|
hackbunny posted:an investigative journalist turned his skin black (temporarily) to live as an african-american for a few weeks. he wrote a book about his experience titled Black Like Me i guess kinda like Günter Wallraff pretending to be a Turkish "guest worker" in Germany 1985?
|
# ? Jun 19, 2017 00:32 |
|
|
# ? May 22, 2024 09:58 |
|
Dad, Dad, why did you let that man Push you around like that? You should have beat him down Down to the ground Down to the ground for that He said son, you're still young And you always jump the gun There's real people in the big, big trucks That you flip off when they get in your way You get so hacked that you pay no mind To the great big sign that says 'oversize load' Do you really think they can go as fast As you in your '87 Trans Am? They know you're in such a terrible rush But they're going just as fast, as fast as they can Dad, Dad, I really don't understand What driving big trucks has to do with that man You shoulda' taught him a lesson about being rude About talking to you with such an attitude He said son, you're still young And you always jump the gun There's real people in the big, big trucks That you flip off when they get in your way You get so hacked that you pay no mind To the great big sign that says 'oversize load' Do you really think they can go as fast As you in your '87 Trans Am? They know you're in such a terrible rush But they're going just as fast, as fast as they can
|
# ? Jun 19, 2017 00:34 |
|
Plorkyeran posted:google seems to do this all the time, so we just need to convince them to use a bit more scientific rigor in conclusion, the choice of programming language is irrelevant to the success of the project, because all efforts are doomed to founder in a swamp of incoherent project management and increasingly bureaucratic interdepartmental sniping about code style
|
# ? Jun 19, 2017 00:55 |
|
rjmccall posted:in conclusion, the choice of programming language is irrelevant to the success of the project, because all efforts are doomed to founder in a swamp of incoherent project management and increasingly bureaucratic interdepartmental sniping about code style
|
# ? Jun 19, 2017 01:17 |
|
rjmccall posted:in conclusion, the choice of programming language is irrelevant to the success of the project, because all efforts are doomed to founder in a swamp of incoherent project management and increasingly bureaucratic interdepartmental sniping about code style
|
# ? Jun 19, 2017 01:22 |
|
cis autodrag posted:beat up 90s gms are the official car of the ghetto because you can buy em for 500 bux cash and fix them with junk yard parts yourself. they attract police attention. dwb is a crime in most major cities. when i was a bit younger it was 1980s american full-size sedans. i got SO many parking lot offers to buy my old 83 caprice classic back when i had it
|
# ? Jun 19, 2017 01:49 |
|
rjmccall posted:in conclusion, the choice of programming language is irrelevant to the success of the project, because all efforts are doomed to founder in a swamp of incoherent project management and increasingly bureaucratic interdepartmental sniping about code style this has been my position all along Here's a throwback to one of my favorite white papers from a consulting group that's definitely not the most trustable thing, but does align with my points super well so why the hell not. MononcQc posted:Related: The Impact of Business Requirements on the Success of Technology Projects by IAG.
|
# ? Jun 19, 2017 02:32 |
|
sometimes you want eval and sometimes you want it to be harder to break poo poo
|
# ? Jun 19, 2017 02:34 |
|
posting on the hacker page
|
# ? Jun 19, 2017 02:40 |
|
MononcQc posted:tl;dr: if IAG's stuff is to be trusted, your project is likely doomed before it even starts. Even when you've got 'excellent' requirements and analysis, ~30% of your projects will still fail. This is in big businesses, but projects that impact fewer people aren't immune to this at all. http://www.systemsguild.com/pdfs/DeMarcoNov2011.pdf If a project offered a value of 10 times its estimated cost, no one would care if the actual cost to get it done were double the estimate. On the other hand, if expected value were only 10 percent greater than expected cost, lateness would be a disaster. Yes it would be a disaster, but instead of obsessing over “What’s the matter with those software folks who didn’t deliver on the schedule we gave them?” we need to ask instead “Why did we ever kick off a project with such marginal expected value?”
|
# ? Jun 19, 2017 02:57 |
|
Bloody posted:posting on the hacker page
|
# ? Jun 19, 2017 02:59 |
|
tef posted:http://www.systemsguild.com/pdfs/DeMarcoNov2011.pdf I also liked the way The Systems Bible (Systemantic) put it: "If it's worth doing at all, it's worth doing poorly", which I figure is meant to be interpreted as "the solution is so worthwhile that even if it's done poorly, it still has value".
|
# ? Jun 19, 2017 03:10 |
|
tef posted:If a project offered a value of 10 times its estimated cost, no one would care if the actual cost to get it done were double the estimate. lol
|
# ? Jun 19, 2017 03:59 |
|
i just recently finished up a project which is projected to earn back 13 times its estimated cost, and one of the things that nearly caused a project delay was a multi-month long fight over adding 25 cents to the bom cost (total bom cost was $41) because it wasnt in the original estimate
|
# ? Jun 19, 2017 04:07 |
|
the beancounters are running the asylum
|
# ? Jun 19, 2017 04:09 |
|
MononcQc posted:this has been my position all along Here's a throwback to one of my favorite white papers from a consulting group that's definitely not the most trustable thing, but does align with my points super well so why the hell not. this is interesting but i was specifically laughing at google
|
# ? Jun 19, 2017 05:05 |
|
i agree that ba's are champions and that everyone needs more of them
|
# ? Jun 19, 2017 07:43 |
|
Bloody posted:posting on the hacker page
|
# ? Jun 19, 2017 08:30 |
|
MALE SHOEGAZE posted:sometimes you want eval and sometimes you want it to be harder to break poo poo certainly helps a lot if the system is designed in a way where it is clear to one and all that "this part is the skeleton of the system and must not be done haphazardly as it'll live forever and be used by all sorts of things" vs. "this part is the little specific realization i need for this task, and it is no big deal if i hack it up as it is an addition off to the side" in fact it is a pretty deep annoyance when things get constrained to a level where the second case cannot even be argued, where one has strict rules applied equally to all parts, as it usually results in poor compliance in all parts
|
# ? Jun 19, 2017 09:09 |
|
MononcQc posted:this has been my position all along Here's a throwback to one of my favorite white papers from a consulting group that's definitely not the most trustable thing, but does align with my points super well so why the hell not. hey so I want to read this paper but the link is a 404 know of anywhere else to get it? google isn't really helping me here
|
# ? Jun 19, 2017 15:07 |
|
Arcsech posted:hey so I want to read this paper but the link is a 404 Oh that's an old post and I didn't check. http://docdro.id/q4GiBAk should be a valid mirror.
|
# ? Jun 19, 2017 15:27 |
|
Powaqoatse posted:i guess kinda like Günter Wallraff pretending to be a Turkish "guest worker" in Germany 1985? or Fabrizio Gatti posing as an iraqi refugee to work as a tomato picking slave, and later as a kurdish refugee to document the immigration management system from the inside
|
# ? Jun 19, 2017 15:47 |
Bloody posted:posting on the hacker page
|
|
# ? Jun 19, 2017 16:47 |
|
compiler question: if you were interested in maximizing your trust in the compiler and its generated output, which optimization level would you use
|
# ? Jun 19, 2017 19:35 |
|
having to step through -O0 code is like watching the worst pedant imaginable interpret your code as slow as possible sure, you can see every arg in a backtrace, but AT WHAT COST??
|
# ? Jun 19, 2017 19:53 |
|
Bloody posted:compiler question: if you were interested in maximizing your trust in the compiler and its generated output, which optimization level would you use there is no way to completely avoid having your broken, fraudulent code get miscompiled. even at -O0 most c/c++ compilers will do at least a few things that are only valid because of the u.b. rules, like re-using stack memory after it goes out of scope with that said, inlining is the main thing that creates new problems at different optimization settings, because it's the main thing that creates new optimization opportunities. generally the optimizer will optimize something if it's happening within a single function and there's nothing in the function which blocks the optimization. inlining both makes brings more code into a single function and removes barriers that block optimizations. higher optimization settings generally correspond to more inlining. whole-module optimization is another thing to watch out for because code that's happily worked for years can suddenly get miscompiled when something gets inlined across files for the first time if you're worried less about u.b. miscompilations but more about compiler bugs, that's different. stay away from gcc's -O3, i guess, which tends to have more experimental optimizations in it. otherwise compiler bugs tend to be in code that isn't specific to any particular optimization setting; they just lurk around waiting for the wrong combination of operations to trigger them the main takeaway here is to never turn on more powerful optimizations right before release
|
# ? Jun 19, 2017 20:18 |
|
rjmccall posted:in conclusion, the choice of programming language is irrelevant to the success of the project, because all efforts are doomed to founder in a swamp of incoherent project management and increasingly bureaucratic interdepartmental sniping about code style [edit] frperg unkbe zrffntr sbe gru yrrg
|
# ? Jun 20, 2017 07:47 |
John Big Booty posted:This is mostly true, but then again coldfusion are we posting rot13 now or something
|
|
# ? Jun 20, 2017 08:20 |
|
hackbunny posted:or Fabrizio Gatti posing as an iraqi refugee to work as a tomato picking slave, and later as a kurdish refugee to document the immigration management system from the inside But what programming language did he use?
|
# ? Jun 20, 2017 11:38 |
|
cinci zoo sniper posted:are we posting rot13 now or something bayl gur ryvgr ba gur unpxre cntr.
|
# ? Jun 20, 2017 15:56 |
|
eww, usenet is leaking
|
# ? Jun 20, 2017 18:37 |
|
redleader posted:i agree that ba's are champions and that everyone needs more of them This can be true, otoh we have zero ba's and it works well too
|
# ? Jun 20, 2017 19:58 |
|
cinci zoo sniper posted:are we posting rot13 now or something don't make me winnuke u
|
# ? Jun 20, 2017 23:15 |
|
MALE SHOEGAZE posted:sometimes you want eval and sometimes you want it to be harder to break poo poo That's why you have strongly-typed eval.
|
# ? Jun 21, 2017 08:29 |
|
Zemyla posted:That's why you have strongly-typed eval. that's... creative
|
# ? Jun 21, 2017 17:06 |
|
Zemyla posted:That's why you have strongly-typed eval.
|
# ? Jun 21, 2017 19:23 |
|
The GHC runtime has within it a bytecode interpreter, which is normally used to make the built-in REPL work. Since GHC is written in Haskell, the hint library simply imports the same libraries and provides a programmer-usable interface to them. What's more: Because the eval function requires the result to be a specific type, and requires you to explicitly import the libraries and functions you want visible from the expression you're evaluating, it's vastly safer than the equivalent in almost any other language.
|
# ? Jun 21, 2017 21:48 |
|
Bloody posted:posting on the hacker page
|
# ? Jun 25, 2017 15:50 |
|
thank you golang for making incorrect floating point rounding a problem in tyool 2017 https://www.cockroachlabs.com/blog/rounding-implementations-in-go/
|
# ? Jul 11, 2017 15:03 |
|
|
# ? May 22, 2024 09:58 |
|
ruby almost changed the default rounding mode in the 2.4 release that would have been fun i imagine
|
# ? Jul 11, 2017 15:31 |