jw2013

>>It may be that bad comments tend to be concentrated on particular threads, and that the right way to implement pending comments is per-thread rather than site-wide.

So what about this? Here comments is a vector of all the scores of comments inside one particular thread.

----

(defn filter? [comments threshold]

  (let [size (count comments)]

    (if (< threshold (/ (reduce + comments) size)) false

       true)))

;;user=> (filter? [1 2 3 4] 2)

;;false

;;user=> (filter? [1 2 3 4] 3)

;;true

Again, thanks very much. Hope can get the chance to work for Stripe one day.

(inc kudos)

I disagreed, and am pretty happy with the openness of current HN system. I'd rather see some negative comments than missing out a few inspiring ones.

What I do find annoying is (as some have mentioned days/weeks ago)- the OP can delete his own post, even though there are many replies below. So now the other hackers have no idea what the rest of people are replying to.

>> I'd argue that your chance of winning the bracket (or, at least, a valuation of it) is much higher than that given your ability to predict all of the matches prior.

I believe the chance to win given you win all but three are higher than 12.5%. But MUCH higher? I am not quite sure. There are too many games this year can go either way, and predicting these games right is more on luck than technique analysis. Though winning all these previous prediction did gave you the credit in your expertise, but I am not quite sure about MUCH higher statement.

Like I have said "Of course the reality is how much you would sell it very much depend on your conviction of how likely your guesses on final 3 games may be correct.". As tournaments goes on, your feeling about we will win it all will change. Say Louisville will all the team 100-0 in previous game, yet you predicted them to lose at final four, you are more likely to sell your stake at a cheaper price than not.

If these four teams are equally competent. Then you have 50% chance to choose each of three correct (2 final four games + 1 title games correctly). So the chance to win it all is (1/2)^3 = 0.125. Your expected return is $1B * 55% * 0.125 = $68.75M.

Or if you prefer a single payout, $300M * 0.125 = $37.5M. So considering inflation for $68.75M is over 40 years, maybe your expected return is between 37.5M to 50M (idk about 50M, I made it up)? Of course the reality is how much you would sell it very much depend on your conviction of how likely your guesses on final 3 games may be correct.

Now they are all done. Thanks, SF Austin. I got it wrong, too.

No. Just took a random look at a couple of them, and one of the poster registered 5 days ago. I am guessing maybe 7 days?

wow, just upvoted your helpful reply. Can I ask what/how to prepare for the Stripe interview? Looks like cramming algorithm questions is not going to work. And it looks like, through your and OP's experience, Stripe is looking for people that are cool to work and hang-out with, self-motivated, and can write clean code.

I will really appreciate your reply.

john resig- secrets of javascript ninja

absolutely the best

I think for the crowd you described, why not use Prolog? Prolog's idea is very much like what Math is to me (I only have a undergrad math degree, so don't count what I think is by any means absolutely correct)- here are some truth/axioms, can you deduce further theorems, and give result whether certain statements are true or false.

They left the most valuable gift to their children - great education from their parents, and they gave the most valuable present to some less priviledged children in the world - the chance to be educated with the help of Bill&Melinda foundation. Well done, Bills.

I deployed many apps on Heroku including but not limited to: Ruby/Rails, Python/Flask&Django, Ring/Clojure, Node.js/vanilla without framework&Express(I haven't tried Hapi there, but yes Heroku support it), PHP/vanilla without framework.. I want to say they are so good at letting you deploying all different kinds of languages/framework at ease.

One caveat is that Heroku is a ephemeral filesystem on each Dyno. So if you want to save the file content on the fly, you probably also need to use services like AWS S3. One time in a hackathon I tried to deployed a shared proxy for letting different users share the same html DOMs, and I need to save all the resources user requested from remote server. Since the time limit I just completely switch the site to be hosted on elastic beanstalk (because I never tried Heroku with S3 before). Using Heroku + S3 is actually easy, later I found out. But having to use another storage host might be something you want to know ahead before you try heroku.

>> Are page speeds okay?

It's very slow for the first request, if your site haven't been visited for a while. ("If your application is unused for a while it gets unloaded from the server memory" Quoted from Stackoverflow.) But afterwards I find it fast enough. But that also depends very much on your backend service and infrastructure, so I can't make any promise even though my guess would be it will be okay.

>> Is the uptime satisfactory?

Yes from my experience. I only had very few downtime. One experience I learnt in the hard way is at least have some mechanisms to alert yourself when the site is down. (So you can get up in 3am to fix the issue!!!)

>> Do you have personal experience with high-traffic sites running on Heroku?

How high is your traffic?

>> Heroku and the promise of no systems administration sound tempting, but...

I love their Add-ons. For example, simple things like a scheduler makes the automation so easy.

Thanks so much! That really helps.

Thanks for the great article and congrats on your Stripe offer. Great company to work for.

Despite the great story of your general interview process, I am quite interested in the hiring process of Stripe. What is its interview like? What do you think probably make you got the offer?

he is probably talking about using hashtable, but that needs linear memory so there is a trade off.

Gmail is free to you, but their ads can earn more money by mining the info in your email. Is this really free here? Your email content has value.

>> "trying to kickstart things again past your formative years seems futile"

The point is almost none of these successes come overnight. Admitted, for example, many of success founders graduated from Stanford. But a lot of them they worked hard in high school to get into Stanford. So you may see A (success), but to get to A, they worked hard on B (some good background to get to A), and to get B they worked hard on C (some good background to get to B). A <- B <- C <-D <- E... Okay, now assume (just assuming, I am not saying that is in fact where you are) if you are at E you might feel A is so far away, and people at B is so lucky to be at B. But if you start doing little things and work hard, you probably (I am not saying definitely because determination is the only factor here) can start to get to D, C, and then B. Who knows? Why not give it a shot? Don't give up. [If you just stare at "A" you probably won't feel you are close, but start with getting to D!] There are plenty of time left. One thing about some media at this age I don't quite like is to idolize those 20-some startup founders, and making it sound like not to be successful at your 20s is a complete failure. That is so not true, there are plenty 40+ internet billionaires out there.

>> "But that hope is quickly vanishing the more and more I read about these individuals' backgrounds."

Does the great background help? Of course. But one thing about the internet age is the cost to start a startup is so low. And you know what the best part is? Pixels don't care. The investors may care about what kind of background you have, but your customers don't. Have you tried, say, make a simple service (e.g. website, or other form even not related to internet) some time? If not, please give it a try. If you never try, you will never fail. And it's okay to fail. None of the startups are guaranteed to be successful. Actually possibly none of your startups will be successful if you are very unlucky. But that shouldn't stop you from trying. Again, don't stare at billion (I am not saying you are), if it helps. That is like winning a lottery, and I don't think that's the right mindset. Do something small. Do something interests you. Build something just because you love building that.

>> I don't mean to rant or derail the topic, and apologies if I do.

Don't be. And thanks for reading all my comments if you made it this far. I hope my comment is not useless. Best luck, man. Don't give up.

I agree with all of what you and Patrick said. Still most likely the student on OPT is in a much better position to start a startup than most people on H1b. Once you are hired by some company on H1b, you want to start your own startup? You need to prove your startup is viable (some funding constraint + providing prevailing wage for yourself I remember). But how you can have a viable startup that has not even been started? That kind of chicken-and-egg problem stuck tons of aspiring entrepreneurs on H1b.

Will really appreciate anyone here that can give some solution under current circumstances.

Thanks for the comment. I am not proposing using OPT as a founder visa, all I am saying is using OPT is one viable alternative, maybe not as cool as a founder visa, but you can still pull it off in the condition I stated in my other response to you.

I am pretty sure even after you graduate and get an OPT, you are still on F1 visa. But the point is 29 months (if you are a STEM major) is enough for your startup to generate some revenue, and then you can apply for H1b (as long as you prove your startup has sufficient fund, your wage from your startup is prevailing wage, and you can be fired from your startup).

If a startup fails, it usually fails quickly. Yes, sometimes a company, like airbnb, went quiet for a very long time before it hit it off, but 29 months are likely enough.

Some points on my mind:

a) Most adventurous student entrepreneurs don't really MUST need a founder visa, they can use OPT. I think founders of Stripe did that.

b) Most non-immigrants already working (mostly with H1 visa) needs a founder visa. To work on a startup you founded in H1b visa is hard, consider H1b visa needs some funding/salary requirement on your startup, and that is hard to get by on the early stage of a startup. Giving these people the visa may be important to the U.S. economy consider that group of budding founders are with the most experiences.

c) Maybe approval of founder visa won't be in the near future. Right now that is really tangled with other messes in the Immigration Act, which seems hopeless at the House in the stage.

d) Really +1 for founder visa. Can't afford to lose foreign entrepreneurs to Canada.

> My training set is specifically designed to be conversational interview and personal questions, but I think a lot of the people who reach the site don't grasp that.

I did not either, as soon as I start to ask about general questions about you (e.g.'what's your email address'), the result got better. Now I know your answer is predefined to your personal info, I know why other questions won't work well.

Perhaps many people, like me, only read the div starting with "Let's chat", but then get started immediately (because the red recording button caught my attention right away) and totally ignored the div "I'll try to answer here" with your intent written.

Wow, ddod your voice>>text is awesome. Did you implement it all on your own?

Some improvement text>>comprehension will be great. Right now it does not understand many of my queries. Keep it up the good work!

For those of you wondering what were they doing in the huge gap from S11 to two months before. Here are some info:

1) The startup they applied to YC is called Yardsale, and it's still alive: https://www.getyardsale.com. "Yardsale, a mobile app to help folks sell goods to local buyers."

2) FOBO is founded by Yardsale, Inc. FOBO received $1.6M in Seed funding. (10/18/13)

3) More stories here: http://josephwalla.com/how-to-hustle-and-launch-in-3-weeks-t...

"It’s been just about two months since FOBO launched its local marketplace app for consumer electronics."

So they were probably making the app before launching.

Is that a coincidence: https://twitter.com/tpdubs2/status/444509011522891776

Just last Friday his wife's startup is accepted to Techstars NYC, and she is going to be at NYC. It may be a good timing for Tom to take a leave to be with his wife.

Just guessing here. I could totally be wrong. No offense to Tom; I looked up to him.

>>> This narrows down the choices: the "founder" who is the cause for all this drama is either PJ Hyett or Tom Preston-Werner.

Any guess?

I am not assuming everything Julie states is right. For statement "the culture of Github needs some fix" to be true, it only needs only some of Julie's statements to be true. And looks like from Chris's statement, that is true (otherwise why did they "punish" founder & his wife).

"the relevant founder has been put on leave... The founder’s wife... will no longer be permitted in the office."

Almost exactly what I expected. As the startup grows larger (200+ employees now for Github?) it's not uncommon to have some cultural setback. +1 for Github trying to fix it asap, but still Chris Wanstrath did not mention how are they exactly going to fix the culture. Putting founder on leave is by no means panacea. The behavior on that one founder is very likely not the cause for the wrong culture, but just the side effect. So more specific plans INSIDE the company, please? (well, actually I think fixing the culture can't be done by any plan, but it must be done by example of founders/high-level executives.)

When you type in some querying strings the backend just fails, for example:

http://inf.demos.dam.io/sdfsdfs

I have yet to find a pattern, but I guess the author did some sort of hashing to the query string and use the hashcode as the seed to generates random text. For some given strings, the backend fail has to do with the use of the hashcode probably. (e.g. use [some_variable_or_constant_here]/[hoshcode - CONSTANT], when hoshcode == CONSTANT, the backend code does not contain the exception handling code.) Just a guess.