there is lots 'coding around the problem' with special cases and branches all over the place

While this is undoubtedly the case, part of the cause is that the world is full of special cases.

The other half of the problem is that many programmers struggle to distinguish between a defect report that is due to a bad assumption about the general case and a defect report that is due to a special case that the general case doesn't handle.

So you get two possible issues - either they change the general case to handle a special case - and break something else along the way (which then gets fixed in its own peculiar way), or they add special case handling and fail to fix the general case. After 20 years of acquiring "bad" fixes like this the code becomes almost unmanageable and even identifying the special cases that would need to be handled in a rewrite are hard to impossible to determine.

Your post sounds to me like one of those "I'm ideologically opposed to the proposal, so I'll think of some problem that it has and invest zero thought into possibilities of how the problem could be solved, because I'm just waiting for you shut up about this issue that I wish we wouldn't even be talking about."

Ha! Precisely the opposite.

But I'm one of those completely state school (free/public for those in the US) educated bods who made it to Oxbridge.

http://www.telegraph.co.uk/edu...

My school advised me (prior to applying to Oxbridge) to drop further maths so I could concentrate on my other exams (which I refused to do). Of course, when I got my offer (Physics) I needed to pass further maths. I probably wouldn't even have got an interview if I'd dropped further maths and I'd have had no idea (nor the school). Quite frankly, I don't think I'd have survived the course without having done further maths in the sixth form.

Well, whether or not there was sexual harassment (illegal as hell, BTW, so that's a serious charge), it sounds like some bully (or bullies) found her achilles' heel, and exploited it.

Sexual harassment, like all other instances of harassment is not the case of "something really bad that should never have happened." That would come under sexual assault, or assault if it isn't sexual.

Harassment is when there's a prolonged course of incidents that individually are not particularly serious but when viewed as a whole, make someone uncomfortable and feel unsafe.

You're allowed to ask that co-worker out. You're allowed to say someone is looking lovely. But you'd better be sensitive to whether that comment is well received because constantly repeating it when it's not appreciated could constitute harassment. And if you cannot tell whether it's well received or not then better to not start at all.

Harassment is Chinese water torture. One drip never hurt anyone.

The problem isn't giving gifted children the opportunity to take advantage of their gifts.

The problem is that the wealthy will use their wealth to coach, and otherwise promote their average or slightly above average children so that they get into those places for gifted children in preference to the gifted poor child who can only score as average or slightly above average due to lack of opportunity and education.

I'm sure the same problem happens in the Asian cultures where this is the norm. There is possibly a difference in that educational excellence is seen as something to boast about and so a poor uneducated peasant who has an exceptional child will still want to see them enrolled in a gifted child program unlike in the west where ignorance is sometimes seen as a badge of honour.

The laws are reasonable - they allow costs to be offset when calculating profits.

The problem occurs because
A can sell to B at cost who can sell to C at the price that C finally sells for and the only person making any profit is B.

Perhaps the market is very competitive, A cannot sell above cost. C cannot buy at below the price B will sell at. (obviously in the real world A and C would need to make some profit but it's not inconceivable that A and C are working with margins of less that 5% while B is working with a 40% margin.)

Or perhaps A, B and C are all subsidiary companies of the same parent. It's still possible that A and C are in a very competitive market, or it's possible that the company is artificially moving profits to where they would prefer them to be.

It tells me they were not following accounting principals and balancing the books at the end of the month.

How do you balance the books? You have a bitcoin which is "stolen" and spent. Surely it's only when you come to try to spend it yourself that you discover that someone else has spent it first and your bitcoin is no longer valid.

It's an interesting problem though. If you give me your bitcoin wallet, (along with the key) then I can make a copy of your wallet and spend the money in my copy. When you ask for your wallet back, I can give it back to you, unchanged. Have I stolen anything?

In real life, if I made a copy of your wallet, and then spent the money in the copy, I'd be guilty of forgery/counterfeiting. I wouldn't be guilty of stealing from you.

And if the recipient forwards it unencrypted then S/MIME or PGP are not very useful.

I was envisioning each person running their own mailserver (as I do) so that the only place mail would be unencrypted would be on their local machine.

Once you're in that position everything gets encrypted and it's invisible to the end user.

If you deliver direct to MX and have TLS setup on both ends then you get this encryption for free.

There are active attacks (MITM) against most setups but passive snooping is prevented.

Although your facts are correct, the reason CO2 causes warming, and the reason it doesn't saturate too, are more complicated.

These complications are the reason why it wasn't until the late 1940s and the advent of high altitude aircraft that these areas of confusion weren't definitively settled.

A better model (one that behaves more like the real world) is to consider the Earth as a black body where the surface is a mile or two up in the atmosphere rather than on the ground.

CO2 (plus water vapour) are what control how high into the atmosphere that surface is.

Because of the lapse rate, the ground will be warmer than the surface of the imaginary black body.

As CO2 increases, the height of that black body surface increases therefore it's temperature decreases. However, if the temperature decreases, the amount of radiation escaping to space decreases while the amount arriving from the sun stays the same, so the ground starts to warm up.

Eventually, the ground warms enough that the black body surface is hot enough to now be in equilibrium with the energy arriving from the sun.

And because that surface raises if CO2 is added regardless of how much CO2 there already is there is no "saturation" point where more CO2 doesn't cause warming.

It's slightly bizarre but sometimes people cannot see that they're being inconsistent.

I wish I'd written it down because it was a perfect example of this - I was approached to make a change to some functionality.

I said - but if we do A then B happens.

To which they said "but we can do C" (which did solve B)

I said - but if we do C then D happens

To which they said "but we can do E" (which did solve D)

I said - but if we do E then F happens

To which they said "but we can do G (which did solve F) except that G was precisely undoing the required effects of A.

It took over two hours going through these simple steps with them before they "accepted" that we couldn't do what they wanted (B, D and F were all unacceptable and no dispute about that) and I think they thought I was playing a trick on them.

That's rather interesting because:

http://www.nabendynamo.de/prod...

Electric Power: 6V / 3W according to German government's road traffic regulations (StVZO)

German Mark of Conformity: ~~~ K 687 for 16"-28" (400-716 mm) in combination with Linkhinweis Edelux

Nonehteless I am betting such light would be forbbidden in many country in europe where the maximum intensity you can pump is limited by law.

It might have changed, but I believe that it's not the maximum light output that is limited but the power input.

So from tungsten filament that the law was written for to the thermodynamic limit gives about a 50x increase in brightness that is allowed.

Similar games for bicycle lights. The reason it's almost impossible to get a bicycle dynamo that will output more than three watts is because that's the legal limit for the front light. In true lawyer fashion, you're allowed to have extra lights (provided they're independently controlled so you can turn them off without your legal light) that are brighter but the lamp the law requires is limited to a maximum of 3W (input)

Many of these laws have been changed relatively recently - for example LED lights weren't allowed at all for bicycles, not sure what the situation for cars was - but they are now.

I think it sounds quite clever.

Typically delivery consists of shipping to a distribution centre and then shipping to the customer.

Anyone who's had a tracked delivery will know that "Arrived at distribution depot" "Out for delivery" steps.

What I think is being suggested here is that you start the shipping process before you know the final destination.

Presumably the distribution depot isn't going to store the parcels (over and above the storage they have to do while waiting to load it onto the van for local delivery)

But here I think you only attach the destination to the parcel when it arrives at the distribution depot. If you have a customer at that point then you attach the customer's address otherwise you put the originating address on the parcel and ship it back.

Get it wrong and it could get very expensive. Get it right and many customers could see very quick delivery of large items that normally would take 2-3 days minimum even for a premium delivery service.

For an odd number of samples consider the median value first.

Obviously for one value, the distance is minimized by picking that value.

Now add the next two points (one on either side) - the extra sum of the absolute value of the distances is the distance between these two points regardless of where we put the value between them- so it's still minimized if it stays at the median. repeat.

For an even number of points the initial point can be anywhere between the two central values.

For completeness you need to consider (and reject) the case that the median lies outside the list completely.