Interesting Links for 27-04-2015

[andrewducker]

• FizzBuzz In Too Much Detail
  (tags: programming )
• the main parties are all staring into a pitch-black night of the soul
  (tags: politics UK )
• 20 figures from renewable energy industry on why they want Lib Dems to play a part in the next government
  (tags: libdem environment green )
• This is what's wrong with your computer
  (tags: computers funny comic

Comments

When you accidentally create an explosive in your home, and nobody can help you...

[cartesiandaemon]

Eek!

I'm glad it turned out ok. I'm glad I didn't experience that metaphorical sinking feeling!

I'm surprised the firebrigade weren't more pro-active. "Hi, um, I think I may have accidentally created something explosive..." seems like it demands a response, even if that's just "knowing who to pass the call on to". And I'd hoped someone would have a response less drastic than the bomb squad evacuating the building and blowing it up, but I agree, I don't know if they actually would...

I also wondered about smashing the flask before it sublimed further, though I don't know at what point it's too late for that!

Re: When you accidentally create an explosive in your home, and nobody can help you...

[steer]

At the point you are considering trying to smash a metal and glass container which is over pressure inside and hence would blow outwards and you're doing that to stop it being pressurised in case it smashes and blows outwards you need to ask yourself what you're trying to achieve. :-)

My first suggestion to pete was "do you have anything that could approximate a sandbag" but he did not really. I've a fair grasp of exploding pressurised cylinders. A 200 bar dive cylinder is thick steel or thicker aluminium. If it explodes (and if they rust, they do) then it's nasty... but this is something much more fragile unlikely to get to 5 bar and much smaller. Don't be holding it and don't have it near anything you like (for example the inside of your flat).

The route to the canal is actually short and fairly unpopulated on a Sunday morning so low risk for others. Would not want to have to do that run though.

Re: When you accidentally create an explosive in your home, and nobody can help you...

[resonant]

Rather than risk carrying it, I'd have chucked it into my bathtub, closed the bathroom door, and hid behind a barricade of books. Of course, I'd then have to stop the water gushing from my shattered toilet, and explain to my landlord why he should not evict me. So I suppose the canal was the best solution.

Re: When you accidentally create an explosive in your home, and nobody can help you...

[steer]

Yes -- I actually thought something similar. I'd actually be 99% confident that it would not even crack the tub but it would be a worrying wait. (Also the people downstairs might not be happy.)

FizzBuzz In Too Much Detail

[cartesiandaemon]

Yeah, that's very well said. I often think of it as, I never want to _close off_ paths to optimisation/encapsulation/etc, but "infinitely much, immediately" isn't a good answer either.

I think someone said, you should encapsulate something the second time you do it. The first time, you'll probably never do it again, and even if you do, you often don't save any work by breaking it off into a function beforehand. The third time, it's probably too late: it's in too different places and you may not be able to refactor them all without risk. But as soon as you do find yourself doing something you've done before, it's time to say, "no, NOW I will move it into a separate function" or whatever

[simont]

I think someone said, you should encapsulate something the second time you do it.

Not least because then you have some idea how to do it. There's often more than one sensible place you can draw the dividing line between stuff you're moving out into the subfunction and stuff you're leaving at the call site, and if you refactor prematurely then you can easily find you've made the wrong guess about where that line should be, and have to do it all over again anyway at some later point.

Once you actually have a second client for your factored-out function, you have real data about what properties of the shared functionally the two clients do and don't have in common.

[chess]

The title of the article about mindfulness irritates me, because it's very easy to pick it up and use it as 'see, you don't need those nasty drugs!' where the facts are more like 'drugs work for some people, mindfulness works for some people, these are not perfectly overlapping sets and it's quite possible drugs will work for you and mindfulness won't, or neither will work for you'.

[andrewducker]

Absolutely - all they've ascertained is that it works for as many people, not that they're the same people, or that a combination wouldn't be better for some/many.

[firecat]

Yes, says one of the 44%.

And another sad thing is, mindfulness training PLUS medication works better for me than either one alone. But the gods of controlling health care costs forbid that people be offered TWO forms of treatment.

[ticktockman]

I've never had a Thermos equipped with a valve. I've only had screw top varieties. I had to check google images to see what this looked like.

[momentsmusicaux]

I'm a big fan of YAGNI.

[andrewducker]

Me too. Except when I'm doing something for the third time. At which point I'm a big fan of either patterns or shared code, depending.