Haha, except AWS doesn't lock your account randomly or stop droplets for benign abuse reports.

Also, OP required Redshift. DO does not offer that.

I have a similar feeling. I don't really like GOPATH, but I'm trying to come to terms with it.

Your Go "project" goes in $GOPATH/src, i.e. your git repo would be in $GOPATH/src/github.com/user/repo

If you have your main pkg in $REPO/service, then you can build like so: go build github.com/user/repo/service (this can be ran from anywhere as it uses $GOPATH)

If you need to add a dependency, use: go get github.com/.... and go get will add it to $GOPATH/src/ which you can then import.

Hmm, ok. I think I get that -- I'll have to see if I can symlink into the GOPATH; I feel like I read somewhere you can't. I'd like to avoid making my GOPATH be a subdir in the containing repo.

Thanks!

This is pretty nice. I was curious how the spread was like for multiple links: http://www.rndmlnk.com/counts/testredirect2

side note, it was fun to see what the max redirects is in Chrome with the redirect loop.

Was it 1000 :-P

Also I should probably stop it from referencing itself

not anymore :(

I participated in ACM-ICPC regionals back in November and it was a lot of fun. I definitely recommend it to anyone that has the chance to go to preliminaries next year.

The time limits are quite brutal for some of the problems @ finals, most of them I'd never have a proper optimal solution for.

Also, the fact that Peking understood and solved A in just 5 minutes is crazy. Sure, I could at least do A, but they practice enormously just for these.

Not denying the rest of what you say -- these teams are scary good, but A was really simple -- lot simpler than the first problems I remember from when I coached our school's team.

I just tried the brute force solution in Java, took me maybe 3-4 mins to code up and it finished with the max value of n in less than 1s.

I agree that A is really simple. I was pretty surprised to see that as a problem.

Was that 3-4 mins from first starting to read it to finishing a solution? If so then great, but personally I would've taken 10-15 minutes because I'd be carefully reading it first and wondering why it sounds so simple.

Yes, I did know that someone did it in 5 mins, so that helped :)

I think it took me 9-10 mins in total, including testing on the provided inputs. I think I could have done it in a few mins faster 15 years ago when I competed (never made it to the Finals though). Also typically: one of the team members would immediately start coding up the first problem while others are working on the next problems, with the assumption that the first one is quite simple.

I was in the regional (Asia-Kuala Lumpur) where Tokyo University team (3rd in the frozen scoreboard) solved all of the 12 problems in 2.5 hours (out of 5 hours). Yeah, it was really demotivating to see someone else have completed all problems yet you just solved 4 problems. In the end, although my team solved 6 problems and finished at 22nd place, I can say that joining ACM-ICPC regionals is one of my greatest moment in this uni life.

Wow that's crazy, no doubt it must have been really demotivating.. I was in the Northeast NA regional and MIT (#11 in the frozen scoreboard) only solved 6/8 of the problems at regionals. Most of us only solved 1-3.

Is it still allowed to have a set of prepared implementations for standard algorithms and data structures? Still impressive to just need 5 minutes.

You're allowed to bring a reference material of 25 pages, single-sided, A4 or letter size.

http://icpc.baylor.edu/worldfinals/on-site-registration

that's amazing. I would love to see an example team's document.

Stanford's is pretty nice: http://stanford.edu/~liszt90/acm/ (notebook files). There's also one from a Korean university there.

> A crossing in central London had programmed intervals for red and green lights, for example. Pushing the button would only impact the length of these intervals between midnight and 7am.

This is the same where I live. The crossing buttons do nothing during the day (walk symbols still come on for the same length of time), but at night they need to be used. I don't push it during the day, but of course most people still do :-)

> I don't push it during the day, but of course most people still do :-)

I always push the button at pedestrian crossings, even though I'm aware it often does nothing.

Is that because it "creates a sense of togetherness with strangers which might otherwise be absent"? Is it because "doing something is better than doing nothing"? Is it because my "attention is on the activity at hand"?

No, it's because at some crossings, at some times, it has an effect; and life's too short to construct a mental record of the reverse-engineered temporal programming of traffic lights.

Plus, it gives you something to do while you wait. More socially acceptable than picking your nose if anything.

If there's a button, I push it. I don't believe it will work all of the time. I believe it will work some of the time.

I also believe that the hardware is there so that the software can be changed.

I've seen plenty of people follow your strategy incorrectly, perhaps following an operational change that they haven't noticed. They are really just waiting for someone else to come and push the button.

Same here - and it's actually really quite bad for safety!

I walk to and from work every day, same route, same crossings, during daylight hours, the same sequence of traffic and crossing, to the degree where one watches the traffic lights and crosses as you see them change, rather than waiting for the pedestrian light.

Anyway, long story short, walked there one evening shortly after getting into this routine, and got halfway across the road before realising that the flow was evidently different at night, and required a button-push for the pedestrian interval.

If they just put up a sign that said "button needs pushing only between hours of X and Y" it'd probably help road safety and save frustration - although that goes contrary to the article. shrug

Then all, the armchair traffic engineers will e complaining that the signs aren't accurate.

It seems to depend on reactivity[1], hence it doesn't block.

Here's an example: https://jsfiddle.net/k61y5sLo/2/

    Pre-callback
    Evaluating getCallbackText()
    Post-callback
    callbackTest => cb("Alice")
    Evaluating getCallbackText()
    callbackTest => cb("Bob")
    Evaluating getCallbackText()
    Callback returned: Alice Bob

https://github.com/aldonline/reactivity

Thanks for the example. Just to clarify, it appears that you used syncify.revert(getCallbackText) to to create a callback from a synchronous function in order to log the final result.

So it looks like Syncify.js evaluates the parent getCallbackText() function each time that one of its child callbackTest("Alice") or callbackTest("Bob") functions fire. Conceptually it replaces instantiations of callbackTest() with their return values. So on the final pass there are no more child functions to evaluate and it can return the final value.

If this all sounds correct (and please correct me if I'm wrong) then the tradeoff is some re-run overhead in order to skip having to deal with promises or yield/generators.

I enjoyed the linked post[1] at the bottom more than the post itself.

    #define while(x) if(x)

Too bad April 1 is already gone, I'm kind of tempted to try and sneak that in.

[1] http://www.cs.earlham.edu/~skylar/humor/Unix/c.language.pran...

Maybe I use Word wrong, but I always hated how it was really easy to screw up formatting and spacing by accident. As well, changing some line spacing consistently across the document was a pain (e.g. list spacing).

With LaTeX I find I make a lot less accidental mistakes. If something goes wrong, I know exactly what caused it (just look at my changes!)

The ability to effectively diff and version control a LaTeX (or Markdown, or reStructuredText, etc) file is, more than anything else, what keeps me away from word processors for anything but throwaway document sketches. Word and similar can do "track changes", but it is extremely clunky by comparison.

You are using Word wrong (like everybody else). Don't manually change the font sizes on anything, instead give the text the correct style (title, subtitle, paragraph, etc) and then change that style to be what you want.

One of the benefits of LaTeX is that it practically forces you to do this.

When the easy way to do something is the wrong way to do it, it indicates a problem in the tool itself.

Edit: Somehow I missed your second sentence, and so we are in agreement. LaTeX is better in this regard, because it forces good practice on the user.

What if the user wants to write a letter?

I fail to see the problem:

\documentclass[a4paper,12pt]{letter} \signature{the user} \address{the user\\right here} \begin{letter}{Receiver\\somewhere over the rainbow} Bla bla bla ...

Not very different and you can ensure your letters look great.

Word: If you work with styles, changes should be rather easy and fast.

Ports inside a container aren't exposed publicly unless explicitly done.

Container linking lets containers connect to each other without exposing ports via docker.

The "linking" is exposed to the container via an /etc/hosts entry, so e.g. --link test-db:db adds "a.b.c.d db" to /etc/hosts which can be used by anything that needs it.

As mentioned by jon-wood, the simple linking approach only works with containers on the same machine unfortunately.

The concepts of Linking are being revisited via the network revamp proposal with the help of Network Plugins : https://github.com/docker/docker/issues/9983. With the plugins providing State Distribution and Service Discovery the containers can be effectively linked across multi-host network.

Why is this an article anyway? They use matching algorithms here in Canada for university co-op (intern) jobs as well. Lots of people don't get jobs, but that's no fault of the matching algorithm.