The implications of the technology are a lot bigger than laptop supplies and I am sure they know that. Laptop supplies seem like a good way to raise some cash without dilution while they work on harder problems.
Link-time optimization is not a very descriptive term. It's usually called whole-program or interprocedural optimization. Link-time optimization is just how it has been implemented.
It uses a lot of memory because it requires keeping a representation of more or less the entire program in memory at one time, in a format which is amenable to analysis. It is not out of the ordinary for an optimizer's internal representation to be on the order of 1000x the size of the source code.
I have a theory about zero tolerance or other types of blanket policies such as mandatory minimum sentencing. They are a way for the management class to divest liability or responsibility for decisions. If there is no room for discretion, there is no real decision to make, and no way to be wrong.
You see this sort of thing in large organizations of all types but government is the one everyone has experience with.
> They are a way for the management class to divest liability or responsibility for decisions
I'd also say they are a way for bureaucrats to blame the victim. Someone's complaining that they were attacked? Kick them out of school for being in a fight. That will shut up the people who want to fix things.
I don't really know that that's a "theory". That's just how you scale, isn't it? "Common sense" works for a company of 25 people when the founder is directly involved with hiring and training of every person. It fails though when you employ 50,000 and the CEO turns over four times in a decade. Putting in "policies" and requiring adherence to these policies is a way to keep the train moving even when the CEO is fired/arrested/etc. "Just follow the SOP and you'll be fine."
I don't fully understand how the decision was made which neighborhoods will be built out. It is obviously not based on demand because the coverage areas  include every single one of Seattle's poorest neighborhoods and virtually none of the richest. Even if you don't know Seattle neighborhoods, there are obvious clues in the map. It seems almost gerrymandered to avoid including non-industrial waterfront. Some of those choices make sense to me (the UW, students are poor). Others less so. I would estimate a majority of the area covered is in lower than average income areas -- the U district and southeast Seattle.
It is obvious that there was some non-market-based decision making here. I am OK with that and Seattleites tend to be as well. I don't think infrastructure should be limited to the rich or that a fiber vendor should be allowed to cherry pick neighborhoods. However I wonder if the right balance was struck.
My suspicion is the 'highest bidder' was picked in part by how many low income neighborhoods they were willing to deploy in, and I fear this has resulted in choosing an organization that overbid because they didn't actually understand what it would take to be successful. Or worse yet is not even going to be capable of doing the buildout.
IIRC the choice to favor neighborhoods like central, chinatown, and beacon hill was intentional. The thinking was that the traditional franchise agreements allowed incumbents to choose to compete in the neighborhoods with the greatest disposable income. Other nieghborhoods were left to providers, like broadstripe and clear, that were unable to provide the significant capital outlays required for infrastructure upgrades. Over the past few decades this led to significant divergence of infrastructure, and ultimately capabilities, between neighborhoods.
One of the "funny" things about Gigabit Seattle is that the last mile is actually supposed to be wireless. Much like that other network pioneer of Seattle, Clearwire.
That is debatable. It's kind of a stretch to call AT&T Bell Labs 'privately funded' as ATT was a government-sanctioned monopoly. Technically it was privately funded, but not through a working market system. Kind of like the other great industrial pure research labs (MS, IBM, Xerox).
It does not seem like that would be very practical.
Very few economically important plants will thrive in anything less than full sunlight. This makes sense because most plants become economically important because they are productive (good at turning sunlight into something we want like calories or wood).
The exceptions are plants that are grown for some valuable quality which is only needed in scarce quantities, like some spices or flowers.
Yeah, I didn't understand the hydroponics comment at the end. How does it work better to grow plants indoors than outside if you're using a limited amount of sunlight indoors? Are there other factors that make it worth it?
White objects are not perfectly reflective and whiteness is not so much a function of reflecting them well, but reflecting them relatively uniformly in the visual range. Not every white object reflects UV.
Black plastic lasts longer because 1) the UV is absorbed by the pigment, not the polymer and 2) the pigment limits UV penetration to the surface.
University of WA cost has not gone up a single penny since 1990. It used to be that the state paid 80% of that, now it is 30%. Which is why the cost the students see has increased so much -- decrease in state funding, and indirectly, people who vote against taxes.
If UW cost had simply kept pace with inflation, it would be $30,000 this year. And if WA residents had maintained their support at 80%, the student bill would be $6000.
It isn't a lack of taxes, it is that Medicaid (a federally mandated program) is out of control and is absolutely destroying state education budgets.
"Governments’ general support for higher education 25 years ago was nearly 50 percent greater than state spending on Medicaid. That relationship has now flipped: Medicaid spending is about 50 percent greater than support for higher education. If higher education’s share of state budgets had remained constant instead of being crowded out by rising Medicaid costs, it would be getting some $30 billion more than it receives today, or more than $2,000 per student."
Many states have significantly reduced funding for public higher education. Much of tax funding now goes towards Medicaid, and perhaps half of that is not for poor people who need medical assistance but rather middle-class people in nursing homes. Nursing home payments are paid out of Medicaid as well as healthcare for the poor.
So, in a sense, the reason why states are spending less on higher education than a decade or two ago is simply because of this middle class entitlement. The middle class should be purchasing nursing home insurance but instead they leave it up to the state to support.
Maybe Akihabara was once like this but it is not anymore. The shops there are clearly boutiques selling to hobbyists, filled with things like expensive vintage vacuum tubes for esoteric audio projects and not the sort of things you would want if you actually wanted to manufacture something.