My home cluster is built from surplus Dell Optiplex desktops that I got from BYU Surplus and added some RAM (before RAM price went totally bananas) and SSDs to.
I spent less than the cost of one of these Pis to acquire all of them together.
I later added a large machine that I used to use as a Linux desktop, with a GPU and 64GB RAM, which I use for generating OpenStreetMap tiles.
I’m not sure why you’re being down voted because you are breaking any rules as far as I can tell, but I must say: after reading your comment, I feel very sorry for you. We only get one chance at life and I want my kids to spend it being happy and fulfilled, doing things they love. I see no reason to prepare them for being Silicon Valley wage slaves at ages 4 and 7. Even if they did want that for themselves, they don’t need to worry about it for a long time. I didn’t own a computer until my junior year of high school and here I am, a principal engineer for a big SFBA tech firm.
I am a parent of two boys—ages 7 and 4–and I hardly recognize the negative forces that this author laments over because my wife and I have simply chosen not to raise our kids that way.
When my boys get home, we get outside and do sports. We’re in the middle of (tackle) football season right now and that’s what my 7 year-old does five days a week for two hours every evening. It’s exercise, fun, camaraderie, and lessons for life, all rolled together. My four year-old spends the practices on the sidelines, digging in the dirt with other little brothers and sisters and chasing the bigger kids around the track when they run laps.
When football season is over, we move on to other team sports. When those are over, we ride our bikes around the neighborhood and play pick-up football games in the front yard.
What we don’t do: iPads, TVs, or smart phones. It’s not that came into parenthood with some anti-technology mindset. Rather, we let our kids have these devices and we saw immediately the negative effects on their behavior, attitude, and motivation, and so we took them away. They get to use them on the rarest of circumstances: during airplane rides or as a brief reward for a week of good behavior.
Adam Carolla once said that the problem with kids isn’t all of the things that we’ve added to childhood, it’s all the things we’ve taken away. Sports, free time, boredom, tree forts, and neighborhood exploration have all been curtailed. For much of this, we have over-involved, over-concerned mothers to thank (and a bunch of dads that think like moms).
It doesn’t have to be this way. Nobody is forcing us to raise our kids in the post-millennium style. You can raise tough, resourceful, happy, outdoors-loving kids if you want to.
If you are a parent of little ones, have a look at this article:
The answer--as I recall--from the Chromium team was that they see trailers access as a security issue. HTTP/2 multiplexes many connections and it was a security risk, they said, to allow one client process to access trailers for connections that may be used by different processes or even entirely different sites behind the same load balancer/reverse proxy/etc.
What I don't understand is why gRPC was designed to depend on trailers. Web gRPC is the killer app for this technology. If they could build a new version of the protocol that doesn't rely on trailers, I would be thrilled.
Sorry, I should update it. All of those public endpoints went down a long time ago. People were donating the endpoint machines and they eventually just disappeared.
You'll have to set your own up on a server somewhere.
Don't use Speedtest.net. The major providers all prioritize their traffic and it gives unrealistic results.
I built a graphical console-based speed tester (client+server) that you can run yourself to get a more accurate picture and test between your home and servers you actually use:
Thanks. It's a 1987 Land Rover 110. It is ex-British Army and I bought it from a government liquidation firm in the UK and imported it myself. I did my own frame and engine swap on it so the body and interior are pretty much the only original bits. It has a Land Rover 300Tdi diesel engine.
I don't blog much but I have done a few stories for Expedition Portal:
I don't disagree at all. But, unless you're Scrooge McDuck, you're going to spend your discretionary income on something so why not a nice watch?
This little assemblage of $500 in parts is worth far more to me than the $8,000 that it's ostensibly worth. It has been there with me for so many of the most important events in my life:
I've travelled to third-world countries [1] with mine and I didn't insure it. I can't say I'd always do this, however. It depends on where you're going and what you're doing. In general, it's not wise to be flashy in a poor place but the stainless steel Sub isn't as obviously expensive as a gold watch and probably wouldn't catch a thief's eye like an iWatch, even though it is well over 10x as valuable.
I treasure my Submariner. It's the one article of fashion I have that goes perfectly with every outfit; it looks great with a tuxedo and it looks just as good with shorts and a t-shirt. It's incredibly well-made: I wear mine every day and after nine years together, it still looks better than the year-old iPhone in my pocket. It's also understated--at least, my stainless steel, black-bezeled Sub Date is--and it doesn't scream "expensive" or "showy" like many watches on both ends of the price spectrum often do.
There are a few advantages to owning a Rolex that may not be so obvious: For starters, it's quickly exchangeable for a high percentage of its value in cash in virtually any major city in the world. Because you're always wearing it, that's $8,000+ in cash that you can have in short order if you needed it badly and circumstances prohibit you from accessing your bank account.
Another advantage: they almost always appreciate in value over time. Very few physical assets can claim similar long-term value. People have been coveting these watches for nearly a century and that's unlikely to change. I will surely give mine to one of my sons someday.
If you can afford it, I highly recommend you treat yourself to one. I get enjoyment every day when I put mine on. I can barely remember any of the cars and trucks I've owned but I spent less on my Sub and I'll never forget it.
For every watt you put in, you need to cool that watt, which requires a high percentage of a second watt. Some of that equipment is almost 20 years old, maybe more. One those old machines could warm an office back in 1994.
On top of the electricity for the servers and for cooling, there is also maintenance: replacing ancient drives and other system components when they fail (which may require paying out the arse for some gear on eBay), fixing cooling equipment (ever replaced a dead AC unit in your house?), network maintenance and upgrades, internet connection, etc., etc. The list goes on and on. $20K seems reasonable to me. He's not asking for $20MM or even $200K, guys. He's not mining bitcoins. Please be reasonable.
I'm sorry, but that is just absurd. Some of the servers in question are coming up on 20 years of age. We didn't have the power-conserving hardware back then that we have now. Those old SPARCstations put out heat like a hair dryer. That stuff uses a lot of power and remember that ever watt of power you put into a server needs to be cooled. So, basically double the electricity costs for any given piece of gear. This rack is not cheap to run or cool, and definitely not in a home basement without access to highly efficient (and huge) datacenter-class cooling and power infrastructure. There are reasons that this gear needs to stay at Theo's place. If you don't like his reasons, you don't have to donate. He didn't email you and ask for a donation; you read the story here.
Thomas, thanks for that comment. If there was a "best of HN", this comment should be a part of it. Good storytelling, a great business lesson tidbit for all of us, interesting technical discussion, and a good reality check.
As a longtime OpenBSD fan and advocate, this has always fascinated me. I loved SGIs back in the day but they are slow as shit today and unusable for any kind of modern desktop usage unless all you do is write code in a terminal. These platforms survive in OpenBSD land because somebody still cares enough about them to enjoy hacking on them. There's no point in saying "Drop them!" because the devs working on them probably could care less what the rest of us think.
Personally, I do wish OpenBSD could somehow regain the popularity it once had and that support for modern hardware like 10GBE and scaling PF throughput w/ multi-core CPUs would improve. I don't know what it would take to bring people back.