“Would it be pos­si­ble to ask the com­pa­nies who par­tic­i­pate in the go!pass pro­gram to tell you which go!passes are being dis­trib­uted to (i) peo­ple who are too poor to drive and (ii) crim­i­nals who have had their licenses taken away?

It would be great to be able to study the bus board­ing pat­terns of the poor peo­ple ver­sus the criminals.”

“I think we can see in these moves a com­mon his­tor­i­cal pat­tern: when the struc­tures that give a pow­er­ful insti­tu­tion strength start to weaken, it reaches for a new level of author­ity not based in the pre­vi­ous struc­ture and there­fore not sus­cep­ti­ble to weakening.”

“What makes this case par­tic­u­larly nasty is that a large com­pany suc­cess­fully forced its will on another com­pany based solely on a spe­cious claim of trade­mark infringe­ment. The next step for an ambi­tious com­pany, of course, is to demand fur­ther con­trol over how a site links to its con­tent. After all, if you can get a judge who doesn’t seem to under­stand the con­cept of hyper­link­ing, who knows what you can get away with under the guise of trade­mark infringement.”

“Along the same lines, Obama forgets—or neglects—that the Civil Rights move­ment not only called on Amer­ica to live up to its con­sti­tu­tional promises of equal­ity; it called on Amer­i­can cit­i­zens to enter into a new way of relat­ing to one another. It did so with a con­stant impa­tience with the way things were, a relent­less, increas­ingly rad­i­cal unwill­ing­ness to accept the sta­tus quo that would quite dis­turb Obama the bipar­ti­san peacemaker.”

“Another com­mon thread in the grad stu­dents’ sto­ries was dis­sua­sion, both pas­sive and active, from engage­ment with the dig­i­tal. From bureau­cratic has­sles to tepid advis­ing to being actu­ally barred from com­put­ing facil­i­ties built for fac­ulty (think about that for a moment; it’s appalling on so very many lev­els), the mes­sage goes out loud and clear: tech­nol­ogy is a toy, it’s a diver­sion, it’s fine for the class­room, but it’s not how you do your work.”

“So why is Linux every­where, and we only hear about 386BSD in his­tor­i­cal con­texts? There is exactly one answer, and it’s what Eric Ray­mond was really talk­ing about in The Cathe­dral and the Bazaar. TCatB has been seen mostly as an argu­ment for open-​​source ver­sus com­mer­cial soft­ware, but what Ray­mond saw was that the real com­pe­ti­tion comes down to an open con­tri­bu­tion model ver­sus closed con­tri­bu­tions. Linus’ promis­cu­ous con­tri­bu­tion pol­icy sim­ply let Linux out-​​evolve 386BSD. More con­trib­u­tors meant more dri­vers, more bug fixes, more enhance­ments… more ideas, ulti­mately. Two peo­ple, no mat­ter how tal­ented, can­not out­code thou­sands of Linux con­trib­u­tors. The best pro­gram­mers are 10 times more pro­duc­tive than the aver­age, and I would rate Bill and Lynne among the very best. But, as of last April, the Linux Foun­da­tion reported that more than 3,600 peo­ple had con­tributed to the ker­nel alone.”

“Archi­tects, design­ers, and devel­op­ers of cor­po­rate sys­tems usu­ally have lit­tle or no voice in what gets built, or how, or why. (Imag­ine the aver­age IT depart­ment meet­ing where one devel­oper says this sys­tem really ought to be built using Scala and Lift.) The don’t sign on, they get assigned. I know that indi­vid­ual devel­op­ers do care pas­sion­ately about their work, but usu­ally have no way to really make a difference.

The net result is that cor­po­rate soft­ware is soft­ware that nobody gives a shit about: not it’s cre­ators, not it’s investors, and not it’s users.”

“But alas, it does not. In fact and here is the cru­cial point, tenure doesn’t enable aca­d­e­mic free­dom, there is no such thing as aca­d­e­mic free­dom, what tenure does is farm the deci­sion of aca­d­e­mic free­dom out to other bod­ies. A major­ity of insti­tu­tions make tenure deci­sions based on pub­lish­ing record, in other words forces out­side the insti­tu­tion which are mak­ing mar­ket deci­sions based on what can be prof­itably sold as an intel­lec­tual com­mod­ity (usu­ally in book form) are decid­ing what aca­d­e­mics can and can­not say.”

“For­tu­nately, we are not yet “per­fected” con­sumers but if we are not vig­i­lant, our atten­tion span will con­tinue to shrink, and those avail­able con­ve­niences that help us force more and more mate­r­ial through our tiny pin­hole of focus will pro­lif­er­ate. (Just as road-​​building wors­ens traf­fic prob­lems, media-​​management and orga­ni­za­tion tools tend to exac­er­bate our atten­tion prob­lems. Hence, I spend as much time edit­ing meta­data as I do con­cen­trat­ing on music I’m lis­ten­ing to.)”

“When viewed in devel­op­ment mode, the per­son div would be over­laid with a translu­cent gray box with the words “Mile­stone 6” in it, as shown below.”

“We’ve been writ­ing a “fea­ture” for every new client request on that project – for each user-​​created ticket we han­dle, we cre­ate a .fea­ture file (and include the ticket num­ber in the fea­ture title), and write steps for that request. This means that we have accep­tance tests for all new client requests on that project. This approach may seem a lit­tle strange, but it’s been help­ful, and we’re very happy with it so far. We’ll likely take a dif­fer­ent approach if we use Cucum­ber on a project from scratch.

Now you have no excuse if your projects aren’t doing any kind of top-​​down test­ing, so get out there and write some accep­tance tests!”

“The Shoulda gem makes it easy to write ele­gant, under­stand­able, and main­tain­able Ruby tests. Shoulda con­sists of test macros, asser­tions, and helpers added on to the Test::Unit frame­work. It’s fully com­pat­i­ble with your exist­ing tests, and requires no retool­ing to use.”

“The sec­ond thing you need to do is push your uncer­tainty thresh­old.
We all have a cer­tain limit, or thresh­old, for the amount of uncer­tainty we can han­dle. For some of us, we have such a low limit, we’re afraid of even sim­ple things, like talk­ing to a stranger. We can’t pre­dict what the per­son we’ll say, so we can’t tol­er­ate the uncer­tainty. This is on the lower end of the spec­trum. The higher end of the scale might be not being able to quit your job and fol­low your pas­sion. There’s no way you can fore­see what will hap­pen, so you let uncer­tainty keep you from tak­ing action.”

“Unions paved the way to the mid­dle class for mil­lions of Amer­i­can work­ers and pio­neered ben­e­fits such as paid health care and pen­sions along the way. Even today, union work­ers earn sig­nif­i­cantly more on aver­age than their non-​​union coun­ter­parts, and union employ­ers are more likely to pro­vide ben­e­fits. And non-​​union workers—particularly in highly union­ized industries—receive finan­cial ben­e­fits from employ­ers who increase wages to match what unions would win in order to avoid unionization.”

“This expe­ri­ence has also altered my view of blog­ging and research. On one hand, I’m very enthu­si­as­tic about research in gen­eral, and my research in par­tic­u­lar, where we are reg­u­larly crack­ing con­ven­tion­ally impos­si­ble prob­lems. On the other hand, it seems that some small num­ber of peo­ple view­ing a dis­cus­sion silently decide they don’t like it, and veto it given the oppor­tu­nity. It only takes one to turn strong paper into a years-​​long odyssey, so pub­lic dis­cus­sion of research direc­tions and top­ics in a vetoc­racy is akin to vol­un­tar­ily wear­ing a “kick me” sign. While this a prob­lem for me, I expect it to be even worse for the mem­bers of a vetoc­racy in the long term.”

“Grad school was even worse. At that level, a self-​​selected bunch of fail­ure avoiders com­peted for fac­ulty approval in a pretty air­less envi­ron­ment for years. By the end, it took an act of will just to put together a declar­a­tive sen­tence. The most damn­ing insult in grad school was “naive,” which was typ­i­cally applied to any­one who actu­ally made some sort of pos­i­tive claim. (“Naive real­ism” was the worst, since it implied the unfor­giv­able sin of claim­ing to actu­ally know some­thing about some­thing.) Self-​​doubt can be taught.

In grad school, too, I recall the fac­ulty being per­plexed as to why so many doc­toral stu­dents seemed oddly hes­i­tant and overly def­er­en­tial dur­ing oral exams. At one panel of grad stu­dent papers, I recall notic­ing that every sin­gle grad stu­dent started her pre­sen­ta­tion with “this is a work in progress.” Trans­lated, that means “please don’t attack me.” These habits are learned.…”

“… The notanem­ployee exam­ple helps these peo­ple to know that some peo­ple are declar­ing that the types of rela­tion­ships in busi­ness are no longer restricted to being hier­ar­chi­cal, and that we can make a choice. And, that inde­pen­dents can work together to help the peo­ple who hire them under­stand that they can get a com­pet­i­tive edge by not try­ing to con­trol those peo­ple who choose to work with them. This makes for bet­ter rela­tion­ships, more adapt­abil­ity and flex­i­bilty, a higher chance of suc­cess. This is a real­is­tic path­way for peo­ple to begin to have the free­dom to start build­ing sys­tems that are com­mons and peer-​​based. 20 years ago, it was unheard of for inde­pen­dent busi­ness peo­ple to work together closely on cre­at­ing an ecolgy of trust, mutual respect, and learning/​participation commons.”

“The Xquartz project is an open-​​source effort to develop a ver­sion of the X.org X Win­dow Sys­tem that runs on Mac OS X. Together with sup­port­ing libraries and appli­ca­tions, it forms the X11.app that Apple has shipped with OS X since ver­sion 10.5.”

“Here’s what dupli­ca­tion removal does, struc­turally. It allows you to pull out redun­dant bits of pulp from big sec­tions, yield­ing smaller sec­tions, but the side effect is that you end up with more fas­cia. Dupli­ca­tion removal increases the ratio of fas­cia to pulp. If the amount of pulp you are able to remove exceeds the size of the fas­cia you intro­duce, the net amount of code decreases, oth­er­wise it might increase.

In gen­eral, I think that a high fas­cia to pulp ratio is bet­ter for main­te­nance. It gives us is a higher sur­face area to vol­ume ratio for our code. This can enhance testa­bil­ity and make it eas­ier to com­pose new soft­ware – we already have smaller more under­stand­able pieces.”

“This may all sound sin­is­ter and manip­u­la­tive, but the impulse behind it is get­ting peo­ple past the blind­ers that inhibit them from help­ing to shape the solu­tion. The point is to enable a con­struc­tive kind of aca­d­e­mic cit­i­zen­ship, rather than the usual dichotomy of either apa­thy or total war. Once they grasp the con­tours of what we’re up against, they’re in a posi­tion to craft actual solu­tions, and to defend their own inter­ests more effec­tively. I want that to hap­pen, since I can’t help but think that we’re smarter together than separately.”

“…But for bank­ing we do need an Atabeg. He (the kind of per­son we’re look­ing for will prob­a­bly be male) should work out of a com­plex of silk tents in Wyoming, which shiv­er­ing sup­pli­ants must reach on foot. The bankers and, why not, deriv­a­tives traders should plead for mercy under the horse­tail ban­ners stream­ing in the steppe wind, close to the white pyra­mid of bleached skulls.”

“I recently got back reviews of a paper in which I used auto­matic dif­fer­en­ti­a­tion. Therein, a reviewer clearly thought I was using finite dif­fer­ence, or “numer­i­cal” dif­fer­en­ti­a­tion. This has led me to won­der­ing: Why don’t machine learn­ing peo­ple use auto­matic dif­fer­en­ti­a­tion more? Why don’t they use it…constantly? Before reck­lessly spec­u­lat­ing on the answer, let me briefly review what auto­matic dif­fer­en­ti­a­tion (hence­forth “autod­iff”) is. Specif­i­cally, I will be talk­ing about reverse-​​mode autodiff.”

“Call to Action: We there­fore urge every U.S. law school to com­mit to end­ing print pub­li­ca­tion of its jour­nals and to mak­ing defin­i­tive ver­sions of jour­nals and other schol­ar­ship pro­duced at the school imme­di­ately avail­able upon pub­li­ca­tion in sta­ble, open, dig­i­tal for­mats, rather than in print.”

““We’ve decided to close your office. We think we might be able to find you another job. Hmmmm.…maybe. Not sure about your staff and patients, but maybe. We’ll see. Gotta go HIRE SOME CARDIOLOGISTS!” (I’m soooooooooooo not kid­ding.) “Thanks for tak­ing it on the chin and not cry­ing like a girl! I’ll be in touch soon, I promise. Bye. Can’t keep the heart guys waiting.””

“I wish you had omit­ted … all those Expres­sions of Resent­ment against your Adver­saries.… In such Cases, the noblest Vic­tory is obtained by Neglect, and by Shin­ing On.”

“Now, I’m in no posi­tion to refute those fig­ures, but I don’t think it takes an eco­nom­ics expert to look at them and realise why the pub­lish­ers are strug­gling at the moment; if their analy­sis peo­ple can only shave off $2 per unit by remov­ing the print­ing, ship­ping, ware­hous­ing and remain­der­ing from the equa­tion, then there’s a busi­ness model that was on shaky ground before the ebook entered the pic­ture. I sus­pect the bits I’ve bolded are where the haem­or­rhag­ing could be stemmed most effectively.”

“A buccaneer-​​scholar is any­one whose love of learn­ing is not muz­zled, yoked or shack­led by any insti­tu­tion or author­ity; whose mind is dri­ven to wan­der and find its own voice and place in the world.”

[indi­rect but key]
“Prac­tice, prac­tice, prac­tice!
Don’t con­fuse expe­ri­ence with exper­tise.
Don’t trust folk­lore — but learn it any­way.
Take noth­ing on faith. Own your method­ol­ogy.
Drive your own edu­ca­tion — no one else will.
Rep­u­ta­tion = Money. Build and pro­tect your rep­u­ta­tion.
Relent­lessly gather resources, mate­ri­als, and tools.
Estab­lish your stan­dards and ethics.
Avoid cer­ti­fi­ca­tions that triv­i­al­ize the craft.
Asso­ciate with demand­ing col­leagues.
Write, speak, and always tell the truth as you see it.”

“She said that Mex­ico and Canada will merge with us and that a new, open source dol­lar called the Amero is going to replace the dol­lar. But the most scary thing is what she told me she’s been doing for the past cou­ple of years. She’s been over­see­ing the con­struc­tion of Learn­ing Object Repos­i­to­ries being built all through­out America.”

“Com­bin­ing these tech­niques can net us some very sophis­ti­cated data manip­u­la­tions. Below is a form that will update a Cre­ator, his address, updates three exist­ing wid­gets adds one wid­get, and deletes two widgets.”

“The Depres­sive Pes­simist will com­plain that the task that they’re doing isn’t enjoy­able, and make state­ments doubt­ing the group’s abil­ity to succeed.

The Jerk will say that other people’s ideas are not ade­quate, but will offer no alter­na­tives him­self. He’ll say “you guys need to lis­ten to the expert: me.”

The Slacker will say “what­ever”, and “I really don’t care.”

gen­er­al­iz­able for many domains, not just soft­ware devel­op­ment: “How­ever, as a pro­gram evolves, there’s a good chance that the Design In Code will not include all the good things we now under­stand. We have a bet­ter Design In Head. When the design in our head is enough bet­ter than the one in the code, it can pay off to bring the code closer to what we now understand.”

“Come on peo­ple, don’t believe that you can cre­ate a world where the bits in the sys­tem are a per­fect mir­ror of who you are. Indi­vid­ual iden­ti­ties are not some­thing that can be reduced to micro­for­mats. I can change my mind, and not have to go back and update a zil­lion web pages to reflect that change of mind. What­ever you are call­ing “iden­tity” here is emphat­i­cally not what human beings think of as their iden­tity; per­haps if you replaced it with “dossier” the nature of the data gath­er­ing would be more clear in a his­tor­i­cal con­text (think Stasi, for instance, instead of Facebook).”

“So that’s my guess about why Hulu blocked Boxee: those ads you see on Heroes are higher mar­gin when you see them on your TV than when you see them on Hulu, and the only rea­son they’re on Hulu is to make money from Heroes when you watch it online, so Apple or Google doesn’t make that money instead. They were meant for your “portable com­put­ing devices” and not your pre­cious TV. Now go back to the couch until we call for you again.”

“Well. As Dave Gray points out, “An unbook’s com­mu­nity is a very real part of the unbook’s devel­op­ment team.” I wouldn’t nec­es­sar­ily have used the phrase “devel­op­ment team,” for the obvi­ous rea­sons, but the point stands. Your voice is a part of this book we’re writ­ing, and not the least sig­nif­i­cant. What do you think?”