HNHacker News
TopNewBestAskShowJobs

nateberkopec

3,366 karma · joined March 23, 2009

submissionscomments
nateberkopec··on U.S. regulators accuse Palantir of bias against Asians
Very interesting line in the complaint that other companies should take note of:

> "In addition, the majority of Palantir's hires into these positions came from an employee referral system that disproportionately excluded Asians."

Also, the complaint says they tried twice to get Palantir in compliance (starting in late '15) and are only now filing this complaint.

nateberkopec··on Computer Vision Syndrome
Time Out for Mac has a "Pro" level feature which waits for a natural break: http://www.dejal.com/timeout/

It's a very customizable tool - I've used it for months.

nateberkopec··on Twitter: It is too late for it to become the giant people expected
When Snapchat (a pretty good comparable for Twitter) raised in 2014, they said they had 100 million monthly active users [1]. According to leaked financials [2] they were doing it on about 10 million a month, or just 30 million per quarter.

Twitter supposedly has 300 million MAUs, so your estimate of 100 million/quarter may be right on the money.

[1] http://blogs.wsj.com/digits/2014/08/26/snapchat-said-to-have...

[2] http://gawker.com/snapchat-lost-a-ton-of-money-last-year-170...

nateberkopec··on Twitter: It is too late for it to become the giant people expected
Twitter had to kill their ecosystem in order to make any money on advertising. There was just no way to sell ads if so many users were going to be using 3rd party clients.

You may be right that killing off the 3rd party ecosystem hurt their usage, but from a business perspective, there wasn't any point keeping it if it was damaging their business model.

nateberkopec··on Twitter: It is too late for it to become the giant people expected
Snapchat is a pretty good comparable for Twitter's business model and employs less than a tenth of the employees, according to Forbes: http://www.forbes.com/sites/liyanchen/2015/08/11/the-most-va...
nateberkopec··on How I spent two weeks hunting a memory leak in Ruby (2015)
This looks really, really cool. You should download the bugged version of this gem and make a screencast showing how your tool makes this process easier.
nateberkopec··on Proposal of a new concurrency model for Ruby 3 [pdf]
> This Ruby 3 proposal says that Ruby 2 compatibility is mission critical

Matz has been very public about his fear of a "Python 3" situation occurring in the Ruby community.

nateberkopec··on Proposal of a new concurrency model for Ruby 3 [pdf]
> If I'm reading this proposal correctly, locks will still be needed within multithreaded guilds to guard mutations against complex object graphs.

That is correct. You'll still need to use locks if doing multi-threading inside of Guilds.

It looks like Guilds can have 1 to X threads.

nateberkopec··on A SpaceX Falcon 9 rocket has exploded at Cape Canaveral
Thanks, that's the answer I was looking for. Other commenters seemed to think I was asking why they test at all!
nateberkopec··on A SpaceX Falcon 9 rocket has exploded at Cape Canaveral
I know nothing about space - why are they doing tests, where the rocket could possibly explode or fail, with the payload on board?
nateberkopec··on How to become the sole owner of your PC [pdf]
Yes. Wiki: https://en.wikipedia.org/wiki/Intel_Active_Management_Techno...

> A Ring -3 rootkit was demonstrated by Invisible Things Lab for the Q35 chipset; it does not work for the later Q45 chipset as Intel implemented additional protections.[39] The exploit worked by remapping the normally protected memory region (top 16 MB of RAM) reserved for the ME. The ME rootkit could be installed regardless of whether the AMT is present or enabled on the system, as the chipset always contains the ARC ME coprocessor. (The "-3" designation was chosen because the ME coprocessor works even when the system is in the S3 state, thus it was considered a layer below the System Management Mode rootkits.[32]) For the vulnerable Q35 chipset, a keystroke logger ME-based rootkit was demonstrated by Patrick Stewin.[40][41]

nateberkopec··on Ruby has been fast enough for 13 years
It's true in practice but not in theory.

Ruby can't defragment memory pages - if there is an active object in any heap page, it cannot be released back to the OS. Technically, Ruby does release empty pages back to the OS, but in practice, heap pages are rarely completely empty.

nateberkopec··on SpareOne: Emergency mobile phone powered by AA batteries
I own a DeLorme inReach - ~$300 device with a ~$15/month plan. Runs on the Iridium network, unlike Spot's proprietary network. Also, inReach devices are duplex, and can send text messages to friends, family, and SAR teams.
nateberkopec··on SpareOne: Emergency mobile phone powered by AA batteries
If it's a phone only for calling 911, why does it require a 2-year cell plan with data?
nateberkopec··on Helix: Rust and Ruby, Without the Glue
Akira was talking about a specific context - view rendering. Which makes sense, who the hell ever renders a view in anything other than UTF-8?

Checking input, however, is a whole 'nother ballgame.

nateberkopec··on F-35 Program Office Signs Off on Air Force 3i Software
> "The best analogy is you are starting up your computer and you want to use Word, Excel, PowerPoint and Outlook, and you are trying to get your work done for the day, and PowerPoint and Outlook came up but you are having an issue with Excel,” Brendan Rhatigan, director of engineering and test operations for the F-35 ITF, told Defense News last week at Edwards. “So you say, I don’t know what’s going on, so let me x out of that, let me restart it again.”

So what did they do with this software update? Just install "CTRL+ALT+DEL"?

nateberkopec··on Magic Roundabout: Circumnavigating the World’s Most Complex Intersection
FTA:

> and has been adapted to create other ring junctions around Britain:

That's a GIF recording of Cities: Skylines, not a roundabout in Britain. Purple pollution in the upper-right gives it away.

nateberkopec··on Richard Stallman's 22-Page Rider (2011)
> In some places, my hosts act as if my every wish were their command. By catering to my every whim, in effect they make me a tyrant over them, which is not a role I like. I start to worry that I might subject them to great burdens without even realizing.

Then:

> I do not eat breakfast. Please do not ask me any questions about what I will do breakfast. Please just do not bring it up.

nateberkopec··on Kindle Unlimited scammers on Amazon
Sure: http://www.amazon.com/Alone-Across-America-Motorcycling-Back...
nateberkopec··on Kindle Unlimited scammers on Amazon
Not at all. It's just a side for me. My sales are 100% thanks to Amazon searches.
nateberkopec··on Kindle Unlimited scammers on Amazon
I suppose so. But they're still charging 10x what I pay per MB for my cellphone. I use Ting, which runs on the same Sprint network as Whispernet, and I get 4G.
nateberkopec··on Kindle Unlimited scammers on Amazon
You're right, I have no idea where I got that figure. Editing.
nateberkopec··on Kindle Unlimited scammers on Amazon
Kindle Unlimited is Amazon's "Spotify/Netflix for Books" product. $10/month to read any book enrolled in the Kindle Unlimited program.
nateberkopec··on Kindle Unlimited scammers on Amazon
Ouch. I wrote a travelogue as a side project/on a bit of a lark, and I enrolled it in "KDP Select" (Kindle Unlimited). I make far more money, it seems, from full-price purchases than my per-page royalties from Kindle Unlimited. It really seems that it doesn't make sense for most authors, since the revenue-per-page is diluted by these scammers.

EDIT: To put some numbers on it, here's my KDP dashboard: http://imgur.com/lDWoQQ6 The book is $4.99, of which I receive 70% ($3.50). I sell about 3 copies per day (making about $10), and generally get ~350 pages read per day (making $1.50). So I make 10x on regular sales what I do on Kindle Unlimited. If it matters, it's a short book of ~300 Kindle-normalized pages.

While writing this up, I realized Amazon charges authors $0.15 per megabyte of book file size per download. My book was 10 MB thanks to a bunch of photos. After extreme JPEG compression I was able to get it down to ~2MB or so. I wonder how many authors, like me, are losing 20% of their royalty without realizing it.

On AWS, you're charged ~$0.09/GB of bandwidth out. How Amazon thinks 1000x pricing is fair here is beyond me.

nateberkopec··on DO Not CLOSE THE ISSUE ASSHOLE
The collaborators in this issue (@flovilmart and @hramos) have way, way more patience than I would have had. Props to them.

That said, why wouldn't they use a software implementation of CRC32? Is it really that big a bottleneck for them?

nateberkopec··on Polaris: Faster Page Loads Using Fine-grained Dependency Tracking [pdf]
I'm curious to know how much this improves page loads when resource hints are used properly. I may be failing to understand the paper, but it seems to me that strong usage of resource hints plus HTTP/2 dependency-based prioritization makes a lot of this technique obsolete.
nateberkopec··on How Changing WebFonts Made Rubygems.org 10x Faster
Worth noting Chrome DevTools makes this super easy now - network throttling is built-in to the tool and can be applied any time DevTools is open.
nateberkopec··on How Changing WebFonts Made Rubygems.org 10x Faster
Hm, you could probably hack something together with the PerformanceTiming API, which makes a lot of these timings available to JavaScript: https://developer.mozilla.org/en-US/docs/Web/API/Performance...
nateberkopec··on The DROWN Attack
> Disabling SSLv2 can be complicated [...]

If users are accidentally enabling a feature that's been insecure for 20 years, the vulnerability is that your configuration mechanism is too complicated to understand.

nateberkopec··on A Message to Our Customers
Is there any doubt that when the FBI brings up a law from the 1700's to justify breaking digital encryption in 2016 that they are completely making it up as they go along?
← PreviousPage 7 of 15Next →