This is a great way of thinking about it. Same goes for "unlimited" plans: the more I use, the lower their profit margins. I'd rather agree on terms and pay for what I use.
I got the offer and really considered it as well! It was the fear of an acquisition that convinced me not to take the offer. I was a TextDrive customer before they were acquired by Joyent, I was a LastPass customer before they were acquired by LogMeIn, and it was CrashPlan's cancellation of CrashPlan for Home that drove me to find rsync.net.
None of those other companies are rsync.net, so it's not really fair of me to consider them when making this decision -- but it made me think it's safer to just pay you guys regularly for a good service. Worst case, I spend a little more over the life of my account, but that money is going to a company I really like, so it's not much of a downside.
> We believe that the risk of "logical failure" of an SSD is higher than the risk of physical failure. This means that some pattern of usage or strange edge-case causes the SSD to die instead of a physical failure. If we are correct, and if we mirror an SSD, then it is possible the two (or three, or four) SSDs will experience identical lifetime usage patterns. To put it simply, it is possible they could all just fail at exactly the same time. The way we mitigate that risk is by building mirrors of SSDs out of similarly spec'd and sized but not identical parts.
This makes sense to me (and is a good example of looking at more abstract failure domains in addition to the basic ones we all know and love) -- I'm curious if there's data to support this. rsync.net is in a good position to possibly collect that data.
I think the "Mix SSD vendors/batch numbers" is a hold over from the very early days of SSDs where a handful of people get seriously slimed by having 90%~ of their same-brand-same-batch drives in a single machine fail at once due to some SSD batch failure.
As a side effect people generally now stagger SSDs a little to avoid something similar happening (ofc if you have multi machine replication this is less of a issue, but still a total machine loss can hurt due to capacity loss or parts shortages in edge locations, etc)
I've personally not seen a synchronised SSD array happen failure for a long time, but it's hard to know how much of that is because people now plan to avoid them.
"I think the "Mix SSD vendors/batch numbers" is a hold over from the very early days of SSDs where a handful of people get seriously slimed by having 90%~ of their same-brand-same-batch drives in a single machine fail at once due to some SSD batch failure."
I want to clarify - there's the issue of a bad batch wherein their longevity is greatly reduced and they fail in a cluster, etc., etc.
But that is not what we are guarding against ...
Instead, the risk we're thinking about is that there is an actual bug in the firmware that causes a particular workload to brick the drive or destroy it or whatever.
The critical point is that if the drives are mirrored then they experience an identical workload over their lifespan and they could fail literally simultaneously.
So by all means - do indeed guard against bad batches or manufacturing defects by mixing drives. Just understand we're talking about something slightly different here ...
I can confirm that e.g. there has definitely been e.g. a batch of enterprise SSDs from Intel a couple years ago which failed en masse after a certain amount of powered-on time.
From my limited experience I did have a pair of Intel SSDs in RAID 1 fail within 2 days of each other in the same way. Thankfully the first was replace and recovered before the second failed.
Why is it even possible to “charge [someone] as an adult”. If you have the concept of “adult” and “child”, and a line in between them, how can the crime dictate someone’s mental maturity?
(Of course the answer is, I’m sure, is selective enforcement. So they can lock up black kids and let white kids off the hook.)
Do you really think that a 14 year old is incapable of understanding that hurting someone is wrong? And that they don't understand that there will be consequences to their actions? Because I find it hard to believe.
>Edit: if you give a child the keys to your car and they hit someone with it, who is at fault?
So this is what HN has degenerated to? I'm sorry, but I remember eagerly reading the front page, learning things from interesting articles, and -- more importantly -- reading well-written commentary from people far more accomplished than I. I lurked, because I couldn't contribute at the level that most of the regular commentors could, but I learned a great deal.
HN has retained its preoccupation with not being reddit (which is noble), but has not retained the quality to justify it.
Great communities are transient, and HN is probably what taught me that. I have nothing but respect for pg, and for the community that once populated HN.
To be honest, I feel the same way, but maybe there's a good explanation for the relative decline of "quality".
I think I started lurking HN back in 2007 (which I also believe is the first year of HN). I remember feeling a euphoric sense of being overwhelmed with great content from knowledgeable people and articles. I learned a lot about start ups. I learned a lot about programming. I learned a lot about partaking and consuming thoughtful commentary in a semi-anonymous setting.
But you know what? All of that content is still here, in different formats. The audience is bigger and so there tends to be more mainstream stuff on HN now, particularly around politics. But I think the biggest change is in myself. I've read the same stuff about lean startups at least 100 times now rehashed in different ways. Even though I've never raised money from a VC before, every time I read something on HN about it, I feel like I've been through the process dozens of times. Every 37signals blog post that gets submitted seems like a rehash of the last one. So most of the submitted content around those topics seem like old-hat.
At some point, when you continue to consume the content from the same sources for years on end, you're going to see more repeat episodes and you'll have fewer stimulating moments. That's just the nature of it - and I think it's just as important to recognize that we're transient individuals. Your probably a different person now with a higher capacity and understanding and that's probably in part thanks to HN.
Your finding the articles less interesting could be a function of you becoming more knowledgable yourself, therefore there is less "low hanging fruit" with regards to new knowledge for you to acquire.
Might be a sign that it's time for you to start posting some well-writen commentary yourself ;)
Dude mkelly has known why a[5] and 5[a] were the same damn thing since before there was an HN. Stop making bad assumptions, he's telling a true story. I'm sorry to see him leave just as I realized he was here. Oh well.
He's not the only one saying these things. It's just how it is. After awhile, what was is no more. HN stood longer than most.
I'm sorry, but I'm happy to get voted down for my tone for once. I remember a time when one could assume that people around HN had a basic CS education and the community didn't get fascinated by stack overflow threads about how C works because we were all thinking about crazy interesting things.
I think anyone who says that people don't have a right to miss that and assumes that there is no problem is part of the problem.
I'm this is true to some extent, but (as much as I'd love to wholeheartedly agree with you), the reason I don't consider this likely is this: the main reason I have this opinion of HN is because of one 3-month period in 2009, where I was off the grid (bought a car in Europe, drove around, drank lots of booze, smoked some mushrooms, &c). I assure you I didn't learn about computers during that time. When I left in June, I thought HN was full of insightful articles and comments; when I returned in September, I thought it was markedly worse. I think HN had its Eternal September then, and it's been steadily downhill since then.
(Um, yes, I was keeping track of replies, and logged back in to post this. Mea culpa.)
My apologies if my post sounded disrespectful in any way though.
I wrote my post from the perspective of being a long time lurker on various tech forums who is only just beginning to feel I may be able to contribute on some topics (and only after many years of coding, reading & sponge-learning from sites like this one).
If it weren't for stories like this & people like yourself engaged in insightful discussion about them I personally would have missed a lot of "assumed knowledge" in the developer world, heck I wouldn't have heard of K&R (the shame!).
So I guess what I'm saying is: please don't leave, stay & help make us noobs less ignorant!
I think that it's general knowledge that a[x] is a + (x * sizeof a). The idea that a[x] could be reversed to x[a] is probably not common knowledge though. I don't necessarily think that a[x] = x[a] would be "C 101." I've certainly never come across any C code that uses this property and I can't really think think of any good uses for this property that don't end up ruining the readability of the code. And in general, it seems a rather useless property that, as others have stated, is just a side-effect.
I could just as easily claim that HN has 'degenerated' if someone posts a Stack Overflow post about what the /o flag does on Perl regexes. Or maybe there will be a Stack Overflow post about how substr('abc', 1, 1) == substr('abc', 0, 1) in Oracle because it helpfully translates the 0 index to a 1 index. I mean, it's right here in the Oracle docs[1]:
When position is 0 (zero), then it is treated as 1.
My thoughts on it: you can find humor anywhere on the internet. It's much harder to find decent discussion. When humor and decent discussion try to coexist, humor tends to crowd out the discussion.
Why bother posting detailed information about what kind of reaction is going on when you can make a simple joke about nuclear power and get as much karma? Why bother posting something intelligent when it will be buried below puns?
I used to think HN downvoting humor was bad, but I've really come to agree with it after looking at some attempts at serious discussion on reddit. Without fail, by the third comment or so, someone will derail it with a predictable joke or pun. Humor is easy. Interesting comments are not.
Depends on how much you care about latency, right?
The easy solution is just, y'know, write the log to a file and scp it back to some central place every so often. But then you have to either (a) keep track of how much of a file you've copied, which is a pain; or (b) only grab files that you're no longer actively writing to (as determined by naming scheme or something), but that introduces some latency, depending on how often you rotate.
I don't, really (you can use syslog pretty successfully, I guess, if you don't mind it), but UDP has certain advantages over TCP, namely that your code keeps running even if the server you're sending to goes down or is unable to respond to messages, it's faster, etc.
Ideally, I'd use a function that sent things to a small server over UDP, which would then put them in Redis. This assumes you don't mind losing a few lines, of course.
If you continuously send UDP messages to a server that isn't accepting them, you will eventually get an error from the sendto() system call; the uninterested receiving host is generating ICMP messages saying "I don't want these". It's true that there is a flavor of sloppy socket coding where that error doesn't manifest itself. But if you're writing good code, TCP is no less fire-and-forget than UDP.
UDP is marginally faster than TCP, but the tradeoff for that is that under heavy load, UDP imposes more costs on the rest of your traffic. Since we're talking about logging, though: who gives a shit how fast it is? With either transport, if logs are taking more than hundreds of milliseconds to clear, you have a problem you need to fix.
Yikes, losing messages? If anyone designs a system that knowingly makes itself harder to diagnose under stress by destroying evidence, I only hope it's themselves and not their successors who are there to endure the pain.
I'll take a shot at this: I think G+ is possibly the most comprehensive (in terms of design and functionality) new product in the social space right now. Since it comes from Google, who have the history of "revolutionizing" various aspects of the web, hackers want to learn more and put G+ in perspective of existing (competing?) products. This makes it very interesting not only to read about how to fit G+ into the current framework (Twitter-but-longer? LinkedIn-killer? Facebook-killer?) but also follow the minutiae of the growth, adoption, and evolution of G+ (much like they would have taken to GMail, if the timelines had overlapped, imo).
Google employees care because Larry Page said your bonus depends on the success of social.
I care because if it fails, then it will be demoralizing to the company. Google will quickly become the new Microsoft, where investors and potential hires see the company as a stodgy old company that can't innovate. The press will pounce on google for Google+/Wave/Buzz. Android will be seen as the only bright spot, and even that is out shined by Apple.
If it is successful, then it means that other social businesses aren't nearly as valuable as VCs thought they were. If Google+ is successful, no one is going to pay close to $50B for facebook when it comes time to IPO. It will be the end of the social tech bubble.
