Live data from Hacker News

GitHub streak: End-game and post-mortem

manishearth.github.io

11–20 of 32 posts

Re: GitHub streak: End-game and post-mortem

#11
post #7

Using a lack of GitHub contributions as a way to disqualify candidates sounds like a really bad idea. Does anyone do that? Seems like a strawman argument.

I honestly don't know; I'm still a student and have never had the chance to experience this first-hand. Various people have mentioned similar things though.

https://twitter.com/sarahmei/status/597276242887348224

https://twitter.com/nick_r_cameron/status/597328012787384320

Re: GitHub streak: End-game and post-mortem

#12

I've written in the past about how I love streaks as a motivating tool.[0] It gives a motivation to get up and do something , when you'd rather do nothing. One thing I try to do on goals that require an action, is not to add stipulations to it like the OP eventually did in the beginning. If you make the hurdle too large you're more likely to break the streak and more likely to fall into inaction, so allowing a small…

I myself find steaks to be a motivating tool. Or at least I'm motivated to eat them.

Re: GitHub streak: End-game and post-mortem

#13

I've written in the past about how I love streaks as a motivating tool.[0] It gives a motivation to get up and do something , when you'd rather do nothing. One thing I try to do on goals that require an action, is not to add stipulations to it like the OP eventually did in the beginning. If you make the hurdle too large you're more likely to break the streak and more likely to fall into inaction, so allowing a small…

> is not to add stipulations to it like the OP eventually did in the beginning.

Agreed; in retrospect setting rules was a bad idea (I wanted to keep it "pure" but that was a very arbitrary standard and hard to keep up). Later in the process I was able to maintain the streak with small things when I had lots of other commitments -- but that didn't make me downspiral into only small contributions -- I bounced back -- which was the important part.

> you're more likely to break the streak and more likely to fall into inaction

This was what I was mostly afraid of once I was well-past the 100 mark. By then I knew that I had gained a lot from it. But I was still afraid that breaking the chain would let me break it again, and again, until it's back to being random bursts (but more regular bursts).

Now, I'm no longer afraid of that, because I know that the streak isn't motivating me at all anymore. I'm being motivated by -- whatever it is, I don't really know :P

> there's no one-size-fits-all when it comes to motivation.

Big +1

Re: GitHub streak: End-game and post-mortem

#14

I wonder how you can go on that long. You can easily see in mine ( https://github.com/daurnimator ) when I went on vacation for 3 weeks in November/December last year.

I'm not sure if you know, but GitHub's stats depend on the viewing user. Most who look at my GitHub profile will see a streak of only 13 days[1], but my coworkers see 171, and I see 871[2]. If the viewer can't see a repo, it won't show up in the stats.

1. https://github.com/ggreer

2. http://geoff.greer.fm/photos/screenshots/Screen%20Shot%20201...

Re: GitHub streak: End-game and post-mortem

#15
post #7

Using a lack of GitHub contributions as a way to disqualify candidates sounds like a really bad idea. Does anyone do that? Seems like a strawman argument.

I honestly don't know; I'm still a student and have never had the chance to experience this first-hand. Various people have mentioned similar things though. https://twitter.com/sarahmei/status/597276242887348224 https://twitter.com/nick_r_cameron/status/597328012787384320

I doubt it, that's a very silly measuring stick.

It's nice when we have a candidate with a GH history, but I'd venture a guess that 90% of the best code written by the engineers we hired was written for some other company in private repos. I can't imagine that being too far outside the norm.

Re: GitHub streak: End-game and post-mortem

#16
FYI, the GitHub streak can be manipulated.[0] My apologies for the lack of documentation. Basically, since you can modify the author and commit date in Git, you can trick GitHub into thinking you've committed something several years in the past[1] or the future[2]. Some people have written things that allow you even write messages into the commit graph[3].

[0] https://github.com/zg/streaker

[1] https://github.com/zg/streaker/commits/master?page=710

[2] https://github.com/zg/streaker/commits/master

[3] https://github.com/jbranchaud/commitart

Re: GitHub streak: End-game and post-mortem

#17

I've written in the past about how I love streaks as a motivating tool.[0] It gives a motivation to get up and do something , when you'd rather do nothing. One thing I try to do on goals that require an action, is not to add stipulations to it like the OP eventually did in the beginning. If you make the hurdle too large you're more likely to break the streak and more likely to fall into inaction, so allowing a small…

I myself find steaks to be a motivating tool. Or at least I'm motivated to eat them.

Steaks for streaks?

Re: GitHub streak: End-game and post-mortem

#18
post #14

I wonder how you can go on that long. You can easily see in mine ( https://github.com/daurnimator ) when I went on vacation for 3 weeks in November/December last year.

I'm not sure if you know, but GitHub's stats depend on the viewing user. Most who look at my GitHub profile will see a streak of only 13 days[1], but my coworkers see 171, and I see 871[2]. If the viewer can't see a repo, it won't show up in the stats. 1. https://github.com/ggreer 2. http://geoff.greer.fm/photos/screenshots/Screen%20Shot%20201...

I always figured this, but never cared to confirm.

Otherwise, you should take a holiday!

Re: GitHub streak: End-game and post-mortem

#19
post #7

Using a lack of GitHub contributions as a way to disqualify candidates sounds like a really bad idea. Does anyone do that? Seems like a strawman argument.

A lot of companies say "Can you link to your GitHub" to applicants -- including applicants fresh out of college -- and the applicants don't really know if "No" or "Yes, but there's nothing there" is an appropriate / acceptable response.

I've had this happen to me, when interviewing (in 2011) with an bachelor's and an expected master's from MIT, and a bunch of other great experience on my resume. I'd had a phone screen, and got an email from one of the engineers asking to see my GitHub. There wasn't very much there at the time, and I was confident I was out of the running without a GitHub presence, so I didn't even reply. A friend at the company said that they liked me and were just curious (and I ended up replying with "well, here's a bunch of patches I sent to the Debian BTS", and ended up getting the offer).

I see MIT classmates with years of real work experience behind them worrying about their GitHub presence and how it affects their future jobs. I see questions on Reddit from students who've been asked this question. I see job postings that ask for a link to your GitHub. I don't know if anyone explicitly disqualifies candidates based on a lack of contributions, but there is absolutely an impression that it's a strong negative signal.

Re: GitHub streak: End-game and post-mortem

#20
post #14

I wonder how you can go on that long. You can easily see in mine ( https://github.com/daurnimator ) when I went on vacation for 3 weeks in November/December last year.

I'm not sure if you know, but GitHub's stats depend on the viewing user. Most who look at my GitHub profile will see a streak of only 13 days[1], but my coworkers see 171, and I see 871[2]. If the viewer can't see a repo, it won't show up in the stats. 1. https://github.com/ggreer 2. http://geoff.greer.fm/photos/screenshots/Screen%20Shot%20201...

Also if you contribute to a clone, it won't show in the stats.

It is a metric for self-motivation, but I guess it can be wrongly used for other things. Is like the "GH is your CV" thing, ignoring that there are other code hosting environments and non public code.

Post reply on HN