narsil

Oh wow I totally missed that. Read it as the same blog post. I jumped right past the date line without even seeing it. The URL's anchor tag didn't update when scrolling, which probably contributed.

There aren't reports of Kalanick himself harassing women.

German isn't "gibberish" to some of us, so I would not describe it that way. It's a German banking API after all. A functional API that is being used is better than no API at all.

It appears the extension would like to read and change the data on any website I visit. I assume users are the product here, so this appears somewhat sketchy since I'm not sure what personal data of mine is being sold.

I'm curious about how the hack occurred.

With the recent Hipchat hack, it is unclear if the login server was compromised or the database dump was obtained some other way.

If the Hipchat webserver that handles sign-in attempts was the one compromised, which is what I've been lead to believe, then it doesn't matter how the passwords were hashed. The plain-texts would be exposed in the majority of use cases. Sure, the DB data could also get stolen triggering reporting requirements in organizations with controls, but that may be less of a concern at that point.

Minor nitpick: I get what you mean, but GitHub might be a poor example since GitLab is a strong competitor in this space that is arguably better in both features (self-hosted) and price (free). That said, I can't think of ones for Basecamp or Trello off the top of my head.

> > Most universities get by with a dozen clustered on campus serving administrative, educational and research objectives > > ... A dozen what?

Looks like GP meant "a dozen data-centers", to compare the costs.

The webserver could have been compromised causing plaintexts for login attempts to be exposed as well. In fact, that is a very plausible explanation for how the database was accessed since it is usually firewalled off.

I would expect it to be ES2017 or similar.

FWIW Comcast claims this will not affect the speed [0]. At the low speeds in the OP's article, this is likely the case.

[0] https://www.xfinity.com/support/internet/xfinity-wifi-hotspo...

Switching to 5 GHz instead of 2.4 GHz has made this magnitude of difference for me in the past.

Today, at home, I haven't had dedicated internet connectivity for 4 years. I log in to the xfinitywifi SSID that some of my neighbors in my apartment complex are probably broadcasting via my parents' comcast account and get good speeds. I'm on it right now and speedtest.net shows 30 Mbps down, 2 Mbps up. Plenty enough for my purposes, including Netflix.

I switch to tethering on my AT&T device if needed for 15-25 GB 4G data each month; the "unlimited" plans offered by carriers today don't provide more than 10 GB of high speed data in contrast. Not sure how much longer this setup will work, but I plan on continuing it for as long as I can; the cost savings really add up.

A side-effect of this that I've noticed is increased operational awareness on my part. Comcast MITMs any unencrypted HTTP traffic and injects in a "powered by xfinitywifi" banner that is a great reminder to not trust the network. :)

It appears to have been a design issue and not necessarily due to the time constraints placed on engineering to launch the spacecraft. Looks like there would have been a strong chance of this occurring anyway, as the initial assumptions of type of data to send back for analysis were insufficient.

The page appears to have been taken down.

Google cache: https://webcache.googleusercontent.com/search?q=cache:XWzfaA...

Nice. Is there an equivalent for Python? A quick Google search doesn't seem to pull up anything obvious.

This is awesome. I developed Chrome/Firefox browser extensions that connected Gmail attachments with cloud storage in the past, and hacking through Gmail's DOM always threw any SLA we could provide out the window.

It's even better that this will run on Android and iOS as well, which truly makes all the "Gmail extensions" from developers such as myself production-ready.

As a sidenote, returning something bought via Instacart is a pain, as I learned. Target refused to accept a defective light-bulb without the original receipt, but did exchange it for store credit fortunately.

This is awesome. Previously, this would have required easily pronounceable/typeable passwords to be able to log in on mobile. I can see this incresing adoption of LastPass, and eventually revenue even if conversion rates drop.

Why do this when letsencrypt.org provides free trusted SSL certs?

I've used https://github.com/alanjds/drf-nested-routers with success.

Yeah, the Stable Marriage problem provides a solution for matching these ordered preferences between two sets of people/elements: https://en.wikipedia.org/wiki/Stable_marriage_problem

Technical solutions are unnecessary. Just make the tickets non-transferrable, and have the ticket holder produce ID confirming the name on the ticket matches the person. Plenty of music festivals, Vegas nightclubs and other places already employ this strategy very effectively.

Thanks for the explanation colmmacc! I agree with the challenge you described, and am not sure what the best approach would be. Perhaps a configurable timeout such as ELBs have?

We recently discovered that the NAT Gateway also terminates connections by issuing a RST packet when it receives the next packet for a connection that it believes to have timed out, effectively causing the new request to fail. The previous recommended approach of NATing in VPC was to use NAT instances, which sent FIN packets when the timeout was hit, cleanly closing the connection. That behavior was far better, since it indicated that a new request should re-connect first.

AWS Support indicated that this was a feature of the new NAT Gateways, even though it breaks outbound connections made by popular implementations such as the Requests python library's urllib3 connection pools. This is pretty unfortunate, and has been a roadblock in migrating to the NAT Gateways.

Not sure what you mean. Plenty, if not most, engineers on HN are compensated at market rate for their region and skill-set. The uproar over the anti-poaching scandal involving Apple, Google, etc. from a few years ago was due to the artificial wage deflation that it created. I can't recall folks from HN siding with the corporations during that time. Sure, sweat equity is valued at early-stage cash-strapped startups, but that is in addition to "passion" and a lower salary.

I prefer Celery for Python. Far more light-weight, and intended for distributed task processing.

http://celery.readthedocs.org/

The home page has a 'Why use Storm' section that you might find helpful: https://storm.apache.org/index.html

For example, Twitter used to use Storm to process analytics on tweet data. They now use Heron. [1] Storm is valuable when your data is constantly being generated as a stream, on which you would like to perform processing immediately. I've previously also used Storm's distributed RPC cluster, where I fire off a task that I would like processed in the background and wait for the response to arrive.

[1] https://blog.twitter.com/2015/flying-faster-with-twitter-her...

Agreed. The DeepMind team have already done this with other Atari games: https://youtu.be/08Cl7ii6viY?t=11m35s (Oct 2015)

"The system only gets raw pixels as input."

The results were pretty great, so it would be fascinating to see this work with Matt's version of SC2 as mentioned elsewhere in this thread: https://news.ycombinator.com/item?id=11326119

I took Matt Walker's class on sleep at UC Berkeley (Psych 133) and it was fascinating. Short-term effects of a lack of sleep are easily observable, and I'm sure we can all relate from personal experience. However, long-term effects of sleep deprivation are harder to determine. One of the few known correlations appears to be that people who are susceptible to psychiatric disorders such as schizophrenia are more likely to experience them as they age, if they suffer from chronic sleep deprivation.

"Sleep problems may increase risk for developing particular mental illnesses, as well as result from such disorders." - http://www.health.harvard.edu/newsletter_article/Sleep-and-m... (2009)

> Off-topic: @dang, can we get a "Upvote this comment if has less than 1 point, else don't upvote it" button?

The font-color of the comment (aside from links already visited) should be enough to indicate if it has positive or negative karma. Individuals can decide whether to upvote or downvote further at that point.

I was curious about the 'nsfw' bit at the end of the lecture as well : )

The Xerox Star was truly ahead of its time. https://en.wikipedia.org/wiki/Xerox_Star