Hacker Newsnew | past | comments | ask | show | jobs | submit | 2012-06-02login
Stories from June 2, 2012
Go back a day, month, or year. Go forward a day, month, or year.
1.Gmaps.js the easiest way to use Google Maps (hpneo.github.com)
276 points by DanielRibeiro on June 2, 2012 | 40 comments
2.WhosHere's response to Brian Hamachek/Who's Near Me (zendesk.com)
222 points by nphase on June 2, 2012 | 138 comments
3.Valve hired their first wave of Linux developers, hiring more (phoronix.com)
216 points by shazow on June 2, 2012 | 87 comments
4.Folly - The Faceboook open source library (facebook.com)
200 points by sidcool on June 2, 2012 | 43 comments
5.Well-Kept Gardens Die By Pacifism (lesswrong.com)
182 points by llambda on June 2, 2012 | 99 comments
6.A Boggling Return to C (thraxil.org)
162 points by ColinWright on June 2, 2012 | 85 comments
7.Prostate cancer drug so effective trial stopped (sfgate.com)
149 points by joeyespo on June 2, 2012 | 54 comments
8.Things the TSA doesn't want you to see (elliott.org)
143 points by diminish on June 2, 2012 | 65 comments
9.HackerNews API: What if HN does not have API? Make API on the fly with APIfy (heroku.com)
132 points by sathish316 on June 2, 2012 | 49 comments
10.The Cloud Is Not for You (justcramer.com)
120 points by d0ugal on June 2, 2012 | 53 comments
11.Cloudflare Attack Post Mortem; Apparent Google Apps/Gmail Vulnerability (cloudflare.com)
118 points by demonfly77 on June 2, 2012 | 63 comments
12.Hacker News Rankings - interactive graphs of an item's fate (hnrankings.info)
103 points by ColinWright on June 2, 2012 | 40 comments

We literally bet our life savings on this and years of zero vacations to bring WhosHere to where it is.

Making a bad business decision and living a shitty life doesn't entitle you to sue people who make a similar app just because their app's name uses some of the same letters as your app's name. (Both companies picked crappy names that are poor brands and poor trademarks. It's like calling your hamburger shop "Hamburger Shop" and then suing anyone who sells hamburgers for trademark infringement. It's not their fault you picked a generic name.)

Would you protect your company and its name?

With the facts I've heard today, absolutely not. Taking a competitor to court is an absolute last resort, done only in the most egregious of cases (serious fraud that people are associating with your business, etc.) While the case has legal merits, it has zero ethical merits. You picked a generic name. The other guy picked a similar generic name. He is not trying to compete and is not negatively affecting your business. You are being jackasses.

Making sacrifices to follow your dream does not give you the right to treat other people badly. You may be proud for the sacrifices you have made and the hard work you have put into your idea, but you deserve nothing for it. You're just another group of people equally good as every other group of people. Act like it.

14.IOS 6: Higher Hanging Fruit (imore.com)
95 points by cmer on June 2, 2012 | 90 comments
15.Software Raises Bar for Hiring (googleusercontent.com)
94 points by imjk on June 2, 2012 | 87 comments
16.Fear and Loathing and Windows 8 (blogspot.com.ar)
86 points by rbanffy on June 2, 2012 | 95 comments
17.Musl libc (etalabs.net)
85 points by pmarin on June 2, 2012 | 22 comments
18.With the #25 issue, Hacker Monthly celebrates its 2 Year Anniversary (hackermonthly.com)
83 points by bearwithclaws on June 2, 2012 | 18 comments
19.Pinterest Growth Hacks: How did it grow so fast? (adambreckler.com)
80 points by abreckle on June 2, 2012 | 6 comments
20.CoffeeScript2 - A Redesign of the CoffeeScript Compiler (github.com/michaelficarra)
76 points by netbyte on June 2, 2012 | 42 comments
21.Peter Thiel's Startup Class 16 Notes: Decoding Ourselves (blakemasters.tumblr.com)
74 points by DanielRibeiro on June 2, 2012 | 25 comments

I was at the announcement of the trial data by the investigators this morning. The results were also discussed by a statistician not involved with the trial directly. Not sure that hacker news needs an in depth analysis of the trial design and outcome, but anyway...

In summary, the trial looked at patients who had incurable prostate cancer that had spread and stopped responding to traditional hormonal treatments. Patients had to have no symptoms to be enrolled on this trial. The trial compared abiraterone, a new type of hormonal treatment plus steroids to just being on steroids alone. The question that the trial was asking can be framed as "Does taking abiraterone prolong the period of time that a patient can be symptom free and their cancer is under control?" That was the primary endpoint. A secondary endpoint was survival, or how long they lived.

Built in to the study design was a planned interim analysis of the data. Normally, the investigators are not allowed to just look at the data as it accrues continuously because it creates bias. It is a standard approach to allow one look at the data when say 40% of the expected events (in this case, development of progressive cancer) have occurred. When they did this, the difference between getting the drug and not getting the drug was considered big enough to unblind the trial. This means that everyone finds out if they were getting placebo or the real drug.

There are multiple reasons why one may decide to do this. First and foremost is that if the drug really works, it would be unethical to leave people on a placebo for long periods, because it is clear they have an inferior outcome. Stopping early also speeds the time to drug approval and market entry for a given indication.

There are somewhat esoteric statistical procedures that can determine at an interim analysis if the accrued data shows a statistically significant difference between groups, which would therefore justify stopping the trial, while adjusting for the impact of repeated testing on the type I error rate (ie the chance of finding an association that doesn't really exist). In this trial, there is no doubt that the primary endpoint was positive and clinically and statistically significant, even with only 40% of people on the trial having had a progression event. So they stopped. The data on whether abiraterone reduces the chance of death are not mature, because not enough of the people on the trial have died yet, so those results will be awaited. Stopping early means that these results may be biased however, because the trial is no longer blinded.

This was a large and well done trial, that employed standard statistical practices.

23.A Group Is Its Own Worst Enemy (shirky.com)
71 points by llambda on June 2, 2012 | 2 comments
24.How Serial Innovators Find The Best Problems To Solve (fastcompany.com)
68 points by sal9000 on June 2, 2012 | 4 comments

Brian's blog posting (now with a donation button):

  "... the deadline to file a response to the lawsuit had passed
   and that WhosHere had requested a motion for default judgment
   (meaning that they would automatically receive everything they
   had asked for - which would effectively bankrupt me..."

  "The lawyers I have spoken to thus far are asking me for at
   least a $10,000 retainer just to get started on the process
   and I don't have those kind of resources laying around."
Compare to Brian's response in the email thread:

  "I was born wealthy; I have an obscene trust fund. I assure you
   that if required, I will obtain the best legal representation
   for corporate litigation in the Silicon Valley. After your
   last proposal, I will resist this legal action well beyond
   what makes any financial sense, simply out of principal."
And, to complement this, a snippet from Judd Weiss' blog [0]:

  "Quick tip: When you do engage the other side (or their
   lawyer), whatever you do, never say the line “I’m taking this
   all the way to trial, I don’t care what this costs me”.
   Everyone says that. Everyone. That doesn’t work with someone
   like me. I smell blood. “Oh really, you don’t care what this
   costs you? Alright then, let’s find out how much you really
   don’t care.” People who say they don’t care about the costs
   often cave sooner, because they are showing that they’re weak.
   They’re showing that they really don’t have much solid to
   fight you with except their loudly stated tolerance for pain.
   That tolerance is easy to test. And it’s usually very low when
   there’s not much else but puffery to back it up. Any modestly
   wise person cares about the financial effects of litigation.
   Don’t try to pretend you’re stupid, or else you’re going to
   look stupid."
[0] http://hustlebear.com/2010/12/14/how-to-handle-lawyers-threa...
26.On Facebook, ‘Likes’ Become Ads (yahoo.com)
64 points by libria on June 2, 2012 | 20 comments
27.Turkey charges pianist Fazil Say over Twitter posts (nytimes.com)
63 points by ozgune on June 2, 2012 | 45 comments
28.Do software engineers need mathematics? (2000) (maa.org)
62 points by henrik_w on June 2, 2012 | 37 comments
29.When is a good time to submit a story to Hacker News? (hnpickup.appspot.com)
53 points by ColinWright on June 2, 2012 | 20 comments

The thing I love about the TSA is that their record of failure is so numerically concrete.

FTA: “The experience to date is 50,000 false positives and 16 known terrorists not flagged,” says Thomson. “No known terrorists have ever been flagged.”

That's some pretty obvious and epic fail on the part of the Agency. Why hasn't it been eliminated or reformed by now?

Government's inability to allow its ideas to fail, or at very least its inability to pivot toward success is so disheartening. This has really been on my mind a lot today. I watched "Waiting for Superman" last night and I can't get the images of the poor kids just looking for a decent public education out of my mind. The crimes against humanity perpetrated upon us with our own tax dollars should really be prosecuted with all vigor.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: