I just came back from having breakfast at midnight at the local diner across the street. It seems like, once you get beyond about ten customers, a diner fails to function properly. I came in, and sat down, and waited. A waiter then comes by and says "I'll be with you in a minute", then takes the order of someone else. You think he'd remember to come back to me right afterwards since he said "I'll be with you right in a minute", but no. Instead, he heads back and starts having one person pay their bill. Then, another. Then, another.
This is taking a few minutes, at which point I think he'll never come by to help again. After all, he has, as I'm sure, completely forgotten what he said three minutes earlier. Now waiters aren't paid a lot, so there's not a high skill level for many of them. The key to good waitering is memorization. In fact, memorization tends to be great for a lot of professions, and is vastly underrated.
A great example of this comes in the film Tampopo about a woman who runs a noodle shop, albeit, rather poorly. A trucker tries to teach her the ways of being a good noodle shop owner. One key is to listen to half a dozen customers rattle off their orders, all with special requests, and her ability to parrot back the orders without ever having written it down. The closest real-world example I know is some lady working at a sushi shop who remembers half a dozen orders, and often remembers orders you made the previous time you were there. It's a bit of a shame that all that memory is in service of waitering, but it's impressive when it happens.
Needless to say, she's like the Einstein of waitering. Everyone else is remedial, at best. For example, I went to Dunkin Donuts, and this lady somehow didn't hear that I ordered a bagel with salmon spread. She was convinced I said sour cream. Furthermore, I was ordering two bagels, one sour cream and onion (this is how she forgot the cream cheese spread) and one with sesame seeds. The final result was backwards. I wanted a sesame seed bagel with salmon spread, and a sour cream and onion steak bagel. Somehow, she could barely remember two orders. This is clearly not the job for her. Needless to say, she didn't write anything down, thus exacerbating her lack of memorization skills.
A second woman did finally come out (where was she hiding?) and I told her I wanted to order right away. She almost seemed surprised by this, and wasn't even ready to take down the order. It was a special, scrambled eggs, and water. I could have said the order the minute I walked in, since I almost always order the special. Still, this crowded evening meant i was dealing with people who couldn't keep track of who came in and when.
My solution to this is to have timers. The instant you sit, the timer starts. Then, you can see how long it takes before they really get to you. I swear, this one addition to waitering would improve it by a lot. You could point to the waiter and say, do you realize it's taken you five minutes to get to me? Now I understand that many restaurants don't like to have too many waiters. Each one you add means less money for everyone. So, one waiter can be extremely overworked, and unable to process which order to take next. Even so, the stopwatch should prove useful. I know. It won't happen, because it would give customers one more thing to complain about.
I wouldn't have even been at the diner had the local coffee place actually kept their kitchens open. For some reason, coffee houses often have subpar food, and really, this place is no exception. Some of the food is horribly overpriced, but I assume that's because it takes SO much time of them to cook food, as opposed to dispensing coffee. Surely, some marketing genius is taking great credit for coming up with the idea of a coffee house, where overpriced coffee and lack of real cooking results in profits galore. Any cooking that does take place in a coffee house is guaranteed to be extra slow, and not up to real high standards.
This leads me to my third watching of Aardvark'd. I've been watching the DVD commentary. I know, statistically, that watching the commentary is almost unheard of. It's possibly even worse when you are watching commentary by the interns, who aren't particularly trained to do commentary. I don't particularly blame them since this is not in their skill set (there, I used "skill set" in a sentence).
I listened to Benjamin Pollack and Michael Lehenbauer in one commentary track. Clearly, of the two, Ben was far more polished, and had lots more to say. Michael would giggle and add some commentary. It's funny what they point out as interesting. The commentary did leave some unanswered questions. For example, it seems Liz was about the only woman there, and by some standards, she's pretty attractive. Yet, the interns were pretty PC in not saying that she was hot, or anything.
The interns tended to comment about themselves, rather than to talk about the others. For example, in that particular commentary track, Yaron comes on. When he's on, neither say much about Yaron. Was he a great guy? Was he insufferable? In fact, until the third time through, I didn't realize he was a marketing guy (I know he mentions it, but it's very briefly done). I don't even know what his responsibilities were. While it sounded like he didn't code much, it also seemed like he was capable of doing it.
The commentary helped because I could finally tell who was who. Since the first commentary was Benjamin and Michael (technically, it's the the third commentary track), I could focus on those two guys. I'm part way through listening to Tyler and Yaron. Their commentary is a little more balanced, but you could tell that Benjamin liked to order Michael around, even though it's not clear in Benjamin and Michael's commentary track. Yaron was super self-concious, and his hope was that he wouldn't seem like a total idiot.
By the third time watching it, you begin to pick up all sorts of odd stuff, like the changing facial hair of Yaron, who is mostly clean shaven in one take, or sporting a grungy beard growth in another. You also see the interns in the backgrounds typing away.
What I found particularly surprising was the number of people at Fog Creek. Not including interns, it's like 4-5 people. That's tiny. I mean, really tiny. I was always under the impression the company had at least 20 people. I had no idea that it barely had anyone. Thus, all the software is probably the result of 3-4 people (at least, up until Project Aardvark). I guess this is why they want to have a bunch of bright programmers. You need far fewer to achieve the same (or better) results.
There was also another segment I didn't quite get until seeing it a second time. There's some guy who grows tomatoes. I had thought it was one of the interns. After all, it's not always clear who is or is not an intern, when all the faces are new. At the end of the documentary, the guy is tearing up the tomato plants (probably because it's cold, and the plant has already borne fruit). I had originally thought there were two different guys, that an intern had planted the tomatoes, and that one of the full timers had ripped it apart. It comes across far more malicious when you think that, then if the same person who grew the tomatoes also tore it down.
I also made a mistake on the music. While the director did compose the music, I think another guy did the lyrics. Yaron points out in his commentary track that it's really cheap to get someone to do music, because so many people want to do it. Even so, I found the music amusing.
In the end, I also realized that Copilot may not be the best of products. While it's a good idea, how often do people really help out their parents, etc. I suppose it seems like a lot, but is it even a dozen times a year? At 2 or 3 times a year, it becomes highly infrequent. I don't have a particular great idea for them to work on for next summer, but I'll ponder it some more, and maybe post it to the forums.
Ah, the forums. Apparently, this is the equivalent of Real World folks being able to hear from fans (which never happens because the filming is done months before it's shown). I should now go back and read the blogs, which I barely paid attention to when it was live. It's archived, so I can reread it all. I'm hoping the blogs offer some insight that I wouldn't already have.
In hindsight, I think the DVD was a great idea by Fog Creek. I'm still not crazy about what made it in there, but at least the people seemed likeable enough (if a bit geeky all around). I'd still love to see if they have good ideas about how to write software, or whether you simply throw bright kids at the problem, and things get magically solved with minimal amounts of planning.
Three opinions on theorems
-
1. Think of theorem statements like an API. Some people feel intimidated by
the prospect of putting a “theorem” into their papers. They feel that their
res...
5 years ago
No comments:
Post a Comment