Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Antisocial Coding: My Year at GitHub

where.coraline.codes

Post reply on HN
Print view View original

191–200 of 572 posts

Re: Antisocial Coding: My Year at GitHub

#191

Post by cholantesh »

specialist wrote: post #182
cholantesh wrote: post #168

Earlier quoted context omitted.

Was there really something wrong with that phrasing? How would you improve it? Genuinely curious; I think I'd have phrased it the same way (though at every place I've worked we'd have done a review like this face to face), and I find it pretty non-confrontational. Enjoyed the letter you shared. Gene is truly missed.

"Hi cholantesh! Just saw your new survey. Looks great! Hey, have you seen the HRC's recommendations for surveying gender? I've used their guidelines in the past. Very helpful! http://www.hrc.org/resources/collecting-transgender-inclusiv... TLDR: They suggest splitting the gender question into two parts. As you know, this is a super important issue for me. Progress! Exciting! If you're busy, I'll make a pull request.…

Whoa, those exclamation marks woke me right up from my midafternoon coma haha. Thanks for this, I'll keep this kind of tone in mind.

Re: Antisocial Coding: My Year at GitHub

#192

Post by eropple »

pc86 wrote: post #129
ohthehugemanate wrote: post #75

Earlier quoted context omitted.

Meritocracy is actually a very damaging thing, the way we tend to implement it in tech. Without concrete and public metrics of "merit", it becomes a buzzword for reinforcing the biases (conscious or not) of the evaluators. Particularly in tech, those biases tend to favor white, upper middle class, males. This kind of bias is demonstrated over and over in studies, even (and in some cases, especially) among people who…

So fix the implementation. Rather than say "well there's not a lot of brown people contributing to this JS package so we need to get some more" why not fix the way the mentioned biases (which I agree 100% exist) are affecting how merit is decided? The idea that the best way to fix biases in favor of wealthy white men is to add new biases against wealthy white men is crazy.

The people who generally push for that "meritocracy" are the ones who benefit from it not being fixed in the first place.

Re: Antisocial Coding: My Year at GitHub

#193

Post by kybernetikos »

openmosix wrote: post #157
RangerScience wrote: post #147

Earlier quoted context omitted.

> she dismisses PRs or code reviews because made by white males. I didn't see this and can't find it. Can you point me at where this happens?

[...] For my first few pull requests, I was getting feedback from literally dozens of engineers (all of whom were male) on other teams, nitpicking the code I had written. [...] Shortly after this happened to me, the code review feature was prioritized. This functionality was rolled out internally pretty quickly. From that point on I didn't get dogpiled anymore [...] Re: who cares if they were all male? If I open a PR…

I took it to indicate that she believes that she is getting so much 'feedback' specifically because of what she is and what she's trying to do.

The fact that all of the people offering feedback were male is weak evidence in favour of this (consider that much of her immediate team is female). She offers as another piece of evidence that she compared notes with a colleague with a similar background who was male and who wasn't getting the same level of attention.

> Re: who cares if they were all male? If I open a PR and I get dozens of feedback, either they are legit feedback or not.

If you're working somewhere where you and people like you are getting 20 people peering over your shoulder uninvited and criticizing your every move and people of a different group only have 2 people reviewing their code then you can legitimately claim to be working in a hostile environment.

Code reviews are always a mixture of objective and subjective feedback, and having to consider detailed comments (objective, subjective, substantive, trivial) from a large number of people not directly involved and without appropriate context would be a stress on anyone (not to mention is a simple drain on productivity).

On a purely technical note, she says nowhere that the PRs should be dismissed because they were from men. I think that was something you read into it. At issue was the unusual quantity of the feedback.

> if a colleague would say "I am getting feedback from literally dozens of engineers (all of who were black women)", wouldn't that raise an eyebrow or more?

If it were fact, then I would assume that there was some way in which this colleague had upset a group of black women. I think a similar conclusion is being offered here (although given the likely employment ratios the black women theory would have a whole lot more evidence).

Re: Antisocial Coding: My Year at GitHub

#194

Post by majewsky »

twerkmonsta wrote: post #72

"The same day that I had this review, I got some devastating personal news. I have bipolar depression and was already in a bad place mentally, so I found myself feeling crushed and hopeless. In an attempt to deal with things I ended up taking a dangerously high dose of my anti-anxiety medication." I'm all for being compassionate and helping people through tough times, but does this sound like a stable person?

She does literally say "I have bipolar depression". Assuming that the company knew this when hiring her, it would be very surprising that they would not be compassionate at this point. I don't know if that assumption holds, though.

Re: Antisocial Coding: My Year at GitHub

#195

Post by tptacek »

DannyBee wrote: post #148
dragonwriter wrote: post #83

Earlier quoted context omitted.

> This practice seems pretty common in California, but given California is an at-will employment state, I'm not even sure why. Because the PIP exists to to support the argument that the firing was not for a prohibited reason in the event that the employee charges that it was. “At-will” doesn't mean there aren't prohibited reasons for firing, and if you don't have any evidence for what the firing reason was, it doesn'…

This is super-cynical. If you want to get rid of them, you don't have to have success criteria at all. You could just document the behavior that makes them want to fire them ("you smell funny") and followups in an email, and call it a day after a week or so. Having success criteria when you just want to get rid of someone means if they do succeed, but you still wanted to just get rid of them, you are worse off as an…

I don't think you're correct here. As I've understood the concept for something like a decade, PIPs are almost universally understood to be a soft form of firing. The correct response to receiving a PIP, probably in most SV companies, is to start looking for another job.

It's easy to turn your own logic back around on you. Since it's a more-or-less open secret that most PIPs are part of managed termination, every competent company that issues one knows they're running a huge risk of sabotaging their relationship with an employee by issuing one. There are lots of ways to manage improved performance from an employee without invoking the dreaded PIP. If the company merely wants to improve performance, they can issue MBOs or schedule a special series of 1:1s.

It's also easy to see why companies would issue PIPs despite potentially backing themselves into a corner when the targeted employee exceeds the stated expectations of the PIP. PIPs are how HR wants employees to be fired; they simple are the whole firing process. But that doesn't mean the people who actually write the PIPs know how to write them effectively. In Coraline's story, you have what reads to me like a pretty standard Kafkaesque PIP story: Github wanted her out, HR demanded they follow the standard process, they PIP'd her, the PIP didn't anticipate that Coraline would keep diligent records, and they were forced to go through contortions to pretend that it was the PIP that had been failed, rather than the "will" part of Github's "at-will employment".

Re: Antisocial Coding: My Year at GitHub

#196

Post by throwaway8800 »

roywiggins wrote: post #104
thsowers wrote: post #38

Earlier quoted context omitted.

You are correct, thank you for pointing that out. However, she is still assuming that the question was phrased in a "Transphobic" way. Perhaps she reads it like that, but I doubt that was the authors (or questions) intention. Perhaps i'm wrong

You can be transphobic by omission. For example, not taking any time to research the issue ("Is 'transgender' a gender?") before writing a study question, and then getting angry when someone tries patiently to educate you. It's not what's in people's hearts that counts. Someone can truly believe that blackface is not racist, but that doesn't mean that it's not racist when they do it.

> You can be transphobic by omission.

I don't agree. Any definition from a reputable source on the term "transphobic" implies or directly states that transphobia is, specifically, a fear or dislike towards trans people. Not researching what transgender means or what a survey should include for genders in the year 2017 may be stupid or ignorant but I just don't see how it meets the bar for being "transphobic", especially when the person who is making the error of omission has no problem with trans people!

As an LGBT person I really think that people need to stop inferring malice where there is none. Not only does it not accomplish anything, it simply aggravates people who would otherwise be friends and allies and creates further divisions. If I pointed out every single time a friend or family member accidentally misgendered someone or said something that I thought was not 100% PC, I would be spending a lot of time alone.

Re: Antisocial Coding: My Year at GitHub

#197

Post by rndmwlk »

ivraatiems wrote: post #190
rndmwlk wrote: post #152

Earlier quoted context omitted.

>I can see how her comments could be perceived as rude, or "unempathetic." I cannot see how - especially at Github, a place not known for cultural sensitivity - they were firable. I agree, I think this is where the "fishy"-ness I spoke of earlier comes in. Giving the benefit of the doubt to Coraline, there is definitely more at play than being rude in a few emails. I just feel uncomfortable jumping from an individual…

I respect that. Personally, I feel that Github has a long enough history of mismanagement and poor culture that sexism and the kinds of issues this post describes aren't difficult to imagine at all.

That's probably my disconnect right there, as I have not been following Github's cultural issues much at all. Unfortunate to hear.

Re: Antisocial Coding: My Year at GitHub

#198

Post by majewsky »

eropple wrote: post #86
throw201707 wrote: post #70

This is not a comment on the experiences described in the article, but it makes me extremely sad that the word "meritocratic/meritocracy" is used pejoratively, almost as if it were damning in itself. To me, meritocracy is something to be cherished, being far preferable to the systems that came before it, and being inherently anti -discriminatory. (Yes, people's hardships which prevent them achieving their full potent…

"Meritocracy" is regarded with derision in no small part because what is actually meant by it, rather than what you describe, is "this person is similar to me and I like them, so I think they do a better job." It's the ingroup club made manifest. What you describe would be great, but it's not what "meritocracy" actually means in practice.

Then we need to call out fake meritocracy as such. I'm actually looking at this situation with some sort of confusion, because I've never seen proclaimed meritocracy to actually mean "boy's club".

But then again, I'm a white caucasian heterosexual male, so maybe I just don't notice. (Previous sentence is meant literally, not ironically.)

Re: Antisocial Coding: My Year at GitHub

#199

Post by majewsky »

specialist wrote: post #179
dragonwriter wrote: post #163

Earlier quoted context omitted.

> Whatever the perceived tone, what Caroline didn't do was suggest improvements, provide examples. It was a teachable moment. Seems to me she directly suggested the improvement, after explaining both the reason the existing version was bad and the reason the improvement was better (the two reasons being the same): Quoting directly from the issue she raised, which was not merely quoted but emphasized as a pull-quote i…

Okay. Continuing to quibble... She said what was wrong, but did not say what was right. Explicitly. To me, it's an example of "Guess what I'm thinking!" Or more famously, that UI designer who said "Don't make me think!" Granted, a "data scientist" should already know how to survey, be open to constructive criticism. I personally would have no idea how to ask the question. Though I'm not a data scientist, I can use go…

> She [...] did not say what was right

> If you want to know if a survey respondent is transgender, you need to explicitly ask that question.

How is this not saying what is right?

Re: Antisocial Coding: My Year at GitHub

#200

Post by DannyBee »

tptacek wrote: post #195
DannyBee wrote: post #148

Earlier quoted context omitted.

This is super-cynical. If you want to get rid of them, you don't have to have success criteria at all. You could just document the behavior that makes them want to fire them ("you smell funny") and followups in an email, and call it a day after a week or so. Having success criteria when you just want to get rid of someone means if they do succeed, but you still wanted to just get rid of them, you are worse off as an…

I don't think you're correct here. As I've understood the concept for something like a decade, PIPs are almost universally understood to be a soft form of firing. The correct response to receiving a PIP, probably in most SV companies, is to start looking for another job. It's easy to turn your own logic back around on you. Since it's a more-or-less open secret that most PIPs are part of managed termination, every com…

"As I've understood the concept for something like a decade, PIPs are almost universally understood to be a soft form of firing."

FWIW, they've never been at any company i've managed at, or org i've belonged to :)

I certainly believe such companies exist, i'm not stupid. I just am not sure I believe they are as prevalent as you do.

The only question in the room (from either HR or the manager) has always been "how do we help this person get better".

In fact, there were cases a PIP was decided against because it wasn't going to be effective in helping.

We simply offered fork in the road instead.

As for "sabotaging relationships", in every successful PIP i've seen, the person is still working at the company years later. So ....

Apparently i'm just very lucky ;)

(which is, of course, within the realm of possibility)

Post reply on HN