Please create an account to participate in the Slashdot moderation system

 



Forgot your password?
typodupeerror

Comment: It's not beside the point (Score 5, Insightful) 141

(Of course, few stopped to think about how Facebook decided what to show people in the first place, but that's beside the point.)

No, it's the entire point. Your stream, to the extent that it's "your" stream, is already manipulated in ways you're not told, based on what Facebook thinks you will find interesting, funny, or engaging enough to come back and see more ads. Experiments have to happen as part of Facebook's desire to expand, if only to see which manipulations mean more ads displayed. The only difference is that now the people who are interested in something closer to science than ad sales at Facebook won't tell us about their results. Are you happy?

Comment: Re:One person a bottleneck doesn't create... (Score 1) 238

by cjc25 (#47071607) Attached to: Google Fiber: No Charge For Peering, No Fast Lanes

Right: TWC sold me "15 Mbps" and Netfix "X Tbps," not me "15 Mbps unless lots of people in your neighborhood are watching Netflix" and Netflix "X Tbps unless you happen to be sending to an individual neighborhood." It's totally plausible that the terms they gave me are very difficult to meet, but that sounds much more like their problem to solve than mine.

If their defense is that they mislead me into thinking they're giving me more than they are ("up to 15 Mbps") I'm not going to be particularly sympathetic to their complaints about the cost of upgrades to handle load.

Comment: Re:1 difference between most, including RH, and Ca (Score 1) 118

by cjc25 (#46257425) Attached to: Why Do You Need License From Canonical To Create Derivatives?
The OP was not talking about trademark, but about copyright on the selection of packages in an argument analagous-ish to one that a mashup can be copyrighted separately from the underlying songs. It's odd and untested but not on it's face definitely wrong under current law, dickishness notwithstanding

Comment: Re:And children of public school cheerleaders (Score 1) 715

by cjc25 (#45941237) Attached to: How Good Are Charter Schools For the Public School System?

The money that goes to support private schools could instead have been taxed and spent on public schools.

I'm pretty sure there's no deduction for a dependent's private school tuition. And while the US tax -> school structure has lots of problems which are fairly off topic, tax deductions wouldn't apply anyway.

Comment: Re:Comparison to Facebook a teensy bit misleading (Score 1) 497

by cjc25 (#45091925) Attached to: Cost of Healthcare.gov: $634 Million — So Far
What makes you think Facebook's architecture doesn't have a similarly complex diagram? Also, it's been more or less admitted that the whole "Eligibility" box was not correctly implemented, which takes a lot of steps out of the chart.

Interop is very difficult when you don't control both ends, you won't find cogent arguments against that, but defining bad interfaces is bad design, not some magical issue that only affects government work. If a bad design results in excess cost, most people would consider that a problem. Within the government, if they can't coordinate enough to redefine interfaces correctly, the issue is a dysfunctional organization.

For interaction with the external insurance agencies, work was probably harder. Supporting disparate systems may have been unavoidable given deadlines. While the PHBs for this project may have said "this is how you get your plan listed on the exchange. If you don't work with our interface, tough." If so, good for them! I've heard too much about how hard it is to work with so many insurance companies to be optimistic about that, however.

Intelligent software design is not optional in any service at scale. The fact that it's hard does not excuse doing it incorrectly at 6x the cost.

Comment: Re:2000 Wyoming (or Montana, or Nebraska) citizens (Score 5, Insightful) 204

by cjc25 (#44777741) Attached to: Humans Choose Friends With Similar DNA

I find this study to be extremely flawed, not to say elitist / racist.

Yes, people who fit a stereotype of those I dislike like to have friends who are similar.

If the study had been conducted with 2000 subjects from places with people like me, I'm sure the results would've been more comforting to me.

FTFY

Comment: Re:Not much of a defense (Score 1) 358

by cjc25 (#44442185) Attached to: NSA Director Defends Surveillance To Unsympathetic Black Hat Crowd

This is seems like a permutation of the Butterfield fallacy

Whichever attack you've decided was the "most visible" was so because it was missed.

Fortunately this doesn't affect arguments regarding the proper scope of surveillance, but unfortunately it underscores that people are often oblivious to their assumptions. In your case, it's that you would have heard of stopped terrorist plots. I'll agree that it's plausible because of the temptation to brag about success, but far from certain.

The tree of research must from time to time be refreshed with the blood of bean counters. -- Alan Kay

Working...