jbellis

27 ms·

361.
by jbellis 12y ago
To a non-expert, it looks like studies on prayer are a mixed bag. This meta-analysis found no effect: http://link.springer.com/article/10.1207/s15324796abm3201_3#... But this one did: http://rsw.sagepu
362.
by jbellis 12y ago
Apple finally admitted that bigger phones are better. Looks like Google is finally admitting that 4:3 is the right tablet aspect ratio. I'm kind of curious what the secret sauce is that makes Google think they are ready to sell a premi
363.
by jbellis 12y ago
Here are the slides from Netflix architect Christos Kalantzis 's talk, "Eventual consistency != hopeful consistency:" http://www.slideshare.net/planetcassandra/c-summit-2013-even... Video: https:/&
364.
by jbellis 12y ago
Right. It's a confusing space because even though couch/riak/mongo are all "document databases" their storage engines are very different. You can't usefully generalize from couch's append-only b-tree to
365.
by jbellis 12y ago
Could be, but that article just says that the update is "in place." Which may mean that just the changed int is written, or it may be the entire document.
366.
by jbellis 12y ago
How sure are you that mongodb doesn't rewrite the entire document with each update? I couldn't find a definitive answer without source diving.
367.
by jbellis 12y ago
I own an Android Wear watch, and the only thing it has in common with Android for phones is the name. "Google says you can shrink smartphone UX to a smartwatch" is a gross misrepresentation.
368.
by jbellis 12y ago
This is running Cassandra 1.2.x which is over 18 months old. Here's some performance notes on the latest (2.1rc4): http://www.datastax.com/dev/blog/cassandra-2-1-now-over-50-f...
369.
by jbellis 12y ago
I'm not as dogmatic as this could sound, but there's some truth to the position that if you're in Asia buying groceries resembling what you'd get in the US, you're doing it wrong. :) (Seriously, native-style meals a
370.
Cassandra 2.1: now over 50% faster(datastax.com)
26 points by jbellis 12y ago|3 comments
371.
by jbellis 12y ago
One of the more interesting discussions about this that I've seen: http://www.reddit.com/r/IAmA/comments/azgs6/iama_guy_who_so/...
372.
by jbellis 12y ago
Some interesting thoughts on better ideas here: http://bleedingheartlibertarians.com/2014/06/why-living-on-t...
373.
by jbellis 12y ago
Of course "limit by time" and "limit by data" are interchangeable; in practice "limit by time" is easier for operators to reason about.
374.
by jbellis 12y ago
Cassandra cannot lose more than 10s of data because it will slow down writes as necessary to make sure it does not. That's why comparing to a system that allows arbitrarily high data loss is unfair. People can and do run Cassandra w
375.
by jbellis 12y ago
(That was a response to a different benchmark, also done by Thumbtack, but it looks like it applies just as well to this one. See my comment at https://news.ycombinator.com/item?id=7944226. )
376.
by jbellis 12y ago
Of course they are not durable until fsync; that is why having an upper bound on how long that will be is important. Out of the box Cassandra defaults to an upper bound of 10s. Couchbase defaults to no upper bound at all -- you can lose ar
377.
by jbellis 12y ago
Token aware is the default now, but Thumbtack benchmarked with the two year old Thrift client instead.
378.
by jbellis 12y ago
Given that Couchbase is very aware of my criticism of the last time Thumbtack benchmarked C* vs Couchbase, it's significant that no mention is made of efforts to have both systems do durable writes. It looks like they made the same mi
379.
by jbellis 12y ago
This was the toughest part to make clear as well as brief. I was not entirely successful. For Cassandra I used the default 50ms batch window. I'm not aware of tunable windows for the others.
380.
by jbellis 12y ago
They are all ops/s.
381.
by jbellis 12y ago
The very big difference is that child-bakes-in-the-car heat deaths are not because Dad left the kids in the car while he ran in to the store for a few minutes, but because Dad forgot the kid was in the car at all. http://www.pare
382.
by jbellis 12y ago
Additionally, in 2013 about 12% of Tesla's revenue came from emissions credits rather than selling cars. That's going to do funny things to marginal profit numbers per car.
383.
by jbellis 12y ago
Why do you say that GlusterFS is not a parallel file system? Certainly it is by the common definition of "file system that spreads data across multiple storage nodes."
384.
by jbellis 13y ago
The incentives in unemployment benefits mitigate against relocation: http://www.washingtonpost.com/opinions/unemployment-benefits...
385.
by jbellis 13y ago
Obsolete advice. "Anchoring research helps clarify the question of whether to make the first offer in a negotiation: by making the first offer, you will anchor the negotiation in your favor. In fact, Mussweiler and I have shown that ma
386.
by jbellis 13y ago
ITT: people who have no idea what board members do.
387.
by jbellis 13y ago
To belabor the obvious: Time Machine saves old versions for you even if you don't have a separate backup disk, subject to available disk space.
388.
by jbellis 13y ago
I don't understand how that undermines the Libertarian point that if a government has no power to make the playing field uneven, it doesn't matter if it's corrupt.
389.
by jbellis 13y ago
Libertarians would say that it's the traditional liberal response (try to make it hard to purchase influence) that is naive: you've only been trying for a hundred years, and it hasn't worked yet.
390.
by jbellis 13y ago
No, the libertarian political philosophy is that incentives matter, so that if companies can make easy money by getting the government to play favorites (primarily by regulating smaller competitors out of the market, as well as straightfo

More ›