Sunday, April 24, 2016

Naming issues, conflicting interests, and their connections to constitutional law

Hope springs eternal in the human breast. Every year will somehow offer more time to do things. One of the main reasons people who will never benefit from policies that benefit only the rich vote for said policies and vote against policies that would benefit those exactly matching their income class / social situation / [etc] is that they believe they will eventually be those rich people and once they get there they don't want to be paying taxes that benefit people like themselves, currently. [Or, similarly, everyone is convinced that at least their children will be doing better than them when in fact lately in general your kids will end up doing worse than you.] I just finish a blog post and think "that was easy; I'll definitely write 5 more next month and be right back on track!". Of course the purpose isn't simply to fill the blogger.com data centres with random words, but content is king, as they say (I think). Hannah will stop crying for no reason all day long any day now; it won't rain every weekend; I'll have way more time to go cycling next week; etc.

At work we are solving the world's problems. Every day the decisions we make are touching fundamental problems/solutions and the paths that we take will have long-term ramifications far beyond anything that we can currently envision or plan for. That's why, this past week, I spent most of my time arguing for the following parameter order for our implementation of the clip function:

clip(x, min_value, max_value)

where x is some tensor (collection of elements with some shape -- like a vector or a matrix, for example), and min_value and max_value are minimum and maximum numbers that the elements in x can have. The output of the function will be the same tensor x, but all of the values will be "clipped" to fall within those min and max values, if need be. Just in case you didn't get it (and because it's so important for the story at hand), here is an example... Let's say x is a vector with x = [1, 2, 3, 4] with min_value = 2 and max_value = 3.5. Then clip(x, 2, 3.5) would result in [2, 2, 3, 3.5].

Anyways, so that's the obvious parameter order. Everyone else does it that way. If you change it you're not doing anything other than confusing people. But, some believe that the data should always be last. So, instead, the signature of the function would be:

clip(min_value, max_value, x)

The argument for the above is that having the data last allows you to do some interesting function chaining where you're setting up a network of functions, etc. So there's this competing question of which is more important? Allowing this (and being consistent about it across our implementation where we would always put the data last), or being consistent with everybody else so if they come over to our tools then they won't be confused. So obviously this is just the same as determining the choice to make when we have conflicting constitutional rights! Like there is a right to free speech but there is also a right to not be subjected to hate speech and you need to balance those interests, etc. And when I wrote the title of this particular blog post I guess I thought I would spend more time on the constitutional law and conflict of rights issue and all that but now I don't feel up for it. Sorry. But these are the important issues that kept me busy at work all week (inter alia, of course).

And here are some completely different random observations that will surely enrich your life:

1. I was sitting back with a (home-brewed) Segafredo1 espresso2 watching La Doyenne ("The Old Lady"), aka Liège-Bastogne-Liège, the last of the Spring classics, and during the break an ad for the shampoo "Alpecin" came on. This in itself is not at all out of the ordinary as Team Giant-Alpecin is a major German UCI cycling team and so during cycling events such as the Tour de France and La Doyenne you will often see ads put on by sponsors of the major teams. Anyways, as far as I can tell they are (or at least should be) known for really dumb ads. But this one took the cake (/"the biscuit" if you're British). The ad starts with a new (I presume -- the 'new' part I mean) slogan that fills the entire screen and has the announcer read out the words:

"Doping für die Haare"

And you don't need to be Jonathan Franzen or Goethe or have a 500+ day streak on German duolingo to quickly figure out that the above reads "Doping for the hair". Even though I don't need to buy shampoo because I didn't do enough doping for my hair (or something), because of this inept / tone-deaf commercial aired during La Doyenne, even if some miracle happens or I make some kind of Faustian deal (weird how we're sort of back to Goethe again, right?) that will give me my hair back in return for my soul then I will never buy this shampoo even if it's the only shampoo in the store because come on.. I guess it's meant to be a joke but even by German standards it's pretty weak. And we're meant to be joking about / glorifying cloaked in humour doping during a cycling event in 2016? Puh-lease.

[FN1] Two of Sandi's cycling heroes3 are the Canadian Ryder Hesjedal and the Swiss Fabian Cancellara. As fate would have it, this year they are both riding on the same team: Trek Segafredo. It is sort of a new team (the teams are always changing around a bit and changing names as sponsors come and go, etc.) because last year they were just called "Trek Factory Racing". But now Segafredo has joined as a major sponsor and so they're, as mentioned, "Trek Segafredo". Since both those guys are on the team, and Sandi didn't have a really cool really professional cycling jersey before, and we both like coffee [see FN2 reference supra, and definition, infra], she got a brand new Trek Segafredo jersey for her birthday and so today she can wear it to cheer on the Trek guys in La Doyenne.

[FN2] Why is there this crazy connection between cycling and espresso? I think both of those things are pretty cool so it works out well for me, but where does it come from? Perhaps a topic for an essay for students to be assigned for some kind of English course? In the meantime I did a very quick Bing search and the Grimpeur Bros. have some common sense suggestions (https://www.grimpeurbros.com/pages/coffee-cycling-the-inextricable-link) on the "inextricable link between cycling and coffee" (but nothing too concrete). These include obvious things like "cycling‘s strong ties to Italy, France, Belgium, and Spain probably have something to do with it" and "on weekends, a post-race coffee or mid-ride espresso break, gives us the opportunity to take a break, share some gossip, nosh on some baked goods, and most importantly, share a laugh and enjoy our friendships — all over coffee." Fair enough. Wait, do any of you care about this at all? I think it's pretty interesting, anyways...

[FN3] Yes, in Europe people have cycling heroes. Kind of like in North America people have hockey and baseball heroes.

2. Ben Harper has a new record and it's really good. Sandi is a pretty lucky girl. In addition to being blessed (in the atheistic equivalent anyways) with a healthy (so far) baby girl, and receiving the aforementioned Trek Segafredo professional cycling jersey, she also received a copy of the new Ben Harper record "Call it what it is" for her birthday and we have been listening to it, like, 40 times a day. It's important to learn all the songs because we want to be able to sing along when we see him in Munich on October 1st.4

[FN4] Sandi and I might have a somewhat deservedly-obtained reputation of not exactly being the best at planning ahead or being prepared for things, etc. Fine. But, in this situation, despite the fact that October 1st is currently more than five months away, we already have experienced babysitters lined up for the night! K&J will be arriving in Munich several days prior and by then will have more than a year of baby girl taking care of experience. Not bad, right?

3. I think I gave Belgium a bad rap. Brussels is not a fun place to visit. And so I kind of hated it. But watching Liège-Bastogne-Liège, it looks like it's not necessarily such a bad place. And they got by for a couple of years without even having a government. So that's not so bad either...

Until next time...

Addendum:
Remember the Krailinger Duathlon race report from last year? Well the 2016 version is coming up next weekend and if it's raining like last year we won't do it.. But if it's nice, we will! If it's nice then, look forward to another upcoming fun-filled race report where Sandi and I might do a relay and hand-off Hannah in the transition zone. Stay tuned...

Saturday, April 2, 2016

Pemulis and Hans

He's been literally counting the minutes tick by for the entire six days that have passed since she told him. Almost every single one of them. Around 8,640 minutes minus a handful where he was able to sleep for the briefest of moments. In fact, most of the time he's even been counting the seconds. He had created a whole swath of games in his head with the numbers. Sometimes he would count up, and other times he would be counting down. Something like "it's been 945 seconds since <some time> and 4 + 5 = 9 and 5 + 4 = 9 and 9 - 4 = 5 and 9 - 5 = 4 and 945 seconds is exactly 15 minutes and 45 seconds which is crazy because the 45 seconds do the same thing as before but now we have the 15 minutes and with 15 you can get 5 - 1 = 4 and then you have a 4 which goes to 4 + 5 = 9, and cetera...". This had actually been happening off and on in his head for as long as he could remember but it had really ramped up and came more to the surface, let's say, over these previous six days. During the wait, he had gone through an entire week of work. She told him on a Sunday, now it was Saturday afternoon, and between those two moments he had somehow attended meetings, written e-mails, conversed with colleagues, and all manner of work stuff had gone down so to say, but all of it was, well, I guess, repressed? It happened completely and utterly capital F-R Freud subconsciously like a dream or even further back than a dream. He knows it happened but he basically just floated through all that stuff that seems to have somehow happened. Each day after work he headed straight home and straight to the couch where he immediately turned on the TV. News, weather, weather, watching the clock, counting seconds, counting minutes, weather, moving pictures that didn't actually help to speed up the clock (or at least the feeling of the clock speeding up) but he felt that if he didn't go straight to the TV that things would feel even slower somehow (even though he didn't try it [not going to the TV, that is]). He took dinner in front of the TV, leaned forward over a short white cheap coffee table. Drank six beers. Seriously. Six every night, and on two of the nights whiskey for dessert. It still couldn't bring him towards sleep, though. Things simply got real hazy and even slower, if you could believe it, and he knew that the beer and the whiskey weren't going to get him to sleep and he even knew that it would make things worse because not only would he not be able to sleep, but he would be drowsy enough that he wouldn't be able to do anything else like read or even watch TV because his eyes would burn so much, but just like with the TV (earlier in the evening) he couldn't NOT do these things because the fear that not doing them would make things even worse was stronger than the rational realization that doing them was making things worse.

Friday night should have been a milestone both because he had somehow made it through the entire work week and he was now only 24 hours away from it. But things felt worse than ever. Since time had somehow exponentially been slowing down, psychologically he was at the furthest point from Saturday night that he'd been since the previous Sunday! That's some wonky space-time physics-voodoo. But still he soldiers on and watches the news and the weather and drinks his six beers and tries to keep watching more TV but his eyes are burning and so, just like every other night in the last week, he turns off the TV and he goes back to counting the seconds and the minutes and after what is truly an infinitely long period each time -- and he knows that a lot of this "literally this" and "literally that" and "infinite this" and "infinite that" is sensationalistic and trust-bleeding in the sense that clearly it wasn't literally every second that he counted and an infinite amount of time by definition will never be reached and all that and so he's sorry that it has to be portrayed this way but THERE'S LITERALLY NO OTHER WAY it was just so so long and the time was going by so so slowly -- he counts the hours too. Finally Saturday morning arrives and the dog wakes him up (yes he actually managed a few minutes sleep somehow) by licking him on the face -- oh what a good boy, he would never do this to me -- and he pats the dog on the head and sits up from his lying position on the couch and turns on the TV and starts watching the news and the weather. It takes just forever, of course, but finally after eons of watching TV and pacing the house and watching more TV, evening arrives. Pemulis feels, in precisely equal amounts, both dread and relief.

His daughter Helga arrives home after being out all afternoon and announces that her new boyfriend Hans is no longer coming over for dinner because she broke up with him and she decides she's never going to have another boyfriend ever again. So Pemulis takes her out for a fancy dinner and buys her a beautiful gold Rolex and a sports car and they play scrabble and he goes to bed and has the most luxuriant sleep of his life and doesn't wake up for 15 hours.