npkarnik

First decide why you want the job. There are good reasons to work at an early-stage startup, the pace of work will stretch your capabilities exponentially more than a stable job will (from personal experience). The skills I developed working at a startup were worth more in hard $$ after I left than I made working there.

If your primary motivation is financial, you need to think carefully. You are taking an irrational and outsized risk as an early employee if cash compensation is below market. But people play negative expected value games all the time at the casino.

Even if the expected value of the equity is "above" market, the variance on that projected equity is significantly worse for you than getting it in cash.

Either way, you should push for roughly market cash compensation, framing it as "your expectations." I have said these exact words before: "The salary figure is definitely below my expectations (which is $___), I'm worried this won't work out." And then pause to get more information from them. Remember, your expectations are framed by your alternatives, which makes it easier to walk away. Best thing is to have another offer, if you don't it's not the end of the world though. They still have to fill the position, which is leverage for you.

I think the point is that people give to charities as a proxy for "helping people." Massive research oriented charities probably won't be displaced by something like GoFundMe, but the idea is that average people can directly fund the needy on a case by case basis (i.e. instead of paying for the administrative fees of a charity that helps sponsor scholarships, you can directly fund the recipient?)

Github Enterprise. Turns out that maintaining a git server and the familiar and great (imo) UI is something that medium-large companies generally have no interest in doing in-house and will gladly pay for.

Many highly valued tech companies do secondary offerings to allow average employees to get some liquidity.

Sure, the general public may not "get in early" but M&A is far less risky for both VCs and general investors. If mostly "sure things" make it to IPO, it's far less likely for the general public to be exposed to the meltdowns that made the headlines circa 2000-2001. The flipside is that until the startups IPO, the VCs and founders are exposed to most of the risk.

Hmm, it depends a lot on the size/type of company and reason for acquisition. If it's closer to a acqui-hire where the employees of the "acquired" company cease development on whatever they were doing and eventually just get staffed on a Google project, then they will MOST LIKELY do technical due diligence on each team member. It's common for only part of the team to get an offer to join.

Fair use does in some cases

Very Relevant: http://www.crypto.com/papers/export.txt

(current Penn professor Matt Blaze's mid-90s account of registering as an arms dealer with US Customs to transport an encrypted telephone to Europe)

Yammer is a team communication solution Microsoft already bought, but is way worse at synchronous communication (chat) than Slack, but also better at at asynchronous communication (threaded conversation, like on Facebook).

Microsoft would want to buy Slack for the same reason they spent $$$ on best-in-class products like Yammer, Sunrise, rebranding Acompli as Outlook. Slack's revenue is incredible for a startup, but a drop in the bucket for MSFT (for now). More immediately, Slack represents an entry point for Microsoft to upsell Slack's customers to other Office365 services and Azure services.

I didn't realize that the Slack CEO was also the Flickr founder, so you might be right in that Slack just isn't for sale to a megacorp.

But Microsoft is the most likely buyer, it's been on a spree with integrating best-in-class enterprise products/productivity tools into Office 365 (Sunrise, Accompli, previously Yammer). Slack fits into that mold, though the sale price would probably be orders of magnitude higher than those acquisitions (likely upwards of $10 billion, possibly much more). At that price tag, it might be untenable.

Not sure what Slack's medium-longer terms goals are, honestly. Their valuation is so high that only megacorps can afford them, but there's still enough risk that public markets might be skeptical (their revenue growth is astronomical, but they haven't been generating revenue for THAT long). In the interview, Stewart mentioned market saturation eventually limiting their crazy growth numbers, but I'd imagine they have a ways to go, given that the total enterprise market has to be well more than 200,000 total seats. Exciting times for them.

Which is preferable to the current model of your-and-my-data-and-eyeballs-to-advertisers-as-a-service

Though Docker's current valuation wasn't disclosed, it was rumored at $400m for the Series C round, and has probably at least doubled since then, given the amount of capital raised right now and the fact that they haven't spent all of the Series B. I seriously doubt that a $1bn exit would make Docker's investors happy

In some cases maybe, but hopefully most founders who obtain some personal liquidity in later rounds are not sadistic/hypocritical enough to deny their employees the same opportunity.

But you're right, one potential large risk is a Chris Sacca -like situation, where one investor/investment group uses many anonymous buying agents to acquire a huge stake in a takeout/IPO candidate, via secondary liquidity. That can mess up a final outcome for whoever thought they had control over the cap table.

?? Barkley retired after the 1999-2000 season.

A lot of the very recent, surprising successes in hardware (Oculus, Pebble) were crowdfunded (specifically for those, Kickstarted). I would go as far to say that using a premier crowdfunding platform to test the market may be the best, lowest risk way to validate a hardware idea. Or at least crowdfunding lowers the barrier for hardware success from "damn near impossible" to "really really hard."

If you have ever successfully done internet (and in particular, email) marketing, you would know that this is not BS.

Targeted emailing works. Your potential customers are engaging with you for a reason, they often need some kind of activation energy (that speaks to their situation) in order to enter their credit card information.

There are more wealthy people suffering through engineering degrees than you'd think.

One of my favorite David Foster Wallace essays is this Esquire piece on an obscure but top 100 tennis player named Michael Joyce (who later coached Sharapova for a while) http://www.esquire.com/features/sports/the-string-theory-079...

Palantir seems to have stopped lettering their rounds after G. Though I'm not sure about the distinction between a $400 million Private Equity raise and a $1.2 billion "Series E" a la Uber

A lot of tech related journalism gets (fairly) maligned. This is truly amazing writing, and I highly recommend reading the whole book, even if you know most of the story.

Extralegal power has always been a mainstay of the federal government. For 20th century examples, the FBI was notorious for monitoring and blackmailing Civil Rights leaders (associating the movement with communism), especially MLK. Japanese internment during World War II was authorized by FDR's executive order 9066.

There is literally no legal basis for most of the NSA's activity. But there is plenty of extra-legal handwaving that justifies the NSA's behavior...executive orders, "secret court" FISA rulings, all culminating in the use of this intelligence to assassinate and indefinitely detain/torture those (citizens and noncitizens alike) who are a "continued and imminent threat" to the United States.

"Slippery slope" arguments used to seem weak and contrived. But I honestly can't help but feel we've reached pretty close to the bottom of that slope. In the sense that I feel like our government just kind of makes stuff up as it goes along and only occasionally responds well to the strongest forms of shame.

Thanks a lot for the catch.

Working on a mixtape right now actually.

I can't really speak to R, having barely used it, but Python is definitely better to know in general, given that it is heavily used for scientific computing/data science and many other domains. Zillabyte's original API is actually for Ruby, and we just started supporting Python recently.

Ultimately, R vs Python comes down to a preference in tools. R has a large community and is specifically designed for statistical analysis, so a lot of statisticians. I would feel weird recommending a tool I don't actually use, but that's not a knock on R, just a lack of need on my end to learn it.

I'd love to improve the UI. My frontend skills need work, but any suggestions on javascript for interactive graphing? I played with sigma.js for about 15 minutes.

Author here, something I've always wondered: how many software engineers are hip hop heads?

I hope this guy succeeds. In most markets, blackouts are rare, but when they do happen, it's nothing more than a disgraceful power play and a relic of rules that were devised 40 years ago. The real crime is publicly financed stadiums.

The tax-payer funded stadiums racket is usually driven by teams explicitly or implicitly threatening to move - the oldest and most effective trick in the book. A recent example was the half a billion dollars the Minnesota legislature coughed up to the keep the Vikings from (probably) leaving for L.A. Remember that the most popular and profitable American sports league doesn't even have a team in the second largest market (even though it used to have 2) and realize that the NFL will continue to be able to hold cities hostage under threat of relocation.

Even the New York Yankees have tried it ( http://nymag.com/nymetro/news/sports/features/2860/ ) and ended up with over $200 million of public financing for New Yankees stadium if I remember correctly. Almost as if someone actually believed that they might become the New Jersey Yankees??

The most egregious recent example is the New Braves Stadium in Cobb county that was subsidized with around $300 million of taxpayer dollars by "tea-party conservative" Cobb county officials. All of this for a stadium that isn't even 20 years old...in a county that had to lay off around 200 teachers and furlough the rest.

Not a bad business, if you're an owner in a North American sports league. The NHL is a different story for most markets, but owning an NFL or MLB team is unquestionably a license to print money. NBA owners love to plead poverty when the CBA is up, but the goddamn Milwaukee Bucks sold for over a half a billion and the wait line to buy an NBA is longer than ever.

Some of the more recent negative stories around the Astros are alluded to in the article, but it's worth emphasizing that Houston is probably the most dysfunctional organization in the majors.

The focus of the piece, GM Jeff Luhnow, has an ongoing and borderline public feud with the manager Bo Porter.

They were the only team in recent memory (and possibly ever) to have 10 months of internal trade talks leaked.

And most recently, they were the first team since 1983 to not agree to terms with the 1st overall pick. Sure, they supposedly identified a red flag in Aiken's elbow, but it seemed driven by their desire to sign specific additional draft picks (none of which they did) and earned them the wrath of the player's union.

Whatsapp is a paid service and was profitable (probably highly highly profitable)

No, unemployed people count as participating. Non participating people don't count towards the official unemployment rate. For participation to drop as unemployment grows is the worst possible combination of trends.

Similar space, but a few major differences.

PiCloud served a different niche, as it provided much lower level access to resources. From what I understand, doing distributed computing on the PiCloud platform required the developer to coordinate job spawning and other lower levels details of distribution.

Zillabyte is designed for distributed computation on demand without requisite distributed systems expertise.

Hadoop MR, Spark, Storm, etc. are all excellent systems for large scale computation, but realistically, the learning curve to administer and run non-trivial jobs on these platforms is quite steep. At larger companies, there are entire teams that specialize in maintaining data infrastructure systems and running jobs on them.

For smaller companies, or those without domain expertise in the "big data" ecosystem, it can be daunting to acquire the capabilities to be Hadoop/Spark experts.

[disclaimer: I work for Zillabyte]