“And what of these lists? They’re long, you never get to the end of them, and half the time the tasks on the list never get done. While it feels good to check items off the list, it feels hor­ri­ble hav­ing items that never get checked off. This is all use­less spend­ing of men­tal energy, because none of it gets you any­where.
The only thing that mat­ters is the actual doing.”

“Lower Uppers are doc­tors, accoun­tants, engi­neers and lawyers. At com­pa­nies they’re mostly peo­ple above the rank of vice pres­i­dent and below the CEO. Their com­rades include well-​​fed mem­bers of the media (and even part-​​time Post colum­nists who earn their liv­ings as con­sul­tants). They include gov­ern­ment offi­cials — and, yes, SEC lawyers — who didn’t make or inherit for­tunes before enter­ing pub­lic ser­vice. Lower Uppers are pro­fes­sion­als who by dint of edu­ca­tion, hard work and good luck are liv­ing bet­ter than 99 per­cent of any­one who has ever walked the planet. They’re also peo­ple who can’t help but notice how many peo­ple with cre­den­tials much like their own seem to be liv­ing in the kind of Gatsby-​​like splen­dor they’ll never enjoy.”

“The main attrac­tion (and last pre­sen­ta­tion) was from my friend Gary Bernhardt1. Gary spent a few months con­tract­ing on a Rails project for four hours in the morn­ing and a Django project for four hours in the after­noon. So when he speaks of either, it’s not in ignorance.

At the end, I couldn’t find a sin­gle crit­i­cism of Ruby that I dis­agreed with. I did pick up a new appre­ci­a­tion for unique fea­tures of the Ruby lan­guage that I pre­vi­ously took for granted.”

“Staged self-​​assembly with RNA removal is a model of tile-​​based algo­rith­mic self-​​assembly that was intro­duced by Abel, Ben­bernou, Damian, Demaine, Demaine, Flat­land, Komin­ers and Schweller (Shape Repli­ca­tion through Self-​​Assembly and RNase Enzymes, SODA 2010) and is a model that allows for the peri­odic removal of all tiles in a given assem­bly that belong to a spe­cially des­ig­nated group of (RNA) tiles. In this paper, we study the self-​​assembly of arbi­trary shapes in staged assem­bly sys­tems with RNA removal. We ana­lyze the per­for­mance of our assem­bly sys­tems with respect to their tile com­plex­ity, stage com­plex­ity as well as the scale fac­tor, con­nec­tiv­ity and address­abil­ity of the uniquely pro­duced final assembly.”

“For me, being a part of the Ruby com­mu­nity feels like get­ting a sneak peek at where soft­ware devel­op­ment is going six months to two years from now. That’s not to say it’s all rain­bows and daisies…the con­stantly chang­ing land­scape requires a real pas­sion­ate ded­i­ca­tion to keep up or you’ll quickly fall behind, and not all tech­nolo­gies are meant to imme­di­ately be deployed to massive-​​scale pro­duc­tion envi­ron­ments (restraint is a skill a good Ruby­ist must learn and exer­cise on a reg­u­lar basis). But I love Ruby because I feel con­fi­dent that I will be made aware of trends in soft­ware devel­op­ment long before I would oth­er­wise be expected to under­stand them.”

“This strongly sug­gests that con­ser­v­a­tives face epis­temic clo­sure, at least on this issue. The more con­ser­v­a­tives ‘know,’ the more likely they are to be wrong.”

“Igno­rance is noth­ing to be ashamed of because we are all, in vary­ing degrees, igno­rant about many things. The impor­tant choice as indi­vid­u­als and as a soci­ety is adopt­ing an epis­te­mol­ogy of rational-​​scientific-​​empiricism that, if steadily applied, allows us to chip away at our igno­rance and become aware of our errors and solve prob­lems. On the other hand, adopt­ing a pos­ture of bel­liger­ent, stub­born, defense of our own igno­rance by evad­ing facts, logic and the con­clu­sions drawn from the evi­dence of expe­ri­ence is the road to cer­tain disaster.”

“Siemens’s plans hinge on a new design that reduces the weight of the system’s gen­er­a­tor. In con­ven­tional wind tur­bines, the gear­box increases the speed of the wind-​​driven rotor sev­eral hun­dred fold, which rad­i­cally reduces the size of the gen­er­a­tor required. Direct-​​drive gen­er­a­tors oper­ate at the same speed as the turbine’s blades and must there­fore be much bigger–over four meters in diam­e­ter for Siemens’s three-​​megawatt tur­bine. Yet Siemens claims that the turbine’s entire nacelle weighs just 73 met­ric tons–12 tons less than that on its less pow­er­ful, gear-​​driven 2.3-megawatt turbines.”

“Com­pu­ta­tional com­plex­ity has in the past adapted well to new com­pu­ta­tion mod­els from the PRAM to bio­log­i­cal and quan­tum com­put­ers. But we are see­ing new com­put­ing par­a­digms in mul­ti­core and cloud com­put­ing and the­ory seems late to the party. There was a nice SODA paper on MapRe­duce, the basic cloud com­put­ing oper­a­tion, but for the most part the­o­rists haven’t tack­led the cloud com­put­ing model and only a few have looked at mul­ti­core. The­ory can say much about new com­pu­ta­tional meth­ods, both in how we can take advan­tage of them and what they can’t do, but only if we make the effort to develop the proper mod­els to cap­ture these new approaches.”

“What is strik­ing is that caveat emp­tor arises as a legal prin­ci­ple mainly because of the tan­gle the courts would get into if they tried to enforce a more ambi­tious stan­dard of right and wrong.
Chief Jus­tice Marshall’s logic surely applies with even greater force to mod­ern deals between invest­ment banks and sophis­ti­cated qual­i­fied investors, both of which will be simul­ta­ne­ously work­ing on many deals, each involv­ing sen­si­tive pro­pri­etary information.”

“BayesMo­tel is a multi-​​variate Bayesian clas­si­fi­ca­tion engine. There are two steps to Bayesian classification:

Train­ing You pro­vide a set of vari­ables along with the proper clas­si­fi­ca­tion for that set.
Run­time You pro­vide a set of vari­ables and ask for the proper clas­si­fi­ca­tion accord­ing to the train­ing in Step 1.
Com­monly this is used for spam detec­tion. You will pro­vide a cor­pus of emails or other data along with a “Spam/​NotSpam” clas­si­fi­ca­tion. The library will deter­mine which vari­ables affect the clas­si­fi­ca­tion and use that to judge future data.”

“Through most of the cen­tury, brack­ets were much closer to a con­tin­u­ous scale. There was a big shift in thought though in the 1980s, when Ronald Rea­gan was elected pres­i­dent. The brack­ets became much more dis­tinct. The idea has more or less stuck over the past two decades.”

“Node.js is evented I/​O for JavaScript, built on top of the blaz­ingly fast V8 engine. It makes han­dling event-​​driven I/​O incred­i­bly sim­ple, and aligns per­fectly with our mani­a­cal focus on sim­plic­ity and devel­oper pro­duc­tiv­ity. The Ruby com­mu­nity has quickly adopted node, and with great rea­son. Com­pli­ment­ing exist­ing apps with node.js for com­po­nents that require real-​​time event han­dling or mas­sive con­cur­rency is both easy and ele­gant – in part thanks to the avail­abil­ity of frame­works such as express.”

“Just as the auto-​​industry is frag­ment­ing as it becomes clear the majors can’t sat­isfy niche demand pro­pelled by per­son­al­iza­tion, the mobile phone indus­try knows that the suc­cess of Apple’s iPhone is a sign that cus­tomiza­tion and per­son­al­iza­tion are now more impor­tant than engi­neer­ing excel­lence. No man­ager, exec­u­tive or designer is unaware of it – though we’ve seen a suc­ces­sion of ultra-​​high func­tion­al­ity phones launched recently, with a phone-​​based Sony playsta­tion device to come.”

“Par­al­lel Global Opti­miza­tion Algo­rithms (PGOA) pro­vide an effi­cient way of deal­ing with hard opti­miza­tion prob­lems. One method of par­al­leliza­tion of GOAs that is fre­quently applied and com­monly found in the con­tem­po­rary lit­er­a­ture is the so-​​called Island Model (IM). In this paper we ana­lyze the impact of the migra­tion topol­ogy on the per­for­mance of a PGOA which uses the Island Model. In par­tic­u­lar we con­sider par­al­lel Dif­fer­en­tial Evo­lu­tion and Sim­u­lated Anneal­ing with Adap­tive Neigh­bor­hood and draw first con­clu­sions that emerge from the con­ducted experiments.”