Hacker News new | past | comments | ask | show | jobs | submit login
Stop looking at HN and go work that thing you promised yourself you'd work on
281 points by sharemywin on Dec 19, 2015 | hide | past | web | favorite | 83 comments
Please ignore if your not me..unless you'll find it useful too.

For me, reading HN is like reading the paper or reading books about subjects that matter to me: it's not time strictly spent on the job I have to do, but it's part of a healthy information diet. Crunching code without taking a look at what's currently happening and being discussed in your professional community is hardly a recipe for self improvement. (It's a thin red line between healthy consumption and abuse, I know, but our job often _requires_ that we keep an eye on what's going on outside our compiler.)

To me, HN has been the MBA which led me to (successfully) start my product. But it has also become way too addictive (probably an effect of loneliness). It's kind of ironic, given HN had no employee for growth hacking and no "we miss you" notifications.

I agree. Reading HN is like taking a break to eat a chocolate chip granola bar. On the other hand, reading reddit is like hooking a syrup IV to your veins.

Rightly said. I use HN to keep me update-to-date. I do spend and read HN articles comments more than article itself, to get a view on how other people thinks. Its proven to be very valuable to me.

Hacker News is the place where I come to learn about new technologies and to see where the industry itself is moving. There is something here for everyone, from CEOs to C++ programmers. Over the years, it has been a very valuable source of news.

If you ignore the first 9 links about any new JS framework, it shouldn't take more than than 10min/day just to keep up with what is happening.

Or is this just not true in your experience?

How long do you spend on HN on average per day?

I can spend a whole Sunday afternoon here if I really dig into 10+ articles on the front page, read the top comments, and contribute to the discussion.

On an average workday though, I look at 3-5 posts and skim the articles. I often find more value in HN's discussion than the article itself.

I spend way too much time on it.

Hacker News is, among its other qualities, one of the few places where the comments on an article are reliably interesting and valuable.

Still, I empathize with the OP... It's often too tempting.

I'm just here because my compile times are slow. 3 hours later… after 12 tweets, reading 6 articles, filing a GitHub issue, and watching a show on Netflix… what was I building?

> 3 hours later… after 12 tweets, reading 6 articles, filing a GitHub issue, and watching a show on Netflix… what was I building?

Judging from the compile times, Gentoo?

Probably just "hello world" for an fpga :>

At our company back in the time we chose to prioritize Delphi as the development language of choice because of compile times (the alternative competing was C++).

Too bad the Delphi team was not able to find a business model that was able to scale their user base (probably some kind of subscription at the price range of JetBrains subscription was needed). Now are now looking back into C++ (and its much longer compile times).

Anyone have any tips for a productive way to deal with long compile times?

1. Write documentation while things are compiling.

2. Finish other work.

3. If you iterate a lot, maybe just reading the code figuring out your next steps.

4. Maybe figure out why it's taking so long, and figure out a way to speed things up.

5. Write more tests.

ccache[1] works for me.


Like all things, it's not a panacea: https://blog.flameeyes.eu/2010/07/debunking-ccache-myths-red...

Of course, if you're rebuilding the same code over and over it can be a net win, but -in the comments- it is mentioned that the obvious things -that one might end up doing somewhat frequently in C/C++ development- invalidate the cache.

Go play a game of ping pong. Saving yourself from RSI and carpel tunnel is considered productive in my book.

Get up, walk around, stretch, think. Talk to other people, unless it's interrupting them.

Large Scale C++ Software Development

John Lakos

The book explains how to design large systems to contain compile times.


If your compiling time is slow, you should work at improving it while it's compiling.

Hmmm....you may have been compiling Scala.

I know, Scala, right?

But if I stop looking at HN, I won't get wonderful tips like: "Stop looking at HN..."

I've drastically cut my HN time in the past few months by following this one weird trick:

Add "news.ycombinator.com##td.subtext" following to your Adblock/uBlock custom filters.

You still get to read the articles and keep up with tech news, but you don't end up in comment threads (unless it's an Ask HN/etc, hence why I'm commenting now). Occasionally I load up HN in a different browser if I want to take a break and read some comments but my idle browsing has been drastically reduced.

There's lots of great content in the comments, but it's also a fantastic way to waste time.

> There's lots of great content in the comments, but it's also a fantastic way to waste time.

But I learn more from the comments than from the articles!

I agree with this. I find it not only interesting, but also informative to hear other people's take on articles, both for their opinion and added knowledge of something that may not be in the article. Sometimes the discussions veer a little bit which adds more background/relevant/related information. At least for me, the more I know about a subject, the more angles I'm aware of, and the more context I have, allows me to not only absorb it better, but it also allows me to develop or realize the theory behind the matter. From there, I can better apply what I'm reading/learning to other situations.

I make sure to only browse HN, youtube, social media etc. on my phone. I don't have my various accounts logged in on the desktop. This way I can take a 5 minute HN/whatever break and actually go back to work again after 5 mins. Even my subconcious knows being on my phone is not working.

I do similar, but mainly by looking at comment count. Typically higher posts have large comments, and I just don't have the energy to read all those comments.

Go to your profile. Set your "noprocrast" to yes. Set you maxvisit to 15. Set your minaway to 60. I'll be back in an aware to see if anyone replies. ;)

Get a load of this guy, typo out in the open and helpless to save his proof reading reputation.

ouch! it's true.

12 years ago, I have decided that I would forever change the way people communicate.

Today, nothing has been done. I'm struggling to get people to relate to this need for a better communication system and language. I am convinced that this is very important, but I fail to get the positive feedback I've always expected.

Should I ignore people and blindly trust my guts? Taking on such a massive task is very difficult for a single person.

> fail to get the positive feedback I've always expected

I don't know what your idea is, but that statement makes me think "confirmation bias". Start looking for something people really want instead, and build from that. Otherwise you're effectively just building a product for yourself, and trying to convince yourself that others will want it.

I went through your comments history and I'm not sure I quite grasp what your idea is or what problem it solves. If I understand correctly, you have this idea for a new type of human-computer interaction based on a visual programming/interaction language? I suggest you start by writing down your thoughts in a blog/wiki for others to better understand it. I'd also start working on an MVP now because in another 12 years, advances in NLP/AI might render your language obsolete :)

I've been practicing something like this myself.

When I have an idea that seems like something people want, I force myself to describe it in a 250-500 word blog post. Most of them I never ship, but it's really about the exercise of conveying your ideas in a way other people can digest quickly and easily.

I have described my vision on IRC, Reddit and HN multiple times.

People hate it because they can't understand it.

I don't think NLP/AI is a threat. Nobody is working on the UI problem.

How are we going to communicate with said AI? Surely not text or speech. Nobody is able to answer this yet. That's why I'm working on it.

You should ignore people enough to try and prove it to yourself, build a proof of concept and the compare it to what you see out there.

If you've been on hacker news one thing is certain most folk will more readily shoot down new or refactored ideas, especially if that's all they are, ideas.

Can't you prototype something small and gather feedback?

All you have to do first is start

i'll give you a recent anecdote

i was working on a project where i needed to convert linear motion into rotational motion

i was working through some schemes and though functional by design they were failing in the real world scenarios i was building for

frustrated i sat down for a break and doped up on some hn for a bit

this post came up:


i clicked through and was surprised by how much i enjoyed the lists found within,

so i clicked on the portrait in the corner to learn more about the person putting the lists together

this led me to this page:


an almost goofy early web days personal page of this nobel prize winning physicist

while scrolling though the personal page, clicking on random things and skimming through them, i came across an animation in the 'unpublished works' section

i looked at it for a second and thought.. is that a rotational motion being converted into a linear motion?

i searched for 'Peaucellier–Lipkin linkage' and found this video: https://www.youtube.com/watch?v=D30iFxhJTQA

and there it was, linear to rotational motion and much more robust than the other efforts i was attempting

i built it and it worked

i've since designed out the need for the conversion but it was an important step along the way in the build

i'd err away from the charge to 'get to work' and suggest one instead: 'enjoy yourself'

It occurs to me that the folk most likely to see this advice are those least likely to follow it.

you caught me

New Year Resolution: turn noprocrast on, with maxvisit 5 and minaway 240.

And for those of you just tuning in (like me) -- noprocrast is a setting in your hackernews profile (upper right corner). Turn it on and it will only allow you on the site for a maximum amount of time (maxvisit) before preventing you from using it for a minimum amount of time (minaway). It has to be turned on manually. The maxvisit 5 and minaway 240 would let you visit for 5 minutes every 4 hours. I just turned mine on and set it to maxvisit 10 minaway 240.

I've used Leechblock (firefox extension) in the past for the same purpose. One feature it has which I found particularly useful is the option to delay pages instead of completely blocking them.

Instead of the transition being an abrupt stop that makes me want to turn the blocking off to finish what I was doing, the delay makes it a gradual frustration, and feels a lot more like getting naturally bored of browsing.

Nice! Thanks for the tip!

Don't wait til the new year. Do it noaw!!

I've known this exists, and that's how I know I have a problem lol. I can't even bring myself to enable the noprocrast

done :)

Or just tweak things for high signal:noise ratio. Scan the front page for only things relevant to your work/research/etc, ignore the clickbait. Read only a selection of your favorite commenters. Etc. This works on other social media like Twitter and Reddit too. You can minimize time spent while extracting max value.

I will keep looking at HN because I learn a lot about the technology and the trend from it. And I also found the comments on HN are high-quality.

The only thing I don't like on HN is the UI. So I created http://hackernewsroom.com based on HN API. Now I read it everyday. But HN API is read only. When I want to comment, I have to click a link to go back to HN to make comments. Another con of hackernewsroom.com is some websites have the same-origin policy which are unfriendly to iframe.

Having the article text and comments on one page is a great feature.

I'd also love to see the community move towards in-text annotations directly on the article page, but it's hard to ease into that.

Exactly. It's inconvenient to read the article and comments in separate pages.

As a professional procrastinator I'll definitely follow your advice...

... tomorrow.


/better at it than you.


Ha! This kinda talk belongs in Reddit.

I was foolish to expect this to link me to a Medium piece or editorial, really..

sorry. I like to keep my advice to 20 words or less.

Do you mean... "Looking at HN" or do you mean "Reading and commenting on HN and getting into arguments with people and checking your karma and replies and upvoting / downvoting / flagging posts and making observations about the correlation between username and karma and generally not even bothering to read TFA or the front page and just trolling the comments section for interesting things to respond to like this"

It probably doesn't make a difference, so cheers.

I wanted to get some code down so I wrote this out to make a statement to the world for me to focus on code.

^ downvoted for referring to being downvoted.

Um, let me rephrase that.

Or use HN to "sharpen the axe": focus on reading stuff on HN that is actually useful to your goals.

Voila! Guilt-free slacking.

That's how I convince myself that the many man-months I have sunk into HN are not in fact wasted but on balance a net positive. Humans are really good at rationalizing away counter-productive behaviour. :)

it's like gambling. click on 10 articles get 5 not worth reading, a couple that are interesting and might be useful later, and once in a while you get something you can use now.

Your thing is too hard man, you need to break it up into bite sized chunks :)

that's really good advice. I keep avoiding this set of classes for notifications maybe I should break it into a smaller goal.

I was, but it's compiling...and now it's done. Back to work.


I'm struggling to push the magic "launch" button.

It's a curse.

If that's all that's left to do - great. You only need 5 seconds of recklessness. If it only really was this easy..

Sounds like something a cocktail can solve.

just do it. you won't know until you do.

Guffaw! Wonderful advice. Now what was I at again?

Fine ill go back to shader writting...

Perfect! Thank you for the reminder.

To me HN is the weapon supplier

You're not my mother.

How'd you know!


You're not my supervisor!

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