Forgot your password?
typodupeerror

Comment: Re:Small-scale, real-time. (Score 1) 502

by hattig (#47612393) Attached to: Why Morgan Stanley Is Betting That Tesla Will Kill Your Power Company

Which is why Solar is working so well in Germany. It doesn't need to be sunny for solar panels to operate well.

The problem is that the grid gets loads of solar power during the day, but peak usage hours are later. There are only so many reservoirs they can pump up mountains with the spare power during the day, and offices that need A/C. Of course hopefully everyone will be charging their electric cars during the day so maybe in the long run it will even out.

Comment: Re:That's not a Exynos SoC (Score 1) 47

by hattig (#47572773) Attached to: Raspberry Pi-Compatible Development Board Released

The article even makes the clear. How the submitter misread that is beyond me.

The advantages of this board: Smaller. eMMC connector. ADC. RTC. Better power management. Small LCD module option ($30 incl. usb hub and ethernet).
Disadvantages: USB/Ethernet on a different board ($20), connectors sold separately ($4).

The RPi B+ resolves the power management issue.

The advantage of being compatible is that the software support for the RPi is actually very good, and there is a massive community.

Comment: Re:When I can play Asteroids on the back of a box. (Score 1) 78

by hattig (#47521179) Attached to: Researchers Print Electronic Memory On Paper

It does seem that they nearly have all the ingredients to make a viable 8-bit computer on a (small) sheet of paper now. I guess an Atari 2600 could fit in a fairly small area with it's 128 bytes of RAM (1 cm^2) and other simple logic. This printed RAM access speed isn't great though - 200us is three orders of magnitude too slow compared to even the memory in those old computers. Hopefully shrinking these RAM dots will also improve speed.

Comment: Re:Who? (Score 1) 245

In addition Inmarsat can surely just also correlate pings from other aircraft with their actual known position to verify their algorithms they have come up with are valid for the MH370 situation where they only have the pings.

I mean, they did do some basic validation like that, right?

In addition I strongly doubt this is the only person to have double checked on the mathematics used, but he's the only one saying its wrong.

Comment: Re:Rewarding the bullies... (Score 5, Insightful) 798

Most likely the bullies are members of the school's football team, hence the protection afforded to them.

So option three, but targeting the Achilles heel tendon or other sport-critical tendon/muscle, is a great option, in my opinion.

Clearly the school has a bullying problem, and a control problem. It's a sick, diseased school run by weak people, and teachers too afraid to do their job to protect students from bullies who are on the school football team. This is something that requires state intervention, I presume the state has school inspection bodies, and the ability to enact punishments? I would suggest a ban in intra-school sporting competition for a couple of years until the school's curriculum has moved back towards education.

Indeed, I think that US school sports is really weirdly venerated. I'd split the two up, schools can have basic sport, but clubs, etc, should be run outside of the school, maybe with loose affiliation, but having no influence on the school's central reason for existence - education.

Comment: Re:Sure, but... (Score 1) 392

by hattig (#46681715) Attached to: How Many People Does It Take To Colonize Another Star System?

Even better, you could probably find a million people willing to pay money to have their DNA stored for use on future colony planets. That could help fund this space programme, a little.

The cost and availability of storage for these genomes isn't really the issue. The automated human-building factories are an issue (synthetic wombs, etc, will probably be a solved issue by the time we can build a 300-year lifespan spacecraft carrying hundreds of people - the surrogate mother issue will probably not be an issue), and the subsequent raising of these children will require real human contact - hence the point of this article which is getting those humans to the end of the journey in a reliable manner (those humans will also monitor and repair the spacecraft during the voyage, even if automated processes are good enough for most problems).

Comment: Re:Sure, but... (Score 1) 392

by hattig (#46681675) Attached to: How Many People Does It Take To Colonize Another Star System?

I believe that was meant to be around 80,000 years ago.

That's a pretty serious long-term colonisation programme. I wouldn't want HP building the computers they needed to last through this process.

A large portion of any spaceship is going to be storage for end-of-journey supplies, and fabrication robots (that fabricate bigger fabricators, etc, until you have colony-builders).

Small villages traded with other small villages, they didn't exist in a vacuum. These people will, however many exabytes of media you supply them with to keep themselves occupied. 150 people have a high risk, on a 300 year journey, of being wiped out, or decimated to a level they can't recover from. 500 people is better, 2000 is nearly okay, 10,000 is basically great, 40,000 is excellent. A convoy of ships is better than a single ship (which makes the single giant asteroid spaceship idea a bit less desirable).

Comment: Re:Sure, but... (Score 1) 392

by hattig (#46681623) Attached to: How Many People Does It Take To Colonize Another Star System?

Who's going to look after the embryo children if the generational population died out, or are crippled by genetic defects?

That's why you need on-ship genetic variation. The on-ship population needs to survive, year after year, century after century.

Sure, the idea that the ship also hatches embryos en-route in case of gaps is fine, except for the catastrophe case (decimation of the population means the population is too weak or non-existent to even raise the in-ship hatchlings).

It seems that 2000 people is a strong enough population, especially if spread over several different spaceships travelling in convoy (which also helps with the Dunbar's number issue). But 150, or 500, is not enough.

Or you wait long enough until spaceships are fast enough to allow the people who want to populate another planet to be the ones who arrive there and populate it. Advances in cryogenic sleep may allow this.

Comment: Re:Sure, but... (Score 1) 392

by hattig (#46681561) Attached to: How Many People Does It Take To Colonize Another Star System?

figure out how to introduce additional genetic diversity once you arrive and establish yourselves.

Keeping to the spirit of the original article (i.e., ignoring embyros and overloading the workload of the colonists to also raise loads of children on the side, rather than just a few children), that could mean staggering the launch of each of the ships, rather than launching together.

By the time the second ship arrives, the crew of the first ship are either all dead (and their roles fall to the second ship), or the first ship has prepared the optimal landing site with all the facilities, cleared the lands for the orchards (which will be mature by the time the second ship arrives) and agricultural needs, and have enough information on living on the new planet to get the new colonisers up to speed rapidly.

Again, human nature is going to be a major problem. A close knit community on the ground, say for 10, or even 20 years, is probably not going to welcome the next spaceship load of humans. But they would like the tools and seeds and animals and facilities they have. There is a strong chance that instead of one large colony, you'd end up with multiple, small, separated colonies. Maybe close enough to trade and therefore amalgamate one day...

Comment: Re:Sure, but... (Score 1) 392

by hattig (#46681529) Attached to: How Many People Does It Take To Colonize Another Star System?

And if you don't think that 150 humans (however advanced and brilliant their on-ship education and upbringing is) aren't going to be selfish and decide that they would rather raise their own children than 10,000 "other peoples" children then you've got another thing coming. It's not like anyone is going to be able to punish them for making a decision like this either.

Disregarding that, these 150 people will be spread over a range of ages and capabilities too. So maybe there will be 50 couples capable of raising an additional child every three years, for the rest of their lives. I.e., you can probably hatch up to 20 additional children a year from the 10,000 at most. Which isn't an ideal rate, I would assume that the extra genetic diversity such a small injection provides would be lost. That's if they're not excluded from society as being "others", blah blah human nature, etc. Of course the population would increase rapidly, so the rate of hatching could increase too.

But then you need people to actually build the civilisation on the new planet. Agriculture, Buildings, Defences. I'm sure the initial landing fleet (for humans and end-of-journey world-building supplies) would suffice initially, but even so a sizeable portion of the population is going to be engaged in colonisation, not parenthood.

Maybe 2000 people + 10,000 embryos would work better. Keeping 2000 people entertained on their asteroid based spaceship is going to be fun.

Comment: Re:That's it (Score 2) 243

by hattig (#46620241) Attached to: Dropbox's New Policy of Scanning Files For DMCA Issues

Or you could read the article and get answers immediately.

They use file hashes of previous DMCA requests when new files are shared. If it transgresses, it's blocked just like this situation.

It's not "policing", it's blacklisting the sharing of specific files via comparing file's hash against a list of blacklisted hashes.

I just hope they're not using CRC16.

Comment: Re:Muh freedoms! (Score 1) 230

by hattig (#46619383) Attached to: Geologists Warned of Washington State Mudslides For Decades

You would have thought that insurance companies would have been quite quick to refuse to cover the properties because of the risk. "We'll have to apply a landslide exemption to this cover" should ring alarm bells, just as "We won't cover flood damage" should be taken to mean "we expect your house to flood" because it's built on a flood plain. Yet people still buy houses on flood plains and then complain bitterly to the media when their house floods.

Comment: Re:CSS variables? (Score 1) 256

by hattig (#46541779) Attached to: Firefox 29 Beta Arrives With UI Overhaul And CSS3 Variables

I don't understand why they didn't just add a "variables" CSS key.

e.g.:

div {
        variables: {
                color: #000000;
        };
        padding: 0px;
}

div.header {
        color: var(color);
}

TBH I don't even think they need to be "variables", just "cascading named constants". And ultimately, because CSS hasn't supported such a feature before, even though back in the 90s people were saying how nice it would be to have such a feature, we have various pre-processors/template driven CSS/in-house solutions.

Comment: Re:Specialism (Score 1) 306

by hattig (#46514415) Attached to: Ask Slashdot: Can an Old Programmer Learn New Tricks?

Yes, hyper-specialism is a problem, as is unwillingness to learn new things. Those Flash devs will not go far unless they push their boundaries.

SVN is not hard. Git is not hard, but a little more complex up front than svn checkout/update/commit. Any intelligent person can get their heads around these in a couple of hours, especially when someone is willing to help them or sit with them through a sample session (an essential part of mentoring).

Comment: Re:what you need them for? (Score 5, Informative) 306

by hattig (#46514399) Attached to: Ask Slashdot: Can an Old Programmer Learn New Tricks?

And which framework was that? If it's targeted at "enterprise" use, then speed and efficiency won't be one of its core features, not with runtime annotation processing...

For example, Apache Wicket is a gross bloated thing to avoid the "horror" of learning how to program a web UI in JS that communicates with the backend server using sane RESTful APIs. OTOH it saves you from writing those APIs and keeps your codebase in a single object oriented language.

Hibernate is a gross-but-cool thing that saves the developer from touching JDBC. It's overhead pales in comparison with the network latency/RTT and database effort though, and it allows the programmer to again do database operations at a decent OO high level. Personally, I prefer JDBC but that can end up with a lot of boilerplate code to do simple operations. But OTOH you could end up with dodgy DB code, failure to try/catch/finally properly, etc. HQL can DIAF.

And Spring ... Spring does everything. Dependency injection is a major advantage (until you use it, you might wonder why your "EntityManager" class is not good enough), interceptors, etc. Ignore the MVC crap, that's old hat.

And tooling is another thing. Maven is essential for the Java developer today. Until I used it, I was happy with Ant and manually updating dependencies. Selenium is an essential web UI integration test tool too. Anything that makes testing, integration testing, etc, easier should be welcomed with open arms. Team-based development is a recipe for breaking code contracts in multiple places.

There are a lot of new tricks that a programmer that has stayed in a comfortable role for a long time could have missed, and find problems when looking for a new job. Luckily, a good C programmer is unlikely to be applying for Java roles, and roles are often now in the embedded marketplace where frameworks are less common over raw C with common libraries.

And there will be plenty of people that disagree with everything I've written. The joy of programming, eh?!

It is not best to swap horses while crossing the river. -- Abraham Lincoln

Working...