Iron Man is now powered by Oracle

spamcourt

Clearly we need a language where you can declare a 0array or a 1array. That would end problems once and for all.

blatant_mcfakename

This is still my favourite though - http://www.youtube.com/watch?v=hkDD03yeLnU

The fun is the fact they probably know it's utter bollocks themselves but they manage to deliver lines like that with 100% seriousness

@boomzilla said:

but it amazes me how many people love to make 1900 part of the 19th century (i.e., the 1800s) as a consequence of counting backwards.

How does that work exactly? Because the way I see it 1900 is part of the 19th century if you count forwards, ie correctly. Year 1 to 100 (1st century), 101 to 200 (2nd), ..., 1801 to 1900 (19th).

I'm guessing you made a typo and meant 1800.

@blatant_mcfakename said:

http://www.youtube.com/watch?v=hkDD03yeLnU

Community server doesn't automatically create links. You can either use an <a href="..."> tag or use TinyMCE's

button.

DCRoss

@da Doctah said:

What I don't get is how product placement can possibly co-exist with the standard "Any resemblance to actual anything is purely coincidental" disclaimer.

I don't see the conflict there. In the movie we see a super powered, nigh-infallible force of computing capable of providing answers to the most complex problems facing the defenders of the free world.

In reality, we have... Oracle. No resemblance at all, other than computers being involved in some way.

lucas1

Mathematicians do start counting at zero ... factorials, fibanacci etc all start at zero . This also makes sense in computer science, since it is mathematically based. Pointer Arithmetic has it basics in mathematics; the clue is in the name.

As for filtering elements in JavaScript or C#, you are doing it wrong. Both have predicate functions for sorting, filtering etc. With JavaScript if the browser doesn't support the newer methods, most popular client side libraries add support.

boomzilla

@Zecc said:

@boomzilla said:
but it amazes me how many people love to make 1900 part of the 19th century (i.e., the 1800s) as a consequence of counting backwards.

How does that work exactly? Because the way I see it 1900 is part of the 19th century if you count forwards, ie correctly. Year 1 to 100 (1st century), 101 to 200 (2nd), ..., 1801 to 1900 (19th).

I'm guessing you made a typo and meant 1800.

If you think it was a typo, then you didn't read what I wrote.

Who starts at the year 1? Why so much emphasis on that? Why should we succumb to ridiculous OCDism regarding a year that no one ever referred to as "year 1" during the actual time period in question. The counting of the years was something created after the fact, and likely isn't even the right year based on the reason for its choice. So your insistence that we count from 1 creates, as I said, an unending procession of special cases.

The point is that it's more intuitive to group things together when they are similar to each other. IOW, 1900 and 1901 are more similar than 1899 and 1900. So why do things backwards? Is 1900 part of the 1900s or not? Will 2020 be part of the decade known as the twenties? Your fascination with the lack of a year 0 does all sorts of counterintuitive things, so why not just have one special case? I think the answer is that it's a rare occasion for normal people to exercise their pedantic dickweedery.

blakeyrat

@morbiuswilters said:

Exactly. It's just sad because we've been doing this long enough that better stuff should have bubbled up. But trends and fads and "make the programmer feel special" seem to do more to drive technology than critical thinking.

It's bubbling up in the wrong direction. The shitty tools are getting promoted, while the good ones are... I dunno... too boring? There is a certain brand of engineer who will ignore and deride anything made by Microsoft, regardless of its quality. There's also a certain brand of engineer who will promote CLI tools over GUI one, in the mistaken and unexamined belief that they are "more efficient". There's also a certain brand of engineer who believes tools should be difficult to use, as to make their jobs more secure.

Those asshats are taking over IT. And tearing down everything great we ever built. Slashdot is winning. "Worse is better" turned out to be right, and things are worse than ever now.

For example, Apple got a load of those asshats from NeXT, and next (haha) thing you know, Mac OS is a ball of shit. What happened to Mac Classic's legendary usability? Gone. Spatial memory? That only benefits non-geeks, so gone. Window borders? Why would you put the close box on the opposite side from the maximize box, it's not like anybody would click that by accident. And why would you give it a meaningful icon instead of making it a ball of red? The red ball vaguely looks like a traffic light so obviously that's superior. Oh and let's make a whole new type of window that looks vaguely like it's a plate of aluminum and behaves totally differently from the existing windows, then let's sprinkle those in our various apps with no rhyme or reason. Is there a single person alive who can internalize the logic of the Dock? There is? Better make it even more confusing then! Notification system? Just because Mac Classic had one, and just because Windows has had one since fucking 1998, we'll take our sweet time and maybe add one in 2011, maybe, if our users are lucky. (And only then because a third-party one got too popular to ignore.)

Apple was on the Right Path. Now they aren't. Ironically, Microsoft is closer to the Right Path than Apple is now. (Of course, even Apple is much closer to the Right Path when it comes to dev tools-- at least they have a decent IDE, that's better than anything in Linux- or Java-land.)

Now I'm depressed.

@boomzilla said:

If you think it was a typo, then you didn't read what I wrote.

I read what you wrote. Doesn't seem I read it as you meant, though. I think we've got a different understanding of what counting "forwards" and "backwards" is.

But your next post clarified things for me, and I agree with you. FWIW, I promote "1900 is in the 19th century" as being "correct" just because there is no year zero. Whether or not there should be a year zero is another, more delicate question.

Btw, wouldn't it be wonderful if the 1800s were the 18th century, even if that meant there was a 0th century? (and a 0th BCE too) We'd be in the first decade of the zeroeth century of the second millenium.

PS, just because: www.timeanddate.com/calendar/monthly.html?year=1582&month=10 (not sure if I was expecting something else)

da_Doctah

@DCRoss said:

@da Doctah said:
What I don't get is how product placement can possibly co-exist with the standard "Any resemblance to actual anything is purely coincidental" disclaimer.

I don't see the conflict there. In the movie we see a super powered, nigh-infallible force of computing capable of providing answers to the most complex problems facing the defenders of the free world.
In reality, we have... Oracle. No resemblance at all, other than computers being involved in some way.

Let's consider a classic film, Mars Attacks. On the one hand, the Slim Whitman people pay big bucks to have their product featured prominently, with the implication that Slim Whitman can put an end to the alien invasion and save the world. Yay for Slim Whitman, and yay for all the people who make money from him.

On the other hand, the disclaimer says that the Slim Whitman in the movie is a work of fiction, and that any resemblance to Slim Whitman living or dead is purely coincidental. For shame! There it is, right at the foot of the credits, an open admission that Slim Whitman doesn't have the ability to repel alien invaders.

QED: direct conceptual conflict.

@Zecc said:

PS, just because: www.timeanddate.com/calendar/monthly.html?year=1582&month=10 (not sure if I was expecting something else)

Actually, I just realized this depends on the country you're in. In France for example, the missing days are(n't) in December; in Denmark, on February of 1700; in the UK, the US and Australia on September 1752.

I did know the Russian October Revolution was in Greg Cal's November though.

Sutherlands

@morbiuswilters said:

I actually think there is a market for a well-designed language that fills a specific need--something high-level, object-oriented, garbage-collected but also compiled, with an eye towards performance and parallelism.

Like C#?

blakeyrat

@Sutherlands said:

@morbiuswilters said:
I actually think there is a market for a well-designed language that fills a specific need--something high-level, object-oriented, garbage-collected but also compiled, with an eye towards performance and parallelism.
Like C#?

Yeah, basically C# and the whole .net framework solves all these problems, the only remaining issue is you can't get the morons to use it because Microsoft is TEH EBIL! (Well, the other problem is some mobile platforms don't support it because Microsoft is TEH COMPETITER!)

The fact that C# and .net are are ECMA and ISO standards seems to not matter in the slightest. Among idiots.

morbiuswilters

@boomzilla said:

You're still in pointer arithmetic mode, and really saying, get me the element that's offset n places from the start. I've come across places where 0-based and 1-based have advantages over each other. In general, I prefer 1-based, because it's more intuitive for the most common cases. 0-based is often easier when you have to calculate things. Anyways, I've had off by one errors in both cases.

True, 0-indexing works a bit better for pointer arithmetic, but that's why I keep qualifying my statement by saying "modern, high-level language". And you can do pointer arithmetic with 1-based indexing, it just requires the compiler to be a bit more clever.

@boomzilla said:

What's really inexcusable is when someone puts months into an enumeration and January is zero. That's when you know that the cargo cultism has gone too far.

Yes. That is flat-out insanity.

morbiuswilters

@Snowyowl said:

Why would you want to jump to the fifth element? Is there something special about the value at the fifth element? Put it in a variable, then, not in an array. Arrays are for iterating over; it shouldn't matter that array[4] is actually the fifth element because there's nothing to distinguish it from array[3] or array[5].

What? No. Arrays are not only for iterating over. Nor are they only accessed in a for (int i = 0; i < size; ++i) loop. Try going some advanced work with arrays that requires you to jump around in the array or jump to a specific point, and you'll quickly learn to curse zero-indexing.

Sutherlands

I honestly can't remember the last time I accessed something by an index.

morbiuswilters

@PSWorx said:

You want to implement a simple XOR-based encryption obfuscation. Or an animation script. Or anything else where you have to continuously loop over your array, depending on a counter.

Which one is more intuitive?

This: currentValue = array[counter % array.size]

Or this: currentValue = array[((counter - 1) % array.size) + 1]

Yes, yes, I figured somebody was going to bring up the modulus thing. But: 1) all other array accesses are far more common than the modulus thing; and 2) it's trivial to implement the modulus case as a function/macro like get_mod_offset(array, counter).

So, no, I don't consider this a valid reason to stick with 0-indexing. It's more important for languages to be intuitive and resistant to bugs than it is to make edge cases look elegant.

morbiuswilters

@lucas said:

They start at zero because mathematicians start counting at zero, coming from a maths heavy background I find it makes sense.

No they don't. They start at zero to make pointer arithmetic easier for the C compiler to implement. 2) This is irrelevant. Modern languages should strive to make bugs difficult, not adhere to some arbitrary mathematics conventions.

@lucas said:

If senior level programmer hasn't grapsed how to write for loop correctly I doubt they should be called "senior".

It's not an issue of "grapsed". It's an issue of "this is implemented in a way that guarantees a certain amount of bugs, even from experienced programmers". You might as well say the same thing about buffer overflows: they're easy to understand and check for, but they still slip through production software created by senior-level programmers all the time. Most modern languages do bounds-checking because humans are not perfect; if your tool is built on the premise that humans aren't going to make mistakes, it's a shoddy tool.

@lucas said:

If you don't like zero based indexing, functional languages or languages that let you do foreach are probably more your cup of tea.

Lots of languages have foreach and zero-based indexing; the former is great, the latter not. Besides which, I do actual work with programming languages, so functional languages are a no-go.

@lucas said:

In C# and JavaScript I rarely have to iterate over a collection while counting, or I manipulate the array/collection before looping.

Then why does it even matter to you? It matters to me because when I do have to use indexes, I find it irritating that we're still stuck with this ass-backwards idiom.

morbiuswilters

@PSWorx said:

...molybdomancy...

Holy crap, my preacher was right when he told me Europeans were witchcraft-practicing devil worshipers.

morbiuswilters

@lucas said:

Pointer Arithmetic has it basics in mathematics; the clue is in the name.

It's also been fucking outmoted for a quarter century. And you don't need 0-based indexes with pointer arithmetic, it just makes the compiler-writer's job easier while making the compiler-user's harder (which is the entire philosophical underpinning of C). Stop with this nonsense.

@morbiuswilters said:

@PSWorx said:
...molybdomancy...

Holy crap, my preacher was right when he told me Europeans were witchcraft-practicing devil worshipers.

Hey, I do not practice witchcraft.

You seem to be agreeing a lot with yourself.

morbiuswilters

@blakeyrat said:

There's also a certain brand of engineer who will promote CLI tools over GUI one, in the mistaken and unexamined belief that they are "more efficient". There's also a certain brand of engineer who believes tools should be difficult to use, as to make their jobs more secure.

Yes. There's also the "let's make the programmer feel like a special little genius" attitude which is killing this industry. So companies pander to this shit with their employees and we end up with "engineers" who chase after every fad technology. That shit drives me crazy.

Look: if you enjoy doing your job, that's great, but your first duty is to do your job. So many software "engineers" today seem to think their primary duty is to have fun, stimulate their brain, and then maybe get some work done if they feel like it. I hate those people. And I pretty much hate computers and programming, I really do, but I'm better than the vast majority of people in this industry just because I actually dedicate myself to being good at it. I don't read Hacker News or Slashdot or any of that junk, because it's just filled with the idiotic braying of jackasses (how much you wanna bet somebody will quote that single line and say something like "Yeah, I wouldn't know what that's like" or if they're less clever "Now you know how we feel" or if they're Zylon "HERP UR TALKIN BOUT URSLF HA HA").

About the only positive thing I can say about software engineering nowadays is that it's improved the quality of french fries. Most of these idiots would be no better working a fryer than they are a debugger and if it weren't for software engineering, they might worm their ways into positions of actual importance where they end up doing real harm, like burning my onion rings.

@blakeyrat said:

For example, Apple got a load of those asshats from NeXT, and next (haha) thing you know, Mac OS is a ball of shit. What happened to Mac Classic's legendary usability? Gone. Spatial memory? That only benefits non-geeks, so gone. Window borders? Why would you put the close box on the opposite side from the maximize box, it's not like anybody would click that by accident. And why would you give it a meaningful icon instead of making it a ball of red? The red ball vaguely looks like a traffic light so obviously that's superior. Oh and let's make a whole new type of window that looks vaguely like it's a plate of aluminum and behaves totally differently from the existing windows, then let's sprinkle those in our various apps with no rhyme or reason. Is there a single person alive who can internalize the logic of the Dock? There is? Better make it even more confusing then! Notification system? Just because Mac Classic had one, and just because Windows has had one since fucking 1998, we'll take our sweet time and maybe add one in 2011, maybe, if our users are lucky. (And only then because a third-party one got too popular to ignore.)

I love when people tell me OSX has the best usability. Then why the fuck is there only a single toolbar at the top of the screen? That shit made sense back when screens were 7" wide and 300px across because it saved space and it was easy to move the mouse up to the toolbar. On a 30" desktop you start getting the feeling Apple has no fucking clue what they're doing anymore, other than turning out overpriced, shiny garbage that hipsters can buy with their parents' credit cards.

Then there's shit too like the built-in terminal is a fucking joke. There are two WTFs in that sentence: the first is that we're still building terminals into modern OSes (I know Microsoft does it, too, but it's much more essential in Unix-based OSes) and the second is that, even though it's fundamental to doing a lot of shit in OSX, they still managed to fuck it up. I mean, seriously, the first time I sat down to use OSX within five minutes I was like "I've got to find a replacement terminal to install". That's some pretty hard-core failure there: five minutes in and I've already used the terminal, found it wanting and decided it's essential enough that I need to find a replacement to keep working.

The iPhone seems okay, although I'd hardly call it great. It's just way better than Android, which is a fucking disaster. When I bought my first Android phone the chick at the Verizon store was like "You should download this 'task killer' app so you can stop any background applications from running and free up memory!" Really? So I'm carrying a goddamn Windows 95 PC in my pants now? And the saddest thing is, even with aggressive killing of tasks it's still constantly running out of memory. "Oh, you want to open another Chrome tab? You got 2 open already, motherfucker. I'm just gonna freeze up for about 30 seconds and then pop up a dialog about Chrome hanging. Then when you go to close it the whole goddamn phone will lock up. Ha ha!"

For tasks I perform rarely, I prefer GUIs. For things I do frequently, command line really is more efficient. I often learn the GUI, then the command line later, because I am faster.

Where I work, we very rarely use Microsoft products, and only on customer request. They're too easy to break.

Like when I build a manufacturing tester on Windows and install the software on a preconfigured user account. Ship it out, and they tell me it doesn't work. 90 minutes of debugging over the phone later, and it turns out they added the machine to their domain because IS/IT told them to. Really, this isn't a computer, it's an appliance. It may look like a computer and act somewhat like a computer, but it must be treated like an appliance. If I use Linux, then they can't break it nearly as easily. This statement has actually been proven.

On the embedded front, device drivers for Windows CE are an absolute pain because they are so frequently obfuscated. Add in that a Windows CE build requires many Gigabytes of disk traffic, and a CE build becomes far more of a disk-bound operation than CPU-bound, and Windows doesn't manage file-system caches well enough to take advantage of large amounts of RAM. An SSD helps, but it's still disk-bound.

On the other hand, most of the stuff we make winds up on MCUs with <128KB flash and <32K RAM, so C# isn't much of an option - .netMicroFramework is way too big.

Regarding MacOS X (note: I don't have a mac, I've just used them), care to explain what you don't like about the Terminal? It seems to work fine for me. As an example of using the command line, a friend had "deleted" a bunch of TimeMachine backups to clear out space on her external hard drive. Well, it turned out they were in the trash, and emptying the trash showed an ETA of over an hour. She wanted to leave the house and have the machine sleep when it was done. Couldn't find a GUI tool to do that, so I wrote a one-line shell script that deleted the contents of the trash folder and told the machine to sleep. I haven't seen many GUI tools that can do sequenced commands like that.