Forgot your password?
typodupeerror

Comment: Re:The press and the people... (Score 1) 228

by lakeland (#45825599) Attached to: USA Today Names Edward Snowden Tech Person of the Year

Ok, so it had a few cliches in it.

But it seems to me that the GP has a point. I'm not old enough to know what it was like during say McCarthy's witch hunts and how many people stood up for what is right. Was it a very small minority while most people just went about their lives, or was there wide public dissension? I do remember in the early 80s that there was fairly active opposition to Apartheid rather than the current defeatist apathy. That was a bit different though because most countries had abandoned it and it was more a matter of imposing our views on an unwilling minority than changing a more powerful group. The largest protest I can remember from recent history is the 99% movement and even that seemed to largely die out after a few weeks.

More succinctly: "Is the general populace's apathy to the issues of the day unique to our generation, or is this normal?"

PS: Cow towing rather than Kowtowing, though since it's from Cantonese I am not sure you could call anything a misspelling.

Comment: Re:how would it work in the real world? (Score 1) 308

by lakeland (#45654733) Attached to: Google's Plan To Kill the Corporate Network

I agree the cost of the computer is effectively a rounding error, but there are non-trivial costs in Window's favour too relating to compatibility.

It is getting a lot better with the rising popularity of Android / iOS meaning that fewer companies target a single platform, but I still find that when I try and take just my mac that I often find I have trouble doing some small thing.

Comment: Re:Why isn't all medical equipment open source? (Score 2) 134

by lakeland (#45450349) Attached to: 12-Lead Clinical ECG Design Open Sourced; Supports Tablets, Too

It's hardly fair to compare a doctor's take home pay to the median income in the US. People going to med school are both smart and hard working, try comparing the income of smart, hard-working non-doctors to doctors and I think you'll find the difference is far smaller.

Yeah, $80k after applicable deductions is a good income. Approximately 12% of Americans make that much or more.

I tried to find some data on the median income of people with similar work ethic and IQ to doctors but couldn't find anything sorry.

Comment: Re:Why do transit smartcards need to be hard? (Score 1) 96

by lakeland (#45385829) Attached to: New Zealand's Hackable Transport Card Grants Free Bus Rides

Because it is hard.

(Disclaimer: I used to work for a company which bid unsuccessfully a few years ago to fix up the Christchurch system)

Probably the hardest part is the decentralised nature. How much money do you have out there? If this card claims to have been topped up by a terminal but you have no record of that, either the terminal is slow at reporting back, or the card is lying. By the time you know, it's too late. We have no way of communicating with a card except when it happens to be brought on bus, and at that point we don't have an internet connection.

Second hardest is probably balancing trust with flexibility. We want to enable internet top-up, but how do we get the money from an authorised transaction to your card without forcing you to buy a USB to NFC adaptor? We don't want to trust the card, or at least we want to test it for hacks. We don't want to trust the terminal, a single break-in there could cost a lot. And most of all we need to be constantly worried about primary keys - a break-in to signing keys would destroy everything.

Third hardest is the cheap hardware. Customers expect to get bus cards for free, which means you can't afford more than about $5 per card. Also with a couple terminals per bus you need perhaps a thousand terminals in christchurch - many of which will only be used a few thousand times in their lifetime and so they can't be expensive either.

Fourth hardest is probably the response time. You have roughly 200ms from card presence to approve/reject. That is not enough time for complicated checking - it is enough to check the has of a card number against a blacklist, or to run a challenge response protocol, but that's about it.

That's just off the top of my head and gives a rough overview. These are largely solved problems, but I can understand why a place like Christchurch with a population of 350k, many of whom don't use PT, would elect to stick with a broken system until a national standard is rolled out.

Comment: Re:Do you work at Microsoft? (Score 2) 433

by lakeland (#42421413) Attached to: Ask Slashdot: CS Degree While Working Full Time?

Most JDs I know require a degree in a relevant field. It's used by HR as a quick filter to avoid wasting time screening out woefully underqualified people.

So what the submitter says makes some sense. If you're dealing with the person who you'll be reporting to then they'll be far more interested in relevant experience. But occasionally you have to get through HR filters too - larger companies require them for major promotions, and he might need to leave his current job if his boss gets replaced with a raving lunatic. Either way it would be very much in his interest to have enough formal education to get past the filter.

That said, I wouldn't pick CS personally. He's already a programmer and not likely to want to get into theoretical CS - there's not much demand for knowledge of asymptotic complexity in a typical programmer's job. I think a technical college or a finance / business degree would be better.

Comment: Re:Promoting Synergistic Synergy (Score 1) 115

by lakeland (#41726239) Attached to: Pols Blur Line Between Data Mining, Cyberstalking

People pad their CVs and attach fancy words to fairly mundane things but I think in your dismissal you're missing the importance of what he is saying.

* Text analysis means he uses programs to read what people type and at least at some level infer meaning, even if it's as simple as relevant and/or sentiment.
* social network analysis means he knows who has a lot of influence so you can concentrate on them.
* web personalisation means everybody who visits the website he creates is given a different version
* Computational advertising means the decisions around ad selection are based on ROI calculations and performance rather than a marketer's gut feel.
* Online experiments is similar and just means he tries different things in parallel to see what works best.

Perhaps more interesting is what is omitted. There is no mention of semantics or indication that the text analysis is beyond superficial, nothing around text generation/synthesis to show any attempt at creating a dialogue, and no conversation planning to indicate he is going to be having a conversation. I'd describe it as very similar to traditional advertising but tuned to the individual.

13. ... r-q1

Working...