Posts by BenWilson
Last ←Newer Page 1 2 3 4 5 Older→ First
-
Heh, and the reverse. When they brought in GST in Oz, a consulting firm was hired by my employer to write universal GST calculator functions. All programmers were directed to find ways to call these functions, which were all developed under the latest paradigms, n-tiered whatnot. For some reason, when it went live, this method seemed to take about a second to work out the GST every time. It was a finance company so it was working out GST thousands of times every second in various apps, web pages, spreadsheets, etc. Of course almost all systems that had been dutifully converted ground to a screaming halt, and screaming was exactly what happened when the stockbrokers couldn't trade.
It was explained to me why it took a computer a second to calculate x * 1.10 which my old Vic20 seemed to be able to do about 1000 times faster in 1982. All the programs had to call wrapper functions, which had to be custom written for every environment. In many cases the operating system had to download the library across the network to use it. This wrapper then called metawrappers which were universally written in Java, which was running in a virtual machine (which usually had to be loaded upon each call to the function). Java then communicated over a specially developed protocol over TCP/IP to a GST server. This GST server's job was to ask the mainframe what GST was, and the interface between the mainframe and the server was apparently DOS batch files. The mainframe answered (via batches) the server spoke TCP/IP back to the metawrappers which spoke to the wrapper which returned a value to the programs. If you can believe this, even programs running on the mainframe were expected to follow this procedure (for standardization), despite the fact that the GST values were actually stored on the mainframe in the first place.
Apparently it was all beautifully documented and peer reviewed. The testing procedures were exemplary, although load testing was never really done properly since no-one could actually believe that computers might actually struggle to calculate x *1.10.
What had not been carefully documented by most of the programmers was that whilst developing the interface to the wrapper functions, they had also written coding stubs, while they were waiting for the contracting team to write the exceedingly difficult function to calculate GST. Most of the time, this coding stub was a barbarically simple x=x*GST_CONSTANT; Disgusting, huh?
On the day it all went to hell, the programmers were told to go back to the undocumented coding stub, and the system fired back up to it's former glory. This GST calculator project was sent back and told to improve it's performance. After a week they had made 10-fold improvement, a magnificent achievement for a hard-working team of contractors (although I'm told the documentation suffered a bit during this week). But unfortunately being able to calculate GST at only 10 times a second was not enough, and again the programmers backrevved after some management apoplexy. This repeated once more before the team decided to sacrifice a little of the abstract perfection of the system. Systems were allowed to cache some of the values returned (which in 100% of the cases turned out to be the value 1.1) and perform the calculation locally. The performance improved drastically. Some machine were able to calculate GST up to a thousand times per second (we've reached 1982 standards by now). This was deemed good enough, and the contractors dispersed to the four winds. When the GST server went down the guy who could fix it turned out to have taken a job contracting in London.
Fortunately he had left 200 pages of excellent doco on how to get this server up and going, which the guy whose job it was looked at, fiddled around a lot with various settings, but couldn't ultimately fix. It seemed that every couple of hours the server would shit itself. So he wrote a batch file to rekick this server every couple of hours. This is apparently the live system today.
Everyone who needs to calc GST more than 1000 times a second (which unsurprisingly is almost everyone) has reverted to the undocumented coding stubs. Some bright spark decided to go around documenting all these stubs, which turned out to be a 2 hour job. If and when GST changes (something that usually comes with several years warning), they will have to pull out his one page document, go through a few dozen programs and change a constant. Hacky, huh? But management decided to live with that.
-
...which I can happily report is an experience akin to real drunkenness. Which is one of the bad things about piss.
-
Jenkem experience report
ROFLMAO. Could hardly get any more ridiculous. It's like getting someone to beat you up so you can feel punch-drunk.
-
It makes me think I'd never, ever hire you in the first place
Why is it I'd want to work for you?
You're following Machiavelli's first law. Always rag on Machiavelli. But the zeroth law is to pretend never to have even heard of him.
Personally I've done quite well as a contractor by having knowledge that there's no point documenting. Every time I even begin to explain it, the employer's eye's glaze over and he just goes 'so it works then?'. 'Yes, sir, it works'. 'Good then, keep up the good work. It's good to know someone understands that shit'. 'Very good, sir'. But I document it anyway, for my own sake.
The number of times I've listened in to other departments go on about their amazing clean-room procedures in the 4th year of not delivering what I've already completed and is being used by thousands of satisfied customers, I can't count on fingers. Then they're always 'hey, we must get together, share the knowledge, write some libraries, do a peer review, work out some standards'. And it's always 'yup, sure, sounds good <goes back to cutting code for the man>'. Perhaps I'm jaded, or perhaps I'm just realistic. 25 years of code cutting will do that to you.
-
Car manufacturers don't need people to die; pot dealers don't need people to be addicted.
No, but it helps. The car manufacturers who have done the best are of course the very ones which have had the most people die in them, since they are the cars which most people have. And a successful pot dealer loves his addicts, who are a walking $ sign.
Totally. The number of people who needed counselling for gambling addictions rose from 923 in 1997 to 4010 in 2004.
How many people were gambling?
I'm not quite sure why you're so against regulation at one level, but happy to let communities decide, but that would be fine with me too.
The bigger the level, the less I like it.
-
You mean the buzz of risk taking, the sparkle of the lights, the agony of the miss and perhaps even the thrill of winning.
Sounds soo much like Frogger.
The worst game from my childhood was Gauntlet. I marveled at the cleverness of the idea that you could continue your game as long as you had coins for. Sheer brilliance in the evil way it exploited all my hard won earnings week after week. At least the old game of 3 lives meant you eventually lost and decided to call it quits. With Gauntlet you only lost when you ran out of money.
That said, I had hundreds of hours of fun with my mates playing it.
-
i guess that is part of the problem - most communities didn't get to decide. how about we make every outlet go through a full resource consent process to make sure that every community does have a say in this.
Very good idea. Works with booze. And it's better than booze cause you can't take the pokies home with you. Or at least the demographic we're trying to make decisions on behalf of probably doesn't.
-
It would be interesting to tot up the number of people who feel harmed by the proliferation of pokies in poor suburbs and compare that to the number who claim enjoyment.
Yes it would. "How harmed" would also be important to establish. I feel harmed every time I hear gangsta rap, but only a very very little bit.
Say there are no pokies in New Zealand -- would you actively bring them in to enhance this public enjoyment you speak of?
Hell no. They're lame. But I wouldn't stop people doing it, so long as the communities concerned were OK with it.
-
And I don't get why you think it's the same
It's the same principle. Unless you don't believe in any principles behind public policy, then tell me by what principle you distinguish between vices. I can accept that your principles are totally piecemeal, and that you don't agree with pokies because they have caused more problem gambling (whether we actually agree on that point or not), but I can't see how you then advocate the availability of any other new vices, like various pills that were also not around in the early 90s. The number of people who got sick from a party pill would only need to be 1 for the proportional increase to be undefinedly large. So what? It's still 1 person.
To take the most obvious example, according to the Problem Gambling Foundation, pokies now deliver more problem gambling referrals than all other forms of gambling combined: by a factor of four. This, from a form of gambling that didn't exist at the beginning of the 1990s.
So it's a more popular form of gambling now. I'm not sure what you're trying to prove by that. I don't play Frogger anymore either.
They say 3% of gamblers have a problem. Why exactly are the other 97% not allowed their fun because of that? Was it a different number before the pokies?
What level of regulation would you favour?
Now for the hard questions.
Would you rather not have had the Gambling Act 2003 and have double the number of machines again by now?
Is there any evidence that there isn't just a fairly fixed number of people who become problem gamblers?
Would you have no restrictions on the number of machines that can be placed in a community?
I'd leave that to the particular community you are talking about to decide. Personally I'd vote against pokies near me, but if I got outvoted then I'd have to swallow that.
No electronic monitoring?
Not sure exactly what you mean. That could mean a lot of things. Are you talking about limiting the losses a person can take? If so, I'd be in favor of that, to mitigate the harm problem gamblers can cause themselves. It would be much like laws against serving alcohol to the intoxicated.
No compulsory notices telling people what the odds really are?
Now there's a tricky one. Yeah, I reckon warning labels are a very good idea, but gambling's a really tricky one because they always invent tricky new games. Perhaps just an 'Average cost per hour' for each game. That would put it in stark relief. But I can see loopholes.
-
So designing a pokie machine to be psychologically addictive, knowing that some relatively set proportion of it's users will be directly harmed by that designed addictiveness is, therefore, wrong?
As I just said at length, no. The small number who receive harm doesn't outweigh the large number who get net enjoyment. And I am taking into account that the harm is severe compared to the enjoyment.
This might seem cold and calculating, but we make these kinds of decisions every single day. People are killed in cars all the time, but the benefit of having them outweighs that. Some people have severe allergic reactions to medication, but we still allow that medication to cure things that would otherwise just take a bit longer for the body to clear up by itself. A lot of TV is of mild entertainment to many, and highly offensive to just a few.
Last ←Newer Page 1 … 1001 1002 1003 1004 1005 … 1066 Older→ First