Sausagely is a very impressive young man, but every so often he still manages to remind us that he's about 14.
He could probably kick the shit out of any 60 year-old pundit. And I don't think you're taking either speed or cardiovascular fitness into account.
Still, you're right. He is underrating the 60 year-old set.
Who cares whether you can take out a sixty-year-old? It's far more important to know how many five year olds you can take out.
I bet he couldn't even take Rocky.
O, do you want me to move my post below yours? God, that sounds so gay.
Also, you're right on the merits. I bet a well-trained and experienced 60-year-old could take out Yglesias, unless Yglesias has some skills we haven't heard about. ("When I was at Harvard, I used to compete in UFC bouts for amusement." Wait, all those words are spelled correctly. Never mind.)
Did anyone else listen to the ISG press conference? There was a wonderful moment when James Baker called on Ackerman, saying "Yes, you, the gentleman with the beard."
3: Yes but how many midgets could take out a lion?
What is the proper way of measuring reach?
OK, what's the point of change then? I feel sure that Yglesias could take out almost all 80 year-olds, and even most 75 year-olds. What about 70 year-olds?
What is the proper way of measuring reach?
From underneath the scrotum.
maybe Matt wants to take on the 92-year-old Jack Lalanne?
Yeah, even Jack Lalane got brittle, right?
whoops, he was reportedly 81 in that photo
Ha. I think the big difference is resiliance-- MY could take a couple of good tackles, but 70 year old bones will crack more easily. Jack is best off sweeping the leg and taking things to the ground, then going for a submission hold.
What is the proper way of measuring reach?
Here.
9: Maybe most, but not all.
Sausagely has no concept of how old or not old sixty is. Anything over 40 probably looks decrepit to him.
I"m under the impression that a lot of aging happens between 60 and 70 nowadays. There are a lot of strong and tough 60 year-olds, but many fewer such 70 year-olds.
15 -- But we must grant that a 70" reach is pretty impressive, in this case.
Yglesias is probably thinking of guys like this (warning: you will hate yourself for laughing).
My dad's led an entirely non-violent life, so he probably couldn't 'take' anyone, but he's in pretty darn good shape and he's 67. I'd say that for a healthy, fit person, the line where you're all of a sudden genuinely old is somewhere in the 70's.
-gg-d thinks bumfights are funny!
Richard Armitage is 62. In this photo he is demonstrating what he will do to Sausagely's head.
Could Sausagely take out anyone in Bush's cabinet? I'm thinking Don Rumsfeld would be the toughest (he's still technically there, right?). Followed by Condi.
Right, but it's not most X year-olds, but all X year-olds. How old before you feel certain that there aren't any obvious outliers who could kick the shit out of Yggy. I say 75; I think dementia might tick up at about that point.
I think Yggy could take 82 year-old Lelane, for Labs's reasons.
Condi? she's like a stick figure. I could take her, easy.
28 -- no, this is what w-lfs-n would do.
I think Sausagely should be Liberalism's champion. We will send him into the arena against Dick Cheney to reclaim America for the pure of heart.
Yggy can't take Chertoff (53). He looks to be the toughest fighter. Everyone else, I think he probably could.
You figure that if McKean can take Spellings, Sausagely probably could too. She's not 60 though, right?
He's a silly kid. A sixty-year-old knows enough to use a .45 when it's indicated.
Only hand-to-hand combat in the struggle for America's soul.
33, the OED says both are acceptable.
Also, I'm sure Sausagely could take this cream-puff.
The problem with old people is that, while their bones are more brittle, they're masters with guilt. I bet they could make you feel really bad about beating them up, getting you off guard. I'd have an easier time beating up the 5 year olds who, while cute, have more of a crazed animal thing about them that would make them easier to hurt.
Which makes me wonder how many 85-year olds a person could take in a fight.
6'1" and 215 sounds a little high on the BMI.
Course, you can get away with that for a while when you're a kid.
We have to remember that Bush generally selects his cabinet members on the basis of their virility, as we learned from the firing of Larry Lindsey. It's possible that the only ones without sufficient OMS to subdue Sausagely are the token minorities, like Toady McCronyism here, or Mr. Rumbold (the token person with scientific experience).
Rereading 42, I should so put that in my online dating profile.
I should so put that in my online dating profile.
Huh? Why are you holding out on us?
Revoke my Unfogged Fan Club membership if necessary, but what's with Matthew Yglesias = Sausagely?
Spell it backwards, then mis-spell it.
Well, that's certainly unnecessary.
Hey, I have an idea: let's all call Bush Shrub!
I think it's for googleproofing. Why googleproofing is desired in this instance, heavens knows.
No one is excused from reading the archives without a note from their doctor.
Yeah, 58 > 52 as a hover-over-text candidate.
55 is now the actual hover text.
"Kohai, what is the first rule of combat?"
"If you encounter an elderly, smiling bald man, you are going to get your ass kicked."
61 would be good hover text for one of the blogs that no longer exists, like "Respectful of Otters".
Hasn't anyone told you, w-lfs-n, that 'sed' is a dead letter?
55: The archives refused to yield to me their mysteries.
No one is excused from reading the archives without a note from their doctor.
Indeed. And start with Comment # 5.
Hasn't anyone told you, w-lfs-n, that 'sed' is a dead letter?
I use sed (and awk!) every now and then. Why dirty my hands with perl?
69: A man's choice of command-line tools is his destiny. But then, I guess you're entitled to go at things your own way.
69 -- why the exclamation mark after awk? it is a magnificently useful tool and I should think anyone that uses a computer much would have it in their arsenal.
i actually meant "everyone", not "anyone", way back up in 27.
arthegall, what do you propose as an alternative? Perl? Really?
clown, I agree. awk is very nice to have around when you really desperately don't want to open up a spreadsheet.
I'm thinking Don Rumsfeld would be the toughest
Rumsfeld is 74 years old. Cranky != tough.
Fucking hoohole. I had to take EXTREME MEASURES to find those.
Focus, people. This thread is about beating up old people, not geeky tech stuff. Unless you're going to talk about beating up Brian Kernighan, take this talk elsewhere. (Bjarne Stroustrup really does have it coming but unfortunately doesn't meet the age 60 threshold.)
Schwarzenegger turns 60 in July. Let's go ahead and arrange the match.
Seriously, what am I going to do with half a goat?
Apo, he has no soul, and I doubt he experiences pain, or if he does, it's not as an unpleasant sensation. There would be no victory over Rumsfeld short of consuming his corporal form. And he'd fight you to the last pinky.
Oh, right. Ship it to Minnesota.
I don't know if Gerry Sussman's over 60, but Sauselgy could definitely take him.
Seriously, what am I going to do with half a goat?
Depends. Which half?
Becks -- what do you have against Bjarne?
I was at a wrestling match recently, and I was impressed by how inscrutable the moves were to someone who doesn't know anything about the sport. Like, why are you doing that to his elbow? Rummy, a collegiate wrestler, probably remembers a few inside moves that would take Sausagely by surprise, and the next thing he knew he's wrapped up like a pretzel.
I'd pay to see a Sausagely v. Sexagenarian match. It'd be even better if Bloggingheads.tv hosted it.
a collegiate wrestler
Yeah, but he's *74*. You just have to distract him with Murder She Wrote or Matlock, then hit him over the head with something heavy.
There is no alternate world in which sausagely loses to Rumsfeld. His skin has thinned to such an extent that you can see his skull. He can't be more than 18 mos. from the grave.
85 -- Who doesn't have a deep and abiding desire to see ole' Bjarne beaten in a fight with Sausegly (or anyone else)?
Three words: "virtual base classes."
Thanks, standpipe. Here I thought "Sausagely" was an obscure expression of disapproval.
What's wrong with the Google where those posts are showing when I search for "Sausagely"? Did somebody robot-zap them to protect the innocent?
I'd be afraid to fight Rumsfeld and I'm not ashamed to admit it.
virtual base classes rock.
The 'inventor' of virtual base classes should be beaten with a rock.
sorry clown; the language is an abomination.
What's wrong with the Google
As SB said in 77: Fucking hoohole.
A bit of a problem. The masters of the blog continue to sacrifice chickens in the hope of rising again.
Yahoo currently has a much more complete index of the site than Google. Why this is so remains unclear.
The Sausagely posts aren't in Yahoo, either, by the way.
Never mind. They're in both Google and Yahoo, but neither of them highlighted my search terms in the result. Because that would have been helpful.
95 -- I've never quite understood this attitude I must confess. Granted that much awful code is written in C++ (equally true of C); but I think it's actually easier to write good code in C++ than it is in C. I've written my share of both types of code in both languages. So it's hard for me to see how C++ is considered a step down from C. If you're saying both C and C++ are an abomination compared to some other language like Smalltalk, well, I'm not qualified to comment on that.
They're there, but not quite. The googleproofing is working (mostly).
Also, templates are teh Pwesome.
It is obvious that both C and C++ are abominations.
C is an abomination when compared to some other language like Scheme. C++ is an abomination when compared to almost any other language. It's the little things, like the so-poorly-designed-that-it-can't-be-well-implemented STL, or the way that templates are 'designed' into the language, or the virtual base class thingy, or the way that C++ exception handling is so poorly grafted to the language that it basically breaks pointers. Which is half the reason to be using C++ in the first place.
I don't know, maybe the C++ community has gotten around these issues in the last five years, since I really used the language? But if they did, it's not because of Stroustrup, who should be outcast from the community of language designers.
I think Objective-C and Dylan look like interesting languages (though I guess if you're going to learn Dylan you might as well go whole hog and learn Scheme or some more parenthetical flavor of Common Lisp).
Or Haskell!
101: C is a good design for a special purpose language. C++ is a bad design for a more general purpose language. It's a hybrid that doesn't quite achieve the strengths of it's progenitors, while creating new weaknesses. It has weak genericity, mediocre object model, several internal inconsistencies (this is crazy!), weak dynamism & introspection... it's just a bad design. Don't get me started on template metaprogramming.
It's a marketable language, but it's a lousy design.
Okay, well granted that the exception handling sucks. I like some things about STL a lot but you're right that it has big problems too.
The word "sausagely" is much commoner than you would think.
106: No, C is a nice little portable assember, when you need such a thing, it's nice.
111: you say this as if it were a bad thing?
108 -- I took a college CS course that was taught in Dylan, and liked it a fair bit: Scheme+objects+types! Yay!
My (more recent) experiences with Objective-C have been less fun, but I only dipped my toes in the pool and could probably be convinced that the water wasn't as cold as I thought it was.
Someone should just suggest "ML," and be done with it.
Yahoo currently has a much more complete index of the site than Google. Why this is so remains unclear.
Actually, it's because Ben screwed up. I hadn't said anything, in the spirit of camaraderie, or some other similar notion that has no meaning in w-lfs-n-world. Ben screwed up. Ben made the hoohole.
108: Dylan isn't really a win over common lisp (which as far as practical general purpose languages go, is probably some sort of local optimum) though, with the possible exception of class sealing. The notational thing is a red herring.
If you like ML, you might like OCAML....
Yggy needs to get out more. There probably isn't a town in America that doesn't have a 60-year-old woman that could boff Matt to death without putting down her beer or her cigarette. Note that I said "could", not "would"; even sexagenarian barflies have standards.
Prolog is a lot more interesting than any of those.
Well, types aren't just a 'notational thing.' I don't know if that's what 116 is referring to...
112: Not saying that C doesn't have its uses in narrow domains (it obviously worked quite well for Kernighan and Ritchie back in the day), but it's continued wide-spread use outside those domains is a menace to civilization.
What a shame, that a brute fact of physiology would prevent Ben from personally delivering me his head on a platter.
114: I'll second 116 and highly recommend Ocaml.
118: Prolog is interesting, but impractical. I don't believe it's more interesting than (general-sense) lisp, but people are less used to it.
121: there is that.
I had to use ML in a compilers class. I thought it was pretty stupid that you had to have a separately-named function for every general-purpose operation (like if you wanted to convert something to a string, you'd have to have like float2str, int2str, list2str, yournewtype2str, foo2str...). I have much less experience with Haskell than with ML (and honestly not much with ML), but its type system, where you can just say hey, this new (or old) type? It's printable, here's how you print it, thereby allowing for generically-named functions, seems much preferable to me.
I imagine that O'Caml is better in this regard, what with being O and all.
121: Wha? In contrast to most other languages mentioned here, ML has a type system.
124: s/impractical/less generally practical/
Aren't the sorts of type systems in ML and Haskell and whatnot basically like mini-prologs?
Well like I say I'm not qualified to comment on most other languages. I feel like a philistine and am not particularly happy about that; but the socket applications I write in C and C++ are good, dammit! Probably oughta learn some other language so I can play with the big boys. (like w-lfs-n, tee-hee.)
Wow, 78 really blew up in Becks' face.
If there's one thing that Becks should have learned by now, it's that being focused is a like like being earnest. Everything post-78 was totally predictable.
Also: 128 is close but not quite right, right? ML lets you reconstruct a certain restricted kind of polymorphic type, and the type reconstruction method will share some characteristics with some of the techniques used in evaluating a prolog program. But I think prolog is a lot more general. You could think of the ML type reconstruction algorithm as a restricted form of prolog program (I think).
125 seems closer to the mark, but ... isn't some kind of reasonable typing scheme with subtypes supposed to let you get around some of this foolishness? I don't have any experience with OCAML -- is this the kind of problem that's fixed in that language?
In general, expressive type systems are awesome. This is the whole reason why C is essentially a 30-to-40-years-out-of-date abomination.
133 -- but it's got pointers!
Okay, so I have a different (earnest) question for Clownae (leaving behind all the typing stuff, which I am learning but am not an expert in by any means)... if you're writing Socket applications, why aren't you writing in (say) Java? Or C# (which I haven't used, but take to be the Microsoft-version-of-Java designed to get them around anti-trust issues)? It's a reasonable language, it's got a good library, it runs on all sorts of systems without all the nasty typedefs and #defines that litter a header file, and it was designed by Scheme guys (and not some random dude from AT&T). Java 1.5 even has a little bit of genericity -- still somewhat stoopid, but not as bad as C++ templates.
All in all, it seems a much more "marketable" language than C++. So (and again I wonder), why not do your network programming in Java?
I'm stuck in an academic setting, away from the wilds of the working world, so I really would like to know what motivates someone who sounds like they're a bit closer to the frontlines than me.
It would sure be fun to use an analogy right now.
why aren't you writing in Java
Because when I started writing socket applications, Java was not a viable language to write in. I am stubborn about changing. Bought a Java book a few months ago though! Planning to read it and teach myself to learn Java! Doesn't look too difficult! No real desire to use C# though.
137: would your analogy involve programming languages?
Yes, and I'm sure we could spend hours picking over whether it was or wasn't appropriate. Yay!
I think that it's not good to have too much faith in robust type systems and stuff like that, because in the end, it's all just bits to the computer. In socket programming, in particular, it pays to be able to see things the way the computer sees them. Java et al. make this a lot more difficult.
I guess my perspective is different because I've lately been programming for a battery-powered embedded device. Every cycle counts!
"I think that it's not good to have too much faith in robust type systems and stuff like that, because in the end, it's all just bits to the computer."
This sounds a lot like saying, "it's not good to have too much faith in logic and stuff like that, because in the end it's all just symbols on a page."
Which I suppose is true but...
I think what Neil means, and this makes sense to me, is that the Achilles' tendon (or whatever) is the point where you translate the stream of bits coming in from your socket or file or () into typed data. If it is on the socket as signed 16-bit integers and you read it as unsigned 16-bit integers or something else, you are likely to have problems that a good programming language will not solve.
Garbage in, garbage out I guess... but if you're reading in unsigned integers, you'll still be right half the time.
This sounds a lot like saying, "it's not good to have too much faith in logic and stuff like that, because in the end it's all just symbols on a page."
Which I suppose is true but...
Will you cut it out with the analogies already? That's not true, and neither is the Corner thing, but I refuse to go into why that is, because I'm a believer in the Ogged Doctrine.
Clownae is basically on the right track. Sure, you can use a language that enforces a strict distinction between strings and integers, or whatever. But the computer doesn't know there's a difference, only the language; the language is an abstract entity which is necessarily an imperfect model for the computer. The computer, on the other hand, is a real entity which is going to actually execute your code. Thus I think that thinking the language's way rather than the computer's way is not productive towards the goal of making good programs.
I guess my perspective is different because I've lately been programming for a battery-powered embedded device.
Like a pacemaker?
145: That's why all the good programs are written in assembly.
All programs are written in assembly. They just don't all get there the same way.
I won't pursue the "Corner thing" either, although I think you might be reading more hostility into my analogy than is actually there.
As for this,
"But the computer doesn't know there's a difference, only the language; the language is an abstract entity which is necessarily an imperfect model for the computer. The computer, on the other hand, is a real entity which is going to actually execute your code."
I think that gets it backwards, to some extent.
Not to say that it isn't right too, but -- comments like "the language is ... an imperfect model for the computer" get you labeled as a C programmer, where I come from.
I'd say it the other way around: the computer is an imperfect model for the function (which is much more precisely expressed, all lambda-calculus-like, by some languages).
Of course, no one's gonna write some kind of network stack in ML anyway. But in the same vein, using C to construct programs that calculate certain functions seems... like going around your elbow to get to your nose.
That's the last analogy, I promise.
No, that's machine language. Quite a different thing.
What you are talking about is the motivation for low level languages; which turned out to not be such a great idea in general. Having the ability to do bit fiddling is useful, but there is a lot more going on that that ... and some of it is much better served by an abstract model of computation that by a CPU design, or whatever.
Neil, the compiler is supposed to do the work of making sure the computer lives up to the language's promises. Not always possible, e.g., in the case that clownae cites, and it's always good to be able to drill down if the compiler fails (or, as Joel Spolsky says, the abstraction leaks), but that doesn't mean clean type systems don't have their place... Which I don't mean to say you were ever arguing.
Now, I'd like to hear how that Corner analogy is "false" as opposed to, say, glib and facile. But I can wait out Ogged's moratorium to find out.
I can't believe I'm reading a programming language thread on Unfogged.
I come here for philosophical discussions I can't possibly begin to understand, not for technical discussions that deal with things I actually know something about, goddammit!
Yeah, this is really disturbing. I mean, what could be more awesome than a thread about which 60-year-old asses Yglesias could kick? And yet it's turned into a thread about programming.
Yglesias vs. Old Chuck Norris. Yglesias vs. Old George Foreman. Yglesias vs. Rick Flair. Woooooo!
If Ric Flair were a programming language, he'd be Javascript. Easy-going and seemingly shallow, but surprisingly powerful when the need arises.
Does that fall into the class of verboten analogies?
If I knew enough about Javascript, I could maybe make some joke about how you spend a lot of time at first just kicking the crap out of it, then it does something surprising and illegal to beat you.
151: Well, the analogy could be apt -- The Corner is a very efficient blog with low overhead, whereas each Unfogged post generates dozens to hundreds of comments, which end up occuping the majority of blogging time.
But the reason it's false is that The Corner is teh suck while C roxx0rz.
There's only so many ways that you can point out how easily a 60-year-old could take Yglesias down. Perhaps it'd be more interesting to come up with a senior citizen who's a perfectly even match?
My recent researchers indicate the LISP programmers are the most misogynist people in the US, on a par with rightwing Christians.
161: Nobody codes in LISP anymore, John. It's been Lisp for decades. I bet we could find an old LISP 1.5 coder who'll take Matt down.
119: Dylan is pretty close to common lisp without parenthesis. It's the lack of parens I was refering to as a notational thing (one that has traditionally be claimed as a problem with lisp, unfairly, I think).
155: You have to understand this is all code for something. Remember some of these languages are 60 years old. So we just have to figure out
Fortran::python ?::Matt
or
Lisp::VB ?::Matt
I think we all know the answer to that, and I'll resist the impulse to express it in set notation.
133 -- but it's got pointers!
This makes it possible to pass complex object hierarchies to a C coder who thinks computer science has made no worthwhile advancements since the invention of the pointer.
-- Gordon McMillan, 30 Jul 1998
(No, I don't know what the "this" refers to.)
Python could kick Fortran's ass any day, and from what I hear, never having used it, Java is a gigantic pain in the ass.
165: does it hurt less if we do
thelonious monk::michael katzheimer ?::Matt
?
Hey Neil, you could send me e-mail if you'd be interested in bitching about these kids and their type-safe programming languages, and I'd be very interested to hear more about embedded-device programming, and so forth. In the realm of gmail, I am: anacreon.
133: thankfully, "mini-prolog" is ambiguous, so I choose to claim that what I meant by 128 is what I said (some techniques overlap but Prolog is more general).
The logic object space of pypy looks neat.
Ah Python, the language that saved me from my humiliating abusive sub-dom relationship with C. I'm just happy that someone finally mentioned a language I'm familiar with other than the mother-bitch of semi-colons.
Hey, that link in 171 is really neat. And oddly enough, I think I was just looking for something like this anyway although I don't think I knew it until I read about the package.
Anyway: thanks w-lfs-n!
re. 148/150 -- "all programs are written in [machine language]" -- Aren't a lot of the programming languages being discussed on this thread either non-compiling or compiling to p-code? Programs written in languages like that are not "written in [machine language]". It's a minor, pedantic point but has been nagging at me this morning.
clownae, the statement is "false" in the sense of there being no nearby possible worlds in which it is even close to true, but if you'd like to salvage what scraps of truth it contains, think of the machine code of the interpreter with a fixed input (the program). In other words, an interpreter plus a partial evaluator equals a compiler.
Hey Chris, is the part of your profile that says you live in Brooklyn still accurate? (I assume the part that says you are spending "the summer" in India is no longer.) If so, we should meet for a beer sometime.
"interpreter plus partial evaluator = compiler" seems wrong to me. Or only true if the whole input is interpreted prior to the start of execution, which I guess is an accurate description of a lot of (or "some") modern interpreters.
clownae, I'm in Brooklyn and I like beer. But I also suffer from unbearable social awkwardness. Email me privately if you'd like to follow up.
