mscman

> any sysadmin worth his salt will have automated the population of his authorized_keys file and alert on any unauthorized changes.

FTFY

We didn't have DST in most of Indiana until about 10 years ago. When we switched, most of us were wondering why the hell we were even doing it. It's really not necessary, it confuses people, leads to screwed up schedules when the switch happens, and the fact that there are still a few counties that don't observe it screws things up even more.

While I agree that over-testing the wrong parts of your program can have an impact on program design, your end user's needs/desires are usually testable things. TDD helps make sure that experience is reproducible, even through sweeping changes to the codebase.

You'll definitely want to add a '&' at the end of that line so that you don't delay user logins if the network is down or mail barfs.

I don't know that a Slack bot is the right type of larger solution though. For instance, I use the ExpenseIt app on my phone because it's paired with Concur. That's got a much more flexible input when it comes time to do the actual expense report, and it's already processed all of those receipts for me. Now I just need to go through and pair it all up with the transactions on the card. Is Concur perfect? No. But the ExpenseIt app works there because it allows me to add data on the fly then toss the receipt.

I don't think I would be compelled to use a Slack bot to do the same thing, even though we use dozens of other Slack integrations.

> The real value is in the presentation.

This is true sometimes, but if it's a design discussion and you're whiteboarding how something should be laid out, or even steps of what's completed and what's left, just a snapshot of the final state of the whiteboard is tremendously useful. It prevents the need for another person jotting down everything that was written up on the board.

Where were your faulty detectors? My ex-girlfriend's dad used to do testing for a smoke alarm company a while ago. I learned an interesting thing from that. You know what they use to simulate smoke when they test smoke detectors? Hair spray. The fine mist is actually pretty good at tricking the sensor into thinking there are smoke particulates in the air. This also leads to an interesting failure scenario for many home smoke alarms: those close to bathrooms or in women's bedrooms tend to get gunked up with hairspray.

Except my phone will get itself in a state where that latency stays high, and bandwidth drops for a long time. Only solution is to turn wifi off/on. That definitely seems like a bug.

I'm fine with lower throughput while discovery is going on, but when nothing's changed on the network I've been on for days, and my phone can't use wifi anymore, that's a bug.

Because it's still technically classified. When you hold a security clearance, you're not only agreeing to keep things classified, but also to not pursue any classified information you don't need to know.

This was a big issue when WikiLeaks started leaking info. There were regular memos circulated in DoD and DoE environments reminding people they weren't to be reading classified documents.

Not just Google. Amazon's "Enhanced networking" uses a feature in 10GbE that's been used in smaller HPC clusters for around 5-10 years now. And MS Azure has InfiniBand backing their highest-tier instance types.

Many datacenters are adopting HPC technologies to reach the scale they need.

Not a Cray... That's gotta hurt. They've been a Cray shop for years now.

I'm guessing ORNL isn't going to move the target as much as NCSA did for Blue Waters.

I hadn't heard about that. That's awesome that the police chief is taking them on about this. It's becoming rare to hear about police chiefs doing the right thing when it comes to abuse of power.

Cloud HPC Solutions Architect here. In this use case, cloud computing wouldn't really suffice. For a company who needs 24x7 access to compute power of that scale, they really will benefit from buying a standalone system like this.

While some cloud providers now have very high-speed, low-latency interconnects, they only scale so wide in a virtual environment. This is due to limitations in the physical datacenter space and the way node provisioning works on the back-end. If you truly need extremely wide, low-latency, high bandwidth interconnect for a large period of time, on-prem is still (probably) your best option.

Given their initial reactions on Twitter, I don't feel bad if they get taken to court over this. As others have mentioned, they only agreed to rapidly reverse course after Microsoft put them in time out.

Except in your shortened version, you've completely eliminated what you're using to measure "achievement".

Plenty of people don't define academic success as achievement. A lot of blue collar communities consider other things as being successful.

It is? I'm not saying you're wrong, but I haven't seen legally organized protests (yes, you do have to file the appropriate paperwork) recently which have actually been shut down by an enormous governmental show of force. Do you have any examples?

I don't think so. Since the assets were on SR servers, which weren't technically Ross's personal property, and the business was used for selling drugs (admittedly along with other things), civil forfeiture laws allow the government to take those assets. Doesn't matter what happens with the lawsuit.

Awesome! Thanks!

I actually disagree with your definition of supercomputer, and liken it to any compute system or cluster with more power than an average enterprise server or workstation. After all, they really exist to solve problems which can't be solved by traditional computing.

I think the metric of it being one of the fastest 500 in the world is a little disingenuous to all of the other supercomputers out there.

As the article mentions, there are also people who lack the money or time to eat healthy all the time. College students in particular are pretty famous for their ramen meals because they're cheap and easy. They're also not filled with nutrients you need.

Soylent proposes an alternative to cheap $0.50 noodles that is a lot healthier. I know if it was available when I was a college student, it would probably have been in my pantry. I haven't tried Soylent yet, but probably will give it a shot to fill in those meals I can't make due to time constraints.

Same here. I work right next to the sliding glass door in our kitchen (door is to the right of me). Didn't read the description fully and jumped when it sounded like someone was knocking at my door, then at the window to the rear-left of me...

I feel some of this is based on personality. It's great that you're able to step away from work periodically and then focus again (I do the same thing working from home). A lot of people have trouble getting back into the work mode again after taking that break. That's what the "work account" idea is designed to help.

This works in a normal bash/zsh prompt as well if you want to set a variable only for that single command. It's why a lot of times you'll see build instructions like:

# CC=gcc make

After the command runs, CC will go back to whatever it was set to before (or unset if it wasn't set). Very handy for some operations.

It's not that the components themselves draw half the power, it's that you don't need to provide as much power to cool them. There's a typical rule of thumb in datacenter planning that for every 1W of compute power, you need 1W to cool it. If you can cut the cooling power, you can cut almost 50% of your datacenter cost.

Obviously you won't eliminate cooling cost entirely, and it's quite possible the overhead of replacing oil and extended maintenance times could offset these power savings.

That's a good point, there are enterprise hosted offerings. I'm just pointing out that the sell to higher level university officials to switch off of internal infrastructure to "cloud" offerings (yes, yes, I realize that word is overused) can be rather difficult.

There's a legal issue with universities though. While the research being done may be public, they could also involve private corporate entities, restricted data, or patentable information. A lot of people forget that, while universities are oftentimes public institutions, they're also very large bureaucracies. Most of them are very sensitive to how data is handled and published, as incorrect findings, proprietary code, and other leaks can be very costly.