neilc

58 ms·

511.
by neilc 17y ago
I still cringe at the thought of paying money for a device that solely allows me to read. I don't see how that's any different than paying for a device that solely allows you to listen to music. It's like building a device that allows me
512.
by neilc 17y ago
Everybody was saying that the new Kindle was terribly important—that it was an alpenhorn blast of post-Gutenbergian revalorization. Ugh, what terrible writing.
513.
by neilc 17y ago
In which case, there is no shortage of analysis available a mere Google search away.
514.
by neilc 17y ago
I thought that was pretty common knowledge. Is it a bad state of affairs?
515.
The Pathologies of Big Data(cacm.acm.org)
2 points by neilc 17y ago|0 comments
516.
by neilc 17y ago
Given how enormously complicated C++98 is, let alone C++0x, I think the last thing C++ needs is to jettison the standardization process in pursuit of faster evolution. The evolution of C++ should be conservative: both because it is widely d
517.
by neilc 17y ago
I think that preference is exactly why you're not a venture capital firm.
518.
Parallelizing the Web Browser [pdf](usenix.org)
1 points by neilc 17y ago|0 comments
519.
by neilc 17y ago
It sort of does - if the money came from existing investors in a follow-on round. It didn't: "The money for this fifth Series E round comes from a Silicon Valley venture firm, Lightspeed Venture Partners ... [there was] interest by Ning i
520.
by neilc 17y ago
750mm sounds crazy until you read that they have raised 119mm so far... thats only a 5.5x ROI if sold at that valuation. Perhaps, but that doesn't make $750 million any less crazy.
521.
by neilc 17y ago
Nevertheless, I have many gripes with BDB which I'll write down some day soon I'm considering BDB as the data storage layer for a new project, so I'd be very curious to hear your experiences with it.
522.
by neilc 17y ago
The "$50 per ounce" is a hypothetical tax on legal marijuana, not the hypothetical price of legal marijuana.
523.
by neilc 17y ago
I think the parent comment was a joke -- "ought to" is perfectly fine usage in that situation, I believe.
524.
States Seek to Jam Prison Cellphone Signals(nytimes.com)
1 points by neilc 17y ago|0 comments
525.
by neilc 17y ago
you don't need to rehash everything whenever your data set grows beyond some size Not necessarily true; linear hashing avoids this requirement, for example. http://en.wikipedia.org/wiki/Linear_hashing
526.
by neilc 17y ago
A hash table isn't much use for queries that return multiple rows, or use aggregate functions. Neither of those claims is true. A hash index is perfectly useful for queries that return multiple rows: the index just stores all the duplicat
527.
by neilc 17y ago
Mainly, though, each b+-tree node is the same size as a disk I/O block. The disk is often the slowest part of the database, so making as few reads as possible will go a long way. If you can't work entirely in memory, that's the next best t
528.
by neilc 17y ago
10% selectivity is the minimum selectivity necessary for a b-tree index to be helpful. That's probably on the high side. Suppose that records are stored in 8KB pages (which is probably too small for a modern system, but nevermind), the se
529.
by neilc 17y ago
Only now, in hindsight, he realises it annoys him that of the great many users, not only do none credit him, but some even blame him. That's pretty harsh. Is it? How many people using Linux know who, say, Ulrich Drepper or Alan Cox is? Ho
530.
by neilc 17y ago
Rumour has it that Google wants to ditch X. If that is the case, no other GUIed nix software will run without a rewrite.* Not exactly: few modern GUI applications are written to use Xlib directly. If Google ditched X for a new window syst
531.
Does Anyone Use Chrome?(avc.com)
2 points by neilc 17y ago|0 comments
532.
Why Hulu Succeeded As Other Video Sites Failed(bits.blogs.nytimes.com)
1 points by neilc 17y ago|0 comments
533.
Fork-join style parallel programming in Java 7(ibm.com)
2 points by neilc 17y ago|0 comments
534.
by neilc 17y ago
I think its unrealistic to build a general web-based solution Maybe not a perfectly "general" solution, but there are lots of domains where data is analyzed that would benefit from a powerful web-based analytic environment. these dataset
535.
Latency Lags Bandwidth, and What It Means For Computing (David Patterson)(ece.eng.wayne.edu)
2 points by neilc 17y ago|0 comments
536.
by neilc 17y ago
Who cares? Government agency has buggy web-accessible implementation of SHA1. Not notable, and a link-bait title to boot.
537.
The Soft Heap: An Approximate Priority Queue with Optimal Error Rate(papersincomputerscience.org)
2 points by neilc 17y ago|0 comments
538.
ACID versus BASE for database transactions(johndcook.com)
11 points by neilc 17y ago|0 comments
539.
by neilc 17y ago
Encrypting and authenticating all Internet packets by adding application-level crypto to each and every protocol in use seems like a Sisyphean task to me. IPSec is more ambitious because it is lower in the protocol stack, but if you really
540.
by neilc 17y ago
At or below the median, that is.

More ›