b" means "a is an independent copy of b." A function can return a value. Static polymorphism (i.e. templates) is king here. And there's a very simple rule for resource management: when a value gets out of scope, the associated resources are automatically deallocated. Everything is an entity / a pointer. "a=b" means "a is another pointer to the entity pointed to by b." A function can return a pointer. This is the realm of inheritance polymorphism (i.e. virtual functions). Typically, you have "Base b= new Derived", and of consequently, deep copy is cumbersome and thus, rare. Of course, for such a model, you need a GC. Both models have their limitations: Sometimes, in the "value" model, you actually need an "entity" object. And things get murky. OTOH, in the "entity" model, sometimes you actually need values. That's why Java has both int and Integer. What surprises me is that the "value" model, despite working very well (as long as you don't want/need Java-style OOP), seems to have been implemented once (in C++), and was then completely abandonned. Why? Excitement for "everything is a value" decreases considerably when one realizes all referential structures are precluded (lists, trees, graphs - and a lot of object models are graphs). That in turn locks one out of a vast application area. Having an object that just refers to another is pretty darn powerful.
How do you think about some standard package manager for D, like npm for Node.js and RubyGems for Ruby? Are there any plans yet? Yep, code.dlang.org is the main candidate.
And: Git/Mercurial? vi/emacs? cats/dogs? We're switching from git to mercurial for scale reasons. I'm fine with anything that works, and love the concept. I give a ton of credit to Linus for having created git back in the day.
We're switching from git to mercurial for scale reasons. Could you tell us more about that point? Git takes long times with our front-end source tree (which is enormous). The matter of customizing git came up and people looked at the code and decided it's pretty convoluted when compared to the Mercurial code.
3. I was very impressed by Sean Parent's talks at this year's Going Native (Link to channel9.msdn.com What is your opinion about the advice he gives? > 3. I was very impressed by Sean Parent's talks at this year's Going Native (Link to channel9.msdn.com ). What is your opinion about the advice he gives?
Following the developement of D you get the feeling that way to many developement resources are spend on new features instead of finishing existing ones. For example a lot of work has been put into user defined attributes while other features are unfinished, unstable or not even usable (alias this, shared, export, structs). Whats your opinion on this? I agree that we should focus more on completing, streamlining, and using what we've got. This is happening already - it's been a while since quality has been at the top of our list, and the positive PR has been visibly improving.
Now that we have more resources there is some amount of parallel work we could do, and the mixed blessing with volunteer work is people work on what they find interesting, not necessarily what's best to do at the moment. This has been a challenge, but at the same time a good problem to have.
In the D bugtracker there are tons of old bugs (1 year or more) that don't get any attention but concern pretty basic language features. What do you think about adding a "old bug of the month" voting to the D developement process to pick a old bug every month that definitely gets fixed? I think that's a good idea. Even before that we should look at a "old pull request of the month". It is a shame that we are unable to accept valuable contributions at a faster rate.
Allow people to set a bug bounty using bitcoin; trust me, this will get a lot more people interested in hitherto uninteresting things. +bitcointip $1 verify. Funny you should mention that. Walter and I are talking about a bounty system, but we didn't think of bitcoins!
Lots of functional programming features are now available in languages such as C++ and D - what do you think are some of the functional features that we'll see incorporated in languages such as those over the next decade or two? There's talk about purity in C++, but beyond that I'm not sure whether there are plans for major FP-related additions.
Of the usual suspects present in FP languages, D notably misses pattern matching. It is in tension with OOP-style (first match vs. best match), and I'm not sure whether or not it's a fundamental feature of functional style. There are no plans to add such at this time.
Worst weakness? Design decision you would like to change? edit: Removed questions skeksis268 link answered. From a pure language standpoint, D's worst weakness is too little control over escaping of addresses. From a larger ecosystem standpoint, D's worst weakness has been quality of implementation.
I have been using D for a while now, and I absolutely love it. What are some improvements that are in the works that you personally are most excited about? Glad to hear that! The most important three features right now are: quality, quality, and quality. We want to finalize the language's ideas to the ultimate detail.
That said, Walter and I are excited about directions like qualifier and attribute inference (more descriptive programs with less source code). Walter has some great insights regarding the relationship between purity and uniqueness, that could greatly improve expressiveness at no cost in complexity - e.g. a pure function returning mutable data must by necessity produce fresh data. Mind = blown.
What does your typical day at Facebook look like? How much of it involves coding? Aside from reading reddit you mean? :o)
Here are a few scattered tidbits.
I use the shuttle to/from San Francisco, and I get to read or get work done on it.
A lot of what I do is coding. I need to wait long times for building and performance testing, which I try to fill by tending email, reading papers, or trying new ideas (most Facebook engineers have multiple parallel git trees so they can work on several things simultaneously.)
We don't have many meetings, and those we do have are actually interesting.
Chats every so often with coworkers.
Reviewing code, perusing internal forums.
But most of it coding.
How much more productive do you think D can really be if adopted at C++ scale? Me and other enthusiasts I know are heavily biased by our positive solo experiences with it. Productivity and its variations are difficult to measure. Build speeds alone, at one order of magnitude speedup, are dramatic enough to exert a change of paradigm. For example, many people say dynamic languages are productive because they have the "right" execution model - save file, hit Refresh. If actual times for a compiled language drop to the point of offering the same model, I think a whole category of perceptions would change.
One thing I noticed with D is its "plasticity". Once you have a body of code that works in C++, the natural tendency is to be conservative about changing it: unit testing is tenuous, subtle failure scenarios upon changes are legion, not to mention build times etc. In D, it's a lot easier to mold and remold designs are you go because you know you wont be penalized for it.
I don't like the separation between Structs and Classes in D. I think that having to avoid classes if you don't want to use the GC could force people to have implemented the same artifact as a Class and as a Struct. Also don't get why a value type / stack object is simply not allowed to use runtime polymorfism. It feels more natural when you can do such distinction when defining a variable type, as you can do in Rust. Could you defend the D approach over the C++ aproach or the Rust approach? Class vs. struct comes up occasionally as a minority opinion. Walter's basic insight was that polymorphism and value semantics rarely mix, and when they do, they rarely mix well: there are about a million ways to declare a data type handling value vs. reference semantics incorrectly in C++. Clearly there are a few legit cases (like allocating a polymorphic type on the stack and aliasing its address), and people feel robbed of that possibility. The matter of fact is that such ambiguous-gender types and their uses are rare enough, and their converse troubles are frequent enough, to warrant designing them away (you can still allocate polymorphic objects on the stack in D, it's just not dead simple, you need to use emplace).
Thanks in advance Andrei, I'm following your work since I've discovered your book Modern C++ Design! For Rust it makes more sense to tackle the problem differently because it gets to tap into its already sophisticated pointeregion paraphernalia. For D, the decision to split struct and class is, I think, a winner. We're very pleased with it.
Have you read Stepanov and McJones's "Elements of Programming"? What are your thoughts on it? Yes. It's a fine book, but (and I'm sure I'll get crucified for it) I think it's a bit on the self-important side.
Last updated: 2013-10-06 17:26 UTC
This post was generated by a robot! Send all complaints to epsy.
submitted by tabledresser to tabled [link] [comments]

A Quick Bitcoin Definition What is Bitcoin Mining? (In Plain English) - YouTube What Is Bitcoin In Tamil  Explanation Of Bitcoin In Tamil ... What is The History of Bitcoin: Super Easy Explanation ... Bitcoin Trading for Beginners (A Guide in Plain English ...

Stack Exchange network consists of 176 Q&A communities including Stack Overflow, the largest, most trusted online community for developers to learn, share their knowledge, and build their careers.. Visit Stack Exchange Urban Dictionary may have originally started as a joke, but the online authority of all things slang is now a legitimate source on what popular sayings mean — and a ton of the words published in ... Question: How much is “two bits” and where did the phrase come from? Answer: Two bits is commonly understood in America to be one quarter. The word “bit” long meant, in England, any coin of a low denomination. In early America, “bit” was used for some Spanish and Mexican coins that circulated and were worth one-eighth of a peso, or about 12 and one-half cents. This idiom is used to express the idea that any unfilled, empty spaces are unnatural and goes against the laws of physics and nature. This is based on Aristotle’s observation. Nature contains no vacuums because the denser surrounding material continuum would immediately fill the rarity of an incipient void. Dive into Bitcoin technology with this hands-on guide from one of the leading teachers on Bitcoin and Bitcoin programming. Author Jimmy Song shows Python programmers and developers how to program a Bitcoin library from scratch. You’ll learn how to work with the basics, including the math, blocks, network, and transactions behind this popular ...

[index] [18311] [7488] [32113] [33917] [23866] [17534] [16792] [34235] [19217] [12822]

A Quick Bitcoin Definition

Start trading Bitcoin and cryptocurrency here: http://bit.ly/2Vptr2X Bitcoin trading is the act of buying low and selling high. Unlike investing, which means... Start trading Bitcoin and cryptocurrency here: http://bit.ly/2Vptr2X Bitcoin is the first decentralized digital currency. All Bitcoin transactions are docume... In this video I'll explain about What Is Bitcoin In Tamil FaceBook: https://goo.gl/pp2d4E Instagram: https://goo.gl/C9zLz2 Tweet Me At: https://goo.gl/j1s2Eq... For more info on Bitcoin, visit weusecoins.com. Green Beret's Ultralight Bug Out Bag with Gear Recommendations - Duration: 18:54. The Gray Bearded Green Beret Recommended for you Baffled by bitcoin? Confused by the concept of crypto-currencies? Well, fear no more. In 190 seconds we explain what bitcoin actually is, where the idea came...

#