mjolk

So how honest are you being here?

Dead honest. I mean a quick async-okay-if-from-phone "hey, here's what I have lined up for this week, can you tell me if this is unreasonable from your perspective" exchange.

Employers and employees can both get caught up in saying things they think the other side wants to hear.. I think it's appropriate to be very explicit. Call out the issue. Tell them you aren't signalling. Tell them you really mean what you say, and that if they're going to get burned out from that, then your shop isn't the place for them.

You nailed it. The problem is that even when being explicit in expectations, candidates will still try to get that offer, even if it's not the right fit at that time. It's really hard to tell if someone means it when they say "oh, that's fine." I think startups have a certain glamour that masks the reality and people see interviews as tests, not conversations.

Further, I'd love to keep that person in the pipeline -- employee happiness and feeling valued is huge for me, so if that person needs to come in at 10 because they need to drop a kid off at school, I can respect that, and hope he/she is available when we can accommodate it.

I think based on your responses you have the right attitude.

I really appreciate you saying this.

It risks creating the typical monoculture of guys (yes unfortunately) between 25 and 35.

Oh yeah, totally. My motivation for asking was to make sure we don't end up in a monoculture of people that stick around after being surprised by the startup grind. That won't work for our business as we _need_ a mix of people, backgrounds, interests to really make it work (opposed to say, a hft/fintech platform, where diversity of thought/life experience isn't crucial).

A ton of people are in this situation, much too many to ignore even for a startup. They might not seem like a good fit for your phase of startup but I think that mode of thought is counterproductive.

Yeah, it's definitely hard. Funding is limited and early stage is about maximizing the value of capital and speed of validating assumptions, and unfortunately, that often means preclusive criteria for people that need to leave "on time" regularly. Not saying it's right, but that's the reality of most startups.

I want to actively combat the bias of just short circuiting to people that are 22-30 and probably without kids, which means being able to have the conversation of "hey, please don't say this expectation is fine if it's not" and knowing how to tell if someone says it's fine for the sake of getting an offer, but it's actually not.

Needing people to occasionally work more is normal. Even having an emergency meeting on a Sunday is normal. Just be clear with what the situation is, what is expected, and make sure that the plan is to never have 50h weeks or weekend meetings. The problem is having the "constant crunch time" culture.

The first 5 you hire are basically hopping into constant crunch with you, which is why their equity should be the carrot to make the stick worthwhile.

I think I rubbed people the wrong way or reminded them of a negative employer, but a weekend meeting for us is a few lines on Slack that essentially serves the purpose of "hey, so I decompressed and reflected, and here's where I'm thinking for this week. is this reasonable?"

I don't call meetings without a purpose and I _definitely_ don't want this to be the norm after we grow.

Not sure I understood the problem here, was it that you would have preferred that to surface in the interview, but it didn't, which caused friction down the line when someone turned out to not want to work more than 40h weeks?

This is exactly it. If a candidate expressed this, I would say "okay, thank you for your time, I hope you're available when we know we can respect this."

I don't know why you're being sarcastic, but I actually wanted to make sure I don't bring someone on to be unhappy.

People over 25 can work more than 40 hours and not get burned out. I feel that you just want to make me out to be a bad guy, so I'm not sure it's worth writing a longer response.

Seriously just hire 5 people instead of 4.

That's actually a lot easier said then done at an early stage startup for reasons of talent and compensation.

The problem is you can't afford to make your company an evening pizza 27 year old bromance company because you likely can't cut out that much of the talent pool without it costing you.

Wow, wait? We're not a "bromance company" and I'm not even sure how to respond to this and I think it's vaguely insulting.

Also: I don't mind people working a lot if compensated well.

We're very clear with compensation and living wage and actual equity is something we make sure is on the table.

I could certainly have worked a ton of hours for a period of my life (before kids etc) but I would have been pretty annoyed had I accepted an offer at a company and later found out that the offer was for an expected 50h and not 40h. Don't have people come and waste time at your interview without knowing what the situation is.

This is really important to me and I make sure the candidate actually knows what he/she is getting into.

This is exactly why I wrote:

"I've never heard a candidate employee express anything close to "I'm in at 9AM and expect to be out by 5:30", even if shortly down the road, it becomes clear that they feel their work/life balance is being infringed upon."

The problem isn't that I'm trying to hustle people, it's that people tend to agree to situations they don't actually want when interviewing.

I'm considering what you say, but it doesn't ring true to my interpersonal experiences at all.

Think about it. You are asking human beings to make personal sacrifices for no personal gain, but for the sole purpose of making your dream a reality.

My team has (real) equity and I like to believe they see the worth of the product. What attracted me to startups when I entered the game was the sense of ownership and agency in projects. That and work can/should come with a sense of reward.

so you could ask if there is any personal trauma they are trying to avoid.

I think this would be somewhat inappropriate to ask "hey, so do you put in extra hours because you don't feel whole?". Everyone I work with knows they can come to me, even if he/she needs to take his/her house keys and to go to a second location to chat.

You could also consider hiring hourly.

There's room for hourly consultants, but that hardly makes a team.

I wouldn't doubt overwork as a factor, but the elephant in the room is meaninglessness... people pretty much realize that the core feature of their jobs is their own economic exploitation...Burnout, like all pain, may be a feature.

Oh god, if you're feeling theatrically unhappy with your job, it's seriously time to make a move.

The core feature of my team is making something we think is awesome and missing from the world, where my role as the head of the startup is to do what needs to get done and make sure my team feels like their work and lives are fulfilling. If any member on my team felt like their work is meaningless or a source of emotional pain, I would not be able to sleep at night.

edit: Wow, I've gotten some negative responses and I'm not sure why. This isn't a "how can I exploit people" question, it's about making sure my interview process doesn't accidentally result in people that will be unhappy. It takes a lot to compete with incumbents and agreeing to work for a startup is the algebra of pros/cons.

I have had startups expect more then the typical 40 hours a week, and yep, I burned out pretty fast.

I'm the head of an early stage startup and have a favor to ask: can you give me advice on how to detect someone with your 40-hours-or-I'll-be-unhappy mindset? I've never heard a candidate employee express anything close to "I'm in at 9AM and expect to be out by 5:30", even if shortly down the road, it becomes clear that they feel their work/life balance is being infringed upon.

I don't mean this in any negative way and I envy people that aren't boolean in or out, but people that would feel burned out on 5x10h days (+ sometimes a quick Sunday Slack session) aren't good fits _right now_ for my team. Reliable "I do my 40, put in good work, and I'll be here for years" people are _great_ for much later stage (post-IPO, for instance), but it would deeply bother me if I found out one of my workers was feeling burned out/demotivated.

I'm not sure that my response helps build a constructive discussion, but I'd suggest that you at least have the direction reversed (lobbyists propose actions, the "assholes" (as you put it) vote as they're told).

near is not the same thing as impossible. I told this story before, but I once read an article about a police officer who said it was impossible for another person to have logged into an account because it was password protected, when we know that's not even close to being true.

That's not even the same realm as this case:

The Forensic examination also disclosed that Doe had downloaded thousands of files known by their “hash” values to be child pornography[0]

Thousands of hash collisions would require prior knowledge of the values and a concerted effort to deceive. It would be more realistic to say that human perception is broken when looking at the media than it is to argue with the mathematical reality at play here.

The requirement should be for them to look at the actual content, not the hash.

Refusing the evidence known to exist and definitely covered by probable cause is why the defendant is still in custody.

[0] https://arstechnica.com/wp-content/uploads/2017/03/rawlsopin...

You're obviously way more legally savvy than I am. Just goes to prove that a _little_ knowledge is a dangerous thing.

Oh no, don't feel that way. The law is a man-made thing at the intersection of logic and opinion, which is why there's so many laws and tests -- if you haven't read the source document that's linked in the Arstechnica article, I would, as it has a lot of important detail.

I'm referring to the 4th, about needing a warrant to intercept communications...Doesn't the 4th -- on the face of it -- preclude any system of wholesale collection of electronic communications?

Law enforcement were specifically targeting traffic expected to have child pornography and the people trying to exchange it on freenet who join very-special-purposed groups. Peer-to-peer platforms depend on people being free to join, and having special-purpose groups really helps with the "probable cause" condition of the 4th.

On the back of that, the defendant gave them confirmation of his illegal acts, so this case is about recovering evidence known to exist.

hashes can be inaccurate, it isn't a foregone conclusion in reality, just in their opinion.

Not really, no. The chance of multiple hash collisions on a set of arbitrary images is a near impossibility.

I'm glad you didn't take offense to me making reference to the fallacy as I appreciate our conversation and wasn't sure how else to express that thought.

If you haven't done so, check out the source document for the article as Arstechnica didn't include some important details (and the headline "Man jailed indefinitely for refusing to decrypt hard drives loses appeal" talks past what is actually happening): https://arstechnica.com/wp-content/uploads/2017/03/rawlsopin...

...but what if an ISP reports that files with those hashes have been downloaded by a particular IP address? ... but can't find the files the ISP says should be there.

I think this case is particular due to the lack of breaks in the chain. In your hypothetical, law enforcement and the prosecution have _vastly less information_ than in this actual case.

Law enforcement knew the path from a remote source, to (presumably dhcp lease based) ISP records, to the laptop that accessed the content (known to be the defendant's), to checksums in logs matching a physical drive (also known to be the defendant's). Coupled with other evidence, the defendant frustrating the process by pretending to no longer know the decryption phrase, and partial admissions of guilt by the defendant, this is a vast distance than a hypothetical case of "someone from this IP address downloaded Game of Thrones Season 1 from bittorrent, so hand over anything that can store bytes" (to use a far less disgusting crime to help keep emotion away from the discussion).

Based on this precedent, I think another judge could find reasonable cause to compel in that scenario.

Luckily, the US justice system is built on nuance; this case wouldn't hold up as a generalizable excuse to compel decryption -- which is why they're invoking the foregone conclusion rule to secure the production of evidence based on the enormity of the other factors.

In this hypothetical case, though, what if LE found OTHER files of child pornography? Would they be admissable?

I honestly don't know. In this case, the defendant is refusing to provide (multiple pieces of) evidence that is known to exist by checksum and direct file path.

Alternatively, if they found other material (e.g, bomb-making), could it be used against him in a separate case?

Having information on how to construct a bomb is not illegal, any more than getting a degree in chemistry is illegal, but plotting to kill people with a bomb is legally actionable.

I'm not sure I trust the government in either one of these situations.

I agree with you, but on a different shade of the argument. I'm suspicious that the ecosystem of justice is built on securing convictions as opposed to seeking objective truths. In this case, I support the government/court based on the information I have.

As with so many other of the Constitutional protections of the Bill of Rights, they've slowly been chipped away in precisely these kinds of legal "corner cases."

I don't know which other cases to which you're referring, but the argument to be made here is that this isn't a corner case. This is having mathematical certainty that the defendant has evidence and is refusing to hand it over.

Sue me for being paranoid.

No law against being paranoid :)

but it's extraordinarily clear that shouldn't be happening in the first place, according The Constitution.

Actual question: where in the constitution is this clearly stated?

No, and thanks for catching that.

My understanding is that the prosecution doesn't need the decrypted data to secure a guilty verdict, but as they're entitled to it, they likely want it for secondary benefit.

They're staying within the law -- the defendant being in violation of the law is why an order to comply was filed and why we have access to the court of appeals document.

If you don't like the process, that's a different conversation.

To me, this whole thing smells of the classic tactic of telling the guy, "We know you're guilty; just confess, and we'll go easy on you." Which, of course, is a lie...If the hash information isn't enough to try him with, then I'd rather he go free, than set a precedent that it's acceptable for a court to compel someone to decrypt information because someone in law enforcement just "knows" the evidence is there.

I'm sympathetic to why you'd be cautious, but that's not fitting in this case -- this is a highly specific case with a number of circumstances that meaningfully differentiate it from the generic case of providing decrypted media. He's guilty and the checksums are enough to convict him (we're talking many checksums, metadata, partial confessions) and this is about him frustrating the discovery process.

Because once this order is allowed to stand, the level of certainty required to compel decryption is going to continually be lowered.

This is a slippery slope fallacy. I had some leaning towards this perspective, but then I read the source document, which goes into far more detail. There's a definite nuance to this case.

The defendant provided the password to his iphone (that contained highly-unsavory media of his nieces), which contained an unlock code for his laptop (filevault backup decrypt key). He connected the external drives to this laptop, and when he'd transfer media from his laptop to the drives, logging would occur with the file checksums. The hash/checksum is on the laptop with the filepath to the identified external drives, and because the hashes match known media of child victimization, the prosecution knows exactly where the evidence exists on the drive, once decrypted.

Subtle point: "they" don't want or care about his credentials -- they want the underlying evidence for which they know exists (check out "The foregone conclusion doctrine" in page 34 of the source document[0]) and they know that he is capable of providing said evidence.

> Here, based on Doe’s own statements, the testimony of his sister, and forensic analysis of the hard drives seized from Doe via a search warrant, the government already knows that Doe possessed and owned the hard drives, that he can decrypt them, and that they contain child pornography.[0]

Based on computer logs (of checksummed files being transferred to drives (and, importantly, knowing those filepaths) he admits to owning), online activity, witnesses, his own admission, and his unlocking of his phone provided the evidence needed to reasonably detain him on suspicion of a serious crime. The defendant is known to collect child pornography, even provably sourcing his own from family members -- again, the source document provides far more detail.

Further, my understanding is that the complication is his refusal is frustrating the process of deciding exactly which crimes for which to charge him and he is acting in defiance of a court order (to produce evidence).

[0] https://cdn.arstechnica.net/wp-content/uploads/2017/02/fedsr... (warning: some parts are sickening)

I think this is just a thing that happens in the movies. Also, check out the source document and let me know if you think someone found guilty of the described acts deserves immunity.

Head's up that the source goes into some detail and is a miserable, sickening read.

Per the source (https://cdn.arstechnica.net/wp-content/uploads/2017/02/fedsr...), the prosecutors already have a case based on checksums of the media that the defendant had downloaded, and per the logs, stored on his external hard drive. I imagine that the prosecution wants the media so they can perform harm reduction services for the identified, affected children and/or improve their data for going after other/future child abusers.

Or, this is a case in which a defendant had a non-foreign-government reason to accept whatever consequences come from not cooperating (thus not having political machinations at work) and it's a case with real human harm that makes it worth pursuing for the prosecutor.

Great clickbait title, venturebeat.

(Google and Microsoft are entering the group chat space..) In other words, Slack doesn’t offer anything that Google and Microsoft can’t replicate.

Of course Google and Microsoft can replicate the functionality of Slack -- they have smart, talented staff and chatrooms are a solved problem. However, where Slack excels is in cutting across organizational units, skill-sets, and getting people to use it. It's persistent-history IRC without a technical barrier and great marketing/user acquisition/word-of-mouth.

Unless you're a Microsoft tech-org (meaning C#/.net is the norm, not exception), there's a high likelihood that the only Microsoft software that engineering and design departments have installed is Skype; if I was at Slack, wouldn't sweat the scenario of sales, marketing, and engineering suddenly adopting Microsoft (in fact, "Microsoft Teams" was ostensibly announced in November and no one in my circles has mentioned it).

Google is a trickier problem for Slack. Everyone uses Gmail and Google is the landing page of the HTTP-internet. However, and maybe this is contentious, Google's need to integrate their myriad services turns users away. Not to beat a dead horse, but Google Plus could not have asked for better conditions for organic growth and it still somehow turned into a ghost town.

Almost every business either has G Suite or Office 365... As a company, why the hell would you shell out for one of them and Slack?

But, to my experience and opinion, "already paid for" and "close enough" may peel away some users from Slack and towards Google -- this is why orgs end up using HipChat.

I'm not playing identity politics. And I didn't mention racism and homophobia because that's not the perspective the author wrote from.

You suggested I'm not qualified to decide if the claims are believable (to me) and suggested that I would be if I was a woman that worked at Uber. "You don't understand X because you're not a Y" is a textbook attempt to reframe an argument away from objective considerations.

Claiming they're "absurd" as a way to present your argument that we should be skeptical is hard to take to seriously unless you're willing to say you worked on her team, or were another female engineering at Uber.

Please note that I didn't claim the author was a liar. I wasn't careful with my words, but: "This submission is just anonymous accusations. Absurd, unrealistic-sounding accusations."

It's truthful to say the accusations are anonymous. The level of misconduct described by the author is absurd and sounds unrealistic -- if these events were proven to have happened, they would still be _absurd_ and _sound unrealistic_.

And yes, I probably escalated things talking about privilege, but I think this is a good example of how we don't always realize the ways our viewpoints affect our opinions.

It's not escalation; invoking claims of "privilege" is head's up that you want to talk in subjective terms ("he/she feels", "other group experiences"), which is fine, but not fitting when I'm suggesting that I have a hard time believing an anonymous author's claims (e.g. "guys" at work writing sexual fantasy stories in private chat that the author somehow read).

An an aside, from reading between the lines, we have similar wants for the future culture of tech, but we're working with different strategies.

Medium isn't a court of law, they don't owe anyone any evidence and you pointing out that there isn't any is implying that they have a reason to hide their name that isn't justified.

You're right. And I don't have to trust what I read on Medium -- what I'm suggesting is that I don't believe _this particular anonymous author_ (and to be careful, I'm entirely willing to believe a culture of unacceptable behavior exists at Uber).

What evidence do you have that it's absurd?

None, which is exactly the level of evidence put forward in the blogpost that it even happened. Susan Fowler's account is a plausible account of a toxic environment; I default to believing her for various reasons. Claims of open/unguarded homophobia racism, sexism, fireable remarks, coworker-sexual-fanfiction are pretty severe as the consequences are steep and financial for the accused -- if this blogpost isn't hyperbole, I hope the author is reading this comment as she should be talking with a lawyer.

Do you think men don't ever...Because I have know a ton of women...

Disproving absolutes is a fool's errand and I'm sure you know the way this is phrased, I can't argue against it.

My own.

I'm not making claims of a sexist, racist, and hostile environment in a company, so for what it's worth, where I work isn't relevant.

... is a privileged viewpoint, and offensive to me.

"Privileged"?

I'm not sure how or why you chose to take offense, but I'm surprised that anything about that was even slightly controversial.

The author doesn't have to prove to your satisfaction that this happened to her to write this article

You're right, but that doesn't mean I have to believe what the author claims and doesn't mean I can't respond to it. The comment section doesn't exist to serve as an echo chamber.

Please tell me what is "vile" about saying "I believe Susan, but this blogpost strains credibility. If the author is being honest, I hope she speaks to an attorney".

so much so that even open source projects use it over IRC.

Remove the small amount of technical know how to "properly" use IRC and make chats (and permissions) persistent, and I imagine you'd end up with something that looks eerily like Slack.

That said, I think the Slack load-time in-browser is silly-long and I get spammed with notifications. As for threads, Zulip (https://github.com/zulip/zulip) does things "right" and I wish Slack would rip off their approach (central topic is the "room" and threads are independent, visually-separate-when-highlighted chats within a central topic).

Congress, for practical purposes, went on strike to attempt to ensure that Obama would not be able to accomplish most of his goals - in effect punishing the US people for political purposes.

Or, if acting in good faith, disagreed that his goals would be good for US people. More laws and actions do not mean the people are better off.

You're wondering whom I'm calling "borderline mentally challenged?" That would be Donald Trump

Thank you for your response, but we're going to just fundamentally disagree and while I'm fine with "in my opinion...", the level of hyperbole involved is enough to make me sit this one out. A "borderline mentally challenged" person would not be able to navigate his way to the position of POTUS.