Posts by Keith Ng

Last ←Newer Page 1 2 3 4 5 Older→ First

  • OnPoint: Ich bin ein Cyberpunk,

    General comment: Baby steps, folks. Tor, Truecrypt, everything else will come.

    Auckland • Since Nov 2006 • 530 posts Report Reply

  • OnPoint: Ich bin ein Cyberpunk, in reply to Thomas Beagle,

    I guess the first part of the article nicely disposes of ex-GCSB head Bruce Ferguson's claim that the GCSB only provided assistance to Police by seconding staff to them.

    Hang on though - this was not quite the same, in that they believed (genuinely or not) that Kim Dotcom was a foreign national, and therefore there was no need for a firewall between them and the investigation, which is what the secondment process is designed for.

    Auckland • Since Nov 2006 • 530 posts Report Reply

  • OnPoint: Ich bin ein Cyberpunk, in reply to B Jones,

    Maybe this is a silly question, but given that I understand encryption that can't be cracked by governments isn't supposed to be publicly available (limit to number of keys) - isn't there a risk with encryption that using it attracts attention in and of itself, and that with small volumes of encrypted traffic, agencies can comfortably handle the processing power to crack it?

    a) I don't think this is true yet, and b) this is why we need to increase the volume of encrypted traffic.

    I've assumed for years that everything I put on the internet, including email, can be harvested, it's just that nobody cares enough to do it provided I don't stick obvious keywords in there, or correspond with known agitators, or have offline activities that might draw attention. Security by obscurity.

    That's not security by obscurity, that's just obscurity.

    Of course, if enough people use encryption, the appearance of secrecy doesn't stand out as noteworthy. Another thought - a really good code doesn't look like a code.

    Steganography!

    Auckland • Since Nov 2006 • 530 posts Report Reply

  • OnPoint: Ich bin ein Cyberpunk, in reply to Moz,

    Moz & Rich: Yes, will get to that in the next part...

    Auckland • Since Nov 2006 • 530 posts Report Reply

  • OnPoint: Ich bin ein Cyberpunk, in reply to HamishFraser,

    Huzzah! It worked. I'd reply in an encrypted message, but I don't have your public key.

    Auckland • Since Nov 2006 • 530 posts Report Reply

  • OnPoint: Government Portfolios for Dummies, in reply to B Jones,

    People making the important decisions in policy ministries don't realise its importance because they're steeped in humanities rather than information science culture, and are mostly older and less likely to have picked up information science through having fun with computers.

    That's why I reckon the best way forward might be to just make stuff like this, to demonstrate that it takes one guy, zero dollars and most of an afternoon to build a searchable all-of-govt (well, most-of-) archive.

    Auckland • Since Nov 2006 • 530 posts Report Reply

  • OnPoint: Government Portfolios for Dummies, in reply to LeslieGoodliffe,

    Is this service likely to be any more reliable than the individual Government departments?

    It looks more haphazard that my own efforts...

    http://nzlc.natlib.govt.nz/vwebv/search?searchArg=briefing+for+incoming+ministers&searchCode=GKEY%5E*&limitTo=&recCount=25&searchType=1&page.search.search.button=Search

    Auckland • Since Nov 2006 • 530 posts Report Reply

  • OnPoint: Quickfisk: Youth Unemployment, in reply to Matthew Hooton,

    Also, I think this is why I did the graph. Here I am faffing about whether the Q1 dip is real, or whether it's a bloody artefact of the seasonal adjustment techniques caused by the the Q4 spikes, when really, none of this is consequential.

    The graph shows, clear as day, that youth unemployment went through the roof post GFC and hasn't budged since. If this decline turns out to be real 6 months down the line, it will show that as well.

    Auckland • Since Nov 2006 • 530 posts Report Reply

  • OnPoint: Quickfisk: Youth Unemployment, in reply to Matthew Hooton,

    These must also be HORRIBlE RIGHTWING LIES from Statistics New Zealand and the OECD.

    Jokes! Just being facetious. For the record, I'm actually pro youth rates, and on the fence about 90-days in principle, but just see no evidence it's working.

    Look at unemployment rates by age and gender and look at the 15-24 age group, all persons. At 15.1% in Q1 2013, we’re now below the OECD average and falling.

    Are we talking about the same 90-days trial period? The one from 3-4 years ago? Because if youth unemployment hasn't moved for three years, and it just dropped for one quarter, it's astounding that you'd credit it - any of it, even as a contributing factor - to the 90-days trial. (Did you? Or is that just a HORRIBLE SOCIALIST LIE?)

    Also, it's a single data point. If we get two more quarters <16%, I will happily concede that youth unemployment is decreasing (still won't credit it to 90-days, for reasons above). Seriously - put it in your diary, I will eat my metaphorical hat and do it with a smile.

    In the year to March 2013, there was a large fall in unemployment for people aged 15–24 years (down 10,500). This fall can be largely attributed to a decrease in unemployed 20–24-year-olds (down 11,200). This was an atypical fall in unemployment, as the number of people unemployed for this age group usually increases during March quarters.

    And in the previous quarter, we had the highest youth unemployment rate since the 90s, and that was in a December quarter, which usually isn't the peak. Was that a sign that things were going very badly? This is why you don't look at single data points in isolation.

    My hunch is that the temporary shift in the peak has messed with the OECD's seasonal adjustments, which is why they're reporting a much larger faller than StatsNZ. Just a hunch though.

    Behind this was a rise in the number of young people outside the labour force who are studying (up 25,000). The number of both 15–19-year-olds and 20–24-year-olds in study rose – up 16,200 and 8,800 respectively.

    Correspondingly, you can see the *decrease* in the employment rates. More youth studying is definitely not a bad thing, but it's also not a sign of the labour market doing well, nor a sign that policies like 90-days trial are working.

    NEET rate declines

    Still not convinced that NEET is the right measure for measuring the success of 90-days trial. But FWIW, they behaved in the same way as youth unemployment - spiked in December, came back down in March, no long-term decline that we can see (yet).

    Auckland • Since Nov 2006 • 530 posts Report Reply

  • OnPoint: Quickfisk: Youth Unemployment, in reply to Brent Jackson,

    a reducing ratio may be (relatively) good for young adults, but it is bad for older adults.

    Yeah - that's a particularly odd measure. It's definitely useful for distinguishing between youth-specific unemployment problems and general unemployment problems, but it definitely needs a lot of context or it's misleading.

    You're right about the "higher = good" next to "higher = bad" thing as well. I actually got it wrong on my first read, and it took a week while for it to click in my head. I don't think it's deliberate though - I think it's just how old-school data nerds roll.

    Auckland • Since Nov 2006 • 530 posts Report Reply

Last ←Newer Page 1 2 3 4 5 53 Older→ First