prompt
stringlengths 4
3.12k
| chosen
stringlengths 3
10.5k
| rejected
stringlengths 3
8.98k
|
|---|---|---|
ask hn: what should i answer to a question about salary?
i am a fresh graduate trying to find a junior developer job.
i often get asked in interviews how much i think i should be paid.<p>how can i know that? what is the best answer i can give so that i won't overestimate and underestimate myself at the same time? | i would answer with "i expect my salary to fall in line with the current market rate" for wherever you live. i would put more of the impetus on them, because giving a number to them rarely works in your favor. i would also try to delay money talk until you are certain you want to work there and they want you to work there. as has been suggested if you want to get an idea of the current market rate glassdoor and other sites will help but may be sparse for your market. | you should know what is fair, and you can get some good feedback likely if you provide your geographic area.in general, when negotiating salary make an employer put a price tag on you first, then if it is fair, you can go for it or negotiate a little more, ask for extra time off etc. if it is a low offer, they may be testing you to see if you know the market, don't be afraid to come back at a fair market value for new grad in your area.and don't accept an offer below market with some unwritten promise to increase your pay if you do xyz. that isn't fair to you and the employer is trying to take advantage of you. if you get a fair offer and you ask for say $2k more, and they say well if you execute over the first 90 days with x y z and those are tangible real goals, not subjective, then that is reasonable. just make sure it is in writing in the offer letter. |
ask hn: what should i answer to a question about salary?
i am a fresh graduate trying to find a junior developer job.
i often get asked in interviews how much i think i should be paid.<p>how can i know that? what is the best answer i can give so that i won't overestimate and underestimate myself at the same time? | you should know what is fair, and you can get some good feedback likely if you provide your geographic area.in general, when negotiating salary make an employer put a price tag on you first, then if it is fair, you can go for it or negotiate a little more, ask for extra time off etc. if it is a low offer, they may be testing you to see if you know the market, don't be afraid to come back at a fair market value for new grad in your area.and don't accept an offer below market with some unwritten promise to increase your pay if you do xyz. that isn't fair to you and the employer is trying to take advantage of you. if you get a fair offer and you ask for say $2k more, and they say well if you execute over the first 90 days with x y z and those are tangible real goals, not subjective, then that is reasonable. just make sure it is in writing in the offer letter. | there are a few ways to try and discover what your market rate is.the first method is a set of websites or statistics that you can find pretty easily. you'll hear glassdoor and salary.com as a couple examples. some schools might have data on their past graduates. i know that carnegie mellon [1] publishes their data, which includes the employer a graduate joined, the location, and some compensation data.unfortunately, much of what you'll find is anonymously self-reported, and when people self-report salary information they are probably more likely to exaggerate up than exaggerate down.a second method of getting numbers on market rate is from talking to people. for you, that may be other graduates who have already landed jobs or people that graduated last year who may be willing to share what they were paid. recruiters who work local markets are also a good resource for this information, as they actually see the job offers and are likely to have more accurate figures than the self-reported data seen on salary websites.as for overestimating or underestimating, most people here are going to tell you to never give a number first. negotiation is described as having a winner and a loser, and if you don't get the maximum amount someone was willing to spend than you lost. there isn't a clear way to determine how much someone would have spent, so approaching negotiation in this win/lose manner isn't all that useful.someone who knows the market rate for their experience and skills is in a much better position, and someone with a long track record of successful employment that knows their market rate is completely set and more able to use tactics in negotiating that are being offered to you here in this thread. unfortunately, you (as an entry-level candidate) won't typically have that same amount of leverage or power in negotiations, as entry-level candidates are generally considered more homogenous.in other words, if a company finds an entry-level candidate being somewhat 'difficult' (for lack of a better word), they may take the attitude that there may be 10 other entry-level candidates who won't be. i'm not saying you shouldn't negotiate - in many cases you should, and in some cases you shouldn't (most obvious being when a company offers you what you requested or even more than you requested).last bit of advice - if and when you give your number, don't follow it with anything. entry-level candidates have a tendency to state a number, and if there is even a few seconds of delay by the other party they add "but that is negotiable". don't do that. there is an article on that a few weeks ago [2] that explains a bit more.1 - <link> - <link> |
ask hn: what should i answer to a question about salary?
i am a fresh graduate trying to find a junior developer job.
i often get asked in interviews how much i think i should be paid.<p>how can i know that? what is the best answer i can give so that i won't overestimate and underestimate myself at the same time? | there are a few ways to try and discover what your market rate is.the first method is a set of websites or statistics that you can find pretty easily. you'll hear glassdoor and salary.com as a couple examples. some schools might have data on their past graduates. i know that carnegie mellon [1] publishes their data, which includes the employer a graduate joined, the location, and some compensation data.unfortunately, much of what you'll find is anonymously self-reported, and when people self-report salary information they are probably more likely to exaggerate up than exaggerate down.a second method of getting numbers on market rate is from talking to people. for you, that may be other graduates who have already landed jobs or people that graduated last year who may be willing to share what they were paid. recruiters who work local markets are also a good resource for this information, as they actually see the job offers and are likely to have more accurate figures than the self-reported data seen on salary websites.as for overestimating or underestimating, most people here are going to tell you to never give a number first. negotiation is described as having a winner and a loser, and if you don't get the maximum amount someone was willing to spend than you lost. there isn't a clear way to determine how much someone would have spent, so approaching negotiation in this win/lose manner isn't all that useful.someone who knows the market rate for their experience and skills is in a much better position, and someone with a long track record of successful employment that knows their market rate is completely set and more able to use tactics in negotiating that are being offered to you here in this thread. unfortunately, you (as an entry-level candidate) won't typically have that same amount of leverage or power in negotiations, as entry-level candidates are generally considered more homogenous.in other words, if a company finds an entry-level candidate being somewhat 'difficult' (for lack of a better word), they may take the attitude that there may be 10 other entry-level candidates who won't be. i'm not saying you shouldn't negotiate - in many cases you should, and in some cases you shouldn't (most obvious being when a company offers you what you requested or even more than you requested).last bit of advice - if and when you give your number, don't follow it with anything. entry-level candidates have a tendency to state a number, and if there is even a few seconds of delay by the other party they add "but that is negotiable". don't do that. there is an article on that a few weeks ago [2] that explains a bit more.1 - <link> - <link> | this is pretty much the definitive answer as far as i'm concerned:<link> |
ask hn: what should i answer to a question about salary?
i am a fresh graduate trying to find a junior developer job.
i often get asked in interviews how much i think i should be paid.<p>how can i know that? what is the best answer i can give so that i won't overestimate and underestimate myself at the same time? | this is pretty much the definitive answer as far as i'm concerned:<link> | while it's good to understand the market prices for your skills, negotiation is a lot about understanding the other party's goals and needs and best-guessing their maximum price. they will have a maximum price in their head. so, determine that maximum price.assume that since you've gotten to the negotiation phase, you've exceeded their expectations.ask, 'how many people are you hiring?' if it's more than one, at the same position, then they may have more demand. it also may imply that their budget for hiring is above average.ask, 'how difficult has your interviewing been?' if it's been difficult, and you've reached this stage, then your in a better position for negotiation.ask, 'what have been your main criteria in hiring?' if it points to a rare skill that you have or some background experience or knowledge that is unique, then you're in a better position.ask, 'are there any other candidates that you're interviewing, and how do i measure up?' if there are no other candidates, and/or you're the best candidate that they have, then you're in a better position.ask, 'do you have upcoming milestones or deadlines?' if so, then may have higher immediate demand.ask, 'what are your goals in hiring me?' if they want to put you to work immediately, then you're in a better position. if they are primarily interested in grooming you, they may have more patience to hire.ask, 'who approves the hiring, and are you able to hire me today?' if someone else approves, it may be a harder negotiation. if they can hire you immediately, then you're in a better position.finally, don't set your price, first. if they ask, 'what is your price?', ask them 'what is your typical range for this type of position?' get them to offer their price first because it's always on the lower end of what their range is. then, consider your position. ask for something that is slightly outrageous. for example, let's say you determine that you're in a good position. they offer you 80,000. say that you were thinking more in the 120,000 range. read their reaction to that. if they seem offended, say 'but, that can be negotiable, based on your benefits, etc.' then, talk to them about the other benefits of the position, and come down slightly. but, once your initial price is set, that will be the baseline for further discussions.always come to the table with as much information as possible.always ask for add-ons. ask for more non-financial benefits. ask for more vacation. ask for a more flexible schedule. if it's a big company, you're not in a good position to ask for extras, however.tell them that you'll give them an answer very soon, and set a specific date. it will show that you're serious. if you plan wisely, you can have multiple offers from different companies at similar time. use these other offers to both gauge the demand and to play each offer against other offers.and, remember, it's always supply and demand. |
test your email server encryption
| i just checked one of my email providers and i get the following results:- hostname: mail.example.com- starttls: supported- certificates: www.spamfirewall.at -certificate does not match hostname certificate chain:
www.spamfirewall.at
337 days remaining 2048 bit sha1withrsaencryption
root certificate unknown
- perfect forward security: unsupported- dane: missingi use a ssl certificate for my domain example.com. is there something i have to do to fix that or is this only something the provider needs to take care of? | i think i win? <link> |
test your email server encryption
| i think i win? <link> | not loving the "shame" list. it would discourage me from trying a public domain on the site. |
test your email server encryption
| not loving the "shame" list. it would discourage me from trying a public domain on the site. | see also (<link> |
test your email server encryption
| see also (<link> | i can send an e-mail to them but they cannot connect to my server. it times out on starttls.nevermind. i penalize suspect servers with a 30 seconds delay and they triggered it. h=ssl-tools.net (ssl-trust.com) [91.202.41.201] warning: remote host presented unverifiable helo/ehlo greeting. |
breaking the silk road's captcha
| fyi: captchas are generally considered "broken" at between 1% and 10% rates of success with automated approaches, because attackers can run hundreds of thousands of requests, generally "for free" at the margin. there is no practical difference in the amount of abuse suffered by a site with a 90% captcha and a 9% captcha -- the first one just requires 10x as many http requests to abuse.this is one of the unfortunate "math favors the bad guy" consequences in a lot of anti-abuse filtering tasks. (anti-spam research has similar problems, which is why the main innovation wasn't making filters better but radically increasing the cost of getting caught, via burning the reputation of the offending ip. ip addresses are a lot more expensive to acquire in quantity than packets.) | this a great writeup and super easy to follow along with. the figures are really nice!one observation: training a neural net to classify segmented characters is probably overkill. the author observed that the font never changed, but never ended up exploiting this fact. after the very effective preprocessing, thresholding, etc the characters are almost identical to the 'average' representations the author generated!i bet it would be enough simply to classify an unknown character by the letter that it shows highest correlation with. |
breaking the silk road's captcha
| this a great writeup and super easy to follow along with. the figures are really nice!one observation: training a neural net to classify segmented characters is probably overkill. the author observed that the font never changed, but never ended up exploiting this fact. after the very effective preprocessing, thresholding, etc the characters are almost identical to the 'average' representations the author generated!i bet it would be enough simply to classify an unknown character by the letter that it shows highest correlation with. | very cool and well-written post.i've created many similar programs to defeat captcha's. i would classify this as a medium severity bug, you would still need to brute force the passwords on a terribly slow and intermittent connection. |
breaking the silk road's captcha
| very cool and well-written post.i've created many similar programs to defeat captcha's. i would classify this as a medium severity bug, you would still need to brute force the passwords on a terribly slow and intermittent connection. | author here. here's the corresponding proggit thread: <link> |
breaking the silk road's captcha
| author here. here's the corresponding proggit thread: <link> | that was a surprisingly simple and easy-to-follow write-up. i will have to try some captcha-breaking for myself soon. |
apple promised an expansive health app, so why can't women track menstruation?
| while i (just like the vast majority of readers of this comment) am not menstruating myself, the small sample of my three sisters (and, as it were, many of their friends) seem to indicate that yes, tracking that information is more relevant to half of the human population than any other health metric is for men.if the question is "why is there no menstruation tracker", the answer (really, the elephant in the room) is somewhere here: <link> course, the unreleased data about gender diversity amongst apple engineers, which we will hopefully get within two years by emulation thanks to google's own interest in the problem, can easily be guessed as well. | do engineers at apple know what menstruation is? |
apple promised an expansive health app, so why can't women track menstruation?
| do engineers at apple know what menstruation is? | is apple the only one who makes a health app that doesn't do this? |
apple promised an expansive health app, so why can't women track menstruation?
| is apple the only one who makes a health app that doesn't do this? | i guess the answer is pretty simple, they didn't want to have a slice of the app that would be irrelevant for half of their customers. of course most people won't use the dozen features of the app anyway, but they could/might regardless of their sex/age/race/ etc. introducing a sex specific feature makes a distinction. maybe they could have done it so you'd chose for the male or female version of the health app or something like that, not saying that doing it well was impossible but it would require focus and dedication for this particular problem. maybe we will see it coming in the future, who knows? but it was certainly not their focus at the release! |
apple promised an expansive health app, so why can't women track menstruation?
| i guess the answer is pretty simple, they didn't want to have a slice of the app that would be irrelevant for half of their customers. of course most people won't use the dozen features of the app anyway, but they could/might regardless of their sex/age/race/ etc. introducing a sex specific feature makes a distinction. maybe they could have done it so you'd chose for the male or female version of the health app or something like that, not saying that doing it well was impossible but it would require focus and dedication for this particular problem. maybe we will see it coming in the future, who knows? but it was certainly not their focus at the release! | what are you talking about this his been a feature on most phones since day one. i think they call it "calendar".also it's sexist to assume only women want to track menstruation isn't it? |
the secret goldman sachs tapes
| i worked at the federal reserve for some time, so i know that the culture is as toxic as the article says it to be. professionally, this experience has been the most aggravating of my career.oh, let me tell you about the laziness! one of the developers i worked with would come to work at 11 am every single day and leave at 5 pm—sharp. he would then take a two hour lunch break.it wouldn’t be so bad if he actually did some work, but there was none. i remember vividly this one incident: we were trying hard to meet a deadline and he was responsible for an important component. every week, he would paint a rosy picture of his progress, saying he was this close to finishing the project. the week of the deadline comes and it’s time for him to lay down his cards—and surprise, surprise—there was no working code! all he produced was a jumbling mishmash of code that had no hope of working.
i spent the next week in the office writing this component.close to 70 hours. unreal.[oh yeah, this guy was getting paid $130k, not to mention those sweet government bennies!]his direct manager would complain about him all the time, and many meetings were asking him to improve his output. and yet, he was never punished. the truth is, when it comes to working at the federal reserve, your work output is not as important as who your friends are. and this guy had friends in high(er) places.this guy is just one employee, but i believe it is just endemic of the kind of culture the fed breeds. one of nepotism, of indolence, of apathy. i am truly scared for the future of this country knowing that these are the people behind the wheel. | there's also a print version of the story at propublica: <link> |
the secret goldman sachs tapes
| there's also a print version of the story at propublica: <link> | audio: <link> <link> |
the secret goldman sachs tapes
| audio: <link> <link> | the ray rice video for the financial sector has arrived.not even close. that metaphor would only be apt if there was literally video of a federal reserve employee accepting cash from an employee of a major bank with an explicit statement from the bank saying something along the lines of"so you're going to keep the interest rates low so we can keep doing bond arbitrage instead of lowering rates for consumer lending right?"even then most people wouldn't care because they have no idea how financial institutions work. |
the secret goldman sachs tapes
| the ray rice video for the financial sector has arrived.not even close. that metaphor would only be apt if there was literally video of a federal reserve employee accepting cash from an employee of a major bank with an explicit statement from the bank saying something along the lines of"so you're going to keep the interest rates low so we can keep doing bond arbitrage instead of lowering rates for consumer lending right?"even then most people wouldn't care because they have no idea how financial institutions work. | if the tapes only contain content describing a general culture of discouragement of rigorous oversight, it's not totally surprising or any more jaw-droppingly incriminating than previous knowledge concerning the fed's relationship with the banks. basically this is what we already knew.the question isn't whether or not regulation is lax and whether or not this happened, it's how to fix it. and this poses a big problem because of the revolving door between regulatory agencies and private banks. no one wants to regulate a company that might offer them a million dollar job one day. it's kind of a tough problem to fix, since the only people who know enough about financial industry to regulate it are the people who work for the industry. and if we prevented people from working for the industry they regulate in any capacity for say, up to 10 years after they leave their government job (is this even constitutional?), much fewer people would want to work for the government or the fed. but even so i think that kind of reform is necessary - because it will attract to government service only people who are serious about regulation. |
entrepreneur is a silly word, and you sound silly when you call yourself one
| damned, i will change my linkedin profile; i didn't know it was sounding pompous :-)for me, an "entrepreneur" (french word) is someone who undertake an "entreprise". an "entreprise" is not necessarily a company or a business. it is a kind of project. you could say that preparing a party or a wedding is an "entreprise". | i am going to guess this is a silicon valley hang up because of people being funded. he kind of makes that point towards the end of the article. if you are someone who starts a business the traditional way, win or fail, i think you earned the right to be called an entrepreneur. you are taking a big risk which is scary and invigorating at the same time. |
entrepreneur is a silly word, and you sound silly when you call yourself one
| i am going to guess this is a silicon valley hang up because of people being funded. he kind of makes that point towards the end of the article. if you are someone who starts a business the traditional way, win or fail, i think you earned the right to be called an entrepreneur. you are taking a big risk which is scary and invigorating at the same time. | fwiw, here is the definition of "entrepreneur" from etmology online.<<< 1828, "manager or promoter of a theatrical production," reborrowing of french entrepreneur "one who undertakes or manages," agent noun from old french entreprendre "undertake" (see enterprise). the word first crossed the channel late 15c. (middle english entreprenour) but did not stay. meaning "business manager" is from 1852. related: entrepreneurship. >>>
<link>;undertake"? "theatrical"? heh. uncharacteristically, etymonline doesn't mention the two french component words, which more literally translate to "between-taker". a middleman, in other words.in more common parlance, anyone who starts their own business of any kind is an entrepreneur. food-truck operator? she's an entrepreneur. hair-salon owner? he's an entrepreneur too. if we wanted to make finer distinctions about degrees of entrepreneurship, those distinctions might be based on risk. buying your own equipment and materials to sell stuff at the local art fair? very entrepreneurial. taking other people's money to sell to an enterprise where you have inside contacts? barely entrepreneurial at all.and that's where i think anger like that in the op comes from. people resent the way that founders of vc-backed high-tech companies seem to act like they're the vanguard or elite of entrepreneurship, when there's often very little entrepreneurial risk involved. for example:* person a goes to work for amagoobook directly.* person b starts a company that fails, then goes to work for amagoobook.* person c starts a company that's acquired by amagoobook.what's the difference? why are b and c entrepreneurs but a isn't? because investors paid their salaries for a couple of years? not all startups are like this, of course, but the question is whether those that are should claim the same "entrepreneur" mantle that others truly deserve. when a startup founder orders pizza, it's highly likely that the guy who owns the pizza shop is the real entrepreneur. |
entrepreneur is a silly word, and you sound silly when you call yourself one
| fwiw, here is the definition of "entrepreneur" from etmology online.<<< 1828, "manager or promoter of a theatrical production," reborrowing of french entrepreneur "one who undertakes or manages," agent noun from old french entreprendre "undertake" (see enterprise). the word first crossed the channel late 15c. (middle english entreprenour) but did not stay. meaning "business manager" is from 1852. related: entrepreneurship. >>>
<link>;undertake"? "theatrical"? heh. uncharacteristically, etymonline doesn't mention the two french component words, which more literally translate to "between-taker". a middleman, in other words.in more common parlance, anyone who starts their own business of any kind is an entrepreneur. food-truck operator? she's an entrepreneur. hair-salon owner? he's an entrepreneur too. if we wanted to make finer distinctions about degrees of entrepreneurship, those distinctions might be based on risk. buying your own equipment and materials to sell stuff at the local art fair? very entrepreneurial. taking other people's money to sell to an enterprise where you have inside contacts? barely entrepreneurial at all.and that's where i think anger like that in the op comes from. people resent the way that founders of vc-backed high-tech companies seem to act like they're the vanguard or elite of entrepreneurship, when there's often very little entrepreneurial risk involved. for example:* person a goes to work for amagoobook directly.* person b starts a company that fails, then goes to work for amagoobook.* person c starts a company that's acquired by amagoobook.what's the difference? why are b and c entrepreneurs but a isn't? because investors paid their salaries for a couple of years? not all startups are like this, of course, but the question is whether those that are should claim the same "entrepreneur" mantle that others truly deserve. when a startup founder orders pizza, it's highly likely that the guy who owns the pizza shop is the real entrepreneur. | same argument can be made about people calling themselves ceo of startups with no employees or product. |
entrepreneur is a silly word, and you sound silly when you call yourself one
| same argument can be made about people calling themselves ceo of startups with no employees or product. | meh.after years of reading hundreds of essays from people complaining about how others use words ("hacker", "professional", "engineer", "artist", "freedom", and hundreds of other words, etc), i've come to the conclusion that the complaint says more about the rigidity of the complainer rather than anything insightful about language evolution.it's ironic that the blog's name has the word "consulting" in it. when i was in my early 20s, i hated that word. whenever i told friends i was a consultant, i'd immediately follow up with a self-deprecation: "that's just a fancy label for white-collar prostitute." at the time, "consulting" to me meant mckinsey or bain and not your bodyshop of visual basic programmers. i preferred "contractor" or "freelancer" to the title "consultant". however, the years have gone by and i no longer care. anyone can call themselves "consultant" and it doesn't bother me at all. it's just bikeshedding now.anyone remember the 1990s complaints about the title "webmaster" because (gasp!) it had the word "master" as a suffix? how pompous of those html authors! we should be thankful that "web-jedi" didn't gain any currency and webmaster has already fallen out of favor.the author suggests "business owner" instead of "entrepreneur". i used to not like that label either. "business owner" had connotations of being some monopolist crushing the protesters on occupy wall street or the slavedriver paying his workers minimum wage while drinking their babies blood. maybe i can call myself "unemployed speculator" or some other circumlocution so as to not offend anyone. personally, i think "entrepreneur" does not convey any false pretenses about prestige or bootstrapping. it's fascinating that some find a word neutral while others perceive it as putting on airs. you can't win.if the author doesn't like "entrepreneur" and i have mixed feelings about "business owner", what's a good compromise?how about other candidate labels such as "wantrepreneur"[1] which is what some of non-hn folks call hn'ers.or how about "wannabe philanthropist"? if high-profile charity is the apotheosis of the journey from student->businessman->philanthropist, is "wannabe philanthropist" self-deprecating enough while simultaneously conveying bold aspirations? if family & friends have that puzzled look and ask, "what the heck is a wannabe philanthropist?", you explain the transformation of bill gates from harvard dropout -> ceo -> gates foundation. as a bonus, you can throw in words like "camel" and "eye of the needle". and after all that, they'll say "oh for fock's sake, just call yourself an entrepreneur!"[1]<link> |
pebilepsy – pebble app for seizure tracking and measurement
| this is a great idea (and name!), but does this fall under fda regulation? i haven't kept up with the fda's current policy on medical apps. | silly to move forward with medication dosage and other management with just generalized tonic clonic seizures. leaving out partial complex seizures and the like would be required information. most temporal lobe seizure patients have a much greater partial to generalized seizure ratio. |
pebilepsy – pebble app for seizure tracking and measurement
| silly to move forward with medication dosage and other management with just generalized tonic clonic seizures. leaving out partial complex seizures and the like would be required information. most temporal lobe seizure patients have a much greater partial to generalized seizure ratio. | automatic notification would be a neat feature.have it text a family member when a seizure is detected so they can follow up with a phone call in a little bit to make sure everything is ok.as more and more 911 centers accept sms messages, it might be cool to allow the user to specify some threshold after which an automated text is sent to 911. the vast majority of epilepsy related seizures do not require an ambulance, but in the event of a seizure that does not stop in a 'normal' time frame[1], it could get help coming that much faster.[1] <link> |
pebilepsy – pebble app for seizure tracking and measurement
| automatic notification would be a neat feature.have it text a family member when a seizure is detected so they can follow up with a phone call in a little bit to make sure everything is ok.as more and more 911 centers accept sms messages, it might be cool to allow the user to specify some threshold after which an automated text is sent to 911. the vast majority of epilepsy related seizures do not require an ambulance, but in the event of a seizure that does not stop in a 'normal' time frame[1], it could get help coming that much faster.[1] <link> | this is a great idea. unfortunately for me i have/had petit-mal seizures, because of that i don't think the method by which this tracks seizures would work for me.i love the name, love the idea, keep it up! |
pebilepsy – pebble app for seizure tracking and measurement
| this is a great idea. unfortunately for me i have/had petit-mal seizures, because of that i don't think the method by which this tracks seizures would work for me.i love the name, love the idea, keep it up! | i've been looking for something like this for ages - i have a family member who suffers from generalized tonic clonic seizures and have been looking at ways to try and do remote tracking.the #1 feature request i would have is the ability to nominate a mobile number to get an sms alert if a seizure is detected.is there any chance of getting early access to this ? i'd be really happy to be an alpha tester and would be highly motivated to give you guys as much feedback as possible from the p.o.v. of a family using it to monitor epilepsy. |
spec-me-maybe: introduces the “maybe” syntax to rspec
| while this is obviously a joke, i can see it having a limited amount of practical value. imagine, for example, being tasked with updating a poorly-maintained legacy code base with a large number of indeterminately failing specs containing order dependencies. you could `rspec::maybe` everything until you're sure you have the order dependencies resolved, then convert to regular `rspec::expectation`s. | everything should also take `.maybe.it_wont` at the end, for completeness. |
spec-me-maybe: introduces the “maybe” syntax to rspec
| everything should also take `.maybe.it_wont` at the end, for completeness. | i can see wondering if this was a joke as 'maybe' is mildly reminiscent of promise[1] and the like but if you got to .on_my_machine and still couldn't figure it out, you got some 'splainin to do. :)[1] <link> |
spec-me-maybe: introduces the “maybe” syntax to rspec
| i can see wondering if this was a joke as 'maybe' is mildly reminiscent of promise[1] and the like but if you got to .on_my_machine and still couldn't figure it out, you got some 'splainin to do. :)[1] <link> | love the build status badge. nice touch. |
spec-me-maybe: introduces the “maybe” syntax to rspec
| love the build status badge. nice touch. | i find it quite funny that people cannot tell whether this is a joke or not, just says something about the state of things... |
from the world of design trends: “flat design” is dying
| this was a pointlessly fluffy article that can be summed up with: "flat design is a fad. just like modern and postmodern movements have come and gone in art — so will this."well, of course movements change, so you could write an article as soon as a trend starts that its "dying."more importantly, human-computer interactions is a relatively new area of design. how do we know that flat design will actually be just a fad rather than a paradigm shift? | site currently down. google cache: <link> |
from the world of design trends: “flat design” is dying
| site currently down. google cache: <link> | flat isn't dying, although maybe the most extreme version of it may be falling out of favor. thick glassy, 3d icons with heavy drop shadows, skeuomorphism, and gradient bevels, protruding interface elements and the like are still as garish now as they were a year ago. the trend toward more subtle depth if any depth at all and only when needed is still very much a thing.much of the new popular ui designs is still pretty flat too: <link> |
from the world of design trends: “flat design” is dying
| flat isn't dying, although maybe the most extreme version of it may be falling out of favor. thick glassy, 3d icons with heavy drop shadows, skeuomorphism, and gradient bevels, protruding interface elements and the like are still as garish now as they were a year ago. the trend toward more subtle depth if any depth at all and only when needed is still very much a thing.much of the new popular ui designs is still pretty flat too: <link> | good riddance.while flat design was definitely done "right" by some (e.g. windows phone) it was done very wrong indeed by others (ios 7, google maps, windows 8, et al). so wrong in fact that i'd go as far as to call it a usability anti-pattern.there's a reason why shadows and borders exist: to distinguish items from one another on a 2d display. without those clues it is hard to distinguish a text label from a clickable button.perfect example:<link> if you had never used windows 8 before, you might not know that "change pc settings" is a button instead of the title of the section, and in my experience that is typical for most users. |
from the world of design trends: “flat design” is dying
| good riddance.while flat design was definitely done "right" by some (e.g. windows phone) it was done very wrong indeed by others (ios 7, google maps, windows 8, et al). so wrong in fact that i'd go as far as to call it a usability anti-pattern.there's a reason why shadows and borders exist: to distinguish items from one another on a 2d display. without those clues it is hard to distinguish a text label from a clickable button.perfect example:<link> if you had never used windows 8 before, you might not know that "change pc settings" is a button instead of the title of the section, and in my experience that is typical for most users. | and i just redesigned my websites from gradients to flat. what will they think of next? |
3rd shellshock vulnerability found
| if you're vulnerable to this, then you're vulnerable to someone setting path and changing all executables that you refer to.that's not the "shock" part of "shellshock"; that's a long known issue. applications which pass arbitrary variables from the network through environment variables already add a unique prefix to the variables, to prevent this kind of problem. that unique prefix also prevents them from accidentally conflicting with the name of an actual executable for this case.now, there are still going to be bugs with suid executables not sanitizing their environment, but those apply in many other ways, such as via path and ld_preload.the "shock" part of shellshock was that it was possible to execute arbitrary code, upon just loading a variable in bash, regardless of what its name was, depending only on the contents, which in many cases can be attacker controlled. | can we go straight to the part where this is disabled behind a --insecure flag and everyone can spend a release cycle migrating untrusted input handling to languages which were designed for it? |
3rd shellshock vulnerability found
| can we go straight to the part where this is disabled behind a --insecure flag and everyone can spend a release cycle migrating untrusted input handling to languages which were designed for it? | this is not a vulnerability, it is intended functionality. some scripts undoubtedly depend on overloading a shell builtin command with a function (even though it is gross).the post's proposed exploit path is not possible on linux and most os's, because you cannot have a setuid script.[1][1] <link> |
3rd shellshock vulnerability found
| this is not a vulnerability, it is intended functionality. some scripts undoubtedly depend on overloading a shell builtin command with a function (even though it is gross).the post's proposed exploit path is not possible on linux and most os's, because you cannot have a setuid script.[1][1] <link> | it's actually not "new" it's a "feature" that everybody became aware it existed since long ago only once shellshock bug became known.christos zoulas from netbsd has the arguments that i really like: it's a "feature" that was widely unknown and that should be disabled by default:<link> |
3rd shellshock vulnerability found
| it's actually not "new" it's a "feature" that everybody became aware it existed since long ago only once shellshock bug became known.christos zoulas from netbsd has the arguments that i really like: it's a "feature" that was widely unknown and that should be disabled by default:<link> | wow this is getting out of hand, that's not really a security vulnerability anymore. that's how bash passes exported functions, i'll get back to that.there is some misunderstanding though regarding the suid shell script. i can't think of any modern unix other than solaris that allows those by default any more. it's just not a concern anymore unless an admin twiddles some sysctl or what not. and in the case of solaris /bin/sh is not bash. also /bin/sh does it's own setuid after comparing to 100 (unless the -p option is passed) so often if people are not adding users to a group or sudo, they would create a ksh or perl script instead. anyway that's just rambling, sorry, the short answer is that i would be really surprised if there was any modern system that used bash as /bin/sh and also allowed suid interpreter scripts. |
why free online classes are still the future of education
| the entire concept of online classes is an anachronism - the way early automobiles resembled horse carriages. this is just schools' futile attempts at staying relevant using some token internet and marketing themselves to people indoctrinated that traditional classes were the only way to learn. lecture is the worst way to teach, and yet people gush about putting videos of lectures online as some sort of innovation. it's sad that we can't think a little more outside of the box.an online education service i'd like to see is not teaching but certification. they could recommend learning resources and offer email and voice communication with subject matter experts for feedback and questions as needed, but the main job would be assessing your competence in a specific area and backing it up with a guarantee. you could get a card showing certification in calculus ii (expires one year after issuance) or early american history, or whatever. but when you pass certification, you actually know the subject - not just sleepwalked through the course doing the minimal busy work. the student may use whatever learning methods that work best for themselves to achieve this and employers looking for specific experts would have confidence in the possession of that expertise. | my big annoyance is with the moocs that still try and fit you into a schedule. many times i have been in a course where i would have burned through the whole thing if i was allowed. but instead they feel the need to stagger the information on a weekly basis and i lose interest. coursera is the one that comes to mind. i also don't care about getting a grade or certificate. just give me all the lectures at once and some exercises so i can practice if i want. gotta let go of the past. |
why free online classes are still the future of education
| my big annoyance is with the moocs that still try and fit you into a schedule. many times i have been in a course where i would have burned through the whole thing if i was allowed. but instead they feel the need to stagger the information on a weekly basis and i lose interest. coursera is the one that comes to mind. i also don't care about getting a grade or certificate. just give me all the lectures at once and some exercises so i can practice if i want. gotta let go of the past. | are they now? part of what you do at a university is build your professional network, both upwards, through your mentors, and sideways, with your peers. you also get introduced to new developments in your field, through informal, unscripted interactions, just walk down the hall and talk to the fellow in the office two doors down. that's a university. you don't get that through online courses. for a trade school, online courses work just fine. and the ruling class needs well-trained, obedient drones. |
why free online classes are still the future of education
| are they now? part of what you do at a university is build your professional network, both upwards, through your mentors, and sideways, with your peers. you also get introduced to new developments in your field, through informal, unscripted interactions, just walk down the hall and talk to the fellow in the office two doors down. that's a university. you don't get that through online courses. for a trade school, online courses work just fine. and the ruling class needs well-trained, obedient drones. | i have mixed feelings about the moocs.on the positive side, the benefits are substantial. 1. democratized access to the world's top educators, 2. low cost distribution of education.on the negative side, the moocs are kind of the large classroom problem taken to the extreme. it's generally agreed that a classroom with more students per teacher is not as good as a lower student/teacher ratio.but to critique my own negativity, the people working on moocs are smart and motivated. it's a bit naive to think that the current mooc is as good as they will ever get. clearly, these are early stage products that have substantial evolution and improvement in their future.the one area in open education that needs to improve is around content licensing. if you look at most open educational content, the licenses are restricted open source (gpl like) and note unrestricted open source (mit/bsd/apache like). i fail to understand how making the content unrestricted would not benefit everyone. |
why free online classes are still the future of education
| i have mixed feelings about the moocs.on the positive side, the benefits are substantial. 1. democratized access to the world's top educators, 2. low cost distribution of education.on the negative side, the moocs are kind of the large classroom problem taken to the extreme. it's generally agreed that a classroom with more students per teacher is not as good as a lower student/teacher ratio.but to critique my own negativity, the people working on moocs are smart and motivated. it's a bit naive to think that the current mooc is as good as they will ever get. clearly, these are early stage products that have substantial evolution and improvement in their future.the one area in open education that needs to improve is around content licensing. if you look at most open educational content, the licenses are restricted open source (gpl like) and note unrestricted open source (mit/bsd/apache like). i fail to understand how making the content unrestricted would not benefit everyone. | i spent a few days looking through online courses from a couple of different vendors, hoping to take course of online class in a new field of study, so i could round myself out or possibly find a new career field.i quickly learned that outside of the software development discipline, and a precious few pure science fields, there are virtually no courses available. if you'd like to look into education or humanities or any of the thousands of other degrees? all you'll find are 101 level courses, if you find anything at all.hopefully this will improve, but i'm not holding my breath. |
marc andreessen sounds warning on startups burning cash
| part of marc andreesen's tweetstorm yesterday said, "worry." andreesen hasn't said "worry" since 2008, so between that and the fact that he is generally very bullish and optimistic (to the point that when a16z makes a bet they can put in an order of magnitude more money than other investors would), i began to stress out more than a little bit, especially having just closed our round.this may be anecdotal to my experience, but there was a time last year when the advice we were getting from mentors changed from, "figure out how much cash you need to meet your next milestone and go raise that with a bit of cushion" to "money is everywhere right now; get as much as you can and build the biggest war chest possible." neither of these approaches are particularly wrong -- it would be weird to turn down money when it's on the table, but we opted to raise significantly less than what we could have raised, and keep our team where we wanted it instead of hiring enough people to become "sexy."part of that is because we didn't raise at a super sexy y-combinator valuation and didn't want to give away the farm, and part of that was intentional. in short, we wouldn't know what to do with those people if we had them.to be clear, we still have plenty of money to get stuff done. we still can pay for all the tools we need, pay ourselves enough to not have to stress about cash, and can hire top talent. we just don't have more money than we know what to do with.we have three people working full-time on product right now, and that will grow to five in the next few months. if we had raised a $3 million seed round (if you can still call that a "seed"), we would have felt compelled to go hire a dozen people. i can think of stuff we'd like to do, but it would probably be an absolute mess if we were building it all in parallel.there may be some companies for whom building a huge war chest makes sense, but the notion of raising money for the sake of raising money (because you can) seems odd and dangerous to me. | vc-funded startups have always burned cash at prodigious rates. there's nothing inherently worrisome about that. what worries andreessen, gurley, wilson, and other investors with decades of experience is that many startups today are burning cash by spending on things that do not increase business value -- and in some cases actually destroy it -- without worrying for even a moment about the possibility of running out of money.in andreessen's words: "lots of people, big shiny office, high expense base = fake 'we’ve made it!' feeling. removes pressure to deliver real results."[1]in gurley's words: "i think that silicon valley as a whole or that the venture-capital community or startup community is taking on an excessive amount of risk right now. unprecedented since '99. in some ways less silly than '99 and in other ways more silly than in '99. ... for the first time since '99, in the past 12 months, i've been in board meetings where the company says, 'our only option is a 10-year lease,' at record pricing on a per square foot basis here in san francisco, which is two or three times what the rent was three years ago. and so this is why it's all cyclical—because the landlords get greedy. they wouldn't do a 10-year lease if they thought that the rates were low. so they're implicitly telling you they want to lock this in for 10 years, which is its own form of greed because what happened in '99 is half the companies went bankrupt and they couldn't pay the lease over the 10-year period."[2]in wilson's words: "i’ve pushed back on long term leases that i thought were outrageous, i’ve pushed back on spending plans that i thought were too aggressive and too risky, i’ve made myself a pain in the ass to more than a few ceos."[3]their worry, in short, is that many vc-funded startups are burning cash stupidly.--[1] <link>[2] <link>[3] <link> |
marc andreessen sounds warning on startups burning cash
| vc-funded startups have always burned cash at prodigious rates. there's nothing inherently worrisome about that. what worries andreessen, gurley, wilson, and other investors with decades of experience is that many startups today are burning cash by spending on things that do not increase business value -- and in some cases actually destroy it -- without worrying for even a moment about the possibility of running out of money.in andreessen's words: "lots of people, big shiny office, high expense base = fake 'we’ve made it!' feeling. removes pressure to deliver real results."[1]in gurley's words: "i think that silicon valley as a whole or that the venture-capital community or startup community is taking on an excessive amount of risk right now. unprecedented since '99. in some ways less silly than '99 and in other ways more silly than in '99. ... for the first time since '99, in the past 12 months, i've been in board meetings where the company says, 'our only option is a 10-year lease,' at record pricing on a per square foot basis here in san francisco, which is two or three times what the rent was three years ago. and so this is why it's all cyclical—because the landlords get greedy. they wouldn't do a 10-year lease if they thought that the rates were low. so they're implicitly telling you they want to lock this in for 10 years, which is its own form of greed because what happened in '99 is half the companies went bankrupt and they couldn't pay the lease over the 10-year period."[2]in wilson's words: "i’ve pushed back on long term leases that i thought were outrageous, i’ve pushed back on spending plans that i thought were too aggressive and too risky, i’ve made myself a pain in the ass to more than a few ceos."[3]their worry, in short, is that many vc-funded startups are burning cash stupidly.--[1] <link>[2] <link>[3] <link> | why now? what about all the years startups have been burning cash? it's nothing new. parties, entitlements like free meals for employees, fancy offices and furniture, too much travel, ads and pr spending, fancy hardware... when a startup has zero or negligible revenue, this kind of excessive spending is absurd. not clear why marc thinks it's time to sound a warning now. |
marc andreessen sounds warning on startups burning cash
| why now? what about all the years startups have been burning cash? it's nothing new. parties, entitlements like free meals for employees, fancy offices and furniture, too much travel, ads and pr spending, fancy hardware... when a startup has zero or negligible revenue, this kind of excessive spending is absurd. not clear why marc thinks it's time to sound a warning now. | i am not really qualified to determine whether there is a bubble nor do i believe there is much value in speculating constantly about it. if there is one, valuations will drop, cash will be harder to get. its not going to change the kind of work i like to pursue or the work that makers, builders and engineers do. it will probably change how much everyone is stuffing in their pockets, fine.one glaringly bizarre thing from my perspective is why anyone would listen to vcs about whether we are or are not in a bubble. i'm sure there are other perspectives, but it would seem someone heavily invested in seeing valuations go up wouldn't want to ring alarm bells... i think instead i would rather gauge things by the numbers which unfortunately, are probably not frequent enough to reach true statistical significance to not have incredibly large confidence ranges. that being said, some of the money flying by on crunchbase every morning is ludicrous. |
marc andreessen sounds warning on startups burning cash
| i am not really qualified to determine whether there is a bubble nor do i believe there is much value in speculating constantly about it. if there is one, valuations will drop, cash will be harder to get. its not going to change the kind of work i like to pursue or the work that makers, builders and engineers do. it will probably change how much everyone is stuffing in their pockets, fine.one glaringly bizarre thing from my perspective is why anyone would listen to vcs about whether we are or are not in a bubble. i'm sure there are other perspectives, but it would seem someone heavily invested in seeing valuations go up wouldn't want to ring alarm bells... i think instead i would rather gauge things by the numbers which unfortunately, are probably not frequent enough to reach true statistical significance to not have incredibly large confidence ranges. that being said, some of the money flying by on crunchbase every morning is ludicrous. | i'd just like to point out that the "why now?" question is probably related to a major shift in america's monetary policy. since the 2008 crash, the fed has been using a policy of "quantitive easing" (qe) to keep interest rates artificially low and thus encourage lending and spending. it's been known for while that the feds were looking to end this policy sometime this year and it's looking like the end of october will be that time. although the feds seem to indicate that they won't actually raise interest rates until sometime next year, halting qe will have the affect of making less money available for lending. no one is really sure what is going to happen but this is an event that will affect the economy as a whole. to me, it's sounding like one of the industries that could be particularly affected by this policy is tech and the result will be a market correction in startup valuations.links!
<link>
<link>
<link>
<link> |
show hn: pine – opentable for massages
| we've been working on this for the last few months and would appreciate feedback. thanks! | i have a back injury so tend to be a frequenter of massage providers. i've tried a bunch of massage services and providers and it's really hit or miss. but this seems way better than trying to book something directly with a masseuse/masseur which usually involves a lot of back and forth for scheduling, rates, clarification of service, etc. so i'm really glad that there is some disruption in this space to make this particular hassle easier. i also think that this model could possibly guarantee a better overall experience as there is someone on the other end that cares about converting you into a regular customer on their platform which may improve quality standards everywhere, and especially for the providers on the particular platform. in my experience, massage providers seldom care if you had a good experience and there is basically nothing you can do if you had a bad one. this is especially true if you go to get massage as a form of physical therapy rather than just someone giving you a relaxation rub.i've actually used pine a few times and i will say that the booking process is surprisingly simple via their app and website. also, i've never been disappointed with the quality of the provider i got - seems like at the very least this startup has really good relationships with quality providers that they are able to bring onto the platform. worth a try in my opinion. |
show hn: pine – opentable for massages
| i have a back injury so tend to be a frequenter of massage providers. i've tried a bunch of massage services and providers and it's really hit or miss. but this seems way better than trying to book something directly with a masseuse/masseur which usually involves a lot of back and forth for scheduling, rates, clarification of service, etc. so i'm really glad that there is some disruption in this space to make this particular hassle easier. i also think that this model could possibly guarantee a better overall experience as there is someone on the other end that cares about converting you into a regular customer on their platform which may improve quality standards everywhere, and especially for the providers on the particular platform. in my experience, massage providers seldom care if you had a good experience and there is basically nothing you can do if you had a bad one. this is especially true if you go to get massage as a form of physical therapy rather than just someone giving you a relaxation rub.i've actually used pine a few times and i will say that the booking process is surprisingly simple via their app and website. also, i've never been disappointed with the quality of the provider i got - seems like at the very least this startup has really good relationships with quality providers that they are able to bring onto the platform. worth a try in my opinion. | is this a/b testing for the other massage post on the front page? or are you a different company? <link> |
show hn: pine – opentable for massages
| is this a/b testing for the other massage post on the front page? or are you a different company? <link> | jesus christ there are lot of 0 day old accounts commenting in here. please go "disrupt" reddit or some other site. |
show hn: pine – opentable for massages
| jesus christ there are lot of 0 day old accounts commenting in here. please go "disrupt" reddit or some other site. | tried it, loved it. the mts are good but it must be hard for them to scale easily |
why doesn't somebody update bash using the exploit?
so, why doesn't someone exploit bash to update bash. might do the world a small favor. i was thinking that you could issue several commands to match a range of systems, and all it would take is one to work and then bash would be updated. | because doing so would be illegal.and why go to jail, literally, for trying to fix other people's mistakes? few have that little to lose.plus companies would claim they spend $$$$ fixing the "damage" you did (e.g. reimage the machine, audits, management meetings about it). they'll claim they "spent" $50k or something stupid after your "illegal break-in" and try to sue you. | because that's hacking and totally illegal in most countries.and this exact question was posted to ask hn yesterday.edit: it looks like yesterday's discussion was flagged out of existence. |
why doesn't somebody update bash using the exploit?
so, why doesn't someone exploit bash to update bash. might do the world a small favor. i was thinking that you could issue several commands to match a range of systems, and all it would take is one to work and then bash would be updated. | because that's hacking and totally illegal in most countries.and this exact question was posted to ask hn yesterday.edit: it looks like yesterday's discussion was flagged out of existence. | as the other commentators mention this would almost certainly be illegal, and would fall under computer-misuse laws.that said the more likely reason is that a lot of the time you couldn't - because the bash-you're exploiting wouldn't have root privileges and would thus be unable to fix the issue. |
why doesn't somebody update bash using the exploit?
so, why doesn't someone exploit bash to update bash. might do the world a small favor. i was thinking that you could issue several commands to match a range of systems, and all it would take is one to work and then bash would be updated. | as the other commentators mention this would almost certainly be illegal, and would fall under computer-misuse laws.that said the more likely reason is that a lot of the time you couldn't - because the bash-you're exploiting wouldn't have root privileges and would thus be unable to fix the issue. | not only would that be illegal, it would be nearly impossible, since you'd have to patch bash at the binary level.iir, compiled software on linux is non-deterministic, so you'd stand an amazingly high chance of damaging the targeted system as opposed to fixing it.so, if you could pull that one off, you'd not only be a god, but you'd be a god that would probably spend the rest of your life defending yourself in court, where ethics and justice are not really the point.just let the process handle it. it's way simpler and far less likely to blow up. |
why doesn't somebody update bash using the exploit?
so, why doesn't someone exploit bash to update bash. might do the world a small favor. i was thinking that you could issue several commands to match a range of systems, and all it would take is one to work and then bash would be updated. | not only would that be illegal, it would be nearly impossible, since you'd have to patch bash at the binary level.iir, compiled software on linux is non-deterministic, so you'd stand an amazingly high chance of damaging the targeted system as opposed to fixing it.so, if you could pull that one off, you'd not only be a god, but you'd be a god that would probably spend the rest of your life defending yourself in court, where ethics and justice are not really the point.just let the process handle it. it's way simpler and far less likely to blow up. | there was actually a great short story in 2600 a few years back about a security researcher using an exploit to create a worm that automatically patched up the vulnerable systems. quite enjoyable, anyone remember the name of it? |
show hn: snitch.io – ssl auditing and alerting
| i think you're trying to solve a non-problem since the company that sells the certificates warns you (sometimes even more than those intervals), afterall, they want you to renew as well.
as for checking for quality, that should be the sys admin task or the webmaster.
good luck though! | this is seems potentially quite useful. it would be nice if it could also notify you if your server is not configured according to best practices in terms of things such as protocol versions and cipher suites. |
show hn: snitch.io – ssl auditing and alerting
| this is seems potentially quite useful. it would be nice if it could also notify you if your server is not configured according to best practices in terms of things such as protocol versions and cipher suites. | idea is great, but pricing seems a bit expensive.have >25 certs? add this check to nagios: <link> you $200/month :) |
show hn: snitch.io – ssl auditing and alerting
| idea is great, but pricing seems a bit expensive.have >25 certs? add this check to nagios: <link> you $200/month :) | great idea, will definitively check it out.where are you incorporated, if? the terms says nothing about it. who is my contract party when i signup? |
show hn: snitch.io – ssl auditing and alerting
| great idea, will definitively check it out.where are you incorporated, if? the terms says nothing about it. who is my contract party when i signup? | we built snitch to make it simple and easy to get a handle on your ssl certificates. our mission is to help people avoid getting blind-sided by ssl issues - losing customers, reputation and business in the process.we've been working on this for a few months and would appreciate any feedback - thanks!if anyone wants to email me directly it is my username at currylabs.com or gmail.comps: if you are an open source project we offer free subscriptions. |
arm’s rebuttal to risc-v: “the case for licensed instruction sets”
| the flip side of "no fragmentation" is, there's no arm core with hardware multithreading, and there won't be one until arm decides to make one, which afaik may well be "never". you can license a core with hardware multithreading if you need it - say, mips from imagination - but the isa won't be arm-compatible. to the extent that arm's "software ecosystem" is valuable, it's a pity nobody can make a multithreaded arm-compatible core. similarly for other features.on the other hand, x86, which is "open" to all of intel, amd and via who reached a patent war stalemate, has the problem of incompatible instruction set extensions, as documented in "stop the instruction set war": <link> problem is worse is hard to decide. | url changed from <link>, which points to this. the article being responded to is <link> |
arm’s rebuttal to risc-v: “the case for licensed instruction sets”
| url changed from <link>, which points to this. the article being responded to is <link> | odd that arm even bothered to reply. and the arguments are the expected stuff you'd expect to see from an entrenched leader. basically it's "benefits of standardization trumps openness", though not stated that way.a few bits struck me as amusing, like the chart with all the extensions to the arm isa over the years, most of which have been abandoned in modern cores. citing the original vfp standard or jazelle as "innovation" seems laughable.if i were to write a arm-vs-risc-v paper, i'd start with the important things that are actually missing from risc-v still, like a mmu spec. |
arm’s rebuttal to risc-v: “the case for licensed instruction sets”
| odd that arm even bothered to reply. and the arguments are the expected stuff you'd expect to see from an entrenched leader. basically it's "benefits of standardization trumps openness", though not stated that way.a few bits struck me as amusing, like the chart with all the extensions to the arm isa over the years, most of which have been abandoned in modern cores. citing the original vfp standard or jazelle as "innovation" seems laughable.if i were to write a arm-vs-risc-v paper, i'd start with the important things that are actually missing from risc-v still, like a mmu spec. | note that the whole arm ecosystem relies on a crapload of open source stuff, like the entire gnu toolchain, kernel and on up.so any criticism of the call for openness in the risc-v paper from the arm camp is sheer, sheer hypocrisy.in economic terms, the free stuff that helps arm be popular is a "complementary good". when you sell something, you want the complementary goods to be commoditized and as cheap as possible, while keeping your ingredient as proprietary as possible.obvious table reversal:"no, no! open source the cpu cores, and buy my compiler instead! that's how you keep costs low, and everything from fragmenting." |
arm’s rebuttal to risc-v: “the case for licensed instruction sets”
| note that the whole arm ecosystem relies on a crapload of open source stuff, like the entire gnu toolchain, kernel and on up.so any criticism of the call for openness in the risc-v paper from the arm camp is sheer, sheer hypocrisy.in economic terms, the free stuff that helps arm be popular is a "complementary good". when you sell something, you want the complementary goods to be commoditized and as cheap as possible, while keeping your ingredient as proprietary as possible.obvious table reversal:"no, no! open source the cpu cores, and buy my compiler instead! that's how you keep costs low, and everything from fragmenting." | if you're interested in the development of an open-source soc using risc-v then do keep an eye on <link> we will have more public soon - i'm talking at the openrisc conference (<link> next month. |
nsa technology transfer program – 2014 technology catalog [pdf]
| i reject the idea that one should have to pay to license technology developed by a (taxpayer funded) federal agency. if tax dollars funded it, it belongs to all of us anyway.i believe i'm going to write my congressional reps later today and suggest pushing for the elimination of licensing fees for tech from nsa, nasa, dod, etc. i'm sure it won't go anywhere, but what can ya do? | some of the nuggets:"digital transcription system" (that can handle foreign languages). this would be awesome to release oss. finally the tech giants wouldn't have a monopoly on good voice recognition."all fiber optically-controlled optical switch" so they have all they need to make optical processors/routers. if this is the declassified tech, it's time to start paying attention to <link>;port protector assembly for d-sub connectors" i haven't heard of much spying happening by someone hooking something up to an unused dvi port. good to know. |
nsa technology transfer program – 2014 technology catalog [pdf]
| some of the nuggets:"digital transcription system" (that can handle foreign languages). this would be awesome to release oss. finally the tech giants wouldn't have a monopoly on good voice recognition."all fiber optically-controlled optical switch" so they have all they need to make optical processors/routers. if this is the declassified tech, it's time to start paying attention to <link>;port protector assembly for d-sub connectors" i haven't heard of much spying happening by someone hooking something up to an unused dvi port. good to know. | another interesting one, i won't post any more. can anyone think of a current consumer use of a technique like this? those garmin gps/radios that track each other is the closest i can think of:determine range and velocity of an object
patent nos.: 7,755,536 and 7,545,325this method provides the ability to determine the
position and velocity of a transmitter over a wide
range of bandwidths. the second invention models
the doppler process as a time shift and change of
scale of the original signal. in this process, the
received signal is delayed in time and dilated. in this
invention, the method of performing the resampling function is to
apply a forward fourier transform of one length to the zero-padded
signal and apply an inverse fourier transform of a different length to
the results of the first transform. this pair of operations results in a
resampling of the signal at a rational multiple of the original sample
rate. the scale correlation is then computed as the normal
correlation of the transmitted signal and the resampled signal.value
accurate model of doppler process
rational signal resampling |
nsa technology transfer program – 2014 technology catalog [pdf]
| another interesting one, i won't post any more. can anyone think of a current consumer use of a technique like this? those garmin gps/radios that track each other is the closest i can think of:determine range and velocity of an object
patent nos.: 7,755,536 and 7,545,325this method provides the ability to determine the
position and velocity of a transmitter over a wide
range of bandwidths. the second invention models
the doppler process as a time shift and change of
scale of the original signal. in this process, the
received signal is delayed in time and dilated. in this
invention, the method of performing the resampling function is to
apply a forward fourier transform of one length to the zero-padded
signal and apply an inverse fourier transform of a different length to
the results of the first transform. this pair of operations results in a
resampling of the signal at a rational multiple of the original sample
rate. the scale correlation is then computed as the normal
correlation of the transmitted signal and the resampled signal.value
accurate model of doppler process
rational signal resampling | one of their offering under technology is converting computer programs with loops to ones without loops. can anyone help me understand what the point of that is? |
nsa technology transfer program – 2014 technology catalog [pdf]
| one of their offering under technology is converting computer programs with loops to ones without loops. can anyone help me understand what the point of that is? | patent 7,945,947lol, they patented a privilege escalation rootkit |
show hn: airtable, a real-time spreadsheet-database hybrid
| this looks like a nice product. software companies have been struggling to make a mass-market database program ever since lotus 1-2-3 (the "3" was a database), but the spreadsheet remains king, despite the fact that for storing structured data, it is almost as bad as a word document with macros. so i'll be rooting for you.one complaint: referring to your basic plan as being "free forever" is a bit disingenuous. in my opinion, the ftc ought to prohibit use of this term by tech companies located in the 650 and 415 area codes. | ah, another dabbledb! i hope you stick around longer. :)<link> |
show hn: airtable, a real-time spreadsheet-database hybrid
| ah, another dabbledb! i hope you stick around longer. :)<link> | this looks fantastic! i was drooling during the demo video - you've improved on so many things from the standard spreadsheet at once.if i could make a request for the api: it's almost impossible to get a simple json serialization of a table in google spreadsheets. there are so many times where i just want to make a really simple mvp with a backend data source, but don't want a full-fledged firebase database. if i could just stick my data in an airtable sheet and point my javascript to load it and create the page dynamically, that would be ideal. |
show hn: airtable, a real-time spreadsheet-database hybrid
| this looks fantastic! i was drooling during the demo video - you've improved on so many things from the standard spreadsheet at once.if i could make a request for the api: it's almost impossible to get a simple json serialization of a table in google spreadsheets. there are so many times where i just want to make a really simple mvp with a backend data source, but don't want a full-fledged firebase database. if i could just stick my data in an airtable sheet and point my javascript to load it and create the page dynamically, that would be ideal. | here's one use case: i tried importing a 500+ row wine spreadsheet into it (it has columns like type, producer, name, vintage, rating, source etc).i had to save the file as csv and import it -- or at least that's what i thought, but filepicker seems to support picking from google drive. i don't know if it would have accepted a spreadsheet.your app seems sluggish to scroll compared to google docs at that size, and the record density seems low (i see 29 records per page vs 50 on google docs). this is using chrome 38 on linux.the "link to another table" seems interesting, but my data came denormalized so i have a column with e.g. 10 different repeated values on 500 rows. maybe it would be nice to automatically clean that up somehow. for example, i could copy the column and have a some paste (unique values only) option. maybe the dialogue that comes up (suggesting to expand the spreadsheet) could even tell you about that. or maybe there could be an option to convert a text field to a separated linked table.i had some conditional formatting set up via google docs, which could be nice to have here; e.g. red wines have a red background in that wine type table.i don't have a simple primary key -- it's really a composite of {wine producer, wine name, vintage}. the app didn't mind importing non-unique values into that first column. i don't know what the alternative might have been -- an auto-generated primary key?having said all of the above, i really like google docs and it will take some amazing features to make me switch to anything else. multi-user planning and documentation via docs is great -- i have a shopping list everyone can update and i can see it change real time on the phone in the shop while someone is changing it from their desktop, and it's always up to date. |
show hn: airtable, a real-time spreadsheet-database hybrid
| here's one use case: i tried importing a 500+ row wine spreadsheet into it (it has columns like type, producer, name, vintage, rating, source etc).i had to save the file as csv and import it -- or at least that's what i thought, but filepicker seems to support picking from google drive. i don't know if it would have accepted a spreadsheet.your app seems sluggish to scroll compared to google docs at that size, and the record density seems low (i see 29 records per page vs 50 on google docs). this is using chrome 38 on linux.the "link to another table" seems interesting, but my data came denormalized so i have a column with e.g. 10 different repeated values on 500 rows. maybe it would be nice to automatically clean that up somehow. for example, i could copy the column and have a some paste (unique values only) option. maybe the dialogue that comes up (suggesting to expand the spreadsheet) could even tell you about that. or maybe there could be an option to convert a text field to a separated linked table.i had some conditional formatting set up via google docs, which could be nice to have here; e.g. red wines have a red background in that wine type table.i don't have a simple primary key -- it's really a composite of {wine producer, wine name, vintage}. the app didn't mind importing non-unique values into that first column. i don't know what the alternative might have been -- an auto-generated primary key?having said all of the above, i really like google docs and it will take some amazing features to make me switch to anything else. multi-user planning and documentation via docs is great -- i have a shopping list everyone can update and i can see it change real time on the phone in the shop while someone is changing it from their desktop, and it's always up to date. | <link> |
25 tools every new developer needs
| "most open source tools and frameworks were developed for osx, then ported to windows or linux"gcc? | obviously the initial comment about most developers using macs was pretty dumb but as a mac user, i do use a lot of these things. maybe it should be titled "tools for new developers using macs"?i don't use sourcetree, alfred, jumpcut, spectacle, screen hero. i also don't think there's any special dependency of a developer on something like file vault or time machine. in fact, using git and a remote server naturally backs up nearly all i do. also, not sure why a developer would have a special need for lastpass or incognito mode?ok, so i use some of these things. :) |
25 tools every new developer needs
| obviously the initial comment about most developers using macs was pretty dumb but as a mac user, i do use a lot of these things. maybe it should be titled "tools for new developers using macs"?i don't use sourcetree, alfred, jumpcut, spectacle, screen hero. i also don't think there's any special dependency of a developer on something like file vault or time machine. in fact, using git and a remote server naturally backs up nearly all i do. also, not sure why a developer would have a special need for lastpass or incognito mode?ok, so i use some of these things. :) | same thing again. "coder" is always a "web coder" now. that's annoying. |
25 tools every new developer needs
| same thing again. "coder" is always a "web coder" now. that's annoying. | rather than using jumpcut, i'd recommend the alfred plugin for clipboard history.<link> |
25 tools every new developer needs
| rather than using jumpcut, i'd recommend the alfred plugin for clipboard history.<link> | "there's a reason most people in tech use macs: most open source tools and frameworks were developed for osx, then ported to windows or linux. "what cloud does this guy live on? |
'battlestar galactica' is leaving netflix. watch it all before october 1st
| why? is battlestar galactica going to another service? | nooooo! i'm sure this screws up a lot of people's to-do lists, like it did mine. |
'battlestar galactica' is leaving netflix. watch it all before october 1st
| nooooo! i'm sure this screws up a lot of people's to-do lists, like it did mine. | man, and i have half of season four to finish... and it's not streaming free on prime either. |
'battlestar galactica' is leaving netflix. watch it all before october 1st
| man, and i have half of season four to finish... and it's not streaming free on prime either. | i'd like to correct the headline to "watch most of it before october 1st".if you haven't finished it, stop when you notice it starts getting bad. it doesn't get better. |
'battlestar galactica' is leaving netflix. watch it all before october 1st
| i'd like to correct the headline to "watch most of it before october 1st".if you haven't finished it, stop when you notice it starts getting bad. it doesn't get better. | i joined netflix way back when, before they had anything decent to stream online. i also have an amazon prime membership, hbo go, as well as a hulu+ subscription. of the four services, netflix held the most promise and is the one that underdelivered the most. i love that they don't bother you with commercials. i love that their service is actually fast and available. their app on roku is faster than hulu's and always seems to select better video quality if bandwidth is constrained. however, i end up almost always reaching for hulu. netflix just stopped expanding, and in some cases started losing content.on the plus side, they have entertaining shows such as house of cards and orange is the new black. however, i think at this point hulu and to an extent amazon has them beat in terms of new content from traditional sources. i will continue paying my $9/month, for now. however, i do hope they improve enough to justify spending money on them. |
juno: a free environment for the julia language
| what's the license? i couldn't find any mention of a license anywhere. lighttable itself seems to be gplv3, so i assume its plugins are too? | mike innes gave a talk/demo of this at the boston/cambridge julia meetup [1] last night and it was really impressive. some highlights:- interactive sliders to manipulate input parameters- support for inline plots (like ijulia)- very nice code completion[1] <link> |
juno: a free environment for the julia language
| mike innes gave a talk/demo of this at the boston/cambridge julia meetup [1] last night and it was really impressive. some highlights:- interactive sliders to manipulate input parameters- support for inline plots (like ijulia)- very nice code completion[1] <link> | julia's been on my to-check-out list for a very long time, as has lighttable. it took me less than five minutes on a brand new machine to install the dependencies and get it working.this is really impressive. |
juno: a free environment for the julia language
| julia's been on my to-check-out list for a very long time, as has lighttable. it took me less than five minutes on a brand new machine to install the dependencies and get it working.this is really impressive. | this is great, as it really lowers the barrier of entering julia.it also offers a level of exploration (with the in-place results), which is a neat thing for students – i've found that interactive environments generate interesting solutions to otherwise boring problems. i'm currently trying to replace matlab as the language i use in our university's probability class and juno adds to "why use julia" argument :)thanks! |
juno: a free environment for the julia language
| this is great, as it really lowers the barrier of entering julia.it also offers a level of exploration (with the in-place results), which is a neat thing for students – i've found that interactive environments generate interesting solutions to otherwise boring problems. i'm currently trying to replace matlab as the language i use in our university's probability class and juno adds to "why use julia" argument :)thanks! | is there a video somewhere of someone using this? i'd love to see this in action, without having to install it myself and read a bunch of docs to see what it's good at. |
annotatious: reddit + delicious
| hey whoever's building this -- please give me a shout. christianpbrink at gmail. if you haven't built it yet, maybe you don't need to, as i already have and may be interested in a partnership. | i have to plug hypothes.is [1] who are doing this in an open source and distributed fashion. they have quite a team and a lot of grant money to make it happen.[1] <link> |
annotatious: reddit + delicious
| i have to plug hypothes.is [1] who are doing this in an open source and distributed fashion. they have quite a team and a lot of grant money to make it happen.[1] <link> | sounds similar to diigo: <link> |
annotatious: reddit + delicious
| sounds similar to diigo: <link> | i wonder if they have a mvp for testing out yet, or if its a collection of emails to register the level of interest before the pilot is built? |
annotatious: reddit + delicious
| i wonder if they have a mvp for testing out yet, or if its a collection of emails to register the level of interest before the pilot is built? | love the idea - thought of it 7+ years ago! hope you guys succeed. |
father and daughter reunion
| url changed from <link> | yeah... my assumption when an executive says they are leaving to "spend more time with my family" is still that they were asked to leave and are trying to be graceful about it.(i mean, i know nothing about bond markets. it could very well be sincere in this case. i'm just saying, my immediate impression, when a top person says they are leaving to spend more time with their family, is to extend my sympathy at, you know, getting canned)maybe it is just my experience at yahoo in the mid aughts. it seemed like every three months, a top exec was "leaving to spend more time with my family"i mean, being graceful about getting pushed out is a good thing for all involved, and it seems like the leaving letters are getting more convincing lately, which is fine, but... yeah, i'm still not buying it. |
father and daughter reunion
| yeah... my assumption when an executive says they are leaving to "spend more time with my family" is still that they were asked to leave and are trying to be graceful about it.(i mean, i know nothing about bond markets. it could very well be sincere in this case. i'm just saying, my immediate impression, when a top person says they are leaving to spend more time with their family, is to extend my sympathy at, you know, getting canned)maybe it is just my experience at yahoo in the mid aughts. it seemed like every three months, a top exec was "leaving to spend more time with my family"i mean, being graceful about getting pushed out is a good thing for all involved, and it seems like the leaving letters are getting more convincing lately, which is fine, but... yeah, i'm still not buying it. | it's definitely one of the better crafted "spend more time with the family" announcements.barring other evidence, there's no particular reason to believe this is different than any other executive who is being encouraged to resign from their organization. |
father and daughter reunion
| it's definitely one of the better crafted "spend more time with the family" announcements.barring other evidence, there's no particular reason to believe this is different than any other executive who is being encouraged to resign from their organization. | for those that don't follow the bond markets, his former company, pimco, would be the google or microsoft of bond traders.tangentially, bill gross, the other pillar of pimco just announced today that he was leaving as well.pimco's funds are reportedly being hammered with withdraws. estimates are that they could loose up to 30% of their managed assets, and with almost 2 trillion managed that's market moving news!!see: <link> if you think you are having a bad day......imagine you are at google and in the span of a few months both larry and then sergey unexpectedly announce they are leaving.> mr el-erian used to leave home for work at 4.30am each morningi always wondered why someone who trades in the markets would live out west where you need to be ready at 6:30 for the market open.i'm guessing someone has a smart ass comment lined up about how being around for your kids should be the default, or how if they made 100 million in a year that they'd quit too, but i think articles like this are good, in that it reminds hard charging people to stop and look around once in a while. |
father and daughter reunion
| for those that don't follow the bond markets, his former company, pimco, would be the google or microsoft of bond traders.tangentially, bill gross, the other pillar of pimco just announced today that he was leaving as well.pimco's funds are reportedly being hammered with withdraws. estimates are that they could loose up to 30% of their managed assets, and with almost 2 trillion managed that's market moving news!!see: <link> if you think you are having a bad day......imagine you are at google and in the span of a few months both larry and then sergey unexpectedly announce they are leaving.> mr el-erian used to leave home for work at 4.30am each morningi always wondered why someone who trades in the markets would live out west where you need to be ready at 6:30 for the market open.i'm guessing someone has a smart ass comment lined up about how being around for your kids should be the default, or how if they made 100 million in a year that they'd quit too, but i think articles like this are good, in that it reminds hard charging people to stop and look around once in a while. | i appreciate the sentiment of the article, but i hate that our culture plays up this idea that being a good parent means being there for all the arbitrary "milestones." i honestly don't remember if my dad was there or not on my first day of school, etc. i sure as hell remember that he called me when i got home from school every day to make sure i was okay and not getting into trouble.you're not going to be able to convince corporate america that a kid's dance recital is more important than a client meeting. all you'll do by stressing that dance recital is make busy parents feel bad for not being able to live up to some ideal, or discourage ambitious people from having kids because they think they can't be good parents too. you'll get a lot further trying to push back on ideas like "face time" or competing on who can take the least vacation. don't have much work? go home at 2 and take your kid to a museum. that's just as valuable as your being there for an arbitrary "milestone." |
translating technological terms throws up some peculiar challenges
| > “windows” became “eyes”.that is quite possibly because the word “window” literally is “wind-ow” or “wind-eye” — i.e. an eye, or opening, for the wind to blow through, which it did before windows had glass, and instead had shutters, like eyelids. | it's an interesting article and i can certainly get behind the sentiment of making software more accessible to non-english speakers. that said, if another language already has a word with the desired meaning, why not use it instead of inventing a new one?this is a two-way street. much of the english vocabulary comes from other languages: trek, pundit, taboo, gauntlet, moped, etc. and many languages use english loanwords for technical terms. just try translating "internet" to other languages. even japanese is "インターネット" (intaanetto)[1].1. <link> |
translating technological terms throws up some peculiar challenges
| it's an interesting article and i can certainly get behind the sentiment of making software more accessible to non-english speakers. that said, if another language already has a word with the desired meaning, why not use it instead of inventing a new one?this is a two-way street. much of the english vocabulary comes from other languages: trek, pundit, taboo, gauntlet, moped, etc. and many languages use english loanwords for technical terms. just try translating "internet" to other languages. even japanese is "インターネット" (intaanetto)[1].1. <link> | a few hopefully relevant thoughts about translating docs1) translating a piece of english ui will almost double the length of it. that's annoying.2) documentation is hard to write and to maintain, and sometimes it's hard to read and use documentation. creating multilingual documentation is even harder, and it's nearly impossible to have an entire coverage of the docs in all languages, which leads to incomplete or outdated/incorrect documentation - which, as the saying goes, is worse than none. multilingual software inevitably makes the documentation situation of a software worse. trying to apply language a docs to a language b interface is hell.3) documentation alone isn't part of the story; forum threads, answered support questions and that kind of discussion-style paradocumentation is ranges in importance from helpful to just full-on essential and that kind of stuff is never really translated. trying to get help from a mostly english community for a software translated is hell.4) the sublanguage used in a given ui isn't written in a given tongue; really, it's written in a pidgin made of the original language of the application plus the slang of the domain of the application, which is usually tied to one, say, culture or spoken language in particular; very large software also typically creates its own sub-language, which separates itself from actual tongue the ui could say to have been written in. after all, any description of a what a software does, expressed in a language evolved for the interaction of human beings in the real, material world, must be a vague and fragile metaphor. translating metaphors is nearly impossible. it's hell. it's poetry, but when poetry is not the goal, it's hell.that also means that to read the documentation and ui of an application amounts to learning the sublanguage of that application, which as i explained is different from the actual human language it was written in. i can understand that "save" means "serialize the relevant subset of the application state to a file" without knowing anything about the "dictionary" meaning of the english word "save". it's not the same thing. "save" could be replaced by "blub" as far as i'm concerned, but it's convenient for people who speak english to associate "serialize" to the concept of preserving, of keeping. but it's not as important as the consistency of the language developed to describe the concepts and operations of that application. a single "blub" is better than "save" and "sauvegarder". duplicating the pidgin a software develops for itself makes using that software a special kind of hell.conclusion: it's easy to end up in situations where i have to use french software here in quebec. most of my friends have laptops and phones with french uis. it makes using them a pain in the ass.i absolutely, deeply despise translated uis; translating a ui is almost systematically a counter-productive activity. it's hell. |
translating technological terms throws up some peculiar challenges
| a few hopefully relevant thoughts about translating docs1) translating a piece of english ui will almost double the length of it. that's annoying.2) documentation is hard to write and to maintain, and sometimes it's hard to read and use documentation. creating multilingual documentation is even harder, and it's nearly impossible to have an entire coverage of the docs in all languages, which leads to incomplete or outdated/incorrect documentation - which, as the saying goes, is worse than none. multilingual software inevitably makes the documentation situation of a software worse. trying to apply language a docs to a language b interface is hell.3) documentation alone isn't part of the story; forum threads, answered support questions and that kind of discussion-style paradocumentation is ranges in importance from helpful to just full-on essential and that kind of stuff is never really translated. trying to get help from a mostly english community for a software translated is hell.4) the sublanguage used in a given ui isn't written in a given tongue; really, it's written in a pidgin made of the original language of the application plus the slang of the domain of the application, which is usually tied to one, say, culture or spoken language in particular; very large software also typically creates its own sub-language, which separates itself from actual tongue the ui could say to have been written in. after all, any description of a what a software does, expressed in a language evolved for the interaction of human beings in the real, material world, must be a vague and fragile metaphor. translating metaphors is nearly impossible. it's hell. it's poetry, but when poetry is not the goal, it's hell.that also means that to read the documentation and ui of an application amounts to learning the sublanguage of that application, which as i explained is different from the actual human language it was written in. i can understand that "save" means "serialize the relevant subset of the application state to a file" without knowing anything about the "dictionary" meaning of the english word "save". it's not the same thing. "save" could be replaced by "blub" as far as i'm concerned, but it's convenient for people who speak english to associate "serialize" to the concept of preserving, of keeping. but it's not as important as the consistency of the language developed to describe the concepts and operations of that application. a single "blub" is better than "save" and "sauvegarder". duplicating the pidgin a software develops for itself makes using that software a special kind of hell.conclusion: it's easy to end up in situations where i have to use french software here in quebec. most of my friends have laptops and phones with french uis. it makes using them a pain in the ass.i absolutely, deeply despise translated uis; translating a ui is almost systematically a counter-productive activity. it's hell. | while i appreciate that they're trying to do good, for people interested in communicating with people from other cultures translated technological terms have always been a massive bane. i live in germany and while germans generally have little patriotism they're fierce defenders of their language, which led to all sorts of domain-specific terms being translated across many different kinds of media. be it movies, books, technical manuals, programming instructions.and for all of them it's a common theme that someone who believes they know a thing, might even consider themselves an expert, they have trouble then branching out of their linguistic bubble. when talking to experts in their domain, from another language, they suddenly cannot communicate anymore, even if they speak the other language fluently, as they need to learn an entirely new vocabulary just for basic interaction.and this is not just a theoretical situation. i've had problems when talking about books i read in my youth, because often important words or even names had been changed and the person i was talking with knew the same characters, concepts or ideas i knew but neither of us knew under what names the other knew them. i learned perl on the internet, on my own, using exclusively the english source documents; when talking to german perl programmers i've repeatedly had issues because they had learned it in german courses which used german cs terms. instead of actually discussing business we spent half the time sorting out what to call a thing. i also translated mmo content professionally for a while. while our translations were great and allowed the germans a perfectly fluid native experience, we found that many germans tended to switch the game to english after a while, because they couldn't talk to other players since the item names, skill names and location names were all different.there are some situations where it's fine to use a translated term. example: mouse. in german the word is written maus and pronounced identically. in such a situation where the difference is minimal, a translatio can be fine.however in all other situation i believe translation of domain-specific words does damage in the long term. |
translating technological terms throws up some peculiar challenges
| while i appreciate that they're trying to do good, for people interested in communicating with people from other cultures translated technological terms have always been a massive bane. i live in germany and while germans generally have little patriotism they're fierce defenders of their language, which led to all sorts of domain-specific terms being translated across many different kinds of media. be it movies, books, technical manuals, programming instructions.and for all of them it's a common theme that someone who believes they know a thing, might even consider themselves an expert, they have trouble then branching out of their linguistic bubble. when talking to experts in their domain, from another language, they suddenly cannot communicate anymore, even if they speak the other language fluently, as they need to learn an entirely new vocabulary just for basic interaction.and this is not just a theoretical situation. i've had problems when talking about books i read in my youth, because often important words or even names had been changed and the person i was talking with knew the same characters, concepts or ideas i knew but neither of us knew under what names the other knew them. i learned perl on the internet, on my own, using exclusively the english source documents; when talking to german perl programmers i've repeatedly had issues because they had learned it in german courses which used german cs terms. instead of actually discussing business we spent half the time sorting out what to call a thing. i also translated mmo content professionally for a while. while our translations were great and allowed the germans a perfectly fluid native experience, we found that many germans tended to switch the game to english after a while, because they couldn't talk to other players since the item names, skill names and location names were all different.there are some situations where it's fine to use a translated term. example: mouse. in german the word is written maus and pronounced identically. in such a situation where the difference is minimal, a translatio can be fine.however in all other situation i believe translation of domain-specific words does damage in the long term. | there's an obvious bias here. the piece ends with "localisation may keep small languages alive". isn't that adding to the problem? we don't try to keep small programming languages alive (except a few diehards). many small languages are the problem itself. they're not something to aim for.i should add that i'm not complaining about making phones more accessible to more people. that's absolutely great. but why try to go further and use it as an excuse to drag out languages that the world would be better off without? |
which gpus to get for deep learning
| is there any reason why there isn't any amd gpu? mantle? | the new maxwell 970/980 kit is very interesting from a compute perspective.970: $329 for 3494 sp / 109 dp gflops @ 145w tdp980: $549 for 4612 sp / 144 dp gflops @ 165w tdp |
which gpus to get for deep learning
| the new maxwell 970/980 kit is very interesting from a compute perspective.970: $329 for 3494 sp / 109 dp gflops @ 145w tdp980: $549 for 4612 sp / 144 dp gflops @ 165w tdp | the best value right now are used amd gpus. gpu cryptocurrency mining has become unprofitable so the used market is saturated with 290, 280x, and other amd gpus. if you have a limited budget this is probably the way to go. |
which gpus to get for deep learning
| the best value right now are used amd gpus. gpu cryptocurrency mining has become unprofitable so the used market is saturated with 290, 280x, and other amd gpus. if you have a limited budget this is probably the way to go. | i disagree with his position on memory (he mentions in the post that anything above 1.5gb should be fine). in my experience, anything below 3gb can be pretty uncomfortable these days, if you want to work on serious problems.it's not just about fitting the parameters into gpu memory, but also all the operations you perform on them, which can require a lot of intermediate storage. the example he gives only has fully-connected layers, but convolutional neural networks tend to require more space, especially some more recent implementations (e.g. fft-based convolutions or the gemm approach used by caffe).he mentions that his network (fully connected) has 52m parameters and compares it to krizhevsky's 2012 imagenet network (convolutional) network, which had 60m. but krizhevsky actually explicitly mentions in his paper that memory was an issue:"a single gtx 580 gpu has only 3gb of memory, which limits the maximum size of the networks that can be trained on it. it turns out that 1.2 million training examples are enough to train networks which are too big to fit on one gpu. therefore we spread the net across two gpus." (from <link> ) |
which gpus to get for deep learning
| i disagree with his position on memory (he mentions in the post that anything above 1.5gb should be fine). in my experience, anything below 3gb can be pretty uncomfortable these days, if you want to work on serious problems.it's not just about fitting the parameters into gpu memory, but also all the operations you perform on them, which can require a lot of intermediate storage. the example he gives only has fully-connected layers, but convolutional neural networks tend to require more space, especially some more recent implementations (e.g. fft-based convolutions or the gemm approach used by caffe).he mentions that his network (fully connected) has 52m parameters and compares it to krizhevsky's 2012 imagenet network (convolutional) network, which had 60m. but krizhevsky actually explicitly mentions in his paper that memory was an issue:"a single gtx 580 gpu has only 3gb of memory, which limits the maximum size of the networks that can be trained on it. it turns out that 1.2 million training examples are enough to train networks which are too big to fit on one gpu. therefore we spread the net across two gpus." (from <link> ) | for those looking to do it on the jvm, i have a prepacked scientific computing framework that might be interesting:<link> is a generic wrapper with ndarrays for cuda and normal blas operations. deeplearning4j (my deeplearning project) also has support for gpus (it's built on nd4j)stable version coming soon =dfor those of you in python land, i would look in to theano |