rbranson

49 ms·

511.
by rbranson 15y ago
Is there any reason this isn't more common? Why aren't more US datacenters in the Chicago metro area where the air is dry and cool, and trillions of gallons of frigid water is available? The geography is mild and natural disasters that effe
512.
by rbranson 15y ago
Lack of any kind of arbitration mechanism can spell the death of a site like this. There's definitely a balancing act.
513.
by rbranson 15y ago
I love how successful architectures like this really cut down a business to it's core offering -- which in Evernote's case, is object storage.
514.
by rbranson 15y ago
The price is set based on market analysis. That's what the company thought they could get for it. The transaction requires willing participants.
515.
by rbranson 15y ago
On top of all of that, using mapped I/O effectively means embracing it directly. Adding an abstraction layer on-top that would allow a quick switch to stdio would negate the benefits. Sucks.
516.
by rbranson 15y ago
Definitely agree with you that abstraction matters. I use C as an example because it's the canonical way in which we as developers can reasonably converse about directly dealing with the machine. C is still important and will continue to be
517.
by rbranson 15y ago
There is an madvise() syscall that roughly translates into an fadvise() for memory-mapped regions as well.
518.
by rbranson 15y ago
... and remember guys, the stock price is just what the public is willing to pay for it, and is not directly bound by any fundamentals. It's not LinkedIn's fault that their stock is overvalued.
519.
by rbranson 15y ago
The mechanism suggested by the author seems a bit absurd. Code would to have to pre-warm regions of memory to be available in a non-blocking fashion before accessing it. One of the major attractions to mmap() is that this is unnecessary, th
520.
by rbranson 15y ago
mmap is great, but it often lures developers into thinking that in-memory and on-disk data structures can be reasonably unified. It's still spinning media, and even SSDs are orders of magnitude slower than regular RAM. Optimized data struct
521.
by rbranson 15y ago
This is solid. The team is not going to leave as a whole as much as any business would abandon a healthy (or maybe not so healthy) customer relationship. I also think that people will leave the team for the same reason they'd leave a health
522.
by rbranson 15y ago
This is a pretty solid hypothesis, but wouldn't the same be true of the exploited cached images? They are essentially the same thing: stale resources that won't be used again, and those regions need to be free'd up.
523.
by rbranson 15y ago
Off and on in the car, but that's irrelevant. I'm not claiming every technology is irreplaceable, but that not all evolutions or supposed improvements of technologies REPLACE their predecessors.
524.
by rbranson 15y ago
I've had Ruby processes run wild (think turning off the garbage collector) and grow to 20-30GB on my MacBook and it didn't make my system unusable. Things slowed down, but I could pretty easily open up the Activity Monitor and kill the offe
525.
by rbranson 15y ago
This makes sense for servers, because they shouldn't really have any significant amount of virtual memory space tied up by non-resident allocations. On a desktop OS, the user has a much larger breadth of software loaded and is only using a
526.
by rbranson 15y ago
There's a light at the end of the tunnel. It's likely whatever energy source for the simulation would be disrupted by some catastrophic event. Even without sentient intervention, over tens of millions of years you've got starvation of some
527.
by rbranson 15y ago
This will replace the terminal as much as the microwave replaced metal over fire for cooking or as much as synthesizers replaced real instruments in music.
528.
by rbranson 15y ago
The 75mph isn't really the hardest part of that. Really, it's indoor coverage with dirt cheap consumer gear that's hard, followed closely by call and data connection handoff. Even if it's free, there's a minimum amount of service quality to
529.
by rbranson 15y ago
It's already hard enough to run a reliable mobile data network built with industrial-grade systems on exclusive spectrum. Unfortunately, moving to a network like you describe, built on commodity gear doesn't translate to quality, redundancy
530.
by rbranson 15y ago
It reminds me of disturbingly common ideological standards at corporations that require all in-house software to be written in the-one-true-and-blessed-language, under the guise of a short sighted notion that this directly translates into u
531.
by rbranson 15y ago
You seem highly optimistic about the JVM's appropriateness here. In reality what the browser needs is something more like the Dalvik VM, which is designed for compactness, efficiency, and multi-process models.
532.
by rbranson 15y ago
If you're just suggesting that we need to ship a common bytecode to browsers, then I agree.
533.
by rbranson 15y ago
You're always going to feel hobbled by a standards-based, compatibility-driven platform like the web where you must maintain compatibility with fairly ancient software. This is just the way things are. If it wasn't the web, it'd be whatever
534.
by rbranson 15y ago
While I don't think the JVM is the savior, applets are a far cry from a good example. Applets usually cause the browser to lazy load the entire JVM as a plug-in, from scratch. Very little effort has been put forth to make this process fast.
535.
by rbranson 15y ago
Seriously? the JVM? You listed the pros, but what about all the cons? Modern JVMs are fast, but highly specialized for long-running applications, which web pages are decidedly not. That's just the tip of the iceberg.
536.
by rbranson 15y ago
I love Scala. So much so that I've wrung my brain trying to find places where it'd really make sense in my toolkit. I keep coming up empty though. I work primarily in Rails on the front end and if I need some "boom-boom" on the backend, gen
537.
by rbranson 15y ago
These are just potentially poor decisions made by people with lots of money. Tech startups are inherently risky -- many more succeed than fail -- even if they are well funded. It is a fundamentally different equation than starting a bagel s
538.
by rbranson 15y ago
Uh, I know sometimes a study is needed to actually bring something to scientific standards, but isn't this sort of thing just obvious? This is what the energy business is all about, creating more output than input.
539.
by rbranson 15y ago
^^^ This. The big Internet companies should take this as a clue and start campaigns to inform users on Internet privacy and actively push for more user friendly privacy management.
540.
by rbranson 15y ago
Same. What about all those people still on 1.5-6mbit DSL, or... shudder dialup?

More ›